v10.1.0 notes

Unpaper accepts PNG as input now, so avoid generating a huge
temporary PPM file if we can. If we
must create a PNG, compress it lightly to keep our temp usage down.

.coveragerc

@@ -1,5 +1,3 @@
-# Coverage isn't really compatible with subprocesses so results are unreliable
-
 [paths]
 source =
 	src
@@ -8,9 +6,11 @@ source =
 [run]
 branch = true
 parallel = true
+concurrency =
+	thread
+	multiprocessing
 source =
 	src/ocrmypdf
-	tests
 omit =
 	tests/spoof/*
 
@@ -23,3 +23,4 @@ exclude_lines =
 	if 0:
 	if False:
 	if __name__ == .__main__.:
+	if TYPE_CHECKING:

.docker/Dockerfile

@@ -1,6 +1,6 @@
 # OCRmyPDF
 #
-FROM ubuntu:19.04 as base
+FROM ubuntu:19.10 as base
 
 FROM base as builder
 
@@ -10,36 +10,37 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
   build-essential autoconf automake libtool \
   libleptonica-dev \
   zlib1g-dev \
-  ocrmypdf \
-  pngquant \
-  python3-pip \
-  python3-venv \
-  tesseract-ocr \
-  unpaper \
-  wget \
+  python3 \
+  python3-distutils \
+  ca-certificates \
+  curl \
   git
 
+# Get the latest pip (Ubuntu version doesn't support manylinux2010)
+RUN \
+  curl https://bootstrap.pypa.io/get-pip.py | python3
 
 # Compile and install jbig2
 # Needs libleptonica-dev, zlib1g-dev
 RUN \
   mkdir jbig2 \
-  && wget -q https://github.com/agl/jbig2enc/archive/0.29.tar.gz -O - | \
+  && curl -L https://github.com/agl/jbig2enc/archive/0.29.tar.gz | \
   tar xz -C jbig2 --strip-components=1 \
   && cd jbig2 \
   && ./autogen.sh && ./configure && make && make install \
   && cd .. \
   && rm -rf jbig2
 
-RUN python3 -m venv /appenv
-
 COPY . /app
 
 WORKDIR /app
 
-RUN . /appenv/bin/activate; \
-  pip install --upgrade pip \
-  && pip install .
+RUN pip3 install --no-cache-dir \
+  -r requirements/main.txt \
+  -r requirements/webservice.txt \
+  -r requirements/test.txt \
+  -r requirements/watcher.txt \
+  .
 
 FROM base
 
@@ -53,7 +54,6 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
   zlib1g \
   pngquant \
   python3 \
-  python3-venv \
   qpdf \
   tesseract-ocr \
   tesseract-ocr-chi-sim \
@@ -62,11 +62,15 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
   tesseract-ocr-fra \
   tesseract-ocr-por \
   tesseract-ocr-spa \
-  unpaper \
-  wget
+  unpaper
+
+WORKDIR /app
+
+COPY --from=builder /usr/local/lib/ /usr/local/lib/
+COPY --from=builder /usr/local/bin/ /usr/local/bin/
 
-# Copy
 COPY --from=builder /app/misc/webservice.py /app/
+COPY --from=builder /app/misc/watcher.py /app/
 
 # Copy minimal project files to get the test suite.
 COPY --from=builder /app/setup.cfg /app/setup.py /app/README.md /app/
@@ -74,7 +78,4 @@ COPY --from=builder /app/requirements /app/requirements
 COPY --from=builder /app/tests /app/tests
 COPY --from=builder /app/src /app/src
 
-COPY --from=builder /appenv /appenv
-COPY --from=builder /usr/local /usr/local
-
-ENTRYPOINT ["/appenv/bin/ocrmypdf"]
+ENTRYPOINT ["/usr/local/bin/ocrmypdf"]

.docker/alpine.dockerfile

@@ -1,91 +0,0 @@
-FROM alpine:3.9 as base
-
-FROM base as builder
-
-ENV LANG=C.UTF-8
-
-# Normally:
-# echo '@testing http://nl.alpinelinux.org/alpine/edge/testing' >> /etc/apk/repositories
-
-RUN \
-  echo -e '@testing http://nl.alpinelinux.org/alpine/edge/testing\n@community http://nl.alpinelinux.org/alpine/edge/community'\
-  >> /etc/apk/repositories \
-  # Add runtime dependencies
-  && apk add --update \
-  python3-dev \
-  py3-setuptools \
-  jbig2enc@testing \
-  ghostscript \
-  qpdf@community \
-  qpdf-dev@community \
-  tesseract-ocr \
-  unpaper \
-  pngquant \
-  libxml2-dev \
-  libxslt-dev \
-  zlib-dev \
-  libffi-dev \
-  leptonica-dev \
-  binutils \
-  && pip3 install --upgrade pip \
-  # Install pybind11 for pikepdf
-  && pip3 install pybind11 \
-  # Install flask for the webservice
-  && pip3 install flask \
-  # Add build dependencies
-  && apk add --virtual build-dependencies \
-  build-base \
-  git
-
-COPY . /app
-
-WORKDIR /app
-
-RUN pip3 install .
-
-FROM base
-
-ENV LANG=C.UTF-8
-
-# Normally:
-# echo '@testing http://nl.alpinelinux.org/alpine/edge/testing' >> /etc/apk/repositories
-
-RUN \
-  echo -e '@testing http://nl.alpinelinux.org/alpine/edge/testing\n@community http://nl.alpinelinux.org/alpine/edge/community'\
-  >> /etc/apk/repositories \
-  # Add runtime dependencies
-  && apk add --update \
-  python3 \
-  jbig2enc@testing \
-  ghostscript \
-  qpdf@community \
-  qpdf-dev@community \
-  tesseract-ocr \
-  tesseract-ocr-data-deu \
-  tesseract-ocr-data-chi_sim \
-  unpaper \
-  pngquant \
-  libxml2 \
-  libxslt \
-  zlib \
-  libffi \
-  leptonica-dev \
-  binutils \
-  && mkdir /app
-
-WORKDIR /app
-
-# Copy build artifacts (python site-packages)
-COPY --from=builder /usr/lib/python3.6/site-packages /usr/lib/python3.6/site-packages
-COPY --from=builder /usr/bin/ocrmypdf /usr/bin/dumppdf.py /usr/bin/latin2ascii.py /usr/bin/pdf2txt.py /usr/bin/img2pdf /usr/bin/chardetect /usr/bin/
-
-# Copy
-COPY --from=builder /app/misc/webservice.py /app/
-
-# Copy minimal project files to get the test suite.
-COPY --from=builder /app/setup.cfg /app/setup.py /app/README.md /app/
-COPY --from=builder /app/requirements /app/requirements
-COPY --from=builder /app/tests /app/tests
-COPY --from=builder /app/src /app/src
-
-ENTRYPOINT ["/usr/bin/ocrmypdf"]

.gitattributes

@@ -9,5 +9,6 @@
 *.png	binary
 *.jpg	binary
 *.bin   binary
+*.afdesign  binary
 
 .git_archival.txt  export-subst

.github/FUNDING.yml

@@ -0,0 +1,12 @@
+# These are supported funding model platforms
+
+github: # Replace with up to 4 GitHub Sponsors-enabled usernames e.g., [user1, user2]
+patreon: # Replace with a single Patreon username
+open_collective: james-barlow
+ko_fi: # Replace with a single Ko-fi username
+tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
+community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
+liberapay: # Replace with a single Liberapay username
+issuehunt: # Replace with a single IssueHunt username
+otechie: # Replace with a single Otechie username
+custom: # Replace with up to 4 custom sponsorship URLs e.g., ['link1', 'link2']

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,40 @@
+---
+name: Bug report
+about: Create a report to help us improve
+title: ''
+labels: ''
+assignees: ''
+
+---
+
+**Describe the bug**
+A clear and concise description of what the bug is.
+
+**To Reproduce**
+What command line or API call were you trying to run?
+
+```bash
+ocrmypdf  ...arguments... input.pdf output.pdf
+```
+
+Run with verbosity or higher `-v1` to see more detailed logging. This information may be helpful.
+
+**Example file**
+Include an input PDF or image that demonstrates your issue.
+
+Please provide an input file with no personal or confidential information. At your option you may `GPG-encrypt the file <https://github.com/jbarlow83/OCRmyPDF/wiki>` for OCRmyPDF's author only.
+
+Links to files hosted elsewhere are perfectly acceptable. You could also look in ``tests/resources`` and see if any of those files reproduce your issue.
+
+(Exceptions: Issues with installation, command line argument parsing, test suite failures.Issues without example files usually cannot be resolved.)
+
+**Expected behavior**
+A clear and concise description of what you expected to happen.
+
+**Screenshots**
+If applicable, add screenshots to help explain your problem.
+
+**System**
+ - OS: [e.g. Linux, Windows, macOS]
+ - OCRmyPDF Version: ``ocrmypdf --version``
+ - How did you install ocrmypdf? Did you use a system package manager, `pip`, or a Docker image?

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,17 @@
+---
+name: Feature request
+about: Suggest an idea for this project
+title: ''
+labels: enhancement
+assignees: ''
+
+---
+
+**Is your feature request related to a problem? Please describe.**
+A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
+
+**Describe the solution you'd like**
+A clear and concise description of what you want to happen.
+
+**Additional context**
+Add any other context or screenshots about the feature request here.

.gitignore

@@ -3,11 +3,11 @@
 .pylintrc
 .pytest_cache/
 .ruffus_history.sqlite
-.venv/
 .venv*/
 *.pyc
 *.sublime-*
 *.DS_Store
+.mypy_cache/
 
 # Package building
 .eggs/
@@ -43,4 +43,5 @@ tests/output/
 tests/resources/private/
 tmp/
 /debug_tests.py
-private/
+*.traineddata
+/private

.pre-commit-config.yaml

@@ -1,6 +1,23 @@
 repos:
--   repo: https://github.com/ambv/black
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v2.4.0
+    hooks:
+      - id: check-case-conflict
+      - id: check-merge-conflict
+      - id: check-toml
+      - id: check-yaml
+      - id: debug-statements
+  - repo: https://github.com/asottile/seed-isort-config
+    rev: v1.9.3
+    hooks:
+      - id: seed-isort-config
+  - repo: https://github.com/pre-commit/mirrors-isort
+    rev: v4.3.21 # pick the isort version you'd like to use from https://github.com/pre-commit/mirrors-isort/releases
+    hooks:
+      - id: isort
+  - repo: https://github.com/psf/black
     rev: stable
     hooks:
-    - id: black
-      language_version: python3.7
+      - id: black
+        language_version: python3.7
+        exclude: ^src/ocrmypdf/lib/_leptonica.py

.travis.yml

@@ -1,146 +0,0 @@
-cache:
-  pip: true
-  directories:
-  - $HOME/Library/Caches/Homebrew
-
-matrix:
-  include:
-    - os: linux
-      dist: trusty
-      sudo: required
-      language: python
-      python: "3.6"
-      env:
-        - DIST=trusty
-      addons: &trusty_apt
-        apt:
-          update: true
-          sources:
-          - sourceline: 'ppa:alex-p/tesseract-ocr'
-          - sourceline: 'ppa:heyarje/libav-11'
-          - sourceline: 'ppa:vshn/ghostscript'
-          packages:
-          - ghostscript
-          - libavcodec56
-          - libavformat56
-          - libavutil54
-          - libexempi3
-          - libffi-dev
-          - pngquant
-          - poppler-utils
-          - qpdf
-          - tesseract-ocr
-          - tesseract-ocr-deu
-          - tesseract-ocr-eng
-          - tesseract-ocr-fra
-    - os: linux
-      dist: xenial
-      sudo: required
-      language: python
-      python: "3.7"
-      env:
-        - DIST=xenial
-      addons:
-        apt:
-          update: true
-          sources:
-          - sourceline: 'ppa:alex-p/tesseract-ocr'
-          packages:
-          - ghostscript
-          - libexempi3
-          - libffi-dev
-          - pngquant
-          - poppler-utils
-          - qpdf
-          - tesseract-ocr
-          - tesseract-ocr-deu
-          - tesseract-ocr-eng
-          - tesseract-ocr-fra
-          - unpaper
-    - os: osx
-      osx_image: xcode9.2
-      language: generic
-      addons:
-        homebrew:
-          update: true
-          packages:
-            - exempi
-            - ghostscript
-            - jbig2enc
-            - leptonica
-            - openjpeg
-            - pngquant
-            - python
-            - qpdf
-            - tesseract
-            - unpaper
-    - os: osx
-      osx_image: xcode9.2
-      language: generic
-      env:
-        - ADD_PDFMINER=1
-      addons:
-        homebrew:
-          update: true
-          packages:
-            - exempi
-            - ghostscript
-            - jbig2enc
-            - leptonica
-            - openjpeg
-            - pngquant
-            - python
-            - qpdf
-            - tesseract
-            - unpaper
-
-before_cache:
-- rm -f $HOME/.cache/pip/log/debug.log
-
-before_install: |
-  mkdir -p bin
-  if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then
-    pip3 install --upgrade pip
-    pip3 install --upgrade wheel
-    if [[ "$DIST" == "trusty" ]]; then
-      mkdir -p packages
-      wget -q 'https://www.dropbox.com/s/vaq0kbwi6e6au80/unpaper_6.1-1.deb?raw=1' -O packages/unpaper_6.1-1.deb
-      sudo dpkg -i packages/unpaper_6.1-1.deb
-    fi
-  elif [[ "$TRAVIS_OS_NAME" == "osx" ]]; then
-    pip3 install --upgrade pip
-    pip3 install wheel
-  fi
-
-install:
-- export PATH=$PWD/bin:$PATH
-- pip3 install pycparser  # py3.7 workaround for https://github.com/eliben/pycparser/issues/251
-- pip3 install -r requirements/main.txt
-- pip3 install --no-deps .
-- |
-  if [[ "$ADD_PDFMINER" == "1" ]]; then
-    pip3 install --no-deps .[pdfminer]
-  fi
-- pip3 install -r requirements/test.txt
-
-script:
-- tesseract --version
-- qpdf --version
-- pytest -n auto
-
-deploy:
-  # release for main pypi
-  # 3.6 is considered the build leader and does the deploy, otherwise there is
-  # a race and all versions will try to deploy
-  # OTOH if we ever need separate binary wheels then each version needs its
-  # own deploy
-- provider: pypi
-  user: ocrmypdf-travis
-  password:
-    secure: "DTFOmmNL6olA0+yXvp4u9jXZlZeqrJsJ0526jzqf4a3gZ6jnGTq5UI6WzRsslSyoMMfXKtHQebqHM6ogSgCZinyZ3ufHJo8fn9brxbEc2gsiWkbj5o3bGwdWMT1vNNE7XW0VCpw87rZ1EEwjl4FJHFudMlPR1yfU5+uq0k0PACo="
-  distributions: "sdist bdist_wheel"
-  on:
-    branch: master
-    tags: true
-    condition: $TRAVIS_PYTHON_VERSION == "3.6" && $TRAVIS_OS_NAME == "linux"
-  skip_upload_docs: true

MANIFEST.in

@@ -1,43 +0,0 @@
-# requirements
-recursive-include requirements *
-
-# git
-include .git_archival.txt
-
-# docker
-include .dockerignore
-recursive-include .docker *
-
-# tests
-include .coveragerc
-recursive-include tests *.bin
-recursive-include tests *.jpg
-recursive-include tests *.jsonl
-recursive-include tests *.png
-recursive-include tests *.pdf
-recursive-include tests *.py
-recursive-include tests *.rst
-recursive-include tests *.txt
-recursive-exclude tests/resources/private *
-
-# documentation
-include LICENSE
-include *.rst
-recursive-exclude .github *
-recursive-include docs *.py
-recursive-include docs *.rst
-recursive-include docs *.svg
-recursive-exclude docs/_build *
-
-
-# support files
-recursive-include src/ocrmypdf/data *
-include *.py
-exclude tasks.py
-recursive-exclude .travis *
-exclude .travis*
-
-
-# code
-exclude src/ocrmypdf/lib/_leptonica.py
-exclude scratch.py

README.md

@@ -1,7 +1,8 @@
-OCRmyPDF
-========
+<img src="docs/images/logo.svg" width="240" alt="OCRmyPDF">
 
-[![Travis build status][travis]](https://travis-ci.org/jbarlow83/OCRmyPDF) [![PyPI version][pypi]](https://pypi.org/project/ocrmypdf/) ![Homebrew version][homebrew] ![ReadTheDocs][docs]
+[![Build Status][azure]](https://dev.azure.com/jim0585/ocrmypdf/_build/latest?definitionId=2&branchName=master) [![PyPI version][pypi]](https://pypi.org/project/ocrmypdf/) ![Homebrew version][homebrew] ![ReadTheDocs][docs] ![Python versions][pyversions]
+
+[azure]: https://dev.azure.com/jim0585/ocrmypdf/_apis/build/status/jbarlow83.OCRmyPDF?branchName=master
 
 [travis]: https://travis-ci.org/jbarlow83/OCRmyPDF.svg?branch=master "Travis build status"
 
@@ -11,6 +12,8 @@ OCRmyPDF
 
 [docs]: https://readthedocs.org/projects/ocrmypdf/badge/?version=latest "RTD"
 
+[pyversions]: https://img.shields.io/pypi/pyversions/ocrmypdf "Supported Python versions"
+
 OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to be searched or copy-pasted.
 
 ```bash
@@ -35,7 +38,7 @@ Main features
 - Keeps the exact resolution of the original embedded images
 - When possible, inserts OCR information as a "lossless" operation without disrupting any other content
 - Optimizes PDF images, often producing files smaller than the input file
-- If requested deskews and/or cleans the image before performing OCR
+- If requested, deskews and/or cleans the image before performing OCR
 - Validates input and output files
 - Distributes work across all available CPU cores
 - Uses [Tesseract OCR](https://github.com/tesseract-ocr/tesseract) engine to recognize more than [100 languages](https://github.com/tesseract-ocr/tessdata)
@@ -47,7 +50,7 @@ For details: please consult the [documentation](https://ocrmypdf.readthedocs.io/
 Motivation
 ----------
 
-I searched the web for a free command line tool to OCR PDF files on Linux/UNIX: I found many, but none of them were really satisfying.
+I searched the web for a free command line tool to OCR PDF files: I found many, but none of them were really satisfying:
 
 - Either they produced PDF files with misplaced text under the image (making copy/paste impossible)
 - Or they did not handle accents and multilingual characters
@@ -62,7 +65,7 @@ I searched the web for a free command line tool to OCR PDF files on Linux/UNIX:
 Installation
 ------------
 
-Linux, UNIX, and macOS are supported. Windows is not directly supported but there is a Docker image available that runs on Windows.
+Linux, Windows, macOS and FreeBSD are supported. Docker images are also available.
 
 Users of Debian 9 or later or Ubuntu 16.10 or later may simply
 
@@ -76,7 +79,7 @@ and users of Fedora 29 or later may simply
 dnf install ocrmypdf
 ```
 
-and macOS users with Homebrew may simply
+and Homebrew users (macOS, Linux, Windows Subsystem for Linux) may simply
 
 ```bash
 brew install ocrmypdf
@@ -94,7 +97,10 @@ OCRmyPDF uses Tesseract for OCR, and relies on its language packs. For Linux use
 apt-cache search tesseract-ocr
 
 # Debian/Ubuntu users
-apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language back
+apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language pack
+
+# Arch Linux users
+pacman -S tesseract-data-eng tesseract-data-deu # Example: Install the English and German language packs
 ```
 
 You can then pass the `-l LANG` argument to OCRmyPDF to give a hint as to what languages it should search for. Multiple languages can be requested.
@@ -102,7 +108,7 @@ You can then pass the `-l LANG` argument to OCRmyPDF to give a hint as to what l
 Documentation and support
 -------------------------
 
-Once ocrmypdf is installed, the built-in help which explains the command syntax and options can be accessed via:
+Once OCRmyPDF is installed, the built-in help which explains the command syntax and options can be accessed via:
 
 ```bash
 ocrmypdf --help
@@ -110,31 +116,26 @@ ocrmypdf --help
 
 Our [documentation is served on Read the Docs](https://ocrmypdf.readthedocs.io/en/latest/index.html).
 
-If you detect an issue, please:
-
-- Check whether your issue is already known
-- If no problem report exists on github, please create one here: <https://github.com/jbarlow83/OCRmyPDF/issues>
-- Describe your problem thoroughly
-- Append the console output of the script when running the debug mode (`-v 1` option)
-- If possible provide your input PDF file as well as the content of the temporary folder (using a file sharing service like Dropbox)
+Please report issues on our [GitHub issues](https://github.com/jbarlow83/OCRmyPDF/issues) page, and follow the issue template for quick response.
 
 Requirements
 ------------
 
-Runs on CPython 3.5, 3.6 and 3.7. Requires external program installations of Ghostscript, Tesseract OCR, QPDF, and Leptonica. ocrmypdf is pure Python, but uses CFFI to portably generate library bindings.
+In addition to the required Python version (3.6+), OCRmyPDF requires external program installations of Ghostscript, Tesseract OCR, QPDF, and Leptonica. OCRmyPDF is pure Python, but uses CFFI to portably generate library bindings. OCRmyPDF works on pretty much everything: Linux, macOS, Windows and FreeBSD.
 
 Press & Media
 -------------
 
 - [Going paperless with OCRmyPDF](https://medium.com/@ikirichenko/going-paperless-with-ocrmypdf-e2f36143f46a)
 - [Converting a scanned document into a compressed searchable PDF with redactions](https://medium.com/@treyharris/converting-a-scanned-document-into-a-compressed-searchable-pdf-with-redactions-63f61c34fe4c)
-- [c't 1-2014, page 59](http://heise.de/-2279695): Detailed presentation of OCRmyPDF v1.0 in the leading German IT magazine c't
-- [heise Open Source, 09/2014: Texterkennung mit OCRmyPDF](http://heise.de/-2356670)
+- [c't 1-2014, page 59](https://heise.de/-2279695): Detailed presentation of OCRmyPDF v1.0 in the leading German IT magazine c't
+- [heise Open Source, 09/2014: Texterkennung mit OCRmyPDF](https://heise.de/-2356670)
+- [heise Durchsuchbare PDF-Dokumente mit OCRmyPDF erstellen](https://www.heise.de/ratgeber/Durchsuchbare-PDF-Dokumente-mit-OCRmyPDF-erstellen-4607592.html)
 
 Business enquiries
 ------------------
 
-OCRmyPDF would not be the software that it is today is without companies and users choosing to provide support for feature development and consulting enquiries. We are happy to discuss all enquiries, whether for extending the existing feature set, or integrating OCRmyPDF into a larger system.
+OCRmyPDF would not be the software that it is today without companies and users choosing to provide support for feature development and consulting enquiries. We are happy to discuss all enquiries, whether for extending the existing feature set, or integrating OCRmyPDF into a larger system.
 
 License
 -------

azure-pipelines.yml

@@ -0,0 +1,253 @@
+trigger:
+  tags:
+    include:
+      - v*
+  branches:
+    include:
+      - "*"
+    exclude:
+      - "travis"
+
+stages:
+  - stage: "Test"
+    jobs:
+      - job: Windows
+        pool:
+          vmImage: "vs2017-win2016"
+        strategy:
+          matrix:
+            Python36:
+              python.version: "3.6"
+            Python37:
+              python.version: "3.7"
+            Python38:
+              python.version: "3.8"
+        steps:
+          - task: UsePythonVersion@0
+            inputs:
+              versionSpec: "$(python.version)"
+          - pwsh: |
+              choco install --yes --no-progress --pre tesseract
+              choco install --yes --no-progress python3
+              choco install --yes --no-progress ghostscript
+              choco install --yes --no-progress pngquant
+            displayName: "Install system packages"
+          - pwsh: |
+              refreshenv
+              python -m pip install --upgrade pip wheel
+              python -m pip install -r requirements/main.txt -r requirements/test.txt .
+            displayName: "Install Python packages"
+          - pwsh: |
+              refreshenv
+              $env:pathext += ';.py'
+              # -n auto helps Windows
+              python -m pytest -n auto --junitxml=test.xml --cov=ocrmypdf --cov-report=xml
+            displayName: "Test"
+          - task: PublishTestResults@2
+            inputs:
+              testResultsFiles: "test.xml"
+              testRunTitle: "$(Agent.OS) - $(Build.DefinitionName) - Python $(python.version)"
+            condition: succeededOrFailed()
+      - job: "Ubuntu_1804"
+        pool:
+          vmImage: "ubuntu-18.04"
+        strategy:
+          matrix:
+            Python36:
+              python.version: "3.6"
+            Python37:
+              python.version: "3.7"
+            Python38:
+              python.version: "3.8"
+        steps:
+          - task: UsePythonVersion@0
+            inputs:
+              versionSpec: "$(python.version)"
+          - bash: |
+              sudo apt-get update
+              sudo apt-get install -y --no-install-recommends \
+                python3-software-properties \
+                curl \
+                ghostscript \
+                img2pdf \
+                libexempi3 \
+                libffi-dev \
+                liblept5 \
+                libsm6 libxext6 libxrender-dev \
+                pngquant \
+                poppler-utils \
+                tesseract-ocr \
+                tesseract-ocr-deu \
+                tesseract-ocr-eng \
+                unpaper \
+                zlib1g
+            displayName: "Install system packages"
+          - bash: |
+              curl https://bootstrap.pypa.io/get-pip.py | python3
+              pip3 install -r requirements/main.txt -r requirements/test.txt .
+            displayName: "Install Python packages"
+          - bash: |
+              tesseract --version
+            displayName: "Record versions"
+          - bash: |
+              # -n auto is slower on Linux and breaks on Python 3.8
+              pytest -n0 --junitxml=test.xml --cov=ocrmypdf --cov-report=xml
+            displayName: "Test"
+          - task: PublishTestResults@2
+            inputs:
+              testResultsFiles: "test.xml"
+              testRunTitle: "$(Agent.OS) - $(Build.DefinitionName) - Python $(python.version)"
+            condition: succeededOrFailed()
+      - job: "Ubuntu_1604"
+        pool:
+          vmImage: "ubuntu-16.04"
+        strategy:
+          matrix:
+            Python36:
+              python.version: "3.6"
+        steps:
+          - task: UsePythonVersion@0
+            inputs:
+              versionSpec: "$(python.version)"
+          - bash: |
+              sudo apt-get update
+              sudo apt-get install -y --no-install-recommends \
+                software-properties-common
+              sudo add-apt-repository -y ppa:alex-p/tesseract-ocr
+              sudo apt-get update
+              sudo apt-get install -y --no-install-recommends \
+                ghostscript \
+                img2pdf \
+                libexempi3 \
+                libffi-dev \
+                liblept5 \
+                libsm6 libxext6 libxrender-dev \
+                pngquant \
+                poppler-utils \
+                tesseract-ocr \
+                tesseract-ocr-deu \
+                tesseract-ocr-eng \
+                unpaper \
+                zlib1g
+            displayName: "Install system packages"
+          - bash: |
+              curl https://bootstrap.pypa.io/get-pip.py | python3
+              pip3 install -r requirements/main.txt -r requirements/test.txt .
+            displayName: "Install Python packages"
+          - bash: |
+              tesseract --version
+            displayName: "Record versions"
+          - bash: |
+              # -n auto is slower on Linux and breaks on Python 3.8
+              pytest -n0 --junitxml=test.xml --cov=ocrmypdf --cov-report=xml
+            displayName: "Test"
+          - task: PublishTestResults@2
+            inputs:
+              testResultsFiles: "test.xml"
+              testRunTitle: "$(Agent.OS) - $(Build.DefinitionName) - Python $(python.version)"
+            condition: succeededOrFailed()
+      - job: "macOS_Mojave"
+        pool:
+          vmImage: "macos-10.14"
+        strategy:
+          matrix:
+            Python37:
+              python.version: "3.7"
+            # Python38:
+            #   python.version: "3.8"
+        steps:
+          # https://github.com/actions/virtual-environments/issues/664
+          # - task: UsePythonVersion@0
+          #   inputs:
+          #     versionSpec: "$(python.version)"
+          - bash: |
+              brew update
+              brew unlink python@2
+              brew install \
+                exempi \
+                ghostscript \
+                jbig2enc \
+                leptonica \
+                openjpeg \
+                pngquant \
+                python \
+                tesseract \
+                unpaper
+            displayName: "Install system packages"
+          - bash: |
+              pip3 install --upgrade pip
+              pip3 install -r requirements/main.txt -r requirements/test.txt .
+            displayName: "Install Python packages"
+          - bash: |
+              tesseract --version
+            displayName: "Record versions"
+          - bash: pytest -nauto --junitxml=test.xml --cov=ocrmypdf --cov-report=xml
+            displayName: "Test"
+          - task: PublishTestResults@2
+            inputs:
+              testResultsFiles: "test.xml"
+              testRunTitle: "$(Agent.OS) - $(Build.DefinitionName) - Python $(python.version)"
+            condition: succeededOrFailed()
+          - task: PublishCodeCoverageResults@1
+            inputs:
+              codeCoverageTool: Cobertura
+              summaryFileLocation: "$(System.DefaultWorkingDirectory)/**/coverage.xml"
+
+  - stage: "Artifacts"
+    jobs:
+      - job: "sdist_wheel"
+        pool:
+          vmImage: "ubuntu-18.04"
+        steps:
+          - task: UsePythonVersion@0
+            inputs:
+              versionSpec: "3.7"
+          - bash: |
+              python -m pip install --upgrade pip wheel
+              python setup.py sdist bdist_wheel
+          - publish: dist
+            artifact: sdist_wheel
+
+  - stage: "Deploy"
+    jobs:
+      - deployment: "PyPI"
+        pool:
+          vmImage: "ubuntu-18.04"
+        environment: "deploy"
+        strategy:
+          runOnce:
+            deploy:
+              steps:
+                - download: current
+                  artifact: sdist_wheel
+                - script: |
+                    mkdir -p dist
+                    mv $(Pipeline.Workspace)/sdist_wheel/* dist
+                  displayName: "Move dist files"
+                - task: UsePythonVersion@0
+                  inputs:
+                    versionSpec: "3.8"
+                    architecture: x64
+                - script: |
+                    pip install --upgrade twine
+                  displayName: "Generate artifacts"
+                - script: |
+                    cat <<FILE >.pypirc
+                    [distutils]
+                    index-servers =
+                        pypi
+
+                    [pypi]
+                    username: __token__
+                    password: $(TOKEN_PYPI)
+
+                    FILE
+                  displayName: "Generate PyPI auth file"
+                - script: |
+                    python -m twine upload --config-file .pypirc dist/*
+                  displayName: "Upload to PyPI"
+                  condition: and(succeeded(), startsWith(variables['Build.SourceBranch'], 'refs/tags/'))
+                - script: |
+                    curl -X POST -d "token=$(TOKEN_RTD)" https://readthedocs.org/api/v2/webhook/pikepdf/39557/
+                  displayName: "Trigger ReadTheDocs"
+                  condition: and(succeeded(), or(startsWith(variables['Build.SourceBranch'], 'refs/tags/'), startsWith(variables['Build.SourceBranch'], 'refs/heads/master')))

debian/copyright

@@ -79,7 +79,7 @@ Copyright: held by the contributors to the Wikipedia article "Optical character
  (epson.pdf generated from Wikipedia article as of 2016-09-14)
 License: CC-BY-SA-3.0
 
-Files: tests/resources/typewriter.png tests/resources/2400dpi.pdf
+Files: tests/resources/typewriter.png tests/resources/2400dpi.pdf tests/resources/3small.pdf
 Copyright: (C) 2005 Ellywa
 License: GFDL-1.2+ or CC-BY-SA-1.0 or CC-BY-SA-2.0 or CC-BY-SA-2.5 or CC-BY-SA-3.0
 
@@ -87,7 +87,7 @@ Files: tests/resources/overlay.pdf
 Copyright: (C) 2017 Max Anderson
 License: Expat
 
-Files: tests/resources/baiona*.png
+Files: tests/resources/baiona*.png tests/resources/3small.pdf
 Copyright: (C) 2014 Euskaldunaa
 License: CC-BY-SA-4.0
 
@@ -95,12 +95,6 @@ Files: tests/resources/vector.pdf
 Copyright: (C) 2018 Catscratch
 License: Expat
 
-Files: test/resources/enron*.pdf
-Copyright: EnronData.org
-License: CC-BY-3.0
- See: https://enrondata.readthedocs.io/en/latest/data/edo-enron-email-pst-dataset/
-Comment: Unprocessed.
-
 Files: src/ocrmypdf/data/sRGB.icc
 Copyright:  Kai-Uwe Behrmann <www.behrmann.name>
             Marti Maria <www.littlecms.com>

docs/advanced.rst

@@ -148,7 +148,8 @@ In addition to tesseract, OCRmyPDF uses the following external binaries:
 
 -  ``gs`` (Ghostscript)
 -  ``unpaper``
--  ``qpdf``
+-  ``pngquant``
+-  ``jbig2``
 
 In each case OCRmyPDF will search the ``PATH`` environment variable to
 locate the binaries.
@@ -322,15 +323,23 @@ working files on a per page basis have the page number as a prefix
 (starting with page 1), an infix indicates the processing stage, and a
 suffix indicates the file type. Some important files include:
 
--  ``.page.png`` - what the input page looks like
--  ``.image`` - the image we will show the user if we are in a mode that
-   changes the final appearance; may be in one of several image formats
--  ``.text.pdf`` - the OCR file; this will load as a blank page but
-   should have visible text if checked with a tool like pdftotext or
-   pdfminder.six
--  ``.ocr.png`` - the file that is sent to Tesseract for OCR; depending
+-  ``_rasterize.png`` - what the input page looks like
+-  ``_ocr.png`` - the file that is sent to Tesseract for OCR; depending
    on arguments this may differ from the presentation image
--  ``layers.rendered.pdf`` - the composite PDF, before metadata repair
-   and optimization
+-  ``_pp_deskew.png`` - the image, after deskewing
+-  ``_pp_clean.png`` - the image, after cleaning with unpaper
+-  ``_ocr_tess.pdf`` - the OCR file; appears as a blank page with invisible
+   text embedded
+-  ``_ocr_tess.txt`` - the OCR text (not necessarily all text on the page,
+   if the page is mixed format)
+-  ``fix_docinfo.pdf`` - a temporary file created to fix the PDF DocumentInfo
+   data structure
+-  ``graft_layers.pdf`` - the rendered PDF with OCR layers grafted on
+-  ``pdfa.pdf`` - ``graft_layers.pdf`` after conversion to PDF/A
+-  ``pdfa.ps`` - a PostScript file used by Ghostscript for PDF/A conversion
+-  ``optimize.pdf`` - the PDF generated before optimization
+-  ``optimize.out.pdf`` - the PDF generated by optimization
+-  ``origin`` - the input file
+-  ``origin.pdf`` - the input file or the input image converted to PDF
 -  ``images/*`` - images extracted during the optimization process; here
    the prefix indicates a PDF object ID not a page number

docs/api.rst

@@ -20,7 +20,7 @@ and largely have the same functions.
 
     import ocrmypdf
 
-    ocrmypdf.run('input.pdf', 'output.pdf', deskew=True)
+    ocrmypdf.ocr('input.pdf', 'output.pdf', deskew=True)
 
 With a few exceptions, all of the command line arguments are available
 and may be passed as equivalent keywords.
@@ -31,13 +31,15 @@ Instead, output should be managed by configuring logging.
 Parent process requirements
 ---------------------------
 
-The :func:`ocrmypdf.run` function runs OCRmyPDF similar to command line
-execution. To do this, it will: - create a monitoring thread - create
-worker processes (forking itself) - manage the signal flags of worker
-processes 0 execute other subprocesses (forking and executing other
-programs)
+The :func:`ocrmypdf.ocr` function runs OCRmyPDF similar to command line
+execution. To do this, it will:
 
-The Python process that calls ``ocrmypdf.run()`` must be sufficiently
+- create a monitoring thread
+- create worker processes (forking itself)
+- manage the signal flags of worker processes
+- execute other subprocesses (forking and executing other programs)
+
+The Python process that calls ``ocrmypdf.ocr()`` must be sufficiently
 privileged to perform these actions. If it is not, ``ocrmypdf()`` will
 fail.
 
@@ -45,10 +47,18 @@ There is no currently no option to manage how jobs are scheduled other
 than the argument ``jobs=`` which will limit the number of worker
 processes.
 
-Forking a child process to call ``ocrmypdf.run()`` is suggested. That
+Forking a child process to call ``ocrmypdf.ocr()`` is suggested. That
 way your application will survive and remain interactive even if
 OCRmyPDF does not.
 
+.. warning::
+
+    On Windows, the script that calls ``ocrmypdf.ocr()`` must be protected
+    by an "ifmain" guard (``if __name__ == '__main__'``) or you must use
+    ``ocrmypdf.ocr(...use_threads=True)``. If you do not take at least one
+    of these steps, Windows process semantics will prevent OCRmyPDF from working
+    correctly.
+
 Logging
 -------
 
@@ -69,7 +79,8 @@ Progress monitoring
 OCRmyPDF uses the ``tqdm`` package to implement its progress bars.
 :func:`ocrmypdf.configure_logging` will set up logging output to
 ``sys.stderr`` in a way that is compatible with the display of the
-progress bar.
+progress bar. Use ``ocrmypdf.ocr(...progress_bar=False)`` to disable
+the progress bar.
 
 Exceptions
 ----------
@@ -88,14 +99,14 @@ When OCRmyPDF succeeds conditionally, it returns an integer exit code.
 Reference
 ---------
 
-.. autofunction:: ocrmypdf.run
+.. autofunction:: ocrmypdf.ocr
 
 .. autoclass:: ocrmypdf.Verbosity
     :members:
     :undoc-members:
 
-.. autoclass:: ocrmypdf.ExitCode
+.. autofunction:: ocrmypdf.configure_logging
+
+.. automodule:: ocrmypdf.exceptions
     :members:
     :undoc-members:
-
-.. autofunction:: ocrmypdf.configure_logging

docs/batch.rst

@@ -47,7 +47,7 @@ where the PDFs are stored):
 
 .. code-block:: bash
 
-   find . -printf '%p' -name '*.pdf' -exec docker run --rm -v <host dir>:<container dir> jbarlow83/ocrmypdf-alpine '<container dir>/{}' '<container dir>/{}' \;
+   find . -printf '%p' -name '*.pdf' -exec docker run --rm -v <host dir>:<container dir> jbarlow83/ocrmypdf '<container dir>/{}' '<container dir>/{}' \;
 
 This only runs one ``ocrmypdf`` process at a time. This variation uses
 ``find`` to create a directory list and ``parallel`` to parallelize runs
@@ -57,59 +57,20 @@ of ``ocrmypdf``, again updating files in place.
 
    find . -name '*.pdf' | parallel --tag -j 2 ocrmypdf '{}' '{}'
 
+In a Windows batch file, use
+
+.. code-block:: bat
+
+   for /r %%f in (*.pdf) do ocrmypdf %%f %%f
+
 Sample script
 -------------
 
 This user contributed script also provides an example of batch
 processing.
 
-.. code-block:: python
-
-   #!/usr/bin/env python3
-   # Walk through directory tree, replacing all files with OCR'd version
-   # Contributed by DeliciousPickle@github
-
-   import logging
-   import os
-   import subprocess
-   import sys
-
-   script_dir = os.path.dirname(os.path.realpath(__file__))
-   print(script_dir + '/ocr-tree.py: Start')
-
-   if len(sys.argv) > 1:
-       start_dir = sys.argv[1]
-   else:
-       start_dir = '.'
-
-   if len(sys.argv) > 2:
-       log_file = sys.argv[2]
-   else:
-       log_file = script_dir + '/ocr-tree.log'
-
-   logging.basicConfig(
-           level=logging.INFO, format='%(asctime)s %(message)s',
-           filename=log_file, filemode='w')
-
-   for dir_name, subdirs, file_list in os.walk(start_dir):
-       logging.info('\n')
-       logging.info(dir_name + '\n')
-       os.chdir(dir_name)
-       for filename in file_list:
-           file_ext = os.path.splitext(filename)[1]
-           if file_ext == '.pdf':
-               full_path = dir_name + '/' + filename
-               print(full_path)
-               cmd = ["ocrmypdf",  "--deskew", filename, filename]
-               logging.info(cmd)
-               proc = subprocess.run(
-                   cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
-               result = proc.stdout
-               if proc.returncode == 6:
-                   print("Skipped document because it already contained text")
-               elif proc.returncode == 0:
-                   print("OCR complete")
-               logging.info(result)
+.. literalinclude:: ../misc/batch.py
+    :caption: misc/batch.py
 
 Synology DiskStations
 ---------------------
@@ -125,62 +86,8 @@ products use ARM or Power processors and do not support Docker. Further
 adjustments might be needed to deal with the Synology's relatively
 limited CPU and RAM.
 
-.. code-block:: python
-
-   #!/bin/env python3
-   # Contributed by github.com/Enantiomerie
-
-   # script needs 2 arguments
-   # 1. source dir with *.pdf - default is location of script
-   # 2. move dir where *.pdf and *_OCR.pdf are moved to
-
-   import logging
-   import os
-   import subprocess
-   import sys
-   import time
-   import shutil
-
-   script_dir = os.path.dirname(os.path.realpath(__file__))
-   timestamp = time.strftime("%Y-%m-%d-%H%M_")
-   log_file = script_dir + '/' + timestamp + 'ocrmypdf.log'
-   logging.basicConfig(level=logging.INFO, format='%(asctime)s %(message)s', filename=log_file, filemode='w')
-
-   if len(sys.argv) > 1:
-       start_dir = sys.argv[1]
-   else:
-       start_dir = '.'
-
-   for dir_name, subdirs, file_list in os.walk(start_dir):
-       logging.info('\n')
-       logging.info(dir_name + '\n')
-       os.chdir(dir_name)
-       for filename in file_list:
-           file_ext = os.path.splitext(filename)[1]
-           if file_ext == '.pdf':
-               full_path = dir_name + '/' + filename
-               file_noext = os.path.splitext(filename)[0]
-               timestamp_OCR = time.strftime("%Y-%m-%d-%H%M_OCR_")
-               filename_OCR = timestamp_OCR + file_noext + '.pdf'
-               docker_mount = dir_name + ':/home/docker'
-   # create string for pdf processing
-   # diskstation needs a user:group docker:docker. find uid:gid of your diskstation docker:docker with id docker.
-   # use this uid:gid in -u flag
-   # rw rights for docker:docker at source dir are also necessary
-   # the script is processed as root user via chron
-               cmd = ['docker', 'run', '--rm', '-v', docker_mount, '-u=1030:65538', 'jbarlow83/ocrmypdf', , '--deskew' , filename, filename_OCR]
-               logging.info(cmd)
-               proc = subprocess.run(cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
-               result = proc.stdout.read()
-               logging.info(result)
-               full_path_OCR = dir_name + '/' + filename_OCR
-               os.chmod(full_path_OCR, 0o666)
-               os.chmod(full_path, 0o666)
-               full_path_OCR_archive = sys.argv[2]
-               full_path_archive = sys.argv[2] + '/no_ocr'
-               shutil.move(full_path_OCR,full_path_OCR_archive)
-               shutil.move(full_path, full_path_archive)
-   logging.info('Finished.\n')
+.. literalinclude:: ../misc/synology.py
+    :caption: misc/synology.py - Sample script for Synology DiskStations
 
 Huge batch jobs
 ---------------
@@ -192,38 +99,83 @@ and all inquiries are appreciated.
 Hot (watched) folders
 =====================
 
-To set up a "hot folder" that will trigger OCR for every file inserted,
-use a program like Python
-`watchdog <https://pypi.python.org/pypi/watchdog>`__ (supports all major
-OS).
+Watched folders with watcher.py
+-------------------------------
 
-One could then configure a scanner to automatically place scanned files
-in a hot folder, so that they will be queued for OCR and copied to the
-destination.
+OCRmyPDF has a folder watcher called watcher.py, which is currently included in source
+distributions but not part of the main program. It may be used natively or may run
+in a Docker container. Native instances tend to give better performance. watcher.py
+works on all platforms.
+
+Users may need to customize the script to meet their requirements.
 
 .. code-block:: bash
 
-   pip install watchdog
+    pip3 install -r requirements/watcher.txt
 
-watchdog installs the command line program ``watchmedo``, which can be
-told to run ``ocrmypdf`` on any .pdf added to the current directory
-(``.``) and place the result in the previously created ``out/`` folder.
+    env OCR_INPUT_DIRECTORY=/mnt/input-pdfs \
+        OCR_OUTPUT_DIRECTORY=/mnt/output-pdfs \
+        OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1 \
+        python3 watcher.py
+
+.. csv-table:: watcher.py environment variables
+    :header: "Environment variable", "Description"
+    :widths: 50, 50
+
+    "OCR_INPUT_DIRECTORY", "Set input directory to monitor (recursive)"
+    "OCR_OUTPUT_DIRECTORY", "Set output directory (should not be under input)"
+    "OCR_ON_SUCCESS_DELETE", "This will delete the input file if the exit code is 0 (OK)"
+    "OCR_OUTPUT_DIRECTORY_YEAR_MONTH", "This will place files in the output in ``{output}/{year}/{month}/{filename}``"
+    "OCR_DESKEW", "Apply deskew to crooked input PDFs"
+    "OCR_JSON_SETTINGS", "A JSON string specifying any other arguments for ``ocrmypdf.ocr``"
+    "OCR_POLL_NEW_FILE_SECONDS", "Polling interval"
+    "OCR_LOGLEVEL", "Level of log messages to report"
+
+One could configure a networked scanner or scanning computer to drop files in the
+watched folder.
+
+Watched folders with Docker
+---------------------------
+
+The watcher service is included in the OCRmyPDF Docker image. To run it:
 
 .. code-block:: bash
 
-   cd hot-folder
-   mkdir out
-   watchmedo shell-command \
-       --patterns="*.pdf" \
-       --ignore-directories \
-       --command='ocrmypdf "${watch_src_path}" "out/${watch_src_path}" ' \
-       .  # don't forget the final dot
+    docker run \
+        -v <path to files to convert>:/input \
+        -v <path to store results>:/output \
+        -e OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1 \
+        -e OCR_ON_SUCCESS_DELETE=1 \
+        -e OCR_DESKEW=1 \
+        -e PYTHONUNBUFFERED=1 \
+        -it --entrypoint python3 \
+        jbarlow83/ocrmypdf \
+        watcher.py
 
-For more complex behavior you can write a Python script around to use
-the watchdog API.
+This service will watch for a file that matches ``/input/\*.pdf`` and will
+convert it to a OCRed PDF in ``/output/``. The parameters to this image are:
 
-On file servers, you could configure watchmedo as a system service so it
-will run all the time.
+.. csv-table:: watcher.py parameters for Docker
+    :header: "Parameter", "Description"
+    :widths: 50, 50
+
+    "``-v <path to files to convert>:/input``", "Files placed in this location will be OCRed"
+    "``-v <path to store results>:/output``", "This is where OCRed files will be stored"
+    "``-e OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1``", "Define environment variable OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1"
+    "``-e OCR_ON_SUCCESS_DELETE=1``", "Define environment variable"
+    "``-e OCR_DESKEW=1``", "Define environment variable"
+    "``-e PYTHONBUFFERED=1``", "This will force STDOUT to be unbuffered and allow you to see messages in docker logs"
+
+This service relies on polling to check for changes to the filesystem. It
+may not be suitable for some environments, such as filesystems shared on a
+slow network.
+
+A configuration manager such as Docker Compose could be used to ensure that the
+service is always available.
+
+.. literalinclude:: ../misc/docker-compose.example.yml
+    :language: yaml
+    :caption: misc/docker-compose.example.yml
 
 Caveats
 -------
@@ -244,9 +196,19 @@ Caveats
 Alternatives
 ------------
 
+-  On Linux, `systemd user services <https://wiki.archlinux.org/index.php/Systemd/User>`__
+   can be configured to automatically perform OCR on a collection of files.
+
 -  `Watchman <https://facebook.github.io/watchman/>`__ is a more
    powerful alternative to ``watchmedo``.
 
+AWS Lambda is not viable
+------------------------
+
+AWS Lambda and its equivalents have low limits on execution time and payload
+size, relative to OCRmyPDF's needs. As of this writing, the request/response
+payload for AWS Lambda was 6 MB, which means many PDFs will not fit.
+
 macOS Automator
 ===============
 
@@ -256,8 +218,7 @@ of Automator, the ``PATH`` may be set differently your Terminal's
 ``PATH``; you may need to explicitly set the PATH to include
 ``ocrmypdf``. The following example may serve as a starting point:
 
-|Example macOS Automator script|
+.. figure:: images/macos-workflow.png
+    :alt: Example macOS Automator workflow
 
 You may customize the command sent to ocrmypdf.
-
-.. |Example macOS Automator script| image:: images/macos-workflow.png

docs/contributing.rst

@@ -0,0 +1,57 @@
+=======================
+Contributing guidelines
+=======================
+
+Contributions are welcome!
+
+Big changes
+===========
+
+Please open a new issue to discuss or propose a major change. Not only is it fun
+to discuss big ideas, but we might save each other's time too. Perhaps some of the
+work you're contemplating is already half-done in a development branch.
+
+Code style
+==========
+
+We use PEP8, ``black`` for code formatting and ``isort`` for import sorting. The
+settings for these programs are in ``pyproject.toml`` and ``setup.cfg``. Pull
+requests should follow the style guide. One difference we use from "black" style
+is that strings shown to the user are always in double quotes (``"``) and strings
+for internal uses are in single quotes (``'``).
+
+Tests
+=====
+
+New features should come with tests that confirm their correctness.
+
+New Python dependencies
+=======================
+
+If you are proposing a change that will require a new Python dependency, we
+prefer dependencies that are already packaged by Debian or Red Hat. This makes
+life much easier for our downstream package maintainers.
+
+Python dependencies must also be GPLv3 compatible.
+
+New non-Python dependencies
+===========================
+
+OCRmyPDF uses several external programs (Tesseract, Ghostscript and others) for
+its functionality. In general we prefer to avoid adding new external programs.
+
+Style guide: Is it OCRmyPDF or ocrmypdf?
+========================================
+
+The program/project is OCRmyPDF and the name of the executable or library is ocrmypdf.
+
+Known ports/packagers
+=====================
+
+OCRmyPDF has been ported to many platforms already. If you are interesting in
+porting to a new platform, check with
+`Repology <https://repology.org/projects/?search=ocrmypdf>`__ to see the status
+of that platform.
+
+Packager maintainers, please ensure that the command line completion scripts in
+``misc/`` are installed.

docs/cookbook.rst

@@ -56,7 +56,11 @@ portrait pages.
     ocrmypdf --rotate-pages myfile.pdf myfile.pdf
 
 You can increase (decrease) the parameter ``--rotate-pages-threshold``
-to make page rotation more (less) aggressive.
+to make page rotation more (less) aggressive. The threshold number is the ratio
+of how confidence the OCR engine is that the document image should be changed,
+compared to kept the same. A value of ``15.0`` is the default, and is fairly
+conservative. A value of ``2.0`` will produce more rotations, and more false
+positives.
 
 If the page is "just a little off horizontal", like a crooked picture,
 then you want ``--deskew``. ``--rotate-pages`` is for when the cardinal
@@ -89,6 +93,18 @@ This produces a file named "output.pdf" and a companion text file named
 
     ocrmypdf --sidecar output.txt input.pdf output.pdf
 
+.. note::
+
+    The sidecar file contains the **OCR text** found by OCRmyPDF. If the document
+    contains pages that already have text, that text will not appear in the
+    sidecar. If the option ``--pages`` is used, only those pages on which OCR
+    was performed will be included in the sidecar. If certain pages were skipped
+    because of options like ``--skip-big`` or ``--tesseract-timeout``, those pages
+    will not be in the sidecar.
+
+    To extract all text from a PDF, whether generated from OCR or otherwise,
+    use a program like Poppler's ``pdftotext`` or ``pdfgrep``.
+
 OCR images, not PDFs
 --------------------
 
@@ -165,8 +181,8 @@ might remove desirable content, especially from poor quality scans.
 -  ``--deskew`` will correct pages were scanned at a skewed angle by
    rotating them back into place. Skew determination and correction is
    performed using `Postl's variance of line
-   sums <http://www.leptonica.com/skew-measurement.html>`__ algorithm as
-   implemented in `Leptonica <http://www.leptonica.com/index.html>`__.
+   sums <http://www.leptonica.org/skew-measurement.html>`__ algorithm as
+   implemented in `Leptonica <http://www.leptonica.org/index.html>`__.
 -  ``--clean`` uses
    `unpaper <https://www.flameeyes.eu/projects/unpaper>`__ to clean up
    pages before OCR, but does not alter the final output. This makes it
@@ -216,6 +232,41 @@ processing or PDF/A conversion.
 
     ocrmypdf --tesseract-timeout=0 --remove-background input.pdf output.pdf
 
+Optimize images without performing OCR
+--------------------------------------
+
+You can also optimize all images without performing any OCR:
+
+.. code-block:: bash
+
+    ocrmypdf --tesseract-timeout=0 --optimize 3 --skip-text input.pdf output.pdf
+
+Perform OCR only certain pages
+------------------------------
+
+You can ask OCRmyPDF to only apply OCR to certain pages.
+
+.. code-block:: bash
+
+    ocrmypdf --pages 2,3,13-17 input.pdf output.pdf
+
+Hyphens denote a range of pages and commas separate page numbers. If you prefer
+to use spaces, quote all of the page numbers: ``--pages '2, 3, 5, 7'``.
+
+OCRmyPDF will warn if your list of page numbers contains duplicates or
+overlap pages. OCRmyPDF does not currently account for document page numbers,
+such as an introduction section of a book that uses Roman numerals. It simply
+counts the number of virtual pieces of paper since the start.
+
+Regardless of the argument to ``--pages``, OCRmyPDF will optimize all pages in
+the file and convert it to PDF/A, unless you disable those options. In this
+example, we want to OCR only the title and otherwise change the PDF as little
+as possible:
+
+.. code-block:: bash
+
+    ocrmypdf --pages 1 --output-type pdf --optimize 0 input.pdf output.pdf
+
 Redo existing OCR
 =================
 

docs/docker.rst

@@ -22,21 +22,20 @@ Installing the Docker image
 If you have `Docker <https://docs.docker.com/>`__ installed on your
 system, you can install a Docker image of the latest release.
 
-The recommended OCRmyPDF Docker image is currently named
-``ocrmypdf-alpine``:
-
-.. code-block:: bash
-
-   docker pull jbarlow83/ocrmypdf-alpine
-
-Follow the Docker installation instructions for your platform. If you
-can run this command successfully, your system is ready to download and
+If you can run this command successfully, your system is ready to download and
 execute the image:
 
 .. code-block:: bash
 
    docker run hello-world
 
+The recommended OCRmyPDF Docker image is currently named ``ocrmypdf``:
+
+.. code-block:: bash
+
+   docker pull jbarlow83/ocrmypdf
+
+
 OCRmyPDF will use all available CPU cores. By default, the VirtualBox
 machine instance on Windows and macOS has only a single CPU core
 enabled. Use the VirtualBox Manager to determine the name of your Docker
@@ -51,64 +50,67 @@ CPUs:
    docker-machine start "yourVM"
    eval $(docker-machine env "yourVM")
 
+See the Docker documentation for
+`adjusting memory and CPU on other platforms <https://docs.docker.com/config/containers/resource_constraints/>`__.
+
 Using the Docker image on the command line
 ==========================================
 
-**Unlike typical Docker containers**, in this mode we are using the
-OCRmyPDF Docker container is intended to be emphemeral – it runs for one
-OCR job and then terminates, just like a command line program. We are
-using Docker as a way of delivering an application, not a server.
+**Unlike typical Docker containers**, in this section the OCRmyPDF Docker
+container is emphemeral – it runs for one OCR job and terminates, just like a
+command line program. We are using Docker to deliver an application (as opposed
+to the more conventional case, where a Docker container runs as a server).
 
 To start a Docker container (instance of the image):
 
 .. code-block:: bash
 
-   docker tag jbarlow83/ocrmypdf-alpine ocrmypdf
-   docker run --rm -i ocrmypdf (... all other arguments here...)
+   docker tag jbarlow83/ocrmypdf ocrmypdf
+   docker run --rm -i ocrmypdf (... all other arguments here...) - -
 
 For convenience, create a shell alias to hide the Docker command. It is
-easier to send the input file to file stdin and read the output from
-stdout – this avoids the occasionally messy permission issues with
-Docker entirely.
+easier to send the input file as stdin and read the output from
+stdout – **this avoids the messy permission issues with Docker entirely**.
 
 .. code-block:: bash
 
-   alias ocrmypdf='docker run --rm -i ocrmypdf'
-   ocrmypdf --version  # runs docker version
-   ocrmypdf <input.pdf >output.pdf
+   alias docker_ocrmypdf='docker run --rm -i ocrmypdf'
+   docker_ocrmypdf --version  # runs docker version
+   docker_ocrmypdf - - <input.pdf >output.pdf
 
 Or in the wonderful `fish shell <https://fishshell.com/>`__:
 
 .. code-block:: fish
 
-   alias ocrmypdf 'docker run --rm ocrmypdf'
-   funcsave ocrmypdf
+   alias docker_ocrmypdf 'docker run --rm ocrmypdf'
+   funcsave docker_ocrmypdf
 
 Alternately, you could mount the local current working directory as a
 Docker volume:
 
 .. code-block:: bash
 
-   docker run --rm -v $(pwd):/data ocrmypdf /data/input.pdf /data/output.pdf
+   alias docker_ocrmypdf='docker run --rm  -i --user "$(id -u):$(id -g)" --workdir /data -v "$PWD:/data" ocrmypdf'
+   docker_ocrmypdf /data/input.pdf /data/output.pdf
 
 .. _docker-lang-packs:
 
 Adding languages to the Docker image
 ====================================
 
-By default the Docker image includes English, German and Simplified
-Chinese, the most popular languages for OCRmyPDF users based on
-feedback. You may add other languages by creating a new Dockerfile based
-on the public one:
+By default the Docker image includes English, German, Simplified Chinese,
+French, Portuguese and Spanish, the most popular languages for OCRmyPDF
+users based on feedback. You may add other languages by creating a new
+Dockerfile based on the public one:
 
 .. code-block:: dockerfile
 
-   FROM jbarlow83/ocrmypdf-alpine
+   FROM jbarlow83/ocrmypdf
 
    # Add French
-   RUN apk add tesseract-ocr-data-fra
+   RUN apt install tesseract-ocr-fra
 
-You can also copy training data to ``/usr/share/tessdata``.
+You can also copy training data to ``/usr/share/tesseract-ocr/<tesseract version>/tessdata``.
 
 Executing the test suite
 ========================
@@ -117,17 +119,16 @@ The OCRmyPDF test suite is installed with image. To run it:
 
 .. code-block:: bash
 
-   docker run --entrypoint python3 jbarlow83/ocrmypdf-alpine setup.py test
+   docker run --entrypoint python3  jbarlow83/ocrmypdf -m pytest
 
 Accessing the shell
 ===================
 
-``bash`` is not installed in the image. To use the busybox shell in the
-Docker image:
+To use the bash shell in the Docker image:
 
 .. code-block:: bash
 
-   docker run -it --entrypoint busybox  jbarlow83/ocrmypdf-alpine sh
+   docker run -it --entrypoint bash  jbarlow83/ocrmypdf
 
 Using the OCRmyPDF web service wrapper
 ======================================
@@ -137,7 +138,12 @@ service. The webservice may be launched as follows:
 
 .. code-block:: bash
 
-   docker run --entrypoint python3 -p 5000:5000 jbarlow83/ocrmypdf-alpine webservice.py
+   docker run --entrypoint python3 -p 5000:5000  jbarlow83/ocrmypdf webservice.py
+
+This will configure the machine to listen on port 5000. On Linux machines
+this is port 5000 of localhost. On macOS or Windows machines running
+Docker, this is port 5000 of the virtual machine that runs your Docker
+images. You can find its IP address using the command ``docker-machine ip``.
 
 Unlike command line usage this program will open a socket and wait for
 connections.
@@ -162,14 +168,3 @@ also licensed in this way.
 
 In addition to the above, please read our
 :ref:`general remarks on using OCRmyPDF as a service <ocr-service>`.
-
-Ubuntu-based Docker image
-=========================
-
-A Ubuntu-based OCRmyPDF image is also available. The main advantage this
-image offers is that it supports manylinux Python wheels (which are not
-supported on Alpine Linux). This may be useful for plugins.
-
-.. code-block:: bash
-
-   docker pull jbarlow83/ocrmypdf

docs/errors.rst

@@ -22,14 +22,20 @@ As the error message suggests, your options are:
 -  ``ocrmypdf --skip-text`` to skip OCR and other processing on any
    pages that contain text. Text pages will be copied into the output
    PDF without modification.
+-  ``ocrmypdf --redo-ocr`` to scan the file for any existing OCR
+   (non-printing text), remove it, and do OCR again. This is one way
+   to take advantage of improvements in OCR accuracy. Printable vector
+   text is excluded from OCR, so this can be used on files that contain
+   a mix of digital and scanned files.
+
 
 Input file 'filename' is not a valid PDF
 ========================================
 
-OCRmyPDF passes files through qpdf, a program that fixes errors in PDFs,
-before it tries to work on them. In most cases this happens because the
-PDF is corrupt and truncated (incomplete file copying) and not much can
-be done.
+OCRmyPDF checks files with pikepdf, a library that in turn uses libqpdf to fixes
+errors in PDFs, before it tries to work on them. In most cases this happens
+because the PDF is corrupt and truncated (incomplete file copying) and not much
+can be done.
 
 You can try rewriting the file with Ghostscript:
 

docs/images/logo.svg

@@ -0,0 +1,75 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<svg width="100%" height="100%" viewBox="0 0 503 227" version="1.1" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" xml:space="preserve" xmlns:serif="http://www.serif.com/" style="fill-rule:evenodd;clip-rule:evenodd;stroke-linecap:round;stroke-linejoin:round;stroke-miterlimit:1.5;">
+    <g id="svg" transform="matrix(0.965977,0,0,0.807602,0,0)">
+        <rect x="0" y="0" width="520" height="280" style="fill:white;"/>
+        <g transform="matrix(1.03522,0,0,1.23823,-69.7528,-83.422)">
+            <g transform="matrix(1,0,0,1,243.977,20.0703)">
+                <g id="Page">
+                    <g transform="matrix(0.961773,0,0,1.05962,6.19811,-3.01071)">
+                        <path d="M328.5,97.682C328.5,96.465 327.983,95.296 327.056,94.418C320.026,87.758 289.442,58.78 282.228,51.944C281.251,51.019 279.901,50.496 278.49,50.496C264.493,50.496 188.083,50.496 167.339,50.496C164.468,50.496 162.141,52.609 162.141,55.214C162.141,83.051 162.141,225.565 162.141,253.4C162.141,256.005 164.468,258.117 167.338,258.117C192.242,258.117 299.159,258.117 323.538,258.117C326.278,258.117 328.5,256.101 328.5,253.613C328.5,229.268 328.5,113.896 328.5,97.682Z" style="fill:rgb(253,253,253);stroke:rgb(51,51,51);stroke-width:3.95px;"/>
+                    </g>
+                    <g id="Dog-ear" serif:id="Dog ear" transform="matrix(1,0,0,1,-4,2)">
+                        <path d="M277.072,48.496L277.072,93.848C277.072,95.172 277.598,96.441 278.534,97.377C279.47,98.313 280.739,98.839 282.063,98.839C294.548,98.839 326.141,98.839 326.141,98.839" style="fill:rgb(245,245,245);stroke:rgb(51,51,51);stroke-width:4px;"/>
+                    </g>
+                </g>
+                <g transform="matrix(1,0,0,1,-29.6816,-0.395178)">
+                    <g transform="matrix(1.00243,0,0,1.11818,-144.72,-8.80181)">
+                        <path d="M465.73,119.654C465.73,117.605 463.874,115.941 461.588,115.941L310.259,115.941C307.973,115.941 306.117,117.605 306.117,119.654L306.117,183.108C306.117,185.157 307.973,186.821 310.259,186.821L461.588,186.821C463.874,186.821 465.73,185.157 465.73,183.108L465.73,119.654Z" style="fill:rgb(248,0,0);stroke:white;stroke-width:3.77px;"/>
+                    </g>
+                    <g transform="matrix(1.24571,0,0,1.35864,116.812,84.3924)">
+                        <g transform="matrix(64,0,0,64,42.1437,77.6203)">
+                            <path d="M0.084,0L0.084,-0.68L0.297,-0.68C0.371,-0.68 0.434,-0.663 0.487,-0.63C0.54,-0.596 0.566,-0.54 0.566,-0.462C0.566,-0.385 0.538,-0.328 0.481,-0.292C0.424,-0.255 0.36,-0.237 0.288,-0.237L0.213,-0.237L0.213,0L0.084,0ZM0.293,-0.572L0.213,-0.572L0.213,-0.344L0.295,-0.344C0.334,-0.344 0.365,-0.353 0.389,-0.371C0.413,-0.388 0.426,-0.416 0.429,-0.454C0.429,-0.498 0.417,-0.529 0.393,-0.546C0.369,-0.563 0.336,-0.572 0.293,-0.572Z" style="fill:white;fill-rule:nonzero;"/>
+                        </g>
+                        <g transform="matrix(64,0,0,64,79.7117,77.6203)">
+                            <path d="M0.332,0L0.084,0L0.084,-0.68L0.336,-0.68C0.441,-0.68 0.518,-0.648 0.569,-0.585C0.62,-0.522 0.645,-0.441 0.645,-0.344C0.645,-0.239 0.618,-0.155 0.563,-0.093C0.508,-0.031 0.431,0 0.332,0ZM0.337,-0.57L0.213,-0.57L0.213,-0.109L0.33,-0.109C0.385,-0.109 0.429,-0.127 0.462,-0.163C0.495,-0.199 0.511,-0.259 0.511,-0.344C0.511,-0.415 0.497,-0.47 0.469,-0.51C0.441,-0.55 0.397,-0.57 0.337,-0.57Z" style="fill:white;fill-rule:nonzero;"/>
+                        </g>
+                        <g transform="matrix(64,0,0,64,123.424,77.6203)">
+                            <path d="M0.405,-0.288L0.213,-0.288L0.213,0L0.084,0L0.084,-0.68L0.469,-0.68L0.489,-0.578L0.213,-0.578L0.213,-0.389L0.386,-0.389L0.405,-0.288Z" style="fill:white;fill-rule:nonzero;"/>
+                        </g>
+                    </g>
+                </g>
+            </g>
+            <g transform="matrix(1,0,0,1.52217,67.3796,10.7507)">
+                <rect x="23.501" y="81.3" width="162.305" height="61.77" style="fill:rgb(180,213,255);"/>
+            </g>
+            <g transform="matrix(0.967536,0,0,0.961535,5.90498,47.9703)">
+                <g transform="matrix(90.4804,0,0,90.4804,82.6698,167.705)">
+                    <path d="M0.057,-0.337C0.057,-0.442 0.084,-0.527 0.139,-0.594C0.194,-0.66 0.271,-0.694 0.37,-0.696C0.477,-0.696 0.556,-0.662 0.607,-0.593C0.658,-0.524 0.684,-0.441 0.684,-0.344C0.684,-0.239 0.657,-0.153 0.602,-0.086C0.547,-0.019 0.469,0.014 0.37,0.014C0.264,0.014 0.185,-0.02 0.134,-0.089C0.083,-0.157 0.057,-0.24 0.057,-0.337ZM0.192,-0.338C0.192,-0.267 0.206,-0.208 0.235,-0.163C0.264,-0.118 0.308,-0.095 0.369,-0.095C0.424,-0.095 0.467,-0.115 0.5,-0.156C0.533,-0.197 0.549,-0.259 0.549,-0.344C0.549,-0.415 0.535,-0.473 0.506,-0.518C0.477,-0.563 0.433,-0.586 0.372,-0.586C0.319,-0.586 0.275,-0.564 0.242,-0.519C0.209,-0.474 0.192,-0.414 0.192,-0.338Z" style="fill:rgb(51,51,51);fill-rule:nonzero;"/>
+                </g>
+                <g transform="matrix(90.4804,0,0,90.4804,147.906,167.705)">
+                    <path d="M0.505,-0.557C0.473,-0.567 0.448,-0.574 0.429,-0.579C0.41,-0.583 0.388,-0.585 0.361,-0.585C0.307,-0.585 0.265,-0.563 0.236,-0.519C0.207,-0.475 0.192,-0.415 0.192,-0.338C0.192,-0.272 0.204,-0.215 0.229,-0.167C0.254,-0.119 0.295,-0.095 0.353,-0.095C0.382,-0.095 0.409,-0.098 0.434,-0.104C0.459,-0.11 0.481,-0.117 0.502,-0.126L0.551,-0.03C0.525,-0.017 0.494,-0.006 0.457,0.002C0.42,0.01 0.388,0.014 0.36,0.014C0.254,0.014 0.177,-0.02 0.129,-0.088C0.081,-0.156 0.057,-0.239 0.057,-0.337C0.057,-0.442 0.084,-0.527 0.137,-0.594C0.19,-0.661 0.266,-0.694 0.365,-0.694C0.385,-0.694 0.413,-0.691 0.448,-0.684C0.483,-0.677 0.516,-0.666 0.545,-0.65L0.505,-0.557Z" style="fill:rgb(51,51,51);fill-rule:nonzero;"/>
+                </g>
+                <g transform="matrix(90.4804,0,0,90.4804,199.751,167.705)">
+                    <path d="M0.293,-0.572L0.213,-0.572L0.213,-0.364L0.295,-0.364C0.334,-0.364 0.366,-0.372 0.391,-0.388C0.416,-0.403 0.429,-0.429 0.429,-0.465C0.429,-0.503 0.417,-0.53 0.393,-0.547C0.369,-0.564 0.336,-0.572 0.293,-0.572ZM0.479,0L0.335,-0.26C0.328,-0.259 0.32,-0.259 0.312,-0.259C0.304,-0.258 0.296,-0.258 0.288,-0.258L0.213,-0.258L0.213,0L0.084,0L0.084,-0.68L0.297,-0.68C0.371,-0.68 0.434,-0.663 0.487,-0.629C0.54,-0.595 0.566,-0.542 0.566,-0.471C0.566,-0.429 0.555,-0.393 0.534,-0.363C0.512,-0.332 0.484,-0.309 0.45,-0.292L0.617,0L0.479,0Z" style="fill:rgb(51,51,51);fill-rule:nonzero;"/>
+                </g>
+            </g>
+            <g transform="matrix(0.916882,0,0,1,121.475,-32.6535)">
+                <g transform="matrix(86.953,0,0,86.953,152.996,241.878)">
+                    <path d="M0.479,-0.428C0.5,-0.451 0.527,-0.47 0.562,-0.484C0.596,-0.497 0.627,-0.504 0.654,-0.504C0.72,-0.504 0.767,-0.485 0.795,-0.446C0.822,-0.407 0.836,-0.36 0.836,-0.304L0.836,0L0.705,0L0.705,-0.298C0.705,-0.329 0.698,-0.352 0.683,-0.369C0.668,-0.385 0.647,-0.393 0.619,-0.393C0.6,-0.393 0.581,-0.388 0.56,-0.378C0.539,-0.368 0.521,-0.357 0.504,-0.344C0.505,-0.337 0.505,-0.331 0.506,-0.324C0.507,-0.317 0.507,-0.311 0.507,-0.304L0.507,0L0.376,0L0.376,-0.298C0.376,-0.329 0.369,-0.352 0.354,-0.369C0.339,-0.385 0.318,-0.393 0.291,-0.393C0.274,-0.393 0.258,-0.39 0.241,-0.383C0.224,-0.376 0.207,-0.367 0.192,-0.356L0.192,0L0.062,0L0.062,-0.485L0.13,-0.485L0.162,-0.441C0.184,-0.461 0.211,-0.476 0.242,-0.488C0.273,-0.499 0.3,-0.504 0.325,-0.504C0.363,-0.504 0.395,-0.497 0.42,-0.484C0.445,-0.47 0.465,-0.451 0.479,-0.428Z" style="fill:rgb(51,51,51);fill-rule:nonzero;"/>
+                </g>
+                <g transform="matrix(86.953,0,0,86.953,228.906,241.878)">
+                    <path d="M0.156,0.023L0.179,-0.034L0.006,-0.467L0.14,-0.485L0.252,-0.191L0.358,-0.485L0.495,-0.485L0.278,0.064C0.263,0.103 0.236,0.137 0.197,0.165C0.158,0.193 0.118,0.212 0.075,0.222L0.029,0.115C0.052,0.105 0.077,0.093 0.104,0.079C0.13,0.064 0.147,0.046 0.156,0.023Z" style="fill:rgb(51,51,51);fill-rule:nonzero;"/>
+                </g>
+            </g>
+            <g id="Selectors" transform="matrix(0.965977,0,0,0.807602,67.3796,67.3718)">
+                <g id="Right-selector" serif:id="Right selector">
+                    <g transform="matrix(1.03522,0,0,1.23823,2.07044,0)">
+                        <path d="M185.806,161.156L185.806,67.132" style="fill:none;stroke:rgb(76,159,255);stroke-width:4px;stroke-linecap:butt;"/>
+                    </g>
+                    <g transform="matrix(1.03522,0,0,1.23823,161.788,169.469)">
+                        <circle cx="31.523" cy="34.314" r="10.021" style="fill:rgb(76,159,255);stroke:rgb(76,159,255);stroke-width:4px;stroke-linecap:butt;"/>
+                    </g>
+                </g>
+                <g id="Left-selector" serif:id="Left selector">
+                    <g transform="matrix(1.03522,0,0,1.23823,-170.092,0)">
+                        <path d="M185.806,161.156L185.806,67.132" style="fill:none;stroke:rgb(76,159,255);stroke-width:4px;stroke-linecap:butt;"/>
+                    </g>
+                    <g transform="matrix(1.03522,0,0,1.23823,-10.3742,28.2274)">
+                        <circle cx="31.523" cy="34.314" r="10.021" style="fill:rgb(76,159,255);stroke:rgb(76,159,255);stroke-width:4px;stroke-linecap:butt;"/>
+                    </g>
+                </g>
+            </g>
+        </g>
+    </g>
+</svg>

docs/index.rst

@@ -1,15 +1,12 @@
-.. ocrmypdf documentation master file, created by
-   sphinx-quickstart on Sun Sep  4 14:29:43 2016.
-   You can adapt this file completely to your liking, but it should at least
-   contain the root `toctree` directive.
-
 OCRmyPDF documentation
 ======================
 
-OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to
-be searched.
+OCRmyPDF adds an optical charcter recognition (OCR) text layer to scanned PDF
+files, allowing them to be searched.
 
-PDF is the best format for storing and exchanging scanned documents.  Unfortunately, PDFs can be difficult to modify. OCRmyPDF makes it easy to apply image processing and OCR to existing PDFs.
+PDF is the best format for storing and exchanging scanned documents.
+Unfortunately, PDFs can be difficult to modify. OCRmyPDF makes it easy to apply
+image processing and OCR to existing PDFs.
 
 .. toctree::
    :maxdepth: 1
@@ -17,6 +14,7 @@ PDF is the best format for storing and exchanging scanned documents.  Unfortunat
    introduction
    release_notes
    installation
+   optimizer
    languages
    jbig2
 
@@ -27,12 +25,18 @@ PDF is the best format for storing and exchanging scanned documents.  Unfortunat
    cookbook
    docker
    advanced
-   api
-   plugins
    batch
-   security
+   performance
+   pdfsecurity
    errors
 
+.. toctree::
+   :caption: Developers
+   :maxdepth: 2
+
+   api
+   plugins
+   contributing
 
 Indices and tables
 ==================

docs/installation.rst

@@ -8,11 +8,20 @@ Installing OCRmyPDF
 |latest|
 
 The easiest way to install OCRmyPDF is to follow the steps for your operating
-system/platform, although sometimes this version may be out of date.
+system/platform, although sometimes this version may be out of date. This
+installation guide provides information allowing you to compare the current
+version to the one provided by your platform.
 
-If you want to use the latest version of OCRmyPDF, your best bet is to install
-the most recent version your platform provides, and then upgrade that version by
-installing the Python binary wheels.
+If you want to use the latest version of OCRmyPDF and all of its optional
+dependencies, the easiest way to get that is install the Homebrew package. Homebrew
+is best known as a macOS package manger, but also works for
+`Linux and Windows Subsystem for Linux <https://docs.brew.sh/Homebrew-on-Linux>`__.
+After Homebrew is installed, simply run ``brew install ocrmypdf``.
+
+You can also use the more detailed procedures here to manually install OCRmyPDF
+from source or with the ``pip`` package manager for binary wheels. The reason
+for these varied steps is that OCRmyPDF requires third-party executables that are
+not part of Python.
 
 .. contents:: Platform-specific steps
     :depth: 2
@@ -21,7 +30,7 @@ installing the Python binary wheels.
 Installing on Linux
 ===================
 
-Debian and Ubuntu 16.10 or newer
+Debian and Ubuntu 18.04 or newer
 --------------------------------
 
 .. |deb-stable| image:: https://repology.org/badge/version-for-repo/debian_stable/ocrmypdf.svg
@@ -33,28 +42,24 @@ Debian and Ubuntu 16.10 or newer
 .. |deb-unstable| image:: https://repology.org/badge/version-for-repo/debian_unstable/ocrmypdf.svg
     :alt: Debian unstable
 
-.. |ubu-1710| image:: https://repology.org/badge/version-for-repo/ubuntu_17_10/ocrmypdf.svg
-    :alt: Ubuntu 17.10
-
 .. |ubu-1804| image:: https://repology.org/badge/version-for-repo/ubuntu_18_04/ocrmypdf.svg
     :alt: Ubuntu 18.04 LTS
 
-.. |ubu-1810| image:: https://repology.org/badge/version-for-repo/ubuntu_18_10/ocrmypdf.svg
-    :alt: Ubuntu 18.10
+.. |ubu-2004| image:: https://repology.org/badge/version-for-repo/ubuntu_20_04/ocrmypdf.svg
+    :alt: Ubuntu 20.04 LTS
 
++-----------------------------------------------+
+| **OCRmyPDF versions in Debian & Ubuntu**      |
++-----------------------------------------------+
+| |latest|                                      |
++-----------------------------------------------+
+| |deb-stable| |deb-testing| |deb-unstable|     |
++-----------------------------------------------+
+| |ubu-1804| |ubu-2004|                         |
++-----------------------------------------------+
 
-+-------------------------------------------+
-| **OCRmyPDF versions in Debian & Ubuntu**  |
-+-------------------------------------------+
-| |latest|                                  |
-+-------------------------------------------+
-| |deb-stable| |deb-testing| |deb-unstable| |
-+-------------------------------------------+
-| |ubu-1710| |ubu-1804| |ubu-1810|          |
-+-------------------------------------------+
-
-Users of Debian 9 ("stretch") or later or Ubuntu 16.10 or later may
-simply
+Users of Debian 9 ("stretch") or later or Ubuntu 18.04 or later, including users
+of Windows Subsystem for Linux, may simply
 
 .. code-block:: bash
 
@@ -64,7 +69,8 @@ As indicated in the table above, Debian and Ubuntu releases may lag
 behind the latest version. If the version available for your platform is
 out of date, you could opt to install the latest version from source.
 See `Installing HEAD revision from
-sources <#installing-head-revision-from-sources>`__.
+sources <#installing-head-revision-from-sources>`__. Ubuntu 16.10 to 17.10
+inclusive also had ocrmypdf, but these versions are end of life.
 
 For full details on version availability for your platform, check the
 `Debian Package Tracker <https://tracker.debian.org/pkg/ocrmypdf>`__ or
@@ -81,22 +87,24 @@ For full details on version availability for your platform, check the
 Fedora 29 or newer
 ------------------
 
-.. |fedora-29| image:: https://repology.org/badge/version-for-repo/fedora29/ocrmypdf.svg
+.. |fedora-29| image:: https://repology.org/badge/version-for-repo/fedora_29/ocrmypdf.svg
     :alt: Fedora 29
 
+.. |fedora-30| image:: https://repology.org/badge/version-for-repo/fedora_30/ocrmypdf.svg
+    :alt: Fedora 30
+
 .. |fedora-rawhide| image:: https://repology.org/badge/version-for-repo/fedora_rawhide/ocrmypdf.svg
     :alt: Fedore Rawhide
 
++-----------------------------------------------+
+| **OCRmyPDF version**                          |
++-----------------------------------------------+
+| |latest|                                      |
++-----------------------------------------------+
+| |fedora-29| |fedora-30| |fedora-rawhide|      |
++-----------------------------------------------+
 
-+------------------------------+
-| **OCRmyPDF version**         |
-+------------------------------+
-| |latest|                     |
-+------------------------------+
-| |fedora-29| |fedora-rawhide| |
-+------------------------------+
-
-Users of Fedora 29 later may simply
+Users of Fedora 29 or later may simply
 
 .. code-block:: bash
 
@@ -117,27 +125,72 @@ from sources <#installing-head-revision-from-sources>`__.
    will automatically detect it on the ``PATH``. To add JBIG2 encoding,
    see `Installing the JBIG2 encoder <jbig2>`__.
 
-Installing the latest version on Ubuntu 18.04 LTS
+.. _ubuntu-lts-latest:
+
+Installing the latest version on Ubuntu 20.04 LTS
 -------------------------------------------------
 
-Ubuntu 18.04 includes ocrmypdf 6.1.2. To install a more recent version,
-first install the system version to get most of the dependencies:
+Ubuntu 20.04 includes ocrmypdf 9.6.0 - you can install that with ``apt``. To
+install a more recent version, uninstall the system-provided version of
+ocrmypdf, and install the following dependencies:
 
 .. code-block:: bash
 
-    sudo apt-get update
-    sudo apt-get install \
-        ocrmypdf \
-        python3-pip
-
-There are a few system dependency changes since ocrmypdf 6.1.2. Let's
-get these, too.
-
-.. code-block:: bash
-
-    sudo apt-get install \
+    sudo apt-get -y remove ocrmypdf  # remove system ocrmypdf, if installed
+    sudo apt-get -y update
+    sudo apt-get -y install \
+        ghostscript \
+        icc-profiles-free \
+        liblept5 \
         libxml2 \
-        pngquant
+        pngquant \
+        python3-pip \
+        tesseract-ocr \
+        zlib1g
+
+To install ocrmypdf for the system:
+
+.. code-block:: bash
+
+    sudo pip3 install ocrmypdf
+
+To install for the current user only:
+
+.. code-block:: bash
+
+    export PATH=$HOME/.local/bin:$PATH
+    pip3 install --user ocrmypdf
+
+Ubuntu 18.04 LTS
+-------------------------------------------------
+
+Ubuntu 18.04 includes ocrmypdf 6.1.2 - you can install that with ``apt``, but
+it is quite old now. To install a more recent version, uninstall the old version
+of ocrmypdf, and install the following dependencies:
+
+.. code-block:: bash
+
+    sudo apt-get -y remove ocrmypdf
+    sudo apt-get -y update
+    sudo apt-get -y install \
+        ghostscript \
+        icc-profiles-free \
+        liblept5 \
+        libxml2 \
+        pngquant \
+        python3-cffi \
+        python3-distutils \
+        python3-pkg-resources \
+        python3-reportlab \
+        qpdf \
+        tesseract-ocr \
+        zlib1g
+
+We will need a newer version of ``pip`` then was available for Ubuntu 18.04:
+
+.. code-block:: bash
+
+    wget https://bootstrap.pypa.io/get-pip.py && python3 get-pip.py
 
 Then install the most recent ocrmypdf for the local user and set the
 user's ``PATH`` to check for the user's Python packages.
@@ -145,7 +198,7 @@ user's ``PATH`` to check for the user's Python packages.
 .. code-block:: bash
 
     export PATH=$HOME/.local/bin:$PATH
-    pip3 install --user ocrmypdf
+    python3 -m pip install --user ocrmypdf
 
 To add JBIG2 encoding, see :ref:`jbig2`.
 
@@ -193,7 +246,8 @@ of ``pip`` at ``/usr/local/bin/pip``.
 **Install OCRmyPDF**
 
 OCRmyPDF requires the locale to be set for UTF-8. **On some minimal
-Ubuntu installations systems**, it may be necessary to set the locale.
+Ubuntu installations**, such as the Ubuntu 16.04 Docker images it may be
+necessary to set the locale.
 
 .. code-block:: bash
 
@@ -207,7 +261,7 @@ environment variable contains ``$HOME/.local/bin``.
 .. code-block:: bash
 
     export PATH=$HOME/.local/bin:$PATH
-    pip3 install --user ocrmypdf
+    pip3.6 install --user ocrmypdf
 
 To add JBIG2 encoding, see :ref:`jbig2`.
 
@@ -265,35 +319,93 @@ Now we need to install ``pip`` and let it install ocrmypdf:
     curl https://bootstrap.pypa.io/ez_setup.py -o - | python3.6 && python3.6 -m easy_install pip
     pip3.6 install ocrmypdf
 
-These installation instructions omit the optional dependency
-``unpaper``, which is only available at version 0.4.2 in Ubuntu 14.04.
-The author could not find a backport of ``unpaper``, and created a .deb
-package to do the job of installing unpaper 6.1 (for x86 64-bit only):
-
-.. code-block:: bash
-
-    wget -q 'https://www.dropbox.com/s/vaq0kbwi6e6au80/unpaper_6.1-1.deb?raw=1' -O unpaper_6.1-1.deb
-    sudo dpkg -i unpaper_6.1-1.deb
+The optional dependency ``unpaper`` is only available at 0.4.2 in Ubuntu 14.04,
+and no backports are available. Previously the author maintained a backported
+.deb package for unpaper 6.1, but since Ubuntu 14.04 is now end of life, this is
+not supported. As such, ``unpaper`` is not available on Ubuntu 14.04 or must by
+compiled by hand.
 
 To add JBIG2 encoding, see :ref:`jbig2`.
 
-ArchLinux (AUR)
----------------
+Arch Linux (AUR)
+----------------
 
 .. image:: https://repology.org/badge/version-for-repo/aur/ocrmypdf.svg
     :alt: ArchLinux
     :target: https://repology.org/metapackage/ocrmypdf
 
-There is an `ArchLinux User Repository package for
-ocrmypdf <https://aur.archlinux.org/packages/ocrmypdf/>`__. You can use
-the following command.
+There is an `Arch User Repository (AUR) package for OCRmyPDF
+<https://aur.archlinux.org/packages/ocrmypdf/>`__.
+
+Installing AUR packages as root is not allowed, so you must first `setup a
+non-root user
+<https://wiki.archlinux.org/index.php/Users_and_groups#User_management>`__ and
+`configure sudo <https://wiki.archlinux.org/index.php/Sudo#Configuration>`__.
+The standard Docker image, ``archlinux/base:latest``, does **not** have a
+non-root user configured, so users of that image must follow these guides. If
+you are using a VM image, such as `the official Vagrant image
+<https://app.vagrantup.com/archlinux/boxes/archlinux>`__, this work may already
+be completed for you.
+
+Next you should install the `base-devel package group
+<https://www.archlinux.org/groups/x86_64/base-devel/>`__. This includes the
+standard tooling needed to build packages, such as a compiler and binary tools.
 
 .. code-block:: bash
 
-    yaourt -S ocrmypdf
+   sudo pacman -S base-devel
 
-If you have any difficulties with installation, check the repository
-package page.
+Now you are ready to install the OCRmyPDF package.
+
+.. code-block:: bash
+
+   curl -O https://aur.archlinux.org/cgit/aur.git/snapshot/ocrmypdf.tar.gz
+   tar xvzf ocrmypdf.tar.gz
+   cd ocrmypdf
+   makepkg -sri
+
+At this point you will have a working install of OCRmyPDF, but the Tesseract
+install won’t include any OCR language data. You can install `the
+tesseract-data package group
+<https://www.archlinux.org/groups/any/tesseract-data/>`__ to add all supported
+languages, or use that package listing to identify the appropriate package for
+your desired language.
+
+.. code-block:: bash
+
+   sudo pacman -S tesseract-data-eng
+
+As an alternative to this manual procedure, consider using an `AUR helper
+<https://wiki.archlinux.org/index.php/AUR_helpers>`__. Such a tool will
+automatically fetch, build and install the AUR package, resolve dependencies
+(including dependencies on AUR packages), and ease the upgrade procedure.
+
+If you have any difficulties with installation, check the repository package
+page.
+
+.. note::
+
+    The OCRmyPDF AUR package currently omits the JBIG2 encoder. OCRmyPDF works
+    fine without it but will produce larger output files. The encoder is
+    available from `the jbig2enc-git AUR package
+    <https://aur.archlinux.org/packages/jbig2enc-git/>`__ and may be installed
+    using the same series of steps as for the installation OCRmyPDF AUR
+    package. Alternatively, it may be built manually from source following the
+    instructions in `Installing the JBIG2 encoder <jbig2>`__.  If JBIG2 is
+    installed, OCRmyPDF 7.0.0 and later will automatically detect it.
+
+Alpine Linux
+------------
+
+.. image:: https://repology.org/badge/version-for-repo/alpine_edge/ocrmypdf.svg
+    :alt: Alpine Linux
+    :target: https://repology.org/metapackage/ocrmypdf
+
+To install OCRmyPDF for Alpine Linux:
+
+.. code-block:: bash
+
+    apk add ocrmypdf
 
 Other Linux packages
 --------------------
@@ -345,7 +457,10 @@ languages you can optionally install them all:
 Manual installation on macOS
 ----------------------------
 
-These instructions probably work on all macOS supported by Homebrew.
+These instructions probably work on all macOS supported by Homebrew, and are
+for installing a more current version of OCRmyPDF than is available from
+Homebrew. Note that the Homebrew versions usually track the release versions
+fairly closely.
 
 If it's not already present, `install Homebrew <http://brew.sh/>`__.
 
@@ -356,14 +471,8 @@ Update Homebrew:
     brew update
 
 Install or upgrade the required Homebrew packages, if any are missing.
-To do this, download the ``Brewfile`` that lists all of the dependencies
-to the current directory, and run ``brew bundle`` to process them
-(installing or upgrading as needed). ``Brewfile`` is a plain text file.
-
-.. code-block:: bash
-
-    wget https://github.com/jbarlow83/OCRmyPDF/raw/master/.travis/Brewfile
-    brew bundle
+To do this, use ``brew edit ocrmypdf`` to obtain a recent list of Homebrew
+dependencies. You could also check the ``azure-pipelines.yml``.
 
 This will include the English, French, German and Spanish language
 packs. If you need other languages you can optionally install them all:
@@ -372,7 +481,7 @@ packs. If you need other languages you can optionally install them all:
 
    .. code-block:: bash
 
-    brew install tesseract --with-all-languages  # Option 2: for all language packs
+    brew install tesseract-lang  # Option 2: for all language packs
 
 Update the homebrew pip:
 
@@ -401,36 +510,145 @@ The command line program should now be available:
 Installing on FreeBSD
 =====================
 
-FreeBSD 11.2 is known to work. Other versions likely work but have not
-been tested.
+.. image:: https://repology.org/badge/version-for-repo/freebsd/python:ocrmypdf.svg
+    :alt: FreeBSD
+    :target: https://repology.org/project/python:ocrmypdf/versions
 
-In general it should work to:
+FreeBSD 11.3, 12.0, 12.1-RELEASE and 13.0-CURRENT are supported. Other
+versions likely work but have not been tested.
 
-#. `Install and build
-   pikepdf <https://pikepdf.readthedocs.io/en/latest/installation.html#installing-on-freebsd-11-2>`__.
-#. Install the equivalent list of dependencies for Linux.
+.. code-block:: bash
+
+    pkg install py37-ocrmypdf
+
+To install a more recent version, you could attempt to first install the system
+version with ``pkg``, then use ``pip install --user ocrmypdf``.
 
 Installing the Docker image
 ===========================
 
 For some users, installing the Docker image will be easier than
-installing all of OCRmyPDF's dependencies. For Windows, it is the only
-option.
+installing all of OCRmyPDF's dependencies.
 
 See `OCRmyPDF Docker Image <docker>`__ for more information.
 
 Installing on Windows
 =====================
 
-Direct installation on Windows is not possible. `Install the
-Docker <docker-install>`__ container as described above. Ensure that
+.. warning::
+
+    Native Windows support is new. Consider it "beta" software. Some
+    functionality is missing or may be more difficult to enable. If you need a
+    production-ready solution, use Windows Subsystem for Linux or a Docker
+    image.
+
+.. note::
+
+    Administrator privileges will be required for some of these steps.
+
+You must install the following for Windows:
+
+* Python 3.7 (64-bit)
+* Tesseract 4.0 or later
+* Ghostscript 9.50 or later
+
+You can install these with the Chocolatey package manager:
+
+* ``choco install python3``
+* ``choco install --pre tesseract``
+* ``choco install ghostscript``
+
+Also consider adding:
+
+* ``choco install pngquant``
+
+Windows 10 64-bit and 64-bit versions of applications are recommended. Earlier
+versions of Windows and 32-bit versions of these programs are not tested, and not
+supported at this time.
+
+OCRmyPDF will check for Tesseract-OCR and Ghostscript in your Program Files folder.
+If they are in some other location, you may need to modify the ``PATH``
+environment variable so Tesseract, Ghostscript, and other any optional executables can
+be found. You can enter it in the command line or
+`follow these directions <https://www.computerhope.com/issues/ch000549.htm#dospath>`_
+to make the change persistent and system-wide.
+
+You may then use pip to install ocrmypdf:
+
+* ``pip install ocrmypdf``
+
+Installing on Windows Subsystem for Linux
+=========================================
+
+#. Install Ubuntu 18.04 for Windows Subsystem for Linux, if not already installed.
+#. Follow the procedure to install :ref:`OCRmyPDF on Ubuntu 18.04 <ubuntu-lts-latest>`.
+#. Open the Windows command prompt and create a symlink:
+
+.. code-block:: powershell
+
+    wsl sudo ln -s  /home/$USER/.local/bin/ocrmypdf /usr/local/bin/ocrmypdf
+
+Then confirm that the expected version from PyPI (|latest|) is installed:
+
+.. code-block:: powershell
+
+    wsl ocrmypdf --version
+
+You can then run OCRmyPDF in the Windows command prompt or Powershell, prefixing
+``wsl``, and call it from Windows programs or batch files.
+
+Docker
+^^^^^^
+
+You can also :ref:`Install the Docker <docker-install>` container on Windows. Ensure that
 your command prompt can run the docker "hello world" container.
 
-It would probably not be too difficult to port on Windows. The main
-reason this has been avoided is the difficulty of packaging and
-installing the various non-Python dependencies: Tesseract, QPDF,
-Ghostscript, Leptonica. Pull requests to add or improve Windows support
-would be quite welcome.
+Installing on Cygwin64 under Windows
+====================================
+
+First install the the following prerequisite Cygwin packages using ``setup-x86_64.exe``::
+
+    python36 (or later)
+    python3?-devel
+    python3?-pip
+    python3?-lxml
+    python3?-imaging
+
+       (where 3? means match the version of python3 you installed)
+
+    gcc-g++
+    ghostscript (<=9.50 or >=9.52-2 see note below)
+    libexempi3
+    libexempi-devel
+    libffi6
+    libffi-devel
+    pngquant
+    qpdf
+    libqpdf-devel
+    tesseract-ocr
+    tesseract-ocr-devel
+
+.. note::
+
+    The Cygwin package for Ghostscript in versions 9.52 and
+    9.52-1 contained a bug that caused an exception to occur when
+    ocrmypdf invoked gs.  Make sure you have either 9.50 (or earlier)
+    or 9.52-2 (or later).
+
+Then open a Cygwin terminal (i.e. ``mintty``), run the following commands. Note
+that if you are using the version of ``pip`` that was installed with the Cygwin
+Python package, the command name will be ``pip3``.  If you have since updated
+``pip`` (with, for instance ``pip3 install --upgrade pip``) the the command is
+likely just ``pip`` instead of ``pip3``:
+
+.. code-block:: bash
+
+    pip3 install wheel
+    pip3 install ocrmypdf
+
+The optional dependency "unpaper" that is currently not available under Cygwin.
+Without it, certain options such as ``--clean`` will produce an error message.
+However, the OCR-to-text-layer functionality is available.
 
 Installing with Python pip
 ==========================
@@ -476,11 +694,11 @@ manager. ``pip`` cannot provide them.
 -  Python 3.6 or newer
 -  Ghostscript 9.15 or newer
 -  qpdf 8.1.0 or newer
--  Tesseract 4.0.0-alpha or newer
+-  Tesseract 4.0.0-beta or newer
 
 As of ocrmypdf 7.2.1, the following versions are recommended:
 
--  Python 3.7
+-  Python 3.7 or 3.8
 -  Ghostscript 9.23 or newer
 -  qpdf 8.2.1
 -  Tesseract 4.0.0 or newer

docs/introduction.rst

@@ -2,7 +2,7 @@
 Introduction
 ============
 
-OCRmyPDF is a Python 3 package that adds OCR layers to PDFs.
+OCRmyPDF is a Python 3 application and library that adds OCR layers to PDFs.
 
 About OCR
 =========
@@ -10,7 +10,7 @@ About OCR
 `Optical character
 recognition <https://en.wikipedia.org/wiki/Optical_character_recognition>`__
 is technology that converts images of typed or handwritten text, such as
-in a scanned document, to computer text that can be searched and copied.
+in a scanned document, to computer text that can be selected, searched and copied.
 
 OCRmyPDF uses
 `Tesseract <https://github.com/tesseract-ocr/tesseract>`__, the best
@@ -161,7 +161,7 @@ OCRmyPDF is also limited by the PDF specification:
    the spaces between words must be derived heuristically. Some PDF
    viewers do a better job of this than others.
 -  Because some popular open source PDF viewers have a particularly hard
-   time with spaces betweem words, OCRmyPDF appends a space to each text
+   time with spaces between words, OCRmyPDF appends a space to each text
    element as a workaround (when using ``--pdf-renderer hocr``). While
    this mixes document structure with graphical information that ideally
    should be left to the PDF viewer to interpret, it improves
@@ -208,7 +208,7 @@ consider one of these similar open source programs:
 Web front-ends
 ==============
 
-The Docker image ``ocrmypdf-alpine`` provides a web service front-end
+The Docker image ``ocrmypdf`` provides a web service front-end
 that allows files to submitted over HTTP and the results "downloaded".
 This is an HTTP server intended to simplify web services deployments; it
 is not intended to be deployed on the public internet and no real

docs/languages.rst

@@ -4,11 +4,20 @@
 Installing additional language packs
 ====================================
 
-OCRmyPDF uses Tesseract for OCR, and relies on its language packs for
-languages other than English.
+OCRmyPDF uses Tesseract for OCR, and relies on its language packs for all languages.
+On most platforms, English is installed with Tesseract by default, but not always.
 
 Tesseract supports `most
 languages <https://github.com/tesseract-ocr/tesseract/blob/master/doc/tesseract.1.asc#languages>`__.
+Languages are identified by standardized three-letter codes (called ISO 639-2 Alpha-3).
+Tesseract's documentation also lists the three-letter code for your language.
+Some are anglicized, e.g. Spanish is ``spa`` rather than ``esp``, while others
+are not, e.g. German is ``deu``.
+
+After you have installed a language pack, you can use it ``ocrmypdf -l <language>``,
+for example ``ocrmypdf -l spa``. For multilingual documents, you can specify
+all languages to be expected, e.g. ``ocrmypdf -l eng+fra`` for English and French.
+English is assumed by default unless other language(s) are specified.
 
 For Linux users, you can often find packages that provide language
 packs:
@@ -57,3 +66,8 @@ Docker users
 Users of the OCRmyPDF Docker image should install language packs into a
 derived Docker image as
 :ref:`described in that section <docker-lang-packs>`.
+
+Windows users
+=============
+
+The Tesseract installer provided by Chocolatey already includes 100 languages.

docs/optimizer.rst

@@ -0,0 +1,75 @@
+================
+PDF optimization
+================
+
+OCRmyPDF includes an image-oriented PDF optimizer. By default, the optimizer
+runs with safe settings with the goal of improving compression at no loss of
+quality. At higher optimization levels, lossy optimizations may be applied and
+tuned. Optimization occurs after OCR, and only if OCR succeeded.  It does not
+perform other possible optimizations such as deduplicating resources,
+consolidating fonts, simplifying vector drawings, or anything of that nature.
+
+Optimization ranges from ``-O0`` through ``-O3``, where ``0`` disables
+optimization and ``3`` implements all options. ``1``, the default, performs only
+safe and lossless optimizations. (This is similar to GCC's optimization
+parameter.) The exact type of optimizations performed will vary over time.
+
+PDF optimization requires third-party, optional tools for certain optimizations.
+If these are not installed or cannot be found by OCRmyPDF, optimization will not
+be as good.
+
+Optimizations that always occurs
+================================
+
+OCRmyPDF will automatically replace obsolete or inferior compression schemes
+such as RLE or LZW with superior schemes such as Deflate and converting
+monochrome images to CCITT G4. Since this is harmless it always occurs and there
+is no way to disable it. Other non-image compressed objects are compressed as
+well.
+
+Fast web view
+=============
+
+OCRmyPDF automatically optimizes PDFs for "fast web view" in Adobe Acrobat's
+parlance, or equivalently, linearizes PDFs so that the resources they reference
+are presented in the order a viewer needs them for sequential display. This
+reduces the latency of viewing a PDF both online and from local storage. This
+actually slightly increases the file size.
+
+To disable this optimization and all others, use ``ocrmypdf --optimize 0 ...``
+or the shorthand ``-O0``.
+
+Lossless optimizations
+======================
+
+At optimization level ``-O1`` (the default), OCRmyPDF will also attempt lossless
+image optimization.
+
+If a JBIG2 encoder is available, then monochrome images will be converted to
+JBIG2, with the potential for huge savings on large black and white images,
+since JBIG2 is far more efficient than any other monochrome (bi-level)
+compression. (All known US patents related to JBIG2 have probably expired, but
+it remains the responsibility of the user to supply a JBIG2 encoder such as
+`jbig2enc <https://github.com/agl/jbig2enc>`__. OCRmyPDF does not implement
+JBIG2 encoding on its own.)
+
+OCRmyPDF currently does not attempt to recompress losslessly compressed objects
+more aggressively.
+
+Lossy optimizations
+===================
+
+At optimization level ``-O2`` and ``-O3``, OCRmyPDF will some attempt lossy
+image optimization.
+
+If ``pngquant`` is installed, OCRmyPDF will use it to perform quantize paletted
+images to reduce their size.
+
+The quality of JPEGs may be lowered, on the assumption that a lower quality
+image may be suitable for storage after OCR.
+
+It is not possible to optimize all image types. Uncommon image types may be
+skipped by the optimizer.
+
+OCRmyPDF provides :ref:`lossy mode JBIG2 <jbig2-lossy>` as an advanced feature
+that additional requires the argument ``--jbig2-lossy``.

docs/pdfsecurity.rst

@@ -68,7 +68,7 @@ license, OCRmyPDF's GPL license, and any other licenses.
 
 Setting aside these concerns, a side effect of OCRmyPDF is it may
 incidentally sanitize PDFs that contain certain types of malware. It
-runs ``qpdf`` to repair the PDF, which could correct malformed PDF
+repairs the PDF with pikepdf/libqpdf, which could correct malformed PDF
 structures that are part of an attack. When PDF/A output is selected
 (the default), the input PDF is partially reconstructed by Ghostscript.
 When ``--force-ocr`` is used, all pages are rasterized and reconverted
@@ -144,10 +144,9 @@ set, the document cannot be viewed without the password.
 Either way, OCRmyPDF does not remove passwords from PDFs and exits with
 an error on encountering them.
 
-``qpdf``, one of OCRmyPDF's dependencies, can remove passwords. If the
-owner and user password are set, a password is required for ``qpdf``. If
-only the owner password is set, then the password can be stripped, even
-if one does not have the owner password.
+``qpdf`` can remove passwords. If the owner and user password are set, a
+password is required for ``qpdf``. If only the owner password is set, then the
+password can be stripped, even if one does not have the owner password.
 
 After OCR is applied, password protection is not permitted on PDF/A
 documents but the file can be converted to regular PDF.

docs/performance.rst

@@ -0,0 +1,22 @@
+===========
+Performance
+===========
+
+Some users have noticed that current versions of OCRmyPDF do not run as quickly
+as some older versions (specifically 6.x and older). This is because OCRmyPDF
+added image optimization as a postprocessing step, and it is enabled by default.
+
+Speed
+=====
+
+If running OCRmyPDF quickly is your main goal, you can use settings such as:
+
+* ``--optimize 0`` to disable file size optimization
+* ``--output-type pdf`` to disable PDF/A generation
+* ``--fast-web-view 0`` to disable fast web view optimization
+* ``--skip-big`` to skip large images, if some pages have large images
+
+You can also avoid:
+
+* ``--force-ocr``
+* Image preprocessing

docs/plugins.rst

@@ -2,23 +2,79 @@
 Plugins
 =======
 
-You can use plugins to customize the behavior of OCRmyPDF at certain
-points of interest.
+You can use plugins to customize the behavior of OCRmyPDF at certain points of
+interest.
 
-Currently, it is possible to: - override the decision for whether or not
-to perform OCR on a particular file - modify the image is about to be
-sent for OCR
+Currently, it is possible to:
+
+- add new command line arguments
+- override the decision for whether or not to perform OCR on a particular file
+- modify the image is about to be sent for OCR
+- modify the page image before it is converted to PDF
+
+OCRmyPDF plugins are based on the Python ``pluggy`` package and conform to its
+conventions. Note that: plugins installed with as setuptools entrypoints are
+not checked currently, because OCRmyPDF assumes you may not want to enable
+plugins for all files. Also, plugins must be functions, not classes.
 
 How plugins are imported
 ========================
 
-Plugins are imported on demand, by the OCRmyPDF worker process that
-needs to use them. As such, plugins cannot share state with each other,
-and will be imported many times, once for each worker process.
+Plugins are imported on demand, by the OCRmyPDF worker process that needs to use
+them. As such, plugins cannot share state with other plugins, cannot rely on
+their module's or the interpreter's global state, and should expect asynchronous
+copies of themselves to be running. Plugins can write intermediate files to the
+folder specified in ``options.work_folder``.
 
-Plugins currently cannot override the same hook.
+Plugins should work whether executed in threads or processes.
 
-How plugins are invoked
-=======================
+Script plugins
+==============
 
-Plugins may be called from the command line:
+Script plugins may be called from the command line, by specifying the name of a file.
+
+.. code-block:: bash
+
+    ocrmypdf --plugin example_plugin.py input.pdf output.pdf
+
+Multiple plugins may be called by issuing the ``--plugin`` argument multiple times.
+
+Packaged plugins
+================
+
+Installed plugins may be installed into the same virtual environment as OCRmyPDF
+is installed into. They may be invoked using Python standard module naming.
+
+.. code-block:: bash
+
+    ocrmypdf --plugin ocrmypdf_fancypants.pockets.contents input.pdf output.pdf
+
+OCRmyPDF does not automatically import plugins, because the assumption is that
+plugins affect different files differently and you may not want them activated
+all the time. The command line or ``ocrmypdf.ocr(plugin='...')`` must call
+for them.
+
+Third parties that wish to distribute packages for ocrmypdf should package them
+as packaged plugins, and these modules should begin with the name ``ocrmypdf_``
+similar to ``pytest`` packages such as ``pytest-cov`` (the package) and
+``pytest_cov`` (the module).
+
+Plugin hooks
+============
+
+A plugin may provide the following hooks. Hooks should be decorated with
+``ocrmypdf.hookimpl``, for example:
+
+.. code-block:: python
+
+    from ocrmpydf import hookimpl
+
+    @hookimpl
+    def add_options(parser):
+        pass
+
+The following is a complete list of hooks that may be installed and when
+they are called.
+
+.. automodule:: ocrmypdf.pluginspec
+    :members:

docs/release_notes.rst

@@ -5,43 +5,355 @@ Release notes
 OCRmyPDF uses `semantic versioning <http://semver.org/>`__ for its
 command line interface and its public API.
 
-The ``ocrmypdf`` package may now be imported. The public API may be
-useful in scripts that launch OCRmyPDF processes or that wish to use
-some of its features for working with PDFs.
+OCRmyPDF's output messages are not considered part of the stable interface -
+that is, output messages may be improved at any release level, so parsing them
+may be unreliable. Use the API to depend on precise behavior.
+
+The public API may be useful in scripts that launch OCRmyPDF processes or that
+wish to use some of its features for working with PDFs.
 
 Note that it is licensed under GPLv3, so scripts that
 ``import ocrmypdf`` and are released publicly should probably also be
 licensed under GPLv3.
 
-next
-====
+v10.1.0
+=======
+
+-  Previously, we ``--clean-final`` would cause an unpaper-cleaned page image to
+   be produced twice, which was necessary in some cases but not in general. We
+   now take this optimization opportunity and reuse the image if possible.
+-  We now provide PNG files as input to unpaper, since it accepts them, instead
+   of generating PPM files which can be very large. This can improve performance
+   and temporary disk usage.
+-  Documentation updated for plugins.
+
+v10.0.1
+=======
+
+-  Fix regression when ``-l lang1+lang2`` is used from command line.
+
+v10.0.0
+=======
 
 **Breaking changes**
 
--   The ``--mask-barcodes`` experimental feature has been dropped due to poor
-    reliability and occasional crashes, both due to the underlying library that
-    implements this feature (Leptonica).
+-  Support for pdfminer.six version 20181108 has been dropped, along with a
+   monkeypatch that made this version work.
+-  Output messages are now displayed in color (when supported by the terminal)
+   and prefixes describing the severity of the message are removed. As such
+   programs that parse OCRmyPDF's log message will need to be revised. (Please
+   consider using OCRmyPDF as a library instead.)
+-  The minimum version for certain dependencies has increased.
+-  Many API changes; see developer changes.
+-  The Python libraries pluggy and coloredlogs are now required.
 
-**Major changes**
+**New features and improvements**
 
--   Added a high level API for applications that want to integrate OCRmyPDF.
-    Special thanks to Martin Wind (@mawi1988) whose made significant contributions
-    to this effort.
--   Added a simple plugin interface that makes certain steps of the pipeline
-    configurable.
--   Added progress bars for long-running steps. As such, the behavior of output
-    messages is different.
--   Dropped the ``ocrmypdf-polyglot`` and ``ocrmypdf-webservice`` images.
--   Removed dependency on ``ruffus``, and with that, the non-reentrancy
-    restrictions that previous made an API impossible.
--   Internal code reorganization.
+-  PDF page scanning is now parallelized across CPUs, speeding up this phase
+   dramatically for files with a high page counts.
+-  PDF page scanning is optimized, addressing some performance regressions.
+-  PDF page scanning is no longer run on pages that are not selected when the
+   ``--pages`` argument is used.
+-  PDF page scanning is now independent of Ghostscript, ending our past reliance
+   on this occasionally unstable feature in Ghostscript.
+-  A plugin architecture has been added, currently allowing one to more easily
+   use a different OCR engine or PDF renderer from Tesseract and Ghostscript,
+   respectively. A plugin can also override some decisions, such changing
+   the OCR settings after initial scanning.
+-  Colored log messages.
 
-**Minor changes**
+**Developer changes**
 
--   Test suite now spawns processes less frequently, allowing more accurate
-    measurement of code coverage.
--   Updated Docker images to use newer versions.
+-  The test spoofing mechanism, used to test correct handling of failures in
+   Tesseract and Ghostscript, has been removed in favor of using plugins for
+   testing. The spoofing mechanism was fairly complex and required many special
+   hacks for Windows.
+-  Code describing the resolution in DPI of images was refactored into a
+   ``ocrmypdf.helpers.Resolution`` class.
+-  The module ``ocrmypdf._exec`` is now private to OCRmyPDF.
+-  The ``ocrmypdf.hocrtransform`` module has been updated to follow PEP8 naming
+   conventions.
+-  Ghostscript is no longer used for finding the location of text in PDFs, and
+   APIs related to this feature have been removed.
+-  Lots of internal reorganization to support plugins.
 
+v9.8.2
+======
+
+-  Fixed an issue where OCRmyPDF would ignore text inside Form XObject when
+   making certain decisions about whether a document already had text.
+-  Fixed file size increase warning to take overhead of small files into account.
+-  Added instructions for installing on Cygwin.
+
+v9.8.1
+======
+
+-  Fixed an issue where unexpected files in the ``%PROGRAMFILES%\gs`` directory
+   (Windows) caused an exception.
+-  Mark pdfminer.six 20200517 as supported.
+-  If jbig2enc is missing and optimization is requested, a warning is issued
+   instead of an error, which was the intended behavior.
+-  Documentation updates.
+
+v9.8.0
+======
+
+-  Fixed issue where only the first PNG (FlateDecode) image in a file would be
+   considered for optimization. File sizes should be improved from here on.
+-  Fixed a startup crash when the chosen language was Japanese (#543).
+-  Added options to configure polling and log level to watcher.py.
+
+v9.7.2
+======
+
+-  Fixed an issue with ``ocrmypdf.ocr(...language=)`` not accepting a list of
+   languages as documented.
+-  Updated setup.py to confirm that pdfminer.six version 20200402 is supported.
+
+v9.7.1
+======
+
+-  Fixed version check failing when used with qpdf 10.0.0.
+-  Added some missing type annotations.
+-  Updated documentation to warn about need for "ifmain" guard and Windows.
+
+v9.7.0
+======
+
+-  Fixed an error in watcher.py if ``OCR_JSON_SETTINGS`` was not defined.
+-  Ghostscript 9.51 is now blacklisted, due to numerous problems with this version.
+-  Added a workaround for a problem with "txtwrite" in Ghostscript 9.52.
+-  Fixed an issue where the incorrect number of threads used was shown when
+   ``OMP_THREAD_LIMIT`` was manipulated.
+-  Removed a possible performance bottlenecks for files that use hundreds to
+   thousands of images on the same page.
+-  Documentation improvements.
+-  Optimization will now be applied to some monochrome images that have a color
+   profile defined instead of only black and white.
+-  ICC profiles are consulted when determining the simplified colorspace of an
+   image.
+
+v9.6.1
+======
+
+-  Documentation improvements - thanks to many users for their contributions!
+
+      - Fixed installation instructions for ArchLinux (@pigmonkey)
+      - Updated installation instructions for FreeBSD and other OSes (@knobix)
+      - Added instructions for using Docker Compose with watchdog (@ianalexander,
+        @deisi)
+      - Other miscellany (@mb720, @toy, @caiofacchinato)
+      - Some scripts provided in the documentation have been migrated out so that
+        they can be copied out as whole files, and to ensure syntax checking
+        is maintained.
+
+-  Fixed an error that caused bash completions to fail on macOS. (#502, #504;
+   @AlexanderWillner)
+-  Fixed a rare case where OCRmyPDF threw an exception while processing a PDF
+   with the wrong object type in its ``/Trailer /Info``. The error is now logged
+   and incorrect object is ignored. (#497)
+-  Removed potentially non-free file ``enron1.pdf`` and simplified the test that
+   used it.
+-  Removed potentially non-free file ``misc/media/logo.afdesign``.
+
+v9.6.0
+======
+
+-  Fixed a regression with transferring metadata from the input PDF to the output
+   PDF in certain situations.
+-  pdfminer.six is now supported up to version 2020-01-24.
+-  Messages are explaining page rotation decisions are now shown at the standard
+   verbosity level again when ``--rotate-pages``. In some previous version they
+   were set to debug level messages that only appeared with the parameter ``-v1``.
+-  Improvements to ``misc/watcher.py``. Thanks to @ianalexander and @svenihoney.
+-  Documentation improvements.
+
+v9.5.0
+======
+
+-  Added API functions to measure OCR quality.
+-  Modest improvements to handling PDFs with difficult/non compliant metadata.
+
+v9.4.0
+======
+
+-  Updated recommended dependency versions.
+-  Improvements to test coverage and changes to facilitate better measurement of
+   test coverage, such as when tests run in subprocesses.
+-  Improvements to error messages when Leptonica is not installed correctly.
+-  Fixed use of pytest "session scope" that may have caused some intermittent
+   CI failures.
+-  When the argument ``--keep-temporary-files`` or verbosity is set to ``-v1``,
+   a debug log file is generated in the working temporary folder.
+
+v9.3.0
+======
+
+-  Improved native Windows support: we now check in the obvious places in
+   the "Program Files" folders installations of Tesseract and Ghostscript,
+   rather than relying on the user to edit ``PATH`` to specify their location.
+   The ``PATH`` environment variable can still be used to differentiate when
+   multiple installations are present or the programs are installed to non-
+   standard locations.
+-  Fixed an exception on parsing Ghostscript error messages.
+-  Added an improved example demonstrating how to set up a watched folder
+   for automated OCR processing (thanks to @ianalexander for the contribution).
+
+v9.2.0
+======
+
+-  Native Windows is now supported.
+-  Continuous integration moved to Azure Pipelines.
+-  Improved test coverage and speed of tests.
+-  Fixed an issue where a page that was originally a JPEG would be saved as a
+   PNG, increasing file size. This occurred only when a preprocessing option
+   was selected along with ``--output-type=pdf`` and all images on the original
+   page were JPEGs. Regression since v7.0.0.
+-  OCRmyPDF no longer depends on the QPDF executable ``qpdf`` or ``libqpdf``.
+   It uses pikepdf (which in turn depends on ``libqpdf``). Package maintainers
+   should adjust dependencies so that OCRmyPDF no longer calls for libqpdf on
+   its own. For users of Python binary wheels, this change means a separate
+   installation of QPDF is no longer necessary. This change is mainly to
+   simplify installation on Windows.
+-  Fixed a rare case where log messages from Tesseract would be discarded.
+-  Fixed incorrect function signature for pixFindPageForeground, causing
+   exceptions on certain platforms/Leptonica versions.
+
+v9.1.1
+======
+
+-  Expand the range of pdfminer.six versions that are supported.
+-  Fixed Docker build when using pikepdf 1.7.0.
+-  Fixed documentation to recommend using pip from get-pip.py.
+
+v9.1.0
+======
+
+-  Improved diagnostics when file size increases at output. Now warns if JBIG2
+   or pngquant were not available.
+-  pikepdf 1.7.0 is now required, to pick up changes that remove the need for
+   a source install on Linux systems running Python 3.8.
+
+v9.0.5
+======
+
+-  The Alpine Docker image (jbarlow83/ocrmypdf-alpine) has been dropped due to
+   the difficulties of supporting Alpine Linux.
+-  The primary Docker image (jbarlow83/ocrmypdf) has been improved to take on
+   the extra features that used to be exclusive to the Alpine image.
+-  No changes to application code.
+-  pdfminer.six version 20191020 is now supported.
+
+v9.0.4
+======
+
+-  Fixed compatibility with Python 3.8 (but requires source install for the moment).
+-  Fixed Tesseract settings for ``--user-words`` and ``--user-patterns``.
+-  Changed to pikepdf 1.6.5 (for Python 3.8).
+-  Changed to Pillow 6.2.0 (to mitigate a security vulnerability in earlier Pillow).
+-  A debug message now mentions when English is automatically selected if the locale
+   is not English.
+
+v9.0.3
+======
+
+-  Embed an encoded version of the sRGB ICC profile in the intermediate
+   Postscript file (used for PDF/A conversion). Previously we included the
+   filename, which required Postscript to run with file access enabled. For
+   security, Ghostscript 9.28 enables ``-dSAFER`` and as such, no longer
+   permits access to any file by default. This fix is necessary for
+   compatibility with Ghostscript 9.28.
+-  Exclude a test that sometimes times out and fails in continuous integration
+   from the standard test suite.
+
+v9.0.2
+======
+
+-  The image optimizer now skips optimizing flate (PNG) encoded images in some
+   situations where the optimization effort was likely wasted.
+-  The image optimizer now ignores images that specify arbitrary decode arrays,
+   since these are rare.
+-  Fixed an issue that caused inversion of black and white in monochrome images.
+   We are not certain but the problem seems to be linked to Leptonica 1.76.0 and
+   older.
+-  Fixed some cases where the test suite failed if
+   English or German Tesseract language packs were not installed.
+-  Fixed a runtime error if the Tesseract English language is not installed.
+-  Improved explicit closing of Pillow images after use.
+-  Actually fixed of Alpine Docker image build.
+-  Changed to pikepdf 1.6.3.
+
+v9.0.1
+======
+
+-  Fixed test suite failing when either of optional dependencies unpaper and
+   pngquant were missing.
+-  Attempted fix of Alpine Docker image build.
+-  Documented that FreeBSD ports are now available.
+-  Changed to pikepdf 1.6.1.
+
+v9.0.0
+======
+
+**Breaking changes**
+
+-  The ``--mask-barcodes`` experimental feature has been dropped due to poor
+   reliability and occasional crashes, both due to the underlying library that
+   implements this feature (Leptonica).
+-  The ``-v`` (verbosity level) parameter now accepts only ``0``, ``1``, and
+   ``2``.
+-  Dropped support for Tesseract 4.00.00-alpha releases. Tesseract 4.0 beta and
+   later remain supported.
+-  Dropped the ``ocrmypdf-polyglot`` and ``ocrmypdf-webservice`` images.
+
+**New features**
+
+-  Added a high level API for applications that want to integrate OCRmyPDF.
+   Special thanks to Martin Wind (@mawi1988) whose made significant contributions
+   to this effort. OCRmyPDF is GPLv3-licensed.
+-  Added progress bars for long-running steps. ■■■■■■■□□
+-  We now create linearized ("fast web view") PDFs by default. The new parameter
+   ``--fast-web-view`` provides control over when this feature is applied.
+-  Added a new ``--pages`` feature to limit OCR to only a specific page range.
+   The list may contain commas or single pages, such as ``1, 3, 5-11``.
+-  When the number of pages is small compared to the number of allowed jobs, we
+   run Tesseract in multithreaded (OpenMP) mode when available. This should
+   improve performance on files with low page counts.
+-  Removed dependency on ``ruffus``, and with that, the non-reentrancy
+   restrictions that previous made an API impossible.
+-  Output and logging messages overhauled so that ocrmypdf may be integrated
+   into applications that use the logging module.
+-  pikepdf 1.6.0 is required.
+-  Added a logo. 😊
+
+**Bug fixes**
+
+-  Pages with vector artwork are treated as full color. Previously, vectors
+   were ignored when considering the colorspace needed to cover a page, which
+   could cause loss of color under certain settings.
+-  Test suite now spawns processes less frequently, allowing more accurate
+   measurement of code coverage.
+-  Improved test coverage.
+-  Fixed a rare division by zero (if optimization produced an invalid file).
+-  Updated Docker images to use newer versions.
+-  Fixed images encoded as JBIG2 with a colorspace other than ``/DeviceGray``
+   were not interpreted correctly.
+-  Fixed a OCR text-image registration (i.e. alignment) problem when the page
+   when MediaBox had a nonzero corner.
+
+v8.3.2
+======
+
+-  Dropped workaround for macOS that allowed it work without pdfminer.six,
+   now a proper sdist release of pdfminer.six is available.
+
+-  pikepdf 1.5.0 is now required.
+
+v8.3.1
+======
+
+-  Fixed an issue where PDFs with malformed metadata would be rendered as
+   blank pages. `#398 <https://github.com/jbarlow83/OCRmyPDF/issues/398>`_.
 
 v8.3.0
 ======

misc/batch.py

@@ -0,0 +1,50 @@
+#!/usr/bin/env python3
+# Original version by DeliciousPickle@github; modified
+
+# This script must be edited to meet your needs.
+
+import logging
+import os
+import sys
+
+import ocrmypdf
+
+# pylint: disable=logging-format-interpolation
+# pylint: disable=logging-not-lazy
+
+script_dir = os.path.dirname(os.path.realpath(__file__))
+print(script_dir + '/batch.py: Start')
+
+if len(sys.argv) > 1:
+    start_dir = sys.argv[1]
+else:
+    start_dir = '.'
+
+if len(sys.argv) > 2:
+    log_file = sys.argv[2]
+else:
+    log_file = script_dir + '/ocr-tree.log'
+
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s %(message)s',
+    filename=log_file,
+    filemode='w',
+)
+
+ocrmypdf.configure_logging(ocrmypdf.Verbosity.default)
+
+for dir_name, subdirs, file_list in os.walk(start_dir):
+    logging.info(dir_name + '\n')
+    os.chdir(dir_name)
+    for filename in file_list:
+        file_ext = os.path.splitext(filename)[1]
+        if file_ext == '.pdf':
+            full_path = dir_name + '/' + filename
+            print(full_path)
+            result = ocrmypdf.ocr(filename, filename, deskew=True)
+            if result == ocrmypdf.ExitCode.already_done_ocr:
+                print("Skipped document because it already contained text")
+            elif result == ocrmypdf.ExitCode.ok:
+                print("OCR complete")
+            logging.info(result)

misc/completion/ocrmypdf.bash

@@ -1,9 +1,37 @@
 # ocrmypdf completion                                     -*- shell-script -*-
 
+set -o errexit
+
 _ocrmypdf()
 {
     local cur prev cword words split
-    _init_completion -s || return
+
+    # Homebrew on Macs have version 1.3 of bash-completion which doesn't include - see #502
+    if declare -F _init_completions >/dev/null 2>&1; then
+        _init_completion -s || return
+    else
+        COMPREPLY=()
+        _get_comp_words_by_ref cur prev words cword
+    fi
+
+    if [[ $cur == -* ]]; then
+        COMPREPLY=( $( compgen -W '--language --image-dpi --output-type
+            --sidecar --version --jobs --quiet --verbose --title --author
+            --subject --keywords --rotate-pages --remove-background --deskew
+            --clean --clean-final --unpaper-args --oversample --remove-vectors
+            --threshold --force-ocr --skip-text --redo-ocr
+            --skip-big --jpeg-quality --png-quality --jbig2-lossy
+            --max-image-mpixels --tesseract-config --tesseract-pagesegmode
+            --help --tesseract-oem --pdf-renderer --tesseract-timeout
+            --rotate-pages-threshold --pdfa-image-compression --user-words
+            --user-patterns --keep-temporary-files --output-type
+            --no-progress-bar --pages --fast-web-view' \
+            --  "$cur" ) )
+        return
+    else
+        _filedir
+        return
+    fi
 
     case $prev in
         --version|-h|--help)
@@ -49,39 +77,23 @@ _ocrmypdf()
             return
             ;;
         -v|--verbose)
-            COMPREPLY=( $( compgen -W '{1..9}' -- "$cur" ) ) # max level ?
+            COMPREPLY=( $( compgen -W '{0..2}' -- "$cur" ) ) # max level ?
             return
             ;;
         --tesseract-pagesegmode)
             COMPREPLY=( $( compgen -W '{1..13}' -- "$cur" ) )
             return
             ;;
-        --sidecar|--title|--author|--subject|--keywords|--unpaper-args)
+        --sidecar|--title|--author|--subject|--keywords|--unpaper-args|--pages|--fast-web-view)
             # argument required but no completions available
             return
             ;;
     esac
 
     $split && return
-
-    if [[ $cur == -* ]]; then
-        COMPREPLY=( $( compgen -W '--language --image-dpi --output-type
-            --sidecar --version --jobs --quiet --verbose --title --author
-            --subject --keywords --rotate-pages --remove-background --deskew
-            --clean --clean-final --unpaper-args --oversample --remove-vectors
-            --mask-barcodes --threshold --force-ocr --skip-text --redo-ocr
-            --skip-big --jpeg-quality --png-quality --jbig2-lossy
-            --max-image-mpixels --tesseract-config --tesseract-pagesegmode
-            --help --tesseract-oem --pdf-renderer --tesseract-timeout
-            --rotate-pages-threshold --pdfa-image-compression --user-words
-            --user-patterns --keep-temporary-files --flowchart --output-type' \
-            --  "$cur" ) )
-        return
-    else
-        _filedir
-        return
-    fi
 } &&
 complete -F _ocrmypdf ocrmypdf
 
+set +o errexit
+
 # ex: filetype=sh

misc/completion/ocrmypdf.fish

@@ -1,15 +1,14 @@
-complete -c ocrmypdf -l version
-complete -c ocrmypdf -l help
+complete -c ocrmypdf -x -n '__fish_is_first_arg' -l version
+complete -c ocrmypdf -x -n '__fish_is_first_arg' -s h -s "?" -l help
 
-complete -c ocrmypdf -l sidecar -r -d "write OCR to text file"
-complete -c ocrmypdf -s q -l quiet
+complete -c ocrmypdf -r -l sidecar -d "write OCR to text file"
+complete -c ocrmypdf -x -s q -l quiet
 
 complete -c ocrmypdf -s r -l rotate-pages -d "rotate pages to correct orientation"
 complete -c ocrmypdf -s d -l deskew -d "fix small horizontal alignment skew"
 complete -c ocrmypdf -s c -l clean -d "clean document images before OCR"
 complete -c ocrmypdf -s i -l clean-final -d "clean document images and keep result"
 complete -c ocrmypdf -l remove-vectors -d "don't send vector objects to OCR"
-complete -c ocrmypdf -l mask-barcodes -d "mask barcodes from OCR"
 complete -c ocrmypdf -l threshold -d "threshold images before OCR"
 
 complete -c ocrmypdf -s f -l force-ocr -d "OCR documents that already have printable text"
@@ -18,8 +17,14 @@ complete -c ocrmypdf -l redo-ocr -d "redo OCR on any pages that seem to have OCR
 
 complete -c ocrmypdf -s k -l keep-temporary-files -d "keep temporary files (debug)"
 
-complete -c ocrmypdf -x -s l -l language -d 'language'
-complete -c ocrmypdf -x -s l -l language -a '(tesseract --list-langs)'
+function __fish_ocrmypdf_languages
+    set langs (tesseract --list-langs ^/dev/null)
+    set arr (string split '\n' $langs)
+    for lang in $arr[2..-1]
+        echo $lang
+    end
+end
+complete -c ocrmypdf -x -s l -l language -a '(__fish_ocrmypdf_languages)' -d "language"
 
 complete -c ocrmypdf -x -l image-dpi -d "assume this DPI if input image DPI is unknown"
 
@@ -37,7 +42,7 @@ function __fish_ocrmypdf_pdf_renderer
     echo -e "hocr\t"(_ "use hocr renderer")
     echo -e "sandwich\t"(_ "use sandwich renderer")
 end
-complete -c ocrmypdf -x -l pdf-render -a '(__fish_ocrmypdf_pdf_renderer)' -d "select PDF renderer options"
+complete -c ocrmypdf -x -l pdf-renderer -a '(__fish_ocrmypdf_pdf_renderer)' -d "select PDF renderer options"
 
 function __fish_ocrmypdf_optimize
     echo -e "0\t"(_ "do not optimize")
@@ -47,8 +52,23 @@ function __fish_ocrmypdf_optimize
 end
 complete -c ocrmypdf -x -s O -l optimize -a '(__fish_ocrmypdf_optimize)' -d "select optimization level"
 
+function __fish_ocrmypdf_verbose
+    echo -e "0\t"(_ "standard output messages")
+    echo -e "1\t"(_ "troubleshooting output messages")
+    echo -e "2\t"(_ "debugging output messages")
+end
+complete -c ocrmypdf -x -s v -l verbose -a '(__fish_ocrmypdf_verbose)' -d "set verbosity level"
+
+complete -c ocrmypdf -x -l no-progress-bar -d "disable the progress bar"
+
+function __fish_ocrmypdf_pdfa_compression
+    echo -e "auto\t"(_ "let Ghostscript decide how to compress images")
+    echo -e "jpeg\t"(_ "convert color and grayscale images to JPEG")
+    echo -e "lossless\t"(_ "convert color and grayscale images to lossless (PNG)")
+end
+complete -c ocrmypdf -x -l pdfa-image-compression -a '(__fish_ocrmypdf_pdfa_compression)' -d "set PDF/A image compression options"
+
 complete -c ocrmypdf -x -s j -l jobs -d "how many worker processes to use"
-complete -c ocrmypdf -x -s v -a '(seq 1 9)'
 complete -c ocrmypdf -x -l title -d "set metadata"
 complete -c ocrmypdf -x -l author -d "set metadata"
 complete -c ocrmypdf -x -l subject -d "set metadata"
@@ -60,11 +80,39 @@ complete -c ocrmypdf -x -l jpeg-quality -d "JPEG quality [0..100]"
 complete -c ocrmypdf -x -l png-quality -d "PNG quality [0..100]"
 complete -c ocrmypdf -x -l jbig2-lossy -d "enable lossy JBIG2 (see docs)"
 complete -c ocrmypdf -x -l max-image-mpixels -d "image decompression bomb threshold"
+complete -c ocrmypdf -x -l pages -d "apply OCR to only the specified pages"
 complete -c ocrmypdf -x -l tesseract-config -d "set custom tesseract config file"
-complete -c ocrmypdf -x -l tesseract-pagesegmode -d "set tesseract --psm"
-complete -c ocrmypdf -x -l tesseract-oem -d "set tesseract --oem"
+
+function __fish_ocrmypdf_tesseract_pagesegmode
+  echo -e "0\t"(_ "orientation and script detection (OSD) only")
+  echo -e "1\t"(_ "automatic page segmentation with OSD")
+  echo -e "2\t"(_ "automatic page segmentation, but no OSD, or OCR")
+  echo -e "3\t"(_ "fully automatic page segmentation, but no OSD (default)")
+  echo -e "4\t"(_ "assume a single column of text of variable sizes")
+  echo -e "5\t"(_ "assume a single uniform block of vertically aligned text")
+  echo -e "6\t"(_ "assume a single uniform block of text")
+  echo -e "7\t"(_ "treat the image as a single text line")
+  echo -e "8\t"(_ "treat the image as a single word")
+  echo -e "9\t"(_ "treat the image as a single word in a circle")
+  echo -e "10\t"(_ "treat the image as a single character")
+  echo -e "11\t"(_ "sparse text - find as much text as possible in no particular order")
+  echo -e "12\t"(_ "sparse text with OSD")
+  echo -e "13\t"(_ "raw line - treat the image as a single text line")
+end
+complete -c ocrmypdf -x -l tesseract-pagesegmode -a '(__fish_ocrmypdf_tesseract_pagesegmode)' -d "set tesseract --psm"
+
+function __fish_ocrmypdf_tesseract_oem
+    echo -e "0\t"(_ "legacy engine only")
+    echo -e "1\t"(_ "neural nets LSTM engine only")
+    echo -e "2\t"(_ "legacy + LSTM engines")
+    echo -e "3\t"(_ "default, based on what is available")
+end
+complete -c ocrmypdf -x -l tesseract-oem -a '(__fish_ocrmypdf_tesseract_oem)' -d "set tesseract --oem"
 complete -c ocrmypdf -x -l tesseract-timeout -d "maximum number of seconds to wait for OCR"
 complete -c ocrmypdf -x -l rotate-pages-threshold -d "page rotation confidence"
-complete -c ocrmypdf -x -l pdfa-image-compression -a 'auto jpeg lossless' -d "set PDF/A image compression options"
+
+complete -c ocrmypdf -r -l user-words -d "specify location of user words file"
+complete -c ocrmypdf -r -l user-patterns -d "specify location of user patterns file"
+complete -c ocrmypdf -x -l fast-web-view -d "if file size if above this amount in MB, linearize PDF"
 
 complete -c ocrmypdf -x -a "(__fish_complete_suffix .pdf)"

misc/docker-compose.example.yml

@@ -0,0 +1,15 @@
+---
+version: "3.3"
+services:
+  ocrmypdf:
+    restart: always
+    container_name: ocrmypdf
+    image: jbarlow83/ocrmypdf
+    volumes:
+      - "/media/scan:/input"
+      - "/mnt/scan:/output"
+    environment:
+      - OCR_OUTPUT_DIRECTORY_YEAR_MONT=0
+    user: "<SET TO YOUR USER ID>:<SET TO YOUR GROUP ID>"
+    entrypoint: python3
+    command: watcher.py

misc/example_plugin.py

@@ -0,0 +1,53 @@
+# © 2020 James R Barlow: https://github.com/jbarlow83
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+
+from PIL import Image
+
+from ocrmypdf import hookimpl
+
+log = logging.getLogger(__name__)
+
+
+@hookimpl
+def add_options(parser):
+    parser.add_argument('--grayscale-ocr', action='store_true')
+
+
+@hookimpl
+def prepare(options):
+    pass
+
+
+@hookimpl
+def validate(pdfinfo, options):
+    pass
+
+
+@hookimpl
+def filter_ocr_image(page, image):
+    if page.options.grayscale_ocr:
+        log.info("graying")
+        return image.convert('L')
+    return image
+
+
+@hookimpl
+def filter_page_image(page, image_filename):
+    output = image_filename.with_suffix('.jpg')
+    with Image.open(image_filename) as im:
+        im.save(output)
+    return output

misc/synology.py

@@ -0,0 +1,72 @@
+#!/bin/env python3
+# Contributed by github.com/Enantiomerie
+
+# This script must be edited to meet your needs.
+
+import logging
+import os
+import shutil
+import subprocess
+import sys
+import time
+
+# pylint: disable=logging-format-interpolation
+# pylint: disable=logging-not-lazy
+
+script_dir = os.path.dirname(os.path.realpath(__file__))
+timestamp = time.strftime("%Y-%m-%d-%H%M_")
+log_file = script_dir + '/' + timestamp + 'ocrmypdf.log'
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s %(message)s',
+    filename=log_file,
+    filemode='w',
+)
+
+if len(sys.argv) > 1:
+    start_dir = sys.argv[1]
+else:
+    start_dir = '.'
+
+for dir_name, subdirs, file_list in os.walk(start_dir):
+    logging.info(dir_name)
+    os.chdir(dir_name)
+    for filename in file_list:
+        file_stem, file_ext = os.path.splitext(filename)
+        if file_ext != '.pdf':
+            continue
+        full_path = os.path.join(dir_name, filename)
+        timestamp_ocr = time.strftime("%Y-%m-%d-%H%M_OCR_")
+        filename_ocr = timestamp_ocr + file_stem + '.pdf'
+        # create string for pdf processing
+        # the script is processed as root user via chron
+        cmd = [
+            'docker',
+            'run',
+            '--rm',
+            '-i',
+            'jbarlow83/ocrmypdf',
+            '--deskew',
+            '-',
+            '-',
+        ]
+        logging.info(cmd)
+        full_path_ocr = os.path.join(dir_name, filename_ocr)
+        with open(filename, 'rb') as input_file, open(
+            full_path_ocr, 'wb'
+        ) as output_file:
+            proc = subprocess.run(
+                cmd,
+                stdin=input_file,
+                stdout=output_file,
+                stderr=subprocess.PIPE,
+                check=False,
+            )
+        logging.info(proc.stderr.read())
+        os.chmod(full_path_ocr, 0o664)
+        os.chmod(full_path, 0o664)
+        full_path_ocr_archive = sys.argv[2]
+        full_path_archive = sys.argv[2] + '/no_ocr'
+        shutil.move(full_path_ocr, full_path_ocr_archive)
+        shutil.move(full_path, full_path_archive)
+logging.info('Finished.\n')

misc/watcher.py

@@ -0,0 +1,156 @@
+# Copyright (C) 2019 Ian Alexander: https://github.com/ianalexander
+# Copyright (C) 2020 James R Barlow: https://github.com/jbarlow83
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+import json
+import logging
+import os
+import sys
+import time
+from datetime import datetime
+from pathlib import Path
+
+import pikepdf
+from watchdog.events import PatternMatchingEventHandler
+from watchdog.observers import Observer
+from watchdog.observers.polling import PollingObserver
+
+import ocrmypdf
+
+# pylint: disable=logging-format-interpolation
+
+INPUT_DIRECTORY = os.getenv('OCR_INPUT_DIRECTORY', '/input')
+OUTPUT_DIRECTORY = os.getenv('OCR_OUTPUT_DIRECTORY', '/output')
+OUTPUT_DIRECTORY_YEAR_MONTH = bool(os.getenv('OCR_OUTPUT_DIRECTORY_YEAR_MONTH', ''))
+ON_SUCCESS_DELETE = bool(os.getenv('OCR_ON_SUCCESS_DELETE', ''))
+DESKEW = bool(os.getenv('OCR_DESKEW', ''))
+OCR_JSON_SETTINGS = json.loads(os.getenv('OCR_JSON_SETTINGS', '{}'))
+POLL_NEW_FILE_SECONDS = int(os.getenv('OCR_POLL_NEW_FILE_SECONDS', '1'))
+USE_POLLING = bool(os.getenv('OCR_USE_POLLING', ''))
+LOGLEVEL = os.getenv('OCR_LOGLEVEL', 'INFO').upper()
+PATTERNS = ['*.pdf']
+
+log = logging.getLogger('ocrmypdf-watcher')
+
+
+def get_output_dir(root, basename):
+    if OUTPUT_DIRECTORY_YEAR_MONTH:
+        today = datetime.today()
+        output_directory_year_month = (
+            Path(root) / str(today.year) / f'{today.month:02d}'
+        )
+        if not output_directory_year_month.exists():
+            output_directory_year_month.mkdir(parents=True, exist_ok=True)
+        output_path = Path(output_directory_year_month) / basename
+    else:
+        output_path = Path(OUTPUT_DIRECTORY) / basename
+    return output_path
+
+
+def wait_for_file_ready(file_path):
+    # This loop waits to make sure that the file is completely loaded on
+    # disk before attempting to read. Docker sometimes will publish the
+    # watchdog event before the file is actually fully on disk, causing
+    # pikepdf to fail.
+
+    retries = 5
+    while retries:
+        try:
+            pdf = pikepdf.open(file_path)
+        except (FileNotFoundError, pikepdf.PdfError) as e:
+            log.info(f"File {file_path} is not ready yet")
+            log.debug("Exception was", exc_info=e)
+            time.sleep(POLL_NEW_FILE_SECONDS)
+            retries -= 1
+        else:
+            pdf.close()
+            return True
+
+    return False
+
+
+def execute_ocrmypdf(file_path):
+    file_path = Path(file_path)
+    output_path = get_output_dir(OUTPUT_DIRECTORY, file_path.name)
+
+    log.info("-" * 20)
+    log.info(f'New file: {file_path}. Waiting until fully loaded...')
+    if not wait_for_file_ready(file_path):
+        log.info(f"Gave up waiting for {file_path} to become ready")
+        return
+    log.info(f'Attempting to OCRmyPDF to: {output_path}')
+    exit_code = ocrmypdf.ocr(
+        input_file=file_path,
+        output_file=output_path,
+        deskew=DESKEW,
+        **OCR_JSON_SETTINGS,
+    )
+    if exit_code == 0 and ON_SUCCESS_DELETE:
+        log.info(f'OCR is done. Deleting: {file_path}')
+        file_path.unlink()
+    else:
+        log.info('OCR is done')
+
+
+class HandleObserverEvent(PatternMatchingEventHandler):
+    def on_any_event(self, event):
+        if event.event_type in ['created']:
+            execute_ocrmypdf(event.src_path)
+
+
+def main():
+    ocrmypdf.configure_logging(
+        verbosity=ocrmypdf.Verbosity.default, manage_root_logger=True
+    )
+    log.setLevel(LOGLEVEL)
+    log.info(
+        f"Starting OCRmyPDF watcher with config:\n"
+        f"Input Directory: {INPUT_DIRECTORY}\n"
+        f"Output Directory: {OUTPUT_DIRECTORY}\n"
+        f"Output Directory Year & Month: {OUTPUT_DIRECTORY_YEAR_MONTH}"
+    )
+    log.debug(
+        f"INPUT_DIRECTORY: {INPUT_DIRECTORY}\n"
+        f"OUTPUT_DIRECTORY: {OUTPUT_DIRECTORY}\n"
+        f"OUTPUT_DIRECTORY_YEAR_MONTH: {OUTPUT_DIRECTORY_YEAR_MONTH}\n"
+        f"ON_SUCCESS_DELETE: {ON_SUCCESS_DELETE}\n"
+        f"DESKEW: {DESKEW}\n"
+        f"ARGS: {OCR_JSON_SETTINGS}\n"
+        f"POLL_NEW_FILE_SECONDS: {POLL_NEW_FILE_SECONDS}\n"
+        f"USE_POLLING: {USE_POLLING}\n"
+        f"LOGLEVEL: {LOGLEVEL}\n"
+    )
+
+    if 'input_file' in OCR_JSON_SETTINGS or 'output_file' in OCR_JSON_SETTINGS:
+        log.error('OCR_JSON_SETTINGS should not specify input file or output file')
+        sys.exit(1)
+
+    handler = HandleObserverEvent(patterns=PATTERNS)
+    if USE_POLLING:
+        observer = PollingObserver()
+    else:
+        observer = Observer()
+    observer.schedule(handler, INPUT_DIRECTORY, recursive=True)
+    observer.start()
+    try:
+        while True:
+            time.sleep(1)
+    except KeyboardInterrupt:
+        observer.stop()
+    observer.join()
+
+
+if __name__ == "__main__":
+    main()

misc/webservice.py

@@ -23,21 +23,22 @@ to emphasize that SaaS deployments should make sure they comply with
 Ghostscript's license as well as OCRmyPDF's.
 """
 
+import os
+import shlex
+from subprocess import PIPE, run
+from tempfile import TemporaryDirectory
+
 from flask import (
     Flask,
     Response,
-    flash,
-    request,
-    redirect,
-    url_for,
     abort,
+    flash,
+    redirect,
+    request,
     send_from_directory,
+    url_for,
 )
-from subprocess import run, PIPE
-from tempfile import TemporaryDirectory
 from werkzeug.utils import secure_filename
-import os
-import shlex
 
 app = Flask(__name__)
 app.secret_key = "secret"

pyproject.toml

@@ -10,7 +10,8 @@ build-backend = "setuptools.build_meta"
 
 [tool.black]
 line-length = 88
-py36 = true
+target-version = ["py36",
+"py37", "py38"]
 skip-string-normalization = true
 include = '\.pyi?$'
 exclude = '''
@@ -28,5 +29,6 @@ exclude = '''
   | docs
   | misc
   | \.egg-info
+  | src/ocrmypdf/lib/_leptonica.py
 )/
 '''

requirements/dev.txt

@@ -1,4 +0,0 @@
-check-manifest >= 0.35
-twine >= 1.8.1
-coverage >= 4.5
-GitPython == 2.1.3

requirements/main.txt

@@ -1,13 +1,12 @@
 # requirements.txt can be used to replicate the developer's build environment
 # setup.py lists a separate set of requirements that are looser to simplify
 # installation
-chardet == 3.0.4
-cffi == 1.12.2
-img2pdf == 0.3.3
-pdfminer.six == 20181108
-pikepdf == 1.3.0
-Pillow >= 5.0.0, != 5.1.0 ; sys_platform == "darwin"
-pycparser == 2.19
-python-xmp-toolkit == 2.0.1
-reportlab == 3.5.13
-tqdm == 4.32.1
+cffi == 1.14.0
+coloredlogs == 14.0  # technically optional
+img2pdf == 0.3.4
+pdfminer.six == 20200517
+pikepdf == 1.14.0
+pluggy == 0.13.1
+Pillow == 7.1.1
+reportlab == 3.5.34
+tqdm == 4.45.0

requirements/test.txt

@@ -1,8 +1,7 @@
-pytest >= 4.4.1, < 5
+pytest >= 5.0.0
 pytest-helpers-namespace >= 2019.1.8
-pytest-xdist == 1.28.0
-pytest-cov >= 2.6.1
-python-xmp-toolkit   # requires apt-get install libexempi3
+pytest-xdist >= 1.31.0
+pytest-cov >= 2.9.0
+python-xmp-toolkit == 2.0.1  # requires apt-get install libexempi3
                      # or brew install exempi
-PyPDF2 >= 1.26.0
 #PyMuPDF == 1.13.4    # optional

requirements/watcher.txt

@@ -0,0 +1 @@
+watchdog >= 0.8.2, < 1.0

requirements/webservice.txt

@@ -0,0 +1 @@
+Flask >= 1, < 2

setup.cfg

@@ -22,6 +22,8 @@ include_trailing_comma=True
 force_grid_wrap=0
 use_parentheses=True
 line_length=88
+known_first_party = ocrmypdf
+known_third_party = PIL,_cffi_backend,cffi,flask,img2pdf,pdfminer,pikepdf,pkg_resources,pluggy,pytest,reportlab,setuptools,sphinx_rtd_theme,tqdm,watchdog,werkzeug
 
 [metadata]
 license_file = LICENSE

setup.py

@@ -21,27 +21,12 @@ from __future__ import print_function, unicode_literals
 
 import sys
 
+from setuptools import find_packages, setup
+
 if sys.version_info < (3, 6):
     print("Python 3.6 or newer is required", file=sys.stderr)
     sys.exit(1)
 
-from setuptools import setup, find_packages
-
-# pylint: disable=w0613
-
-
-command = next((arg for arg in sys.argv[1:] if not arg.startswith('-')), '')
-if command.startswith('install') or command in [
-    'check',
-    'test',
-    'nosetests',
-    'easy_install',
-]:
-    forced = '--force' in sys.argv
-    if forced:
-        print("The argument --force is deprecated. Please discontinue use.")
-
-
 if 'upload' in sys.argv[1:]:
     print('Use twine to upload the package - setup.py upload is insecure')
     sys.exit(1)
@@ -61,13 +46,14 @@ setup(
     long_description_content_type='text/markdown',
     url='https://github.com/jbarlow83/OCRmyPDF',
     author='James R. Barlow',
-    author_email='jim@purplerock.ca',
+    author_email='james@purplerock.ca',
     packages=find_packages('src', exclude=["tests", "tests.*"]),
     package_dir={'': 'src'},
     keywords=['PDF', 'OCR', 'optical character recognition', 'PDF/A', 'scanning'],
     classifiers=[
         "Programming Language :: Python :: 3.6",
         "Programming Language :: Python :: 3.7",
+        "Programming Language :: Python :: 3.8",
         "Development Status :: 5 - Production/Stable",
         "Environment :: Console",
         "Intended Audience :: End Users/Desktop",
@@ -75,6 +61,7 @@ setup(
         "Intended Audience :: System Administrators",
         "License :: OSI Approved :: GNU General Public License v3 (GPLv3)",
         "Operating System :: MacOS :: MacOS X",
+        "Operating System :: Microsoft :: Windows :: Windows 10",
         "Operating System :: POSIX",
         "Operating System :: POSIX :: BSD",
         "Operating System :: POSIX :: Linux",
@@ -92,18 +79,16 @@ setup(
     use_scm_version={'version_scheme': 'post-release'},
     cffi_modules=['src/ocrmypdf/lib/compile_leptonica.py:ffibuilder'],
     install_requires=[
-        'chardet >= 3.0.4, < 4',  # unlisted requirement of pdfminer.six 20181108
         'cffi >= 1.9.1',  # must be a setup and install requirement
+        'coloredlogs >= 14.0',  # strictly optional
         'img2pdf >= 0.3.0, < 0.4',  # pure Python, so track HEAD closely
-        'pdfminer.six == 20181108 ; sys_platform != "darwin"',
-        'pikepdf >= 1.3.0, < 2',
-        'Pillow >= 4.0.0, != 5.1.0 ; sys_platform == "darwin"',
-        # Pillow < 4 has BytesIO/TIFF bug w/img2pdf 0.2.3
-        # block 5.1.0, broken wheels
+        'pdfminer.six >= 20191110, <= 20200517',
+        'pikepdf >= 1.14.0, < 2',
+        'Pillow >= 7.0.0',
+        'pluggy >= 0.13.0',
         'reportlab >= 3.3.0',  # oldest released version with sane image handling
         'tqdm >= 4',
     ],
-    extras_require={'pdfminer': ['pdfminer.six == 20181108']},
     tests_require=tests_require,
     entry_points={'console_scripts': ['ocrmypdf = ocrmypdf.__main__:run']},
     package_data={'ocrmypdf': ['data/sRGB.icc']},

src/ocrmypdf/__init__.py

@@ -15,33 +15,27 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import pkg_resources
 
-PROGRAM_NAME = 'ocrmypdf'
+from pluggy import HookimplMarker as _HookimplMarker
 
-# Official PEP 396
-__version__ = pkg_resources.get_distribution('ocrmypdf').version
-
-VERSION = __version__
-
-from .exceptions import (
-    ExitCode,
+from ocrmypdf import helpers, hocrtransform, leptonica, pdfa, pdfinfo
+from ocrmypdf._version import PROGRAM_NAME, __version__
+from ocrmypdf.api import Verbosity, configure_logging, ocr
+from ocrmypdf.exceptions import (
     BadArgsError,
-    PdfMergeFailedError,
-    MissingDependencyError,
-    UnsupportedImageFormatError,
     DpiError,
-    OutputFileAccessError,
-    PriorOcrFoundError,
-    InputFileError,
-    SubprocessOutputError,
     EncryptedPdfError,
+    ExitCode,
+    ExitCodeException,
+    InputFileError,
+    MissingDependencyError,
+    OutputFileAccessError,
+    PdfMergeFailedError,
+    PriorOcrFoundError,
+    SubprocessOutputError,
     TesseractConfigError,
+    UnsupportedImageFormatError,
 )
+from ocrmypdf.pluginspec import OcrEngine, OrientationConfidence
 
-from . import helpers
-from . import hocrtransform
-from . import leptonica
-from . import pdfa
-from . import pdfinfo
-from .api import run, configure_logging, Verbosity
+hookimpl = _HookimplMarker('ocrmypdf')

src/ocrmypdf/__main__.py

@@ -19,18 +19,20 @@
 import logging
 import os
 import sys
+from multiprocessing import set_start_method
 
-from . import __version__
-from .cli import parser
-from .api import configure_logging, Verbosity
-from ._jobcontext import make_logger
-from ._sync import run_pipeline
-from ._validation import check_closed_streams, check_options
-from .exceptions import ExitCode, BadArgsError, MissingDependencyError
+from ocrmypdf import __version__
+from ocrmypdf._plugin_manager import get_parser_options_plugins
+from ocrmypdf._sync import run_pipeline
+from ocrmypdf._validation import check_closed_streams, check_options
+from ocrmypdf.api import Verbosity, configure_logging
+from ocrmypdf.exceptions import BadArgsError, ExitCode, MissingDependencyError
+
+log = logging.getLogger('ocrmypdf')
 
 
 def run(args=None):
-    options = parser.parse_args(args=args)
+    _parser, options, plugin_manager = get_parser_options_plugins(args=args)
 
     if not check_closed_streams(options):
         return ExitCode.bad_args
@@ -47,10 +49,9 @@ def run(args=None):
     configure_logging(
         verbosity, progress_bar_friendly=options.progress_bar, manage_root_logger=True
     )
-    log = make_logger('ocrmypdf')
-    log.debug('ocrmypdf ' + __version__)
+    log.debug('ocrmypdf %s', __version__)
     try:
-        check_options(options)
+        check_options(options, plugin_manager)
     except ValueError as e:
         log.error(e)
         return ExitCode.bad_args
@@ -61,9 +62,11 @@ def run(args=None):
         log.error(e)
         return ExitCode.missing_dependency
 
-    result = run_pipeline(options=options)
+    result = run_pipeline(options=options, plugin_manager=plugin_manager)
     return result
 
 
 if __name__ == '__main__':
+    if sys.platform == 'darwin' and sys.version_info < (3, 8):
+        set_start_method('spawn')  # see python bpo-33725
     sys.exit(run())

src/ocrmypdf/_concurrent.py

@@ -0,0 +1,135 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+import logging.handlers
+import multiprocessing
+import os
+import signal
+import sys
+import threading
+from multiprocessing import Pool as ProcessPool
+from multiprocessing.dummy import Pool as ThreadPool
+from typing import Callable, Iterable, Optional
+
+from tqdm import tqdm
+
+
+def log_listener(queue):
+    """Listen to the worker processes and forward the messages to logging
+
+    For simplicity this is a thread rather than a process. Only one process
+    should actually write to sys.stderr or whatever we're using, so if this is
+    made into a process the main application needs to be directed to it.
+
+    See https://docs.python.org/3/howto/logging-cookbook.html#logging-to-a-single-file-from-multiple-processes
+    """
+
+    while True:
+        try:
+            record = queue.get()
+            if record is None:
+                break
+            logger = logging.getLogger(record.name)
+            logger.handle(record)
+        except Exception:  # pylint: disable=broad-except
+            import traceback  # pylint: disable=import-outside-toplevel
+
+            print("Logging problem", file=sys.stderr)
+            traceback.print_exc(file=sys.stderr)
+
+
+def process_init(queue, user_init):
+    """Initialize a process pool worker"""
+
+    # Ignore SIGINT (our parent process will kill us gracefully)
+    signal.signal(signal.SIGINT, signal.SIG_IGN)
+
+    # Reconfigure the root logger for this process to send all messages to a queue
+    h = logging.handlers.QueueHandler(queue)
+    root = logging.getLogger()
+    root.handlers = []
+    root.addHandler(h)
+
+    if user_init:
+        user_init()
+
+
+def thread_init(_queue, user_init):
+    if user_init:
+        user_init()
+
+
+def exec_progress_pool(
+    *,
+    use_threads: bool,
+    max_workers: int,
+    tqdm_kwargs: dict,
+    task_initializer: Optional[Callable] = None,
+    task: Optional[Callable] = None,
+    task_arguments: Optional[Iterable] = None,
+    task_finished: Optional[Callable] = None,
+):
+    log_queue = multiprocessing.Queue(-1)
+    listener = threading.Thread(target=log_listener, args=(log_queue,))
+
+    if use_threads:
+        pool_class = ThreadPool
+        initializer = thread_init
+    else:
+        pool_class = ProcessPool
+        initializer = process_init
+    listener.start()
+
+    with tqdm(**tqdm_kwargs) as pbar:
+        pool = pool_class(
+            processes=max_workers,
+            initializer=initializer,
+            initargs=(log_queue, task_initializer),
+        )
+        try:
+            results = pool.imap_unordered(task, task_arguments)
+            while True:
+                try:
+                    result = results.next()
+                    if task_finished:
+                        task_finished(result, pbar)
+                    else:
+                        pbar.update()
+                except StopIteration:
+                    break
+        except KeyboardInterrupt:
+            # Terminate pool so we exit instantly
+            pool.terminate()
+            # Don't try listener.join() here, will deadlock
+            raise
+        except Exception:
+            if not os.environ.get("PYTEST_CURRENT_TEST", ""):
+                # Unless inside pytest, exit immediately because no one wants
+                # to wait for child processes to finalize results that will be
+                # thrown away. Inside pytest, we want child processes to exit
+                # cleanly so that they output an error messages or coverage data
+                # we need from them.
+                pool.terminate()
+            raise
+        finally:
+            # Terminate log listener
+            log_queue.put_nowait(None)
+            pool.close()
+            pool.join()
+
+    listener.join()

src/ocrmypdf/_exec/__init__.py

@@ -0,0 +1,18 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Manage third party executables"""

src/ocrmypdf/_exec/ghostscript.py

@@ -0,0 +1,244 @@
+# © 2017 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+"""Interface to Ghostscript executable"""
+
+import logging
+import os
+import re
+from io import BytesIO
+from os import fspath
+from pathlib import Path
+from shutil import which
+from subprocess import PIPE, CalledProcessError
+
+from PIL import Image
+
+from ocrmypdf.exceptions import MissingDependencyError, SubprocessOutputError
+from ocrmypdf.helpers import Resolution
+from ocrmypdf.subprocess import get_version, run
+
+log = logging.getLogger(__name__)
+
+GS = 'gs'
+if os.name == 'nt':
+    GS = which('gswin64c')
+    if not GS:
+        GS = which('gswin32c')
+        if not GS:
+            raise MissingDependencyError(
+                """
+                ---------------------------------------------------------------------
+                This error normally occurs when ocrmypdf can't Ghostscript.  Please
+                ensure Ghostscript is installed and its location is added to the
+                system PATH environment variable.
+
+                For details see:
+                    https://ocrmypdf.readthedocs.io/en/latest/installation.html
+                ---------------------------------------------------------------------
+                """
+            )
+    GS = Path(GS).stem
+
+
+def version():
+    return get_version(GS)
+
+
+def jpeg_passthrough_available() -> bool:
+    """Returns True if the installed version of Ghostscript supports JPEG passthru
+
+    Prior to 9.23, Ghostscript decode and re-encoded JPEGs internally. In 9.23
+    it gained the ability to keep JPEGs unmodified. However, the 9.23
+    implementation was buggy and would deletes the last two bytes of images in
+    some cases, as reported here.
+    https://bugs.ghostscript.com/show_bug.cgi?id=699216
+
+    The issue was fixed for 9.24, hence that is the first version we consider
+    the feature available. (However, we don't use 9.24 at all, so the first
+    version that allows JPEG passthrough is 9.25.
+
+    """
+    return version() >= '9.24'
+
+
+def _gs_error_reported(stream) -> bool:
+    return re.search(r'error', stream, flags=re.IGNORECASE)
+
+
+def rasterize_pdf(
+    input_file: os.PathLike,
+    output_file: os.PathLike,
+    *,
+    raster_device: str,
+    raster_dpi: Resolution,
+    pageno: int = 1,
+    page_dpi: Resolution = None,
+    rotation: int = None,
+    filter_vector: bool = False,
+):
+    """Rasterize one page of a PDF at resolution raster_dpi in canvas units."""
+    raster_dpi = raster_dpi.round(6)
+    if not page_dpi:
+        page_dpi = raster_dpi
+
+    args_gs = (
+        [
+            GS,
+            '-dQUIET',
+            '-dSAFER',
+            '-dBATCH',
+            '-dNOPAUSE',
+            f'-sDEVICE={raster_device}',
+            f'-dFirstPage={pageno}',
+            f'-dLastPage={pageno}',
+            f'-r{raster_dpi.x:f}x{raster_dpi.y:f}',
+        ]
+        + (['-dFILTERVECTOR'] if filter_vector else [])
+        + [
+            '-o',
+            '-',
+            '-sstdout=%stderr',
+            '-dAutoRotatePages=/None',  # Probably has no effect on raster
+            '-f',
+            fspath(input_file),
+        ]
+    )
+
+    try:
+        p = run(args_gs, stdout=PIPE, stderr=PIPE, check=True)
+    except CalledProcessError as e:
+        log.error(e.stderr.decode(errors='replace'))
+        raise SubprocessOutputError('Ghostscript rasterizing failed')
+    else:
+        stderr = p.stderr.decode(errors='replace')
+        if _gs_error_reported(stderr):
+            log.error(stderr)
+        elif stderr:
+            log.debug(stderr)
+
+    with Image.open(BytesIO(p.stdout)) as im:
+        if rotation is not None:
+            log.debug("Rotating output by %i", rotation)
+            # rotation is a clockwise angle and Image.ROTATE_* is
+            # counterclockwise so this cancels out the rotation
+            if rotation == 90:
+                im = im.transpose(Image.ROTATE_90)
+            elif rotation == 180:
+                im = im.transpose(Image.ROTATE_180)
+            elif rotation == 270:
+                im = im.transpose(Image.ROTATE_270)
+            if rotation % 180 == 90:
+                page_dpi = page_dpi.flip_axis()
+        im.save(fspath(output_file), dpi=page_dpi)
+
+
+def generate_pdfa(
+    pdf_pages,
+    output_file: os.PathLike,
+    compression: str,
+    pdf_version: str = '1.5',
+    pdfa_part: str = '2',
+):
+    compression_args = []
+    if compression == 'jpeg':
+        compression_args = [
+            "-dAutoFilterColorImages=false",
+            "-dColorImageFilter=/DCTEncode",
+            "-dAutoFilterGrayImages=false",
+            "-dGrayImageFilter=/DCTEncode",
+        ]
+    elif compression == 'lossless':
+        compression_args = [
+            "-dAutoFilterColorImages=false",
+            "-dColorImageFilter=/FlateEncode",
+            "-dAutoFilterGrayImages=false",
+            "-dGrayImageFilter=/FlateEncode",
+        ]
+    else:
+        compression_args = [
+            "-dAutoFilterColorImages=true",
+            "-dAutoFilterGrayImages=true",
+        ]
+
+    # Older versions of Ghostscript expect a leading slash in
+    # sColorConversionStrategy, newer ones should not have it. See Ghostscript
+    # git commit fe1c025d.
+    strategy = 'RGB' if version() >= '9.19' else '/RGB'
+
+    if version() == '9.23':
+        # 9.23: new feature JPEG passthrough is broken in some cases, best to
+        # disable it always
+        # https://bugs.ghostscript.com/show_bug.cgi?id=699216
+        compression_args.append('-dPassThroughJPEGImages=false')
+
+    # nb no need to specify ProcessColorModel when ColorConversionStrategy
+    # is set; see:
+    # https://bugs.ghostscript.com/show_bug.cgi?id=699392
+    args_gs = (
+        [
+            GS,
+            "-dQUIET",
+            "-dBATCH",
+            "-dNOPAUSE",
+            "-dSAFER",
+            "-dCompatibilityLevel=" + str(pdf_version),
+            "-sDEVICE=pdfwrite",
+            "-dAutoRotatePages=/None",
+            "-sColorConversionStrategy=" + strategy,
+        ]
+        + compression_args
+        + [
+            "-dJPEGQ=95",
+            "-dPDFA=" + pdfa_part,
+            "-dPDFACompatibilityPolicy=1",
+            "-o",
+            "-",
+            "-sstdout=%stderr",
+        ]
+    )
+    args_gs.extend(fspath(s) for s in pdf_pages)  # Stringify Path objs
+    try:
+        with Path(output_file).open('wb') as output:
+            p = run(args_gs, stdout=output, stderr=PIPE, check=True)
+    except CalledProcessError as e:
+        # Ghostscript does not change return code when it fails to create
+        # PDF/A - check PDF/A status elsewhere
+        log.error(e.stderr.decode(errors='replace'))
+        raise SubprocessOutputError('Ghostscript PDF/A rendering failed')
+    else:
+        stderr = p.stderr.decode('utf-8', errors='replace')
+        if _gs_error_reported(stderr):
+            last_part = None
+            repcount = 0
+            for part in stderr.split('****'):
+                if part != last_part:
+                    if repcount > 1:
+                        log.error(f"(previous error message repeated {repcount} times)")
+                        repcount = 0
+                    log.error(part)
+                else:
+                    repcount += 1
+                last_part = part
+        elif 'overprint mode not set' in stderr:
+            # Unless someone is going to print PDF/A documents on a
+            # magical sRGB printer I can't see the removal of overprinting
+            # being a problem....
+            log.debug(
+                "Ghostscript had to remove PDF 'overprinting' from the "
+                "input file to complete PDF/A conversion. "
+            )

src/ocrmypdf/_exec/jbig2enc.py

@@ -15,14 +15,14 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from functools import lru_cache
-from subprocess import PIPE, run
+"""Interface to jbig2 executable"""
 
-from . import get_version
-from ..exceptions import MissingDependencyError
+from subprocess import PIPE
+
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.subprocess import get_version, run
 
 
-@lru_cache(maxsize=1)
 def version():
     return get_version('jbig2', regex=r'jbig2enc (\d+(\.\d+)*).*')
 

src/ocrmypdf/_exec/pngquant.py

@@ -15,17 +15,16 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from functools import lru_cache
-from subprocess import run
+"""Interface to pngquant executable"""
+
 from tempfile import NamedTemporaryFile
 
 from PIL import Image
 
-from . import get_version
-from ..exceptions import MissingDependencyError
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.subprocess import get_version, run
 
 
-@lru_cache(maxsize=1)
 def version():
     return get_version('pngquant', regex=r'(\d+(\.\d+)*).*')
 
@@ -40,8 +39,7 @@ def available():
 
 def quantize(input_file, output_file, quality_min, quality_max):
     if input_file.endswith('.jpg'):
-        im = Image.open(input_file)
-        with NamedTemporaryFile(suffix='.png') as tmp:
+        with Image.open(input_file) as im, NamedTemporaryFile(suffix='.png') as tmp:
             im.save(tmp)
             args = [
                 'pngquant',

src/ocrmypdf/_exec/tesseract.py

@@ -15,22 +15,27 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
+"""Interface to Tesseract executable"""
+
+import logging
 import os
 import shutil
-import sys
 from collections import namedtuple
-from contextlib import suppress
-from functools import lru_cache
 from os import fspath
-from subprocess import PIPE, STDOUT, CalledProcessError, TimeoutExpired, run
+from pathlib import Path
+from subprocess import PIPE, STDOUT, CalledProcessError, TimeoutExpired
+from typing import List
 
-from . import get_version
-from ..exceptions import (
+from PIL import Image
+
+from ocrmypdf.exceptions import (
     MissingDependencyError,
-    TesseractConfigError,
     SubprocessOutputError,
+    TesseractConfigError,
 )
-from ..helpers import page_number
+from ocrmypdf.subprocess import get_version, run
+
+log = logging.getLogger(__name__)
 
 OrientationConfidence = namedtuple('OrientationConfidence', ('angle', 'confidence'))
 
@@ -52,43 +57,47 @@ HOCR_TEMPLATE = """<?xml version="1.0" encoding="UTF-8"?>
 """
 
 
-def version(tesseract_env=None):
-    return get_version('tesseract', regex=r'tesseract\s(.+)', env=tesseract_env)
+class TesseractLoggerAdapter(logging.LoggerAdapter):
+    def process(self, msg, kwargs):
+        kwargs['extra'] = self.extra
+        return '[tesseract] %s' % (msg), kwargs
 
 
-def v4(tesseract_env=None):
-    "Is this Tesseract v4.0?"
-    return version(tesseract_env) >= '4'
+def version():
+    return get_version('tesseract', regex=r'tesseract\s(.+)')
 
 
-def has_textonly_pdf(tesseract_env=None):
+def has_textonly_pdf(langs=None):
     """Does Tesseract have textonly_pdf capability?
 
     Available in v4.00.00alpha since January 2017. Best to
-    parse the parameter list
+    parse the parameter list.
     """
-    args_tess = ['tesseract', '--print-parameters', 'pdf']
+    args_tess = tess_base_args(langs, engine_mode=None) + ['--print-parameters', 'pdf']
     params = ''
     try:
-        proc = run(
-            args_tess,
-            check=True,
-            universal_newlines=True,
-            stdout=PIPE,
-            stderr=STDOUT,
-            env=tesseract_env,
-        )
+        proc = run(args_tess, check=True, stdout=PIPE, stderr=STDOUT)
         params = proc.stdout
     except CalledProcessError as e:
         raise MissingDependencyError(
-            "Could not --print-parameters from tesseract"
+            "Could not --print-parameters from tesseract. This can happen if the "
+            "TESSDATA_PREFIX environment is not set to a valid tessdata folder. "
         ) from e
-    if 'textonly_pdf' in params:
+    if b'textonly_pdf' in params:
         return True
     return False
 
 
-def languages(tesseract_env=None):
+def has_user_words():
+    """Does Tesseract have --user-words capability?
+
+    Not available in 4.0, but available in 4.1. Also available in 3.x, but
+    we no longer support 3.x.
+    """
+    return version() >= '4.1'
+
+
+def get_languages():
     def lang_error(output):
         msg = (
             "Tesseract failed to report available languages.\n"
@@ -101,33 +110,29 @@ def languages(tesseract_env=None):
     args_tess = ['tesseract', '--list-langs']
     try:
         proc = run(
-            args_tess,
-            universal_newlines=True,
-            stdout=PIPE,
-            stderr=STDOUT,
-            check=True,
-            env=tesseract_env,
+            args_tess, universal_newlines=True, stdout=PIPE, stderr=STDOUT, check=True
         )
         output = proc.stdout
     except CalledProcessError as e:
         raise MissingDependencyError(lang_error(e.output)) from e
 
-    header, *rest = output.splitlines()
-    if not header.startswith('List of available languages'):
-        raise MissingDependencyError(lang_error(output))
+    for line in output.splitlines():
+        if line.startswith('Error'):
+            raise MissingDependencyError(lang_error(output))
+    _header, *rest = output.splitlines()
     return set(lang.strip() for lang in rest)
 
 
-def tess_base_args(langs, engine_mode):
+def tess_base_args(langs: List[str], engine_mode) -> List[str]:
     args = ['tesseract']
     if langs:
         args.extend(['-l', '+'.join(langs)])
-    if engine_mode is not None and v4():
+    if engine_mode is not None:
         args.extend(['--oem', str(engine_mode)])
     return args
 
 
-def get_orientation(input_file, engine_mode, timeout: float, log, tesseract_env=None):
+def get_orientation(input_file: Path, engine_mode, timeout: float):
     args_tesseract = tess_base_args(['osd'], engine_mode) + [
         '--psm',
         '0',
@@ -136,19 +141,13 @@ def get_orientation(input_file, engine_mode, timeout: float, log, tesseract_env=
     ]
 
     try:
-        p = run(
-            args_tesseract,
-            stdout=PIPE,
-            stderr=STDOUT,
-            timeout=timeout,
-            check=True,
-            env=tesseract_env,
-        )
+        p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
         stdout = p.stdout
     except TimeoutExpired:
         return OrientationConfidence(angle=0, confidence=0.0)
     except CalledProcessError as e:
-        tesseract_log_output(log, e.output, input_file)
+        tesseract_log_output(e.stdout)
+        tesseract_log_output(e.stderr)
         if (
             b'Too few characters. Skipping this page' in e.output
             or b'Image too large' in e.output
@@ -170,19 +169,17 @@ def get_orientation(input_file, engine_mode, timeout: float, log, tesseract_env=
         return oc
 
 
-def tesseract_log_output(log, stdout, input_file):
-    prefix = "[tesseract] "
+def tesseract_log_output(stream):
+    tlog = TesseractLoggerAdapter(
+        log, extra=log.extra if hasattr(log, 'extra') else None
+    )
 
+    if not stream:
+        return
     try:
-        text = stdout.decode()
+        text = stream.decode()
     except UnicodeDecodeError:
-        log.error(
-            prefix
-            + "command line output was not utf-8. "
-            + "This usually means Tesseract's language packs do not match "
-            "the installed version of Tesseract."
-        )
-        text = stdout.decode('utf-8', 'backslashreplace')
+        text = stream.decode('utf-8', 'ignore')
 
     lines = text.splitlines()
     for line in lines:
@@ -191,67 +188,58 @@ def tesseract_log_output(log, stdout, input_file):
         elif line.startswith("Warning in pixReadMem"):
             continue
         elif 'diacritics' in line:
-            log.warning(prefix + "lots of diacritics - possibly poor OCR")
+            tlog.warning("lots of diacritics - possibly poor OCR")
         elif line.startswith('OSD: Weak margin'):
-            log.warning(prefix + "unsure about page orientation")
+            tlog.warning("unsure about page orientation")
         elif 'Error in pixScanForForeground' in line:
             pass  # Appears to be spurious/problem with nonwhite borders
         elif 'Error in boxClipToRectangle' in line:
             pass  # Always appears with pixScanForForeground message
         elif 'parameter not found: ' in line.lower():
-            log.error(prefix + line.strip())
+            tlog.error(line.strip())
             problem = line.split('found: ')[1]
             raise TesseractConfigError(problem)
         elif 'error' in line.lower() or 'exception' in line.lower():
-            log.error(prefix + line.strip())
+            tlog.error(line.strip())
         elif 'warning' in line.lower():
-            log.warning(prefix + line.strip())
+            tlog.warning(line.strip())
         elif 'read_params_file' in line.lower():
-            log.error(prefix + line.strip())
+            tlog.error(line.strip())
         else:
-            log.info(prefix + line.strip())
+            tlog.info(line.strip())
 
 
-def page_timedout(log, input_file, timeout):
+def page_timedout(timeout):
     if timeout == 0:
         return
-    prefix = f"{(page_number(input_file)):4d}: [tesseract] "
-    log.warning(prefix + " took too long to OCR - skipping")
+    log.warning("[tesseract] took too long to OCR - skipping")
 
 
-def _generate_null_hocr(output_hocr, output_sidecar, image):
+def _generate_null_hocr(output_hocr, output_text, image):
     """Produce a .hocr file that reports no text detected on a page that is
     the same size as the input image."""
-    from PIL import Image
+    with Image.open(image) as im:
+        w, h = im.size
 
-    im = Image.open(image)
-    w, h = im.size
-
-    with open(output_hocr, 'w', encoding="utf-8") as f:
-        f.write(HOCR_TEMPLATE.format(w, h))
-    with open(output_sidecar, 'w', encoding='utf-8') as f:
-        f.write('[skipped page]')
+    output_hocr.write_text(HOCR_TEMPLATE.format(w, h), encoding='utf-8')
+    output_text.write_text('[skipped page]', encoding='utf-8')
 
 
 def generate_hocr(
-    input_file,
-    output_files,
-    language: list,
+    input_file: Path,
+    output_hocr: Path,
+    output_text: Path,
+    languages: list,
     engine_mode,
     tessconfig: list,
     timeout: float,
     pagesegmode: int,
     user_words,
     user_patterns,
-    tesseract_env,
-    log,
 ):
+    prefix = output_hocr.with_suffix('')
 
-    output_hocr = next(o for o in output_files if o.endswith('.hocr'))
-    output_sidecar = next(o for o in output_files if o.endswith('.txt'))
-    prefix = os.path.splitext(output_hocr)[0]
-
-    args_tesseract = tess_base_args(language, engine_mode)
+    args_tesseract = tess_base_args(languages, engine_mode)
 
     if pagesegmode is not None:
         args_tesseract.extend(['--psm', str(pagesegmode)])
@@ -262,95 +250,70 @@ def generate_hocr(
     if user_patterns:
         args_tesseract.extend(['--user-patterns', user_patterns])
 
-    # Reminder: test suite tesseract spoofers will break after any changes
+    # Reminder: test suite tesseract test plugins will break after any changes
     # to the number of order parameters here
     args_tesseract.extend([input_file, prefix, 'hocr', 'txt'] + tessconfig)
     try:
-        log.debug(args_tesseract)
-        p = run(
-            args_tesseract,
-            stdout=PIPE,
-            stderr=STDOUT,
-            timeout=timeout,
-            check=True,
-            env=tesseract_env,
-        )
+        p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
         stdout = p.stdout
     except TimeoutExpired:
         # Generate a HOCR file with no recognized text if tesseract times out
         # Temporary workaround to hocrTransform not being able to function if
         # it does not have a valid hOCR file.
-        page_timedout(log, input_file, timeout)
-        _generate_null_hocr(output_hocr, output_sidecar, input_file)
+        page_timedout(timeout)
+        _generate_null_hocr(output_hocr, output_text, input_file)
     except CalledProcessError as e:
-        tesseract_log_output(log, e.output, input_file)
+        tesseract_log_output(e.output)
         if b'Image too large' in e.output:
-            _generate_null_hocr(output_hocr, output_sidecar, input_file)
+            _generate_null_hocr(output_hocr, output_text, input_file)
             return
 
         raise SubprocessOutputError() from e
     else:
-        tesseract_log_output(log, stdout, input_file)
+        tesseract_log_output(stdout)
         # The sidecar text file will get the suffix .txt; rename it to
         # whatever caller wants it named
-        if os.path.exists(prefix + '.txt'):
-            shutil.move(prefix + '.txt', output_sidecar)
+        if prefix.with_suffix('.txt').exists():
+            shutil.move(prefix.with_suffix('.txt'), output_text)
 
 
-def use_skip_page(text_only, skip_pdf, output_pdf, output_text):
-    with open(output_text, 'w') as f:
-        f.write('[skipped page]')
+def use_skip_page(output_pdf, output_text):
+    output_text.write_text('[skipped page]', encoding='utf-8')
 
-    if skip_pdf and not text_only:
-        # Substitute a "skipped page"
-        with suppress(FileNotFoundError):
-            os.remove(output_pdf)  # In case it was partially created
-        os.symlink(skip_pdf, output_pdf)
-        return
-
-    # Or normally, just write a 0 byte file to the output to indicate a skip
-    with open(output_pdf, 'wb') as out:
-        out.write(b'')
+    # A 0 byte file to the output to indicate a skip
+    output_pdf.write_bytes(b'')
 
 
 def generate_pdf(
     *,
-    input_image,
-    skip_pdf=None,
-    output_pdf,
-    output_text,
-    language: list,
+    input_file: Path,
+    output_pdf: Path,
+    output_text: Path,
+    languages: List[str],
     engine_mode,
-    text_only: bool,
-    tessconfig: list,
+    tessconfig: List[str],
     timeout: float,
     pagesegmode: int,
     user_words,
     user_patterns,
-    tesseract_env,
-    log,
 ):
     """Use Tesseract to render a PDF.
 
-    input_image -- image to analyze
-    skip_pdf -- if we time out, use this file as output
+    input_file -- image to analyze
     output_pdf -- file to generate
     output_text -- OCR text file
-    language -- list of languages to consider
+    languages -- list of languages to consider
     engine_mode -- engine mode argument for tess v4
-    text_only -- enable tesseract text only mode?
     tessconfig -- tesseract configuration
     timeout -- timeout (seconds)
-    log -- logger object
     """
 
-    args_tesseract = tess_base_args(language, engine_mode)
+    args_tesseract = tess_base_args(languages, engine_mode)
 
     if pagesegmode is not None:
         args_tesseract.extend(['--psm', str(pagesegmode)])
 
-    if text_only and has_textonly_pdf(tesseract_env):
-        args_tesseract.extend(['-c', 'textonly_pdf=1'])
+    args_tesseract.extend(['-c', 'textonly_pdf=1'])
 
     if user_words:
         args_tesseract.extend(['--user-words', user_words])
@@ -360,31 +323,23 @@ def generate_pdf(
 
     prefix = os.path.splitext(output_pdf)[0]  # Tesseract appends suffixes
 
-    # Reminder: test suite tesseract spoofers might break after any changes
+    # Reminder: test suite tesseract test plugins might break after any changes
     # to the number of order parameters here
 
-    args_tesseract.extend([input_image, prefix, 'pdf', 'txt'] + tessconfig)
+    args_tesseract.extend([input_file, prefix, 'pdf', 'txt'] + tessconfig)
     try:
-        log.debug(args_tesseract)
-        p = run(
-            args_tesseract,
-            stdout=PIPE,
-            stderr=STDOUT,
-            timeout=timeout,
-            check=True,
-            env=tesseract_env,
-        )
+        p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
         stdout = p.stdout
         if os.path.exists(prefix + '.txt'):
             shutil.move(prefix + '.txt', output_text)
     except TimeoutExpired:
-        page_timedout(log, input_image, timeout)
-        use_skip_page(text_only, skip_pdf, output_pdf, output_text)
+        page_timedout(timeout)
+        use_skip_page(output_pdf, output_text)
     except CalledProcessError as e:
-        tesseract_log_output(log, e.output, input_image)
+        tesseract_log_output(e.output)
         if b'Image too large' in e.output:
-            use_skip_page(text_only, skip_pdf, output_pdf, output_text)
+            use_skip_page(output_pdf, output_text)
             return
         raise SubprocessOutputError() from e
     else:
-        tesseract_log_output(log, stdout, input_image)
+        tesseract_log_output(stdout)

src/ocrmypdf/_exec/unpaper.py

@@ -18,57 +18,68 @@
 # unpaper documentation:
 # https://github.com/Flameeyes/unpaper/blob/master/doc/basic-concepts.md
 
+"""Interface to unpaper executable"""
+
+import logging
 import os
 import shlex
-import subprocess
-import sys
-from functools import lru_cache
+from pathlib import Path
 from subprocess import PIPE, STDOUT, CalledProcessError
 from tempfile import TemporaryDirectory
-
-from . import get_version
-from ..exceptions import MissingDependencyError, SubprocessOutputError
+from typing import Tuple
 
 from PIL import Image
 
+from ocrmypdf.exceptions import MissingDependencyError, SubprocessOutputError
+from ocrmypdf.subprocess import get_version
+from ocrmypdf.subprocess import run as external_run
+
+log = logging.getLogger(__name__)
+
 
-@lru_cache(maxsize=1)
 def version():
     return get_version('unpaper')
 
 
-def run(input_file, output_file, dpi, log, mode_args):
-    args_unpaper = ['unpaper', '-v', '--dpi', str(dpi)] + mode_args
-
+def _setup_unpaper_io(tmpdir: Path, input_file: Path) -> Tuple[Path, Path]:
     SUFFIXES = {'1': '.pbm', 'L': '.pgm', 'RGB': '.ppm'}
-
-    im = Image.open(input_file)
-    if im.mode not in SUFFIXES.keys():
-        log.info("Converting image to other colorspace")
-        try:
-            if im.mode == 'P' and len(im.getcolors()) == 2:
-                im = im.convert(mode='1')
+    with Image.open(input_file) as im:
+        im_modified = False
+        if im.mode not in SUFFIXES:
+            log.info("Converting image to other colorspace")
+            try:
+                if im.mode == 'P' and len(im.getcolors()) == 2:
+                    im = im.convert(mode='1')
+                else:
+                    im = im.convert(mode='RGB')
+            except IOError as e:
+                raise MissingDependencyError(
+                    "Could not convert image with type " + im.mode
+                ) from e
             else:
-                im = im.convert(mode='RGB')
-        except IOError as e:
-            im.close()
+                im_modified = True
+        try:
+            suffix = SUFFIXES[im.mode]
+        except KeyError:
             raise MissingDependencyError(
-                "Could not convert image with type " + im.mode
+                "Failed to convert image to a supported format."
             ) from e
 
-    try:
-        suffix = SUFFIXES[im.mode]
-    except KeyError:
-        im.close()
-        raise MissingDependencyError(
-            "Failed to convert image to a supported format."
-        ) from e
+        if im_modified or input_file.suffix != '.png':
+            input_png = tmpdir / 'input.png'
+            im.save(input_png, format='PNG', compress_level=1)
+        else:
+            # No changes, PNG input, just use the file we already have
+            input_png = input_file
+        output_pnm = tmpdir / f'output{suffix}'
+    return input_png, output_pnm
+
+
+def run(input_file, output_file, dpi, mode_args):
+    args_unpaper = ['unpaper', '-v', '--dpi', str(dpi)] + mode_args
 
     with TemporaryDirectory() as tmpdir:
-        input_pnm = os.path.join(tmpdir, f'input{suffix}')
-        output_pnm = os.path.join(tmpdir, f'output{suffix}')
-        im.save(input_pnm, format='PPM')
-        im.close()
+        input_png, output_pnm = _setup_unpaper_io(Path(tmpdir), input_file)
 
         # To prevent any shenanigans from accepting arbitrary parameters in
         # --unpaper-args, we:
@@ -77,28 +88,29 @@ def run(input_file, output_file, dpi, log, mode_args):
         # 3) append absolute paths for the input and output file
         # This should ensure that a user cannot clobber some other file with
         # their unpaper arguments (whether intentionally or otherwise)
-        args_unpaper.extend([input_pnm, output_pnm])
+        args_unpaper.extend([os.fspath(input_png), os.fspath(output_pnm)])
         try:
-            proc = subprocess.run(
+            proc = external_run(
                 args_unpaper,
                 check=True,
                 close_fds=True,
                 universal_newlines=True,
-                stderr=STDOUT,
-                cwd=tmpdir,
+                stderr=STDOUT,  # unpaper writes logging output to stdout and stderr
+                cwd=tmpdir,  # and cannot send file output to stdout
                 stdout=PIPE,
             )
         except CalledProcessError as e:
-            log.debug(e.output)
+            log.debug(e.stderr)
             raise e from e
         else:
-            log.debug(proc.stdout)
-            # unpaper sets dpi to 72; fix this
+            log.debug(proc.stderr)
             try:
-                Image.open(output_pnm).save(output_file, dpi=(dpi, dpi))
+                with Image.open(output_pnm) as imout:
+                    imout.save(output_file, dpi=(dpi, dpi))
             except (FileNotFoundError, OSError):
                 raise SubprocessOutputError(
-                    "unpaper: failed to produce the expected output file. Called with: "
+                    "unpaper: failed to produce the expected output file. "
+                    + " Called with: "
                     + str(args_unpaper)
                 ) from None
 
@@ -110,7 +122,7 @@ def validate_custom_args(args: str):
     return unpaper_args
 
 
-def clean(input_file, output_file, dpi, log, unpaper_args=None):
+def clean(input_file, output_file, dpi, unpaper_args=None):
     default_args = [
         '--layout',
         'none',
@@ -124,4 +136,4 @@ def clean(input_file, output_file, dpi, log, unpaper_args=None):
     ]
     if not unpaper_args:
         unpaper_args = default_args
-    run(input_file, output_file, dpi, log, unpaper_args)
+    run(input_file, output_file, dpi, unpaper_args)

src/ocrmypdf/_graft.py

@@ -15,15 +15,15 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import os
+import logging
 from contextlib import suppress
 from pathlib import Path
+from typing import Optional
 
 import pikepdf
 
-from .exec import tesseract
-
-MAX_REPLACE_PAGES = int(os.environ.get('_OCRMYPDF_MAX_REPLACE_PAGES', 100))
+log = logging.getLogger(__name__)
+MAX_REPLACE_PAGES = 100
 
 
 def _update_page_resources(*, page, font, font_key, procset):
@@ -90,97 +90,10 @@ def strip_invisible_text(pdf, page):
     page.Contents = pikepdf.Stream(pdf, content_stream)
 
 
-def _graft_text_layer(
-    *, pdf_base, page_num, text, font, font_key, procset, rotation, strip_old_text, log
-):
-    """Insert the text layer from text page 0 on to pdf_base at page_num"""
-
-    log.debug("Grafting")
-    if Path(text).stat().st_size == 0:
-        return
-
-    # This is a pointer indicating a specific page in the base file
-    pdf_text = pikepdf.open(text)
-    pdf_text_contents = pdf_text.pages[0].Contents.read_bytes()
-
-    base_page = pdf_base.pages.p(page_num)
-
-    # The text page always will be oriented up by this stage but the original
-    # content may have a rotation applied. Wrap the text stream with a rotation
-    # so it will be oriented the same way as the rest of the page content.
-    # (Previous versions OCRmyPDF rotated the content layer to match the text.)
-    mediabox = [float(pdf_text.pages[0].MediaBox[v]) for v in range(4)]
-    wt, ht = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
-
-    mediabox = [float(base_page.MediaBox[v]) for v in range(4)]
-    wp, hp = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
-
-    translate = pikepdf.PdfMatrix().translated(-wt / 2, -ht / 2)
-    untranslate = pikepdf.PdfMatrix().translated(wp / 2, hp / 2)
-    # -rotation because the input is a clockwise angle and this formula
-    # uses CCW
-    rotation = -rotation % 360
-    rotate = pikepdf.PdfMatrix().rotated(rotation)
-
-    # Because of rounding of DPI, we might get a text layer that is not
-    # identically sized to the target page. Scale to adjust. Normally this
-    # is within 0.998.
-    if rotation in (90, 270):
-        wt, ht = ht, wt
-    scale_x = wp / wt
-    scale_y = hp / ht
-
-    # log.debug('%r', scale_x, scale_y)
-    scale = pikepdf.PdfMatrix().scaled(scale_x, scale_y)
-
-    # Translate the text so it is centered at (0, 0), rotate it there, adjust
-    # for a size different between initial and text PDF, then untranslate
-    ctm = translate @ rotate @ scale @ untranslate
-
-    pdf_text_contents = b'q %s cm\n' % ctm.encode() + pdf_text_contents + b'\nQ\n'
-
-    new_text_layer = pikepdf.Stream(pdf_base, pdf_text_contents)
-
-    if strip_old_text:
-        strip_invisible_text(pdf_base, base_page)
-
-    base_page.page_contents_add(new_text_layer, prepend=True)
-
-    _update_page_resources(
-        page=base_page, font=font, font_key=font_key, procset=procset
-    )
-    pdf_text.close()
-
-
-def _find_font(text, pdf_base):
-    """Copy a font from the filename text into pdf_base"""
-
-    font, font_key = None, None
-    possible_font_names = ('/f-0-0', '/F1')
-    try:
-        with pikepdf.open(text) as pdf_text:
-            try:
-                pdf_text_fonts = pdf_text.pages[0].Resources.get('/Font', {})
-            except (AttributeError, IndexError, KeyError):
-                return None, None
-            for f in possible_font_names:
-                pdf_text_font = pdf_text_fonts.get(f, None)
-                if pdf_text_font is not None:
-                    font_key = f
-                    break
-            if pdf_text_font:
-                font = pdf_base.copy_foreign(pdf_text_font)
-            return font, font_key
-    except (FileNotFoundError, pikepdf.PdfError):
-        # PdfError occurs if a 0-length file is written e.g. due to OCR timeout
-        return None, None
-
-
 class OcrGrafter:
     def __init__(self, context):
         self.context = context
-        self.log = context.log
-        self.path_base = Path(context.origin).resolve()
+        self.path_base = context.origin
 
         self.pdf_base = pikepdf.open(self.path_base)
         self.font, self.font_key = None, None
@@ -195,10 +108,16 @@ class OcrGrafter:
         self.emplacements = 1
         self.interim_count = 0
 
-    def graft_page(self, page_result):
-        pageno, image, text, _sidecar, autorotate_correction = page_result
-        if text and not self.font:
-            self.font, self.font_key = _find_font(text, self.pdf_base)
+    def graft_page(
+        self,
+        *,
+        pageno: int,
+        image: Optional[Path],
+        textpdf: Optional[Path],
+        autorotate_correction: int,
+    ):
+        if textpdf and not self.font:
+            self.font, self.font_key = self._find_font(textpdf)
 
         emplaced_page = False
         content_rotation = self.pdfinfo[pageno].rotation
@@ -206,7 +125,7 @@ class OcrGrafter:
         if path_image is not None and path_image != self.path_base:
             # We are updating the old page with a rasterized PDF of the new
             # page (without changing objgen, to preserve references)
-            self.log.debug("Emplacement update")
+            log.debug("Emplacement update")
             with pikepdf.open(image) as pdf_image:
                 self.emplacements += 1
                 foreign_image_page = pdf_image.pages[0]
@@ -220,24 +139,23 @@ class OcrGrafter:
             content_rotation = autorotate_correction
         text_rotation = autorotate_correction
         text_misaligned = (text_rotation - content_rotation) % 360
-        self.log.debug(
-            '%r',
-            [text_rotation, autorotate_correction, text_misaligned, content_rotation],
+        log.debug(
+            f"Rotations for page {pageno}: [text, auto, misalign, content] = "
+            f"{text_rotation}, {autorotate_correction}, "
+            f"{text_misaligned}, {content_rotation}"
         )
 
-        if text and self.font:
+        if textpdf and self.font:
             # Graft the text layer onto this page, whether new or old
             strip_old = self.context.options.redo_ocr
-            _graft_text_layer(
-                pdf_base=self.pdf_base,
+            self._graft_text_layer(
                 page_num=pageno + 1,
-                text=text,
+                textpdf=textpdf,
                 font=self.font,
                 font_key=self.font_key,
                 rotation=text_misaligned,
                 procset=self.procset,
                 strip_old_text=strip_old,
-                log=self.log,
             )
 
         # Correct the rotation if applicable
@@ -249,10 +167,13 @@ class OcrGrafter:
             self.save_and_reload()
 
     def save_and_reload(self):
-        # Periodically save and reload the Pdf object. This will keep a
-        # lid on our memory usage for very large files. Attach the font to
-        # page 1 even if page 1 doesn't use it, so we have a way to get it
-        # back.
+        """Save and reload the Pdf.
+
+        This will keep a lid on our memory usage for very large files. Attach
+        the font to page 1 even if page 1 doesn't use it, so we have a way to get it
+        back.
+        """
+
         page0 = self.pdf_base.pages[0]
         _update_page_resources(
             page=page0, font=self.font, font_key=self.font_key, procset=self.procset
@@ -263,12 +184,14 @@ class OcrGrafter:
         # {interim_count} is the opened file we were updateing
         # {interim_count - 1} can be deleted
         # {interim_count + 1} is the new file will produce and open
-        old_file = self.output_file + f'_working{self.interim_count - 1}.pdf'
+        old_file = self.output_file.with_suffix(f'.working{self.interim_count - 1}.pdf')
         if not self.context.options.keep_temporary_files:
             with suppress(FileNotFoundError):
-                os.unlink(old_file)
+                old_file.unlink()
 
-        next_file = self.output_file + f'_working{self.interim_count + 1}.pdf'
+        next_file = self.output_file.with_suffix(
+            f'.working{self.interim_count + 1}.pdf'
+        )
         self.pdf_base.save(next_file)
         self.pdf_base.close()
 
@@ -281,3 +204,98 @@ class OcrGrafter:
         self.pdf_base.save(self.output_file)
         self.pdf_base.close()
         return self.output_file
+
+    def _find_font(self, text):
+        """Copy a font from the filename text into pdf_base"""
+
+        font, font_key = None, None
+        possible_font_names = ('/f-0-0', '/F1')
+        try:
+            with pikepdf.open(text) as pdf_text:
+                try:
+                    pdf_text_fonts = pdf_text.pages[0].Resources.get('/Font', {})
+                except (AttributeError, IndexError, KeyError):
+                    return None, None
+                for f in possible_font_names:
+                    pdf_text_font = pdf_text_fonts.get(f, None)
+                    if pdf_text_font is not None:
+                        font_key = f
+                        break
+                if pdf_text_font:
+                    font = self.pdf_base.copy_foreign(pdf_text_font)
+                return font, font_key
+        except (FileNotFoundError, pikepdf.PdfError):
+            # PdfError occurs if a 0-length file is written e.g. due to OCR timeout
+            return None, None
+
+    def _graft_text_layer(
+        self,
+        *,
+        page_num: int,
+        textpdf: Path,
+        font: pikepdf.Object,
+        font_key: pikepdf.Object,
+        procset: pikepdf.Object,
+        rotation: int,
+        strip_old_text: bool,
+    ):
+        """Insert the text layer from text page 0 on to pdf_base at page_num"""
+
+        log.debug("Grafting")
+        if Path(textpdf).stat().st_size == 0:
+            return
+
+        # This is a pointer indicating a specific page in the base file
+        with pikepdf.open(textpdf) as pdf_text:
+            pdf_text_contents = pdf_text.pages[0].Contents.read_bytes()
+
+            base_page = self.pdf_base.pages.p(page_num)
+
+            # The text page always will be oriented up by this stage but the original
+            # content may have a rotation applied. Wrap the text stream with a rotation
+            # so it will be oriented the same way as the rest of the page content.
+            # (Previous versions OCRmyPDF rotated the content layer to match the text.)
+            mediabox = [float(pdf_text.pages[0].MediaBox[v]) for v in range(4)]
+            wt, ht = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
+
+            mediabox = [float(base_page.MediaBox[v]) for v in range(4)]
+            wp, hp = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
+
+            translate = pikepdf.PdfMatrix().translated(-wt / 2, -ht / 2)
+            untranslate = pikepdf.PdfMatrix().translated(wp / 2, hp / 2)
+            corner = pikepdf.PdfMatrix().translated(mediabox[0], mediabox[1])
+            # -rotation because the input is a clockwise angle and this formula
+            # uses CCW
+            rotation = -rotation % 360
+            rotate = pikepdf.PdfMatrix().rotated(rotation)
+
+            # Because of rounding of DPI, we might get a text layer that is not
+            # identically sized to the target page. Scale to adjust. Normally this
+            # is within 0.998.
+            if rotation in (90, 270):
+                wt, ht = ht, wt
+            scale_x = wp / wt
+            scale_y = hp / ht
+
+            # log.debug('%r', scale_x, scale_y)
+            scale = pikepdf.PdfMatrix().scaled(scale_x, scale_y)
+
+            # Translate the text so it is centered at (0, 0), rotate it there, adjust
+            # for a size different between initial and text PDF, then untranslate, and
+            # finally move the lower left corner to match the mediabox
+            ctm = translate @ rotate @ scale @ untranslate @ corner
+
+            pdf_text_contents = (
+                b'q %s cm\n' % ctm.encode() + pdf_text_contents + b'\nQ\n'
+            )
+
+            new_text_layer = pikepdf.Stream(self.pdf_base, pdf_text_contents)
+
+            if strip_old_text:
+                strip_invisible_text(self.pdf_base, base_page)
+
+            base_page.page_contents_add(new_text_layer, prepend=True)
+
+            _update_page_resources(
+                page=base_page, font=font, font_key=font_key, procset=procset
+            )

src/ocrmypdf/_jobcontext.py

@@ -15,40 +15,25 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import logging
+import os
 import shutil
 import sys
-import os
+from pathlib import Path
+
+from ocrmypdf._plugin_manager import get_plugin_manager
 
 
-class PicklableLoggerMixin:
-    def __init__(self):
-        self._log = None
-
-    @property
-    def log(self):
-        if not self._log:
-            self._log = self.get_logger()
-        return self._log
-
-    def __getstate__(self):
-        # Python 3.6 is incapable of pickling a logger and marshalling it to another
-        # process (threading._RLock error), so we disconnect it before pickling,
-        # and create a new logger in the worker process.
-        state = self.__dict__.copy()
-        state['_log'] = None
-        return state
-
-
-class PDFContext(PicklableLoggerMixin):
+class PdfContext:
     """Holds our context for a particular run of the pipeline"""
 
-    def __init__(self, options, work_folder, origin, pdfinfo):
-        PicklableLoggerMixin.__init__(self)
+    def __init__(
+        self, options, work_folder: Path, origin: Path, pdfinfo, plugin_manager
+    ):
         self.options = options
-        self.work_folder = work_folder
-        self.origin = origin
+        self.work_folder = Path(work_folder)
+        self.origin = Path(origin)
         self.pdfinfo = pdfinfo
+        self.plugin_manager = plugin_manager
         if options:
             self.name = os.path.basename(options.input_file)
         else:
@@ -56,11 +41,8 @@ class PDFContext(PicklableLoggerMixin):
         if self.name == '-':
             self.name = 'stdin'
 
-    def get_logger(self):
-        return make_logger(self.options, filename=self.name)
-
-    def get_path(self, name):
-        return os.path.join(self.work_folder, name)
+    def get_path(self, name: str) -> Path:
+        return self.work_folder / name
 
     def get_page_contexts(self):
         npages = len(self.pdfinfo)
@@ -68,27 +50,33 @@ class PDFContext(PicklableLoggerMixin):
             yield PageContext(self, n)
 
 
-class PageContext(PicklableLoggerMixin):
+class PageContext:
     """Holds our context for a page
 
     Must be pickable, so only store intrinsic/simple data elements
     """
 
-    def __init__(self, pdf_context, pageno):
-        PicklableLoggerMixin.__init__(self)
+    def __init__(self, pdf_context: PdfContext, pageno):
         self.work_folder = pdf_context.work_folder
         self.origin = pdf_context.origin
         self.options = pdf_context.options
         self.name = pdf_context.name
         self.pageno = pageno
         self.pageinfo = pdf_context.pdfinfo[pageno]
-        self._log = None
+        self.plugin_manager = pdf_context.plugin_manager
 
-    def get_logger(self):
-        return make_logger(self.options, filename=self.name, page=self.pageno + 1)
+    def get_path(self, name: str) -> Path:
+        return self.work_folder / ("%06d_%s" % (self.pageno + 1, name))
 
-    def get_path(self, name):
-        return os.path.join(self.work_folder, "%06d_%s" % (self.pageno + 1, name))
+    def __getstate__(self):
+        state = self.__dict__.copy()
+        if state['plugin_manager'] is not None:
+            state['plugin_manager'] = None
+        return state
+
+    def __setstate__(self, state):
+        self.__dict__.update(state)
+        self.plugin_manager = get_plugin_manager(self.options.plugins)
 
 
 def cleanup_working_files(work_folder, options):
@@ -96,29 +84,3 @@ def cleanup_working_files(work_folder, options):
         print(f"Temporary working files retained at:\n{work_folder}", file=sys.stderr)
     else:
         shutil.rmtree(work_folder, ignore_errors=True)
-
-
-class LogNameAdapter(logging.LoggerAdapter):
-    def process(self, msg, kwargs):
-        # return '[%s] %s' % (self.extra['filename'], msg), kwargs
-        return '%s' % (msg,), kwargs
-
-
-class LogNamePageAdapter(logging.LoggerAdapter):
-    def process(self, msg, kwargs):
-        return (
-            #'[%s:%05u] %s' % (self.extra['filename'], self.extra['page'], msg),
-            '%4u: %s' % (self.extra['page'], msg),
-            kwargs,
-        )
-
-
-def make_logger(options=None, prefix='ocrmypdf', filename=None, page=None):
-    log = logging.getLogger(prefix)
-    if filename and page:
-        adapter = LogNamePageAdapter(log, dict(filename=filename, page=page))
-    elif filename:
-        adapter = LogNameAdapter(log, dict(filename=filename))
-    else:
-        adapter = log
-    return adapter

src/ocrmypdf/_logging.py

@@ -0,0 +1,60 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+import sys
+from contextlib import suppress
+
+from tqdm import tqdm
+
+
+class PageNumberFilter(logging.Filter):
+    def filter(self, record):
+        pageno = getattr(record, 'pageno', None)
+        if isinstance(pageno, int):
+            record.pageno = f'{pageno:5d} '
+        elif pageno is None:
+            record.pageno = ''
+        return True
+
+
+class TqdmConsole:
+    """Wrapper to log messages in a way that is compatible with tqdm progress bar
+
+    This routes log messages through tqdm so that it can print them above the
+    progress bar, and then refresh the progress bar, rather than overwriting
+    it which looks messy.
+
+    For some reason Python 3.6 prints extra empty messages from time to time,
+    so we suppress those.
+    """
+
+    def __init__(self, file):
+        self.file = file
+        self.py36 = sys.version_info[0:2] == (3, 6)
+
+    def write(self, msg):
+        # When no progress bar is active, tqdm.write() routes to print()
+        if self.py36:
+            if msg.strip() != '':
+                tqdm.write(msg.rstrip(), end='\n', file=self.file)
+        else:
+            tqdm.write(msg.rstrip(), end='\n', file=self.file)
+
+    def flush(self):
+        with suppress(AttributeError):
+            self.file.flush()

src/ocrmypdf/_pipeline.py

@@ -15,53 +15,58 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
+import logging
 import os
 import re
 import sys
 from datetime import datetime, timezone
+from pathlib import Path
 from shutil import copyfileobj
+from typing import Dict, Iterable, Optional
 
 import img2pdf
-from PIL import Image
-
 import pikepdf
 from pikepdf.models.metadata import encode_pdf_date
+from PIL import Image, ImageColor, ImageDraw
 
-from . import PROGRAM_NAME, VERSION, leptonica
-
-from ._plugins import load_plugin
-from .exceptions import (
+from ocrmypdf import leptonica
+from ocrmypdf._exec import unpaper
+from ocrmypdf._jobcontext import PageContext, PdfContext
+from ocrmypdf._version import PROGRAM_NAME
+from ocrmypdf._version import __version__ as VERSION
+from ocrmypdf.exceptions import (
     DpiError,
     EncryptedPdfError,
     InputFileError,
-    UnsupportedImageFormatError,
     PriorOcrFoundError,
+    UnsupportedImageFormatError,
 )
-from .exec import ghostscript, tesseract
-from .helpers import re_symlink
-from .hocrtransform import HocrTransform
-from .optimize import optimize
-from .pdfa import generate_pdfa_ps
-from .pdfinfo import Colorspace, PdfInfo
+from ocrmypdf.helpers import Resolution, safe_symlink
+from ocrmypdf.hocrtransform import HocrTransform
+from ocrmypdf.optimize import optimize
+from ocrmypdf.pdfa import generate_pdfa_ps
+from ocrmypdf.pdfinfo import Colorspace, Encoding, PdfInfo
+
+log = logging.getLogger(__name__)
 
 VECTOR_PAGE_DPI = 400
 
 
-def triage_image_file(input_file, output_file, options, log):
+def triage_image_file(input_file, output_file, options):
+    log.info("Input file is not a PDF, checking if it is an image...")
     try:
-        log.info("Input file is not a PDF, checking if it is an image...")
         im = Image.open(input_file)
     except EnvironmentError as e:
         # Recover the original filename
-        log.error(str(e).replace(input_file, options.input_file))
+        log.error(str(e).replace(str(input_file), str(options.input_file)))
         raise UnsupportedImageFormatError() from e
-    else:
-        log.info("Input file is an image")
 
+    with im:
+        log.info("Input file is an image")
         if 'dpi' in im.info:
             if im.info['dpi'] <= (96, 96) and not options.image_dpi:
-                log.info("Image size: (%d, %d)" % im.size)
-                log.info("Image resolution: (%d, %d)" % im.info['dpi'])
+                log.info("Image size: (%d, %d)", *im.size)
+                log.info("Image resolution: (%d, %d)", *im.info['dpi'])
                 log.error(
                     "Input file is an image, but the resolution (DPI) is "
                     "not credible.  Estimate the resolution at which the "
@@ -69,7 +74,7 @@ def triage_image_file(input_file, output_file, options, log):
                 )
                 raise DpiError()
         elif not options.image_dpi:
-            log.info("Image size: (%d, %d)" % im.size)
+            log.info("Image size: (%d, %d)", *im.size)
             log.error(
                 "Input file is an image, but has no resolution (DPI) "
                 "in its metadata.  Estimate the resolution at which "
@@ -86,22 +91,24 @@ def triage_image_file(input_file, output_file, options, log):
 
         if 'iccprofile' not in im.info:
             if im.mode == 'RGB':
-                log.info('Input image has no ICC profile, assuming sRGB')
+                log.info("Input image has no ICC profile, assuming sRGB")
             elif im.mode == 'CMYK':
-                log.info('Input CMYK image has no ICC profile, not usable')
+                log.error("Input CMYK image has no ICC profile, not usable")
                 raise UnsupportedImageFormatError()
-        im.close()
 
     try:
         log.info("Image seems valid. Try converting to PDF...")
         layout_fun = img2pdf.default_layout_fun
         if options.image_dpi:
             layout_fun = img2pdf.get_fixed_dpi_layout_fun(
-                (options.image_dpi, options.image_dpi)
+                Resolution(options.image_dpi, options.image_dpi)
             )
         with open(output_file, 'wb') as outf:
             img2pdf.convert(
-                input_file, layout_fun=layout_fun, with_pdfrw=False, outputstream=outf
+                os.fspath(input_file),
+                layout_fun=layout_fun,
+                with_pdfrw=False,
+                outputstream=outf,
             )
         log.info("Successfully converted to PDF, processing...")
     except img2pdf.ImageOpenError as e:
@@ -125,29 +132,40 @@ def _pdf_guess_version(input_file, search_window=1024):
     return ''
 
 
-def triage(input_file, output_file, options, log):
+def triage(original_filename, input_file, output_file, options):
     try:
         if _pdf_guess_version(input_file):
             if options.image_dpi:
                 log.warning(
-                    "Argument --image-dpi ignored because the "
+                    "Argument --image-dpi is being ignored because the "
                     "input file is a PDF, not an image."
                 )
             # Origin file is a pdf create a symlink with pdf extension
-            re_symlink(input_file, output_file)
+            safe_symlink(input_file, output_file)
             return output_file
     except EnvironmentError as e:
-        log.error(e)
-        raise InputFileError() from e
+        log.debug(f"Temporary file was at: {input_file}")
+        msg = str(e).replace(str(input_file), original_filename)
+        raise InputFileError(msg) from e
 
-    triage_image_file(input_file, output_file, options, log)
+    triage_image_file(input_file, output_file, options)
     return output_file
 
 
-def get_pdfinfo(input_file, detailed_page_analysis=False, progbar=False):
+def get_pdfinfo(
+    input_file,
+    detailed_analysis=False,
+    progbar=False,
+    max_workers=None,
+    check_pages=None,
+):
     try:
         return PdfInfo(
-            input_file, detailed_page_analysis=detailed_page_analysis, progbar=progbar
+            input_file,
+            detailed_analysis=detailed_analysis,
+            progbar=progbar,
+            max_workers=max_workers,
+            check_pages=check_pages,
         )
     except pikepdf.PasswordError:
         raise EncryptedPdfError()
@@ -155,17 +173,10 @@ def get_pdfinfo(input_file, detailed_page_analysis=False, progbar=False):
         raise InputFileError()
 
 
-def validate_pdfinfo_options(context):
-    log = context.log
+def validate_pdfinfo_options(context: PdfContext):
     pdfinfo = context.pdfinfo
     options = context.options
 
-    if options.plugin_validation:
-        validate = load_plugin(options.plugin_validation)
-        result = validate(context)
-        if result is not None:
-            return result
-
     if pdfinfo.needs_rendering:
         log.error(
             "This PDF contains dynamic XFA forms created by Adobe LiveCycle "
@@ -189,7 +200,7 @@ def validate_pdfinfo_options(context):
             )
             raise InputFileError()
         else:
-            log.warn(
+            log.warning(
                 "This PDF has a fillable form. "
                 "Chances are it is a pure digital "
                 "document that does not need OCR."
@@ -200,54 +211,56 @@ def validate_pdfinfo_options(context):
                     "form and all filled form fields. The output PDF will be "
                     "'flattened' and will no longer be fillable."
                 )
+    context.plugin_manager.hook.validate(pdfinfo=pdfinfo, options=options)
 
 
 def get_page_dpi(pageinfo, options):
     "Get the DPI when nonsquare DPI is tolerable"
     xres = max(
-        pageinfo.xres or VECTOR_PAGE_DPI,
-        options.oversample or 0,
-        VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
+        pageinfo.dpi.x or VECTOR_PAGE_DPI,
+        options.oversample or 0.0,
+        VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
     )
     yres = max(
-        pageinfo.yres or VECTOR_PAGE_DPI,
+        pageinfo.dpi.y or VECTOR_PAGE_DPI,
         options.oversample or 0,
-        VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
+        VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
     )
-    return (float(xres), float(yres))
+    return Resolution(float(xres), float(yres))
 
 
-def get_page_square_dpi(pageinfo, options):
+def get_page_square_dpi(pageinfo, options) -> Resolution:
     "Get the DPI when we require xres == yres, scaled to physical units"
-    xres = pageinfo.xres or 0
-    yres = pageinfo.yres or 0
-    userunit = pageinfo.userunit or 1
-    return float(
+    xres = pageinfo.dpi.x or 0.0
+    yres = pageinfo.dpi.y or 0.0
+    userunit = float(pageinfo.userunit) or 1.0
+    units = float(
         max(
             (xres * userunit) or VECTOR_PAGE_DPI,
             (yres * userunit) or VECTOR_PAGE_DPI,
-            VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
-            options.oversample or 0,
+            VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
+            options.oversample or 0.0,
         )
     )
+    return Resolution(units, units)
 
 
-def get_canvas_square_dpi(pageinfo, options):
+def get_canvas_square_dpi(pageinfo, options) -> Resolution:
     """Get the DPI when we require xres == yres, in Postscript units"""
-    return float(
+    units = float(
         max(
-            (pageinfo.xres) or VECTOR_PAGE_DPI,
-            (pageinfo.yres) or VECTOR_PAGE_DPI,
-            VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
-            options.oversample or 0,
+            (pageinfo.dpi.x) or VECTOR_PAGE_DPI,
+            (pageinfo.dpi.y) or VECTOR_PAGE_DPI,
+            VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
+            options.oversample or 0.0,
         )
     )
+    return Resolution(units, units)
 
 
-def is_ocr_required(page_context):
+def is_ocr_required(page_context: PageContext):
     pageinfo = page_context.pageinfo
     options = page_context.options
-    log = page_context.log
 
     ocr_required = True
 
@@ -257,14 +270,15 @@ def is_ocr_required(page_context):
     elif pageinfo.has_text:
         if not options.force_ocr and not (options.skip_text or options.redo_ocr):
             raise PriorOcrFoundError(
-                "page already has text! - aborting (use --force-ocr to force OCR)"
+                "page already has text! - aborting (use --force-ocr to force OCR; "
+                " see also help for the arguments --skip-text and --redo-ocr"
             )
         elif options.force_ocr:
             log.info("page already has text! - rasterizing text and running OCR anyway")
             ocr_required = True
         elif options.redo_ocr:
             if pageinfo.has_corrupt_text:
-                log.warn(
+                log.warning(
                     "some text on this page cannot be mapped to characters: "
                     "consider using --force-ocr instead"
                 )
@@ -291,7 +305,7 @@ def is_ocr_required(page_context):
             )
         elif options.force_ocr:
             # Warn the user they might not want to do this
-            log.warn(
+            log.warning(
                 "page has no images - "
                 "all vector content will be "
                 f"rasterized at {VECTOR_PAGE_DPI} DPI, losing some resolution and likely "
@@ -311,99 +325,94 @@ def is_ocr_required(page_context):
         pixel_count = pageinfo.width_pixels * pageinfo.height_pixels
         if pixel_count > (options.skip_big * 1_000_000):
             ocr_required = False
-            log.warn(
+            log.warning(
                 "page too big, skipping OCR "
                 f"({(pixel_count / 1_000_000):.1f} MPixels > {options.skip_big:.1f} MPixels --skip-big)"
             )
     return ocr_required
 
 
-def rasterize_preview(input_file, page_context):
+def rasterize_preview(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('rasterize_preview.jpg')
     canvas_dpi = get_canvas_square_dpi(page_context.pageinfo, page_context.options)
     page_dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    ghostscript.rasterize_pdf(
-        input_file,
-        output_file,
-        xres=canvas_dpi,
-        yres=canvas_dpi,
+    page_context.plugin_manager.hook.rasterize_pdf_page(
+        input_file=input_file,
+        output_file=output_file,
         raster_device='jpeggray',
-        log=page_context.log,
-        page_dpi=(page_dpi, page_dpi),
+        raster_dpi=canvas_dpi,
+        page_dpi=page_dpi,
         pageno=page_context.pageinfo.pageno + 1,
     )
     return output_file
 
 
-def get_orientation_correction(preview, page_context):
+def describe_rotation(page_context: PageContext, orient_conf, correction: int):
     """
-    Work out orientation correct for each page.
+    Describe the page rotation we are going to perform.
+    """
+    direction = {0: '⇧', 90: '⇨', 180: '⇩', 270: '⇦'}
+    turns = {0: ' ', 90: '⬏', 180: '↻', 270: '⬑'}
+
+    existing_rotation = page_context.pageinfo.rotation
+    action = ''
+    if orient_conf.confidence >= page_context.options.rotate_pages_threshold:
+        if correction != 0:
+            action = 'will rotate ' + turns[correction]
+        else:
+            action = 'rotation appears correct'
+    else:
+        if correction != 0:
+            action = 'confidence too low to rotate'
+        else:
+            action = 'no change'
+
+    facing = ''
+
+    if existing_rotation != 0:
+        facing = f"with existing rotation {direction.get(existing_rotation, '?')}, "
+    facing += f"page is facing {direction.get(orient_conf.angle, '?')}"
+
+    return f"{facing}, confidence {orient_conf.confidence:.2f} - {action}"
+
+
+def get_orientation_correction(preview: Path, page_context: PageContext):
+    """Work out orientation correct for each page.
 
     We ask Ghostscript to draw a preview page, which will rasterize with the
-    current /Rotate applied, and then ask Tesseract which way the page is
+    current /Rotate applied, and then ask OCR which way the page is
     oriented. If the value of /Rotate is correct (e.g., a user already
-    manually fixed rotation), then Tesseract will say the page is pointing
+    manually fixed rotation), then OCR will say the page is pointing
     up and the correction is zero. Otherwise, the orientation found by
-    Tesseract represents the clockwise rotation, or the counterclockwise
+    OCR represents the clockwise rotation, or the counterclockwise
     correction to rotation.
 
     When we draw the real page for OCR, we rotate it by the CCW correction,
     which points it (hopefully) upright. _graft.py takes care of the orienting
     the image and text layers.
-
     """
 
-    orient_conf = tesseract.get_orientation(
-        preview,
-        engine_mode=page_context.options.tesseract_oem,
-        timeout=page_context.options.tesseract_timeout,
-        log=page_context.log,
-        tesseract_env=page_context.options.tesseract_env,
+    orient_conf = page_context.plugin_manager.hook.get_ocr_engine().get_orientation(
+        preview, page_context.options
     )
 
-    direction = {0: '⇧', 90: '⇨', 180: '⇩', 270: '⇦'}
-
-    existing_rotation = page_context.pageinfo.rotation
-
     correction = orient_conf.angle % 360
-
-    apply_correction = False
-    action = ''
-    if orient_conf.confidence >= page_context.options.rotate_pages_threshold:
-        if correction != 0:
-            apply_correction = True
-            action = ' - will rotate'
-        else:
-            action = ' - rotation appears correct'
-    else:
-        if correction != 0:
-            action = ' - confidence too low to rotate'
-        else:
-            action = ' - no change'
-
-    facing = ''
-    if existing_rotation != 0:
-        facing = 'with existing rotation {}, '.format(
-            direction.get(existing_rotation, '?')
-        )
-    facing += 'page is facing {}'.format(direction.get(orient_conf.angle, '?'))
-
-    page_context.log.debug(
-        '{pagenum:4d}: {facing}, confidence {conf:.2f}{action}'.format(
-            pagenum=page_context.pageinfo.pageno,
-            facing=facing,
-            conf=orient_conf.confidence,
-            action=action,
-        )
-    )
-
-    if apply_correction:
+    log.info(describe_rotation(page_context, orient_conf, correction))
+    if (
+        orient_conf.confidence >= page_context.options.rotate_pages_threshold
+        and correction != 0
+    ):
         return correction
+
     return 0
 
 
 def rasterize(
-    input_file, page_context, correction=0, output_tag='', remove_vectors=None
+    input_file: Path,
+    page_context: PageContext,
+    correction: int = 0,
+    output_tag: str = '',
+    remove_vectors=None,
 ):
     colorspaces = ['pngmono', 'pnggray', 'png256', 'png16m']
     device_idx = 0
@@ -433,21 +442,19 @@ def rasterize(
 
     device = colorspaces[device_idx]
 
-    page_context.log.debug(f"Rasterize with {device}")
+    log.debug(f"Rasterize with {device}")
 
     # Produce the page image with square resolution or else deskew and OCR
     # will not work properly.
     canvas_dpi = get_canvas_square_dpi(pageinfo, page_context.options)
     page_dpi = get_page_square_dpi(pageinfo, page_context.options)
 
-    ghostscript.rasterize_pdf(
-        input_file,
-        output_file,
-        xres=canvas_dpi,
-        yres=canvas_dpi,
+    page_context.plugin_manager.hook.rasterize_pdf_page(
+        input_file=input_file,
+        output_file=output_file,
         raster_device=device,
-        log=page_context.log,
-        page_dpi=(page_dpi, page_dpi),
+        raster_dpi=canvas_dpi,
+        page_dpi=page_dpi,
         pageno=pageinfo.pageno + 1,
         rotation=correction,
         filter_vector=remove_vectors,
@@ -455,39 +462,31 @@ def rasterize(
     return output_file
 
 
-def preprocess_remove_background(input_file, page_context):
+def preprocess_remove_background(input_file: Path, page_context: PageContext):
     if any(image.bpc > 1 for image in page_context.pageinfo.images):
         output_file = page_context.get_path('pp_rm_bg.png')
         leptonica.remove_background(input_file, output_file)
         return output_file
     else:
-        page_context.log.info("background removal skipped on mono page")
+        log.info("background removal skipped on mono page")
         return input_file
 
 
-def preprocess_deskew(input_file, page_context):
+def preprocess_deskew(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('pp_deskew.png')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    leptonica.deskew(input_file, output_file, dpi)
+    leptonica.deskew(input_file, output_file, dpi.x)
     return output_file
 
 
-def preprocess_clean(input_file, page_context):
-    from .exec import unpaper
-
+def preprocess_clean(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('pp_clean.png')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    unpaper.clean(
-        input_file,
-        output_file,
-        dpi,
-        page_context.log,
-        page_context.options.unpaper_args,
-    )
+    unpaper.clean(input_file, output_file, dpi.x, page_context.options.unpaper_args)
     return output_file
 
 
-def create_ocr_image(image, page_context):
+def create_ocr_image(image: Path, page_context: PageContext):
     """Create the image we send for OCR. May not be the same as the display
     image depending on preprocessing. This image will never be shown to the
     user."""
@@ -495,15 +494,11 @@ def create_ocr_image(image, page_context):
     output_file = page_context.get_path('ocr.png')
     options = page_context.options
     with Image.open(image) as im:
-        from PIL import ImageColor
-        from PIL import ImageDraw
-
         white = ImageColor.getcolor('#ffffff', im.mode)
         # pink = ImageColor.getcolor('#ff0080', im.mode)
         draw = ImageDraw.ImageDraw(im)
 
-        xres, yres = im.info['dpi']
-        page_context.log.debug('resolution %r %r' % (xres, yres))
+        log.debug('resolution %r', im.info['dpi'])
 
         if not options.force_ocr:
             # Do not mask text areas when forcing OCR, because we need to OCR
@@ -519,15 +514,15 @@ def create_ocr_image(image, page_context):
                 # without regard whatever resolution is in pageinfo (may differ or
                 # be None)
                 bbox = [float(v) for v in textarea]
-                xscale, yscale = float(xres) / 72.0, float(yres) / 72.0
+                xyscale = tuple(float(coord) / 72.0 for coord in im.info['dpi'])
                 pixcoords = [
-                    bbox[0] * xscale,
-                    im.height - bbox[3] * yscale,
-                    bbox[2] * xscale,
-                    im.height - bbox[1] * yscale,
+                    bbox[0] * xyscale[0],
+                    im.height - bbox[3] * xyscale[1],
+                    bbox[2] * xyscale[0],
+                    im.height - bbox[1] * xyscale[1],
                 ]
                 pixcoords = [int(round(c)) for c in pixcoords]
-                page_context.log.debug('blanking %r', pixcoords)
+                log.debug('blanking %r', pixcoords)
                 draw.rectangle(pixcoords, fill=white)
                 # draw.rectangle(pixcoords, outline=pink)
 
@@ -536,60 +531,60 @@ def create_ocr_image(image, page_context):
             pix = pix.masked_threshold_on_background_norm()
             im = pix.topil()
 
-        if options.filter_ocr_image:
-            filt = load_plugin(options.filter_ocr_image)
-            im = filt(im)
-
         del draw
+
+        filter_im = page_context.plugin_manager.hook.filter_ocr_image(
+            page=page_context, image=im
+        )
+        if filter_im is not None:
+            im = filter_im
+
         # Pillow requires integer DPI
-        dpi = round(xres), round(yres)
+        dpi = tuple(round(coord) for coord in im.info['dpi'])
         im.save(output_file, dpi=dpi)
     return output_file
 
 
-def ocr_tesseract_hocr(input_file, page_context):
+def ocr_engine_hocr(input_file: Path, page_context: PageContext):
     hocr_out = page_context.get_path('ocr_hocr.hocr')
     hocr_text_out = page_context.get_path('ocr_hocr.txt')
     options = page_context.options
-    tesseract.generate_hocr(
+
+    ocr_engine = page_context.plugin_manager.hook.get_ocr_engine()
+    ocr_engine.generate_hocr(
         input_file=input_file,
-        output_files=[hocr_out, hocr_text_out],
-        language=options.language,
-        engine_mode=options.tesseract_oem,
-        tessconfig=options.tesseract_config,
-        timeout=options.tesseract_timeout,
-        pagesegmode=options.tesseract_pagesegmode,
-        user_words=options.user_words,
-        user_patterns=options.user_patterns,
-        tesseract_env=options.tesseract_env,
-        log=page_context.log,
+        output_hocr=hocr_out,
+        output_text=hocr_text_out,
+        options=options,
     )
     return (hocr_out, hocr_text_out)
 
 
 def should_visible_page_image_use_jpg(pageinfo):
     # If all images were JPEGs originally, produce a JPEG as output
-    return pageinfo.images and all(im.enc == 'jpeg' for im in pageinfo.images)
+    return pageinfo.images and all(im.enc == Encoding.jpeg for im in pageinfo.images)
 
 
-def create_visible_page_jpg(image, page_context):
+def create_visible_page_jpg(image: Path, page_context: PageContext) -> Path:
     output_file = page_context.get_path('visible.jpg')
     with Image.open(image) as im:
         # At this point the image should be a .png, but deskew, unpaper
         # might have removed the DPI information. In this case, fall back to
         # square DPI used to rasterize. When the preview image was
         # rasterized, it was also converted to square resolution, which is
-        # what we want to give tesseract, so keep it square.
-        fallback_dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-        dpi = im.info.get('dpi', (fallback_dpi, fallback_dpi))
+        # what we want to give to the OCR engine, so keep it square.
+        if 'dpi' in im.info:
+            dpi = Resolution(*im.info['dpi'])
+        else:
+            # Fallback to page-implied DPI
+            dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
 
         # Pillow requires integer DPI
-        dpi = round(dpi[0]), round(dpi[1])
-        im.save(output_file, format='JPEG', dpi=dpi)
+        im.save(output_file, format='JPEG', dpi=dpi.to_int())
     return output_file
 
 
-def create_pdf_page_from_image(image, page_context):
+def create_pdf_page_from_image(image: Path, page_context: PageContext):
     # We rasterize a square DPI version of each page because most image
     # processing tools don't support rectangular DPI. Use the square DPI as it
     # accurately describes the image. It would be possible to resample the image
@@ -598,56 +593,50 @@ def create_pdf_page_from_image(image, page_context):
     # sandwich renderer would be fine.
     output_file = page_context.get_path('visible.pdf')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    layout_fun = img2pdf.get_fixed_dpi_layout_fun((dpi, dpi))
+    layout_fun = img2pdf.get_fixed_dpi_layout_fun(dpi)
 
     # This create a single page PDF
     with open(image, 'rb') as imfile, open(output_file, 'wb') as pdf:
-        page_context.log.debug('convert')
+        log.debug('convert')
         img2pdf.convert(
             imfile, with_pdfrw=False, layout_fun=layout_fun, outputstream=pdf
         )
-        page_context.log.debug('convert done')
+        log.debug('convert done')
     return output_file
 
 
-def render_hocr_page(hocr, page_context):
+def render_hocr_page(hocr: Path, page_context: PageContext):
     output_file = page_context.get_path('ocr_hocr.pdf')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    hocrtransform = HocrTransform(hocr, dpi)
+    hocrtransform = HocrTransform(hocr, dpi.x)  # square
     hocrtransform.to_pdf(
         output_file,
-        imageFileName=None,
-        showBoundingboxes=False,
-        invisibleText=True,
-        interwordSpaces=True,
+        image_filename=None,
+        show_bounding_boxes=False,
+        invisible_text=True,
+        interword_spaces=True,
     )
     return output_file
 
 
-def ocr_tesseract_textonly_pdf(input_image, page_context):
+def ocr_engine_textonly_pdf(input_image: Path, page_context: PageContext):
     output_pdf = page_context.get_path('ocr_tess.pdf')
     output_text = page_context.get_path('ocr_tess.txt')
     options = page_context.options
-    tesseract.generate_pdf(
-        input_image=input_image,
-        skip_pdf=None,
+
+    ocr_engine = page_context.plugin_manager.hook.get_ocr_engine()
+    ocr_engine.generate_pdf(
+        input_file=input_image,
         output_pdf=output_pdf,
         output_text=output_text,
-        language=options.language,
-        engine_mode=options.tesseract_oem,
-        text_only=True,
-        tessconfig=options.tesseract_config,
-        timeout=options.tesseract_timeout,
-        pagesegmode=options.tesseract_pagesegmode,
-        user_words=options.user_words,
-        user_patterns=options.user_patterns,
-        tesseract_env=options.tesseract_env,
-        log=page_context.log,
+        options=options,
     )
     return (output_pdf, output_text)
 
 
-def get_docinfo(base_pdf, options):
+def get_docinfo(base_pdf: pikepdf.Pdf, context: PdfContext) -> Dict[str, str]:
+    options = context.options
+
     def from_document_info(key):
         try:
             s = base_pdf.docinfo[key]
@@ -659,7 +648,6 @@ def get_docinfo(base_pdf, options):
         k: from_document_info(k)
         for k in ('/Title', '/Author', '/Keywords', '/Subject', '/CreationDate')
     }
-    renderer_tag = 'OCR'
     if options is not None:
         if options.title:
             pdfmark['/Title'] = options.title
@@ -670,12 +658,9 @@ def get_docinfo(base_pdf, options):
         if options.subject:
             pdfmark['/Subject'] = options.subject
 
-        if options.pdf_renderer == 'sandwich':
-            renderer_tag = 'OCR-PDF'
+    creator_tag = context.plugin_manager.hook.get_ocr_engine().creator_tag(options)
 
-    pdfmark['/Creator'] = (
-        f'{PROGRAM_NAME} {VERSION} / ' f'Tesseract {renderer_tag} {tesseract.version()}'
-    )
+    pdfmark['/Creator'] = f'{PROGRAM_NAME} {VERSION} / {creator_tag}'
     pdfmark['/Producer'] = f'pikepdf {pikepdf.__version__}'
     if 'OCRMYPDF_CREATOR' in os.environ:
         pdfmark['/Creator'] = os.environ['OCRMYPDF_CREATOR']
@@ -686,15 +671,16 @@ def get_docinfo(base_pdf, options):
     return pdfmark
 
 
-def generate_postscript_stub(context):
+def generate_postscript_stub(context: PdfContext):
     output_file = context.get_path('pdfa.ps')
     generate_pdfa_ps(output_file)
     return output_file
 
 
-def convert_to_pdfa(input_pdf, input_ps_stub, context):
+def convert_to_pdfa(input_pdf: Path, input_ps_stub: Path, context: PdfContext):
     options = context.options
     input_pdfinfo = context.pdfinfo
+    fix_docinfo_file = context.get_path('fix_docinfo.pdf')
     output_file = context.get_path('pdfa.pdf')
 
     # If the DocumentInfo record contains NUL characters, Ghostscript will
@@ -702,79 +688,106 @@ def convert_to_pdfa(input_pdf, input_ps_stub, context):
     # NULs in DocumentInfo seem to be common since older Acrobats included them.
     # pikepdf can deal with this, but we make the world a better place by
     # stamping them out as soon as possible.
+    modified = False
     with pikepdf.open(input_pdf) as pdf_file:
-        if pdf_file.docinfo:
-            modified = False
-            for k, v in pdf_file.docinfo.items():
-                if b'\x00' in bytes(v):
-                    pdf_file.docinfo[k] = bytes(v).replace(b'\x00', b'')
-                    modified = True
-            if modified:
-                pdf_file.save(input_pdf)
+        try:
+            len(pdf_file.docinfo)
+        except TypeError:
+            log.error(
+                "File contains a malformed DocumentInfo block - continuing anyway"
+            )
+        else:
+            if pdf_file.docinfo:
+                for k, v in pdf_file.docinfo.items():
+                    if b'\x00' in bytes(v):
+                        pdf_file.docinfo[k] = bytes(v).replace(b'\x00', b'')
+                        modified = True
+        if modified:
+            pdf_file.save(fix_docinfo_file)
+        else:
+            safe_symlink(input_pdf, fix_docinfo_file)
 
-    ghostscript.generate_pdfa(
+    context.plugin_manager.hook.generate_pdfa(
         pdf_version=input_pdfinfo.min_version,
-        pdf_pages=[input_pdf, input_ps_stub],
+        pdf_pages=[fix_docinfo_file],
+        pdfmark=input_ps_stub,
         output_file=output_file,
         compression=options.pdfa_image_compression,
-        log=context.log,
         pdfa_part=options.output_type[-1],  # is pdfa-1, pdfa-2, or pdfa-3
     )
 
     return output_file
 
 
-def metadata_fixup(working_file, context):
+def should_linearize(working_file: Path, context: PdfContext):
+    filesize = os.stat(working_file).st_size
+    if filesize > (context.options.fast_web_view * 1_000_000):
+        return True
+    return False
+
+
+def metadata_fixup(working_file: Path, context: PdfContext):
     output_file = context.get_path('metafix.pdf')
     options = context.options
-    original = pikepdf.open(context.origin)
-    docinfo = get_docinfo(original, options)
-    pdf = pikepdf.open(working_file)
-    with pdf.open_metadata() as meta:
-        meta.load_from_docinfo(docinfo, delete_missing=False)
-        # If xmp:CreateDate is missing, set it to the modify date to
-        # match Ghostscript, for consistency
-        if 'xmp:CreateDate' not in meta:
-            meta['xmp:CreateDate'] = meta.get('xmp:ModifyDate', '')
 
-        meta_original = original.open_metadata()
-        not_copied = set(meta_original.keys()) - set(meta.keys())
-        if not_copied:
-            if options.output_type.startswith('pdfa'):
-                context.log.warning(
-                    "Some input metadata could not be copied because it is not "
-                    "permitted in PDF/A. You may wish to examine the output "
-                    "PDF's XMP metadata."
-                )
-                context.log.debug(
-                    "The following metadata fields were not copied: %r", not_copied
-                )
-            else:
-                context.log.error(
-                    "Some input metadata could not be copied."
-                    "You may wish to examine the output PDF's XMP metadata."
-                )
-                context.log.info(
-                    "The following metadata fields were not copied: %r", not_copied
-                )
+    def report_on_metadata(missing):
+        if not missing:
+            return
+        if options.output_type.startswith('pdfa'):
+            log.warning(
+                "Some input metadata could not be copied because it is not "
+                "permitted in PDF/A. You may wish to examine the output "
+                "PDF's XMP metadata."
+            )
+            log.debug("The following metadata fields were not copied: %r", missing)
+        else:
+            log.error(
+                "Some input metadata could not be copied."
+                "You may wish to examine the output PDF's XMP metadata."
+            )
+            log.info("The following metadata fields were not copied: %r", missing)
+
+    with pikepdf.open(context.origin) as original, pikepdf.open(working_file) as pdf:
+        docinfo = get_docinfo(original, context)
+        with pdf.open_metadata() as meta:
+            meta.load_from_docinfo(docinfo, delete_missing=False, raise_failure=False)
+            # If xmp:CreateDate is missing, set it to the modify date to
+            # match Ghostscript, for consistency
+            if 'xmp:CreateDate' not in meta:
+                meta['xmp:CreateDate'] = meta.get('xmp:ModifyDate', '')
+
+            meta_original = original.open_metadata()
+            missing = set(meta_original.keys()) - set(meta.keys())
+            report_on_metadata(missing)
+
+        pdf.save(
+            output_file,
+            compress_streams=True,
+            preserve_pdfa=True,
+            object_stream_mode=pikepdf.ObjectStreamMode.generate,
+            linearize=(  # Don't linearize if optimize() will be linearizing too
+                should_linearize(working_file, context)
+                if options.optimize == 0
+                else False
+            ),
+        )
 
-    pdf.save(
-        output_file,
-        compress_streams=True,
-        object_stream_mode=pikepdf.ObjectStreamMode.generate,
-    )
-    original.close()
-    pdf.close()
     return output_file
 
 
-def optimize_pdf(input_file, context):
+def optimize_pdf(input_file: Path, context: PdfContext):
     output_file = context.get_path('optimize.pdf')
-    optimize(input_file, output_file, context)
+    save_settings = dict(
+        compress_streams=True,
+        preserve_pdfa=True,
+        object_stream_mode=pikepdf.ObjectStreamMode.generate,
+        linearize=should_linearize(input_file, context),
+    )
+    optimize(input_file, output_file, context, save_settings)
     return output_file
 
 
-def merge_sidecars(txt_files, context):
+def merge_sidecars(txt_files: Iterable[Optional[Path]], context: PdfContext):
     output_file = context.get_path('sidecar.txt')
     with open(output_file, 'w', encoding="utf-8") as stream:
         for page_num, txt_file in enumerate(txt_files):
@@ -783,11 +796,9 @@ def merge_sidecars(txt_files, context):
             if txt_file:
                 with open(txt_file, 'r', encoding="utf-8") as in_:
                     txt = in_.read()
-                    # Tesseract v4 alpha started adding form feeds in
-                    # commit aa6eb6b
-                    # No obvious way to detect what binaries will do this, so
-                    # for consistency just ignore its form feeds and insert our
-                    # own
+                    # Some OCR engines (e.g. Tesseract v4 alpha) add form feeds
+                    # between pages, and some do not. For consistency, we ignore
+                    # any added by the OCR engine and them on our own.
                     if txt.endswith('\f'):
                         stream.write(txt[:-1])
                     else:
@@ -797,8 +808,8 @@ def merge_sidecars(txt_files, context):
     return output_file
 
 
-def copy_final(input_file, output_file, context):
-    context.log.debug('%s -> %s', input_file, output_file)
+def copy_final(input_file: Path, output_file: Path, _context: PdfContext):
+    log.debug('%s -> %s', input_file, output_file)
     with open(input_file, 'rb') as input_stream:
         if output_file == '-':
             copyfileobj(input_stream, sys.stdout.buffer)

src/ocrmypdf/_plugin_manager.py

@@ -0,0 +1,67 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import argparse
+import importlib
+import importlib.util
+import sys
+from pathlib import Path
+from typing import List, Tuple
+
+import pluggy
+
+from ocrmypdf import pluginspec
+from ocrmypdf.cli import get_parser, plugins_only_parser
+
+
+def get_plugin_manager(plugins: List[str], builtins=True):
+    pm = pluggy.PluginManager('ocrmypdf')
+    pm.add_hookspecs(pluginspec)
+
+    if builtins:
+        all_plugins = [
+            'ocrmypdf.builtin_plugins.ghostscript',
+            'ocrmypdf.builtin_plugins.tesseract_ocr',
+        ] + plugins
+    else:
+        all_plugins = plugins
+    for name in all_plugins:
+        if name.endswith('.py'):
+            # Import by filename
+            module_name = Path(name).stem
+            spec = importlib.util.spec_from_file_location(module_name, name)
+            module = importlib.util.module_from_spec(spec)
+            sys.modules[module_name] = module
+            spec.loader.exec_module(module)
+        else:
+            # Import by dotted module name
+            module = importlib.import_module(name)
+        pm.register(module)
+    return pm
+
+
+def get_parser_options_plugins(
+    args,
+) -> Tuple[argparse.ArgumentParser, argparse.Namespace, pluggy.PluginManager]:
+    pre_options, _unused = plugins_only_parser.parse_known_args(args=args)
+    plugin_manager = get_plugin_manager(pre_options.plugins)
+
+    parser = get_parser()
+    plugin_manager.hook.add_options(parser=parser)  # pylint: disable=no-member
+
+    options = parser.parse_args(args=args)
+    return parser, options, plugin_manager

src/ocrmypdf/_plugins.py

@@ -1,79 +0,0 @@
-# © 2019 James R. Barlow: github.com/jbarlow83
-#
-# This file is part of OCRmyPDF.
-#
-# OCRmyPDF is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# OCRmyPDF is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
-
-import logging
-import importlib
-import os
-import sys
-from pathlib import Path
-
-log = logging.getLogger(__name__)
-
-
-def _load_object_from_module(location):
-    """Load a object given a module location
-
-    For location=a.b.c, will effectively run "from a.b import c"
-
-    Example:
-        _load_object_from_module("a.b.c")
-
-    """
-    module_parts = location.split('.')
-    module_name = '.'.join(module_parts[:-1])
-    object_name = module_parts[-1]
-    module = importlib.import_module(module_name)
-    obj = getattr(module, object_name)
-    log.debug(f"Loaded object: from {module_name} import {object_name}")
-    return obj
-
-
-def _load_object_from_pyfile(location):
-    """Load a object from a file
-
-    Example:
-        _load_object_from_pyfile("test.py::blur_filter")
-    """
-    filename, object_name = location.split('::', maxsplit=1)
-    log.debug(f"Loading object {object_name} from {filename}")
-
-    module_name = Path(filename).stem
-    spec = importlib.util.spec_from_file_location(module_name, filename)
-    module = importlib.util.module_from_spec(spec)
-    spec.loader.exec_module(module)
-    obj = getattr(module, object_name)
-    return obj
-
-
-def load_plugin(plugin):
-    if callable(plugin):
-        return plugin
-
-    if not isinstance(plugin, str):
-        raise TypeError()
-
-    if '::' not in plugin:
-        plugin = _load_object_from_module(plugin)
-    else:
-        plugin = _load_object_from_pyfile(plugin)
-
-    return plugin
-
-
-def check_plugin_loadable(plugin):
-    load_plugin(plugin)
-    return plugin

src/ocrmypdf/_sync.py

@@ -17,19 +17,20 @@
 
 import logging
 import logging.handlers
-import multiprocessing
 import os
-import signal
 import sys
 import threading
-from collections import namedtuple
+from functools import partial
+from pathlib import Path
 from tempfile import mkdtemp
+from typing import List, NamedTuple, Optional, Tuple
 
-from tqdm import tqdm
+import PIL
 
-from . import __version__
-from ._jobcontext import PDFContext, cleanup_working_files, make_logger
-from ._pipeline import (
+from ocrmypdf._concurrent import exec_progress_pool
+from ocrmypdf._graft import OcrGrafter
+from ocrmypdf._jobcontext import PageContext, PdfContext, cleanup_working_files
+from ocrmypdf._pipeline import (
     convert_to_pdfa,
     copy_final,
     create_ocr_image,
@@ -41,8 +42,8 @@ from ._pipeline import (
     is_ocr_required,
     merge_sidecars,
     metadata_fixup,
-    ocr_tesseract_hocr,
-    ocr_tesseract_textonly_pdf,
+    ocr_engine_hocr,
+    ocr_engine_textonly_pdf,
     optimize_pdf,
     preprocess_clean,
     preprocess_deskew,
@@ -54,23 +55,51 @@ from ._pipeline import (
     triage,
     validate_pdfinfo_options,
 )
-from ._validation import (
+from ocrmypdf._plugin_manager import get_plugin_manager
+from ocrmypdf._validation import (
     check_requested_output_file,
     create_input_file,
     report_output_file_size,
 )
-from ._graft import OcrGrafter
-from .exceptions import ExitCode, ExitCodeException
-from .exec import qpdf
-from .helpers import available_cpu_count
-from .pdfa import file_claims_pdfa
+from ocrmypdf.exceptions import ExitCode, ExitCodeException
+from ocrmypdf.helpers import available_cpu_count, check_pdf, samefile
+from ocrmypdf.pdfa import file_claims_pdfa
 
-PageResult = namedtuple(
-    'PageResult', 'pageno, pdf_page_from_image, ocr, text, orientation_correction'
-)
+log = logging.getLogger(__name__)
 
 
-def preprocess(page_context, image, remove_background, deskew, clean):
+class PageResult(NamedTuple):
+    pageno: int
+    pdf_page_from_image: Optional[Path]
+    ocr: Optional[Path]
+    text: Optional[Path]
+    orientation_correction: int
+
+
+tls = threading.local()
+tls.pageno = None
+
+
+old_factory = logging.getLogRecordFactory()
+
+
+def record_factory(*args, **kwargs):
+    record = old_factory(*args, **kwargs)
+    if hasattr(tls, 'pageno'):
+        record.pageno = tls.pageno
+    return record
+
+
+logging.setLogRecordFactory(record_factory)
+
+
+def preprocess(
+    page_context: PageContext,
+    image: Path,
+    remove_background: bool,
+    deskew: bool,
+    clean: bool,
+) -> Path:
     if remove_background:
         image = preprocess_remove_background(image, page_context)
     if deskew:
@@ -80,54 +109,55 @@ def preprocess(page_context, image, remove_background, deskew, clean):
     return image
 
 
-def exec_page_sync(page_context):
+def make_intermediate_images(
+    page_context: PageContext, orientation_correction: int
+) -> Tuple[Path, Optional[Path]]:
     options = page_context.options
-    orientation_correction = 0
-    pdf_page_from_image_out = None
-    ocr_out = None
-    text_out = None
-    if is_ocr_required(page_context):
-        if options.rotate_pages:
-            # Rasterize
-            rasterize_preview_out = rasterize_preview(page_context.origin, page_context)
-            orientation_correction = get_orientation_correction(
-                rasterize_preview_out, page_context
-            )
 
-        rasterize_out = rasterize(
-            page_context.origin,
+    ocr_image = preprocess_out = None
+    rasterize_out = rasterize(
+        page_context.origin,
+        page_context,
+        correction=orientation_correction,
+        remove_vectors=False,
+    )
+
+    if not any([options.clean, options.clean_final, options.remove_vectors]):
+        ocr_image = preprocess_out = preprocess(
             page_context,
-            correction=orientation_correction,
-            remove_vectors=False,
+            rasterize_out,
+            options.remove_background,
+            options.deskew,
+            clean=False,
         )
-
-        if not any([options.clean, options.clean_final, options.remove_vectors]):
-            ocr_image = preprocess_out = preprocess(
+    else:
+        if not options.lossless_reconstruction:
+            preprocess_out = preprocess(
                 page_context,
                 rasterize_out,
                 options.remove_background,
                 options.deskew,
-                clean=False,
+                clean=options.clean_final,
+            )
+        if options.remove_vectors:
+            rasterize_ocr_out = rasterize(
+                page_context.origin,
+                page_context,
+                correction=orientation_correction,
+                remove_vectors=True,
+                output_tag='_ocr',
             )
         else:
-            if not options.lossless_reconstruction:
-                preprocess_out = preprocess(
-                    page_context,
-                    rasterize_out,
-                    options.remove_background,
-                    options.deskew,
-                    clean=options.clean_final,
-                )
-            if options.remove_vectors:
-                rasterize_ocr_out = rasterize(
-                    page_context.origin,
-                    page_context,
-                    correction=orientation_correction,
-                    remove_vectors=True,
-                    output_tag='_ocr',
-                )
-            else:
-                rasterize_ocr_out = rasterize_out
+            rasterize_ocr_out = rasterize_out
+
+        if (
+            preprocess_out
+            and rasterize_ocr_out == rasterize_out
+            and options.clean == options.clean_final
+        ):
+            # Optimization: image for OCR is identical to presentation image
+            ocr_image = preprocess_out
+        else:
             ocr_image = preprocess(
                 page_context,
                 rasterize_ocr_out,
@@ -135,28 +165,56 @@ def exec_page_sync(page_context):
                 options.deskew,
                 clean=options.clean,
             )
+    return ocr_image, preprocess_out
 
-        ocr_image_out = create_ocr_image(ocr_image, page_context)
 
-        pdf_page_from_image_out = None
-        if not options.lossless_reconstruction:
-            visible_image_out = preprocess_out
-            if should_visible_page_image_use_jpg(page_context.pageinfo):
-                visible_image_out = create_visible_page_jpg(
-                    visible_image_out, page_context
-                )
-            pdf_page_from_image_out = create_pdf_page_from_image(
-                visible_image_out, page_context
-            )
+def exec_page_sync(page_context: PageContext):
+    options = page_context.options
+    tls.pageno = page_context.pageno + 1
 
-        if options.pdf_renderer == 'hocr':
-            (hocr_out, text_out) = ocr_tesseract_hocr(ocr_image_out, page_context)
-            ocr_out = render_hocr_page(hocr_out, page_context)
+    if not is_ocr_required(page_context):
+        return PageResult(
+            pageno=page_context.pageno,
+            pdf_page_from_image=None,
+            ocr=None,
+            text=None,
+            orientation_correction=0,
+        )
 
-        if options.pdf_renderer == 'sandwich':
-            (ocr_out, text_out) = ocr_tesseract_textonly_pdf(
-                ocr_image_out, page_context
-            )
+    orientation_correction = 0
+    if options.rotate_pages:
+        # Rasterize
+        rasterize_preview_out = rasterize_preview(page_context.origin, page_context)
+        orientation_correction = get_orientation_correction(
+            rasterize_preview_out, page_context
+        )
+
+    ocr_image, preprocess_out = make_intermediate_images(
+        page_context, orientation_correction
+    )
+    ocr_image_out = create_ocr_image(ocr_image, page_context)
+
+    pdf_page_from_image_out = None
+    if not options.lossless_reconstruction:
+        assert preprocess_out
+        visible_image_out = preprocess_out
+        if should_visible_page_image_use_jpg(page_context.pageinfo):
+            visible_image_out = create_visible_page_jpg(visible_image_out, page_context)
+        filtered_image = page_context.plugin_manager.hook.filter_page_image(
+            page=page_context, image_filename=visible_image_out
+        )
+        if filtered_image:
+            visible_image_out = filtered_image
+        pdf_page_from_image_out = create_pdf_page_from_image(
+            visible_image_out, page_context
+        )
+
+    if options.pdf_renderer == 'hocr':
+        (hocr_out, text_out) = ocr_engine_hocr(ocr_image_out, page_context)
+        ocr_out = render_hocr_page(hocr_out, page_context)
+
+    if options.pdf_renderer == 'sandwich':
+        (ocr_out, text_out) = ocr_engine_textonly_pdf(ocr_image_out, page_context)
 
     return PageResult(
         pageno=page_context.pageno,
@@ -167,7 +225,7 @@ def exec_page_sync(page_context):
     )
 
 
-def post_process(pdf_file, context):
+def post_process(pdf_file, context: PdfContext):
     pdf_out = pdf_file
     if context.options.output_type.startswith('pdfa'):
         ps_stub_out = generate_postscript_stub(context)
@@ -177,111 +235,50 @@ def post_process(pdf_file, context):
     return optimize_pdf(pdf_out, context)
 
 
-def worker_init(queue):
-    """Initialize a process pool worker"""
-
-    # Ignore SIGINT (our parent process will kill us gracefully)
-    signal.signal(signal.SIGINT, signal.SIG_IGN)
-
-    # Reconfigure the root logger for this process to send all messages to a queue
-    h = logging.handlers.QueueHandler(queue)
-    root = logging.getLogger()
-    root.handlers = []
-    root.addHandler(h)
+def worker_init(max_pixels: int):
+    # In Windows, child process will not inherit our change to this value in
+    # the parent process, so ensure workers get it set. Not needed when running
+    # threaded, but harmless to set again.
+    PIL.Image.MAX_IMAGE_PIXELS = max_pixels
 
 
-def worker_thread_init(queue):
-    pass
-
-
-def log_listener(queue):
-    """Listen to the worker processes and forward the messages to logging
-
-    For simplicity this is a thread rather than a process. Only one process
-    should actually write to sys.stderr or whatever we're using, so if this is
-    made into a process the main application needs to be directed to it.
-
-    See https://docs.python.org/3/howto/logging-cookbook.html#logging-to-a-single-file-from-multiple-processes
-    """
-
-    while True:
-        try:
-            record = queue.get()
-            if record is None:
-                break
-            logger = logging.getLogger(record.name)
-            logger.handle(record)
-        except Exception:
-            import traceback
-
-            print("Logging problem", file=sys.stderr)
-            traceback.print_exc(file=sys.stderr)
-
-
-def exec_concurrent(context):
+def exec_concurrent(context: PdfContext):
     """Execute the pipeline concurrently"""
 
     # Run exec_page_sync on every page context
     max_workers = min(len(context.pdfinfo), context.options.jobs)
     if max_workers > 1:
-        context.log.info("Start processing %d pages concurrent", max_workers)
+        log.info("Start processing %d pages concurrently", max_workers)
 
-    # Tesseract 4.0 is multithreaded, and we also run multiple workers. We want to
-    # avoid the situation where we end up trying to run NxN jobs on N CPU cores,
-    # as that gives poor performance. Performance testing shows we're better off
-    # parallelizing ocrmypdf and forcing Tesseract to be single threaded, which we
-    # get by setting the envvar OMP_THREAD_LIMIT to 1. But if the page count of the
-    # input file is small, then we allow Tesseract to use threads, subject to the
-    # constraint: (ocrmypdf workers) * (tesseract threads) <= max_workers and limiting
-    # Tesseract to 4 threads.
-    tess_threads = min(4, context.options.jobs // max_workers)
-    if context.options.tesseract_env is None:
-        context.options.tesseract_env = os.environ.copy()
-    context.options.tesseract_env.setdefault('OMP_THREAD_LIMIT', str(tess_threads))
-    if tess_threads > 1:
-        context.log.info("Using Tesseract OpenMP thread limit %d", tess_threads)
-
-    if context.options.use_threads:
-        from multiprocessing.dummy import Pool
-
-        initializer = worker_thread_init
-    else:
-        Pool = multiprocessing.Pool
-        initializer = worker_init
-
-    sidecars = [None] * len(context.pdfinfo)
+    sidecars: List[Optional[Path]] = [None] * len(context.pdfinfo)
     ocrgraft = OcrGrafter(context)
 
-    log_queue = multiprocessing.Queue(-1)
-    listener = threading.Thread(target=log_listener, args=(log_queue,))
-    listener.start()
-    with tqdm(
-        total=(2 * len(context.pdfinfo)),
-        desc='OCR',
-        unit='page',
-        unit_scale=0.5,
-        disable=not context.options.progress_bar,
-    ) as pbar, Pool(
-        processes=max_workers, initializer=initializer, initargs=(log_queue,)
-    ) as pool:
-        results = pool.imap_unordered(exec_page_sync, context.get_page_contexts())
-        while True:
-            try:
-                page_result = results.next()
-                sidecars[page_result.pageno] = page_result.text
-                pbar.update()
-                ocrgraft.graft_page(page_result)
-                pbar.update()
-            except StopIteration:
-                break
-            except (Exception, KeyboardInterrupt):
-                pool.terminate()
-                log_queue.put_nowait(None)  # Terminate log listener
-                # Don't try listener.join() here, will deadlock
-                raise
+    def update_page(result: PageResult, pbar):
+        sidecars[result.pageno] = result.text
+        pbar.update()
+        ocrgraft.graft_page(
+            pageno=result.pageno,
+            image=result.pdf_page_from_image,
+            textpdf=result.ocr,
+            autorotate_correction=result.orientation_correction,
+        )
+        pbar.update()
 
-    log_queue.put_nowait(None)
-    listener.join()
+    exec_progress_pool(
+        use_threads=context.options.use_threads,
+        max_workers=max_workers,
+        tqdm_kwargs=dict(
+            total=(2 * len(context.pdfinfo)),
+            desc='OCR',
+            unit='page',
+            unit_scale=0.5,
+            disable=not context.options.progress_bar,
+        ),
+        task_initializer=partial(worker_init, PIL.Image.MAX_IMAGE_PIXELS),
+        task=exec_page_sync,
+        task_arguments=context.get_page_contexts(),
+        task_finished=update_page,
+    )
 
     # Output sidecar text
     if context.options.sidecar:
@@ -302,41 +299,84 @@ def exec_concurrent(context):
 class NeverRaise(Exception):
     """An exception that is never raised"""
 
-    pass
+    pass  # pylint: disable=unnecessary-pass
 
 
-def run_pipeline(options, api=False):
-    log = make_logger(options, __name__)
+def configure_debug_logging(log_filename, prefix=''):
+    log_file_handler = logging.FileHandler(log_filename, delay=True)
+    log_file_handler.setLevel(logging.DEBUG)
+    formatter = logging.Formatter(
+        '[%(asctime)s] - %(name)s - %(levelname)7s -%(pageno)s %(message)s'
+    )
+    log_file_handler.setFormatter(formatter)
+    logging.getLogger(prefix).addHandler(log_file_handler)
+    return log_file_handler
 
+
+def run_pipeline(options, *, plugin_manager, api=False):
     # Any changes to options will not take effect for options that are already
     # bound to function parameters in the pipeline. (For example
     # options.input_file, options.pdf_renderer are already bound.)
     if not options.jobs:
         options.jobs = available_cpu_count()
+    if not plugin_manager:
+        plugin_manager = get_plugin_manager(options.plugins)
+
+    work_folder = Path(mkdtemp(prefix="com.github.ocrmypdf."))
+    debug_log_handler = None
+    if (options.keep_temporary_files or options.verbose >= 1) and not os.environ.get(
+        'PYTEST_CURRENT_TEST', ''
+    ):
+        debug_log_handler = configure_debug_logging(Path(work_folder) / "debug.log")
 
-    work_folder = mkdtemp(prefix="com.github.ocrmypdf.")
     try:
         check_requested_output_file(options)
-        start_input_file = create_input_file(options, work_folder)
+        start_input_file, original_filename = create_input_file(options, work_folder)
 
         # Triage image or pdf
         origin_pdf = triage(
-            start_input_file, os.path.join(work_folder, 'origin.pdf'), options, log
+            original_filename, start_input_file, work_folder / 'origin.pdf', options
         )
 
         # Gather pdfinfo and create context
         pdfinfo = get_pdfinfo(
             origin_pdf,
-            detailed_page_analysis=options.redo_ocr,
+            detailed_analysis=options.redo_ocr,
             progbar=options.progress_bar,
+            max_workers=options.jobs if not options.use_threads else 1,  # To help debug
+            check_pages=options.pages,
         )
-        context = PDFContext(options, work_folder, origin_pdf, pdfinfo)
+
+        context = PdfContext(options, work_folder, origin_pdf, pdfinfo, plugin_manager)
 
         # Validate options are okay for this pdf
         validate_pdfinfo_options(context)
 
         # Execute the pipeline
         exec_concurrent(context)
+
+        if options.output_file == '-':
+            log.info("Output sent to stdout")
+        elif samefile(options.output_file, os.devnull):
+            pass  # Say nothing when sending to dev null
+        else:
+            if options.output_type.startswith('pdfa'):
+                pdfa_info = file_claims_pdfa(options.output_file)
+                if pdfa_info['pass']:
+                    log.info(
+                        "Output file is a %s (as expected)", pdfa_info['conformance']
+                    )
+                else:
+                    log.warning(
+                        "Output file is okay but is not PDF/A (seems to be %s)",
+                        pdfa_info['conformance'],
+                    )
+                    return ExitCode.pdfa_conversion_failed
+            if not check_pdf(options.output_file):
+                log.warning('Output file: The generated PDF is INVALID')
+                return ExitCode.invalid_output_pdf
+            report_output_file_size(options, start_input_file, options.output_file)
+
     except (KeyboardInterrupt if not api else NeverRaise) as e:
         if options.verbose >= 1:
             log.exception("KeyboardInterrupt")
@@ -349,33 +389,16 @@ def run_pipeline(options, api=False):
         else:
             log.error(type(e).__name__)
         return e.exit_code
-    except (Exception if not api else NeverRaise) as e:
+    except (Exception if not api else NeverRaise) as e:  # pylint: disable=broad-except
         log.exception("An exception occurred while executing the pipeline")
         return ExitCode.other_error
     finally:
+        if debug_log_handler:
+            try:
+                debug_log_handler.close()
+                log.removeHandler(debug_log_handler)
+            except EnvironmentError as e:
+                print(e, file=sys.stderr)
         cleanup_working_files(work_folder, options)
 
-    if options.output_file == '-':
-        log.info("Output sent to stdout")
-    elif os.path.samefile(options.output_file, os.devnull):
-        pass  # Say nothing when sending to dev null
-    else:
-        if options.output_type.startswith('pdfa'):
-            pdfa_info = file_claims_pdfa(options.output_file)
-            if pdfa_info['pass']:
-                msg = f"Output file is a {pdfa_info['conformance']} (as expected)"
-                log.info(msg)
-            else:
-                msg = (
-                    f"Output file is okay but is not PDF/A "
-                    f"(seems to be {pdfa_info['conformance']})"
-                )
-                log.warning(msg)
-                return ExitCode.pdfa_conversion_failed
-        if not qpdf.check(options.output_file, log):
-            log.warning('Output file: The generated PDF is INVALID')
-            return ExitCode.invalid_output_pdf
-
-        report_output_file_size(options, start_input_file, options.output_file)
-
     return ExitCode.ok

src/ocrmypdf/_validation.py

@@ -17,36 +17,39 @@
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
 
+import locale
 import logging
 import os
 import sys
+import unicodedata
 from pathlib import Path
 from shutil import copyfileobj
+from typing import Tuple
 
+import pikepdf
 import PIL
 
-from ._unicodefun import verify_python3_env
-from .exceptions import (
+from ocrmypdf._exec import jbig2enc, pngquant, unpaper
+from ocrmypdf._unicodefun import verify_python3_env
+from ocrmypdf.exceptions import (
     BadArgsError,
     InputFileError,
     MissingDependencyError,
     OutputFileAccessError,
 )
-from .exec import (
-    check_external_program,
-    ghostscript,
-    jbig2enc,
-    pngquant,
-    qpdf,
-    tesseract,
-    unpaper,
+from ocrmypdf.helpers import (
+    is_file_writable,
+    is_iterable_notstr,
+    monotonic,
+    safe_symlink,
 )
-from .helpers import is_file_writable, re_symlink, is_iterable_notstr, monotonic
+from ocrmypdf.subprocess import check_external_program
 
 # -------------
 # External dependencies
 
 HOCR_OK_LANGS = frozenset(['eng', 'deu', 'spa', 'ita', 'por'])
+DEFAULT_LANGUAGE = 'eng'  # Enforce English hegemony
 
 log = logging.getLogger(__name__)
 
@@ -56,32 +59,35 @@ log = logging.getLogger(__name__)
 verify_python3_env()
 
 
-def check_options_languages(options):
-    if not options.language:
-        options.language = ['eng']  # Enforce English hegemony
-
-    # Support v2.x "eng+deu" language syntax
-    if '+' in options.language[0]:
-        options.language = options.language[0].split('+')
-
-    languages = set(options.language)
-    if not languages.issubset(tesseract.languages()):
-        msg = (
-            "The installed version of tesseract does not have language "
-            "data for the following requested languages: \n"
+def check_platform():
+    if os.name == 'nt' and sys.maxsize <= 2 ** 32:  # pragma: no cover
+        # 32-bit interpreter on Windows
+        log.error(
+            "You are running OCRmyPDF in a 32-bit (x86) Python interpreter."
+            "Please use a 64-bit (x86-64) version of Python."
         )
-        for lang in languages - tesseract.languages():
+
+
+def check_options_languages(options, ocr_engine_languages):
+    if not options.languages:
+        options.languages = {DEFAULT_LANGUAGE}
+        system_lang = locale.getlocale()[0]
+        if system_lang and not system_lang.startswith('en'):
+            log.debug("No language specified; assuming --language %s", DEFAULT_LANGUAGE)
+    if not ocr_engine_languages:
+        return
+    if not options.languages.issubset(ocr_engine_languages):
+        msg = (
+            f"OCR engine does not have language data for the following "
+            "requested languages: \n"
+        )
+        for lang in options.languages - ocr_engine_languages:
             msg += lang + '\n'
         raise MissingDependencyError(msg)
 
 
 def check_options_output(options):
-    # We have these constraints to check for.
-    # 1. Ghostscript < 9.20 mangles multibyte Unicode
-    # 2. hocr doesn't work on non-Latin languages (so don't select it)
-
-    languages = set(options.language)
-    is_latin = languages.issubset(HOCR_OK_LANGS)
+    is_latin = options.languages.issubset(HOCR_OK_LANGS)
 
     if options.pdf_renderer == 'hocr' and not is_latin:
         msg = (
@@ -91,37 +97,6 @@ def check_options_output(options):
         )
         log.warning(msg)
 
-    if ghostscript.version() < '9.20' and options.output_type != 'pdf' and not is_latin:
-        # https://bugs.ghostscript.com/show_bug.cgi?id=696874
-        # Ghostscript < 9.20 fails to encode multibyte characters properly
-        msg = (
-            "The installed version of Ghostscript does not work correctly "
-            "with the OCR languages you specified. Use --output-type pdf or "
-            "upgrade to Ghostscript 9.20 or later to avoid this issue."
-        )
-        msg += f"Found Ghostscript {ghostscript.version()}"
-        log.warning(msg)
-
-    # Decide on what renderer to use
-    if options.pdf_renderer == 'auto':
-        options.pdf_renderer = 'sandwich'
-
-    if options.pdf_renderer == 'sandwich' and not tesseract.has_textonly_pdf(
-        options.tesseract_env
-    ):
-        raise MissingDependencyError(
-            "You are using an alpha version of Tesseract 4.0 that does not support "
-            "the textonly_pdf parameter. We don't support versions this old."
-        )
-
-    if options.output_type == 'pdfa':
-        options.output_type = 'pdfa-2'
-
-    if options.output_type == 'pdfa-3' and ghostscript.version() < '9.19':
-        raise MissingDependencyError(
-            "--output-type pdfa-3 requires Ghostscript 9.19 or later"
-        )
-
     lossless_reconstruction = False
     if not any(
         (
@@ -185,7 +160,10 @@ def _pages_from_ranges(ranges):
         except ValueError:
             pages.append(int(g) - 1)
         else:
-            pages.extend(range(int(start) - 1, int(end)))
+            try:
+                pages.extend(range(int(start) - 1, int(end)))
+            except ValueError:
+                raise BadArgsError("invalid page range")
 
     if not monotonic(pages):
         log.warning(
@@ -253,15 +231,9 @@ def check_options_advanced(options):
             "--pdfa-image-compression argument has no effect when "
             "--output-type is not 'pdfa', 'pdfa-1', or 'pdfa-2'"
         )
-    if tesseract.v4(options.tesseract_env) and (
-        options.user_words or options.user_patterns
-    ):
-        log.warning('Tesseract 4.x ignores --user-words, so this has no effect')
 
 
 def check_options_metadata(options):
-    import unicodedata
-
     docinfo = [options.title, options.author, options.keywords, options.subject]
     for s in (m for m in docinfo if m):
         for c in s:
@@ -280,8 +252,9 @@ def check_options_pillow(options):
         PIL.Image.MAX_IMAGE_PIXELS = None
 
 
-def check_options(options):
-    check_options_languages(options)
+def _check_options(options, plugin_manager, ocr_engine_languages):
+    check_platform()
+    check_options_languages(options, ocr_engine_languages)
     check_options_metadata(options)
     check_options_output(options)
     check_options_sidecar(options)
@@ -290,10 +263,15 @@ def check_options(options):
     check_options_optimizing(options)
     check_options_advanced(options)
     check_options_pillow(options)
-    check_dependency_versions(options)
+    plugin_manager.hook.check_options(options=options)
 
 
-def check_closed_streams(options):
+def check_options(options, plugin_manager):
+    ocr_engine_languages = plugin_manager.hook.get_ocr_engine().languages(options)
+    _check_options(options, plugin_manager, ocr_engine_languages)
+
+
+def check_closed_streams(options):  # pragma: no cover
     """Work around Python issue with multiprocessing forking on closed streams
 
     https://bugs.python.org/issue28326
@@ -352,19 +330,19 @@ def log_page_orientations(pdfinfo):
         log.info('Page orientations detected: %s', ' '.join(orientations))
 
 
-def create_input_file(options, work_folder):
+def create_input_file(options, work_folder: Path) -> Tuple[Path, str]:
     if options.input_file == '-':
         # stdin
         log.info('reading file from standard input')
-        target = os.path.join(work_folder, 'stdin')
+        target = work_folder / 'stdin'
         with open(target, 'wb') as stream_buffer:
             copyfileobj(sys.stdin.buffer, stream_buffer)
-        return target
+        return target, "<stdin>"
     else:
         try:
-            target = os.path.join(work_folder, 'origin')
-            re_symlink(options.input_file, target)
-            return target
+            target = work_folder / 'origin'
+            safe_symlink(options.input_file, target)
+            return target, os.fspath(options.input_file)
         except FileNotFoundError:
             raise InputFileError(f"File not found - {options.input_file}")
 
@@ -389,8 +367,15 @@ def report_output_file_size(options, input_file, output_file):
         input_size = Path(input_file).stat().st_size
     except FileNotFoundError:
         return  # Outputting to stream or something
+    with pikepdf.open(output_file) as p:
+        # Overhead constants obtained by estimating amount of data added by OCR
+        # PDF/A conversion, and possible XMP metadata addition, with compression
+        FILE_OVERHEAD = 4000
+        OCR_PER_PAGE_OVERHEAD = 3000
+        reasonable_overhead = FILE_OVERHEAD + OCR_PER_PAGE_OVERHEAD * len(p.pages)
     ratio = output_size / input_size
-    if ratio < 1.35 or input_size < 25000:
+    reasonable_ratio = output_size / (input_size + reasonable_overhead)
+    if reasonable_ratio < 1.35 or input_size < 25000:
         return  # Seems fine
 
     reasons = []
@@ -407,6 +392,20 @@ def report_output_file_size(options, input_file, output_file):
                 f"The argument --{arg.replace('_', '-')} was issued, causing transcoding."
             )
 
+    if options.optimize == 0:
+        reasons.append("Optimization was disabled.")
+    else:
+        image_optimizers = {
+            'jbig2': jbig2enc.available(),
+            'pngquant': pngquant.available(),
+        }
+        for name, available in image_optimizers.items():
+            if not available:
+                reasons.append(
+                    f"The optional dependency '{name}' was not found, so some image "
+                    f"optimizations could not be attempted."
+                )
+
     if reasons:
         explanation = "Possible reasons for this include:\n" + '\n'.join(reasons) + "\n"
     else:
@@ -416,30 +415,3 @@ def report_output_file_size(options, input_file, output_file):
         f"The output file size is {ratio:.2f}× larger than the input file.\n"
         f"{explanation}"
     )
-
-
-def check_dependency_versions(options):
-    check_external_program(
-        program='tesseract',
-        package={'darwin': 'tesseract', 'linux': 'tesseract-ocr'},
-        version_checker=tesseract.version,
-        need_version='4.0.0',  # using backport for Travis CI
-    )
-    check_external_program(
-        program='gs',
-        package='ghostscript',
-        version_checker=ghostscript.version,
-        need_version='9.15',  # limited by Travis CI / Ubuntu 14.04 backports
-    )
-    if ghostscript.version() == '9.24':
-        raise MissingDependencyError(
-            "Ghostscript 9.24 contains serious regressions and is not "
-            "supported. Please upgrade to Ghostscript 9.25 or use an older "
-            "version."
-        )
-    check_external_program(
-        program='qpdf',
-        package='qpdf',
-        version_checker=qpdf.version,
-        need_version='8.0.2',
-    )

src/ocrmypdf/_version.py

@@ -0,0 +1,24 @@
+# © 2017 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+
+import pkg_resources
+
+PROGRAM_NAME = 'ocrmypdf'
+
+# Official PEP 396
+__version__ = pkg_resources.get_distribution('ocrmypdf').version

src/ocrmypdf/api.py

@@ -18,34 +18,21 @@
 import logging
 import os
 import sys
+from argparse import ArgumentParser
 from enum import IntEnum
 from pathlib import Path
+from typing import Iterable
 
-from tqdm import tqdm
+from ocrmypdf._logging import PageNumberFilter, TqdmConsole
+from ocrmypdf._plugin_manager import get_plugin_manager
+from ocrmypdf._sync import run_pipeline
+from ocrmypdf._validation import check_options
+from ocrmypdf.cli import get_parser
 
-from .cli import parser
-from ._sync import run_pipeline
-from ._validation import check_options
-
-
-class TqdmConsole:
-    """Wrapper to log messages in a way that is compatible with tqdm progress bar"""
-
-    def __init__(self, file):
-        self.file = file
-        self.py36 = sys.version_info >= (3, 6)
-
-    def write(self, msg):
-        # When no progress bar is active, tqdm.write() routes to print()
-        if self.py36:
-            if msg.strip() != '':
-                tqdm.write(msg.rstrip(), end='\n', file=self.file)
-        else:
-            tqdm.write(msg.rstrip(), end='\n', file=self.file)
-
-    def flush(self):
-        if hasattr(self.file, "flush"):
-            self.file.flush()
+try:
+    import coloredlogs
+except ModuleNotFoundError:
+    coloredlogs = None
 
 
 class Verbosity(IntEnum):
@@ -57,17 +44,21 @@ class Verbosity(IntEnum):
     debug_all = 2  #: More detailed debugging from ocrmypdf and dependent modules
 
 
-def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=False):
+def configure_logging(
+    verbosity: Verbosity,
+    progress_bar_friendly: bool = True,
+    manage_root_logger: bool = False,
+):
     """Set up logging.
 
     Library users may wish to use this function if they want their log output to be
     similar to ocrmypdf command line interface. If not used, the external application
     should configure logging on its own.
 
-    ocrmypdf will perform all of its logging under the `"ocrmypdf"` logging namespace.
-    In addition, ocrmypdf imports pdfminer, which logs under `"pdfminer"`. A library
+    ocrmypdf will perform all of its logging under the ``"ocrmypdf"`` logging namespace.
+    In addition, ocrmypdf imports pdfminer, which logs under ``"pdfminer"``. A library
     user may wish to configure both; note that pdfminer is extremely chatty at the log
-    level logging.INFO.
+    level ``logging.INFO``.
 
     Library users may perform additional configuration afterwards.
 
@@ -78,11 +69,15 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
             overwrite the progress bar
         manage_root_logger (bool): Configure the process's root logger, to ensure
             all log output is sent through
+
+    Returns:
+        The toplevel logger for ocrmypdf (or the root logger, if we are managing it).
     """
 
     prefix = '' if manage_root_logger else 'ocrmypdf'
+
     log = logging.getLogger(prefix)
-    log.setLevel(logging.INFO)
+    log.setLevel(logging.DEBUG)
 
     if progress_bar_friendly:
         console = logging.StreamHandler(stream=TqdmConsole(sys.stderr))
@@ -96,11 +91,25 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
     else:
         console.setLevel(logging.INFO)
 
-    formatter = logging.Formatter('%(levelname)7s - %(message)s')
-    if verbosity >= 1:
-        log.setLevel(logging.DEBUG)
+    console.addFilter(PageNumberFilter())
+
     if verbosity >= 2:
-        formatter = logging.Formatter('%(name)s - %(levelname)7s - %(message)s')
+        fmt = '%(levelname)7s %(name)s -%(pageno)s %(message)s'
+    else:
+        fmt = '%(pageno)s%(message)s'
+
+    use_colors = progress_bar_friendly
+    if not coloredlogs:
+        use_colors = False
+    if use_colors:
+        if os.name == 'nt':
+            use_colors = coloredlogs.enable_ansi_support()
+        if use_colors:
+            use_colors = coloredlogs.terminal_supports_colors()
+    if use_colors:
+        formatter = coloredlogs.ColoredFormatter(fmt=fmt)
+    else:
+        formatter = logging.Formatter(fmt=fmt)
 
     console.setFormatter(formatter)
     log.addHandler(console)
@@ -114,26 +123,45 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
     if manage_root_logger:
         logging.captureWarnings(True)
 
+    return log
 
-def create_options(*, input_file, output_file, **kwargs):
+
+def create_options(
+    *,
+    input_file: os.PathLike,
+    output_file: os.PathLike,
+    parser: ArgumentParser,
+    **kwargs,
+):
     cmdline = []
     deferred = []
 
     for arg, val in kwargs.items():
         if val is None:
             continue
-        if (arg.startswith('plugin') or arg.startswith('filter')) and (
-            callable(val) or isinstance(val, str)
-        ):
-            deferred.append((arg, val))
-            continue
-        elif arg == 'tesseract_env':
+
+        # These arguments with special handling for which we bypass
+        # argparse
+        if arg in {'progress_bar', 'plugins'}:
             deferred.append((arg, val))
             continue
+
         cmd_style_arg = arg.replace('_', '-')
-        cmdline.append(f"--{cmd_style_arg}")
+
+        # Booleans are special: add only if True, omit for False
         if isinstance(val, bool):
+            if val:
+                cmdline.append(f"--{cmd_style_arg}")
             continue
+
+        if isinstance(val, Iterable) and not isinstance(val, str):
+            for elem in val:
+                cmdline.append(f"--{cmd_style_arg}")
+                cmdline.append(elem)
+            continue
+
+        # We have a parameter
+        cmdline.append(f"--{cmd_style_arg}")
         if isinstance(val, (int, float)):
             cmdline.append(str(val))
         elif isinstance(val, str):
@@ -146,66 +174,61 @@ def create_options(*, input_file, output_file, **kwargs):
     cmdline.append(str(input_file))
     cmdline.append(str(output_file))
 
-    parser.api_mode = True
+    parser._api_mode = True
     options = parser.parse_args(cmdline)
     for keyword, val in deferred:
         setattr(options, keyword, val)
-
-    # If we are running a Tesseract spoof, ensure it knows what the input file is
-    if os.environ.get('PYTEST_CURRENT_TEST') and options.tesseract_env:
-        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = input_file
-
     return options
 
 
-def run(  # pylint: disable=unused-argument
-    input_file,
-    output_file,
+def ocr(  # pylint: disable=unused-argument
+    input_file: os.PathLike,
+    output_file: os.PathLike,
     *,
-    language=None,
-    image_dpi=None,
+    language: Iterable[str] = None,
+    image_dpi: int = None,
     output_type=None,
-    sidecar=None,
-    jobs=None,
-    use_threads=None,
-    title=None,
-    author=None,
-    subject=None,
-    keywords=None,
-    rotate_pages=None,
-    remove_background=None,
-    deskew=None,
-    clean=None,
-    clean_final=None,
-    unpaper_args=None,
-    oversample=None,
-    remove_vectors=None,
-    threshold=None,
-    force_ocr=None,
-    skip_text=None,
-    redo_ocr=None,
-    skip_big=None,
-    optimize=None,
-    jpg_quality=None,
-    png_quality=None,
-    jbig2_lossy=None,
-    jbig2_page_group_size=None,
-    pages=None,
-    max_image_mpixels=None,
-    tesseract_config=None,
-    tesseract_pagesegmode=None,
-    tesseract_oem=None,
+    sidecar: os.PathLike = None,
+    jobs: int = None,
+    use_threads: bool = None,
+    title: str = None,
+    author: str = None,
+    subject: str = None,
+    keywords: str = None,
+    rotate_pages: bool = None,
+    remove_background: bool = None,
+    deskew: bool = None,
+    clean: bool = None,
+    clean_final: bool = None,
+    unpaper_args: str = None,
+    oversample: int = None,
+    remove_vectors: bool = None,
+    threshold: bool = None,
+    force_ocr: bool = None,
+    skip_text: bool = None,
+    redo_ocr: bool = None,
+    skip_big: float = None,
+    optimize: int = None,
+    jpg_quality: int = None,
+    png_quality: int = None,
+    jbig2_lossy: bool = None,
+    jbig2_page_group_size: int = None,
+    pages: str = None,
+    max_image_mpixels: float = None,
+    tesseract_config: Iterable[str] = None,
+    tesseract_pagesegmode: int = None,
+    tesseract_oem: int = None,
     pdf_renderer=None,
-    tesseract_timeout=None,
-    rotate_pages_threshold=None,
+    tesseract_timeout: float = None,
+    rotate_pages_threshold: float = None,
     pdfa_image_compression=None,
-    user_words=None,
-    user_patterns=None,
-    keep_temporary_files=None,
-    progress_bar=None,
-    filter_ocr_image=None,
-    plugin_validation=None,
-    tesseract_env=None,
+    user_words: os.PathLike = None,
+    user_patterns: os.PathLike = None,
+    fast_web_view: float = None,
+    plugins: Iterable[str] = None,
+    keep_temporary_files: bool = None,
+    progress_bar: bool = None,
+    **kwargs,
 ):
     """Run OCRmyPDF on one PDF or image.
 
@@ -216,7 +239,6 @@ def run(  # pylint: disable=unused-argument
         use_threads (bool): Use worker threads instead of processes. This reduces
             performance but may make debugging easier since it is easier to set
             breakpoints.
-        tesseract_env (dict): Override environment variables for Tesseract
     Raises:
         ocrmypdf.PdfMergeFailedError: If the input PDF is malformed, preventing merging
             with the OCR layer.
@@ -240,7 +262,20 @@ def run(  # pylint: disable=unused-argument
     Returns:
         :class:`ocrmypdf.ExitCode`
     """
+    if not plugins:
+        plugins = []
+    else:
+        plugins = list(plugins)
 
-    options = create_options(**locals())
-    check_options(options)
-    return run_pipeline(options, api=True)
+    parser = get_parser()
+    _plugin_manager = get_plugin_manager(plugins)
+    _plugin_manager.hook.add_options(parser=parser)  # pylint: disable=no-member
+
+    create_options_kwargs = {
+        k: v for k, v in locals().items() if not k.startswith('_') and k != 'kwargs'
+    }
+    create_options_kwargs.update(kwargs)
+
+    options = create_options(**create_options_kwargs)
+    check_options(options, _plugin_manager)
+    return run_pipeline(options=options, plugin_manager=_plugin_manager, api=True)

src/ocrmypdf/builtin_plugins/__init__.py

@@ -0,0 +1,16 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.

src/ocrmypdf/builtin_plugins/ghostscript.py

@@ -0,0 +1,102 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+
+from ocrmypdf import hookimpl
+from ocrmypdf._exec import ghostscript
+from ocrmypdf._validation import HOCR_OK_LANGS
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.subprocess import check_external_program
+
+log = logging.getLogger(__name__)
+
+
+@hookimpl
+def check_options(options):
+    gs_version = ghostscript.version()
+    check_external_program(
+        program='gs',
+        package='ghostscript',
+        version_checker=gs_version,
+        need_version='9.15',  # limited by Travis CI / Ubuntu 14.04 backports
+    )
+    if gs_version in ('9.24', '9.51'):
+        raise MissingDependencyError(
+            f"Ghostscript {gs_version} contains serious regressions and is not "
+            "supported. Please upgrade to a newer version, or downgrade to the "
+            "previous version."
+        )
+
+    # We have these constraints to check for.
+    # 1. Ghostscript < 9.20 mangles multibyte Unicode
+    # 2. hocr doesn't work on non-Latin languages (so don't select it)
+    is_latin = options.languages.issubset(HOCR_OK_LANGS)
+    if gs_version < '9.20' and options.output_type != 'pdf' and not is_latin:
+        # https://bugs.ghostscript.com/show_bug.cgi?id=696874
+        # Ghostscript < 9.20 fails to encode multibyte characters properly
+        msg = (
+            "The installed version of Ghostscript does not work correctly "
+            "with the OCR languages you specified. Use --output-type pdf or "
+            "upgrade to Ghostscript 9.20 or later to avoid this issue."
+        )
+        msg += f"Found Ghostscript {gs_version}"
+        log.warning(msg)
+
+    if options.output_type == 'pdfa':
+        options.output_type = 'pdfa-2'
+
+    if options.output_type == 'pdfa-3' and ghostscript.version() < '9.19':
+        raise MissingDependencyError(
+            "--output-type pdfa-3 requires Ghostscript 9.19 or later"
+        )
+
+
+@hookimpl
+def rasterize_pdf_page(
+    input_file,
+    output_file,
+    raster_device,
+    raster_dpi,
+    pageno,
+    page_dpi=None,
+    rotation=None,
+    filter_vector=False,
+):
+    ghostscript.rasterize_pdf(
+        input_file,
+        output_file,
+        raster_device=raster_device,
+        raster_dpi=raster_dpi,
+        pageno=pageno,
+        page_dpi=page_dpi,
+        rotation=rotation,
+        filter_vector=filter_vector,
+    )
+    return output_file
+
+
+@hookimpl
+def generate_pdfa(pdf_pages, pdfmark, output_file, compression, pdf_version, pdfa_part):
+    ghostscript.generate_pdfa(
+        pdf_pages=[*pdf_pages, pdfmark],
+        output_file=output_file,
+        compression=compression,
+        pdf_version=pdf_version,
+        pdfa_part=pdfa_part,
+    )
+    return output_file

src/ocrmypdf/builtin_plugins/tesseract_ocr.py

@@ -0,0 +1,196 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+import os
+
+from ocrmypdf import hookimpl
+from ocrmypdf._exec import tesseract
+from ocrmypdf.cli import numeric
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.pluginspec import OcrEngine
+from ocrmypdf.subprocess import check_external_program
+
+log = logging.getLogger(__name__)
+
+
+@hookimpl
+def add_options(parser):
+    tess = parser.add_argument_group("Tesseract", "Advanced control of Tesseract OCR")
+    tess.add_argument(
+        '--tesseract-config',
+        action='append',
+        metavar='CFG',
+        default=[],
+        help="Additional Tesseract configuration files -- see documentation",
+    )
+    tess.add_argument(
+        '--tesseract-pagesegmode',
+        action='store',
+        type=int,
+        metavar='PSM',
+        choices=range(0, 14),
+        help="Set Tesseract page segmentation mode (see tesseract --help)",
+    )
+    tess.add_argument(
+        '--tesseract-oem',
+        action='store',
+        type=int,
+        metavar='MODE',
+        choices=range(0, 4),
+        help=(
+            "Set Tesseract 4.0 OCR engine mode: "
+            "0 - original Tesseract only; "
+            "1 - neural nets LSTM only; "
+            "2 - Tesseract + LSTM; "
+            "3 - default."
+        ),
+    )
+    tess.add_argument(
+        '--tesseract-timeout',
+        default=180.0,
+        type=numeric(float, 0),
+        metavar='SECONDS',
+        help='Give up on OCR after the timeout, but copy the preprocessed page '
+        'into the final output',
+    )
+    tess.add_argument(
+        '--user-words',
+        metavar='FILE',
+        help="Specify the location of the Tesseract user words file. This is a "
+        "list of words Tesseract should consider while performing OCR in "
+        "addition to its standard language dictionaries. This can improve "
+        "OCR quality especially for specialized and technical documents.",
+    )
+    tess.add_argument(
+        '--user-patterns',
+        metavar='FILE',
+        help="Specify the location of the Tesseract user patterns file.",
+    )
+
+
+@hookimpl
+def check_options(options):
+    check_external_program(
+        program='tesseract',
+        package={'linux': 'tesseract-ocr'},
+        version_checker=tesseract.version,
+        need_version='4.0.0',  # using backport for Travis CI
+    )
+
+    # Decide on what renderer to use
+    if options.pdf_renderer == 'auto':
+        options.pdf_renderer = 'sandwich'
+
+    if options.pdf_renderer == 'sandwich' and not tesseract.has_textonly_pdf(
+        set(options.languages)
+    ):
+        raise MissingDependencyError(
+            "You are using an alpha version of Tesseract 4.0 that does not support "
+            "the textonly_pdf parameter. We don't support versions this old."
+        )
+    if not tesseract.has_user_words() and (options.user_words or options.user_patterns):
+        log.warning(
+            "Tesseract 4.0 ignores --user-words and --user-patterns, so these "
+            "arguments have no effect."
+        )
+    if options.tesseract_pagesegmode in (0, 2):
+        log.warning(
+            "The --tesseract-pagesegmode argument you select will disable OCR. "
+            "This may cause processing to fail."
+        )
+
+
+@hookimpl
+def validate(pdfinfo, options):
+    # Tesseract 4.x can be multithreaded, and we also run multiple workers. We want
+    # to manage how many threads it uses to avoid creating total threads than cores.
+    # Performance testing shows we're better off
+    # parallelizing ocrmypdf and forcing Tesseract to be single threaded, which we
+    # get by setting the envvar OMP_THREAD_LIMIT to 1. But if the page count of the
+    # input file is small, then we allow Tesseract to use threads, subject to the
+    # constraint: (ocrmypdf workers) * (tesseract threads) <= max_workers.
+    # As of Tesseract 4.1, 3 threads is the most effective on a 4 core/8 thread system.
+    if not os.environ.get('OMP_THREAD_LIMIT', '').isnumeric():
+        tess_threads = min(3, options.jobs // len(pdfinfo), len(pdfinfo))
+        os.environ['OMP_THREAD_LIMIT'] = str(tess_threads)
+    else:
+        tess_threads = int(os.environ['OMP_THREAD_LIMIT'])
+
+    if tess_threads > 1:
+        log.info("Using Tesseract OpenMP thread limit %d", tess_threads)
+
+
+class TesseractOcrEngine(OcrEngine):
+    @staticmethod
+    def version():
+        return tesseract.version()
+
+    @staticmethod
+    def creator_tag(options):
+        tag = '-PDF' if options.pdf_renderer == 'sandwich' else ''
+        return f"Tesseract OCR{tag} {TesseractOcrEngine.version()}"
+
+    def __str__(self):
+        return f"Tesseract OCR {TesseractOcrEngine.version()}"
+
+    @staticmethod
+    def languages(options):
+        return tesseract.get_languages()
+
+    @staticmethod
+    def get_orientation(input_file, options):
+        return tesseract.get_orientation(
+            input_file,
+            engine_mode=options.tesseract_oem,
+            timeout=options.tesseract_timeout,
+        )
+
+    @staticmethod
+    def generate_hocr(input_file, output_hocr, output_text, options):
+        tesseract.generate_hocr(
+            input_file=input_file,
+            output_hocr=output_hocr,
+            output_text=output_text,
+            languages=options.languages,
+            engine_mode=options.tesseract_oem,
+            tessconfig=options.tesseract_config,
+            timeout=options.tesseract_timeout,
+            pagesegmode=options.tesseract_pagesegmode,
+            user_words=options.user_words,
+            user_patterns=options.user_patterns,
+        )
+
+    @staticmethod
+    def generate_pdf(input_file, output_pdf, output_text, options):
+        tesseract.generate_pdf(
+            input_file=input_file,
+            output_pdf=output_pdf,
+            output_text=output_text,
+            languages=options.languages,
+            engine_mode=options.tesseract_oem,
+            tessconfig=options.tesseract_config,
+            timeout=options.tesseract_timeout,
+            pagesegmode=options.tesseract_pagesegmode,
+            user_words=options.user_words,
+            user_patterns=options.user_patterns,
+        )
+
+
+@hookimpl
+def get_ocr_engine():
+    return TesseractOcrEngine()

src/ocrmypdf/cli.py

@@ -17,8 +17,8 @@
 
 import argparse
 
-from . import PROGRAM_NAME, VERSION
-from ._plugins import check_plugin_loadable
+from ocrmypdf._version import PROGRAM_NAME as _PROGRAM_NAME
+from ocrmypdf._version import __version__ as _VERSION
 
 
 def numeric(basetype, min_=None, max_=None):
@@ -45,27 +45,43 @@ class ArgumentParser(argparse.ArgumentParser):
 
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
-        self.api_mode = False
+        self._api_mode = False
 
     def error(self, message):
-        if not self.api_mode:
+        if not self._api_mode:
             super().error(message)
             return
         raise ValueError(message)
 
 
-parser = ArgumentParser(
-    prog=PROGRAM_NAME,
-    fromfile_prefix_chars='@',
-    formatter_class=argparse.RawDescriptionHelpFormatter,
-    description="""\
+class LanguageSetAction(argparse.Action):
+    def __init__(self, option_strings, dest, default=None, **kwargs):
+        if default is None:
+            default = set()
+        super().__init__(option_strings, dest, default=default, **kwargs)
+
+    def __call__(self, parser, namespace, values, option_string=None):
+        dest = getattr(namespace, self.dest)
+        if '+' in values:
+            dest.update(lang for lang in values.split('+'))
+        else:
+            dest.add(values)
+
+
+def get_parser():
+    parser = ArgumentParser(
+        prog=_PROGRAM_NAME,
+        allow_abbrev=True,
+        fromfile_prefix_chars='@',
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        description="""\
 Generates a searchable PDF or PDF/A from a regular PDF.
 
 OCRmyPDF rasterizes each page of the input PDF, optionally corrects page
 rotation and performs image processing, runs the Tesseract OCR engine on the
 image, and then creates a PDF from the OCR information.
 """,
-    epilog="""\
+        epilog="""\
 OCRmyPDF attempts to keep the output file at about the same size.  If a file
 contains losslessly compressed images, and output file will be losslessly
 compressed as well.
@@ -106,381 +122,368 @@ Online documentation is located at:
     https://ocrmypdf.readthedocs.io/en/latest/introduction.html
 
 """,
-)
+    )
 
-parser.add_argument(
-    'input_file',
-    metavar="input_pdf_or_image",
-    help="PDF file containing the images to be OCRed (or '-' to read from "
-    "standard input)",
+    parser.add_argument(
+        'input_file',
+        metavar="input_pdf_or_image",
+        help="PDF file containing the images to be OCRed (or '-' to read from "
+        "standard input)",
+    )
+    parser.add_argument(
+        'output_file',
+        metavar="output_pdf",
+        help="Output searchable PDF file (or '-' to write to standard output). "
+        "Existing files will be ovewritten. If same as input file, the "
+        "input file will be updated only if processing is successful.",
+    )
+    parser.add_argument(
+        '-l',
+        '--language',
+        dest='languages',
+        action=LanguageSetAction,
+        help="Language(s) of the file to be OCRed (see tesseract --list-langs for "
+        "all language packs installed in your system). Use -l eng+deu for "
+        "multiple languages.",
+    )
+    parser.add_argument(
+        '--image-dpi',
+        metavar='DPI',
+        type=int,
+        help="For input image instead of PDF, use this DPI instead of file's.",
+    )
+    parser.add_argument(
+        '--output-type',
+        choices=['pdfa', 'pdf', 'pdfa-1', 'pdfa-2', 'pdfa-3'],
+        default='pdfa',
+        help="Choose output type. 'pdfa' creates a PDF/A-2b compliant file for "
+        "long term archiving (default, recommended) but may not suitable "
+        "for users who want their file altered as little as possible. 'pdfa' "
+        "also has problems with full Unicode text. 'pdf' attempts to "
+        "preserve file contents as much as possible. 'pdf-a1' creates a "
+        "PDF/A1-b file. 'pdf-a2' is equivalent to 'pdfa'. 'pdf-a3' creates a "
+        "PDF/A3-b file.",
+    )
+
+    # Use null string '\0' as sentinel to indicate the user supplied no argument,
+    # since that is the only invalid character for filepaths on all platforms
+    # bool('\0') is True in Python
+    parser.add_argument(
+        '--sidecar',
+        nargs='?',
+        const='\0',
+        default=None,
+        metavar='FILE',
+        help="Generate sidecar text files that contain the same text recognized "
+        "by Tesseract. This may be useful for building a OCR text database. "
+        "If FILE is omitted, the sidecar file be named {output_file}.txt "
+        "If FILE is set to '-', the sidecar is written to stdout (a "
+        "convenient way to preview OCR quality). The output file and sidecar "
+        "may not both use stdout at the same time.",
+    )
+
+    parser.add_argument(
+        '--version',
+        action='version',
+        version=_VERSION,
+        help="Print program version and exit",
+    )
+
+    jobcontrol = parser.add_argument_group("Job control options")
+    jobcontrol.add_argument(
+        '-j',
+        '--jobs',
+        metavar='N',
+        type=numeric(int, 0, 256),
+        help="Use up to N CPU cores simultaneously (default: use all).",
+    )
+    jobcontrol.add_argument(
+        '-q', '--quiet', action='store_true', help="Suppress INFO messages"
+    )
+    jobcontrol.add_argument(
+        '-v',
+        '--verbose',
+        type=numeric(int, 0, 2),
+        default=0,
+        const=1,
+        nargs='?',
+        help="Print more verbose messages for each additional verbose level. Use "
+        "`-v 1` typically for much more detailed logging. Higher numbers "
+        "are probably only useful in debugging.",
+    )
+    jobcontrol.add_argument(
+        '--no-progress-bar',
+        action='store_false',
+        dest='progress_bar',
+        help=argparse.SUPPRESS,
+    )
+    jobcontrol.add_argument(
+        '--use-threads', action='store_true', help=argparse.SUPPRESS
+    )
+
+    metadata = parser.add_argument_group(
+        "Metadata options",
+        "Set output PDF/A metadata (default: copy input document's metadata)",
+    )
+    metadata.add_argument(
+        '--title', type=str, help="Set document title (place multiple words in quotes)"
+    )
+    metadata.add_argument('--author', type=str, help="Set document author")
+    metadata.add_argument(
+        '--subject', type=str, help="Set document subject description"
+    )
+    metadata.add_argument('--keywords', type=str, help="Set document keywords")
+
+    preprocessing = parser.add_argument_group(
+        "Image preprocessing options",
+        "Options to improve the quality of the final PDF and OCR",
+    )
+    preprocessing.add_argument(
+        '-r',
+        '--rotate-pages',
+        action='store_true',
+        help="Automatically rotate pages based on detected text orientation",
+    )
+    preprocessing.add_argument(
+        '--remove-background',
+        action='store_true',
+        help="Attempt to remove background from gray or color pages, setting it "
+        "to white ",
+    )
+    preprocessing.add_argument(
+        '-d',
+        '--deskew',
+        action='store_true',
+        help="Deskew each page before performing OCR",
+    )
+    preprocessing.add_argument(
+        '-c',
+        '--clean',
+        action='store_true',
+        help="Clean pages from scanning artifacts before performing OCR, and send "
+        "the cleaned page to OCR, but do not include the cleaned page in "
+        "the output",
+    )
+    preprocessing.add_argument(
+        '-i',
+        '--clean-final',
+        action='store_true',
+        help="Clean page as above, and incorporate the cleaned image in the final "
+        "PDF.  Might remove desired content.",
+    )
+    preprocessing.add_argument(
+        '--unpaper-args',
+        type=str,
+        default=None,
+        help="A quoted string of arguments to pass to unpaper. Requires --clean. "
+        "Example: --unpaper-args '--layout double'.",
+    )
+    preprocessing.add_argument(
+        '--oversample',
+        metavar='DPI',
+        type=numeric(int, 0, 5000),
+        default=0,
+        help="Oversample images to at least the specified DPI, to improve OCR "
+        "results slightly",
+    )
+    preprocessing.add_argument(
+        '--remove-vectors',
+        action='store_true',
+        help="EXPERIMENTAL. Mask out any vector objects in the PDF so that they "
+        "will not be included in OCR. This can eliminate false characters.",
+    )
+    preprocessing.add_argument(
+        '--threshold',
+        action='store_true',
+        help=(
+            "EXPERIMENTAL. Threshold image to 1bpp before sending it to Tesseract "
+            "for OCR. Can improve OCR quality compared to Tesseract's thresholder."
+        ),
+    )
+
+    ocrsettings = parser.add_argument_group("OCR options", "Control how OCR is applied")
+    ocrsettings.add_argument(
+        '-f',
+        '--force-ocr',
+        action='store_true',
+        help="Rasterize any text or vector objects on each page, apply OCR, and "
+        "save the rastered output (this rewrites the PDF)",
+    )
+    ocrsettings.add_argument(
+        '-s',
+        '--skip-text',
+        action='store_true',
+        help="Skip OCR on any pages that already contain text, but include the "
+        "page in final output; useful for PDFs that contain a mix of "
+        "images, text pages, and/or previously OCRed pages",
+    )
+    ocrsettings.add_argument(
+        '--redo-ocr',
+        action='store_true',
+        help="Attempt to detect and remove the hidden OCR layer from files that "
+        "were previously OCRed with OCRmyPDF or another program. Apply OCR "
+        "to text found in raster images. Existing visible text objects will "
+        "not be changed. If there is no existing OCR, OCR will be added.",
+    )
+    ocrsettings.add_argument(
+        '--skip-big',
+        type=numeric(float, 0, 5000),
+        metavar='MPixels',
+        help="Skip OCR on pages larger than the specified amount of megapixels, "
+        "but include skipped pages in final output",
+    )
+
+    optimizing = parser.add_argument_group(
+        "Optimization options", "Control how the PDF is optimized after OCR"
+    )
+    optimizing.add_argument(
+        '-O',
+        '--optimize',
+        type=int,
+        choices=range(0, 4),
+        default=1,
+        help=(
+            "Control how PDF is optimized after processing:"
+            "0 - do not optimize; "
+            "1 - do safe, lossless optimizations (default); "
+            "2 - do some lossy optimizations; "
+            "3 - do aggressive lossy optimizations (including lossy JBIG2)"
+        ),
+    )
+    optimizing.add_argument(
+        '--jpeg-quality',
+        type=numeric(int, 0, 100),
+        default=0,
+        metavar='Q',
+        help=(
+            "Adjust JPEG quality level for JPEG optimization. "
+            "100 is best quality and largest output size; "
+            "1 is lowest quality and smallest output; "
+            "0 uses the default."
+        ),
+    )
+    optimizing.add_argument(
+        '--jpg-quality',
+        type=numeric(int, 0, 100),
+        default=0,
+        metavar='Q',
+        dest='jpeg_quality',
+        help=argparse.SUPPRESS,  # Alias for --jpeg-quality
+    )
+    optimizing.add_argument(
+        '--png-quality',
+        type=numeric(int, 0, 100),
+        default=0,
+        metavar='Q',
+        help=(
+            "Adjust PNG quality level to use when quantizing PNGs. "
+            "Values have same meaning as with --jpeg-quality"
+        ),
+    )
+    optimizing.add_argument(
+        '--jbig2-lossy',
+        action='store_true',
+        help=(
+            "Enable JBIG2 lossy mode (better compression, not suitable for some "
+            "use cases - see documentation)."
+        ),
+    )
+    optimizing.add_argument(
+        '--jbig2-page-group-size',
+        type=numeric(int, 1, 10000),
+        default=0,
+        metavar='N',
+        # Adjust number of pages to consider at once for JBIG2 compression
+        help=argparse.SUPPRESS,
+    )
+
+    advanced = parser.add_argument_group(
+        "Advanced", "Advanced options to control OCRmyPDF"
+    )
+    advanced.add_argument(
+        '--pages',
+        type=str,
+        help=(
+            "Limit OCR to the specified pages (ranges or comma separated), "
+            "skipping others"
+        ),
+    )
+    advanced.add_argument(
+        '--max-image-mpixels',
+        action='store',
+        type=numeric(float, 0),
+        metavar='MPixels',
+        help="Set maximum number of pixels to unpack before treating an image as a "
+        "decompression bomb",
+        default=128.0,
+    )
+    advanced.add_argument(
+        '--pdf-renderer',
+        choices=['auto', 'hocr', 'sandwich'],
+        default='auto',
+        help="Choose OCR PDF renderer - the default option is to let OCRmyPDF "
+        "choose.  See documentation for discussion.",
+    )
+    advanced.add_argument(
+        '--rotate-pages-threshold',
+        default=14.0,
+        type=numeric(float, 0, 1000),
+        metavar='CONFIDENCE',
+        help="Only rotate pages when confidence is above this value (arbitrary "
+        "units reported by tesseract)",
+    )
+    advanced.add_argument(
+        '--pdfa-image-compression',
+        choices=['auto', 'jpeg', 'lossless'],
+        default='auto',
+        help="Specify how to compress images in the output PDF/A. 'auto' lets "
+        "OCRmyPDF decide.  'jpeg' changes all grayscale and color images to "
+        "JPEG compression.  'lossless' uses PNG-style lossless compression "
+        "for all images.  Monochrome images are always compressed using a "
+        "lossless codec.  Compression settings "
+        "are applied to all pages, including those for which OCR was "
+        "skipped.  Not supported for --output-type=pdf ; that setting "
+        "preserves the original compression of all images.",
+    )
+    advanced.add_argument(
+        '--fast-web-view',
+        type=numeric(float, 0),
+        default=1.0,
+        metavar="MEGABYTES",
+        help="If the size of file is more than this threshold (in MB), then "
+        "linearize the PDF for fast web viewing. This allows the PDF to be "
+        "displayed before it is fully downloaded in web browsers, but increases "
+        "the space required slightly. By default we skip this for small files "
+        "which do not benefit. If the threshold is 0 it will be apply to all files. "
+        "Set the threshold very high to disable.",
+    )
+    advanced.add_argument(
+        '--plugin',
+        dest='plugins',
+        action='append',
+        default=[],
+        help="Name of plugin to import.",
+    )
+
+    debugging = parser.add_argument_group(
+        "Debugging", "Arguments to help with troubleshooting and debugging"
+    )
+    debugging.add_argument(
+        '-k',
+        '--keep-temporary-files',
+        action='store_true',
+        help="Keep temporary files (helpful for debugging)",
+    )
+    return parser
+
+
+plugins_only_parser = ArgumentParser(
+    prog=_PROGRAM_NAME, fromfile_prefix_chars='@', add_help=False, allow_abbrev=False
 )
-parser.add_argument(
-    'output_file',
-    metavar="output_pdf",
-    help="Output searchable PDF file (or '-' to write to standard output). "
-    "Existing files will be ovewritten. If same as input file, the "
-    "input file will be updated only if processing is successful.",
-)
-parser.add_argument(
-    '-l',
-    '--language',
+plugins_only_parser.add_argument(
+    '--plugin',
+    dest='plugins',
     action='append',
-    help="Language(s) of the file to be OCRed (see tesseract --list-langs for "
-    "all language packs installed in your system). Use -l eng+deu for "
-    "multiple languages.",
-)
-parser.add_argument(
-    '--image-dpi',
-    metavar='DPI',
-    type=int,
-    help="For input image instead of PDF, use this DPI instead of file's.",
-)
-parser.add_argument(
-    '--output-type',
-    choices=['pdfa', 'pdf', 'pdfa-1', 'pdfa-2', 'pdfa-3'],
-    default='pdfa',
-    help="Choose output type. 'pdfa' creates a PDF/A-2b compliant file for "
-    "long term archiving (default, recommended) but may not suitable "
-    "for users who want their file altered as little as possible. 'pdfa' "
-    "also has problems with full Unicode text. 'pdf' attempts to "
-    "preserve file contents as much as possible. 'pdf-a1' creates a "
-    "PDF/A1-b file. 'pdf-a2' is equivalent to 'pdfa'. 'pdf-a3' creates a "
-    "PDF/A3-b file.",
-)
-
-# Use null string '\0' as sentinel to indicate the user supplied no argument,
-# since that is the only invalid character for filepaths on all platforms
-# bool('\0') is True in Python
-parser.add_argument(
-    '--sidecar',
-    nargs='?',
-    const='\0',
-    default=None,
-    metavar='FILE',
-    help="Generate sidecar text files that contain the same text recognized "
-    "by Tesseract. This may be useful for building a OCR text database. "
-    "If FILE is omitted, the sidecar file be named {output_file}.txt "
-    "If FILE is set to '-', the sidecar is written to stdout (a "
-    "convenient way to preview OCR quality). The output file and sidecar "
-    "may not both use stdout at the same time.",
-)
-
-parser.add_argument(
-    '--version',
-    action='version',
-    version=VERSION,
-    help="Print program version and exit",
-)
-
-jobcontrol = parser.add_argument_group("Job control options")
-jobcontrol.add_argument(
-    '-j',
-    '--jobs',
-    metavar='N',
-    type=numeric(int, 0, 256),
-    help="Use up to N CPU cores simultaneously (default: use all).",
-)
-jobcontrol.add_argument(
-    '-q', '--quiet', action='store_true', help="Suppress INFO messages"
-)
-jobcontrol.add_argument(
-    '-v',
-    '--verbose',
-    type=int,
-    default=0,
-    nargs='?',
-    help="Print more verbose messages for each additional verbose level. Use "
-    "`-v 1` typically for much more detailed logging. Higher numbers "
-    "are probably only useful in debugging.",
-)
-jobcontrol.add_argument(
-    '--no-progress-bar',
-    action='store_false',
-    dest='progress_bar',
-    help=argparse.SUPPRESS,
-)
-jobcontrol.add_argument('--use-threads', action='store_true', help=argparse.SUPPRESS)
-
-metadata = parser.add_argument_group(
-    "Metadata options",
-    "Set output PDF/A metadata (default: copy input document's metadata)",
-)
-metadata.add_argument(
-    '--title', type=str, help="Set document title (place multiple words in quotes)"
-)
-metadata.add_argument('--author', type=str, help="Set document author")
-metadata.add_argument('--subject', type=str, help="Set document subject description")
-metadata.add_argument('--keywords', type=str, help="Set document keywords")
-
-preprocessing = parser.add_argument_group(
-    "Image preprocessing options",
-    "Options to improve the quality of the final PDF and OCR",
-)
-preprocessing.add_argument(
-    '-r',
-    '--rotate-pages',
-    action='store_true',
-    help="Automatically rotate pages based on detected text orientation",
-)
-preprocessing.add_argument(
-    '--remove-background',
-    action='store_true',
-    help="Attempt to remove background from gray or color pages, setting it "
-    "to white ",
-)
-preprocessing.add_argument(
-    '-d', '--deskew', action='store_true', help="Deskew each page before performing OCR"
-)
-preprocessing.add_argument(
-    '-c',
-    '--clean',
-    action='store_true',
-    help="Clean pages from scanning artifacts before performing OCR, and send "
-    "the cleaned page to OCR, but do not include the cleaned page in "
-    "the output",
-)
-preprocessing.add_argument(
-    '-i',
-    '--clean-final',
-    action='store_true',
-    help="Clean page as above, and incorporate the cleaned image in the final "
-    "PDF.  Might remove desired content.",
-)
-preprocessing.add_argument(
-    '--unpaper-args',
-    type=str,
-    default=None,
-    help="A quoted string of arguments to pass to unpaper. Requires --clean. "
-    "Example: --unpaper-args '--layout double'.",
-)
-preprocessing.add_argument(
-    '--oversample',
-    metavar='DPI',
-    type=numeric(int, 0, 5000),
-    default=0,
-    help="Oversample images to at least the specified DPI, to improve OCR "
-    "results slightly",
-)
-preprocessing.add_argument(
-    '--remove-vectors',
-    action='store_true',
-    help="EXPERIMENTAL. Mask out any vector objects in the PDF so that they "
-    "will not be included in OCR. This can eliminate false characters.",
-)
-preprocessing.add_argument(
-    '--threshold',
-    action='store_true',
-    help="EXPERIMENTAL. Threshold image to 1bpp before sending it to Tesseract for OCR. Can "
-    "improve OCR quality compared to Tesseract's thresholder.",
-)
-
-ocrsettings = parser.add_argument_group("OCR options", "Control how OCR is applied")
-ocrsettings.add_argument(
-    '-f',
-    '--force-ocr',
-    action='store_true',
-    help="Rasterize any text or vector objects on each page, apply OCR, and "
-    "save the rastered output (this rewrites the PDF)",
-)
-ocrsettings.add_argument(
-    '-s',
-    '--skip-text',
-    action='store_true',
-    help="Skip OCR on any pages that already contain text, but include the "
-    "page in final output; useful for PDFs that contain a mix of "
-    "images, text pages, and/or previously OCRed pages",
-)
-ocrsettings.add_argument(
-    '--redo-ocr',
-    action='store_true',
-    help="Attempt to detect and remove the hidden OCR layer from files that "
-    "were previously OCRed with OCRmyPDF or another program. Apply OCR "
-    "to text found in raster images. Existing visible text objects will "
-    "not be changed. If there is no existing OCR, OCR will be added.",
-)
-ocrsettings.add_argument(
-    '--skip-big',
-    type=numeric(float, 0, 5000),
-    metavar='MPixels',
-    help="Skip OCR on pages larger than the specified amount of megapixels, "
-    "but include skipped pages in final output",
-)
-
-optimizing = parser.add_argument_group(
-    "Optimization options", "Control how the PDF is optimized after OCR"
-)
-optimizing.add_argument(
-    '-O',
-    '--optimize',
-    type=int,
-    choices=range(0, 4),
-    default=1,
-    help=(
-        "Control how PDF is optimized after processing:"
-        "0 - do not optimize; "
-        "1 - do safe, lossless optimizations (default); "
-        "2 - do some lossy optimizations; "
-        "3 - do aggressive lossy optimizations (including lossy JBIG2)"
-    ),
-)
-optimizing.add_argument(
-    '--jpeg-quality',
-    type=numeric(int, 0, 100),
-    default=0,
-    metavar='Q',
-    help=(
-        "Adjust JPEG quality level for JPEG optimization. "
-        "100 is best quality and largest output size; "
-        "1 is lowest quality and smallest output; "
-        "0 uses the default."
-    ),
-)
-optimizing.add_argument(
-    '--jpg-quality',
-    type=numeric(int, 0, 100),
-    default=0,
-    metavar='Q',
-    dest='jpeg_quality',
-    help=argparse.SUPPRESS,  # Alias for --jpeg-quality
-)
-optimizing.add_argument(
-    '--png-quality',
-    type=numeric(int, 0, 100),
-    default=0,
-    metavar='Q',
-    help=(
-        "Adjust PNG quality level to use when quantizing PNGs. "
-        "Values have same meaning as with --jpeg-quality"
-    ),
-)
-optimizing.add_argument(
-    '--jbig2-lossy',
-    action='store_true',
-    help=(
-        "Enable JBIG2 lossy mode (better compression, not suitable for some "
-        "use cases - see documentation)."
-    ),
-)
-optimizing.add_argument(
-    '--jbig2-page-group-size',
-    type=numeric(int, 1, 10000),
-    default=0,
-    metavar='N',
-    # Adjust number of pages to consider at once for JBIG2 compression
-    help=argparse.SUPPRESS,
-)
-
-advanced = parser.add_argument_group(
-    "Advanced", "Advanced options to control Tesseract's OCR behavior"
-)
-advanced.add_argument(
-    '--pages',
-    type=str,
-    help="Limit OCR to the specified pages (ranges or comma separated), skipping others",
-)
-advanced.add_argument(
-    '--max-image-mpixels',
-    action='store',
-    type=numeric(float, 0),
-    metavar='MPixels',
-    help="Set maximum number of pixels to unpack before treating an image as a "
-    "decompression bomb",
-    default=128.0,
-)
-advanced.add_argument(
-    '--tesseract-config',
-    action='append',
-    metavar='CFG',
     default=[],
-    help="Additional Tesseract configuration files -- see documentation",
+    help="Name of plugin to import.",
 )
-advanced.add_argument(
-    '--tesseract-pagesegmode',
-    action='store',
-    type=int,
-    metavar='PSM',
-    choices=range(0, 14),
-    help="Set Tesseract page segmentation mode (see tesseract --help)",
-)
-advanced.add_argument(
-    '--tesseract-oem',
-    action='store',
-    type=int,
-    metavar='MODE',
-    choices=range(0, 4),
-    help=(
-        "Set Tesseract 4.0 OCR engine mode: "
-        "0 - original Tesseract only; "
-        "1 - neural nets LSTM only; "
-        "2 - Tesseract + LSTM; "
-        "3 - default."
-    ),
-)
-advanced.add_argument(
-    '--pdf-renderer',
-    choices=['auto', 'hocr', 'sandwich'],
-    default='auto',
-    help="Choose OCR PDF renderer - the default option is to let OCRmyPDF "
-    "choose.  See documentation for discussion.",
-)
-advanced.add_argument(
-    '--tesseract-timeout',
-    default=180.0,
-    type=numeric(float, 0),
-    metavar='SECONDS',
-    help='Give up on OCR after the timeout, but copy the preprocessed page '
-    'into the final output',
-)
-advanced.add_argument(
-    '--rotate-pages-threshold',
-    default=14.0,
-    type=numeric(float, 0, 1000),
-    metavar='CONFIDENCE',
-    help="Only rotate pages when confidence is above this value (arbitrary "
-    "units reported by tesseract)",
-)
-advanced.add_argument(
-    '--pdfa-image-compression',
-    choices=['auto', 'jpeg', 'lossless'],
-    default='auto',
-    help="Specify how to compress images in the output PDF/A. 'auto' lets "
-    "OCRmyPDF decide.  'jpeg' changes all grayscale and color images to "
-    "JPEG compression.  'lossless' uses PNG-style lossless compression "
-    "for all images.  Monochrome images are always compressed using a "
-    "lossless codec.  Compression settings "
-    "are applied to all pages, including those for which OCR was "
-    "skipped.  Not supported for --output-type=pdf ; that setting "
-    "preserves the original compression of all images.",
-)
-advanced.add_argument(
-    '--user-words',
-    metavar='FILE',
-    help="Specify the location of the Tesseract user words file. This is a "
-    "list of words Tesseract should consider while performing OCR in "
-    "addition to its standard language dictionaries. This can improve "
-    "OCR quality especially for specialized and technical documents.",
-)
-advanced.add_argument(
-    '--user-patterns',
-    metavar='FILE',
-    help="Specify the location of the Tesseract user patterns file.",
-)
-
-plugins = parser.add_argument_group("Filters and Plugins", argparse.SUPPRESS)
-plugins.add_argument(
-    '--filter-ocr-image', help=argparse.SUPPRESS, type=check_plugin_loadable
-)
-plugins.add_argument(
-    '--plugin-validation', help=argparse.SUPPRESS, type=check_plugin_loadable
-)
-
-debugging = parser.add_argument_group(
-    "Debugging", "Arguments to help with troubleshooting and debugging"
-)
-debugging.add_argument(
-    '-k',
-    '--keep-temporary-files',
-    action='store_true',
-    help="Keep temporary files (helpful for debugging)",
-)
-debugging.add_argument('--tesseract-env', type=str, help=argparse.SUPPRESS)

src/ocrmypdf/exec/ghostscript.py

@@ -1,298 +0,0 @@
-# © 2017 James R. Barlow: github.com/jbarlow83
-#
-# This file is part of OCRmyPDF.
-#
-# OCRmyPDF is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# OCRmyPDF is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
-
-import logging
-import re
-from functools import lru_cache
-from os import fspath
-from shutil import copy
-from subprocess import PIPE, STDOUT, run
-from tempfile import NamedTemporaryFile
-
-from PIL import Image
-
-from ..exceptions import SubprocessOutputError
-from . import get_version
-
-
-gslog = logging.getLogger()
-
-
-@lru_cache(maxsize=1)
-def version():
-    return get_version('gs')
-
-
-def jpeg_passthrough_available():
-    """Returns True if the installed version of Ghostscript supports JPEG passthru
-
-    Prior to 9.23, Ghostscript decode and re-encoded JPEGs internally. In 9.23
-    it gained the ability to keep JPEGs unmodified. However, the 9.23
-    implementation was buggy and would deletes the last two bytes of images in
-    some cases, as reported here.
-    https://bugs.ghostscript.com/show_bug.cgi?id=699216
-
-    The issue was fixed for 9.24, hence that is the first version we consider
-    the feature available. (However, we don't use 9.24 at all, so the first
-    version that allows JPEG passthrough is 9.25.
-
-    """
-    return version() >= '9.24'
-
-
-def _gs_error_reported(stream):
-    return re.search(r'error', stream, flags=re.IGNORECASE)
-
-
-def extract_text(input_file, pageno=1):
-    """Use the txtwrite device to get text layout information out
-
-    For details on options of -dTextFormat see
-    https://www.ghostscript.com/doc/current/VectorDevices.htm#TXT
-
-    Format is like
-    <page>
-    <line>
-    <span bbox="left top right bottom" font="..." size="...">
-    <char bbox="...." c="X"/>
-
-    :param pageno: number of page to extract, or all pages if None
-    :return: XML-ish text representation in bytes
-    """
-
-    if pageno is not None:
-        pages = ['-dFirstPage=%i' % pageno, '-dLastPage=%i' % pageno]
-    else:
-        pages = []
-
-    args_gs = (
-        [
-            'gs',
-            '-dQUIET',
-            '-dSAFER',
-            '-dBATCH',
-            '-dNOPAUSE',
-            '-sDEVICE=txtwrite',
-            '-dTextFormat=0',
-        ]
-        + pages
-        + ['-o', '-', fspath(input_file)]
-    )
-
-    p = run(args_gs, stdout=PIPE, stderr=PIPE)
-    if p.returncode != 0:
-        raise SubprocessOutputError(
-            'Ghostscript text extraction failed\n%s\n%s\n%s'
-            % (input_file, p.stdout.decode(), p.stderr.decode())
-        )
-
-    return p.stdout
-
-
-def rasterize_pdf(
-    input_file,
-    output_file,
-    xres,
-    yres,
-    raster_device,
-    log,
-    pageno=1,
-    page_dpi=None,
-    rotation=None,
-    filter_vector=False,
-):
-    """Rasterize one page of a PDF at resolution (xres, yres) in canvas units.
-
-    The image is sized to match the integer pixels dimensions implied by
-    (xres, yres) even if those numbers are noninteger. The image's DPI will
-     be overridden with the values in page_dpi.
-
-    :param input_file: pathlike
-    :param output_file: pathlike
-    :param xres: resolution at which to rasterize page
-    :param yres:
-    :param raster_device:
-    :param log:
-    :param pageno: page number to rasterize (beginning at page 1)
-    :param page_dpi: resolution tuple (x, y) overriding output image DPI
-    :param rotation: 0, 90, 180, 270: clockwise angle to rotate page
-    :param filter_vector: if True, remove vector graphics objects
-    :return:
-    """
-    res = round(xres, 6), round(yres, 6)
-    if not page_dpi:
-        page_dpi = res
-    if not log:
-        log = gslog
-
-    with NamedTemporaryFile(delete=True) as tmp:
-        args_gs = (
-            [
-                'gs',
-                '-dQUIET',
-                '-dSAFER',
-                '-dBATCH',
-                '-dNOPAUSE',
-                f'-sDEVICE={raster_device}',
-                f'-dFirstPage={pageno}',
-                f'-dLastPage={pageno}',
-                f'-r{res[0]:f}x{res[1]:f}',
-            ]
-            + (['-dFILTERVECTOR'] if filter_vector else [])
-            + [
-                '-o',
-                tmp.name,
-                '-dAutoRotatePages=/None',  # Probably has no effect on raster
-                '-f',
-                fspath(input_file),
-            ]
-        )
-
-        log.debug(args_gs)
-        p = run(args_gs, stdout=PIPE, stderr=STDOUT, universal_newlines=True)
-        if _gs_error_reported(p.stdout):
-            log.error(p.stdout)
-        else:
-            log.debug(p.stdout)
-
-        if p.returncode != 0:
-            raise SubprocessOutputError('Ghostscript rasterizing failed')
-
-        tmp.seek(0)
-        with Image.open(tmp) as im:
-            if rotation is not None:
-                log.debug("Rotating output by %i", rotation)
-                # rotation is a clockwise angle and Image.ROTATE_* is
-                # counterclockwise so this cancels out the rotation
-                if rotation == 90:
-                    im = im.transpose(Image.ROTATE_90)
-                elif rotation == 180:
-                    im = im.transpose(Image.ROTATE_180)
-                elif rotation == 270:
-                    im = im.transpose(Image.ROTATE_270)
-                if rotation % 180 == 90:
-                    page_dpi = page_dpi[1], page_dpi[0]
-            im.save(fspath(output_file), dpi=page_dpi)
-
-
-def generate_pdfa(
-    pdf_pages,
-    output_file,
-    compression,
-    log,
-    threads=1,
-    pdf_version='1.5',
-    pdfa_part='2',
-):
-    """Generate a PDF/A.
-
-    The pdf_pages, a list files, will be merged into output_file. One or more
-    PDF files may be merged. One of the files in this list must be a pdfmark
-    file that provides Ghostscript with details on how to perform the PDF/A
-    conversion. By default with we pick PDF/A-2b, but this works for 1 or 3.
-
-    compression can be 'jpeg', 'lossless', or an empty string. In 'jpeg',
-    Ghostscript is instructed to convert color and grayscale images to DCT
-    (JPEG encoding). In 'lossless' Ghostscript is told to convert images to
-    Flate (lossless/PNG). If the parameter is omitted Ghostscript is left to
-    make its own decisions about how to encode images; it appears to use a
-    heuristic to decide how to encode images. As of Ghostscript 9.25, we
-    support passthrough JPEG which allows Ghostscript to avoid transcoding
-    images entirely. (The feature was added in 9.23 but broken, and the 9.24
-    release of Ghostscript had regressions, so we don't support it until 9.25.)
-    """
-    if not log:
-        log = gslog
-
-    compression_args = []
-    if compression == 'jpeg':
-        compression_args = [
-            "-dAutoFilterColorImages=false",
-            "-dColorImageFilter=/DCTEncode",
-            "-dAutoFilterGrayImages=false",
-            "-dGrayImageFilter=/DCTEncode",
-        ]
-    elif compression == 'lossless':
-        compression_args = [
-            "-dAutoFilterColorImages=false",
-            "-dColorImageFilter=/FlateEncode",
-            "-dAutoFilterGrayImages=false",
-            "-dGrayImageFilter=/FlateEncode",
-        ]
-    else:
-        compression_args = [
-            "-dAutoFilterColorImages=true",
-            "-dAutoFilterGrayImages=true",
-        ]
-
-    # Older versions of Ghostscript expect a leading slash in
-    # sColorConversionStrategy, newer ones should not have it. See Ghostscript
-    # git commit fe1c025d.
-    strategy = 'RGB' if version() >= '9.19' else '/RGB'
-
-    if version() == '9.23':
-        # 9.23: new feature JPEG passthrough is broken in some cases, best to
-        # disable it always
-        # https://bugs.ghostscript.com/show_bug.cgi?id=699216
-        compression_args.append('-dPassThroughJPEGImages=false')
-
-    with NamedTemporaryFile(delete=True) as gs_pdf:
-        # nb no need to specify ProcessColorModel when ColorConversionStrategy
-        # is set; see:
-        # https://bugs.ghostscript.com/show_bug.cgi?id=699392
-        args_gs = (
-            [
-                "gs",
-                "-dQUIET",
-                "-dBATCH",
-                "-dNOPAUSE",
-                "-dCompatibilityLevel=" + str(pdf_version),
-                "-sDEVICE=pdfwrite",
-                "-dAutoRotatePages=/None",
-                "-sColorConversionStrategy=" + strategy,
-            ]
-            + compression_args
-            + [
-                "-dJPEGQ=95",
-                "-dPDFA=" + pdfa_part,
-                "-dPDFACompatibilityPolicy=1",
-                "-sOutputFile=" + gs_pdf.name,
-            ]
-        )
-        args_gs.extend(fspath(s) for s in pdf_pages)  # Stringify Path objs
-        log.debug(args_gs)
-        p = run(args_gs, stdout=PIPE, stderr=STDOUT, universal_newlines=True)
-
-        if _gs_error_reported(p.stdout):
-            log.error(p.stdout)
-        elif 'overprint mode not set' in p.stdout:
-            # Unless someone is going to print PDF/A documents on a
-            # magical sRGB printer I can't see the removal of overprinting
-            # being a problem....
-            log.debug(
-                "Ghostscript had to remove PDF 'overprinting' from the "
-                "input file to complete PDF/A conversion. "
-            )
-        else:
-            log.debug(p.stdout)
-
-        if p.returncode == 0:
-            # Ghostscript does not change return code when it fails to create
-            # PDF/A - check PDF/A status elsewhere
-            copy(gs_pdf.name, fspath(output_file))
-        else:
-            raise SubprocessOutputError('Ghostscript PDF/A rendering failed')

src/ocrmypdf/exec/qpdf.py

@@ -1,49 +0,0 @@
-# © 2017 James R. Barlow: github.com/jbarlow83
-#
-# This file is part of OCRmyPDF.
-#
-# OCRmyPDF is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# OCRmyPDF is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
-
-from functools import lru_cache
-from os import fspath
-from subprocess import PIPE, STDOUT, CalledProcessError, run
-
-from . import get_version
-
-
-@lru_cache(maxsize=1)
-def version():
-    return get_version('qpdf', regex=r'qpdf version (.+)')
-
-
-def check(input_file, log=None):
-    args_qpdf = ['qpdf', '--check', fspath(input_file)]
-
-    if log is None:
-        import logging as log
-
-    try:
-        run(args_qpdf, stderr=STDOUT, stdout=PIPE, universal_newlines=True, check=True)
-    except CalledProcessError as e:
-        if e.returncode == 2:
-            log.error("%s: not a valid PDF, and could not repair it.", input_file)
-            log.error("Details:")
-            log.error(e.output)
-        elif e.returncode == 3:
-            log.info("qpdf --check returned warnings:")
-            log.info(e.output)
-        else:
-            log.warning(e.output)
-        return False
-    return True

src/ocrmypdf/filters.py

@@ -1,10 +0,0 @@
-from PIL import Image
-import PIL.ImageOps
-
-
-def invert(im):
-    return PIL.ImageOps.invert(im.convert('L'))
-
-
-def whiteout(im):
-    return Image.new(im.mode, im.size)

src/ocrmypdf/helpers.py

@@ -18,24 +18,58 @@
 import logging
 import multiprocessing
 import os
+import shutil
 import warnings
+from collections import namedtuple
 from collections.abc import Iterable
 from contextlib import suppress
 from functools import wraps
+from io import StringIO
+from math import isclose
 from pathlib import Path
+from typing import Any, Sequence
+
+import pikepdf
 
 log = logging.getLogger(__name__)
 
 
-def re_symlink(input_file, soft_link_name, *args, **kwargs):
+class Resolution(namedtuple('Resolution', ('x', 'y'))):
+    __slots__ = ()
+
+    def round(self, ndigits):
+        return Resolution(round(self.x, ndigits), round(self.y, ndigits))
+
+    def to_int(self):
+        return Resolution(int(round(self.x)), int(round(self.y)))
+
+    @property
+    def is_square(self):
+        return isclose(self.x, self.y, rel_tol=1e-3)
+
+    def take_max(self, vals, yvals=None):
+        if yvals is not None:
+            return Resolution(max(self.x, *vals), max(self.y, *yvals))
+        max_x, max_y = self.x, self.y
+        for x, y in vals:
+            max_x = max(x, max_x)
+            max_y = max(y, max_y)
+        return Resolution(max_x, max_y)
+
+    def flip_axis(self):
+        return Resolution(self.y, self.x)
+
+    def __str__(self):
+        return f"{self.x:f}x{self.y:f}"
+
+    def __repr__(self):
+        return f"Resolution({self.x}x{self.y} dpi)"
+
+
+def safe_symlink(input_file: os.PathLike, soft_link_name: os.PathLike):
     """
     Helper function: relinks soft symbolic link if necessary
     """
-    if len(args) == 1 and isinstance(args[0], logging.Logger):
-        log.warning("Deprecated: re_symlink(,log)")
-    if 'log' in kwargs:
-        log.warning('Deprecated: re_symlink(...log=)')
-
     input_file = os.fspath(input_file)
     soft_link_name = os.fspath(soft_link_name)
 
@@ -60,27 +94,39 @@ def re_symlink(input_file, soft_link_name, *args, **kwargs):
     if not os.path.exists(input_file):
         raise FileNotFoundError(f"trying to create a broken symlink to {input_file}")
 
+    if os.name == 'nt':
+        # Don't actually use symlinks on Windows due to permission issues
+        shutil.copyfile(input_file, soft_link_name)
+        return
+
     log.debug("os.symlink(%s, %s)", input_file, soft_link_name)
 
     # Create symbolic link using absolute path
     os.symlink(os.path.abspath(input_file), soft_link_name)
 
 
-def is_iterable_notstr(thing):
+def samefile(f1: os.PathLike, f2: os.PathLike):
+    if os.name == 'nt':
+        return f1 == f2
+    else:
+        return os.path.samefile(f1, f2)
+
+
+def is_iterable_notstr(thing: Any) -> bool:
     return isinstance(thing, Iterable) and not isinstance(thing, str)
 
 
-def monotonic(L):
+def monotonic(L: Sequence) -> bool:
     """Does list increase monotonically?"""
     return all(b > a for a, b in zip(L, L[1:]))
 
 
-def page_number(input_file):
+def page_number(input_file: os.PathLike) -> int:
     """Get one-based page number implied by filename (000002.pdf -> 2)"""
     return int(os.path.basename(os.fspath(input_file))[0:6])
 
 
-def available_cpu_count():
+def available_cpu_count() -> int:
     try:
         return multiprocessing.cpu_count()
     except NotImplementedError:
@@ -91,35 +137,73 @@ def available_cpu_count():
     return 1
 
 
-def is_file_writable(test_file):
+def is_file_writable(test_file: os.PathLike) -> bool:
     """Intentionally racy test if target is writable.
 
     We intend to write to the output file if and only if we succeed and
     can replace it atomically. Before doing the OCR work, make sure
     the location is writable.
     """
-    p = Path(test_file)
+    try:
+        p = Path(test_file)
+        if p.is_symlink():
+            p = p.resolve(strict=False)
 
-    if p.is_symlink():
-        p = p.resolve(strict=False)
-
-    # p.is_file() throws an exception in some cases
-    if p.exists() and p.is_file():
-        return os.access(
-            os.fspath(p),
-            os.W_OK,
-            effective_ids=(os.access in os.supports_effective_ids),
-        )
-    else:
-        try:
-            fp = p.open('wb')
-        except OSError:
-            return False
+        # p.is_file() throws an exception in some cases
+        if p.exists() and p.is_file():
+            return os.access(
+                os.fspath(p),
+                os.W_OK,
+                effective_ids=(os.access in os.supports_effective_ids),
+            )
         else:
-            fp.close()
-            with suppress(OSError):
-                p.unlink()
-        return True
+            try:
+                fp = p.open('wb')
+            except OSError:
+                return False
+            else:
+                fp.close()
+                with suppress(OSError):
+                    p.unlink()
+            return True
+    except (EnvironmentError, RuntimeError) as e:
+        log.debug(e)
+        log.error(str(e))
+        return False
+
+
+def check_pdf(input_file: Path) -> bool:
+    pdf = None
+    try:
+        pdf = pikepdf.open(input_file)
+    except pikepdf.PdfError as e:
+        log.error(e)
+        return False
+    else:
+        messages = pdf.check()
+        for msg in messages:
+            if 'error' in msg.lower():
+                log.error(msg)
+            else:
+                log.warning(msg)
+
+        sio = StringIO()
+        linearize = None
+        try:
+            pdf.check_linearization(sio)
+        except RuntimeError:
+            pass
+        else:
+            linearize = sio.getvalue()
+            if linearize:
+                log.warning(linearize)
+
+        if not messages and not linearize:
+            return True
+        return False
+    finally:
+        if pdf:
+            pdf.close()
 
 
 def deprecated(func):

src/ocrmypdf/hocrtransform.py

@@ -29,9 +29,12 @@
 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 import argparse
+import os
 import re
 from collections import namedtuple
 from math import atan, cos, sin
+from pathlib import Path
+from typing import Union
 from xml.etree import ElementTree
 
 from reportlab.lib.units import inch
@@ -64,9 +67,9 @@ class HocrTransform:
         {'ff': 'ff', 'ffi': 'f‌f‌i', 'ffl': 'f‌f‌l', 'fi': 'fi', 'fl': 'fl'}
     )
 
-    def __init__(self, hocrFileName, dpi):
+    def __init__(self, hocr_filename: Union[str, Path], dpi: float):
         self.dpi = dpi
-        self.hocr = ElementTree.parse(hocrFileName)
+        self.hocr = ElementTree.parse(os.fspath(hocr_filename))
 
         # if the hOCR file has a namespace, ElementTree requires its use to
         # find elements
@@ -88,7 +91,7 @@ class HocrTransform:
         if self.width is None or self.height is None:
             raise HocrTransformError("hocr file is missing page dimensions")
 
-    def __str__(self):
+    def __str__(self):  # pragma: no cover
         """
         Return the textual content of the HTML body
         """
@@ -114,12 +117,12 @@ class HocrTransform:
         return text
 
     @classmethod
-    def element_coordinates(cls, element):
+    def element_coordinates(cls, element) -> Rect:
         """
         Returns a tuple containing the coordinates of the bounding box around
         an element
         """
-        out = (0, 0, 0, 0)
+        out = Rect._make(0 for _ in range(4))
         if 'title' in element.attrib:
             matches = cls.box_pattern.search(element.attrib['title'])
             if matches:
@@ -136,7 +139,7 @@ class HocrTransform:
             matches = cls.baseline_pattern.search(element.attrib['title'])
             if matches:
                 return float(matches.group(1)), int(matches.group(2))
-        return (0, 0)
+        return (0.0, 0.0)
 
     def pt_from_pixel(self, pxl):
         """
@@ -145,7 +148,7 @@ class HocrTransform:
         return Rect._make((c / self.dpi * inch) for c in pxl)
 
     @classmethod
-    def replace_unsupported_chars(cls, s):
+    def replace_unsupported_chars(cls, s: str):
         """
         Given an input string, returns the corresponding string that:
         - is available in the helvetica facetype
@@ -155,12 +158,12 @@ class HocrTransform:
 
     def to_pdf(
         self,
-        outFileName,
-        imageFileName=None,
-        showBoundingboxes=False,
-        fontname="Helvetica",
-        invisibleText=False,
-        interwordSpaces=False,
+        out_filename: Path,
+        image_filename: Path = None,
+        show_bounding_boxes: bool = False,
+        fontname: str = "Helvetica",
+        invisible_text: bool = False,
+        interword_spaces: bool = False,
     ):
         """
         Creates a PDF file with an image superimposed on top of the text.
@@ -172,7 +175,11 @@ class HocrTransform:
         """
         # create the PDF file
         # page size in points (1/72 in.)
-        pdf = Canvas(outFileName, pagesize=(self.width, self.height), pageCompression=1)
+        pdf = Canvas(
+            os.fspath(out_filename),
+            pagesize=(self.width, self.height),
+            pageCompression=1,
+        )
 
         # draw bounding box for each paragraph
         # light blue for bounding box of paragraph
@@ -190,7 +197,7 @@ class HocrTransform:
             pt = self.pt_from_pixel(pxl_coords)
 
             # draw the bbox border
-            if showBoundingboxes:
+            if show_bounding_boxes:  # pragma: no cover
                 pdf.rect(
                     pt.x1, self.height - pt.y2, pt.x2 - pt.x1, pt.y2 - pt.y1, fill=1
                 )
@@ -205,9 +212,9 @@ class HocrTransform:
                 line,
                 "ocrx_word",
                 fontname,
-                invisibleText,
-                interwordSpaces,
-                showBoundingboxes,
+                invisible_text,
+                interword_spaces,
+                show_bounding_boxes,
             )
 
         if not found_lines:
@@ -218,31 +225,33 @@ class HocrTransform:
                 root,
                 "ocrx_word",
                 fontname,
-                invisibleText,
-                interwordSpaces,
-                showBoundingboxes,
+                invisible_text,
+                interword_spaces,
+                show_bounding_boxes,
             )
         # put the image on the page, scaled to fill the page
-        if imageFileName is not None:
-            pdf.drawImage(imageFileName, 0, 0, width=self.width, height=self.height)
+        if image_filename is not None:
+            pdf.drawImage(
+                os.fspath(image_filename), 0, 0, width=self.width, height=self.height
+            )
 
         # finish up the page and save it
         pdf.showPage()
         pdf.save()
 
     @classmethod
-    def polyval(cls, poly, x):
+    def polyval(cls, poly, x):  # pragma: no cover
         return x * poly[0] + poly[1]
 
     def _do_line(
         self,
-        pdf,
+        pdf: Canvas,
         line,
-        elemclass,
-        fontname,
-        invisibleText,
-        interwordSpaces,
-        showBoundingboxes,
+        elemclass: str,
+        fontname: str,
+        invisible_text: bool,
+        interword_spaces: bool,
+        show_bounding_boxes: bool,
     ):
         pxl_line_coords = self.element_coordinates(line)
         line_box = self.pt_from_pixel(pxl_line_coords)
@@ -262,14 +271,14 @@ class HocrTransform:
         # on a sloped baseline and the edge of the bounding box.
         fontsize = (line_height - abs(intercept)) / cos_a
         text.setFont(fontname, fontsize)
-        if invisibleText:
+        if invisible_text:
             text.setTextRenderMode(3)  # Invisible (indicates OCR text)
 
         # Intercept is normally negative, so this places it above the bottom
         # of the line box
         baseline_y2 = self.height - (line_box.y2 + intercept)
 
-        if showBoundingboxes:
+        if show_bounding_boxes:  # pragma: no cover
             # draw the baseline in magenta, dashed
             pdf.setDash()
             pdf.setStrokeColorRGB(0.95, 0.65, 0.95)
@@ -298,7 +307,7 @@ class HocrTransform:
 
             pxl_coords = self.element_coordinates(elem)
             box = self.pt_from_pixel(pxl_coords)
-            if interwordSpaces:
+            if interword_spaces:
                 # if  `--interword-spaces` is true, append a space
                 # to the end of each text element to allow simpler PDF viewers
                 # such as PDF.js to better recognize words in search and copy
@@ -318,7 +327,7 @@ class HocrTransform:
             font_width = pdf.stringWidth(elemtxt, fontname, fontsize)
 
             # draw the bbox border
-            if showBoundingboxes:
+            if show_bounding_boxes:  # pragma: no cover
                 pdf.rect(
                     box.x1, self.height - line_box.y2, box_width, line_height, fill=0
                 )
@@ -385,5 +394,5 @@ if __name__ == "__main__":
         args.outputfile,
         args.image,
         args.boundingboxes,
-        interwordSpaces=args.interword_spaces,
+        interword_spaces=args.interword_spaces,
     )

src/ocrmypdf/leptonica.py

@@ -29,18 +29,50 @@ from collections.abc import Sequence
 from contextlib import suppress
 from ctypes.util import find_library
 from functools import lru_cache
-from io import BytesIO
+from io import BytesIO, UnsupportedOperation
 from os import fspath
 from tempfile import TemporaryFile
 
-from .lib._leptonica import ffi
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.lib._leptonica import ffi
+from ocrmypdf.subprocess import shim_paths_with_program_files
 
 # pylint: disable=protected-access
 
 logger = logging.getLogger(__name__)
 
-lept = ffi.dlopen(find_library('lept'))
-lept.setMsgSeverity(lept.L_SEVERITY_WARNING)
+if os.name == 'nt':
+    libname = 'liblept-5'
+    os.environ['PATH'] = shim_paths_with_program_files()
+else:
+    libname = 'lept'
+_libpath = find_library(libname)
+if not _libpath:
+    raise MissingDependencyError(
+        """
+        ---------------------------------------------------------------------
+        This error normally occurs when ocrmypdf can't find the Leptonica
+        library, which is usually installed with Tesseract OCR. It could be that
+        Tesseract is not installed properly, we can't find the installation
+        on your system PATH environment variable.
+
+        The library we are looking for is usually called:
+            liblept-5.dll   (Windows)
+            liblept*.dylib  (macOS)
+            liblept*.so     (Linux/BSD)
+
+        Please review our installation procedures to find a solution:
+            https://ocrmypdf.readthedocs.io/en/latest/installation.html
+        ---------------------------------------------------------------------
+        """
+    )
+try:
+    lept = ffi.dlopen(_libpath)
+    lept.setMsgSeverity(lept.L_SEVERITY_WARNING)
+except ffi.error as e:
+    raise MissingDependencyError(
+        f"Leptonica library found at {_libpath}, but we could not access it"
+    ) from e
 
 
 class _LeptonicaErrorTrap:
@@ -64,7 +96,6 @@ class _LeptonicaErrorTrap:
         self.no_stderr = False
 
     def __enter__(self):
-        from io import UnsupportedOperation
 
         self.tmpfile = TemporaryFile()
 
@@ -292,9 +323,11 @@ class Pix(LeptonicaObject):
         Leptonica can load TIFF, PNM (PBM, PGM, PPM), PNG, and JPEG.  If
         loading fails then the object will wrap a C null pointer.
         """
-        filename = fspath(path)
-        with _LeptonicaErrorTrap():
-            return cls(lept.pixRead(os.fsencode(filename)))
+        with open(path, 'rb') as py_file:
+            data = py_file.read()
+            buffer = ffi.from_buffer(data)
+            with _LeptonicaErrorTrap():
+                return cls(lept.pixReadMem(buffer, len(buffer)))
 
     def write_implied_format(self, path, jpeg_quality=0, jpeg_progressive=0):
         """Write pix to the filename, with the extension indicating format.
@@ -302,14 +335,22 @@ class Pix(LeptonicaObject):
         jpeg_quality -- quality (iff JPEG; 1 - 100, 0 for default)
         jpeg_progressive -- (iff JPEG; 0 for baseline seq., 1 for progressive)
         """
-        filename = fspath(path)
-        with _LeptonicaErrorTrap():
-            lept.pixWriteImpliedFormat(
-                os.fsencode(filename), self._cdata, jpeg_quality, jpeg_progressive
-            )
+        lept_format = lept.getImpliedFileFormat(os.fsencode(path))
+        with open(path, 'wb') as py_file:
+            data = ffi.new('l_uint8 **pdata')
+            size = ffi.new('size_t *psize')
+            with _LeptonicaErrorTrap():
+                if lept_format == lept.L_JPEG_ENCODE:
+                    lept.pixWriteMemJpeg(
+                        data, size, self._cdata, jpeg_quality, jpeg_progressive
+                    )
+                else:
+                    lept.pixWriteMem(data, size, self._cdata, lept_format)
+            buffer = ffi.buffer(data[0], size[0])
+            py_file.write(buffer)
 
     @classmethod
-    def frompil(self, pillow_image):
+    def frompil(cls, pillow_image):
         """Create a copy of a PIL.Image from this Pix"""
         bio = BytesIO()
         pillow_image.save(bio, format='png', compress_level=1)
@@ -321,7 +362,7 @@ class Pix(LeptonicaObject):
 
     def topil(self):
         """Returns a PIL.Image version of this Pix"""
-        from PIL import Image
+        from PIL import Image  # pylint: disable=import-outside-toplevel
 
         # Leptonica manages data in words, so it implicitly does an endian
         # swap.  Tell Pillow about this when it reads the data.
@@ -492,27 +533,15 @@ class Pix(LeptonicaObject):
             )
             return Pix(thresh_pix)
 
-    def crop_to_foreground(
-        self,
-        threshold=128,
-        mindist=70,
-        erasedist=30,
-        pagenum=0,
-        showmorph=0,
-        display=0,
-        pdfdir=ffi.NULL,
-    ):
+    def crop_to_foreground(self, threshold=128, mindist=70, erasedist=30, showmorph=0):
+        if get_leptonica_version() < 'leptonica-1.76':
+            # Leptonica 1.76 changed the API for pixFindPageForeground; we don't
+            # support the old version
+            raise LeptonicaError("Not available in this version of Leptonica")
         with _LeptonicaErrorTrap():
             cropbox = Box(
                 lept.pixFindPageForeground(
-                    self._cdata,
-                    threshold,
-                    mindist,
-                    erasedist,
-                    pagenum,
-                    showmorph,
-                    display,
-                    pdfdir,
+                    self._cdata, threshold, mindist, erasedist, showmorph, ffi.NULL
                 )
             )
 

src/ocrmypdf/lib/compile_leptonica.py

@@ -16,6 +16,8 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
+from pathlib import Path
+
 from cffi import FFI
 
 ffibuilder = FFI()
@@ -74,6 +76,17 @@ struct Pixa
 };
 typedef struct Pixa PIXA;
 
+/*! Array of compressed pix */
+struct PixaComp
+{
+    l_int32              n;         /*!< number of PixComp in ptr array    */
+    l_int32              nalloc;    /*!< number of PixComp ptrs allocated  */
+    l_int32              offset;    /*!< indexing offset into ptr array    */
+    struct PixComp     **pixc;      /*!< the array of ptrs to PixComp      */
+    struct Boxa         *boxa;      /*!< array of boxes                    */
+};
+typedef struct PixaComp PIXAC;
+
 struct Box
 {
     l_int32            x;
@@ -210,9 +223,15 @@ ffibuilder.cdef(
     """
 PIX * pixRead ( const char *filename );
 PIX * pixReadMem ( const l_uint8 *data, size_t size );
+PIX * pixReadStream ( FILE *fp, l_int32 hint );
 PIX * pixScale ( PIX *pixs, l_float32 scalex, l_float32 scaley );
 l_int32 pixFindSkew ( PIX *pixs, l_float32 *pangle, l_float32 *pconf );
 l_int32 pixWriteImpliedFormat ( const char *filename, PIX *pix, l_int32 quality, l_int32 progressive );
+l_int32 getImpliedFileFormat ( const char *filename );
+l_ok pixWriteStream ( FILE *fp, PIX *pix, l_int32 format );
+l_ok pixWriteStreamJpeg ( FILE *fp, PIX *pixs, l_int32 quality, l_int32 progressive );
+l_ok pixWriteMem ( l_uint8 **pdata, size_t *psize, PIX *pix, l_int32 format );
+l_ok pixWriteMemJpeg ( l_uint8 **pdata, size_t *psize, PIX *pix, l_int32 quality, l_int32 progressive );
 l_int32
 pixWriteMemPng(l_uint8  **pdata,
                size_t    *psize,
@@ -294,14 +313,12 @@ pixCleanBackgroundToWhite(PIX       *pixs,
                           l_int32    whiteval);
 
 BOX *
-pixFindPageForeground(PIX         *pixs,
-                      l_int32      threshold,
-                      l_int32      mindist,
-                      l_int32      erasedist,
-                      l_int32      pagenum,
-                      l_int32      showmorph,
-                      l_int32      display,
-                      const char  *pdfdir);
+pixFindPageForeground ( PIX *pixs,
+                        l_int32 threshold,
+                        l_int32 mindist,
+                        l_int32 erasedist,
+                        l_int32 showmorph,
+                        PIXAC *pixac );
 
 PIX *
 pixClipRectangle(PIX   *pixs,
@@ -414,7 +431,10 @@ pixExtractBarcodes(PIX     *pixs,
                    l_int32 debugflag);
 
 BOXA *
-pixLocateBarcodes ( PIX *pixs, l_int32 thresh, PIX **ppixb, PIX **ppixm );
+pixLocateBarcodes ( PIX *pixs,
+ l_int32 thresh,
+ PIX **ppixb,
+ PIX **ppixm );
 
 SARRAY *
 pixReadBarcodes(PIXA     *pixa,
@@ -423,6 +443,12 @@ pixReadBarcodes(PIXA     *pixa,
                 SARRAY  **psaw,
                 l_int32 debugflag);
 
+PIX *
+pixGenHalftoneMask(PIX      *pixs,
+                        PIX     **ppixtext,
+                        l_int32  *phtfound,
+                        PIXA     *pixadb);
+
 l_int32
 l_generateCIDataForPdf(const char *fname,
                        PIX *pix,
@@ -491,3 +517,8 @@ ffibuilder.set_source("ocrmypdf.lib._leptonica", None)
 
 if __name__ == '__main__':
     ffibuilder.compile(verbose=True)
+    if Path('ocrmypdf/lib/_leptonica.py').exists() and Path('src/ocrmypdf').exists():
+        output = Path('ocrmypdf/lib/_leptonica.py')
+        output.rename('src/ocrmypdf/lib/_leptonica.py')
+        Path('ocrmypdf/lib').rmdir()
+        Path('ocrmypdf').rmdir()

src/ocrmypdf/optimize.py

@@ -15,23 +15,27 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import concurrent.futures
+import logging
 import sys
 import tempfile
 from collections import defaultdict
+from functools import partial
 from os import fspath
 from pathlib import Path
 
+import pikepdf
+from pikepdf import Dictionary, Name
 from PIL import Image
 from tqdm import tqdm
-import pikepdf
-from pikepdf import Name, Dictionary
 
-from . import leptonica
-from ._jobcontext import PDFContext
-from .exec import jbig2enc, pngquant
-from .exceptions import OutputFileAccessError
-from .helpers import re_symlink
+from ocrmypdf import leptonica
+from ocrmypdf._concurrent import exec_progress_pool
+from ocrmypdf._exec import jbig2enc, pngquant
+from ocrmypdf._jobcontext import PdfContext
+from ocrmypdf.exceptions import OutputFileAccessError
+from ocrmypdf.helpers import safe_symlink
+
+log = logging.getLogger(__name__)
 
 DEFAULT_JPEG_QUALITY = 75
 DEFAULT_PNG_QUALITY = 70
@@ -53,7 +57,7 @@ def tif_name(root, xref):
     return img_name(root, xref, '.tif')
 
 
-def extract_image_filter(pike, root, log, image, xref):
+def extract_image_filter(pike, root, image, xref):
     if image.Subtype != Name.Image:
         return None
     if image.Length < 100:
@@ -73,11 +77,14 @@ def extract_image_filter(pike, root, log, image, xref):
     if filtdp[0] == Name.JPXDecode:
         return None  # Don't do JPEG2000
 
+    if Name.Decode in image:
+        return None  # Don't mess with custom Decode tables
+
     return pim, filtdp
 
 
-def extract_image_jbig2(*, pike, root, log, image, xref, options):
-    result = extract_image_filter(pike, root, log, image, xref)
+def extract_image_jbig2(*, pike, root, image, xref, options):
+    result = extract_image_filter(pike, root, image, xref)
     if result is None:
         return None
     pim, filtdp = result
@@ -98,12 +105,21 @@ def extract_image_jbig2(*, pike, root, log, image, xref, options):
     return None
 
 
-def extract_image_generic(*, pike, root, log, image, xref, options):
-    result = extract_image_filter(pike, root, log, image, xref)
+def extract_image_generic(*, pike, root, image, xref, options):
+    result = extract_image_filter(pike, root, image, xref)
     if result is None:
         return None
     pim, filtdp = result
 
+    # Don't try to PNG-optimize 1bpp images, since JBIG2 does it better.
+    if pim.bits_per_component == 1:
+        return None
+
+    try:
+        pim.indexed  # pikepdf 1.6.3 can't handle [/Indexed [/Array...]]
+    except NotImplementedError:
+        return None
+
     if filtdp[0] == Name.DCTDecode and options.optimize >= 2:
         # This is a simple heuristic derived from some training data, that has
         # about a 70% chance of guessing whether the JPEG is high quality,
@@ -143,11 +159,22 @@ def extract_image_generic(*, pike, root, log, image, xref, options):
         # generating a PNG from compressed data
         pim.as_pil_image().save(png_name(root, xref))
         return xref, '.png'
+    elif (
+        not pim.indexed
+        and pim.colorspace == Name.ICCBased
+        and pim.bits_per_component == 1
+        and not options.jbig2_lossy
+    ):
+        # We can losslessly optimize 1-bit images to CCITT or JBIG2 without
+        # paying any attention to the ICC profile, provided we're not doing
+        # lossy JBIG2
+        pim.as_pil_image().save(png_name(root, xref))
+        return xref, '.png'
 
     return None
 
 
-def extract_images(pike, root, log, options, extract_fn):
+def extract_images(pike, root, options, extract_fn):
     """Extract image using extract_fn
 
     Enumerate images on each page, lookup their xref/ID number in the PDF.
@@ -189,9 +216,9 @@ def extract_images(pike, root, log, options, extract_fn):
         image = pike.get_object((xref, 0))
         try:
             result = extract_fn(
-                pike=pike, root=root, log=log, image=image, xref=xref, options=options
+                pike=pike, root=root, image=image, xref=xref, options=options
             )
-        except Exception as e:
+        except Exception as e:  # pylint: disable=broad-except
             log.debug("Image xref %s, error %s", xref, repr(e))
             errors += 1
         else:
@@ -200,12 +227,12 @@ def extract_images(pike, root, log, options, extract_fn):
                 yield pageno_for_xref[xref], xref, ext
 
 
-def extract_images_generic(pike, root, log, options):
+def extract_images_generic(pike, root, options):
     """Extract any >=2bpp image we think we can improve"""
 
     jpegs = []
     pngs = []
-    for _, xref, ext in extract_images(pike, root, log, options, extract_image_generic):
+    for _, xref, ext in extract_images(pike, root, options, extract_image_generic):
         log.debug('xref = %s ext = %s', xref, ext)
         if ext == '.png':
             pngs.append(xref)
@@ -215,13 +242,11 @@ def extract_images_generic(pike, root, log, options):
     return jpegs, pngs
 
 
-def extract_images_jbig2(pike, root, log, options):
+def extract_images_jbig2(pike, root, options):
     """Extract any bitonal image that we think we can improve as JBIG2"""
 
     jbig2_groups = defaultdict(list)
-    for pageno, xref, ext in extract_images(
-        pike, root, log, options, extract_image_jbig2
-    ):
+    for pageno, xref, ext in extract_images(pike, root, options, extract_image_jbig2):
         group = pageno // options.jbig2_page_group_size
         jbig2_groups[group].append((xref, ext))
 
@@ -233,54 +258,55 @@ def extract_images_jbig2(pike, root, log, options):
     return jbig2_groups
 
 
-def _produce_jbig2_images(jbig2_groups, root, log, options):
+def _produce_jbig2_images(jbig2_groups, root, options):
     """Produce JBIG2 images from their groups"""
 
-    def jbig2_group_futures(executor, root, groups):
+    def jbig2_group_args(root, groups):
         for group, xref_exts in groups.items():
             prefix = f'group{group:08d}'
-            future = executor.submit(
-                jbig2enc.convert_group,
+            yield dict(
                 cwd=fspath(root),
                 infiles=(img_name(root, xref, ext) for xref, ext in xref_exts),
                 out_prefix=prefix,
             )
-            yield future
 
-    def jbig2_single_futures(executor, root, groups):
+    def jbig2_single_args(root, groups):
         for group, xref_exts in groups.items():
             prefix = f'group{group:08d}'
             # Second loop is to ensure multiple images per page are unpacked
             for n, xref_ext in enumerate(xref_exts):
                 xref, ext = xref_ext
-                future = executor.submit(
-                    jbig2enc.convert_single,
+                yield dict(
                     cwd=fspath(root),
                     infile=img_name(root, xref, ext),
                     outfile=root / f'{prefix}.{n:04d}',
                 )
-                yield future
+
+    def convert_generic(fn, kwargs_dict):
+        return fn(**kwargs_dict)
 
     if options.jbig2_page_group_size > 1:
-        jbig2_futures = jbig2_group_futures
+        jbig2_args = jbig2_group_args
+        jbig2_convert = partial(convert_generic, jbig2enc.convert_group)
     else:
-        jbig2_futures = jbig2_single_futures
+        jbig2_args = jbig2_single_args
+        jbig2_convert = partial(convert_generic, jbig2enc.convert_single)
 
-    with concurrent.futures.ThreadPoolExecutor(max_workers=options.jobs) as executor:
-        futures = jbig2_futures(executor, root, jbig2_groups)
-        with tqdm(
+    exec_progress_pool(
+        use_threads=True,
+        max_workers=options.jobs,
+        tqdm_kwargs=dict(
             total=len(jbig2_groups),
             desc="JBIG2",
             unit='item',
             disable=not options.progress_bar,
-        ) as pbar:
-            for future in concurrent.futures.as_completed(futures):
-                proc = future.result()
-                log.debug(proc.stderr.decode())
-                pbar.update()
+        ),
+        task=jbig2_convert,
+        task_arguments=jbig2_args(root, jbig2_groups),
+    )
 
 
-def convert_to_jbig2(pike, jbig2_groups, root, log, options):
+def convert_to_jbig2(pike, jbig2_groups, root, options):
     """Convert images to JBIG2 and insert into PDF.
 
     When the JBIG2 page group size is > 1 we do several JBIG2 images at once
@@ -294,7 +320,7 @@ def convert_to_jbig2(pike, jbig2_groups, root, log, options):
     and needs no dictionary. Currently this must be lossless JBIG2.
     """
 
-    _produce_jbig2_images(jbig2_groups, root, log, options)
+    _produce_jbig2_images(jbig2_groups, root, options)
 
     for group, xref_exts in jbig2_groups.items():
         prefix = f'group{group:08d}'
@@ -318,7 +344,7 @@ def convert_to_jbig2(pike, jbig2_groups, root, log, options):
             )
 
 
-def transcode_jpegs(pike, jpegs, root, log, options):
+def transcode_jpegs(pike, jpegs, root, options):
     for xref in tqdm(
         jpegs, desc="JPEGs", unit='image', disable=not options.progress_bar
     ):
@@ -341,40 +367,49 @@ def transcode_jpegs(pike, jpegs, root, log, options):
         im_obj.write(compdata.read(), filter=Name.DCTDecode)
 
 
-def transcode_pngs(pike, images, image_name_fn, root, log, options):
+def transcode_pngs(pike, images, image_name_fn, root, options):
+    modified = set()
     if options.optimize >= 2:
         png_quality = (
             max(10, options.png_quality - 10),
             min(100, options.png_quality + 10),
         )
-        with concurrent.futures.ThreadPoolExecutor(
-            max_workers=options.jobs
-        ) as executor:
-            futures = []
+
+        def pngquant_args():
             for xref in images:
                 log.debug(image_name_fn(root, xref))
-                futures.append(
-                    executor.submit(
-                        pngquant.quantize,
-                        image_name_fn(root, xref),
-                        png_name(root, xref),
-                        png_quality[0],
-                        png_quality[1],
-                    )
+                yield (
+                    image_name_fn(root, xref),
+                    png_name(root, xref),
+                    png_quality[0],
+                    png_quality[1],
                 )
-            with tqdm(
+                modified.add(xref)
+
+        def pngquant_fn(args):
+            pngquant.quantize(*args)
+
+        exec_progress_pool(
+            use_threads=True,
+            max_workers=options.jobs,
+            tqdm_kwargs=dict(
                 desc="PNGs",
-                total=len(futures),
+                total=len(images),
                 unit='image',
                 disable=not options.progress_bar,
-            ) as pbar:
-                for _future in concurrent.futures.as_completed(futures):
-                    pbar.update()
+            ),
+            task=pngquant_fn,
+            task_arguments=pngquant_args(),
+        )
 
-    for xref in images:
+    for xref in modified:
         im_obj = pike.get_object(xref, 0)
         try:
-            compdata = leptonica.CompressedData.open(png_name(root, xref))
+            pix = leptonica.Pix.open(png_name(root, xref))
+            if pix.mode == '1':
+                compdata = pix.generate_pdf_ci_data(leptonica.lept.L_G4_ENCODE, 0)
+            else:
+                compdata = leptonica.CompressedData.open(png_name(root, xref))
         except leptonica.LeptonicaError as e:
             # Most likely this means file not found, i.e. quantize did not
             # produce an improved version
@@ -390,69 +425,89 @@ def transcode_pngs(pike, images, image_name_fn, root, log, options):
                 f"{len(compdata)} > {int(im_obj.stream_dict.Length)}"
             )
             continue
-
-        # When a PNG is inserted into a PDF, we more or less copy the IDAT section from
-        # the PDF and transfer the rest of the PNG headers to PDF image metadata.
-        # One thing we have to do is tell the PDF reader whether a predictor was used
-        # on the image before Flate encoding. (Typically one is.)
-        # According to Leptonica source, PDF readers don't actually need us
-        # to specify the correct predictor, they just need a value of either:
-        #   1 - no predictor
-        #   10-14 - there is a predictor
-        # Leptonica's compdata->predictor only tells TRUE or FALSE
-        # From there the PNG decoder can infer the rest from the file.
-        # In practice the predictor should be Paeth, 14, so we'll use that.
-        # See:
-        #   - PDF RM 7.4.4.4 Table 10
-        #   - https://github.com/DanBloomberg/leptonica/blob/master/src/pdfio2.c#L757
-        predictor = 14 if compdata.predictor > 0 else 1
-        dparms = Dictionary(Predictor=predictor)
-        if predictor > 1:
-            dparms.BitsPerComponent = compdata.bps  # Yes, this is redundant
-            dparms.Colors = compdata.spp
-            dparms.Columns = compdata.w
-
-        im_obj.BitsPerComponent = compdata.bps
-        im_obj.Width = compdata.w
-        im_obj.Height = compdata.h
-
-        if compdata.ncolors > 0:
-            # .ncolors is the number of colors in the palette, not the number of
-            # colors used in a true color image
-            palette_pdf_string = compdata.get_palette_pdf_string()
-            palette_data = pikepdf.Object.parse(palette_pdf_string)
-            palette_stream = pikepdf.Stream(pike, bytes(palette_data))
-            palette = [
-                Name.Indexed,
-                Name.DeviceRGB,
-                compdata.ncolors - 1,
-                palette_stream,
-            ]
-            cs = palette
-        else:
-            if compdata.spp == 1:
-                # PDF interprets binary-1 as black in 1bpp, but PNG sets
-                # black to 0 for 1bpp. Create a palette that informs the PDF
-                # of the mapping - seems cleaner to go this way but pikepdf
-                # needs to be patched to support it.
-                # palette = [Name.Indexed, Name.DeviceGray, 1, b"\xff\x00"]
-                # cs = palette
-                cs = Name.DeviceGray
-            elif compdata.spp == 3:
-                cs = Name.DeviceRGB
-            elif compdata.spp == 4:
-                cs = Name.DeviceCMYK
-        if compdata.bps == 1:
-            im_obj.Decode = [1, 0]  # Bit of a kludge but this inverts photometric too
-        im_obj.ColorSpace = cs
-        im_obj.write(compdata.read(), filter=Name.FlateDecode, decode_parms=dparms)
+        if compdata.type == leptonica.lept.L_FLATE_ENCODE:
+            rewrite_png(pike, im_obj, compdata)
+        elif compdata.type == leptonica.lept.L_G4_ENCODE:
+            rewrite_png_as_g4(pike, im_obj, compdata)
 
 
-def optimize(input_file, output_file, context):
-    log = context.log
+def rewrite_png_as_g4(pike, im_obj, compdata):
+    im_obj.BitsPerComponent = 1
+    im_obj.Width = compdata.w
+    im_obj.Height = compdata.h
+
+    im_obj.write(compdata.read())
+
+    log.debug(f"PNG to G4 {im_obj.objgen}")
+    if Name.Predictor in im_obj:
+        del im_obj.Predictor
+    if Name.DecodeParms in im_obj:
+        del im_obj.DecodeParms
+    im_obj.DecodeParms = Dictionary(
+        K=-1, BlackIs1=bool(compdata.minisblack), Columns=compdata.w
+    )
+
+    im_obj.Filter = Name.CCITTFaxDecode
+    return
+
+
+def rewrite_png(pike, im_obj, compdata):
+    # When a PNG is inserted into a PDF, we more or less copy the IDAT section from
+    # the PDF and transfer the rest of the PNG headers to PDF image metadata.
+    # One thing we have to do is tell the PDF reader whether a predictor was used
+    # on the image before Flate encoding. (Typically one is.)
+    # According to Leptonica source, PDF readers don't actually need us
+    # to specify the correct predictor, they just need a value of either:
+    #   1 - no predictor
+    #   10-14 - there is a predictor
+    # Leptonica's compdata->predictor only tells TRUE or FALSE
+    # 10-14 means the actual predictor is specified in the data, so for any
+    # number >= 10 the PDF reader will use whatever the PNG data specifies.
+    # In practice Leptonica should use Paeth, 14, but 15 seems to be the
+    # designated value for "optimal". So we will use 15.
+    # See:
+    #   - PDF RM 7.4.4.4 Table 10
+    #   - https://github.com/DanBloomberg/leptonica/blob/master/src/pdfio2.c#L757
+    predictor = 15 if compdata.predictor > 0 else 1
+    dparms = Dictionary(Predictor=predictor)
+    if predictor > 1:
+        dparms.BitsPerComponent = compdata.bps  # Yes, this is redundant
+        dparms.Colors = compdata.spp
+        dparms.Columns = compdata.w
+
+    im_obj.BitsPerComponent = compdata.bps
+    im_obj.Width = compdata.w
+    im_obj.Height = compdata.h
+
+    log.debug(
+        f"PNG {im_obj.objgen}: palette={compdata.ncolors} spp={compdata.spp} bps={compdata.bps}"
+    )
+    if compdata.ncolors > 0:
+        # .ncolors is the number of colors in the palette, not the number of
+        # colors used in a true color image. The palette string is always
+        # given as RGB tuples even when the image is grayscale; see
+        # https://github.com/DanBloomberg/leptonica/blob/master/src/colormap.c#L2067
+        palette_pdf_string = compdata.get_palette_pdf_string()
+        palette_data = pikepdf.Object.parse(palette_pdf_string)
+        palette_stream = pikepdf.Stream(pike, bytes(palette_data))
+        palette = [Name.Indexed, Name.DeviceRGB, compdata.ncolors - 1, palette_stream]
+        cs = palette
+    else:
+        # ncolors == 0 means we are using a colorspace without a palette
+        if compdata.spp == 1:
+            cs = Name.DeviceGray
+        elif compdata.spp == 3:
+            cs = Name.DeviceRGB
+        elif compdata.spp == 4:
+            cs = Name.DeviceCMYK
+    im_obj.ColorSpace = cs
+    im_obj.write(compdata.read(), filter=Name.FlateDecode, decode_parms=dparms)
+
+
+def optimize(input_file, output_file, context, save_settings):
     options = context.options
     if options.optimize == 0:
-        re_symlink(input_file, output_file)
+        safe_symlink(input_file, output_file)
         return
 
     if options.jpeg_quality == 0:
@@ -466,23 +521,19 @@ def optimize(input_file, output_file, context):
         root = Path(output_file).parent / 'images'
         root.mkdir(exist_ok=True)
 
-        jpegs, pngs = extract_images_generic(pike, root, log, options)
-        transcode_jpegs(pike, jpegs, root, log, options)
+        jpegs, pngs = extract_images_generic(pike, root, options)
+        transcode_jpegs(pike, jpegs, root, options)
         # if options.optimize >= 2:
         # Try pngifying the jpegs
-        #    transcode_pngs(pike, jpegs, jpg_name, root, log, options)
-        transcode_pngs(pike, pngs, png_name, root, log, options)
+        #    transcode_pngs(pike, jpegs, jpg_name, root, options)
+        transcode_pngs(pike, pngs, png_name, root, options)
 
-        jbig2_groups = extract_images_jbig2(pike, root, log, options)
-        convert_to_jbig2(pike, jbig2_groups, root, log, options)
+        jbig2_groups = extract_images_jbig2(pike, root, options)
+        convert_to_jbig2(pike, jbig2_groups, root, options)
 
         target_file = Path(output_file).with_suffix('.opt.pdf')
         pike.remove_unreferenced_resources()
-        pike.save(
-            target_file,
-            preserve_pdfa=True,
-            object_stream_mode=pikepdf.ObjectStreamMode.generate,
-        )
+        pike.save(target_file, **save_settings)
 
     input_size = Path(input_file).stat().st_size
     output_size = Path(target_file).stat().st_size
@@ -496,25 +547,28 @@ def optimize(input_file, output_file, context):
     log.info(f"Optimize ratio: {ratio:.2f} savings: {(100 * savings):.1f}%")
 
     if savings < 0:
-        log.info("Optimize did not improve the file - discarded")
-        re_symlink(input_file, output_file)
+        log.info("Image optimization did not improve the file - discarded")
+        # We still need to save the file
+        with pikepdf.open(input_file) as pike:
+            pike.remove_unreferenced_resources()
+            pike.save(output_file, **save_settings)
     else:
-        re_symlink(target_file, output_file)
+        safe_symlink(target_file, output_file)
 
 
 def main(infile, outfile, level, jobs=1):
-    from tempfile import TemporaryDirectory
-    from shutil import copy
+    from tempfile import TemporaryDirectory  # pylint: disable=import-outside-toplevel
+    from shutil import copy  # pylint: disable=import-outside-toplevel
 
     class OptimizeOptions:
         """Emulate ocrmypdf's options"""
 
         def __init__(
-            self, input_file, jobs, optimize, jpeg_quality, png_quality, jb2lossy
+            self, input_file, jobs, optimize_, jpeg_quality, png_quality, jb2lossy
         ):
             self.input_file = input_file
             self.jobs = jobs
-            self.optimize = optimize
+            self.optimize = optimize_
             self.jpeg_quality = jpeg_quality
             self.png_quality = png_quality
             self.jbig2_page_group_size = 0
@@ -525,16 +579,25 @@ def main(infile, outfile, level, jobs=1):
     options = OptimizeOptions(
         input_file=infile,
         jobs=jobs,
-        optimize=int(level),
+        optimize_=int(level),
         jpeg_quality=0,  # Use default
         png_quality=0,
         jb2lossy=False,
     )
 
     with TemporaryDirectory() as td:
-        context = PDFContext(options, td, infile, None)
+        context = PdfContext(options, td, infile, None, None)
         tmpout = Path(td) / 'out.pdf'
-        optimize(infile, tmpout, context)
+        optimize(
+            infile,
+            tmpout,
+            context,
+            dict(
+                compress_streams=True,
+                preserve_pdfa=True,
+                object_stream_mode=pikepdf.ObjectStreamMode.generate,
+            ),
+        )
         copy(fspath(tmpout), fspath(outfile))
 
 

src/ocrmypdf/pdfa.py

@@ -31,14 +31,12 @@ Ghostscript's handling of pdfmark.
 
 """
 
-import os
-from binascii import hexlify
+import base64
 from pathlib import Path
 from string import Template
 
-import pkg_resources
-
 import pikepdf
+import pkg_resources
 
 ICC_PROFILE_RELPATH = 'data/sRGB.icc'
 
@@ -49,26 +47,13 @@ SRGB_ICC_PROFILE = pkg_resources.resource_filename('ocrmypdf', ICC_PROFILE_RELPA
 # files, from the Ghostscript documentation. Lines beginning with % are
 # comments. Python substitution variables have a '$' prefix.
 pdfa_def_template = u"""%!
-% Define entries in the document Info dictionary :
+% Define an ICC profile :
 /ICCProfile $icc_profile
 def
 
-% Define an ICC profile :
-
 [/_objdef {icc_PDFA} /type /stream /OBJ pdfmark
-[{icc_PDFA}
-<<
-  /N currentpagedevice /ProcessColorModel known {
-    currentpagedevice /ProcessColorModel get dup /DeviceGray eq
-    {pop 1} {
-      /DeviceRGB eq
-      {3}{4} ifelse
-    } ifelse
-  } {
-    (ERROR, unable to determine ProcessColorModel) == flush
-  } ifelse
->> /PUT pdfmark
-[{icc_PDFA} ICCProfile (r) file /PUT pdfmark
+[{icc_PDFA} << /N 3 >> /PUT pdfmark
+[{icc_PDFA} ICCProfile /PUT pdfmark
 
 % Define the output intent dictionary :
 
@@ -97,24 +82,16 @@ def generate_pdfa_ps(target_filename, icc='sRGB'):
 
     :param target_filename: filename to save
     :param icc: ICC identifier such as 'sRGB'
-
-    :returns: a string containing the entire pdfmark
     """
     if icc == 'sRGB':
         icc_profile = SRGB_ICC_PROFILE
     else:
         raise NotImplementedError("Only supporting sRGB")
 
-    # pdfmark must contain the full path to the ICC profile, and pdfmark must be
-    # also encoded in ASCII. ocrmypdf can be installed anywhere, including to
-    # paths that have a non-ASCII character in the filename. Ghostscript
-    # accepts hex-encoded strings and converts them to byte strings, so
-    # we encode the path with fsencode() and use the hex representation.
-    # UTF-16 not accepted here. (Even though ASCII encodable is the usual case,
-    # do this always to avoid making it a rare conditional.)
-    bytes_icc_profile = os.fsencode(icc_profile)
-    hex_icc_profile = hexlify(bytes_icc_profile)
-    icc_profile = '<' + hex_icc_profile.decode('ascii') + '>'
+    # Read the ICC profile, encode as ASCII85 and convert to a string which we
+    # will insert in the .ps file
+    bytes_icc_profile = Path(icc_profile).read_bytes()
+    icc_profile = base64.a85encode(bytes_icc_profile, adobe=True).decode('ascii')
 
     t = Template(pdfa_def_template)
     ps = t.substitute(icc_profile=icc_profile, icc_identifier=icc)
@@ -122,6 +99,7 @@ def generate_pdfa_ps(target_filename, icc='sRGB'):
     # We should have encoded everything to pure ASCII by this point, and
     # to be safe, only allow ASCII in PostScript
     Path(target_filename).write_text(ps, encoding='ascii')
+    return target_filename
 
 
 def file_claims_pdfa(filename):

src/ocrmypdf/pdfinfo/__init__.py

@@ -16,4 +16,4 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from .info import PdfInfo, Colorspace, Encoding
+from ocrmypdf.pdfinfo.info import Colorspace, Encoding, PdfInfo

src/ocrmypdf/pdfinfo/ghosttext.py

@@ -1,101 +0,0 @@
-# © 2018 James R. Barlow: github.com/jbarlow83
-#
-# This file is part of OCRmyPDF.
-#
-# OCRmyPDF is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# OCRmyPDF is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
-
-import logging
-import re
-import xml.etree.ElementTree as ET
-
-from ..exec import ghostscript
-
-gslog = logging.getLogger()
-
-# Forgive me for I have sinned
-# I am using regular expressions to parse XML. However the XML in this case,
-# generated by Ghostscript, is self-consistent enough to be parseable.
-regex_remove_char_tags = re.compile(
-    br"""
-    <char\b
-    (?:   [^>]   # anything single character but >
-        | \">\"  # special case: trap ">"
-    )*
-    />           # terminate with '/>'
-""",
-    re.VERBOSE,
-)
-
-
-def page_get_textblocks(infile, pageno, xmltext, height):
-    """Get text boxes out of Ghostscript txtwrite xml"""
-
-    root = xmltext
-    if not hasattr(xmltext, 'findall'):
-        return []
-
-    def blocks():
-        for span in root.findall('.//span'):
-            bbox_str = span.attrib['bbox']
-            font_size = span.attrib['size']
-            pts = [int(pt) for pt in bbox_str.split()]
-            pts[1] = pts[1] - int(float(font_size) + 0.5)
-            bbox_topdown = tuple(pts)
-            bb = bbox_topdown
-            bbox_bottomup = (bb[0], height - bb[3], bb[2], height - bb[1])
-            yield bbox_bottomup
-
-    def joined_blocks():
-        prev = None
-        for bbox in blocks():
-            if prev is None:
-                prev = bbox
-            if bbox[1] == prev[1] and bbox[3] == prev[3]:
-                gap = prev[2] - bbox[0]
-                height = abs(bbox[3] - bbox[1])
-                if gap < height:
-                    # Join boxes
-                    prev = (prev[0], prev[1], bbox[2], bbox[3])
-                    continue
-            # yield previously joined bboxes and start anew
-            yield prev
-            prev = bbox
-        if prev is not None:
-            yield prev
-
-    return [block for block in joined_blocks()]
-
-
-def extract_text_xml(infile, pdf, pageno=None, log=gslog):
-    existing_text = ghostscript.extract_text(infile, pageno=None)
-    existing_text = regex_remove_char_tags.sub(b' ', existing_text)
-
-    try:
-        root = ET.fromstringlist([b'<document>\n', existing_text, b'</document>\n'])
-        page_xml = root.findall('page')
-    except ET.ParseError as e:
-        log.error(
-            "An error occurred while attempting to retrieve existing text in "
-            "the input file. Will attempt to continue assuming that there is "
-            "no existing text in the file. The error was:"
-        )
-        log.error(e)
-        page_xml = [None] * len(pdf.pages)
-
-    page_count_difference = len(pdf.pages) - len(page_xml)
-    if page_count_difference != 0:
-        log.error("The number of pages in the input file is inconsistent.")
-        if page_count_difference > 0:
-            page_xml.extend([None] * page_count_difference)
-    return page_xml

src/ocrmypdf/pdfinfo/info.py

@@ -16,29 +16,31 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from collections import namedtuple
+import logging
+import re
+from collections import defaultdict, namedtuple
 from decimal import Decimal
 from enum import Enum
-import logging
+from functools import partial
 from math import hypot, isclose
-from os import fspath
+from os import PathLike
 from pathlib import Path
 from warnings import warn
-import re
 
-from pikepdf import PdfMatrix
 import pikepdf
-from tqdm import tqdm
+from pikepdf import PdfMatrix
 
-from . import ghosttext
-from ocrmypdf.exceptions import EncryptedPdfError, MissingDependencyError
+from ocrmypdf._concurrent import exec_progress_pool
+from ocrmypdf.exceptions import EncryptedPdfError
+from ocrmypdf.helpers import Resolution, available_cpu_count
+from ocrmypdf.pdfinfo.layout import get_page_analysis, get_text_boxes
 
 logger = logging.getLogger()
 
 Colorspace = Enum('Colorspace', 'gray rgb cmyk lab icc index sep devn pattern jpeg2000')
 
 Encoding = Enum(
-    'Encoding', 'ccitt jpeg jpeg2000 jbig2 asciihex ascii85 lzw flate ' + 'runlength'
+    'Encoding', 'ccitt jpeg jpeg2000 jbig2 asciihex ascii85 lzw flate runlength'
 )
 
 FRIENDLY_COLORSPACE = {
@@ -96,15 +98,19 @@ XobjectSettings = namedtuple('XobjectSettings', ['name', 'shorthand', 'stack_dep
 InlineSettings = namedtuple('InlineSettings', ['iimage', 'shorthand', 'stack_depth'])
 
 ContentsInfo = namedtuple(
-    'ContentsInfo', ['xobject_settings', 'inline_images', 'found_vector']
+    'ContentsInfo',
+    ['xobject_settings', 'inline_images', 'found_vector', 'found_text', 'name_index'],
 )
 
 TextboxInfo = namedtuple('TextboxInfo', ['bbox', 'is_visible', 'is_corrupt'])
 
 
-class VectorInfo:
-    def __init__(self):
-        pass
+class VectorMarker:
+    pass
+
+
+class TextMarker:
+    pass
 
 
 def _normalize_stack(graphobjs):
@@ -149,10 +155,13 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
     ctm = PdfMatrix(initial_shorthand)
     xobject_settings = []
     inline_images = []
+    name_index = defaultdict(lambda: [])
     found_vector = False
+    found_text = False
     vector_ops = set('S s f F f* B B* b b*'.split())
+    text_showing_ops = set("""TJ Tj " '""".split())
     image_ops = set('BI ID EI q Q Do cm'.split())
-    operator_whitelist = ' '.join(vector_ops | image_ops)
+    operator_whitelist = ' '.join(vector_ops | text_showing_ops | image_ops)
 
     for n, graphobj in enumerate(
         _normalize_stack(
@@ -183,6 +192,7 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
                 name=image_name, shorthand=ctm.shorthand, stack_depth=len(stack)
             )
             xobject_settings.append(settings)
+            name_index[image_name].append(settings)
         elif operator == 'INLINE IMAGE':  # BI/ID/EI are grouped into this
             iimage = operands[0]
             inline = InlineSettings(
@@ -191,11 +201,15 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
             inline_images.append(inline)
         elif operator in vector_ops:
             found_vector = True
+        elif operator in text_showing_ops:
+            found_text = True
 
     return ContentsInfo(
         xobject_settings=xobject_settings,
         inline_images=inline_images,
         found_vector=found_vector,
+        found_text=found_text,
+        name_index=name_index,
     )
 
 
@@ -260,7 +274,7 @@ def _get_dpi(ctm_shorthand, image_size):
     dpi_w = scale_w * 72.0
     dpi_h = scale_h * 72.0
 
-    return dpi_w, dpi_h
+    return Resolution(dpi_w, dpi_h)
 
 
 class ImageInfo:
@@ -301,14 +315,22 @@ class ImageInfo:
         if self._enc == Encoding.jpeg2000:
             self._color = Colorspace.jpeg2000
 
-        self._comp = FRIENDLY_COMP.get(self._color, '?')
+        if self._color == Colorspace.icc:
+            # Check the ICC profile to determine actual colorspace
+            pim_icc = pim.icc
+            if pim_icc.profile.xcolor_space == 'GRAY':
+                self._comp = 1
+            elif pim_icc.profile.xcolor_space == 'CMYK':
+                self._comp = 4
+            else:
+                self._comp = 3
+        else:
+            self._comp = FRIENDLY_COMP.get(self._color, '?')
 
-        # Bit of a hack... infer grayscale if component count is uncertain
-        # but encoding must be monochrome. This happens if a monochrome image
-        # has an ICC profile attached. Better solution would be to examine
-        # the ICC profile.
-        if self._comp == '?' and self._enc in (Encoding.ccitt, 'jbig2'):
-            self._comp = FRIENDLY_COMP[Colorspace.gray]
+            # Bit of a hack... infer grayscale if component count is uncertain
+            # but encoding only supports monochrome.
+            if self._comp == '?' and self._enc in (Encoding.ccitt, Encoding.jbig2):
+                self._comp = FRIENDLY_COMP[Colorspace.gray]
 
     @property
     def name(self):
@@ -343,12 +365,8 @@ class ImageInfo:
         return self._enc
 
     @property
-    def xres(self):
-        return _get_dpi(self._shorthand, (self._width, self._height))[0]
-
-    @property
-    def yres(self):
-        return _get_dpi(self._shorthand, (self._width, self._height))[1]
+    def dpi(self):
+        return _get_dpi(self._shorthand, (self._width, self._height))
 
     def __repr__(self):
         class_locals = {
@@ -358,7 +376,7 @@ class ImageInfo:
         }
         return (
             "<ImageInfo '{name}' {type_} {width}x{height} {color} "
-            "{comp} {bpc} {enc} {xres}x{yres}>"
+            "{comp} {bpc} {enc} {dpi}>"
         ).format(**class_locals)
 
 
@@ -409,13 +427,9 @@ def _find_regular_images(container, contentsinfo):
     """
 
     for pdfimage, xobj in _image_xobjects(container):
-
-        # For each image that is drawn on this, check if we drawing the
-        # current image - yes this is O(n^2), but n == 1 almost always
-        for draw in contentsinfo.xobject_settings:
-            if draw.name != xobj:
-                continue
-
+        if xobj not in contentsinfo.name_index:
+            continue
+        for draw in contentsinfo.name_index[xobj]:
             if draw.stack_depth == 0 and _is_unit_square(draw.shorthand):
                 # At least one PDF in the wild (and test suite) draws an image
                 # when the graphics stack depth is 0, meaning that the image
@@ -500,7 +514,9 @@ def _process_content_streams(*, pdf, container, shorthand=None):
     contentsinfo = _interpret_contents(container, initial_shorthand)
 
     if contentsinfo.found_vector:
-        yield VectorInfo()
+        yield VectorMarker()
+    if contentsinfo.found_text:
+        yield TextMarker()
     yield from _find_inline_images(contentsinfo)
     yield from _find_regular_images(container, contentsinfo)
     yield from _find_form_xobject_images(pdf, container, contentsinfo)
@@ -549,7 +565,9 @@ def simplify_textboxes(miner, textbox_getter):
         yield TextboxInfo(box.bbox, visible, corrupt)
 
 
-def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
+def _pdf_get_pageinfo(
+    pdf, pageno: int, infile: PathLike, check_pages, detailed_analysis
+):
     pageinfo = {}
     pageinfo['pageno'] = pageno
     pageinfo['images'] = []
@@ -559,26 +577,18 @@ def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
     width_pt = mediabox[2] - mediabox[0]
     height_pt = mediabox[3] - mediabox[1]
 
-    if xmltext is not None:
-        bboxes = ghosttext.page_get_textblocks(
-            fspath(infile), pageno, xmltext=xmltext, height=height_pt
-        )
-        pageinfo['bboxes'] = bboxes
-    else:
-        # pdfminer required for this section
-        try:
-            from .layout import get_page_analysis, get_text_boxes
-        except ImportError:
-            raise MissingDependencyError(
-                "pdfminer is required for this feature. Your distribution "
-                "may not have installed it."
-            )
+    check_this_page = pageno in check_pages
+
+    if check_this_page and detailed_analysis:
         pscript5_mode = str(pdf.docinfo.get('/Creator')).startswith('PScript5')
         miner = get_page_analysis(infile, pageno, pscript5_mode)
         pageinfo['textboxes'] = list(simplify_textboxes(miner, get_text_boxes))
         bboxes = (box.bbox for box in pageinfo['textboxes'])
 
-    pageinfo['has_text'] = _page_has_text(bboxes, width_pt, height_pt)
+        pageinfo['has_text'] = _page_has_text(bboxes, width_pt, height_pt)
+    else:
+        pageinfo['textboxes'] = []
+        pageinfo['has_text'] = None  # i.e. "no information"
 
     userunit = page.get('/UserUnit', Decimal(1.0))
     if not isinstance(userunit, Decimal):
@@ -593,59 +603,96 @@ def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
         pageinfo['rotate'] = 0
 
     userunit_shorthand = (userunit, 0, 0, userunit, 0, 0)
-    contentsinfo = [
-        ci
+
+    if check_this_page:
+        pageinfo['has_vector'] = False
+        pageinfo['has_text'] = False
+        pageinfo['images'] = []
         for ci in _process_content_streams(
             pdf=pdf, container=page, shorthand=userunit_shorthand
-        )
-    ]
+        ):
+            if isinstance(ci, VectorMarker):
+                pageinfo['has_vector'] = True
+            elif isinstance(ci, TextMarker):
+                pageinfo['has_text'] = True
+            elif isinstance(ci, ImageInfo):
+                pageinfo['images'].append(ci)
+            else:
+                raise NotImplementedError()
+    else:
+        pageinfo['has_vector'] = None  # i.e. "no information"
+        pageinfo['has_text'] = None
+        pageinfo['images'] = None
 
-    pageinfo['has_vector'] = False
-    if any(isinstance(ci, VectorInfo) for ci in contentsinfo):
-        pageinfo['has_vector'] = True
-
-    pageinfo['images'] = [im for im in contentsinfo if isinstance(im, ImageInfo)]
     if pageinfo['images']:
-        xres = Decimal(max(image.xres for image in pageinfo['images']))
-        yres = Decimal(max(image.yres for image in pageinfo['images']))
-        pageinfo['xres'], pageinfo['yres'] = xres, yres
-        pageinfo['width_pixels'] = int(round(xres * pageinfo['width_inches']))
-        pageinfo['height_pixels'] = int(round(yres * pageinfo['height_inches']))
+        dpi = Resolution(0.0, 0.0).take_max(image.dpi for image in pageinfo['images'])
+        pageinfo['dpi'] = dpi
+        pageinfo['width_pixels'] = int(round(dpi.x * float(pageinfo['width_inches'])))
+        pageinfo['height_pixels'] = int(round(dpi.y * float(pageinfo['height_inches'])))
 
     return pageinfo
 
 
-def _pdf_get_all_pageinfo(infile, detailed_analysis=False, log=None, progbar=False):
-    pdf = pikepdf.open(infile)  # Do not close in this function
-    if pdf.is_encrypted:
-        pdf.close()
-        raise EncryptedPdfError()  # Triggered by encryption with empty passwd
-    if detailed_analysis:
-        pages_xml = None
-    else:
-        pages_xml = ghosttext.extract_text_xml(infile, pdf, pageno=None, log=log)
+worker_pdf = None
 
-    pages = []
-    for n, _ in tqdm(
-        enumerate(pdf.pages),
-        total=len(pdf.pages),
-        desc="Scan",
-        unit='page',
-        disable=not progbar,
-    ):
-        page_xml = pages_xml[n] if pages_xml else None
-        page = PageInfo(pdf, n, infile, page_xml, detailed_analysis)
-        pages.append(page)
 
-    return pages, pdf
+def _pdf_pageinfo_sync_init(infile):
+    global worker_pdf  # pylint: disable=global-statement
+    worker_pdf = pikepdf.open(infile)
+
+
+def _pdf_pageinfo_sync(args):
+    global worker_pdf  # pylint: disable=global-statement
+    pageno, infile, check_pages, detailed_analysis = args
+    page = PageInfo(worker_pdf, pageno, infile, check_pages, detailed_analysis)
+    return page
+
+
+def _pdf_pageinfo_concurrent(
+    pdf, infile, progbar, max_workers, check_pages, detailed_analysis=False
+):
+    pages = [None] * len(pdf.pages)
+
+    def update_pageinfo(result, pbar):
+        page = result
+        pages[page.pageno] = page
+        pbar.update()
+
+    if max_workers is None:
+        max_workers = available_cpu_count()
+
+    total = len(pdf.pages)
+    contexts = ((n, infile, check_pages, detailed_analysis) for n in range(total))
+
+    use_threads = False  # No performance gain if threaded due to GIL
+    n_workers = min(1 + len(pages) // 4, max_workers)
+    if n_workers == 1:
+        # But if we decided on only one worker, there is no point in using
+        # a separate process.
+        use_threads = True
+
+    exec_progress_pool(
+        use_threads=use_threads,
+        max_workers=n_workers,
+        tqdm_kwargs=dict(
+            total=total, desc="Scanning contents", unit='page', disable=not progbar
+        ),
+        task_initializer=partial(_pdf_pageinfo_sync_init, infile),
+        task=_pdf_pageinfo_sync,
+        task_arguments=contexts,
+        task_finished=update_pageinfo,
+    )
+    return pages
 
 
 class PageInfo:
-    def __init__(self, pdf, pageno, infile, xmltext, detailed_analysis=False):
+    def __init__(self, pdf, pageno, infile, check_pages, detailed_analysis=False):
         self._pageno = pageno
         self._infile = infile
-        self._pageinfo = _pdf_get_pageinfo(pdf, pageno, infile, xmltext)
         self._detailed_analysis = detailed_analysis
+        self._pageinfo = _pdf_get_pageinfo(
+            pdf, pageno, infile, check_pages, detailed_analysis
+        )
 
     @property
     def pageno(self):
@@ -675,11 +722,11 @@ class PageInfo:
 
     @property
     def width_pixels(self):
-        return int(round(self.width_inches * self.xres))
+        return int(round(float(self.width_inches) * self.dpi.x))
 
     @property
     def height_pixels(self):
-        return int(round(self.height_inches * self.yres))
+        return int(round(float(self.height_inches) * self.dpi.y))
 
     @property
     def rotation(self):
@@ -709,7 +756,7 @@ class PageInfo:
 
         if 'textboxes' not in self._pageinfo:
             if visible is not None and corrupt is not None:
-                raise NotImplementedError('Ghostscript textboxes cannot be classified')
+                raise NotImplementedError('Incomplete information on textboxes')
             return self._pageinfo['bboxes']
 
         return (
@@ -719,12 +766,8 @@ class PageInfo:
         )
 
     @property
-    def xres(self):
-        return self._pageinfo.get('xres', None)
-
-    @property
-    def yres(self):
-        return self._pageinfo.get('yres', None)
+    def dpi(self):
+        return self._pageinfo.get('dpi', Resolution(0.0, 0.0))
 
     @property
     def userunit(self):
@@ -739,34 +782,45 @@ class PageInfo:
 
     def __repr__(self):
         return (
-            '<PageInfo ' 'pageno={} {}"x{}" rotation={} res={}x{} has_text={}>'
-        ).format(
-            self.pageno,
-            self.width_inches,
-            self.height_inches,
-            self.rotation,
-            self.xres,
-            self.yres,
-            self.has_text,
+            f'<PageInfo '
+            f'pageno={self.pageno} {self.width_inches}"x{self.height_inches}" '
+            f'rotation={self.rotation} dpi={self.dpi} has_text={self.has_text}>'
         )
 
 
 class PdfInfo:
     """Get summary information about a PDF"""
 
-    def __init__(self, infile, detailed_page_analysis=False, log=logger, progbar=False):
+    def __init__(
+        self,
+        infile,
+        detailed_analysis=False,
+        progbar=False,
+        max_workers=None,
+        check_pages=None,
+    ):
         self._infile = infile
-        self._pages, pdf = _pdf_get_all_pageinfo(
-            infile, detailed_page_analysis, log=log, progbar=progbar
-        )
-        self._needs_rendering = pdf.root.get('/NeedsRendering', False)
-        self._has_acroform = False
-        if '/AcroForm' in pdf.root:
-            if len(pdf.root.AcroForm.get('/Fields', [])) > 0:
-                self._has_acroform = True
-            elif '/XFA' in pdf.root.AcroForm:
-                self._has_acroform = True
-        pdf.close()
+        if check_pages is None:
+            check_pages = range(0, 1_000_000_000)
+
+        with pikepdf.open(infile) as pdf:
+            if pdf.is_encrypted:
+                raise EncryptedPdfError()  # Triggered by encryption with empty passwd
+            self._pages = _pdf_pageinfo_concurrent(
+                pdf,
+                infile,
+                progbar,
+                max_workers,
+                check_pages=check_pages,
+                detailed_analysis=detailed_analysis,
+            )
+            self._needs_rendering = pdf.root.get('/NeedsRendering', False)
+            self._has_acroform = False
+            if '/AcroForm' in pdf.root:
+                if len(pdf.root.AcroForm.get('/Fields', [])) > 0:
+                    self._has_acroform = True
+                elif '/XFA' in pdf.root.AcroForm:
+                    self._has_acroform = True
 
     @property
     def pages(self):
@@ -806,15 +860,19 @@ class PdfInfo:
 
 
 def main():
-    import argparse
+    import argparse  # pylint: disable=import-outside-toplevel
+    from pprint import pprint  # pylint: disable=import-outside-toplevel
 
     parser = argparse.ArgumentParser()
     parser.add_argument('infile')
     args = parser.parse_args()
-    info = _pdf_get_all_pageinfo(args.infile)
-    from pprint import pprint
+    pdfinfo = PdfInfo(args.infile)
 
-    pprint(info)
+    pprint(pdfinfo)
+    for page in pdfinfo.pages:
+        pprint(page)
+        for im in page.images:
+            pprint(im)
 
 
 if __name__ == '__main__':

src/ocrmypdf/pdfinfo/layout.py

@@ -20,66 +20,21 @@ from math import copysign
 from pathlib import Path
 from unittest.mock import patch
 
+import pdfminer
 import pdfminer.encodingdb
 import pdfminer.pdfdevice
 import pdfminer.pdfinterp
 from pdfminer.converter import PDFLayoutAnalyzer
-from pdfminer.glyphlist import glyphname2unicode
 from pdfminer.layout import LAParams, LTChar, LTPage, LTTextBox
 from pdfminer.pdfdocument import PDFTextExtractionNotAllowed
-from pdfminer.pdffont import PDFFont, PDFSimpleFont, PDFUnicodeNotDefined
+from pdfminer.pdffont import PDFSimpleFont, PDFUnicodeNotDefined
 from pdfminer.pdfpage import PDFPage
 from pdfminer.utils import bbox2str, matrix2str
 
-from ..exceptions import EncryptedPdfError
+from ocrmypdf.exceptions import EncryptedPdfError
 
 STRIP_NAME = re.compile(r'[0-9]+')
 
-#
-# Unconditional pdfminer patches
-#
-
-
-def name2unicode(name):
-    """Fix pdfminer's name2unicode function
-
-    Font cids that are mapped to names of the form /g123 seem to be, by convention
-    characters with no corresponding Unicode entry. These can be subsetted fonts
-    or symbolic fonts. There seems to be no way to map /g123 fonts to Unicode,
-    barring a ToUnicode data structure.
-    """
-    if name in glyphname2unicode:
-        return glyphname2unicode[name]
-    if name.startswith('g') or name.startswith('a'):
-        raise KeyError(name)
-    if name.startswith('uni'):
-        try:
-            return chr(int(name[3:], 16))
-        except ValueError:  # Not hexadecimal
-            raise KeyError(name)
-    m = STRIP_NAME.search(name)
-    if not m:
-        raise KeyError(name)
-    return chr(int(m.group(0)))
-
-
-pdfminer.encodingdb.name2unicode = name2unicode
-
-original_PDFFont_init = PDFFont.__init__
-
-
-def PDFFont__init__(self, descriptor, widths, default_width=None):
-    original_PDFFont_init(self, descriptor, widths, default_width)
-    # PDF spec says descent should be negative
-    # A font with a positive descent implies it floats entirely above the
-    # baseline, i.e. it's not really a baseline anymore. I have fonts that
-    # claim a positive descent, but treating descent as positive always seems
-    # to misposition text.
-    if self.descent > 0:
-        self.descent = -self.descent
-
-
-PDFFont.__init__ = PDFFont__init__
 
 original_PDFSimpleFont_init = PDFSimpleFont.__init__
 
@@ -97,6 +52,7 @@ def PDFSimpleFont__init__(self, descriptor, widths, spec):
 
 
 PDFSimpleFont.__init__ = PDFSimpleFont__init__
+
 #
 # pdfminer patches when creator is PScript5.dll
 #
@@ -207,6 +163,7 @@ class TextPositionTracker(PDFLayoutAnalyzer):
         super().__init__(rsrcmgr, pageno, laparams)
         self.textstate = None
         self.result = None
+        self.cur_item = None  # not defined in pdfminer code as it should be
 
     def begin_page(self, page, ctm):
         super().begin_page(page, ctm)
@@ -263,7 +220,9 @@ class TextPositionTracker(PDFLayoutAnalyzer):
 
 def get_page_analysis(infile, pageno, pscript5_mode):
     rman = pdfminer.pdfinterp.PDFResourceManager(caching=True)
-    dev = TextPositionTracker(rman, laparams=LAParams())
+    dev = TextPositionTracker(
+        rman, laparams=LAParams(all_texts=True, detect_vertical=True)
+    )
     interp = pdfminer.pdfinterp.PDFPageInterpreter(rman, dev)
 
     if pscript5_mode:

src/ocrmypdf/pluginspec.py

@@ -0,0 +1,252 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+from abc import ABC, abstractmethod, abstractstaticmethod
+from argparse import ArgumentParser, Namespace
+from collections import namedtuple
+from pathlib import Path
+from typing import TYPE_CHECKING, AbstractSet, List, Optional
+
+import pluggy
+from PIL import Image
+
+from ocrmypdf.helpers import Resolution
+
+if TYPE_CHECKING:
+    from ocrmypdf._jobcontext import PageContext
+    from ocrmypdf.pdfinfo import PdfInfo
+
+hookspec = pluggy.HookspecMarker('ocrmypdf')
+
+# pylint: disable=unused-argument
+
+
+@hookspec
+def add_options(parser: ArgumentParser) -> None:
+    """Allows the plugin to add its own command line and API arguments.
+
+    OCRmyPDF converts command line arguments to API arguments, so adding
+    arguments here will cause new arguments to be processed for API calls
+    to ``ocrmypdf.ocr``, or when invoked on the command line.
+
+    Note:
+        This hook will be called from the main process, and may modify global state
+        before child worker processes are forked.
+    """
+
+
+@hookspec
+def check_options(options: Namespace) -> None:
+    """Called to ask the plugin to check all of the options.
+
+    The plugin may check if options that it added are valid.
+
+    Warnings or other messages may be passed to the user by creating a logger
+    object using ``log = logging.getLogger(__name__)`` and logging to this.
+
+    The plugin may also modify the *options*. All objects that are in options
+    must be picklable so they can be marshalled to child worker processes.
+
+    Raises:
+        ocrmypdf.exceptions.ExitCodeException: If options are not acceptable
+            and the application should terminate gracefully with an informative
+            message and error code.
+    Note:
+        This hook will be called from the main process, and may modify global state
+        before child worker processes are forked.
+	"""
+
+
+@hookspec
+def validate(pdfinfo: 'PdfInfo', options: Namespace) -> None:
+    """Called to give a plugin an opportunity to review *options* and *pdfinfo*.
+
+    *options* contains the "work order" to process a particular file. *pdfinfo*
+    contains information about the input file obtained after loading and
+    parsing. The plugin may modify the *options*. For example, you could decide
+    that a certain type of file should be treated with ``options.force_ocr = True``
+    based on information in its *pdfinfo*.
+
+    Raises:
+        ocrmypdf.exceptions.ExitCodeException: If options or pdfinfo are not acceptable
+            and the application should terminate gracefully with an informative
+            message and error code.
+    Note:
+        This hook will be called from the main process, and may modify global state
+        before child worker processes are forked.
+    """
+
+
+@hookspec(firstresult=True)
+def rasterize_pdf_page(
+    input_file: Path,
+    output_file: Path,
+    raster_device: str,
+    raster_dpi: Resolution,
+    pageno: int,
+    page_dpi: Optional[Resolution] = None,
+    rotation: Optional[int] = None,
+    filter_vector: bool = False,
+) -> Path:
+    """Rasterize one page of a PDF at resolution raster_dpi in canvas units.
+
+    The image is sized to match the integer pixels dimensions implied by
+    raster_dpi even if those numbers are noninteger. The image's DPI will
+    be overridden with the values in page_dpi.
+
+    Args:
+        input_file: The PDF to rasterize.
+        output_file: The desired name of the rasterized image.
+        raster_device: Type of image to produce at output_file
+        raster_dpi: Resolution at which to rasterize page
+        pageno: Page number to rasterize (beginning at page 1)
+        page_dpi: Resolution, overriding output image DPI
+        rotation: Cardinal angle, clockwise, to rotate page
+        filter_vector: If True, remove vector graphics objects
+    Returns:
+        output_file
+    Note:
+        This hook will be called from child processes. Modifying global state
+        will not affect the main process or other child processes.
+    """
+
+
+@hookspec(firstresult=True)
+def filter_ocr_image(page: 'PageContext', image: Image) -> Image:
+    """Called to filter the image before it is sent to OCR.
+
+    This is the image that OCR sees, not what the user sees when they view the
+    PDF.
+
+    Note:
+        This hook will be called from child processes. Modifying global state
+        will not affect the main process or other child processes.
+    """
+
+
+@hookspec(firstresult=True)
+def filter_page_image(page: 'PageContext', image_filename: Path) -> Path:
+    """Called to filter the whole page before it is inserted into the PDF.
+
+    A whole page image is only produced when preprocessing command line arguments
+    are issued or when ``--force-ocr`` is issued. If no whole page is image is
+    produced for a given page, this function will not be called. This is not
+    the image that will be shown to OCR.
+
+    ocrmypdf will create the PDF page based on the image format used. If you
+    convert the image to a JPEG, the output page will be created as a JPEG, etc.
+    Note that the ocrmypdf image optimization stage may ultimately chose a
+    different format.
+
+    Note:
+        This hook will be called from child processes. Modifying global state
+        will not affect the main process or other child processes.
+    """
+
+
+OrientationConfidence = namedtuple('OrientationConfidence', ('angle', 'confidence'))
+
+
+class OcrEngine(ABC):
+    @abstractstaticmethod
+    def version() -> str:
+        """Returns the version of the OCR engine."""
+
+    @abstractstaticmethod
+    def creator_tag(options: Namespace) -> str:
+        """Returns the creator tag to identify this software's role in creating the PDF."""
+
+    @abstractmethod
+    def __str__(self):
+        """Returns name of OCR engine and version."""
+
+    @abstractstaticmethod
+    def languages(options: Namespace) -> AbstractSet[str]:
+        """Returns the set of all languages that are supported by the engine.
+
+        Languages are typically given in 3-letter ISO 3166-1 codes, but actually
+        can be any value understood by the OCR engine."""
+
+    @abstractstaticmethod
+    def get_orientation(input_file: Path, options: Namespace) -> OrientationConfidence:
+        """Returns the orientation of the image."""
+
+    @abstractstaticmethod
+    def generate_hocr(
+        input_file: Path, output_hocr: Path, output_text: Path, options: Namespace
+    ) -> None:
+        """Called to produce a hOCR file and sidecar text file."""
+
+    @abstractstaticmethod
+    def generate_pdf(
+        input_file: Path, output_pdf: Path, output_text: Path, options: Namespace
+    ) -> None:
+        """Called to produce a text only PDF.
+
+        Args:
+            input_file: A page image on which to perform OCR.
+            output_pdf: The expected name of the output PDF, which must be
+                a single page PDF with no visible content of any kind, sized
+                to the dimensions implied by the input_file's width, height
+                and DPI. The image will be grafted onto the input PDF page.
+        """
+
+
+@hookspec(firstresult=True)
+def get_ocr_engine() -> OcrEngine:
+    """Returns an OcrEngine to use for processing this file.
+
+    The OcrEngine may be instantiated multiple times, by both the main process
+    and child process. As such, it must be obtain store any state in ``options``
+    or some common location.
+    """
+
+
+@hookspec(firstresult=True)
+def generate_pdfa(
+    pdf_pages: List[Path],
+    pdfmark: Path,
+    output_file: Path,
+    compression: str,
+    pdf_version: str,
+    pdfa_part: str,
+) -> Path:
+    """Generate a PDF/A.
+
+    This API strongly assumes a PDF/A generator with Ghostscript's semantics.
+
+    OCRmyPDF will modify the metadata and possibly linearize the PDF/A after it
+    is generated.
+
+    Arguments:
+        pdf_pages: A list of one or more filenames, will be merged into output_file.
+        pdfmark: A PostScript file intended for Ghostscript with details on
+            how to perform the PDF/A conversion.
+        output_file: The name of the desired output file.
+        compression: One of ``'jpeg'``, ``'lossless'``, ``''``. For ``'jpeg'``,
+            the PDF/A generator should convert all images to JPEG encoding where
+            possible. For lossless, all images should be converted to FlateEncode
+            (lossless PNG). If an empty string, the PDF generator should make its
+            own decisions about how to encode images.
+        pdf_version: The minimum PDF version that the output file should be.
+            At its own discretion, the PDF/A generator may raise the version,
+            but should not lower it.
+        pdfa_part: The desired PDF/A compliance level, such as ``'2B'``.
+
+    Returns:
+        output_file: If successful, the hook should return ``output_file``.
+    """

src/ocrmypdf/quality.py

@@ -0,0 +1,60 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import re
+from typing import Iterable
+
+"""Utilities to measure OCR quality"""
+
+
+class OcrQualityDictionary:
+    """Manages a dictionary for simple OCR quality checks."""
+
+    def __init__(self, *, wordlist: Iterable[str] = []):
+        """Construct a dictionary from a list of words.
+
+        Words for which capitalization is important should be capitalized in the
+        dictionary. Words that contain spaces or other punctuation will never match.
+        """
+        self.dictionary = set()
+        self.dictionary.update(w for w in wordlist)
+
+    def measure_words_matched(self, ocr_text: str) -> float:
+        """Check how many unique words in the OCR text match a dictionary.
+
+        Words with mixed capitalized are only considered a match if the test word
+        matches that capitalization.
+
+        Returns:
+            number of words that match / number
+        """
+        text = re.sub(r"[0-9_]+", ' ', ocr_text)
+        text = re.sub(r'\W+', ' ', text)
+        text_words_list = re.split(r'\s+', text)
+        text_words = {w for w in text_words_list if len(w) >= 3}
+
+        matches = 0
+        for w in text_words:
+            if w in self.dictionary or (
+                w != w.lower() and w.lower() in self.dictionary
+            ):
+                matches += 1
+        if matches > 0:
+            hit_ratio = matches / len(text_words)
+        else:
+            hit_ratio = 0.0
+        return hit_ratio

src/ocrmypdf/subprocess.py

@@ -1,4 +1,4 @@
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # This file is part of OCRmyPDF.
 #
@@ -20,16 +20,94 @@
 import logging
 import os
 import re
+import shutil
 import sys
-from subprocess import run, STDOUT, PIPE, CalledProcessError
-from ..exceptions import MissingDependencyError, ExitCode
 from collections.abc import Mapping
+from contextlib import suppress
+from distutils.version import LooseVersion
+from functools import lru_cache
+from pathlib import Path
+from subprocess import PIPE, STDOUT, CalledProcessError
+from subprocess import run as subprocess_run
 
-log = logging.Logger(__name__)
+from ocrmypdf.exceptions import MissingDependencyError
+
+log = logging.getLogger(__name__)
 
 
+def run(args, *, env=None, **kwargs):
+    """Wrapper around subprocess.run()
+
+    The main purpose of this wrapper is to log subprocess output.
+
+    Secondly we have to account for behavioral differences in Windows in particular.
+    Creating symbolic links in Windows requires administrator privileges and
+    may not work if for some reason we're using a FAT file system or the temporary
+    folder is on a different drive from the working folder. The test suite
+    works around this by creating shim Python scripts that perform the same function
+    as a symbolic link, but those shims require support on this side, to ensure
+    we call them with Python.
+
+    """
+    if not env:
+        env = os.environ
+
+    # Search in spoof path if necessary
+    program = args[0]
+
+    if os.name == 'nt':
+        args = _fix_windows_args(program, args, env)
+
+    log.debug("Running: %s", args)
+    process_log = log.getChild('subprocess.' + os.path.basename(program))
+    if sys.version_info < (3, 7) and os.name == 'nt':
+        # Can't use close_fds=True on Windows with Python 3.6 or older
+        # https://bugs.python.org/issue19575, etc.
+        kwargs['close_fds'] = False
+
+    stderr = None
+    try:
+        proc = subprocess_run(args, env=env, **kwargs)
+    except CalledProcessError as e:
+        stderr = getattr(e, 'stderr', None)
+        raise
+    else:
+        stderr = getattr(proc, 'stderr', None)
+    finally:
+        if process_log.isEnabledFor(logging.DEBUG) and stderr:
+            with suppress(AttributeError, UnicodeDecodeError):
+                stderr = stderr.decode('utf-8', 'replace')
+            process_log.debug("stderr = %s", stderr)
+    return proc
+
+
+def _fix_windows_args(program, args, env):
+    """Adjust our desired program and command line arguments for use on Windows"""
+
+    if sys.version_info < (3, 8):
+        # bpo-33617 - Windows needs manual Path -> str conversion
+        args = [os.fspath(arg) for arg in args]
+        program = os.fspath(program)
+
+    # If we are running a .py on Windows, ensure we call it with this Python
+    # (to support test suite shims)
+    if program.lower().endswith('.py'):
+        args = [sys.executable] + args
+
+    paths = os.pathsep.join(os.get_exec_path(env))
+    if not shutil.which(args[0], path=paths):
+        # If the program we want is not on the PATH, add some interesting
+        # locations in %PROGRAMFILES% to the PATH and try again
+        shimmed_path = shim_paths_with_program_files(env)
+        new_args0 = shutil.which(args[0], path=shimmed_path)
+        if new_args0:
+            args[0] = new_args0
+    return args
+
+
+@lru_cache(maxsize=None)
 def get_version(program, *, version_arg='--version', regex=r'(\d+(\.\d+)*)', env=None):
-    "Get the version of the specified program"
+    """Get the version of the specified program"""
     args_prog = [program, version_arg]
     try:
         proc = run(
@@ -65,6 +143,33 @@ def get_version(program, *, version_arg='--version', regex=r'(\d+(\.\d+)*)', env
     return version
 
 
+def shim_paths_with_program_files(env=None):
+    if not env:
+        env = os.environ
+    program_files = env.get('PROGRAMFILES', '')
+    if not program_files:
+        return env.get('PATH', '')
+
+    def path_walker():
+        for path in Path(program_files).iterdir():
+            if not path.is_dir():
+                continue
+            if path.name.lower() == 'tesseract-ocr':
+                yield path
+            elif path.name.lower() == 'gs':
+                yield from (p for p in path.glob('**/bin') if p.is_dir())
+
+    paths = sorted(
+        (p for p in path_walker()), key=lambda p: (p.name, p.parent.name), reverse=True
+    )
+    paths.extend(
+        Path(str_path)
+        for str_path in os.get_exec_path(env)
+        if Path(str_path) not in set(paths)
+    )
+    return os.pathsep.join(str(p) for p in paths)
+
+
 missing_program = '''
 The program '{program}' could not be executed or was not found on your
 system PATH.
@@ -110,30 +215,40 @@ On RPM-based systems (Red Hat, Fedora), search for instructions on
 installing the RPM for {program}.
 '''
 
+windows_install_advice = '''
+If not already installed, install the Chocolatey package manager. Then use
+a command prompt to install the missing package:
+    choco install {package}
+'''
+
 
 def _get_platform():
     if sys.platform.startswith('freebsd'):
         return 'freebsd'
     elif sys.platform.startswith('linux'):
         return 'linux'
+    elif sys.platform.startswith('win'):
+        return 'windows'
     return sys.platform
 
 
 def _error_trailer(program, package, **kwargs):
     if isinstance(package, Mapping):
-        package = package[_get_platform()]
+        package = package.get(_get_platform(), program)
 
     if _get_platform() == 'darwin':
         log.info(osx_install_advice.format(**locals()))
     elif _get_platform() == 'linux':
         log.info(linux_install_advice.format(**locals()))
+    elif _get_platform() == 'windows':
+        log.info(windows_install_advice.format(**locals()))
 
 
 def _error_missing_program(program, package, required_for, recommended):
-    if required_for:
+    if recommended:
+        log.warning(missing_recommend_program.format(**locals()))
+    elif required_for:
         log.error(missing_optional_program.format(**locals()))
-    elif recommended:
-        log.info(missing_recommend_program.format(**locals()))
     else:
         log.error(missing_program.format(**locals()))
     _error_trailer(**locals())
@@ -155,20 +270,27 @@ def check_external_program(
     need_version,
     required_for=None,
     recommended=False,
-    **kwargs,  # To consume log parameter
 ):
-    if kwargs:
-        if not 'log' in kwargs:
-            log.warning('check_external_program(log=...) is deprecated')
     try:
-        found_version = version_checker()
+        if callable(version_checker):
+            found_version = version_checker()
+        else:
+            found_version = version_checker
     except (CalledProcessError, FileNotFoundError, MissingDependencyError):
         _error_missing_program(program, package, required_for, recommended)
         if not recommended:
             raise MissingDependencyError()
         return
 
-    if found_version < need_version:
+    def remove_leading_v(s):
+        if s.startswith('v'):
+            return s[1:]
+        return s
+
+    found_version = remove_leading_v(found_version)
+    need_version = remove_leading_v(need_version)
+
+    if found_version and LooseVersion(found_version) < LooseVersion(need_version):
         _error_old_version(program, package, need_version, found_version, required_for)
         if not recommended:
             raise MissingDependencyError()

tests/cache/3small/__-l__eng__000001_ocr.png__000001_ocr_tess__pdf__txt/stderr.bin

@@ -0,0 +1 @@
+Tesseract Open Source OCR Engine v4.1.0 with Leptonica

tests/cache/3small/__-l__eng__000001_ocr.png__000001_ocr_tess__pdf__txt/txt.bin

@@ -0,0 +1,30 @@
+Tarnose
+
+  
+  
+
+ 
+
+Bokale oa
+
+ 
+
+Lehuntze
+
+   
+  
+ 
+
+Mugerre
+
+ 
+  
+
+Milafranga Komunikabideak
+
+BAIONA i zeettnansise —
+
+1 Trenbideak -- ~~~
+
+t\ Basusarri — spmsans20141004 se: . a ~
+

tests/cache/3small/__-l__eng__000002_ocr.png__000002_ocr_tess__pdf__txt/stderr.bin

@@ -0,0 +1 @@
+Tesseract Open Source OCR Engine v4.1.0 with Leptonica

tests/cache/3small/__-l__eng__000002_ocr.png__000002_ocr_tess__pdf__txt/txt.bin

@@ -0,0 +1,2 @@
+Covfefe is a perfectly cromulent word.
+

tests/cache/3small/__-l__eng__000003_ocr.png__000003_ocr_tess__pdf__txt/stderr.bin

@@ -0,0 +1 @@
+Tesseract Open Source OCR Engine v4.1.0 with Leptonica

tests/cache/3small/__-l__eng__000003_ocr.png__000003_ocr_tess__pdf__txt/txt.bin

@@ -0,0 +1,27 @@
+Linzensoep a la Waterman
+
+ 
+
+4 ons linzen
+
+3 liter water
+
+3 uien
+
+bloem, boter
+
+2 kopjes melk
+
+laurier, kruidnagel, kerrie, zout
+
+De linzgen wassen en in -l liter kokend wa-
+ter 1 dag laten weken, 2 liter water bij
+de linzen voegen, zonder het water waarin
+ze geweekt zijn af te gieten, De helft van
+de uien bakken met laurier en Kruicdnagel.
+Alle uien, kerrie en gout bij de linzen
+voegen, Alles aan de kook brengen,. Van de
+bloem met boter en melk een papje maken en
+verder afmaken met de soep, Als de linzen
+gaar Zijn is de soep klaar.
+