pytest picky about list vs tuple

Fixes #523

.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/*
 

.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
 
@@ -49,10 +50,10 @@ RUN apt-get update && apt-get install -y --no-install-recommends \
   ghostscript \
   img2pdf \
   liblept5 \
+  libsm6 libxext6 libxrender-dev \
   zlib1g \
   pngquant \
   python3 \
-  python3-venv \
   qpdf \
   tesseract-ocr \
   tesseract-ocr-chi-sim \
@@ -61,11 +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/
@@ -73,7 +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 +1,36 @@
+---
+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**
+Please include an example *input* PDF (or image). The input file is more helpful.
+
+If possible, use an input file with no personal or confidential information. At your option you may GPG-encrypt the file for OCRmyPDF's author only.
+
+**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 +1,251 @@
+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 $(python.version)"
+            condition: succeededOrFailed()
+      - job: "macOS_Mojave"
+        pool:
+          vmImage: "macos-10.14"
+        strategy:
+          matrix:
+            Python37:
+              python.version: "3.7"
+            Python38:
+              python.version: "3.8"
+        steps:
+          - task: UsePythonVersion@0
+            inputs:
+              versionSpec: "$(python.version)"
+          - bash: |
+              brew update
+              brew unlink python@2
+              brew install \
+                exempi \
+                ghostscript \
+                jbig2enc \
+                leptonica \
+                openjpeg \
+                pngquant \
+                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

@@ -1,16 +1,34 @@
+=================
 Advanced features
 =================
 
 Control of unpaper
-------------------
+==================
 
-OCRmyPDF uses ``unpaper`` to provide the implementation of the ``--clean`` and ``--clean-final`` arguments. `unpaper <https://github.com/Flameeyes/unpaper/blob/master/doc/basic-concepts.md>`_ provides a variety of image processing filters to improve images.
+OCRmyPDF uses ``unpaper`` to provide the implementation of the
+``--clean`` and ``--clean-final`` arguments.
+`unpaper <https://github.com/Flameeyes/unpaper/blob/master/doc/basic-concepts.md>`__
+provides a variety of image processing filters to improve images.
 
-By default, OCRmyPDF uses only ``unpaper`` arguments that were found to be safe to use on almost all files without having to inspect every page of the file afterwards. This is particularly true when only ``--clean`` is used, since that instructs OCRmyPDF to only clean the image before OCR and not the final image.
+By default, OCRmyPDF uses only ``unpaper`` arguments that were found to
+be safe to use on almost all files without having to inspect every page
+of the file afterwards. This is particularly true when only ``--clean``
+is used, since that instructs OCRmyPDF to only clean the image before
+OCR and not the final image.
 
-However, if you wish to use the more aggressive options in ``unpaper``, you may use ``--unpaper-args '...'`` to override the OCRmyPDF's defaults and forward other arguments to unpaper. This option will forward arguments to ``unpaper`` without any knowledge of what that program considers to be valid arguments. The string of arguments must be quoted as shown in the examples below. No filename arguments may be included. OCRmyPDF will assume it can append input and output filename of intermediate images to the ``--unpaper-args`` string.
+However, if you wish to use the more aggressive options in ``unpaper``,
+you may use ``--unpaper-args '...'`` to override the OCRmyPDF's defaults
+and forward other arguments to unpaper. This option will forward
+arguments to ``unpaper`` without any knowledge of what that program
+considers to be valid arguments. The string of arguments must be quoted
+as shown in the examples below. No filename arguments may be included.
+OCRmyPDF will assume it can append input and output filename of
+intermediate images to the ``--unpaper-args`` string.
 
-In this example, we tell ``unpaper`` to expect two pages of text on a sheet (image), such as occurs when two facing pages of a book are scanned. ``unpaper`` uses this information to deskew each independently and clean up the margins of both.
+In this example, we tell ``unpaper`` to expect two pages of text on a
+sheet (image), such as occurs when two facing pages of a book are
+scanned. ``unpaper`` uses this information to deskew each independently
+and clean up the margins of both.
 
 .. code-block:: bash
 
@@ -19,40 +37,71 @@ In this example, we tell ``unpaper`` to expect two pages of text on a sheet (ima
 
 .. warning::
 
-    Some ``unpaper`` features will reposition text within the image. ``--clean-final`` is recommended to avoid this issue.
+   Some ``unpaper`` features will reposition text within the image.
+   ``--clean-final`` is recommended to avoid this issue.
 
 .. warning::
 
-    Some ``unpaper`` features cause multiple input or output files to be consumed or produced. OCRmyPDF requires ``unpaper`` to consume one file and produce one file. An deviation from that condition will result in errors.
+   Some ``unpaper`` features cause multiple input or output files to be
+   consumed or produced. OCRmyPDF requires ``unpaper`` to consume one
+   file and produce one file. An deviation from that condition will
+   result in errors.
 
 .. note::
 
-    ``unpaper`` uses uncompressed PBM/PGM/PPM files for its intermediate files. For large images or documents, it can take a lot of temporary disk space.
+   ``unpaper`` uses uncompressed PBM/PGM/PPM files for its intermediate
+   files. For large images or documents, it can take a lot of temporary
+   disk space.
 
 Control of OCR options
-----------------------
+======================
 
-OCRmyPDF provides many features to control the behavior of the OCR engine, Tesseract.
+OCRmyPDF provides many features to control the behavior of the OCR
+engine, Tesseract.
 
 When OCR is skipped
-"""""""""""""""""""
+-------------------
 
-If a page in a PDF seems to have text, by default OCRmyPDF will exit without modifying the PDF. This is to ensure that PDFs that were previously OCRed or were "born digital" rather than scanned are not processed.
+If a page in a PDF seems to have text, by default OCRmyPDF will exit
+without modifying the PDF. This is to ensure that PDFs that were
+previously OCRed or were "born digital" rather than scanned are not
+processed.
 
-If ``--skip-text`` is issued, then no OCR will be performed on pages that already have text. The page will be copied to the output. This may be useful for documents that contain both "born digital" and scanned content, or to use OCRmyPDF to normalize and convert to PDF/A regardless of their contents.
+If ``--skip-text`` is issued, then no OCR will be performed on pages
+that already have text. The page will be copied to the output. This may
+be useful for documents that contain both "born digital" and scanned
+content, or to use OCRmyPDF to normalize and convert to PDF/A regardless
+of their contents.
 
-If ``--redo-ocr`` is issued, then a detailed text analysis is performed. Text is categorized as either visible or invisible. Invisible text (OCR) is stripped out. Then an image of each page is created with visible text masked out. The page image is sent for OCR, and any additional text is inserted as OCR. If a file contains a mix of text and bitmap images that contain text, OCRmyPDF will locate the additional text in images without disrupting the existing text.
+If ``--redo-ocr`` is issued, then a detailed text analysis is performed.
+Text is categorized as either visible or invisible. Invisible text (OCR)
+is stripped out. Then an image of each page is created with visible text
+masked out. The page image is sent for OCR, and any additional text is
+inserted as OCR. If a file contains a mix of text and bitmap images that
+contain text, OCRmyPDF will locate the additional text in images without
+disrupting the existing text.
 
-If ``--force-ocr`` is issued, then all pages will be rasterized to images, discarding any hidden OCR text, and rasterizing any printable text. This is useful for redoing OCR, for fixing OCR text with a damaged character map (text is selectable but not searchable), and destroying redacted information. Any forms and vector graphics will be rasterized as well.
+If ``--force-ocr`` is issued, then all pages will be rasterized to
+images, discarding any hidden OCR text, and rasterizing any printable
+text. This is useful for redoing OCR, for fixing OCR text with a damaged
+character map (text is selectable but not searchable), and destroying
+redacted information. Any forms and vector graphics will be rasterized
+as well.
 
 Time and image size limits
-""""""""""""""""""""""""""
+--------------------------
 
-By default, OCRmyPDF permits tesseract to run for three minutes (180 seconds) per page. This is usually more than enough time to find all text on a reasonably sized page with modern hardware.
+By default, OCRmyPDF permits tesseract to run for three minutes (180
+seconds) per page. This is usually more than enough time to find all
+text on a reasonably sized page with modern hardware.
 
-If a page is skipped, it will be inserted without OCR. If preprocessing was requested, the preprocessed image layer will be inserted.
+If a page is skipped, it will be inserted without OCR. If preprocessing
+was requested, the preprocessed image layer will be inserted.
 
-If you want to adjust the amount of time spent on OCR, change ``--tesseract-timeout``.  You can also automatically skip images that exceed a certain number of megapixels with ``--skip-big``. (A 300 DPI, 8.5×11" page is 8.4 megapixels.)
+If you want to adjust the amount of time spent on OCR, change
+``--tesseract-timeout``. You can also automatically skip images that
+exceed a certain number of megapixels with ``--skip-big``. (A 300 DPI,
+8.5×11" page is 8.4 megapixels.)
 
 .. code-block:: bash
 
@@ -60,21 +109,27 @@ If you want to adjust the amount of time spent on OCR, change ``--tesseract-time
     ocrmypdf --tesseract-timeout 300 --skip-big 50 bigfile.pdf output.pdf
 
 Overriding default tesseract
-""""""""""""""""""""""""""""
+----------------------------
 
 OCRmyPDF checks the system ``PATH`` for the ``tesseract`` binary.
 
-Some relevant environment variables that influence Tesseract's behavior include:
+Some relevant environment variables that influence Tesseract's behavior
+include:
 
 .. envvar:: TESSDATA_PREFIX
 
-    Overrides the path to Tesseract's data files. This can allow simultaneous installation of the "best" and "fast" training data sets. OCRmyPDF does not manage this environment variable.
+   Overrides the path to Tesseract's data files. This can allow
+   simultaneous installation of the "best" and "fast" training data
+   sets. OCRmyPDF does not manage this environment variable.
 
 .. envvar:: OMP_THREAD_LIMIT
 
-    Controls the number of threads Tesseract will use. OCRmyPDF will manage this environment if it is not already set. (Currently, it will set it to 1 because this gives the best results in testing.)
+   Controls the number of threads Tesseract will use. OCRmyPDF will
+   manage this environment if it is not already set. (Currently, it will
+   set it to 1 because this gives the best results in testing.)
 
-For example, if you have a development build of Tesseract don't wish to use the system installation, you can launch OCRmyPDF as follows:
+For example, if you have a development build of Tesseract don't wish to
+use the system installation, you can launch OCRmyPDF as follows:
 
 .. code-block:: bash
 
@@ -83,26 +138,33 @@ For example, if you have a development build of Tesseract don't wish to use the
         TESSDATA_PREFIX=/home/user/src/tesseract \
         ocrmypdf input.pdf output.pdf
 
-In this example ``TESSDATA_PREFIX`` is required to redirect Tesseract to an alternate folder for its "tessdata" files.
+In this example ``TESSDATA_PREFIX`` is required to redirect Tesseract to
+an alternate folder for its "tessdata" files.
 
 Overriding other support programs
-"""""""""""""""""""""""""""""""""
+---------------------------------
 
 In addition to tesseract, OCRmyPDF uses the following external binaries:
 
-* ``gs`` (Ghostscript)
-* ``unpaper``
-* ``qpdf``
-
-In each case OCRmyPDF will search the ``PATH`` environment variable to locate the binaries.
+-  ``gs`` (Ghostscript)
+-  ``unpaper``
+-  ``qpdf``
 
+In each case OCRmyPDF will search the ``PATH`` environment variable to
+locate the binaries.
 
 Changing tesseract configuration variables
-""""""""""""""""""""""""""""""""""""""""""
+------------------------------------------
 
-You can override tesseract's default `control parameters <https://github.com/tesseract-ocr/tesseract/wiki/ControlParams>`_ with a configuration file.
+You can override tesseract's default `control
+parameters <https://github.com/tesseract-ocr/tesseract/wiki/ControlParams>`__
+with a configuration file.
 
-As an example, this configuration will disable Tesseract's dictionary for current language. Normally the dictionary is helpful for interpolating words that are unclear, but it may interfere with OCR if the document does not contain many words (for example, a list of part numbers).
+As an example, this configuration will disable Tesseract's dictionary
+for current language. Normally the dictionary is helpful for
+interpolating words that are unclear, but it may interfere with OCR if
+the document does not contain many words (for example, a list of part
+numbers).
 
 Create a file named "no-dict.cfg" with these contents:
 
@@ -120,11 +182,11 @@ then run ocrmypdf as follows (along with any other desired arguments):
 
 .. warning::
 
-    Some combinations of control parameters will break Tesseract or break assumptions that OCRmyPDF makes about Tesseract's output.
-
+   Some combinations of control parameters will break Tesseract or break
+   assumptions that OCRmyPDF makes about Tesseract's output.
 
 Changing the PDF renderer
--------------------------
+=========================
 
 rasterizing
   Converting a PDF to an image for display.
@@ -132,42 +194,63 @@ rasterizing
 rendering
   Creating a new PDF from other data (such as an existing PDF).
 
-
-OCRmyPDF has these PDF renderers: ``sandwich`` and ``hocr``. The renderer may be selected using ``--pdf-renderer``. The default is ``auto`` which lets OCRmyPDF select the renderer to use. Currently, ``auto`` always selects ``sandwich``.
+OCRmyPDF has these PDF renderers: ``sandwich`` and ``hocr``. The
+renderer may be selected using ``--pdf-renderer``. The default is
+``auto`` which lets OCRmyPDF select the renderer to use. Currently,
+``auto`` always selects ``sandwich``.
 
 The ``sandwich`` renderer
-"""""""""""""""""""""""""
+-------------------------
 
-The ``sandwich`` renderer uses Tesseract's new text-only PDF feature, which produces a PDF page that lays out the OCR in invisible text. This page is then "sandwiched" onto the original PDF page, allowing lossless application of OCR even to PDF pages that contain other vector objects.
+The ``sandwich`` renderer uses Tesseract's new text-only PDF feature,
+which produces a PDF page that lays out the OCR in invisible text. This
+page is then "sandwiched" onto the original PDF page, allowing lossless
+application of OCR even to PDF pages that contain other vector objects.
 
-Currently this is the best renderer for most uses, however it is implemented in Tesseract so OCRmyPDF cannot influence it. Currently some problematic PDF viewers like Mozilla PDF.js and macOS Preview have problems with segmenting its text output, and mightrunseveralwordstogether.
+Currently this is the best renderer for most uses, however it is
+implemented in Tesseract so OCRmyPDF cannot influence it. Currently some
+problematic PDF viewers like Mozilla PDF.js and macOS Preview have
+problems with segmenting its text output, and
+mightrunseveralwordstogether.
 
-When image preprocessing features like ``--deskew`` are used, the original PDF will be rendered as a full page and the OCR layer will be placed on top.
+When image preprocessing features like ``--deskew`` are used, the
+original PDF will be rendered as a full page and the OCR layer will be
+placed on top.
 
 The ``hocr`` renderer
-"""""""""""""""""""""
+---------------------
 
-The ``hocr`` renderer works with older versions of Tesseract. The image layer is copied from the original PDF page if possible, avoiding potentially lossy transcoding or loss of other PDF information. If preprocessing is specified, then the image layer is a new PDF.
+The ``hocr`` renderer works with older versions of Tesseract. The image
+layer is copied from the original PDF page if possible, avoiding
+potentially lossy transcoding or loss of other PDF information. If
+preprocessing is specified, then the image layer is a new PDF.
 
-Unlike ``sandwich`` this renderer is implemented within OCRmyPDF; anyone looking to customize how OCR is presented should look here. A major disadvantage of this renderer is it not capable of correctly handling text outside the Latin alphabet. Pull requests to improve the situation are welcome.
+Unlike ``sandwich`` this renderer is implemented within OCRmyPDF; anyone
+looking to customize how OCR is presented should look here. A major
+disadvantage of this renderer is it not capable of correctly handling
+text outside the Latin alphabet. Pull requests to improve the situation
+are welcome.
 
-Currently, this renderer has the best compatibility with Mozilla's PDF.js viewer.
+Currently, this renderer has the best compatibility with Mozilla's
+PDF.js viewer.
 
 This works in all versions of Tesseract.
 
 The ``tesseract`` renderer
-""""""""""""""""""""""""""
+--------------------------
 
-The ``tesseract`` renderer was removed. OCRmyPDF's new approach to text layer grafting makes it functionally equivalent to ``sandwich``.
+The ``tesseract`` renderer was removed. OCRmyPDF's new approach to text
+layer grafting makes it functionally equivalent to ``sandwich``.
 
 Return code policy
-------------------
+==================
 
-OCRmyPDF writes all messages to ``stderr``.  ``stdout`` is reserved for piping
-output files.  ``stdin`` is reserved for piping input files.
+OCRmyPDF writes all messages to ``stderr``. ``stdout`` is reserved for
+piping output files. ``stdin`` is reserved for piping input files.
 
-The return codes generated by the OCRmyPDF are considered part of the stable
-user interface.  They may be imported from ``ocrmypdf.exceptions``.
+The return codes generated by the OCRmyPDF are considered part of the
+stable user interface. They may be imported from
+``ocrmypdf.exceptions``.
 
 .. list-table:: Return codes
     :widths: 5 35 60
@@ -218,22 +301,44 @@ user interface.  They may be imported from ``ocrmypdf.exceptions``.
 
 
 Debugging the intermediate files
---------------------------------
+================================
 
-OCRmyPDF normally saves its intermediate results to a temporary folder and deletes this folder when it exits, whether it succeeded or failed.
+OCRmyPDF normally saves its intermediate results to a temporary folder
+and deletes this folder when it exits, whether it succeeded or failed.
 
-If the ``-k`` argument is issued on the command line, OCRmyPDF will keep the temporary folder and print the location, whether it succeeded or failed (provided the Python interpreter did not crash). An example message is:
+If the ``-k`` argument is issued on the command line, OCRmyPDF will keep
+the temporary folder and print the location, whether it succeeded or
+failed (provided the Python interpreter did not crash). An example
+message is:
 
 .. code-block:: none
 
-    Temporary working files saved at:
+    Temporary working files retained at:
     /tmp/com.github.ocrmypdf.u20wpz07
 
-The organization of this folder is an implementation detail and subject to change between releases. However the general organization is that 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:
+The organization of this folder is an implementation detail and subject
+to change between releases. However the general organization is that
+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 on arguments this may differ from the presentation image
-* ``layers.rendered.pdf`` - the composite PDF, before metadata repair and optimization
-* ``images/*`` - images extracted during the optimization process; here the prefix indicates a PDF object ID not a page number
+-  ``_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
+-  ``_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

@@ -1,65 +1,105 @@
+======================
 Using the OCRmyPDF API
 ======================
 
-OCRmyPDF originated as a command line program and continues to have this legacy, but parts of it can be imported and used in other Python applications.
+OCRmyPDF originated as a command line program and continues to have this
+legacy, but parts of it can be imported and used in other Python
+applications.
 
-Some applications may want to consider running ocrmypdf from a subprocess call anyway, as this provides isolation of its activities.
+Some applications may want to consider running ocrmypdf from a
+subprocess call anyway, as this provides isolation of its activities.
 
 Example
--------
+=======
 
-OCRmyPDF one high-level function to run its main engine from an application. The parameters are symmetric to the command line arguments and largely have the same functions.
+OCRmyPDF one high-level function to run its main engine from an
+application. The parameters are symmetric to the command line arguments
+and largely have the same functions.
 
 .. code-block:: python
 
-    from ocrmypdf import ocrmypdf
+    import ocrmypdf
 
-    ocrmypdf('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.
+With a few exceptions, all of the command line arguments are available
+and may be passed as equivalent keywords.
 
-A few differences are that ``verbose`` and ``quiet`` are not available. Instead, output should be managed by configuring logging.
+A few differences are that ``verbose`` and ``quiet`` are not available.
+Instead, output should be managed by configuring logging.
 
 Parent process requirements
-^^^^^^^^^^^^^^^^^^^^^^^^^^^
+---------------------------
+
+The :func:`ocrmypdf.ocr` function runs OCRmyPDF similar to command line
+execution. To do this, it will:
 
-The :func:`ocrmypdf.ocrmypdf` 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)
+- execute other subprocesses (forking and executing other programs)
 
-The Python process that calls ``ocrmypdf()`` must be sufficiently privileged to perform these actions. If it is not, ``ocrmypdf()`` will fail.
+The Python process that calls ``ocrmypdf.ocr()`` must be sufficiently
+privileged to perform these actions. If it is not, ``ocrmypdf()`` will
+fail.
 
-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.
+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()`` is suggested. That way your application will survive even if OCRmyPDF does not.
+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 fork semantics will prevent OCRmyPDF from working
+    correct.
 
 Logging
-^^^^^^^
+-------
 
-OCRmyPDF will log under loggers named ``ocrmypdf``. In addition, it imports ``pdfminer`` and ``PIL``, both of which post log messages under those logging namespaces.
+OCRmyPDF will log under loggers named ``ocrmypdf``. In addition, it
+imports ``pdfminer`` and ``PIL``, both of which post log messages under
+those logging namespaces.
 
-You can configure the logging as desired for your application or call :func:`ocrmypdf.configure_logging` to configure logging the same way OCRmyPDF itself does. The command line parameters such as ``--quiet`` and ``--verbose`` have no equivalents in the API; you must configure logging.
+You can configure the logging as desired for your application or call
+:func:`ocrmypdf.configure_logging` to configure logging the same way
+OCRmyPDF itself does. The command line parameters such as ``--quiet``
+and ``--verbose`` have no equivalents in the API; you must use the
+provided configuration function or do configuration in a way that suits
+your use case.
 
 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.
+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. Use ``ocrmypdf.ocr(...progress_bar=False)`` to disable
+the progress bar.
 
 Exceptions
-^^^^^^^^^^
+----------
 
-OCRmyPDF may throw standard Python exceptions, ``ocrmypdf.exceptions.*`` exceptions, some exceptions related to multiprocessing, and ``KeyboardInterrupt``. The parent process should provide an exception handler. OCRmyPDF will clean up its temporary files and worker processes automatically when an exception occurs.
+OCRmyPDF may throw standard Python exceptions, ``ocrmypdf.exceptions.*``
+exceptions, some exceptions related to multiprocessing, and
+``KeyboardInterrupt``. The parent process should provide an exception
+handler. OCRmyPDF will clean up its temporary files and worker processes
+automatically when an exception occurs.
 
-Programs that call OCRmyPDF should consider trapping KeyboardInterrupt so that they allow OCR to terminate with the whole program terminating.
+Programs that call OCRmyPDF should consider trapping KeyboardInterrupt
+so that they allow OCR to terminate with the whole program terminating.
 
-When OCRmyPDF succeeds conditionally, it may return an integer exit code.
+When OCRmyPDF succeeds conditionally, it returns an integer exit code.
 
 Reference
 ---------
 
-.. autofunction:: ocrmypdf.ocrmypdf
+.. autofunction:: ocrmypdf.ocr
 
 .. autoclass:: ocrmypdf.Verbosity
     :members:

docs/batch.rst

@@ -1,215 +1,224 @@
+================
 Batch processing
 ================
 
-This article provides information about running OCRmyPDF on multiple files or configuring it as a service triggered by file system events.
+This article provides information about running OCRmyPDF on multiple
+files or configuring it as a service triggered by file system events.
 
 Batch jobs
-----------
+==========
 
-Consider using the excellent `GNU Parallel <https://www.gnu.org/software/parallel/>`_ to apply OCRmyPDF to multiple files at once.
+Consider using the excellent `GNU
+Parallel <https://www.gnu.org/software/parallel/>`__ to apply OCRmyPDF
+to multiple files at once.
 
-Both ``parallel`` and ``ocrmypdf`` will try to use all available processors. To maximize parallelism without overloading your system with processes, consider using ``parallel -j 2`` to limit parallel to running two jobs at once.
+Both ``parallel`` and ``ocrmypdf`` will try to use all available
+processors. To maximize parallelism without overloading your system with
+processes, consider using ``parallel -j 2`` to limit parallel to running
+two jobs at once.
 
-This command will run all ocrmypdf all files named ``*.pdf`` in the current directory and write them to the previous created ``output/`` folder. It will not search subdirectories.
+This command will run all ocrmypdf all files named ``*.pdf`` in the
+current directory and write them to the previous created ``output/``
+folder. It will not search subdirectories.
 
-The ``--tag`` argument tells parallel to print the filename as a prefix whenever a message is printed, so that one can trace any errors to the file that produced them.
+The ``--tag`` argument tells parallel to print the filename as a prefix
+whenever a message is printed, so that one can trace any errors to the
+file that produced them.
 
 .. code-block:: bash
 
-	parallel --tag -j 2 ocrmypdf '{}' 'output/{}' ::: *.pdf
+   parallel --tag -j 2 ocrmypdf '{}' 'output/{}' ::: *.pdf
 
-OCRmyPDF automatically repairs PDFs before parsing and gathering information from them.
+OCRmyPDF automatically repairs PDFs before parsing and gathering
+information from them.
 
 Directory trees
----------------
+===============
 
-This will walk through a directory tree and run OCR on all files in place, printing the output in a way that makes
+This will walk through a directory tree and run OCR on all files in
+place, printing the output in a way that makes
 
 .. code-block:: bash
 
-	find . -printf '%p' -name '*.pdf' -exec ocrmypdf '{}' '{}' \;
+   find . -printf '%p' -name '*.pdf' -exec ocrmypdf '{}' '{}' \;
 
-Alternatively, with a docker container (mounts a volume to the container where the PDFs are stored):
+Alternatively, with a docker container (mounts a volume to the container
+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 of ``ocrmypdf``, again updating files in place.
+This only runs one ``ocrmypdf`` process at a time. This variation uses
+``find`` to create a directory list and ``parallel`` to parallelize runs
+of ``ocrmypdf``, again updating files in place.
 
 .. code-block:: bash
 
-	find . -name '*.pdf' | parallel --tag -j 2 ocrmypdf '{}' '{}'
+   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.
+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
-"""""""""""""""""""""
-
-Synology DiskStations (Network Attached Storage devices) can run the Docker image of OCRmyPDF if the Synology `Docker package <https://www.synology.com/en-global/dsm/packages/Docker>`_ is installed. Attached is a script to address particular quirks of using OCRmyPDF on one of these devices.
-
-This is only possible for x86-based Synology products. Some Synology 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')
-
-Huge batch jobs
-"""""""""""""""
-
-If you have thousands of files to work with, contact the author. Consulting work related to OCRmyPDF helps fund this open source project 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).
+Synology DiskStations (Network Attached Storage devices) can run the
+Docker image of OCRmyPDF if the Synology `Docker
+package <https://www.synology.com/en-global/dsm/packages/Docker>`__ is
+installed. Attached is a script to address particular quirks of using
+OCRmyPDF on one of these devices.
 
-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.
+This is only possible for x86-based Synology products. Some Synology
+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:: bash
+.. literalinclude:: ../misc/synology.py
+    :caption: misc/synology.py - Sample script for Synology DiskStations
 
-	pip install watchdog
-
-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.
-
-.. 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
-
-For more complex behavior you can write a Python script around to use the watchdog API.
-
-On file servers, you could configure watchmedo as a system service so it will run all the time.
-
-Caveats
-"""""""
-
-* ``watchmedo`` may not work properly on a networked file system, depending on the capabilities of the file system client and server.
-* This simple recipe does not filter for the type of file system event, so file copies, deletes and moves, and directory operations, will all be sent to ocrmypdf, producing errors in several cases. Disable your watched folder if you are doing anything other than copying files to it.
-* If the source and destination directory are the same, watchmedo may create an infinite loop.
-* On BSD, FreeBSD and older versions of macOS, you may need to increase the number of file descriptors to monitor more files, using ``ulimit -n 1024`` to watch a folder of up to 1024 files.
-
-Alternatives
-""""""""""""
-
-* `Watchman <https://facebook.github.io/watchman/>`_ is a more powerful alternative to ``watchmedo``.
-
-macOS Automator
+Huge batch jobs
 ---------------
 
-You can use the Automator app with macOS, to create a Workflow or Quick Action. Use a *Run Shell Script* action in your workflow. In the context 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:
+If you have thousands of files to work with, contact the author.
+Consulting work related to OCRmyPDF helps fund this open source project
+and all inquiries are appreciated.
 
-.. image:: images/macos-workflow.png
-	:alt: Example macOS Automator script
+Hot (watched) folders
+=====================
+
+Watched folders with watcher.py
+-------------------------------
+
+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
+
+    pip3 install -r requirements/watcher.txt
+
+    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
+
+    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
+
+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:
+
+.. 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
+-------
+
+-  ``watchmedo`` may not work properly on a networked file system,
+   depending on the capabilities of the file system client and server.
+-  This simple recipe does not filter for the type of file system event,
+   so file copies, deletes and moves, and directory operations, will all
+   be sent to ocrmypdf, producing errors in several cases. Disable your
+   watched folder if you are doing anything other than copying files to
+   it.
+-  If the source and destination directory are the same, watchmedo may
+   create an infinite loop.
+-  On BSD, FreeBSD and older versions of macOS, you may need to increase
+   the number of file descriptors to monitor more files, using
+   ``ulimit -n 1024`` to watch a folder of up to 1024 files.
+
+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
+===============
+
+You can use the Automator app with macOS, to create a Workflow or Quick
+Action. Use a *Run Shell Script* action in your workflow. In the context
+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:
+
+.. figure:: images/macos-workflow.png
+    :alt: Example macOS Automator workflow
 
 You may customize the command sent to ocrmypdf.

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

@@ -1,11 +1,12 @@
+========
 Cookbook
 ========
 
 Basic examples
---------------
+==============
 
 Help!
-^^^^^
+-----
 
 ocrmypdf has built-in help.
 
@@ -13,30 +14,29 @@ ocrmypdf has built-in help.
 
     ocrmypdf --help
 
-
 Add an OCR layer and convert to PDF/A
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+-------------------------------------
 
 .. code-block:: bash
 
     ocrmypdf input.pdf output.pdf
 
 Add an OCR layer and output a standard PDF
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+------------------------------------------
 
 .. code-block:: bash
 
     ocrmypdf --output-type pdf input.pdf output.pdf
 
 Create a PDF/A with all color and grayscale images converted to JPEG
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+--------------------------------------------------------------------
 
 .. code-block:: bash
 
     ocrmypdf --output-type pdfa --pdfa-image-compression jpeg input.pdf output.pdf
 
 Modify a file in place
-^^^^^^^^^^^^^^^^^^^^^^
+----------------------
 
 The file will only be overwritten if OCRmyPDF is successful.
 
@@ -45,48 +45,70 @@ The file will only be overwritten if OCRmyPDF is successful.
     ocrmypdf myfile.pdf myfile.pdf
 
 Correct page rotation
-^^^^^^^^^^^^^^^^^^^^^
+---------------------
 
-OCR will attempt to automatic correct the rotation of each page. This can help fix a scanning job that contains a mix of landscape and portrait pages.
+OCR will attempt to automatic correct the rotation of each page. This
+can help fix a scanning job that contains a mix of landscape and
+portrait pages.
 
 .. code-block:: bash
 
     ocrmypdf --rotate-pages myfile.pdf myfile.pdf
 
-You can increase (decrease) the parameter ``--rotate-pages-threshold`` to make page rotation more (less) aggressive.
+You can increase (decrease) the parameter ``--rotate-pages-threshold``
+to make page rotation more (less) aggressive.
 
-If the page is "just a little off horizontal", like a crooked picture, then you want ``--deskew``. ``--rotate-pages`` is for when the cardinal angle is wrong.
+If the page is "just a little off horizontal", like a crooked picture,
+then you want ``--deskew``. ``--rotate-pages`` is for when the cardinal
+angle is wrong.
 
 OCR languages other than English
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+--------------------------------
 
-OCRmyPDF assumes the document is in English unless told otherwise. OCR quality may be poor if the wrong language is used.
+OCRmyPDF assumes the document is in English unless told otherwise. OCR
+quality may be poor if the wrong language is used.
 
 .. code-block:: bash
 
     ocrmypdf -l fra LeParisien.pdf LeParisien.pdf
     ocrmypdf -l eng+fra Bilingual-English-French.pdf Bilingual-English-French.pdf
 
-Language packs must be installed for all languages specified. See :ref:`Installing additional language packs <lang-packs>`.
+Language packs must be installed for all languages specified. See
+:ref:`Installing additional language packs <lang-packs>`.
 
-Unfortunately, the Tesseract OCR engine has no ability to detect the language when it is unknown.
+Unfortunately, the Tesseract OCR engine has no ability to detect the
+language when it is unknown.
 
 Produce PDF and text file containing OCR text
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+---------------------------------------------
 
-This produces a file named "output.pdf" and a companion text file named "output.txt".
+This produces a file named "output.pdf" and a companion text file named
+"output.txt".
 
 .. code-block:: bash
 
     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
-^^^^^^^^^^^^^^^^^^^^
+--------------------
 
 Option: use Tesseract
-"""""""""""""""""""""
+~~~~~~~~~~~~~~~~~~~~~
 
-If you are starting with images, you can just use Tesseract directly to convert images to PDFs:
+If you are starting with images, you can just use Tesseract directly to
+convert images to PDFs:
 
 .. code-block:: bash
 
@@ -97,62 +119,88 @@ If you are starting with images, you can just use Tesseract directly to convert
     # When there are multiple images
     tesseract text-file-containing-list-of-image-filenames.txt output-prefix pdf
 
-Tesseract's PDF output is quite good – OCRmyPDF uses it internally, in some cases. However, OCRmyPDF has many features not available in Tesseract like image processing, metadata control, and PDF/A generation.
+Tesseract's PDF output is quite good – OCRmyPDF uses it internally, in
+some cases. However, OCRmyPDF has many features not available in
+Tesseract like image processing, metadata control, and PDF/A generation.
 
 Option: use img2pdf
-"""""""""""""""""""
+~~~~~~~~~~~~~~~~~~~
 
-You can also use a program like `img2pdf <https://gitlab.mister-muffin.de/josch/img2pdf>`_ to convert your images to PDFs, and then pipe the results to run ocrmypdf.  The ``-`` tells ocrmypdf to read standard input.
+You can also use a program like
+`img2pdf <https://gitlab.mister-muffin.de/josch/img2pdf>`__ to convert
+your images to PDFs, and then pipe the results to run ocrmypdf. The
+``-`` tells ocrmypdf to read standard input.
 
 .. code-block:: bash
 
     img2pdf my-images*.jpg | ocrmypdf - myfile.pdf
 
-``img2pdf`` is recommended because it does an excellent job at generating PDFs without transcoding images.
+``img2pdf`` is recommended because it does an excellent job at
+generating PDFs without transcoding images.
 
 Option: use OCRmyPDF (single images only)
-"""""""""""""""""""""""""""""""""""""""""
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-For convenience, OCRmyPDF can also convert single images to PDFs on its own. If the resolution (dots per inch, DPI) of an image is not set or is incorrect, it can be overridden with ``--image-dpi``. (As 1 inch is 2.54 cm, 1 dpi = 0.39 dpcm).
+For convenience, OCRmyPDF can also convert single images to PDFs on its
+own. If the resolution (dots per inch, DPI) of an image is not set or is
+incorrect, it can be overridden with ``--image-dpi``. (As 1 inch is 2.54
+cm, 1 dpi = 0.39 dpcm).
 
 .. code-block:: bash
 
     ocrmypdf --image-dpi 300 image.png myfile.pdf
 
-If you have multiple images, you must use ``img2pdf`` to convert the images to PDF.
+If you have multiple images, you must use ``img2pdf`` to convert the
+images to PDF.
 
 Not recommended
-"""""""""""""""
+~~~~~~~~~~~~~~~
 
-We caution against using ImageMagick or Ghostscript to convert images to PDF, since they may transcode images or produce downsampled images, sometimes without warning.
+We caution against using ImageMagick or Ghostscript to convert images to
+PDF, since they may transcode images or produce downsampled images,
+sometimes without warning.
 
 Image processing
-----------------
+================
 
-OCRmyPDF perform some image processing on each page of a PDF, if desired.  The same processing is applied to each page.  It is suggested that the user review files after image processing as these commands might remove desirable content, especially from poor quality scans.
+OCRmyPDF perform some image processing on each page of a PDF, if
+desired. The same processing is applied to each page. It is suggested
+that the user review files after image processing as these commands
+might remove desirable content, especially from poor quality scans.
 
-* ``--rotate-pages`` attempts to determine the correct orientation for each page and rotates the page if necessary.
-
-* ``--remove-background`` attempts to detect and remove a noisy background from grayscale or color images.  Monochrome images are ignored. This should not be used on documents that contain color photos as it may remove them.
-
-* ``--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>`_.
-
-* ``--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 less likely that OCR will try to find text in background noise.
-
-* ``--clean-final`` uses unpaper to clean up pages before OCR and inserts the page into the final output.  You will want to review each page to ensure that unpaper did not remove something important.
-
-* ``--mask-barcodes`` will suppress any barcodes detected in a page image. Barcodes are known to confuse Tesseract OCR and interfere with the recognition of text on the same baseline as a barcode. The output file will contain the unaltered image of the barcode.
+-  ``--rotate-pages`` attempts to determine the correct orientation for
+   each page and rotates the page if necessary.
+-  ``--remove-background`` attempts to detect and remove a noisy
+   background from grayscale or color images. Monochrome images are
+   ignored. This should not be used on documents that contain color
+   photos as it may remove them.
+-  ``--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.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
+   less likely that OCR will try to find text in background noise.
+-  ``--clean-final`` uses unpaper to clean up pages before OCR and
+   inserts the page into the final output. You will want to review each
+   page to ensure that unpaper did not remove something important.
 
 .. note::
 
-    In many cases image processing will rasterize PDF pages as images, potentially losing quality.
+   In many cases image processing will rasterize PDF pages as images,
+   potentially losing quality.
 
 .. warning::
 
-    ``--clean-final`` and ``-remove-background`` may leave undesirable visual artifacts in some images where their algorithms have shortcomings. Files should be visually reviewed after using these options.
+   ``--clean-final`` and ``-remove-background`` may leave undesirable
+   visual artifacts in some images where their algorithms have
+   shortcomings. Files should be visually reviewed after using these
+   options.
 
 Example: OCR and correct document skew (crooked scan)
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+-----------------------------------------------------
 
 Deskew:
 
@@ -160,55 +208,118 @@ Deskew:
 
     ocrmypdf --deskew input.pdf output.pdf
 
-Image processing commands can be combined. The order in which options are given does not matter. OCRmyPDF always applies the steps of the image processing pipeline in the same order (rotate, remove background, deskew, clean).
+Image processing commands can be combined. The order in which options
+are given does not matter. OCRmyPDF always applies the steps of the
+image processing pipeline in the same order (rotate, remove background,
+deskew, clean).
 
 .. code-block:: bash
 
     ocrmypdf --deskew --clean --rotate-pages input.pdf output.pdf
 
-
 Don't actually OCR my PDF
--------------------------
+=========================
 
-If you set ``--tesseract-timeout 0`` OCRmyPDF will apply its image processing without performing OCR, if all you want to is to apply image processing or PDF/A conversion.
+If you set ``--tesseract-timeout 0`` OCRmyPDF will apply its image
+processing without performing OCR, if all you want to is to apply image
+processing or PDF/A conversion.
 
 .. code-block:: bash
 
     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
------------------
+=================
 
-To redo OCR on a file OCRed with other OCR software or a previous version of OCRmyPDF and/or Tesseract, you may use the ``--redo-ocr`` argument. (Normally, OCRmyPDF will exit with an error if asked to modify a file with OCR.)
+To redo OCR on a file OCRed with other OCR software or a previous
+version of OCRmyPDF and/or Tesseract, you may use the ``--redo-ocr``
+argument. (Normally, OCRmyPDF will exit with an error if asked to modify
+a file with OCR.)
 
-This may be helpful for users who want to take advantage of accuracy improvements in Tesseract 4.0 for files they previously OCRed with an earlier version of Tesseract and OCRmyPDF.
+This may be helpful for users who want to take advantage of accuracy
+improvements in Tesseract 4.0 for files they previously OCRed with an
+earlier version of Tesseract and OCRmyPDF.
 
 .. code-block:: bash
 
     ocrmypdf --redo-ocr input.pdf output.pdf
 
-This method will replace OCR without rasterizing, reducing quality or removing vector content. If a file contains a mix of pure digital text and OCR, digital text will be ignored and OCR will be replaced. As such this mode is incompatible with image processing options, since they alter the appearance of the file.
+This method will replace OCR without rasterizing, reducing quality or
+removing vector content. If a file contains a mix of pure digital text
+and OCR, digital text will be ignored and OCR will be replaced. As such
+this mode is incompatible with image processing options, since they
+alter the appearance of the file.
 
-In some cases, existing OCR cannot be detected or replaced. Files produced by OCRmyPDF v2.2 or earlier, for example, are internally represented as having visible text with an opaque image drawn on top. This situation cannot be detected.
+In some cases, existing OCR cannot be detected or replaced. Files
+produced by OCRmyPDF v2.2 or earlier, for example, are internally
+represented as having visible text with an opaque image drawn on top.
+This situation cannot be detected.
 
-If ``--redo-ocr`` does not work, you can use ``--force-ocr``, which will force rasterization of all pages, potentially reducing quality or losing vector content.
+If ``--redo-ocr`` does not work, you can use ``--force-ocr``, which will
+force rasterization of all pages, potentially reducing quality or losing
+vector content.
 
 Improving OCR quality
----------------------
+=====================
 
-The `Image processing`_ features can improve OCR quality.
+The `Image processing <#image-processing>`__ features can improve OCR
+quality.
 
-Rotating pages and deskewing helps to ensure that the page orientation is correct before OCR begins. Removing the background and/or cleaning the page can also improve results. The ``--oversample DPI`` argument can be specified to resample images to higher resolution before attempting OCR; this can improve results as well.
+Rotating pages and deskewing helps to ensure that the page orientation
+is correct before OCR begins. Removing the background and/or cleaning
+the page can also improve results. The ``--oversample DPI`` argument can
+be specified to resample images to higher resolution before attempting
+OCR; this can improve results as well.
 
-OCR quality will suffer if the resolution of input images is not correct (since the range of pixel sizes that will be checked for possible fonts will also be incorrect).
+OCR quality will suffer if the resolution of input images is not correct
+(since the range of pixel sizes that will be checked for possible fonts
+will also be incorrect).
 
 PDF optimization
-----------------
+================
 
-By default OCRmyPDF will attempt to perform lossless optimizations on the images inside PDFs after OCR is complete. Optimization is performed even if no OCR text is found.
+By default OCRmyPDF will attempt to perform lossless optimizations on
+the images inside PDFs after OCR is complete. Optimization is performed
+even if no OCR text is found.
 
-The ``--optimize N`` (short form ``-O``) argument controls optimization, where ``N`` ranges from 0 to 3 inclusive, analogous to the optimization levels in the GCC compiler.
+The ``--optimize N`` (short form ``-O``) argument controls optimization,
+where ``N`` ranges from 0 to 3 inclusive, analogous to the optimization
+levels in the GCC compiler.
 
 .. list-table::
     :widths: auto
@@ -227,9 +338,15 @@ The ``--optimize N`` (short form ``-O``) argument controls optimization, where `
     *   - ``--optimize 3``
         - All of the above, and enables more aggressive optimizations and targets lower image quality.
 
-Optimization is improved when a JBIG2 encoder is available and when ``pngquant`` is installed. If either of these components are missing, then some types of images cannot be optimized.
+Optimization is improved when a JBIG2 encoder is available and when
+``pngquant`` is installed. If either of these components are missing,
+then some types of images cannot be optimized.
 
-The types of optimization available may expand over time. By default, OCRmyPDF compresses data streams inside PDFs, and will change inefficient compression modes to more modern versions. A program like ``qpdf`` can be used to change encodings, e.g. to inspect the internals fo a PDF.
+The types of optimization available may expand over time. By default,
+OCRmyPDF compresses data streams inside PDFs, and will change
+inefficient compression modes to more modern versions. A program like
+``qpdf`` can be used to change encodings, e.g. to inspect the internals
+fo a PDF.
 
 .. code-block:: bash
 

docs/docker.rst

@@ -1,134 +1,170 @@
+=====================
 OCRmyPDF Docker image
 =====================
 
-OCRmyPDF is also available in a Docker image that packages recent versions of all dependencies.
+OCRmyPDF is also available in a Docker image that packages recent
+versions of all dependencies.
 
-For users who already have Docker installed this may be an easy and convenient option. However, it is less performant than a system installation and may require Docker engine configuration.
+For users who already have Docker installed this may be an easy and
+convenient option. However, it is less performant than a system
+installation and may require Docker engine configuration.
 
-OCRmyPDF needs a generous amount of RAM, CPU cores, temporary storage space, whether running in a Docker container or on its own. It may be necessary to ensure the container is provisioned with additional resources.
+OCRmyPDF needs a generous amount of RAM, CPU cores, temporary storage
+space, whether running in a Docker container or on its own. It may be
+necessary to ensure the container is provisioned with additional
+resources.
 
 .. _docker-install:
 
 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.
+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``:
+If you can run this command successfully, your system is ready to download and
+execute the image:
 
 .. code-block:: bash
 
-    docker pull jbarlow83/ocrmypdf-alpine
+   docker run hello-world
 
-Follow the Docker installation instructions for your platform.  If you can run this command successfully, your system is ready to download and execute the image:
+The recommended OCRmyPDF Docker image is currently named ``ocrmypdf``:
 
 .. code-block:: bash
 
-    docker run hello-world
+   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 engine host, and then follow these optional steps to enable multiple CPUs:
+
+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
+engine host, and then follow these optional steps to enable multiple
+CPUs:
 
 .. code-block:: bash
 
-    # Optional step for Mac OS X users
-    docker-machine stop "yourVM"
-    VBoxManage modifyvm "yourVM" --cpus 2  # or whatever number of core is desired
-    docker-machine start "yourVM"
-    eval $(docker-machine env "yourVM")
+   # Optional step for Mac OS X users
+   docker-machine stop "yourVM"
+   VBoxManage modifyvm "yourVM" --cpus 2  # or whatever number of core is desired
+   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.
+For convenience, create a shell alias to hide the Docker command. It is
+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/>`_:
+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:
+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
+   # Add French
+   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
-------------------------
+========================
 
-The OCRmyPDF test suite is installed with image.  To run it:
+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
---------------------------------------
+======================================
 
-The OCRmyPDF Docker image includes an example, barebones HTTP web service. The webservice may be launched as follows:
+The OCRmyPDF Docker image includes an example, barebones HTTP web
+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
 
-Unlike command line usage this program will open a socket and wait for connections.
+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.
 
 .. warning::
 
-    The OCRmyPDF web service wrapper is intended for demonstration or development. It provides no security, no authentication, no protection against denial of service attacks, and no load balancing. The default Flask WSGI server is used, which is intended for development only. The server is single-threaded and so can respond to only one client at a time. While running OCR, it cannot respond to any other clients.
+   The OCRmyPDF web service wrapper is intended for demonstration or
+   development. It provides no security, no authentication, no
+   protection against denial of service attacks, and no load balancing.
+   The default Flask WSGI server is used, which is intended for
+   development only. The server is single-threaded and so can respond to
+   only one client at a time. While running OCR, it cannot respond to
+   any other clients.
 
-Clients must keep their open connection while waiting for OCR to complete. This may entail setting a long timeout; this interface is more useful for internal HTTP API calls.
+Clients must keep their open connection while waiting for OCR to
+complete. This may entail setting a long timeout; this interface is more
+useful for internal HTTP API calls.
 
-Unlike the rest of OCRmyPDF, this web service is licensed under the Affero GPLv3 (AGPLv3) since Ghostscript, a dependency of OCRmyPDF, is also licensed in this way.
+Unlike the rest of OCRmyPDF, this web service is licensed under the
+Affero GPLv3 (AGPLv3) since Ghostscript, a dependency of OCRmyPDF, is
+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
+In addition to the above, please read our
+:ref:`general remarks on using OCRmyPDF as a service <ocr-service>`.

docs/errors.rst

@@ -1,33 +1,47 @@
+=====================
 Common error messages
 =====================
 
 Page already has text
----------------------
+=====================
 
-.. code::
+.. code-block::
 
-	ERROR -    1: page already has text! – aborting (use --force-ocr to force OCR)
+   ERROR -    1: page already has text! – aborting (use --force-ocr to force OCR)
 
-You ran ocrmypdf on a file that already contains printable text or a hidden OCR text layer (it can't quite tell the difference). You probably don't want to do this, because the file is already searchable.
+You ran ocrmypdf on a file that already contains printable text or a
+hidden OCR text layer (it can't quite tell the difference). You probably
+don't want to do this, because the file is already searchable.
 
 As the error message suggests, your options are:
 
-- ``ocrmypdf --force-ocr`` to :ref:`rasterize <raster-vector>` all vector content and run OCR on the images. This is useful if a previous OCR program failed, or if the document contains a text watermark.
-
-- ``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 --force-ocr`` to :ref:`rasterize <raster-vector>` all
+   vector content and run OCR on the images. This is useful if a
+   previous OCR program failed, or if the document contains a text
+   watermark.
+-  ``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.
 
 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 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.
 
-You can try rewriting the file with Ghostscript or pdftk:
+You can try rewriting the file with Ghostscript:
 
-- ``gs -o output.pdf -dSAFER -sDEVICE=pdfwrite input.pdf``
+.. code-block:: bash
 
-- ``pdftk input.pdf cat output output.pdf``
+    gs -o output.pdf -dSAFER -sDEVICE=pdfwrite input.pdf
 
-Sometimes Acrobat can repair PDFs with its `Preflight tool <https://helpx.adobe.com/acrobat/using/correcting-problem-areas-preflight-tool.html>`_.
+``pdftk`` can also rewrite PDFs:
 
+.. code-block:: bash
+
+    pdftk input.pdf cat output output.pdf
+
+Sometimes Acrobat can repair PDFs with its `Preflight
+tool <https://helpx.adobe.com/acrobat/using/correcting-problem-areas-preflight-tool.html>`__.

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
 
@@ -27,11 +25,16 @@ PDF is the best format for storing and exchanging scanned documents.  Unfortunat
    cookbook
    docker
    advanced
-   api
    batch
    security
    errors
 
+.. toctree::
+   :caption: Developers
+   :maxdepth: 2
+
+   api
+   contributing
 
 Indices and tables
 ==================

docs/installation.rst

@@ -1,3 +1,4 @@
+===================
 Installing OCRmyPDF
 ===================
 
@@ -7,21 +8,30 @@ 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
     :local:
 
 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
     :alt: Debian 9 stable ("stretch")
@@ -32,118 +42,145 @@ 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-1904| image:: https://repology.org/badge/version-for-repo/ubuntu_19_04/ocrmypdf.svg
+    :alt: Ubuntu 19.04
 
-+-------------------------------------------+
-| **OCRmyPDF versions in Debian & Ubuntu**  |
-+-------------------------------------------+
-| |latest|                                  |
-+-------------------------------------------+
-| |deb-stable| |deb-testing| |deb-unstable| |
-+-------------------------------------------+
-| |ubu-1710| |ubu-1804| |ubu-1810|          |
-+-------------------------------------------+
+.. |ubu-1910| image:: https://repology.org/badge/version-for-repo/ubuntu_19_10/ocrmypdf.svg
+    :alt: Ubuntu 19.10
 
-Users of Debian 9 ("stretch") or later or Ubuntu 16.10 or later may simply
++-----------------------------------------------+
+| **OCRmyPDF versions in Debian & Ubuntu**      |
++-----------------------------------------------+
+| |latest|                                      |
++-----------------------------------------------+
+| |deb-stable| |deb-testing| |deb-unstable|     |
++-----------------------------------------------+
+| |ubu-1804| |ubu-1810| |ubu-1904| |ubu-1910|   |
++-----------------------------------------------+
+
+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
 
     apt-get install ocrmypdf
 
-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`_.
+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>`__. 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 `Ubuntu launchpad.net <https://launchpad.net/ocrmypdf>`_.
+For full details on version availability for your platform, check the
+`Debian Package Tracker <https://tracker.debian.org/pkg/ocrmypdf>`__ or
+`Ubuntu launchpad.net <https://launchpad.net/ocrmypdf>`__.
 
 .. note::
 
-    OCRmyPDF for Debian and Ubuntu currently omit the JBIG2 encoder. OCRmyPDF works fine without it but will produce larger output files. If you build jbig2enc from source, ocrmypdf 7.0.0 and later will automatically detect it (specifically the ``jbig2`` binary) on the ``PATH``. To add JBIG2 encoding, see :ref:`jbig2`.
+   OCRmyPDF for Debian and Ubuntu currently omit the JBIG2 encoder.
+   OCRmyPDF works fine without it but will produce larger output files.
+   If you build jbig2enc from source, ocrmypdf 7.0.0 and later will
+   automatically detect it (specifically the ``jbig2`` binary) on the
+   ``PATH``. To add JBIG2 encoding, see :ref:`jbig2`.
 
 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
 
     dnf install ocrmypdf
 
-For full details on version availability, check the `Fedora Package Tracker
-<https://apps.fedoraproject.org/packages/ocrmypdf>`_.
+For full details on version availability, check the `Fedora Package
+Tracker <https://apps.fedoraproject.org/packages/ocrmypdf>`__.
 
-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`_.
+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>`__.
 
 .. note::
 
-    OCRmyPDF for Fedora currently omits the JBIG2 encoder due to patent issues.
-    OCRmyPDF works fine without it but will produce larger output files. If you
-    build jbig2enc from source, ocrmypdf 7.0.0 and later will automatically
-    detect it on the ``PATH``. To add JBIG2 encoding, see `Installing the JBIG2
-    encoder <jbig2>`_.
+   OCRmyPDF for Fedora currently omits the JBIG2 encoder due to patent
+   issues. OCRmyPDF works fine without it but will produce larger output
+   files. If you build jbig2enc from source, ocrmypdf 7.0.0 and later
+   will automatically detect it on the ``PATH``. To add JBIG2 encoding,
+   see `Installing the JBIG2 encoder <jbig2>`__.
+
+.. _ubuntu-lts-latest:
 
 Installing the latest version on Ubuntu 18.04 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 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, first install several
+system 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 update
+    sudo apt-get -y install \
+        ghostscript \
+        icc-profiles-free \
+        liblept5 \
         libxml2 \
-        pngquant
+        pngquant \
+        python3-cffi \
+        python3-distutils \
+        python3-pkg-resources \
+        python3-reportlab \
+        qpdf \
+        tesseract-ocr \
+        zlib1g
 
-Then install the most recent ocrmypdf for the local user and set the user's ``PATH`` to check for the user's Python packages.
+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.
 
 .. 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`.
 
 Ubuntu 16.04 LTS
-^^^^^^^^^^^^^^^^
+----------------
 
-No package is available for Ubuntu 16.04. OCRmyPDF 8.0 and newer require Python
-3.6. Ubuntu 16.04 ships Python 3.5, but you can install Python 3.6 on it. Or,
-you can skip Python 3.6 and install OCRmyPDF 7.x or older - for that procedure,
-please see the installation documentation for the version of OCRmyPDF you plan
-to use.
+No package is available for Ubuntu 16.04. OCRmyPDF 8.0 and newer require
+Python 3.6. Ubuntu 16.04 ships Python 3.5, but you can install Python
+3.6 on it. Or, you can skip Python 3.6 and install OCRmyPDF 7.x or older
+- for that procedure, please see the installation documentation for the
+version of OCRmyPDF you plan to use.
 
 **Install system packages for OCRmyPDF**
 
@@ -165,13 +202,13 @@ to use.
         tesseract-ocr \
         unpaper
 
-This will install a Python 3.6 binary at ``/usr/bin/python3.6`` alongside the
-system's Python 3.5. Do not remove the system Python. This will also install
-Tesseract 4.0 from a PPA, since the version available in Ubuntu 16.04 is too old
-for OCRmyPDF.
+This will install a Python 3.6 binary at ``/usr/bin/python3.6``
+alongside the system's Python 3.5. Do not remove the system Python. This
+will also install Tesseract 4.0 from a PPA, since the version available
+in Ubuntu 16.04 is too old for OCRmyPDF.
 
-Now install pip for Python 3.6. This will install the Python 3.6 version of
-``pip`` at ``/usr/local/bin/pip``.
+Now install pip for Python 3.6. This will install the Python 3.6 version
+of ``pip`` at ``/usr/local/bin/pip``.
 
 .. code-block:: bash
 
@@ -179,8 +216,9 @@ Now install pip for Python 3.6. This will install the Python 3.6 version of
 
 **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.
+OCRmyPDF requires the locale to be set for UTF-8. **On some minimal
+Ubuntu installations**, such as the Ubuntu 16.04 Docker images it may be
+necessary to set the locale.
 
 .. code-block:: bash
 
@@ -194,16 +232,17 @@ 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`.
 
 Ubuntu 14.04 LTS
-^^^^^^^^^^^^^^^^
+----------------
 
-Installing on Ubuntu 14.04 LTS (trusty) is more difficult than some other
-options, because of its age. Several backports are required. For explanations of
-some steps of this procedure, see the similar steps for Ubuntu 16.04.
+Installing on Ubuntu 14.04 LTS (trusty) is more difficult than some
+other options, because of its age. Several backports are required. For
+explanations of some steps of this procedure, see the similar steps for
+Ubuntu 16.04.
 
 Install system dependencies:
 
@@ -220,12 +259,12 @@ Install system dependencies:
         qpdf
 
 We will need backports of Ghostscript 9.16, libav-11 (for unpaper 6.1),
-Tesseract 4.00 (alpha), and Python 3.6. This will replace Ghostscript and
-Tesseract 3.x on your system. Python 3.6 will be installed alongside the system
-Python 3.4.
+Tesseract 4.00 (alpha), and Python 3.6. This will replace Ghostscript
+and Tesseract 3.x on your system. Python 3.6 will be installed alongside
+the system Python 3.4.
 
-If you prefer to not modify your system in this matter, consider using a Docker
-container.
+If you prefer to not modify your system in this matter, consider using a
+Docker container.
 
 .. code-block:: bash
 
@@ -251,54 +290,137 @@ 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.
+The OCRmyPDF package depends on `the python-pdfminer.six AUR package
+<https://aur.archlinux.org/packages/python-pdfminer.six/>`__.  Dependencies on
+AUR packages are not automatically resolved, so this package must be manually
+installed first.
+
+.. code-block:: bash
+
+   curl -O https://aur.archlinux.org/cgit/aur.git/snapshot/python-pdfminer.six.tar.gz
+   tar xvzf python-pdfminer.six.tar.gz
+   cd python-pdfminer.six
+   makepkg -sri
+
+With that complete you can then repeat the same series of steps for 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 of the pdfminer.six
+    and OCRmyPDF AUR packages. 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
-^^^^^^^^^^^^^^^^^^^^
+--------------------
 
-See the `Repology <https://repology.org/metapackage/ocrmypdf/versions>`_ page.
+See the
+`Repology <https://repology.org/metapackage/ocrmypdf/versions>`__ page.
 
-In general, first install the OCRmyPDF package for your system, then optionally use the procedure `Installing with Python pip`_ to install a more recent version.
+In general, first install the OCRmyPDF package for your system, then
+optionally use the procedure `Installing with Python
+pip <#installing-with-python-pip>`__ to install a more recent version.
 
 Installing on macOS
--------------------
+===================
 
 Homebrew
-^^^^^^^^
+--------
 
 .. image:: https://img.shields.io/homebrew/v/ocrmypdf.svg
     :alt: homebrew
     :target: http://brewformulas.org/Ocrmypdf
 
-OCRmyPDF is now a standard `Homebrew <https://brew.sh>`_ formula. To install on macOS:
+OCRmyPDF is now a standard `Homebrew <https://brew.sh>`__ formula. To
+install on macOS:
 
 .. code-block:: bash
 
     brew install ocrmypdf
 
-This will include only the English language pack. If you need other languages you can optionally install them all:
+This will include only the English language pack. If you need other
+languages you can optionally install them all:
 
 .. code-block:: bash
 
@@ -306,18 +428,23 @@ This will include only the English language pack. If you need other languages yo
 
 .. note::
 
-    Users who previously installed OCRmyPDF on macOS using ``pip install ocrmypdf`` should remove the pip version (``pip3 uninstall ocrmypdf``) before switching to the Homebrew version.
+   Users who previously installed OCRmyPDF on macOS using
+   ``pip install ocrmypdf`` should remove the pip version
+   (``pip3 uninstall ocrmypdf``) before switching to the Homebrew
+   version.
 
 .. note::
 
-    Users who previously installed OCRmyPDF from the private tap should switch to the mainline version (``brew untap jbarlow83/ocrmypdf``) and install from there.
+   Users who previously installed OCRmyPDF from the private tap should
+   switch to the mainline version (``brew untap jbarlow83/ocrmypdf``)
+   and install from there.
 
 Manual installation on macOS
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+----------------------------
 
 These instructions probably work on all macOS supported by Homebrew.
 
-If it's not already present, `install Homebrew <http://brew.sh/>`_.
+If it's not already present, `install Homebrew <http://brew.sh/>`__.
 
 Update Homebrew:
 
@@ -325,20 +452,24 @@ 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.
+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
 
-This will include the English, French, German and Spanish language packs. If you need other languages you can optionally install them all:
+This will include the English, French, German and Spanish language
+packs. If you need other languages you can optionally install them all:
 
 .. _macos-all-languages:
 
-.. code-block:: bash
+   .. 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:
 
@@ -365,103 +496,209 @@ The command line program should now be available:
     ocrmypdf --help
 
 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.
+For some users, installing the Docker image will be easier than
+installing all of OCRmyPDF's dependencies.
 
-See `OCRmyPDF Docker Image <docker>`_ for more information.
+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 your command prompt can run the docker "hello world" container.
+.. warning::
 
-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.
+    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.
 
 Installing with Python pip
---------------------------
+==========================
 
-OCRmyPDF is delivered by PyPI because it is a convenient way to install the latest version. However, PyPI and ``pip`` cannot address the fact that ``ocrmypdf`` depends on certain non-Python system libraries and programs being instsalled.
+OCRmyPDF is delivered by PyPI because it is a convenient way to install
+the latest version. However, PyPI and ``pip`` cannot address the fact
+that ``ocrmypdf`` depends on certain non-Python system libraries and
+programs being instsalled.
 
-For best results, first install `your platform's version <https://repology.org/metapackage/ocrmypdf/versions>`_ of ``ocrmypdf``, using the instructions elsewhere in this document. Then you can use ``pip`` to get the latest version if your platform version is out of date. Chances are that this will satisfy most dependencies.
+For best results, first install `your platform's
+version <https://repology.org/metapackage/ocrmypdf/versions>`__ of
+``ocrmypdf``, using the instructions elsewhere in this document. Then
+you can use ``pip`` to get the latest version if your platform version
+is out of date. Chances are that this will satisfy most dependencies.
 
 Use ``ocrmypdf --version`` to confirm what version was installed.
 
-Then you can install the latest OCRmyPDF from the Python wheels. First try:
+Then you can install the latest OCRmyPDF from the Python wheels. First
+try:
 
 .. code-block:: bash
 
     pip3 install --user ocrmypdf
 
-You should then be able to run ``ocrmypdf --version`` and see that the latest version was located.
+You should then be able to run ``ocrmypdf --version`` and see that the
+latest version was located.
 
-Since ``pip3 install --user`` does not work correctly on some platforms, notably Ubuntu 16.04 and older, and the Homebrew version of Python, instead use this for a system wide installation:
+Since ``pip3 install --user`` does not work correctly on some platforms,
+notably Ubuntu 16.04 and older, and the Homebrew version of Python,
+instead use this for a system wide installation:
 
 .. code-block:: bash
 
     pip3 install ocrmypdf
 
 Requirements for pip and HEAD install
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+-------------------------------------
 
-OCRmyPDF currently requires these external programs and libraries to be installed, and must be satisfied using the operating system package manager. ``pip`` cannot provide them.
+OCRmyPDF currently requires these external programs and libraries to be
+installed, and must be satisfied using the operating system package
+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
+-  Python 3.6 or newer
+-  Ghostscript 9.15 or newer
+-  qpdf 8.1.0 or newer
+-  Tesseract 4.0.0-beta or newer
 
 As of ocrmypdf 7.2.1, the following versions are recommended:
 
-- Python 3.7
-- Ghostscript 9.23 or newer
-- qpdf 8.2.1
-- Tesseract 4.0.0 or newer
-- jbig2enc 0.29 or newer
-- pngquant 2.5 or newer
-- unpaper 6.1
+-  Python 3.7 or 3.8
+-  Ghostscript 9.23 or newer
+-  qpdf 8.2.1
+-  Tesseract 4.0.0 or newer
+-  jbig2enc 0.29 or newer
+-  pngquant 2.5 or newer
+-  unpaper 6.1
 
-jbig2enc, pngquant, and unpaper are optional. If missing certain features are disabled. OCRmyPDF will discover them as soon as they are available.
+jbig2enc, pngquant, and unpaper are optional. If missing certain
+features are disabled. OCRmyPDF will discover them as soon as they are
+available.
 
-**jbig2enc**, if present, will be used to optimize the encoding of monochrome images.  This can significantly reduce the file size of the output file.  It is not required.  `jbig2enc <https://github.com/agl/jbig2enc>`_ is not generally available for Ubuntu or Debian due to lingering concerns about patent issues, but can easily be built from source. To add JBIG2 encoding, see :ref:`jbig2`.
+**jbig2enc**, if present, will be used to optimize the encoding of
+monochrome images. This can significantly reduce the file size of the
+output file. It is not required.
+`jbig2enc <https://github.com/agl/jbig2enc>`__ is not generally
+available for Ubuntu or Debian due to lingering concerns about patent
+issues, but can easily be built from source. To add JBIG2 encoding, see
+:ref:`jbig2`.
 
-**pngquant**, if present, is optionally used to optimize the encoding of PNG-style images in PDFs (actually, any that are that losslessly encoded) by lossily quantizing to a smaller color palette. It is only activated then the ``--optimize`` argument is ``2`` or ``3``.
+**pngquant**, if present, is optionally used to optimize the encoding of
+PNG-style images in PDFs (actually, any that are that losslessly
+encoded) by lossily quantizing to a smaller color palette. It is only
+activated then the ``--optimize`` argument is ``2`` or ``3``.
 
-**unpaper**, if present, enables the ``--clean`` and ``--clean-final`` command line options.
-
-These are in addition to the Python packaging dependencies, meaning that unfortunately, the ``pip install`` command cannot satisfy all of them.
+**unpaper**, if present, enables the ``--clean`` and ``--clean-final``
+command line options.
 
+These are in addition to the Python packaging dependencies, meaning that
+unfortunately, the ``pip install`` command cannot satisfy all of them.
 
 Installing HEAD revision from sources
--------------------------------------
+=====================================
 
-If you have ``git`` and Python 3.6 or newer installed, you can install from source. When the ``pip`` installer runs, it will alert you if dependencies are missing.
+If you have ``git`` and Python 3.6 or newer installed, you can install
+from source. When the ``pip`` installer runs, it will alert you if
+dependencies are missing.
 
-If you prefer to build every from source, you will need to `build pikepdf from source <https://pikepdf.readthedocs.io/en/latest/installation.html#building-from-source>`_. First ensure you can build and install pikepdf.
+If you prefer to build every from source, you will need to `build
+pikepdf from
+source <https://pikepdf.readthedocs.io/en/latest/installation.html#building-from-source>`__.
+First ensure you can build and install pikepdf.
 
-To install the HEAD revision from sources in the current Python 3 environment:
+To install the HEAD revision from sources in the current Python 3
+environment:
 
 .. code-block:: bash
 
     pip3 install git+https://github.com/jbarlow83/OCRmyPDF.git
 
-Or, to install in `development mode <https://pythonhosted.org/setuptools/setuptools.html#development-mode>`_,  allowing customization of OCRmyPDF, use the ``-e`` flag:
+Or, to install in `development
+mode <https://pythonhosted.org/setuptools/setuptools.html#development-mode>`__,
+allowing customization of OCRmyPDF, use the ``-e`` flag:
 
 .. code-block:: bash
 
     pip3 install -e git+https://github.com/jbarlow83/OCRmyPDF.git
 
-You may find it easiest to install in a virtual environment, rather than system-wide:
+You may find it easiest to install in a virtual environment, rather than
+system-wide:
 
 .. code-block:: bash
 
@@ -471,8 +708,8 @@ You may find it easiest to install in a virtual environment, rather than system-
     cd OCRmyPDF
     pip3 install .
 
-However, ``ocrmypdf`` will only be accessible on the system PATH
-when you activate the virtual environment.
+However, ``ocrmypdf`` will only be accessible on the system PATH when
+you activate the virtual environment.
 
 To run the program:
 
@@ -486,7 +723,7 @@ dependencies. Older version than the ones mentioned in the release notes
 are likely not to be compatible to OCRmyPDF.
 
 For development
-^^^^^^^^^^^^^^^
+---------------
 
 To install all of the development and test requirements:
 
@@ -502,15 +739,17 @@ To install all of the development and test requirements:
 To add JBIG2 encoding, see :ref:`jbig2`.
 
 Shell completions
------------------
+=================
 
 Completions for ``bash`` and ``fish`` are available in the project's
 ``misc/completion`` folder. The ``bash`` completions are likely ``zsh``
-compatible but this has not been confirmed. Package maintainers, please install
-these at the appropriate locations for your system.
+compatible but this has not been confirmed. Package maintainers, please
+install these at the appropriate locations for your system.
 
-To manually install the ``bash`` completion, copy ``misc/completion/ocrmypdf.bash`` to
-``/etc/bash_completion.d/ocrmypdf`` (rename the file).
+To manually install the ``bash`` completion, copy
+``misc/completion/ocrmypdf.bash`` to ``/etc/bash_completion.d/ocrmypdf``
+(rename the file).
 
-To manually install the ``fish`` completion, copy ``misc/completion/ocrmypdf.fish`` to
+To manually install the ``fish`` completion, copy
+``misc/completion/ocrmypdf.fish`` to
 ``~/.config/fish/completions/ocrmypdf.fish``.

docs/introduction.rst

@@ -1,118 +1,228 @@
+============
 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
----------
+=========
 
-`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.
+`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 selected, searched and copied.
 
-OCRmyPDF uses `Tesseract <https://github.com/tesseract-ocr/tesseract>`_, the best available open source OCR engine, to perform OCR.
+OCRmyPDF uses
+`Tesseract <https://github.com/tesseract-ocr/tesseract>`__, the best
+available open source OCR engine, to perform OCR.
 
 .. _raster-vector:
 
 About PDFs
-----------
+==========
 
-PDFs are page description files that attempts to preserve a layout exactly. They  contain `vector graphics <http://vector-conversions.com/vectorizing/raster_vs_vector.html>`_ that can contain raster objects such as scanned images. Because PDFs can contain multiple pages (unlike many image formats) and can contain fonts and text, it is a good formats for exchanging scanned documents.
+PDFs are page description files that attempts to preserve a layout
+exactly. They contain `vector
+graphics <http://vector-conversions.com/vectorizing/raster_vs_vector.html>`__
+that can contain raster objects such as scanned images. Because PDFs can
+contain multiple pages (unlike many image formats) and can contain fonts
+and text, it is a good formats for exchanging scanned documents.
 
-.. image:: images/bitmap_vs_svg.svg
+|image|
 
-A PDF page might contain multiple images, even if it only appears to have one image.  Some scanners or scanning software will segment pages into monochromatic text and color regions for example, to improve the compression ratio and appearance of the page.
-
-Rasterizing a PDF is the process of generating an image suitable for display or analyzing with an OCR engine.  OCR engines like Tesseract work with images, not vector objects.
+A PDF page might contain multiple images, even if it only appears to
+have one image. Some scanners or scanning software will segment pages
+into monochromatic text and color regions for example, to improve the
+compression ratio and appearance of the page.
 
+Rasterizing a PDF is the process of generating an image suitable for
+display or analyzing with an OCR engine. OCR engines like Tesseract work
+with images, not vector objects.
 
 About PDF/A
------------
+===========
 
-`PDF/A <https://en.wikipedia.org/wiki/PDF/A>`_ is an ISO-standardized subset of the full PDF specification that is designed for archiving (the 'A' stands for Archive).  PDF/A differs from PDF primarily by omitting features that would make it difficult to read the file in the future, such as embedded Javascript, video, audio and references to external fonts.  All fonts and resources needed to interpret the PDF must be contained within it. Because PDF/A disables Javascript and other types of embedded content, it is probably more secure.
+`PDF/A <https://en.wikipedia.org/wiki/PDF/A>`__ is an ISO-standardized
+subset of the full PDF specification that is designed for archiving (the
+'A' stands for Archive). PDF/A differs from PDF primarily by omitting
+features that would make it difficult to read the file in the future,
+such as embedded Javascript, video, audio and references to external
+fonts. All fonts and resources needed to interpret the PDF must be
+contained within it. Because PDF/A disables Javascript and other types
+of embedded content, it is probably more secure.
 
 There are various conformance levels and versions, such as "PDF/A-2b".
 
-Generally speaking, the best format for scanned documents is PDF/A. Some governments and jurisdictions, US Courts in particular, `mandate the use of PDF/A <https://pdfblog.com/2012/02/13/what-is-pdfa/>`_ for scanned documents.
+Generally speaking, the best format for scanned documents is PDF/A. Some
+governments and jurisdictions, US Courts in particular, `mandate the use
+of PDF/A <https://pdfblog.com/2012/02/13/what-is-pdfa/>`__ for scanned
+documents.
 
-Since most people who scan documents are interested in reading them indefinitely into the future, OCRmyPDF generates PDF/A-2b by default.
-
-PDF/A has a few drawbacks.  Some PDF viewers include an alert that the file is a PDF/A, which may confuse some users.  It also tends to produce larger files than PDF, because it embeds certain resources even if they are commonly available. PDF/A files can be digitally signed, but may not be encrypted, to ensure they can be read in the future.  Fortunately, converting from PDF/A to a regular PDF is trivial, and any PDF viewer can view PDF/A.
+Since most people who scan documents are interested in reading them
+indefinitely into the future, OCRmyPDF generates PDF/A-2b by default.
 
+PDF/A has a few drawbacks. Some PDF viewers include an alert that the
+file is a PDF/A, which may confuse some users. It also tends to produce
+larger files than PDF, because it embeds certain resources even if they
+are commonly available. PDF/A files can be digitally signed, but may not
+be encrypted, to ensure they can be read in the future. Fortunately,
+converting from PDF/A to a regular PDF is trivial, and any PDF viewer
+can view PDF/A.
 
 What OCRmyPDF does
-------------------
+==================
 
-OCRmyPDF analyzes each page of a PDF to determine the colorspace and resolution (DPI) needed to capture all of the information on that page without losing content.  It uses `Ghostscript <http://ghostscript.com/>`_ to rasterize the page, and then performs on OCR on the rasterized image to create an OCR "layer". The layer is then grafted back onto the original PDF.
+OCRmyPDF analyzes each page of a PDF to determine the colorspace and
+resolution (DPI) needed to capture all of the information on that page
+without losing content. It uses
+`Ghostscript <http://ghostscript.com/>`__ to rasterize the page, and
+then performs on OCR on the rasterized image to create an OCR "layer".
+The layer is then grafted back onto the original PDF.
 
-While one can use a program like Ghostscript or ImageMagick to get an image and put the image through Tesseract, that actually creates a new PDF and many details may be lost. OCRmyPDF can produce a minimally changed PDF as output.
+While one can use a program like Ghostscript or ImageMagick to get an
+image and put the image through Tesseract, that actually creates a new
+PDF and many details may be lost. OCRmyPDF can produce a minimally
+changed PDF as output.
 
-OCRmyPDF also some image processing options like deskew which improve the appearance of files and quality of OCR. When these are used, the OCR layer is grafted onto the processed image instead.
-
-By default, OCRmyPDF produces archival PDFs – PDF/A, which are a stricter subset of PDF features designed for long term archives. If regular PDFs are desired, this can be disabled with ``--output-type pdf``.
+OCRmyPDF also some image processing options like deskew which improve
+the appearance of files and quality of OCR. When these are used, the OCR
+layer is grafted onto the processed image instead.
 
+By default, OCRmyPDF produces archival PDFs – PDF/A, which are a
+stricter subset of PDF features designed for long term archives. If
+regular PDFs are desired, this can be disabled with
+``--output-type pdf``.
 
 Why you shouldn't do this manually
-----------------------------------
+==================================
 
-A PDF is similar to an HTML file, in that it contains document structure along with images.  Sometimes a PDF does nothing more than present a full page image, but often there is additional content that would be lost.
+A PDF is similar to an HTML file, in that it contains document structure
+along with images. Sometimes a PDF does nothing more than present a full
+page image, but often there is additional content that would be lost.
 
 A manual process could work like either of these:
 
-1. Rasterize each page as an image, OCR the images, and combine the output into a PDF. This preserves the layout of each page, but resamples all images (possibly losing quality, increasing file size, introducing compression artifacts, etc.).
+1. Rasterize each page as an image, OCR the images, and combine the
+   output into a PDF. This preserves the layout of each page, but
+   resamples all images (possibly losing quality, increasing file size,
+   introducing compression artifacts, etc.).
+2. Extract each image, OCR, and combine the output into a PDF. This
+   loses the context in which images are used in the PDF, meaning that
+   cropping, rotation and scaling of pages may be lost. Some scanned
+   PDFs use multiple images segmented into black and white, grayscale
+   and color regions, with stencil masks to prevent overlap, as this can
+   enhance the appearance of a file while reducing file size. Clearly,
+   reassembling these images will be easy. This also loses and text or
+   vector art on any pages in a PDF with both scanned and pure digital
+   content.
 
-2. Extract each image, OCR, and combine the output into a PDF. This loses the context in which images are used in the PDF, meaning that cropping, rotation and scaling of pages may be lost. Some scanned PDFs use multiple images segmented into black and white, grayscale and color regions, with stencil masks to prevent overlap, as this can enhance the appearance of a file while reducing file size. Clearly, reassembling these images will be easy. This also loses and text or vector art on any pages in a PDF with both scanned and pure digital content.
+In the case of a PDF that is nothing other than a container of images
+(no rotation, scaling, cropping, one image per page), the second
+approach can be lossless.
 
-In the case of a PDF that is nothing other than a container of images (no rotation, scaling, cropping, one image per page), the second approach can be lossless.
-
-OCRmyPDF uses several strategies depending on input options and the input PDF itself, but generally speaking it rasterizes a page for OCR and then grafts the OCR back onto the original. As such it can handle complex PDFs and still preserve their contents as much as possible.
-
-OCRmyPDF also supports a many, many edge cases that have cropped over several years of development. We support PDF features like images inside of Form XObjects, and pages with UserUnit scaling. We support rare image formats like non-monochrome 1-bit images. We warn about files you may not to OCR. Thanks to pikepdf and QPDF, we auto-repair PDFs that are damaged. (Not that you need to know what any of these are! You should be able to throw any PDF at it.)
+OCRmyPDF uses several strategies depending on input options and the
+input PDF itself, but generally speaking it rasterizes a page for OCR
+and then grafts the OCR back onto the original. As such it can handle
+complex PDFs and still preserve their contents as much as possible.
 
+OCRmyPDF also supports a many, many edge cases that have cropped over
+several years of development. We support PDF features like images inside
+of Form XObjects, and pages with UserUnit scaling. We support rare image
+formats like non-monochrome 1-bit images. We warn about files you may
+not to OCR. Thanks to pikepdf and QPDF, we auto-repair PDFs that are
+damaged. (Not that you need to know what any of these are! You should be
+able to throw any PDF at it.)
 
 Limitations
------------
+===========
 
-OCRmyPDF is limited by the Tesseract OCR engine.  As such it experiences these limitations, as do any other programs that rely on Tesseract:
+OCRmyPDF is limited by the Tesseract OCR engine. As such it experiences
+these limitations, as do any other programs that rely on Tesseract:
 
-* The OCR is not as accurate as commercial solutions such as Abbyy.
-* It is not capable of recognizing handwriting.
-* It may find gibberish and report this as OCR output.
-* If a document contains languages outside of those given in the ``-l LANG`` arguments, results may be poor.
-* It is not always good at analyzing the natural reading order of documents. For example, it may fail to recognize that a document contains two columns, and may try to join text across columns.
-* Poor quality scans may produce poor quality OCR. Garbage in, garbage out.
-* It does not expose information about what font family text belongs to.
+-  The OCR is not as accurate as commercial solutions such as Abbyy.
+-  It is not capable of recognizing handwriting.
+-  It may find gibberish and report this as OCR output.
+-  If a document contains languages outside of those given in the
+   ``-l LANG`` arguments, results may be poor.
+-  It is not always good at analyzing the natural reading order of
+   documents. For example, it may fail to recognize that a document
+   contains two columns, and may try to join text across columns.
+-  Poor quality scans may produce poor quality OCR. Garbage in, garbage
+   out.
+-  It does not expose information about what font family text belongs
+   to.
 
 OCRmyPDF is also limited by the PDF specification:
 
-* PDF encodes the position of text glyphs but does not encode document structure.  There is no markup that divides a document in sections, paragraphs, sentences, or even words (since blank spaces are not represented). As such all elements of document structure including 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 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 compatibility with some viewers and does not cause problems for better ones.
+-  PDF encodes the position of text glyphs but does not encode document
+   structure. There is no markup that divides a document in sections,
+   paragraphs, sentences, or even words (since blank spaces are not
+   represented). As such all elements of document structure including
+   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 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
+   compatibility with some viewers and does not cause problems for
+   better ones.
 
 Ghostscript also imposes some limitations:
 
-* PDFs containing JBIG2-encoded content will be converted to CCITT Group4 encoding, which has lower compression ratios, if Ghostscript PDF/A is enabled.
-* PDFs containing JPEG 2000-encoded content will be converted to JPEG encoding, which may introduce compression artifacts, if Ghostscript PDF/A is enabled.
-* Ghostscript may transcode grayscale and color images, either lossy to lossless or lossless to lossy, based on an internal algorithm. This behavior can be suppressed by setting ``--pdfa-image-compression`` to ``jpeg`` or ``lossless`` to set all images to one type or the other. Ghostscript has no option to maintain the input image's format. (Ghostscript 9.25+ can copy JPEG images without transcoding them; earlier versions will transcode.)
-* Ghostscript's PDF/A conversion removes any XMP metadata that is not one of the standard XMP metadata namespaces for PDFs. In particular, PRISM Metdata is removed.
+-  PDFs containing JBIG2-encoded content will be converted to CCITT
+   Group4 encoding, which has lower compression ratios, if Ghostscript
+   PDF/A is enabled.
+-  PDFs containing JPEG 2000-encoded content will be converted to JPEG
+   encoding, which may introduce compression artifacts, if Ghostscript
+   PDF/A is enabled.
+-  Ghostscript may transcode grayscale and color images, either lossy to
+   lossless or lossless to lossy, based on an internal algorithm. This
+   behavior can be suppressed by setting ``--pdfa-image-compression`` to
+   ``jpeg`` or ``lossless`` to set all images to one type or the other.
+   Ghostscript has no option to maintain the input image's format.
+   (Ghostscript 9.25+ can copy JPEG images without transcoding them;
+   earlier versions will transcode.)
+-  Ghostscript's PDF/A conversion removes any XMP metadata that is not
+   one of the standard XMP metadata namespaces for PDFs. In particular,
+   PRISM Metdata is removed.
 
 Regarding OCRmyPDF itself:
 
-* PDFs that use transparency are not currently represented in the test suite
+-  PDFs that use transparency are not currently represented in the test
+   suite
 
 Similar programs
-----------------
+================
 
-To the author's knowledge, OCRmyPDF is the most feature-rich and thoroughly tested command line OCR PDF conversion tool. If it does not meet your needs, contributions and suggestions are welcome. If not, consider one of these similar open source programs:
+To the author's knowledge, OCRmyPDF is the most feature-rich and
+thoroughly tested command line OCR PDF conversion tool. If it does not
+meet your needs, contributions and suggestions are welcome. If not,
+consider one of these similar open source programs:
 
-* pdf2pdfocr
-* pdfsandwich
-* pypdfocr
-* pdfbeads
+-  pdf2pdfocr
+-  pdfsandwich
+-  pypdfocr
+-  pdfbeads
 
 Web front-ends
---------------
+==============
 
-The Docker image ``ocrmypdf-alpine`` 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 security measures to speak of.
+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
+security measures to speak of.
 
 In addition, the following third-party integrations are available:
 
-* `Nextcloud OCR <https://github.com/janis91/ocr>`_ is a free software plugin for the Nextcloud private cloud software
+-  `Nextcloud OCR <https://github.com/janis91/ocr>`__ is a free software
+   plugin for the Nextcloud private cloud software
 
-OCRmyPDF is not designed to be secure against malware-bearing PDFs (see `Using OCRmyPDF online <ocr-service>`_). Users should ensure they comply with OCRmyPDF's licenses and the licenses of all dependencies. In particular, OCRmyPDF requires Ghostscript, which is licensed under AGPLv3.
+OCRmyPDF is not designed to be secure against malware-bearing PDFs (see
+`Using OCRmyPDF online <ocr-service>`__). Users should ensure they
+comply with OCRmyPDF's licenses and the licenses of all dependencies. In
+particular, OCRmyPDF requires Ghostscript, which is licensed under
+AGPLv3.
+
+.. |image| image:: images/bitmap_vs_svg.svg

docs/jbig2.rst

@@ -1,35 +1,55 @@
 .. _jbig2:
 
+============================
 Installing the JBIG2 encoder
 ============================
 
-Most Linux distributions do not include a JBIG2 encoder since JBIG2 encoding was patented for a long time. All known JBIG2 US patents have expired as of 2017, but it is possible that unknown patents exist.
+Most Linux distributions do not include a JBIG2 encoder since JBIG2
+encoding was patented for a long time. All known JBIG2 US patents have
+expired as of 2017, but it is possible that unknown patents exist.
 
-JBIG2 encoding is recommended for OCRmyPDF and is used to losslessly create smaller PDFs. If JBIG2 encoding not available, lower quality encodings will be used.
+JBIG2 encoding is recommended for OCRmyPDF and is used to losslessly
+create smaller PDFs. If JBIG2 encoding not available, lower quality
+encodings will be used.
 
-JBIG2 decoding is not patented and is performed automatically by most PDF viewers. It is widely supported has been part of the PDF specification since 2001.
+JBIG2 decoding is not patented and is performed automatically by most
+PDF viewers. It is widely supported has been part of the PDF
+specification since 2001.
 
-On macOS, Homebrew packages jbig2enc and OCRmyPDF includes it by default. The Docker image for OCRmyPDF also builds its own JBIG2 encoder from source.
+On macOS, Homebrew packages jbig2enc and OCRmyPDF includes it by
+default. The Docker image for OCRmyPDF also builds its own JBIG2 encoder
+from source.
 
 For all other Linux, you must build a JBIG2 encoder from source:
 
 .. code-block:: bash
 
-    git clone https://github.com/agl/jbig2enc
-    cd jbig2enc
-    ./autogen.sh
-    ./configure && make
-    [sudo] make install
+   git clone https://github.com/agl/jbig2enc
+   cd jbig2enc
+   ./autogen.sh
+   ./configure && make
+   [sudo] make install
 
 .. _jbig2-lossy:
 
 Lossy mode JBIG2
-----------------
+================
 
-OCRmyPDF provides lossy mode JBIG2 as an advanced feature. Users should `review the technical concerns with JBIG2 in lossy mode <https://abbyy.technology/en:kb:tip:jbig2_compression_and_ocr>`_ and decide if this feature is acceptable for their use case.
+OCRmyPDF provides lossy mode JBIG2 as an advanced feature. Users should
+`review the technical concerns with JBIG2 in lossy
+mode <https://abbyy.technology/en:kb:tip:jbig2_compression_and_ocr>`__
+and decide if this feature is acceptable for their use case.
 
-JBIG2 lossy mode does achieve higher compression ratios than any other monochrome (bitonal) compression technology; for large text documents the savings are considerable. JBIG2 lossless still gives great compression ratios and is a major improvement over the older CCITT G4 standard. As explained above, there is some risk of substitution errors.
+JBIG2 lossy mode does achieve higher compression ratios than any other
+monochrome (bitonal) compression technology; for large text documents
+the savings are considerable. JBIG2 lossless still gives great
+compression ratios and is a major improvement over the older CCITT G4
+standard. As explained above, there is some risk of substitution errors.
 
-To turn on JBIG2 lossy mode, add the argument ``--jbig2-lossy``. ``--optimize {1,2,3}`` are necessary for the argument to take effect also required. Also, a JBIG2 encoder must be installed as described in the previous section.
+To turn on JBIG2 lossy mode, add the argument ``--jbig2-lossy``.
+``--optimize {1,2,3}`` are necessary for the argument to take effect
+also required. Also, a JBIG2 encoder must be installed as described in
+the previous section.
 
-*Due to an oversight, ocrmypdf v7.0 and v7.1 used lossy mode by default.*
+*Due to an oversight, ocrmypdf v7.0 and v7.1 used lossy mode by
+default.*

docs/languages.rst

@@ -1,16 +1,29 @@
 .. _lang-packs:
 
+====================================
 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>`_.
+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``.
 
-For Linux users, you can often find packages that provide language packs:
+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:
 
 Debian and Ubuntu users
------------------------
+=======================
 
 .. code-block:: bash
 
@@ -20,11 +33,13 @@ Debian and Ubuntu users
    # Install Chinese Simplified language pack
    apt-get install tesseract-ocr-chi-sim
 
-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 using either ``-l eng+fre`` (English and French) or ``-l eng -l fre``.
+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 using either ``-l eng+fre`` (English and French) or
+``-l eng -l fre``.
 
 Fedora users
-------------
+============
 
 .. code-block:: bash
 
@@ -34,16 +49,25 @@ Fedora users
    # Install Chinese Simplified language pack
    dnf install tesseract-langpack-chi_sim
 
-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 using
-either ``-l eng+fre`` (English and French) or ``-l eng -l fre``.
+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 using either ``-l eng+fre`` (English and French) or
+``-l eng -l fre``.
 
 macOS users
------------
+===========
 
-You can install additional language packs by :ref:`installing Tesseract using Homebrew with all language packs <macos-all-languages>`.
+You can install additional language packs by
+:ref:`installing Tesseract using Homebrew with all language packs <macos-all-languages>`.
 
 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>`.
+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/plugins.rst

@@ -0,0 +1,24 @@
+=======
+Plugins
+=======
+
+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
+
+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 currently cannot override the same hook.
+
+How plugins are invoked
+=======================
+
+Plugins may be called from the command line:

docs/security.rst

@@ -1,79 +1,162 @@
+===================
 PDF security issues
 ===================
 
-	OCRmyPDF should only be used on PDFs you trust. It is not designed to protect you against malware.
+   OCRmyPDF should only be used on PDFs you trust. It is not designed to
+   protect you against malware.
 
-Recognizing that many users have an interest in handling PDFs and applying OCR to PDFs they did not generate themselves, this article discusses the security implications of PDFs and how users can protect themselves.
+Recognizing that many users have an interest in handling PDFs and
+applying OCR to PDFs they did not generate themselves, this article
+discusses the security implications of PDFs and how users can protect
+themselves.
 
 The disclaimer applies: this software has no warranties of any kind.
 
 PDFs may contain malware
-------------------------
+========================
 
-PDF is a rich, complex file format. The official PDF 1.7 specification, ISO 32000:2008, is hundreds of pages long and references several annexes each of which are similar in length. PDFs can contain video, audio, XML, JavaScript and other programming, and forms. In some cases, they can open internet connections to pre-selected URLs. All of these possible attack vectors.
+PDF is a rich, complex file format. The official PDF 1.7 specification,
+ISO 32000:2008, is hundreds of pages long and references several annexes
+each of which are similar in length. PDFs can contain video, audio, XML,
+JavaScript and other programming, and forms. In some cases, they can
+open internet connections to pre-selected URLs. All of these possible
+attack vectors.
 
-In short, PDFs `may contain viruses <https://security.stackexchange.com/questions/64052/can-a-pdf-file-contain-a-virus>`_.
+In short, PDFs `may contain
+viruses <https://security.stackexchange.com/questions/64052/can-a-pdf-file-contain-a-virus>`__.
 
-This `article <https://theinvisiblethings.blogspot.ca/2013/02/converting-untrusted-pdfs-into-trusted.html>`_ describes a high-paranoia method which allows potentially hostile PDFs to be viewed and rasterized safely in a disposable virtual machine. A trusted PDF created in this manner is converted to images and loses all information making it searchable and losing all compression. OCRmyPDF could be used restore searchability.
+This
+`article <https://theinvisiblethings.blogspot.ca/2013/02/converting-untrusted-pdfs-into-trusted.html>`__
+describes a high-paranoia method which allows potentially hostile PDFs
+to be viewed and rasterized safely in a disposable virtual machine. A
+trusted PDF created in this manner is converted to images and loses all
+information making it searchable and losing all compression. OCRmyPDF
+could be used restore searchability.
 
 How OCRmyPDF processes PDFs
----------------------------
+===========================
 
-OCRmyPDF must open and interpret your PDF in order to insert an OCR layer. First, it runs all PDFs through `pikepdf <https://github.com/pikepdf/pikepdf>`_, a library based on `qpdf <https://github.com/qpdf/qpdf>`_, a program that repairs PDFs with syntax errors. This is done because, in the author's experience, a significant number of PDFs in the wild especially those created by scanners are not well-formed files. qpdf makes it more likely that OCRmyPDF will succeed, but offers no security guarantees. qpdf is also used to split the PDF into single page PDFs.
+OCRmyPDF must open and interpret your PDF in order to insert an OCR
+layer. First, it runs all PDFs through
+`pikepdf <https://github.com/pikepdf/pikepdf>`__, a library based on
+`qpdf <https://github.com/qpdf/qpdf>`__, a program that repairs PDFs
+with syntax errors. This is done because, in the author's experience, a
+significant number of PDFs in the wild especially those created by
+scanners are not well-formed files. qpdf makes it more likely that
+OCRmyPDF will succeed, but offers no security guarantees. qpdf is also
+used to split the PDF into single page PDFs.
 
-Finally, OCRmyPDF rasterizes each page of the PDF using `Ghostscript <http://ghostscript.com/>`_ in ``-dSAFER`` mode.
+Finally, OCRmyPDF rasterizes each page of the PDF using
+`Ghostscript <http://ghostscript.com/>`__ in ``-dSAFER`` mode.
 
-Depending on the options specified, OCRmyPDF may graft the OCR layer into the existing PDF or it may essentially reconstruct ("re-fry") a visually identical PDF that may be quite different at the binary level. That said, OCRmyPDF is not a tool designed for sanitizing PDFs.
+Depending on the options specified, OCRmyPDF may graft the OCR layer
+into the existing PDF or it may essentially reconstruct ("re-fry") a
+visually identical PDF that may be quite different at the binary level.
+That said, OCRmyPDF is not a tool designed for sanitizing PDFs.
 
 .. _ocr-service:
 
 Using OCRmyPDF online or as a service
--------------------------------------
+=====================================
 
-OCRmyPDF is not designed for use as a public web service where a malicious user could upload a chosen PDF. In particular, it is not necessarily secure against PDF malware or PDFs that cause denial of service. OCRmyPDF relies on Ghostscript, and therefore, if deployed online one should be prepared to comply with Ghostscript's Affero GPL license, OCRmyPDF's GPL license, and any other licenses.
+OCRmyPDF is not designed for use as a public web service where a
+malicious user could upload a chosen PDF. In particular, it is not
+necessarily secure against PDF malware or PDFs that cause denial of
+service. OCRmyPDF relies on Ghostscript, and therefore, if deployed
+online one should be prepared to comply with Ghostscript's Affero GPL
+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 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 to PDF, which could remove malware in embedded images.
+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
+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
+to PDF, which could remove malware in embedded images.
 
-OCRmyPDF should be relatively safe to use in a trusted intranet, with some considerations:
+OCRmyPDF should be relatively safe to use in a trusted intranet, with
+some considerations:
 
 Limiting CPU usage
-^^^^^^^^^^^^^^^^^^
+------------------
 
-OCRmyPDF will attempt to use all available CPUs and storage, so executing ``nice ocrmypdf`` or limiting the number of jobs with the ``-j`` argument may ensure the server remains available. Another option would be run OCRmyPDF jobs inside a Docker container, a virtual machine, or a cloud instance, which can impose its own limits on CPU usage and be terminated "from orbit" if it fails to complete.
+OCRmyPDF will attempt to use all available CPUs and storage, so
+executing ``nice ocrmypdf`` or limiting the number of jobs with the
+``-j`` argument may ensure the server remains available. Another option
+would be run OCRmyPDF jobs inside a Docker container, a virtual machine,
+or a cloud instance, which can impose its own limits on CPU usage and be
+terminated "from orbit" if it fails to complete.
 
 Temporary storage requirements
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+------------------------------
 
-OCRmyPDF will use a large amount of temporary storage for its work, proportional to the total number of pixels needed to rasterize the PDF. The raster image of a 8.5×11" color page at 300 DPI takes 25 MB uncompressed; OCRmyPDF saves its intermediates as PNG, but that still means it requires about 9 MB per intermediate based on average compression ratios. Multiple intermediates per page are also required, depending on the command line given. A rule of thumb would be to allow 100 MB of temporary storage per page in a file – meaning that a small cloud servers or small VM partitions should be provisioned with plenty of extra space, if say, a 500 page file might be sent.
+OCRmyPDF will use a large amount of temporary storage for its work,
+proportional to the total number of pixels needed to rasterize the PDF.
+The raster image of a 8.5×11" color page at 300 DPI takes 25 MB
+uncompressed; OCRmyPDF saves its intermediates as PNG, but that still
+means it requires about 9 MB per intermediate based on average
+compression ratios. Multiple intermediates per page are also required,
+depending on the command line given. A rule of thumb would be to allow
+100 MB of temporary storage per page in a file – meaning that a small
+cloud servers or small VM partitions should be provisioned with plenty
+of extra space, if say, a 500 page file might be sent.
 
-To check temporary storage usage on actual files, run ``ocrmypdf -k ...`` which will preserve and print the path to temporary storage when the job is done.
+To check temporary storage usage on actual files, run
+``ocrmypdf -k ...`` which will preserve and print the path to temporary
+storage when the job is done.
 
-To change where temporary files are stored, change the ``TMPDIR`` environment variable for ocrmypdf's environment. (Python's ``tempfile.gettempdir()`` returns the root directory in which temporary files will be stored.) For example, one could redirect ``TMPDIR`` to a large RAM disk to avoid wear on HDD/SSD and potentially improve performance. On Amazon Web Services, ``TMPDIR`` can be set to `empheral storage <https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/InstanceStorage.html>`_.
+To change where temporary files are stored, change the ``TMPDIR``
+environment variable for ocrmypdf's environment. (Python's
+``tempfile.gettempdir()`` returns the root directory in which temporary
+files will be stored.) For example, one could redirect ``TMPDIR`` to a
+large RAM disk to avoid wear on HDD/SSD and potentially improve
+performance. On Amazon Web Services, ``TMPDIR`` can be set to `empheral
+storage <https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/InstanceStorage.html>`__.
 
 Timeouts
-^^^^^^^^
+--------
 
-To prevent excessively long OCR jobs consider setting ``--tesseract-timeout`` and/or ``--skip-big`` arguments. ``--skip-big`` is particularly helpful if your PDFs include documents such as reports on standard page sizes with large images attached - often large images are not worth OCR'ing anyway.
+To prevent excessively long OCR jobs consider setting
+``--tesseract-timeout`` and/or ``--skip-big`` arguments. ``--skip-big``
+is particularly helpful if your PDFs include documents such as reports
+on standard page sizes with large images attached - often large images
+are not worth OCR'ing anyway.
 
 Commercial alternatives
-^^^^^^^^^^^^^^^^^^^^^^^
+-----------------------
 
-The author also provides professional services that include OCR and building databases around PDFs, and is happy to provide consultation.
-
-Abbyy Cloud OCR is a viable commercial alternative with a web services API.
+The author also provides professional services that include OCR and
+building databases around PDFs, and is happy to provide consultation.
 
+Abbyy Cloud OCR is a viable commercial alternative with a web services
+API.
 
 Password protection, digital signatures and certification
----------------------------------------------------------
+=========================================================
 
-Password protected PDFs usually have two passwords, and owner and user password. When the user password is set to empty, PDF readers will open the file automatically and marked it as "(SECURED)". While not as reliable as a digital signature, this indicates that whoever set the password approved of the file at that time. When the user password is set, the document cannot be viewed without the password.
+Password protected PDFs usually have two passwords, and owner and user
+password. When the user password is set to empty, PDF readers will open
+the file automatically and marked it as "(SECURED)". While not as
+reliable as a digital signature, this indicates that whoever set the
+password approved of the file at that time. When the user password is
+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.
+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``, 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.
 
-After OCR is applied, password protection is not permitted on PDF/A documents but the file can be converted to regular PDF.
+After OCR is applied, password protection is not permitted on PDF/A
+documents but the file can be converted to regular PDF.
 
-Many programs exist which are capable of inserting an image of someone's signature. On its own, this offers no security guarantees. It is trivial to remove the signature image and apply it to other files. This practice offers no real security.
+Many programs exist which are capable of inserting an image of someone's
+signature. On its own, this offers no security guarantees. It is trivial
+to remove the signature image and apply it to other files. This practice
+offers no real security.
 
-Important documents can be digitally signed and certified to attest to their authorship. OCRmyPDF cannot do this. Open source tools such as pdfbox (Java) have this capability as does Adobe Acrobat.
+Important documents can be digitally signed and certified to attest to
+their authorship. OCRmyPDF cannot do this. Open source tools such as
+pdfbox (Java) have this capability as does Adobe Acrobat.

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/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 +1,149 @@
+# 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
+
+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', False))
+ON_SUCCESS_DELETE = bool(os.getenv('OCR_ON_SUCCESS_DELETE', False))
+DESKEW = bool(os.getenv('OCR_DESKEW', False))
+OCR_JSON_SETTINGS = json.loads(os.getenv('OCR_JSON_SETTINGS', '{}'))
+POLL_NEW_FILE_SECONDS = os.getenv('OCR_POLL_NEW_FILE_SECONDS', 1)
+LOGLEVEL = os.environ.get('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.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"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)
+    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 +1,2 @@
-check-manifest >= 0.35
 twine >= 1.8.1
 coverage >= 4.5
-GitPython == 2.1.3

requirements/main.txt

@@ -1,13 +1,10 @@
 # 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
+cffi == 1.14.0
 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
+pdfminer.six == 20200402
+pikepdf == 1.10.2
+Pillow == 7.0.0
+reportlab == 3.5.34
+tqdm == 4.42.1

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 >= 2.8.0
+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,gs,img2pdf,pdfminer,pikepdf,pkg_resources,pytest,reportlab,setuptools,sphinx_rtd_theme,tqdm,watchdog,werkzeug
 
 [metadata]
 license_file = LICENSE

setup.py

@@ -21,11 +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
 
@@ -68,6 +69,7 @@ setup(
     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 +77,7 @@ 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",
@@ -95,15 +98,12 @@ setup(
         'chardet >= 3.0.4, < 4',  # unlisted requirement of pdfminer.six 20181108
         'cffi >= 1.9.1',  # must be a setup and install requirement
         '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 >= 20181108, <= 20200402',
+        'pikepdf >= 1.8.1, < 2',
+        'Pillow >= 6.2.0',
         '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 +15,21 @@
 # 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
-
-VERSION = __version__
-
+from . import helpers, hocrtransform, leptonica, pdfa, pdfinfo
+from ._version import PROGRAM_NAME, __version__
+from .api import Verbosity, configure_logging, ocr
 from .exceptions import (
-    ExitCode,
     BadArgsError,
-    PdfMergeFailedError,
-    MissingDependencyError,
-    UnsupportedImageFormatError,
     DpiError,
-    OutputFileAccessError,
-    PriorOcrFoundError,
-    InputFileError,
-    SubprocessOutputError,
     EncryptedPdfError,
+    ExitCode,
+    ExitCodeException,
+    InputFileError,
+    MissingDependencyError,
+    OutputFileAccessError,
+    PdfMergeFailedError,
+    PriorOcrFoundError,
+    SubprocessOutputError,
     TesseractConfigError,
+    UnsupportedImageFormatError,
 )
-
-from . import helpers
-from . import hocrtransform
-from . import leptonica
-from . import pdfa
-from . import pdfinfo
-from .api import ocrmypdf, configure_logging, Verbosity

src/ocrmypdf/__main__.py

@@ -21,12 +21,12 @@ import os
 import sys
 
 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 .api import Verbosity, configure_logging
+from .cli import parser
+from .exceptions import BadArgsError, ExitCode, MissingDependencyError
 
 
 def run(args=None):

src/ocrmypdf/_graft.py

@@ -21,9 +21,7 @@ from pathlib import Path
 
 import pikepdf
 
-from .exec import tesseract
-
-MAX_REPLACE_PAGES = int(os.environ.get('_OCRMYPDF_MAX_REPLACE_PAGES', 100))
+MAX_REPLACE_PAGES = 100
 
 
 def _update_page_resources(*, page, font, font_key, procset):
@@ -90,7 +88,7 @@ def strip_invisible_text(pdf, page):
     page.Contents = pikepdf.Stream(pdf, content_stream)
 
 
-def _weave_layers_graft(
+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"""
@@ -117,6 +115,7 @@ def _weave_layers_graft(
 
     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
@@ -134,8 +133,9 @@ def _weave_layers_graft(
     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
+    # 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'
 
@@ -186,7 +186,7 @@ class OcrGrafter:
         self.font, self.font_key = None, None
 
         self.pdfinfo = context.pdfinfo
-        self.output_file = context.get_path('weave_layers.pdf')
+        self.output_file = context.get_path('graft_layers.pdf')
 
         self.procset = self.pdf_base.make_indirect(
             pikepdf.Object.parse(b'[ /PDF /Text /ImageB /ImageC /ImageI ]')
@@ -221,14 +221,15 @@ class OcrGrafter:
         text_rotation = autorotate_correction
         text_misaligned = (text_rotation - content_rotation) % 360
         self.log.debug(
-            '%r',
-            [text_rotation, autorotate_correction, text_misaligned, content_rotation],
+            f"Rotations for page {pageno}: [text, auto, misalign, content] = "
+            f"{text_rotation}, {autorotate_correction}, "
+            f"{text_misaligned}, {content_rotation}"
         )
 
         if text and self.font:
             # Graft the text layer onto this page, whether new or old
             strip_old = self.context.options.redo_ocr
-            _weave_layers_graft(
+            _graft_text_layer(
                 pdf_base=self.pdf_base,
                 page_num=pageno + 1,
                 text=text,

src/ocrmypdf/_jobcontext.py

@@ -16,9 +16,9 @@
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
 import logging
+import os
 import shutil
 import sys
-import os
 
 
 class PicklableLoggerMixin:
@@ -93,21 +93,21 @@ class PageContext(PicklableLoggerMixin):
 
 def cleanup_working_files(work_folder, options):
     if options.keep_temporary_files:
-        print(f"Temporary working files saved at:\n{work_folder}", file=sys.stderr)
+        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] %s' % (self.extra['input_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),
+            #'[%s:%05u] %s' % (self.extra['input_filename'], self.extra['page'], msg),
             '%4u: %s' % (self.extra['page'], msg),
             kwargs,
         )
@@ -116,9 +116,9 @@ class LogNamePageAdapter(logging.LoggerAdapter):
 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))
+        adapter = LogNamePageAdapter(log, dict(input_filename=filename, page=page))
     elif filename:
-        adapter = LogNameAdapter(log, dict(filename=filename))
+        adapter = LogNameAdapter(log, dict(input_filename=filename))
     else:
         adapter = log
     return adapter

src/ocrmypdf/_pipeline.py

@@ -19,45 +19,45 @@ import os
 import re
 import sys
 from datetime import datetime, timezone
+from pathlib import Path
 from shutil import copyfileobj
 
 import img2pdf
-from PIL import Image
-
 import pikepdf
 from pikepdf.models.metadata import encode_pdf_date
+from PIL import Image
 
-from . import PROGRAM_NAME, VERSION, leptonica
-
-from ._plugins import load_plugin
+from . import leptonica
+from ._version import PROGRAM_NAME
+from ._version import __version__ as VERSION
 from .exceptions import (
     DpiError,
     EncryptedPdfError,
     InputFileError,
-    UnsupportedImageFormatError,
     PriorOcrFoundError,
+    UnsupportedImageFormatError,
 )
 from .exec import ghostscript, tesseract
-from .helpers import re_symlink
+from .helpers import safe_symlink
 from .hocrtransform import HocrTransform
 from .optimize import optimize
 from .pdfa import generate_pdfa_ps
-from .pdfinfo import Colorspace, PdfInfo
+from .pdfinfo import Colorspace, Encoding, PdfInfo
 
 VECTOR_PAGE_DPI = 400
 
 
 def triage_image_file(input_file, output_file, options, log):
+    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))
         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)
@@ -86,11 +86,10 @@ 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...")
@@ -125,20 +124,21 @@ 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, log):
     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(input_file, original_filename)
+        raise InputFileError(msg) from e
 
     triage_image_file(input_file, output_file, options, log)
     return output_file
@@ -160,12 +160,6 @@ def validate_pdfinfo_options(context):
     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 +183,7 @@ 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."
@@ -251,12 +245,14 @@ def is_ocr_required(page_context):
 
     ocr_required = True
 
-    if pageinfo.has_text:
+    if options.pages and pageinfo.pageno not in options.pages:
+        log.debug(f"skipped {pageinfo.pageno} as requested by --pages {options.pages}")
+        ocr_required = False
+    elif pageinfo.has_text:
         if not options.force_ocr and not (options.skip_text or options.redo_ocr):
-            log.error(
+            raise PriorOcrFoundError(
                 "page already has text! - aborting (use --force-ocr to force OCR)"
             )
-            raise PriorOcrFoundError()
         elif options.force_ocr:
             log.info("page already has text! - rasterizing text and running OCR anyway")
             ocr_required = True
@@ -333,6 +329,35 @@ def rasterize_preview(input_file, page_context):
     return output_file
 
 
+def describe_rotation(page_context, orient_conf, correction):
+    """
+    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, page_context):
     """
     Work out orientation correct for each page.
@@ -346,7 +371,7 @@ def get_orientation_correction(preview, page_context):
     correction to rotation.
 
     When we draw the real page for OCR, we rotate it by the CCW correction,
-    which points it (hopefully) upright. _weave.py takes care of the orienting
+    which points it (hopefully) upright. _graft.py takes care of the orienting
     the image and text layers.
 
     """
@@ -359,44 +384,14 @@ def get_orientation_correction(preview, page_context):
         tesseract_env=page_context.options.tesseract_env,
     )
 
-    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:
+    page_context.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
 
 
@@ -426,6 +421,9 @@ def rasterize(
             else:
                 device_idx = at_least('png16m')
 
+    if pageinfo.has_vector:
+        device_idx = at_least('png16m')
+
     device = colorspaces[device_idx]
 
     page_context.log.debug(f"Rasterize with {device}")
@@ -526,22 +524,11 @@ def create_ocr_image(image, page_context):
                 draw.rectangle(pixcoords, fill=white)
                 # draw.rectangle(pixcoords, outline=pink)
 
-        if options.mask_barcodes or options.threshold:
+        if options.threshold:
             pix = leptonica.Pix.frompil(im)
-            if options.threshold:
-                pix = pix.masked_threshold_on_background_norm()
-            if options.mask_barcodes:
-                barcodes = pix.locate_barcodes()
-                for barcode in barcodes:
-                    decoded, rect = barcode
-                    page_context.log.debug('masking barcode %s %r', decoded, rect)
-                    draw.rectangle(rect, fill=white)
+            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
         # Pillow requires integer DPI
         dpi = round(xres), round(yres)
@@ -571,7 +558,7 @@ def ocr_tesseract_hocr(input_file, page_context):
 
 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):
@@ -697,6 +684,7 @@ def generate_postscript_stub(context):
 def convert_to_pdfa(input_pdf, input_ps_stub, context):
     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
@@ -704,19 +692,28 @@ 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:
+            context.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(
         pdf_version=input_pdfinfo.min_version,
-        pdf_pages=[input_pdf, input_ps_stub],
+        pdf_pages=[fix_docinfo_file, input_ps_stub],
         output_file=output_file,
         compression=options.pdfa_image_compression,
         log=context.log,
@@ -726,53 +723,75 @@ def convert_to_pdfa(input_pdf, input_ps_stub, context):
     return output_file
 
 
+def should_linearize(working_file, context):
+    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, context):
     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'):
+            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", missing
+            )
+        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", missing
+            )
+
+    with pikepdf.open(context.origin) as original, pikepdf.open(working_file) as pdf:
+        docinfo = get_docinfo(original, options)
+        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):
     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
 
 

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_function_from_module(location):
-    """Load a function given a module location
-
-    For location=a.b.c, will effectively run "from a.b import c"
-
-    Example:
-        _load_function_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)
-    fn = getattr(module, object_name)
-    log.debug(f"Loaded function: from {module_name} import {object_name}")
-    return fn
-
-
-def _load_function_from_pyfile(location):
-    """Load a function from a file
-
-    Example:
-        _load_function_from_pyfile("test.py::blur_filter")
-    """
-    filename, object_name = location.split('::', maxsplit=1)
-    log.debug(f"Loading function {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)
-    fn = getattr(module, object_name)
-    return fn
-
-
-def load_plugin(plugin):
-    if callable(plugin):
-        return plugin
-
-    if not isinstance(plugin, str):
-        raise TypeError()
-
-    if '::' not in plugin:
-        plugin = _load_function_from_module(plugin)
-    else:
-        plugin = _load_function_from_pyfile(plugin)
-
-    return plugin
-
-
-def check_plugin_loadable(plugin):
-    load_plugin(plugin)
-    return plugin

src/ocrmypdf/_sync.py

@@ -23,11 +23,13 @@ import signal
 import sys
 import threading
 from collections import namedtuple
+from pathlib import Path
 from tempfile import mkdtemp
 
+import PIL
 from tqdm import tqdm
 
-from . import __version__
+from ._graft import OcrGrafter
 from ._jobcontext import PDFContext, cleanup_working_files, make_logger
 from ._pipeline import (
     convert_to_pdfa,
@@ -59,7 +61,6 @@ from ._validation import (
     create_input_file,
     report_output_file_size,
 )
-from ._weave import OcrGrafter
 from .exceptions import ExitCode, ExitCodeException
 from .exec import qpdf
 from .helpers import available_cpu_count
@@ -177,7 +178,7 @@ def post_process(pdf_file, context):
     return optimize_pdf(pdf_out, context)
 
 
-def worker_init(queue):
+def worker_init(queue, max_pixels):
     """Initialize a process pool worker"""
 
     # Ignore SIGINT (our parent process will kill us gracefully)
@@ -189,9 +190,15 @@ def worker_init(queue):
     root.handlers = []
     root.addHandler(h)
 
+    # In Windows, child process will not inherit our change to this value in
+    # the parent process, so ensure workers get it set
+    PIL.Image.MAX_IMAGE_PIXELS = max_pixels
 
-def worker_thread_init(queue):
-    pass
+
+def worker_thread_init(_queue, max_pixels):
+    # This is probably not needed since threads should all see the same memory,
+    # but done for consistency.
+    PIL.Image.MAX_IMAGE_PIXELS = max_pixels
 
 
 def log_listener(queue):
@@ -224,7 +231,26 @@ def exec_concurrent(context):
     # 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)
+        context.log.info("Start processing %d pages concurrently", max_workers)
+
+    # 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.
+    tess_threads = min(3, 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))
+    try:
+        tess_threads = int(context.options.tesseract_env['OMP_THREAD_LIMIT'])
+    except ValueError:  # OMP_THREAD_LIMIT initialized to non-numeric
+        context.log.error("Environment variable OMP_THREAD_LIMIT is not numeric")
+    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
@@ -246,26 +272,43 @@ def exec_concurrent(context):
         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):
+    ) as pbar:
+        pool = Pool(
+            processes=max_workers,
+            initializer=initializer,
+            initargs=(log_queue, PIL.Image.MAX_IMAGE_PIXELS),
+        )
+        try:
+            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 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()
-                log_queue.put_nowait(None)  # Terminate log listener
-                # Don't try listener.join() here, will deadlock
-                raise
+            raise
+        finally:
+            # Terminate log listener
+            log_queue.put_nowait(None)
+            pool.close()
+            pool.join()
 
-    log_queue.put_nowait(None)
     listener.join()
 
     # Output sidecar text
@@ -275,7 +318,6 @@ def exec_concurrent(context):
         copy_final(text, context.options.sidecar, context)
 
     # Merge layers to one single pdf
-    # pdf = weave_layers(layers, context)
     pdf = ocrgraft.finalize()
 
     # PDF/A and metadata
@@ -288,7 +330,25 @@ def exec_concurrent(context):
 class NeverRaise(Exception):
     """An exception that is never raised"""
 
-    pass
+    pass  # pylint: disable=unnecessary-pass
+
+
+def samefile(f1, f2):
+    if os.name == 'nt':
+        return f1 == f2
+    else:
+        return os.path.samefile(f1, f2)
+
+
+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 - %(message)s'
+    )
+    log_file_handler.setFormatter(formatter)
+    logging.getLogger(prefix).addHandler(log_file_handler)
+    return log_file_handler
 
 
 def run_pipeline(options, api=False):
@@ -300,20 +360,24 @@ def run_pipeline(options, api=False):
     if not options.jobs:
         options.jobs = available_cpu_count()
 
-    # Performance is improved by setting Tesseract to single threaded. In tests
-    # this gives better throughput than letting a smaller number of Tesseract
-    # jobs run multithreaded. Same story for pngquant. Tess <4 ignores this
-    # variable, but harmless to set if ignored.
-    os.environ.setdefault('OMP_THREAD_LIMIT', '1')
-
     work_folder = 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")
+
     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,
+            os.path.join(work_folder, 'origin.pdf'),
+            options,
+            log,
         )
 
         # Gather pdfinfo and create context
@@ -322,6 +386,7 @@ def run_pipeline(options, api=False):
             detailed_page_analysis=options.redo_ocr,
             progbar=options.progress_bar,
         )
+
         context = PDFContext(options, work_folder, origin_pdf, pdfinfo)
 
         # Validate options are okay for this pdf
@@ -329,8 +394,34 @@ def run_pipeline(options, api=False):
 
         # 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 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)
+
     except (KeyboardInterrupt if not api else NeverRaise) as e:
-        log.error("KeyboardInterrupt")
+        if options.verbose >= 1:
+            log.exception("KeyboardInterrupt")
+        else:
+            log.error("KeyboardInterrupt")
         return ExitCode.ctrl_c
     except (ExitCodeException if not api else NeverRaise) as e:
         if str(e):
@@ -342,29 +433,12 @@ def run_pipeline(options, api=False):
         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

@@ -17,6 +17,7 @@
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
 
+import locale
 import logging
 import os
 import sys
@@ -41,12 +42,13 @@ from .exec import (
     tesseract,
     unpaper,
 )
-from .helpers import is_file_writable, re_symlink
+from .helpers import is_file_writable, is_iterable_notstr, monotonic, safe_symlink
 
 # -------------
 # External dependencies
 
 HOCR_OK_LANGS = frozenset(['eng', 'deu', 'spa', 'ita', 'por'])
+DEFAULT_LANGUAGE = 'eng'  # Enforce English hegemony
 
 log = logging.getLogger(__name__)
 
@@ -56,9 +58,21 @@ log = logging.getLogger(__name__)
 verify_python3_env()
 
 
+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."
+        )
+
+
 def check_options_languages(options):
     if not options.language:
-        options.language = ['eng']  # Enforce English hegemony
+        options.language = [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)
 
     # Support v2.x "eng+deu" language syntax
     if '+' in options.language[0]:
@@ -107,7 +121,7 @@ def check_options_output(options):
         options.pdf_renderer = 'sandwich'
 
     if options.pdf_renderer == 'sandwich' and not tesseract.has_textonly_pdf(
-        options.tesseract_env
+        options.tesseract_env, languages
     ):
         raise MissingDependencyError(
             "You are using an alpha version of Tesseract 4.0 that does not support "
@@ -172,6 +186,36 @@ def check_options_preprocessing(options):
             raise BadArgsError(str(e))
 
 
+def _pages_from_ranges(ranges):
+    if is_iterable_notstr(ranges):
+        return set(ranges)
+    pages = []
+    page_groups = ranges.replace(' ', '').split(',')
+    for g in page_groups:
+        if not g:
+            continue
+        try:
+            start, end = g.split('-')
+        except ValueError:
+            pages.append(int(g) - 1)
+        else:
+            try:
+                pages.extend(range(int(start) - 1, int(end)))
+            except ValueError:
+                raise BadArgsError("invalid page range")
+
+    if not monotonic(pages):
+        log.warning(
+            "List of pages to process contains duplicate pages, or pages that are "
+            "out of order"
+        )
+    if any(page < 0 for page in pages):
+        raise BadArgsError("pages refers to a page number less than 1")
+
+    log.debug("OCRing only these pages: %s", pages)
+    return set(pages)
+
+
 def check_options_ocr_behavior(options):
     exclusive_options = sum(
         [
@@ -180,9 +224,11 @@ def check_options_ocr_behavior(options):
         ]
     )
     if exclusive_options >= 2:
-        raise BadArgsError(
-            "Error: choose only one of --force-ocr, --skip-text, --redo-ocr."
-        )
+        raise BadArgsError("Choose only one of --force-ocr, --skip-text, --redo-ocr.")
+    if options.pages and options.sidecar:
+        raise BadArgsError("--pages and --sidecar are mutually exclusive")
+    if options.pages:
+        options.pages = _pages_from_ranges(options.pages)
 
 
 def check_options_optimizing(options):
@@ -224,10 +270,13 @@ 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 (
+    if not tesseract.has_user_words(options.tesseract_env) and (
         options.user_words or options.user_patterns
     ):
-        log.warning('Tesseract 4.x ignores --user-words, so this has no effect')
+        log.warning(
+            "Tesseract 4.0 ignores --user-words and --user-patterns, so these "
+            "arguments have no effect."
+        )
 
 
 def check_options_metadata(options):
@@ -252,6 +301,7 @@ def check_options_pillow(options):
 
 
 def check_options(options):
+    check_platform()
     check_options_languages(options)
     check_options_metadata(options)
     check_options_output(options)
@@ -262,10 +312,9 @@ def check_options(options):
     check_options_advanced(options)
     check_options_pillow(options)
     check_dependency_versions(options)
-    check_environ(options)
 
 
-def check_closed_streams(options):
+def check_closed_streams(options):  # pragma: no cover
     """Work around Python issue with multiprocessing forking on closed streams
 
     https://bugs.python.org/issue28326
@@ -324,21 +373,6 @@ def log_page_orientations(pdfinfo):
         log.info('Page orientations detected: %s', ' '.join(orientations))
 
 
-def check_environ(options):
-    old_envvars = (
-        'OCRMYPDF_TESSERACT',
-        'OCRMYPDF_QPDF',
-        'OCRMYPDF_GS',
-        'OCRMYPDF_UNPAPER',
-    )
-    for k in old_envvars:
-        if k in os.environ:
-            log.warning(
-                "OCRmyPDF no longer uses the environment variable {k}."
-                "Change PATH to select alternate programs."
-            )
-
-
 def create_input_file(options, work_folder):
     if options.input_file == '-':
         # stdin
@@ -346,12 +380,12 @@ def create_input_file(options, work_folder):
         target = os.path.join(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
+            safe_symlink(options.input_file, target)
+            return target, os.fspath(options.input_file)
         except FileNotFoundError:
             raise InputFileError(f"File not found - {options.input_file}")
 
@@ -389,12 +423,24 @@ def report_output_file_size(options, input_file, output_file):
         'force_ocr',
     }
     for arg in image_preproc:
-        attr = getattr(options, arg, None)
-        if not attr:
-            continue
-        reasons.append(
-            f"The argument --{arg.replace('_', '-')} was issued, causing transcoding."
-        )
+        if getattr(options, arg, False):
+            reasons.append(
+                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"
@@ -410,7 +456,7 @@ def report_output_file_size(options, input_file, output_file):
 def check_dependency_versions(options):
     check_external_program(
         program='tesseract',
-        package={'darwin': 'tesseract', 'linux': 'tesseract-ocr'},
+        package={'linux': 'tesseract-ocr'},
         version_checker=tesseract.version,
         need_version='4.0.0',  # using backport for Travis CI
     )
@@ -420,11 +466,12 @@ def check_dependency_versions(options):
         version_checker=ghostscript.version,
         need_version='9.15',  # limited by Travis CI / Ubuntu 14.04 backports
     )
-    if ghostscript.version() == '9.24':
+    gs_version = ghostscript.version()
+    if gs_version in ('9.24', '9.51'):
         raise MissingDependencyError(
-            "Ghostscript 9.24 contains serious regressions and is not "
-            "supported. Please upgrade to Ghostscript 9.25 or use an older "
-            "version."
+            f"Ghostscript {gs_version} contains serious regressions and is not "
+            "supported. Please upgrade to a newer version, or downgrade to the "
+            "previous version."
         )
     check_external_program(
         program='qpdf',

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

@@ -18,22 +18,32 @@
 import logging
 import os
 import sys
+from contextlib import suppress
 from enum import IntEnum
 from pathlib import Path
+from typing import Dict, Iterable
 
 from tqdm import tqdm
 
-from .cli import parser
 from ._sync import run_pipeline
 from ._validation import check_options
+from .cli import parser
 
 
 class TqdmConsole:
-    """Wrapper to log messages in a way that is compatible with tqdm progress bar"""
+    """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 >= (3, 6)
+        self.py36 = sys.version_info[0:2] == (3, 6)
 
     def write(self, msg):
         # When no progress bar is active, tqdm.write() routes to print()
@@ -44,7 +54,7 @@ class TqdmConsole:
             tqdm.write(msg.rstrip(), end='\n', file=self.file)
 
     def flush(self):
-        if hasattr(self.file, "flush"):
+        with suppress(AttributeError):
             self.file.flush()
 
 
@@ -57,17 +67,21 @@ 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 +92,14 @@ 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))
@@ -97,8 +114,6 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
         console.setLevel(logging.INFO)
 
     formatter = logging.Formatter('%(levelname)7s - %(message)s')
-    if verbosity >= 1:
-        log.setLevel(logging.DEBUG)
     if verbosity >= 2:
         formatter = logging.Formatter('%(name)s - %(levelname)7s - %(message)s')
 
@@ -111,26 +126,42 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
         pil_log = logging.getLogger('PIL')
         pil_log.setLevel(logging.INFO)
 
+    if manage_root_logger:
+        logging.captureWarnings(True)
 
-def create_options(*, input_file, output_file, **kwargs):
+    return log
+
+
+def create_options(*, input_file: os.PathLike, output_file: os.PathLike, **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 {'tesseract_env', 'progress_bar'}:
             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):
@@ -150,59 +181,58 @@ def create_options(*, input_file, output_file, **kwargs):
 
     # 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
+        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(input_file)
 
     return options
 
 
-def ocrmypdf(  # 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,
-    mask_barcodes=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,
-    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,
+    keep_temporary_files: bool = None,
+    progress_bar: bool = None,
+    tesseract_env: Dict[str, str] = None,
 ):
     """Run OCRmyPDF on one PDF or image.
 

src/ocrmypdf/cli.py

@@ -17,8 +17,10 @@
 
 import argparse
 
-from . import PROGRAM_NAME, VERSION
-from ._plugins import check_plugin_loadable
+from ._version import PROGRAM_NAME as _PROGRAM_NAME
+from ._version import __version__ as _VERSION
+
+__all__ = ['parser']
 
 
 def numeric(basetype, min_=None, max_=None):
@@ -55,7 +57,7 @@ class ArgumentParser(argparse.ArgumentParser):
 
 
 parser = ArgumentParser(
-    prog=PROGRAM_NAME,
+    prog=_PROGRAM_NAME,
     fromfile_prefix_chars='@',
     formatter_class=argparse.RawDescriptionHelpFormatter,
     description="""\
@@ -168,7 +170,7 @@ parser.add_argument(
 parser.add_argument(
     '--version',
     action='version',
-    version=VERSION,
+    version=_VERSION,
     help="Print program version and exit",
 )
 
@@ -186,8 +188,9 @@ jobcontrol.add_argument(
 jobcontrol.add_argument(
     '-v',
     '--verbose',
-    type=int,
+    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 "
@@ -267,13 +270,6 @@ preprocessing.add_argument(
     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(
-    '--mask-barcodes',
-    action='store_true',
-    help="EXPERIMENTAL. Mask out any barcodes that appear in the PDF so they are not "
-    "considered during OCR. Barcodes can introduce false characters into "
-    "OCR.",
-)
 preprocessing.add_argument(
     '--threshold',
     action='store_true',
@@ -380,6 +376,11 @@ optimizing.add_argument(
 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',
@@ -467,13 +468,17 @@ advanced.add_argument(
     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
+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.",
 )
 
 debugging = parser.add_argument_group(

src/ocrmypdf/exec/__init__.py

@@ -20,16 +20,83 @@
 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 distutils.version import LooseVersion
+from functools import lru_cache
+from subprocess import PIPE, STDOUT, CalledProcessError
+from subprocess import run as subprocess_run
 
-log = logging.Logger(__name__)
+from ..exceptions import ExitCode, MissingDependencyError
+
+log = logging.getLogger(__name__)
+
+
+def _get_program(args, env=None):
+    program = args[0]
+    test_path = env.get('_OCRMYPDF_TEST_PATH', '')
+    if test_path:
+        program = shutil.which(program, path=test_path)
+    return program
+
+
+def run(args, *, env=None, **kwargs):
+    """Wrapper around subprocess.run()
+
+    The main purpose of this wrapper is to allow us to substitute the main program
+    for a spoof in the test suite. The hidden variable _OCRMYPDF_TEST_PATH replaces
+    the main PATH as a location to check for programs to run.
+
+    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 = _get_program(args, env)
+
+    # If we are running a .py on Windows, ensure we call it with this Python
+    # (to support test suite shims)
+    if os.name == 'nt' and program.lower().endswith('.py'):
+        args = [sys.executable, program] + args[1:]
+    else:
+        args = [program] + args[1:]
+
+    if os.name == 'nt':
+        paths = os.pathsep.join(os.get_exec_path(env))
+        if not shutil.which(args[0], path=paths):
+            shimmed_path = shim_paths_with_program_files(env)
+            new_args0 = shutil.which(args[0], path=shimmed_path)
+            if new_args0:
+                args[0] = new_args0
+
+    process_log = log.getChild(os.path.basename(program))
+    process_log.debug("Running: %s", args)
+    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
+    proc = subprocess_run(args, env=env, **kwargs)
+    if process_log.isEnabledFor(logging.DEBUG):
+        try:
+            stderr = proc.stderr.decode('utf-8', 'replace')
+        except AttributeError:
+            stderr = proc.stderr
+        if stderr:
+            process_log.debug("stderr = %s", stderr)
+    return proc
 
 
 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 +132,32 @@ 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', '')
+    paths = []
+    try:
+        for dirname in os.listdir(program_files):
+            if dirname.lower() == 'tesseract-ocr':
+                paths.append(os.path.join(program_files, dirname))
+            elif dirname.lower() == 'gs':
+                try:
+                    latest_gs = max(
+                        os.listdir(os.path.join(program_files, dirname)),
+                        key=lambda d: float(d[2:]),
+                    )
+                except (FileNotFoundError, NotADirectoryError):
+                    continue
+                paths.append(os.path.join(program_files, dirname, latest_gs, 'bin'))
+    except EnvironmentError:
+        pass
+    paths.extend(path for path in os.get_exec_path(env) if path not in set(paths))
+    return os.pathsep.join(paths)
+
+
 missing_program = '''
 The program '{program}' could not be executed or was not found on your
 system PATH.
@@ -110,23 +203,33 @@ 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):
@@ -168,7 +271,15 @@ def check_external_program(
             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 LooseVersion(found_version) < LooseVersion(need_version):
         _error_old_version(program, package, need_version, found_version, required_for)
         if not recommended:
             raise MissingDependencyError()

src/ocrmypdf/exec/ghostscript.py

@@ -15,26 +15,51 @@
 # 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
+import warnings
+from contextlib import suppress
 from functools import lru_cache
+from io import BytesIO
 from os import fspath
-from shutil import copy
-from subprocess import PIPE, STDOUT, run
-from tempfile import NamedTemporaryFile
+from pathlib import Path
+from shutil import which
+from subprocess import PIPE, CalledProcessError
 
 from PIL import Image
 
-from ..exceptions import SubprocessOutputError
-from . import get_version
-
+from ..exceptions import MissingDependencyError, SubprocessOutputError
+from . import get_version, run
 
 gslog = logging.getLogger()
 
+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
+
 
 @lru_cache(maxsize=1)
 def version():
-    return get_version('gs')
+    return get_version(GS)
 
 
 def jpeg_passthrough_available():
@@ -79,9 +104,12 @@ def extract_text(input_file, pageno=1):
     else:
         pages = []
 
+    # Note due to bug https://bugs.ghostscript.com/show_bug.cgi?id=701971
+    # Ghostscript <= 9.50 will truncate output unless we write to stdout, so
+    # don't write to a file.
     args_gs = (
         [
-            'gs',
+            GS,
             '-dQUIET',
             '-dSAFER',
             '-dBATCH',
@@ -90,14 +118,15 @@ def extract_text(input_file, pageno=1):
             '-dTextFormat=0',
         ]
         + pages
-        + ['-o', '-', fspath(input_file)]
+        + ['-o', '-', fspath(input_file), "-sstdout=%stderr"]
     )
 
-    p = run(args_gs, stdout=PIPE, stderr=PIPE)
-    if p.returncode != 0:
+    try:
+        p = run(args_gs, stdout=PIPE, stderr=PIPE, check=True)
+    except CalledProcessError as e:
         raise SubprocessOutputError(
-            'Ghostscript text extraction failed\n%s\n%s\n%s'
-            % (input_file, p.stdout.decode(), p.stderr.decode())
+            'Ghostscript text extraction failed\n%s\n%s'
+            % (input_file, e.stderr.decode(errors='replace'))
         )
 
     return p.stdout
@@ -139,54 +168,56 @@ def rasterize_pdf(
     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),
-            ]
-        )
+    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',
+            '-',
+            '-sstdout=%stderr',
+            '-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)
+    log.debug(args_gs)
+    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)
 
-        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)
+    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[1], page_dpi[0]
+        im.save(fspath(output_file), dpi=page_dpi)
 
 
 def generate_pdfa(
@@ -194,7 +225,7 @@ def generate_pdfa(
     output_file,
     compression,
     log,
-    threads=1,
+    threads=None,  # deprecated parameter
     pdf_version='1.5',
     pdfa_part='2',
 ):
@@ -217,6 +248,10 @@ def generate_pdfa(
     """
     if not log:
         log = gslog
+    if threads is not None:
+        warnings.warn(
+            "use of deprecated parameter 'threads'", category=DeprecationWarning
+        )
 
     compression_args = []
     if compression == 'jpeg':
@@ -250,36 +285,55 @@ def generate_pdfa(
         # 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:
+    # 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....
@@ -287,12 +341,3 @@ def generate_pdfa(
                 "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/jbig2enc.py

@@ -15,11 +15,13 @@
 # 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 jbig2 executable"""
+
 from functools import lru_cache
-from subprocess import PIPE, run
+from subprocess import PIPE
 
-from . import get_version
 from ..exceptions import MissingDependencyError
+from . import get_version, run
 
 
 @lru_cache(maxsize=1)

src/ocrmypdf/exec/pngquant.py

@@ -15,14 +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/>.
 
+"""Interface to pngquant executable"""
+
 from functools import lru_cache
 from subprocess import run
 from tempfile import NamedTemporaryFile
 
 from PIL import Image
 
-from . import get_version
 from ..exceptions import MissingDependencyError
+from . import get_version
 
 
 @lru_cache(maxsize=1)
@@ -40,8 +42,7 @@ 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/qpdf.py

@@ -15,35 +15,47 @@
 # 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
+"""Interface to qpdf executable"""
 
-from . import get_version
+from io import StringIO
+
+import pikepdf
 
 
-@lru_cache(maxsize=1)
 def version():
-    return get_version('qpdf', regex=r'qpdf version (.+)')
+    return pikepdf.__libqpdf_version__
 
 
 def check(input_file, log=None):
-    args_qpdf = ['qpdf', '--check', fspath(input_file)]
-
-    if log is None:
-        import logging as log
-
+    pdf = None
     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)
+        pdf = pikepdf.open(input_file)
+    except pikepdf.PdfError as e:
+        if log:
+            log.error(e)
         return False
-    return True
+    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()

src/ocrmypdf/exec/tesseract.py

@@ -15,22 +15,23 @@
 # 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 subprocess import PIPE, STDOUT, CalledProcessError, TimeoutExpired
 
-from . import get_version
 from ..exceptions import (
     MissingDependencyError,
-    TesseractConfigError,
     SubprocessOutputError,
+    TesseractConfigError,
 )
-from ..helpers import page_number
+from ..helpers import page_number, safe_symlink
+from . import get_version, run
 
 OrientationConfidence = namedtuple('OrientationConfidence', ('angle', 'confidence'))
 
@@ -52,6 +53,12 @@ HOCR_TEMPLATE = """<?xml version="1.0" encoding="UTF-8"?>
 """
 
 
+class TesseractLoggerAdapter(logging.LoggerAdapter):
+    def process(self, msg, kwargs):
+        kwargs['extra'] = self.extra
+        return '[tesseract] %s' % (msg), kwargs
+
+
 def version(tesseract_env=None):
     return get_version('tesseract', regex=r'tesseract\s(.+)', env=tesseract_env)
 
@@ -61,13 +68,13 @@ def v4(tesseract_env=None):
     return version(tesseract_env) >= '4'
 
 
-def has_textonly_pdf(tesseract_env=None):
+def has_textonly_pdf(tesseract_env=None, 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(
@@ -88,6 +95,15 @@ def has_textonly_pdf(tesseract_env=None):
     return False
 
 
+def has_user_words(tesseract_env=None):
+    """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(tesseract_env) >= '4.1'
+
+
 def languages(tesseract_env=None):
     def lang_error(output):
         msg = (
@@ -112,9 +128,10 @@ def languages(tesseract_env=None):
     except CalledProcessError as e:
         raise MissingDependencyError(lang_error(e.output)) from e
 
+    for line in output.splitlines():
+        if line.startswith('Error'):
+            raise MissingDependencyError(lang_error(output))
     header, *rest = output.splitlines()
-    if not header.startswith('List of available languages'):
-        raise MissingDependencyError(lang_error(output))
     return set(lang.strip() for lang in rest)
 
 
@@ -122,7 +139,7 @@ def tess_base_args(langs, engine_mode):
     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
 
@@ -170,19 +187,15 @@ 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(mainlog, stdout, input_file):
+    log = TesseractLoggerAdapter(
+        mainlog, extra=mainlog.extra if hasattr(mainlog, 'extra') else None
+    )
 
     try:
         text = stdout.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 = stdout.decode('utf-8', 'ignore')
 
     lines = text.splitlines()
     for line in lines:
@@ -191,25 +204,25 @@ 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")
+            log.warning("lots of diacritics - possibly poor OCR")
         elif line.startswith('OSD: Weak margin'):
-            log.warning(prefix + "unsure about page orientation")
+            log.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())
+            log.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())
+            log.error(line.strip())
         elif 'warning' in line.lower():
-            log.warning(prefix + line.strip())
+            log.warning(line.strip())
         elif 'read_params_file' in line.lower():
-            log.error(prefix + line.strip())
+            log.error(line.strip())
         else:
-            log.info(prefix + line.strip())
+            log.info(line.strip())
 
 
 def page_timedout(log, input_file, timeout):
@@ -224,8 +237,8 @@ def _generate_null_hocr(output_hocr, output_sidecar, image):
     the same size as the input image."""
     from PIL import Image
 
-    im = Image.open(image)
-    w, h = im.size
+    with Image.open(image) as im:
+        w, h = im.size
 
     with open(output_hocr, 'w', encoding="utf-8") as f:
         f.write(HOCR_TEMPLATE.format(w, h))
@@ -247,8 +260,8 @@ def generate_hocr(
     log,
 ):
 
-    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'))
+    output_hocr = next(o for o in output_files if fspath(o).endswith('.hocr'))
+    output_sidecar = next(o for o in output_files if fspath(o).endswith('.txt'))
     prefix = os.path.splitext(output_hocr)[0]
 
     args_tesseract = tess_base_args(language, engine_mode)
@@ -266,7 +279,6 @@ def generate_hocr(
     # 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,
@@ -305,7 +317,7 @@ def use_skip_page(text_only, skip_pdf, output_pdf, output_text):
         # Substitute a "skipped page"
         with suppress(FileNotFoundError):
             os.remove(output_pdf)  # In case it was partially created
-        os.symlink(skip_pdf, output_pdf)
+        safe_symlink(skip_pdf, output_pdf)
         return
 
     # Or normally, just write a 0 byte file to the output to indicate a skip
@@ -349,7 +361,7 @@ def generate_pdf(
     if pagesegmode is not None:
         args_tesseract.extend(['--psm', str(pagesegmode)])
 
-    if text_only and has_textonly_pdf(tesseract_env):
+    if text_only and has_textonly_pdf(tesseract_env, language):
         args_tesseract.extend(['-c', 'textonly_pdf=1'])
 
     if user_words:
@@ -365,7 +377,6 @@ def generate_pdf(
 
     args_tesseract.extend([input_image, prefix, 'pdf', 'txt'] + tessconfig)
     try:
-        log.debug(args_tesseract)
         p = run(
             args_tesseract,
             stdout=PIPE,

src/ocrmypdf/exec/unpaper.py

@@ -18,19 +18,20 @@
 # unpaper documentation:
 # https://github.com/Flameeyes/unpaper/blob/master/doc/basic-concepts.md
 
+"""Interface to unpaper executable"""
+
 import os
 import shlex
-import subprocess
-import sys
 from functools import lru_cache
 from subprocess import PIPE, STDOUT, CalledProcessError
 from tempfile import TemporaryDirectory
 
-from . import get_version
-from ..exceptions import MissingDependencyError, SubprocessOutputError
-
 from PIL import Image
 
+from ..exceptions import MissingDependencyError, SubprocessOutputError
+from . import get_version
+from . import run as external_run
+
 
 @lru_cache(maxsize=1)
 def version():
@@ -42,33 +43,30 @@ def run(input_file, output_file, dpi, log, mode_args):
 
     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")
+    with TemporaryDirectory() as tmpdir, Image.open(input_file) as im:
+        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')
+                else:
+                    im = im.convert(mode='RGB')
+            except IOError as e:
+                im.close()
+                raise MissingDependencyError(
+                    "Could not convert image with type " + im.mode
+                ) from e
+
         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:
-            im.close()
+            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
-
-    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()
 
         # To prevent any shenanigans from accepting arbitrary parameters in
         # --unpaper-args, we:
@@ -79,7 +77,7 @@ def run(input_file, output_file, dpi, log, mode_args):
         # their unpaper arguments (whether intentionally or otherwise)
         args_unpaper.extend([input_pnm, output_pnm])
         try:
-            proc = subprocess.run(
+            proc = external_run(
                 args_unpaper,
                 check=True,
                 close_fds=True,
@@ -95,10 +93,12 @@ def run(input_file, output_file, dpi, log, mode_args):
             log.debug(proc.stdout)
             # unpaper sets dpi to 72; fix this
             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
 

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

@@ -18,6 +18,7 @@
 import logging
 import multiprocessing
 import os
+import shutil
 import warnings
 from collections.abc import Iterable
 from contextlib import suppress
@@ -27,14 +28,14 @@ from pathlib import Path
 log = logging.getLogger(__name__)
 
 
-def re_symlink(input_file, soft_link_name, *args, **kwargs):
+def safe_symlink(input_file: os.PathLike, soft_link_name: os.PathLike, *args, **kwargs):
     """
     Helper function: relinks soft symbolic link if necessary
     """
     if len(args) == 1 and isinstance(args[0], logging.Logger):
-        log.warning("Deprecated: re_symlink(,log)")
+        log.warning("Deprecated: safe_symlink(,log)")
     if 'log' in kwargs:
-        log.warning('Deprecated: re_symlink(...log=)')
+        log.warning('Deprecated: safe_symlink(...log=)')
 
     input_file = os.fspath(input_file)
     soft_link_name = os.fspath(soft_link_name)
@@ -60,13 +61,27 @@ 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 page_number(input_file):
+def is_iterable_notstr(thing):
+    return isinstance(thing, Iterable) and not isinstance(thing, str)
+
+
+def monotonic(L: Iterable):
+    """Does list increase monotonically?"""
+    return all(b > a for a, b in zip(L, L[1:]))
+
+
+def page_number(input_file: os.PathLike):
     """Get one-based page number implied by filename (000002.pdf -> 2)"""
     return int(os.path.basename(os.fspath(input_file))[0:6])
 
@@ -76,49 +91,49 @@ def available_cpu_count():
         return multiprocessing.cpu_count()
     except NotImplementedError:
         pass
-
-    try:
-        import psutil
-
-        return psutil.cpu_count()
-    except (ImportError, AttributeError):
-        pass
-
     warnings.warn(
         "Could not get CPU count.  Assuming one (1) CPU." "Use -j N to set manually."
     )
     return 1
 
 
-def is_file_writable(test_file):
+def is_file_writable(test_file: os.PathLike):
     """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)
-
-    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
+    try:
+        if not isinstance(test_file, Path):
+            p = Path(test_file)
         else:
-            fp.close()
-            with suppress(OSError):
-                p.unlink()
-        return True
+            p = test_file
+
+        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
+            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 deprecated(func):

src/ocrmypdf/hocrtransform.py

@@ -88,7 +88,7 @@ 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
         """
@@ -190,7 +190,7 @@ class HocrTransform:
             pt = self.pt_from_pixel(pxl_coords)
 
             # draw the bbox border
-            if showBoundingboxes:
+            if showBoundingboxes:  # pragma: no cover
                 pdf.rect(
                     pt.x1, self.height - pt.y2, pt.x2 - pt.x1, pt.y2 - pt.y1, fill=1
                 )
@@ -231,7 +231,7 @@ class HocrTransform:
         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(
@@ -269,7 +269,7 @@ class HocrTransform:
         # of the line box
         baseline_y2 = self.height - (line_box.y2 + intercept)
 
-        if showBoundingboxes:
+        if showBoundingboxes:  # pragma: no cover
             # draw the baseline in magenta, dashed
             pdf.setDash()
             pdf.setStrokeColorRGB(0.95, 0.65, 0.95)
@@ -318,7 +318,7 @@ class HocrTransform:
             font_width = pdf.stringWidth(elemtxt, fontname, fontsize)
 
             # draw the bbox border
-            if showBoundingboxes:
+            if showBoundingboxes:  # pragma: no cover
                 pdf.rect(
                     box.x1, self.height - line_box.y2, box_width, line_height, fill=0
                 )

src/ocrmypdf/leptonica.py

@@ -33,14 +33,46 @@ from io import BytesIO
 from os import fspath
 from tempfile import TemporaryFile
 
+from .exceptions import MissingDependencyError
+from .exec import shim_paths_with_program_files
 from .lib._leptonica import ffi
 
 # 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:
@@ -292,9 +324,11 @@ 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.
@@ -302,11 +336,19 @@ 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):
@@ -502,17 +544,14 @@ class Pix(LeptonicaObject):
         display=0,
         pdfdir=ffi.NULL,
     ):
+        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

@@ -22,16 +22,16 @@ from collections import defaultdict
 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 .exec import jbig2enc, pngquant
+from .helpers import safe_symlink
 
 DEFAULT_JPEG_QUALITY = 75
 DEFAULT_PNG_QUALITY = 70
@@ -73,6 +73,9 @@ 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
 
 
@@ -104,6 +107,15 @@ def extract_image_generic(*, pike, root, log, image, xref, options):
         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,
@@ -143,6 +155,17 @@ 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
 
@@ -276,7 +299,8 @@ def _produce_jbig2_images(jbig2_groups, root, log, options):
         ) as pbar:
             for future in concurrent.futures.as_completed(futures):
                 proc = future.result()
-                log.debug(proc.stderr.decode())
+                if proc.stderr:
+                    log.debug(proc.stderr.decode())
                 pbar.update()
 
 
@@ -342,6 +366,7 @@ def transcode_jpegs(pike, jpegs, root, log, options):
 
 
 def transcode_pngs(pike, images, image_name_fn, root, log, options):
+    modified = set()
     if options.optimize >= 2:
         png_quality = (
             max(10, options.png_quality - 10),
@@ -362,6 +387,7 @@ def transcode_pngs(pike, images, image_name_fn, root, log, options):
                         png_quality[1],
                     )
                 )
+                modified.add(xref)
             with tqdm(
                 desc="PNGs",
                 total=len(futures),
@@ -371,10 +397,14 @@ def transcode_pngs(pike, images, image_name_fn, root, log, options):
                 for _future in concurrent.futures.as_completed(futures):
                     pbar.update()
 
-    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 +420,90 @@ 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:
+            return rewrite_png(pike, im_obj, compdata, log)
+        elif compdata.type == leptonica.lept.L_G4_ENCODE:
+            return rewrite_png_as_g4(pike, im_obj, compdata, log)
 
 
-def optimize(input_file, output_file, context):
+def rewrite_png_as_g4(pike, im_obj, compdata, log):
+    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, log):
+    # 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):
     log = context.log
     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:
@@ -478,11 +529,7 @@ def optimize(input_file, output_file, context):
 
         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
@@ -496,10 +543,13 @@ 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):
@@ -510,11 +560,11 @@ def main(infile, outfile, level, jobs=1):
         """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
@@ -525,7 +575,7 @@ 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,
@@ -534,7 +584,16 @@ def main(infile, outfile, level, jobs=1):
     with TemporaryDirectory() as td:
         context = PDFContext(options, td, infile, 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

@@ -16,808 +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 collections import namedtuple
-from decimal import Decimal
-from enum import Enum
-import logging
-from math import hypot, isclose
-from os import fspath
-from pathlib import Path
-from warnings import warn
-import re
-
-from pikepdf import PdfMatrix
-import pikepdf
-from tqdm import tqdm
-
-from . import ghosttext
-
-from ..exceptions import EncryptedPdfError, MissingDependencyError
-
-
-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'
-)
-
-FRIENDLY_COLORSPACE = {
-    '/DeviceGray': Colorspace.gray,
-    '/CalGray': Colorspace.gray,
-    '/DeviceRGB': Colorspace.rgb,
-    '/CalRGB': Colorspace.rgb,
-    '/DeviceCMYK': Colorspace.cmyk,
-    '/Lab': Colorspace.lab,
-    '/ICCBased': Colorspace.icc,
-    '/Indexed': Colorspace.index,
-    '/Separation': Colorspace.sep,
-    '/DeviceN': Colorspace.devn,
-    '/Pattern': Colorspace.pattern,
-    '/G': Colorspace.gray,  # Abbreviations permitted in inline images
-    '/RGB': Colorspace.rgb,
-    '/CMYK': Colorspace.cmyk,
-    '/I': Colorspace.index,
-}
-
-FRIENDLY_ENCODING = {
-    '/CCITTFaxDecode': Encoding.ccitt,
-    '/DCTDecode': Encoding.jpeg,
-    '/JPXDecode': Encoding.jpeg2000,
-    '/JBIG2Decode': Encoding.jbig2,
-    '/CCF': Encoding.ccitt,  # Abbreviations permitted in inline images
-    '/DCT': Encoding.jpeg,
-    '/AHx': Encoding.asciihex,
-    '/A85': Encoding.ascii85,
-    '/LZW': Encoding.lzw,
-    '/Fl': Encoding.flate,
-    '/RL': Encoding.runlength,
-}
-
-FRIENDLY_COMP = {
-    Colorspace.gray: 1,
-    Colorspace.rgb: 3,
-    Colorspace.cmyk: 4,
-    Colorspace.lab: 3,
-    Colorspace.index: 1,
-}
-
-
-UNIT_SQUARE = (1.0, 0.0, 0.0, 1.0, 0.0, 0.0)
-
-
-def _is_unit_square(shorthand):
-    values = map(float, shorthand)
-    pairwise = zip(values, UNIT_SQUARE)
-    return all([isclose(a, b, rel_tol=1e-3) for a, b in pairwise])
-
-
-XobjectSettings = namedtuple('XobjectSettings', ['name', 'shorthand', 'stack_depth'])
-
-InlineSettings = namedtuple('InlineSettings', ['iimage', 'shorthand', 'stack_depth'])
-
-ContentsInfo = namedtuple(
-    'ContentsInfo', ['xobject_settings', 'inline_images', 'found_vector']
-)
-
-TextboxInfo = namedtuple('TextboxInfo', ['bbox', 'is_visible', 'is_corrupt'])
-
-
-class VectorInfo:
-    def __init__(self):
-        pass
-
-
-def _normalize_stack(graphobjs):
-    """Convert runs of qQ's in the stack into single graphobjs"""
-    for operands, operator in graphobjs:
-        operator = str(operator)
-        if re.match(r'Q*q+$', operator):  # Zero or more Q, one or more q
-            for char in operator:  # Split into individual
-                yield ([], char)  # Yield individual
-        else:
-            yield (operands, operator)
-
-
-def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
-    """Interpret the PDF content stream.
-
-    The stack represents the state of the PDF graphics stack.  We are only
-    interested in the current transformation matrix (CTM) so we only track
-    this object; a full implementation would need to track many other items.
-
-    The CTM is initialized to the mapping from user space to device space.
-    PDF units are 1/72".  In a PDF viewer or printer this matrix is initialized
-    to the transformation to device space.  For example if set to
-    (1/72, 0, 0, 1/72, 0, 0) then all units would be calculated in inches.
-
-    Images are always considered to be (0, 0) -> (1, 1).  Before drawing an
-    image there should be a 'cm' that sets up an image coordinate system
-    where drawing from (0, 0) -> (1, 1) will draw on the desired area of the
-    page.
-
-    PDF units suit our needs so we initialize ctm to the identity matrix.
-
-    According to the PDF specification, the maximum stack depth is 32. Other
-    viewers tolerate some amount beyond this.  We issue a warning if the
-    stack depth exceeds the spec limit and set a hard limit beyond this to
-    bound our memory requirements.  If the stack underflows behavior is
-    undefined in the spec, but we just pretend nothing happened and leave the
-    CTM unchanged.
-    """
-
-    stack = []
-    ctm = PdfMatrix(initial_shorthand)
-    xobject_settings = []
-    inline_images = []
-    found_vector = False
-    vector_ops = set('S s f F f* B B* b b*'.split())
-    image_ops = set('BI ID EI q Q Do cm'.split())
-    operator_whitelist = ' '.join(vector_ops | image_ops)
-
-    for n, graphobj in enumerate(
-        _normalize_stack(
-            pikepdf.parse_content_stream(contentstream, operator_whitelist)
-        )
-    ):
-        operands, operator = graphobj
-        if operator == 'q':
-            stack.append(ctm)
-            if len(stack) > 32:  # See docstring
-                if len(stack) > 128:
-                    raise RuntimeError(
-                        "PDF graphics stack overflowed hard limit, operator %i" % n
-                    )
-                warn("PDF graphics stack overflowed spec limit")
-        elif operator == 'Q':
-            try:
-                ctm = stack.pop()
-            except IndexError:
-                # Keeping the ctm the same seems to be the only sensible thing
-                # to do. Just pretend nothing happened, keep calm and carry on.
-                warn("PDF graphics stack underflowed - PDF may be malformed")
-        elif operator == 'cm':
-            ctm = PdfMatrix(operands) @ ctm
-        elif operator == 'Do':
-            image_name = operands[0]
-            settings = XobjectSettings(
-                name=image_name, shorthand=ctm.shorthand, stack_depth=len(stack)
-            )
-            xobject_settings.append(settings)
-        elif operator == 'INLINE IMAGE':  # BI/ID/EI are grouped into this
-            iimage = operands[0]
-            inline = InlineSettings(
-                iimage=iimage, shorthand=ctm.shorthand, stack_depth=len(stack)
-            )
-            inline_images.append(inline)
-        elif operator in vector_ops:
-            found_vector = True
-
-    return ContentsInfo(
-        xobject_settings=xobject_settings,
-        inline_images=inline_images,
-        found_vector=found_vector,
-    )
-
-
-def _get_dpi(ctm_shorthand, image_size):
-    """Given the transformation matrix and image size, find the image DPI.
-
-    PDFs do not include image resolution information within image data.
-    Instead, the PDF page content stream describes the location where the
-    image will be rasterized, and the effective resolution is the ratio of the
-    pixel size to raster target size.
-
-    Normally a scanned PDF has the paper size set appropriately but this is
-    not guaranteed. The most common case is a cropped image will change the
-    page size (/CropBox) without altering the page content stream. That means
-    it is not sufficient to assume that the image fills the page, even though
-    that is the most common case.
-
-    A PDF image may be scaled (always), cropped, translated, rotated in place
-    to an arbitrary angle (rarely) and skewed. Only equal area mappings can
-    be expressed, that is, it is not necessary to consider distortions where
-    the effective DPI varies with position.
-
-    To determine the image scale, transform an offset axis vector v0 (0, 0),
-    width-axis vector v0 (1, 0), height-axis vector vh (0, 1) with the matrix,
-    which gives the dimensions of the image in PDF units. From there we can
-    compare to actual image dimensions. PDF uses
-    row vector * matrix_tranposed unlike the traditional
-    matrix * column vector.
-
-    The offset, width and height vectors can be combined in a matrix and
-    multiplied by the transform matrix. Then we want to calculated
-        magnitude(width_vector - offset_vector)
-    and
-        magnitude(height_vector - offset_vector)
-
-    When the above is worked out algebraically, the effect of translation
-    cancels out, and the vector magnitudes become functions of the nonzero
-    transformation matrix indices. The results of the derivation are used
-    in this code.
-
-    pdfimages -list does calculate the DPI in some way that is not completely
-    naive, but it does not get the DPI of rotated images right, so cannot be
-    used anymore to validate this. Photoshop works, or using Acrobat to
-    rotate the image back to normal.
-
-    It does not matter if the image is partially cropped, or even out of the
-    /MediaBox.
-
-    """
-
-    a, b, c, d, _, _ = ctm_shorthand
-
-    # Calculate the width and height of the image in PDF units
-    image_drawn_width = hypot(a, b)
-    image_drawn_height = hypot(c, d)
-
-    # The scale of the image is pixels per unit of default user space (1/72")
-    scale_w = image_size[0] / image_drawn_width
-    scale_h = image_size[1] / image_drawn_height
-
-    # DPI = scale * 72
-    dpi_w = scale_w * 72.0
-    dpi_h = scale_h * 72.0
-
-    return dpi_w, dpi_h
-
-
-class ImageInfo:
-    DPI_PREC = Decimal('1.000')
-
-    def __init__(self, *, name='', pdfimage=None, inline=None, shorthand=None):
-
-        self._name = str(name)
-        self._shorthand = shorthand
-
-        if inline is not None:
-            self._origin = 'inline'
-            pim = inline.iimage
-        elif pdfimage is not None:
-            self._origin = 'xobject'
-            pim = pikepdf.PdfImage(pdfimage)
-        self._width = pim.width
-        self._height = pim.height
-
-        # If /ImageMask is true, then this image is a stencil mask
-        # (Images that draw with this stencil mask will have a reference to
-        # it in their /Mask, but we don't actually need that information)
-        if pim.image_mask:
-            self._type = 'stencil'
-        else:
-            self._type = 'image'
-
-        self._bpc = int(pim.bits_per_component)
-        try:
-            self._enc = FRIENDLY_ENCODING.get(pim.filters[0], 'image')
-        except IndexError:
-            self._enc = '?'
-
-        try:
-            self._color = FRIENDLY_COLORSPACE.get(pim.colorspace, '?')
-        except NotImplementedError:
-            self._color = '?'
-        if self._enc == Encoding.jpeg2000:
-            self._color = Colorspace.jpeg2000
-
-        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]
-
-    @property
-    def name(self):
-        return self._name
-
-    @property
-    def type_(self):
-        return self._type
-
-    @property
-    def width(self):
-        return self._width
-
-    @property
-    def height(self):
-        return self._height
-
-    @property
-    def bpc(self):
-        return self._bpc
-
-    @property
-    def color(self):
-        return self._color
-
-    @property
-    def comp(self):
-        return self._comp
-
-    @property
-    def enc(self):
-        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 __repr__(self):
-        class_locals = {
-            attr: getattr(self, attr, None)
-            for attr in dir(self)
-            if not attr.startswith('_')
-        }
-        return (
-            "<ImageInfo '{name}' {type_} {width}x{height} {color} "
-            "{comp} {bpc} {enc} {xres}x{yres}>"
-        ).format(**class_locals)
-
-
-def _find_inline_images(contentsinfo):
-    "Find inline images in the contentstream"
-
-    for n, inline in enumerate(contentsinfo.inline_images):
-        yield ImageInfo(
-            name='inline-%02d' % n, shorthand=inline.shorthand, inline=inline
-        )
-
-
-def _image_xobjects(container):
-    """Search for all XObject-based images in the container
-
-    Usually the container is a page, but it could also be a Form XObject
-    that contains images. Filter out the Form XObjects which are dealt with
-    elsewhere.
-
-    Generate a sequence of tuples (image, xobj container), where container,
-    where xobj is the name of the object and image is the object itself,
-    since the object does not know its own name.
-
-    """
-
-    if '/Resources' not in container:
-        return
-    resources = container['/Resources']
-    if '/XObject' not in resources:
-        return
-    xobjs = resources['/XObject'].as_dict()
-    for xobj in xobjs:
-        candidate = xobjs[xobj]
-        if not '/Subtype' in candidate:
-            continue
-        if candidate['/Subtype'] == '/Image':
-            pdfimage = candidate
-            yield (pdfimage, xobj)
-
-
-def _find_regular_images(container, contentsinfo):
-    """Find images stored in the container's /Resources /XObject
-
-    Usually the container is a page, but it could also be a Form XObject
-    that contains images.
-
-    Generates images with their DPI at time of drawing.
-    """
-
-    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 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
-                # gets drawn into a square of 1x1 PDF units (or 1/72",
-                # or 0.35 mm).  The equivalent DPI will be >100,000.  Exclude
-                # these from our DPI calculation for the page.
-                continue
-
-            yield ImageInfo(name=draw.name, pdfimage=pdfimage, shorthand=draw.shorthand)
-
-
-def _find_form_xobject_images(pdf, container, contentsinfo):
-    """Find any images that are in Form XObjects in the container
-
-    The container may be a page, or a parent Form XObject.
-
-    """
-    if '/Resources' not in container:
-        return
-    resources = container['/Resources']
-    if '/XObject' not in resources:
-        return
-    xobjs = resources['/XObject'].as_dict()
-    for xobj in xobjs:
-        candidate = xobjs[xobj]
-        if candidate['/Subtype'] != '/Form':
-            continue
-
-        form_xobject = candidate
-        for settings in contentsinfo.xobject_settings:
-            if settings.name != xobj:
-                continue
-
-            # Find images once for each time this Form XObject is drawn.
-            # This could be optimized to cache the multiple drawing events
-            # but in practice both Form XObjects and multiple drawing of the
-            # same object are both very rare.
-            ctm_shorthand = settings.shorthand
-            yield from _process_content_streams(
-                pdf=pdf, container=form_xobject, shorthand=ctm_shorthand
-            )
-
-
-def _process_content_streams(*, pdf, container, shorthand=None):
-    """Find all individual instances of images drawn in the container
-
-    Usually the container is a page, but it may also be a Form XObject.
-
-    On a typical page images are stored inline or as regular images
-    in an XObject.
-
-    Form XObjects may include inline images, XObject images,
-    and recursively, other Form XObjects; and also vector graphic objects.
-
-    Every instance of an image being drawn somewhere is flattened and
-    treated as a unique image, since if the same image is drawn multiple times
-    on one page it may be drawn at differing resolutions, and our objective
-    is to find the resolution at which the page can be rastered without
-    downsampling.
-
-    """
-
-    if container.get('/Type') == '/Page' and '/Contents' in container:
-        initial_shorthand = shorthand or UNIT_SQUARE
-    elif container.get('/Type') == '/XObject' and container['/Subtype'] == '/Form':
-        # Set the CTM to the state it was when the "Do" operator was
-        # encountered that is drawing this instance of the Form XObject
-        ctm = PdfMatrix(shorthand) if shorthand else PdfMatrix.identity()
-
-        # A Form XObject may provide its own matrix to map form space into
-        # user space. Get this if one exists
-        form_shorthand = container.get('/Matrix', PdfMatrix.identity())
-        form_matrix = PdfMatrix(form_shorthand)
-
-        # Concatenate form matrix with CTM to ensure CTM is correct for
-        # drawing this instance of the XObject
-        ctm = form_matrix @ ctm
-        initial_shorthand = ctm.shorthand
-    else:
-        return
-
-    contentsinfo = _interpret_contents(container, initial_shorthand)
-
-    if contentsinfo.found_vector:
-        yield VectorInfo()
-    yield from _find_inline_images(contentsinfo)
-    yield from _find_regular_images(container, contentsinfo)
-    yield from _find_form_xobject_images(pdf, container, contentsinfo)
-
-
-def _page_has_text(text_blocks, page_width, page_height):
-    """Smarter text detection that ignores text in margins"""
-
-    pw, ph = float(page_width), float(page_height)
-
-    margin_ratio = 0.125
-    interior_bbox = (
-        margin_ratio * pw,  # left
-        (1 - margin_ratio) * ph,  # top
-        (1 - margin_ratio) * pw,  # right
-        margin_ratio * ph,  # bottom  (first quadrant: bottom < top)
-    )
-
-    def rects_intersect(a, b):
-        """
-        Where (a,b) are 4-tuple rects (left-0, top-1, right-2, bottom-3)
-        https://stackoverflow.com/questions/306316/determine-if-two-rectangles-overlap-each-other
-        Formula assumes all boxes are in first quadrant
-        """
-        return a[0] < b[2] and a[2] > b[0] and a[1] > b[3] and a[3] < b[1]
-
-    has_text = False
-    for bbox in text_blocks:
-        if rects_intersect(bbox, interior_bbox):
-            has_text = True
-            break
-    return has_text
-
-
-def simplify_textboxes(miner, textbox_getter):
-    """Extract only limited content from text boxes
-
-    We do this to save memory and ensure that our objects are pickleable.
-    """
-    for box in textbox_getter(miner):
-        first_line = box._objs[0]
-        first_char = first_line._objs[0]
-
-        visible = first_char.rendermode != 3
-        corrupt = first_char.get_text() == '\ufffd'
-        yield TextboxInfo(box.bbox, visible, corrupt)
-
-
-def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
-    pageinfo = {}
-    pageinfo['pageno'] = pageno
-    pageinfo['images'] = []
-
-    page = pdf.pages[pageno]
-    mediabox = [Decimal(d) for d in page.MediaBox.as_list()]
-    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."
-            )
-        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)
-
-    userunit = page.get('/UserUnit', Decimal(1.0))
-    if not isinstance(userunit, Decimal):
-        userunit = Decimal(userunit)
-    pageinfo['userunit'] = userunit
-    pageinfo['width_inches'] = width_pt * userunit / Decimal(72.0)
-    pageinfo['height_inches'] = height_pt * userunit / Decimal(72.0)
-
-    try:
-        pageinfo['rotate'] = int(page['/Rotate'])
-    except KeyError:
-        pageinfo['rotate'] = 0
-
-    userunit_shorthand = (userunit, 0, 0, userunit, 0, 0)
-    contentsinfo = [
-        ci
-        for ci in _process_content_streams(
-            pdf=pdf, container=page, shorthand=userunit_shorthand
-        )
-    ]
-
-    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']))
-
-    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)
-
-    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
-
-
-class PageInfo:
-    def __init__(self, pdf, pageno, infile, xmltext, detailed_analysis=False):
-        self._pageno = pageno
-        self._infile = infile
-        self._pageinfo = _pdf_get_pageinfo(pdf, pageno, infile, xmltext)
-        self._detailed_analysis = detailed_analysis
-
-    @property
-    def pageno(self):
-        return self._pageno
-
-    @property
-    def has_text(self):
-        return self._pageinfo['has_text']
-
-    @property
-    def has_corrupt_text(self):
-        if not self._detailed_analysis:
-            raise NotImplementedError('Did not do detailed analysis')
-        return any(tbox.is_corrupt for tbox in self._pageinfo['textboxes'])
-
-    @property
-    def has_vector(self):
-        return self._pageinfo['has_vector']
-
-    @property
-    def width_inches(self):
-        return self._pageinfo['width_inches']
-
-    @property
-    def height_inches(self):
-        return self._pageinfo['height_inches']
-
-    @property
-    def width_pixels(self):
-        return int(round(self.width_inches * self.xres))
-
-    @property
-    def height_pixels(self):
-        return int(round(self.height_inches * self.yres))
-
-    @property
-    def rotation(self):
-        return self._pageinfo.get('rotate', None)
-
-    @rotation.setter
-    def rotation(self, value):
-        if value in (0, 90, 180, 270, 360, -90, -180, -270):
-            self._pageinfo['rotate'] = value
-        else:
-            raise ValueError("rotation must be a cardinal angle")
-
-    @property
-    def images(self):
-        return self._pageinfo['images']
-
-    def get_textareas(self, visible=None, corrupt=None):
-        def predicate(obj, want_visible, want_corrupt):
-            result = True
-            if want_visible is not None:
-                if obj.is_visible != want_visible:
-                    result = False
-            if want_corrupt is not None:
-                if obj.is_corrupt != want_corrupt:
-                    result = False
-            return result
-
-        if 'textboxes' not in self._pageinfo:
-            if visible is not None and corrupt is not None:
-                raise NotImplementedError('Ghostscript textboxes cannot be classified')
-            return self._pageinfo['bboxes']
-
-        return (
-            obj.bbox
-            for obj in self._pageinfo['textboxes']
-            if predicate(obj, visible, corrupt)
-        )
-
-    @property
-    def xres(self):
-        return self._pageinfo.get('xres', None)
-
-    @property
-    def yres(self):
-        return self._pageinfo.get('yres', None)
-
-    @property
-    def userunit(self):
-        return self._pageinfo.get('userunit', None)
-
-    @property
-    def min_version(self):
-        if self.userunit is not None:
-            return '1.6'
-        else:
-            return '1.5'
-
-    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,
-        )
-
-
-class PdfInfo:
-    """Get summary information about a PDF"""
-
-    def __init__(self, infile, detailed_page_analysis=False, log=logger, progbar=False):
-        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()
-
-    @property
-    def pages(self):
-        return self._pages
-
-    @property
-    def min_version(self):
-        # The minimum PDF is the maximum version that any particular page needs
-        return max(page.min_version for page in self.pages)
-
-    @property
-    def has_userunit(self):
-        return any(page.userunit != 1.0 for page in self.pages)
-
-    @property
-    def has_acroform(self):
-        return self._has_acroform
-
-    @property
-    def filename(self):
-        if not isinstance(self._infile, (str, Path)):
-            raise NotImplementedError("can't get filename from stream")
-        return self._infile
-
-    @property
-    def needs_rendering(self):
-        return self._needs_rendering
-
-    def __getitem__(self, item):
-        return self._pages[item]
-
-    def __len__(self):
-        return len(self._pages)
-
-    def __repr__(self):
-        return f"<PdfInfo('...'), page count={len(self)}>"
-
-
-def main():
-    import argparse
-
-    parser = argparse.ArgumentParser()
-    parser.add_argument('infile')
-    args = parser.parse_args()
-    info = _pdf_get_all_pageinfo(args.infile)
-    from pprint import pprint
-
-    pprint(info)
-
-
-if __name__ == '__main__':
-    main()
+from .info import Colorspace, Encoding, PdfInfo

src/ocrmypdf/pdfinfo/ghosttext.py

@@ -96,6 +96,7 @@ def extract_text_xml(infile, pdf, pageno=None, log=gslog):
     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.")
+        log.error(f"Expected {len(pdf.pages)}, txtwrite says {len(page_xml)}")
         if page_count_difference > 0:
             page_xml.extend([None] * page_count_difference)
     return page_xml

src/ocrmypdf/pdfinfo/info.py

@@ -0,0 +1,831 @@
+#!/usr/bin/env python3
+# © 2015 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 collections import defaultdict, namedtuple
+from decimal import Decimal
+from enum import Enum
+from math import hypot, isclose
+from os import PathLike, fspath
+from pathlib import Path
+from warnings import warn
+
+import pikepdf
+from pikepdf import PdfMatrix
+from tqdm import tqdm
+
+from ocrmypdf.exceptions import EncryptedPdfError
+from ocrmypdf.exec import ghostscript
+from ocrmypdf.pdfinfo import ghosttext
+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'
+)
+
+FRIENDLY_COLORSPACE = {
+    '/DeviceGray': Colorspace.gray,
+    '/CalGray': Colorspace.gray,
+    '/DeviceRGB': Colorspace.rgb,
+    '/CalRGB': Colorspace.rgb,
+    '/DeviceCMYK': Colorspace.cmyk,
+    '/Lab': Colorspace.lab,
+    '/ICCBased': Colorspace.icc,
+    '/Indexed': Colorspace.index,
+    '/Separation': Colorspace.sep,
+    '/DeviceN': Colorspace.devn,
+    '/Pattern': Colorspace.pattern,
+    '/G': Colorspace.gray,  # Abbreviations permitted in inline images
+    '/RGB': Colorspace.rgb,
+    '/CMYK': Colorspace.cmyk,
+    '/I': Colorspace.index,
+}
+
+FRIENDLY_ENCODING = {
+    '/CCITTFaxDecode': Encoding.ccitt,
+    '/DCTDecode': Encoding.jpeg,
+    '/JPXDecode': Encoding.jpeg2000,
+    '/JBIG2Decode': Encoding.jbig2,
+    '/CCF': Encoding.ccitt,  # Abbreviations permitted in inline images
+    '/DCT': Encoding.jpeg,
+    '/AHx': Encoding.asciihex,
+    '/A85': Encoding.ascii85,
+    '/LZW': Encoding.lzw,
+    '/Fl': Encoding.flate,
+    '/RL': Encoding.runlength,
+}
+
+FRIENDLY_COMP = {
+    Colorspace.gray: 1,
+    Colorspace.rgb: 3,
+    Colorspace.cmyk: 4,
+    Colorspace.lab: 3,
+    Colorspace.index: 1,
+}
+
+
+UNIT_SQUARE = (1.0, 0.0, 0.0, 1.0, 0.0, 0.0)
+
+
+def _is_unit_square(shorthand):
+    values = map(float, shorthand)
+    pairwise = zip(values, UNIT_SQUARE)
+    return all([isclose(a, b, rel_tol=1e-3) for a, b in pairwise])
+
+
+XobjectSettings = namedtuple('XobjectSettings', ['name', 'shorthand', 'stack_depth'])
+
+InlineSettings = namedtuple('InlineSettings', ['iimage', 'shorthand', 'stack_depth'])
+
+ContentsInfo = namedtuple(
+    'ContentsInfo', ['xobject_settings', 'inline_images', 'found_vector', 'name_index']
+)
+
+TextboxInfo = namedtuple('TextboxInfo', ['bbox', 'is_visible', 'is_corrupt'])
+
+
+class VectorInfo:
+    def __init__(self):
+        pass
+
+
+def _normalize_stack(graphobjs):
+    """Convert runs of qQ's in the stack into single graphobjs"""
+    for operands, operator in graphobjs:
+        operator = str(operator)
+        if re.match(r'Q*q+$', operator):  # Zero or more Q, one or more q
+            for char in operator:  # Split into individual
+                yield ([], char)  # Yield individual
+        else:
+            yield (operands, operator)
+
+
+def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
+    """Interpret the PDF content stream.
+
+    The stack represents the state of the PDF graphics stack.  We are only
+    interested in the current transformation matrix (CTM) so we only track
+    this object; a full implementation would need to track many other items.
+
+    The CTM is initialized to the mapping from user space to device space.
+    PDF units are 1/72".  In a PDF viewer or printer this matrix is initialized
+    to the transformation to device space.  For example if set to
+    (1/72, 0, 0, 1/72, 0, 0) then all units would be calculated in inches.
+
+    Images are always considered to be (0, 0) -> (1, 1).  Before drawing an
+    image there should be a 'cm' that sets up an image coordinate system
+    where drawing from (0, 0) -> (1, 1) will draw on the desired area of the
+    page.
+
+    PDF units suit our needs so we initialize ctm to the identity matrix.
+
+    According to the PDF specification, the maximum stack depth is 32. Other
+    viewers tolerate some amount beyond this.  We issue a warning if the
+    stack depth exceeds the spec limit and set a hard limit beyond this to
+    bound our memory requirements.  If the stack underflows behavior is
+    undefined in the spec, but we just pretend nothing happened and leave the
+    CTM unchanged.
+    """
+
+    stack = []
+    ctm = PdfMatrix(initial_shorthand)
+    xobject_settings = []
+    inline_images = []
+    name_index = defaultdict(lambda: [])
+    found_vector = False
+    vector_ops = set('S s f F f* B B* b b*'.split())
+    image_ops = set('BI ID EI q Q Do cm'.split())
+    operator_whitelist = ' '.join(vector_ops | image_ops)
+
+    for n, graphobj in enumerate(
+        _normalize_stack(
+            pikepdf.parse_content_stream(contentstream, operator_whitelist)
+        )
+    ):
+        operands, operator = graphobj
+        if operator == 'q':
+            stack.append(ctm)
+            if len(stack) > 32:  # See docstring
+                if len(stack) > 128:
+                    raise RuntimeError(
+                        "PDF graphics stack overflowed hard limit, operator %i" % n
+                    )
+                warn("PDF graphics stack overflowed spec limit")
+        elif operator == 'Q':
+            try:
+                ctm = stack.pop()
+            except IndexError:
+                # Keeping the ctm the same seems to be the only sensible thing
+                # to do. Just pretend nothing happened, keep calm and carry on.
+                warn("PDF graphics stack underflowed - PDF may be malformed")
+        elif operator == 'cm':
+            ctm = PdfMatrix(operands) @ ctm
+        elif operator == 'Do':
+            image_name = operands[0]
+            settings = XobjectSettings(
+                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(
+                iimage=iimage, shorthand=ctm.shorthand, stack_depth=len(stack)
+            )
+            inline_images.append(inline)
+        elif operator in vector_ops:
+            found_vector = True
+
+    return ContentsInfo(
+        xobject_settings=xobject_settings,
+        inline_images=inline_images,
+        found_vector=found_vector,
+        name_index=name_index,
+    )
+
+
+def _get_dpi(ctm_shorthand, image_size):
+    """Given the transformation matrix and image size, find the image DPI.
+
+    PDFs do not include image resolution information within image data.
+    Instead, the PDF page content stream describes the location where the
+    image will be rasterized, and the effective resolution is the ratio of the
+    pixel size to raster target size.
+
+    Normally a scanned PDF has the paper size set appropriately but this is
+    not guaranteed. The most common case is a cropped image will change the
+    page size (/CropBox) without altering the page content stream. That means
+    it is not sufficient to assume that the image fills the page, even though
+    that is the most common case.
+
+    A PDF image may be scaled (always), cropped, translated, rotated in place
+    to an arbitrary angle (rarely) and skewed. Only equal area mappings can
+    be expressed, that is, it is not necessary to consider distortions where
+    the effective DPI varies with position.
+
+    To determine the image scale, transform an offset axis vector v0 (0, 0),
+    width-axis vector v0 (1, 0), height-axis vector vh (0, 1) with the matrix,
+    which gives the dimensions of the image in PDF units. From there we can
+    compare to actual image dimensions. PDF uses
+    row vector * matrix_tranposed unlike the traditional
+    matrix * column vector.
+
+    The offset, width and height vectors can be combined in a matrix and
+    multiplied by the transform matrix. Then we want to calculated
+        magnitude(width_vector - offset_vector)
+    and
+        magnitude(height_vector - offset_vector)
+
+    When the above is worked out algebraically, the effect of translation
+    cancels out, and the vector magnitudes become functions of the nonzero
+    transformation matrix indices. The results of the derivation are used
+    in this code.
+
+    pdfimages -list does calculate the DPI in some way that is not completely
+    naive, but it does not get the DPI of rotated images right, so cannot be
+    used anymore to validate this. Photoshop works, or using Acrobat to
+    rotate the image back to normal.
+
+    It does not matter if the image is partially cropped, or even out of the
+    /MediaBox.
+
+    """
+
+    a, b, c, d, _, _ = ctm_shorthand
+
+    # Calculate the width and height of the image in PDF units
+    image_drawn_width = hypot(a, b)
+    image_drawn_height = hypot(c, d)
+
+    # The scale of the image is pixels per unit of default user space (1/72")
+    scale_w = image_size[0] / image_drawn_width
+    scale_h = image_size[1] / image_drawn_height
+
+    # DPI = scale * 72
+    dpi_w = scale_w * 72.0
+    dpi_h = scale_h * 72.0
+
+    return dpi_w, dpi_h
+
+
+class ImageInfo:
+    DPI_PREC = Decimal('1.000')
+
+    def __init__(self, *, name='', pdfimage=None, inline=None, shorthand=None):
+
+        self._name = str(name)
+        self._shorthand = shorthand
+
+        if inline is not None:
+            self._origin = 'inline'
+            pim = inline.iimage
+        elif pdfimage is not None:
+            self._origin = 'xobject'
+            pim = pikepdf.PdfImage(pdfimage)
+        self._width = pim.width
+        self._height = pim.height
+
+        # If /ImageMask is true, then this image is a stencil mask
+        # (Images that draw with this stencil mask will have a reference to
+        # it in their /Mask, but we don't actually need that information)
+        if pim.image_mask:
+            self._type = 'stencil'
+        else:
+            self._type = 'image'
+
+        self._bpc = int(pim.bits_per_component)
+        try:
+            self._enc = FRIENDLY_ENCODING.get(pim.filters[0], 'image')
+        except IndexError:
+            self._enc = '?'
+
+        try:
+            self._color = FRIENDLY_COLORSPACE.get(pim.colorspace, '?')
+        except NotImplementedError:
+            self._color = '?'
+        if self._enc == Encoding.jpeg2000:
+            self._color = Colorspace.jpeg2000
+
+        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 only supports monochrome.
+            if self._comp == '?' and self._enc in (Encoding.ccitt, Encoding.jbig2):
+                self._comp = FRIENDLY_COMP[Colorspace.gray]
+
+    @property
+    def name(self):
+        return self._name
+
+    @property
+    def type_(self):
+        return self._type
+
+    @property
+    def width(self):
+        return self._width
+
+    @property
+    def height(self):
+        return self._height
+
+    @property
+    def bpc(self):
+        return self._bpc
+
+    @property
+    def color(self):
+        return self._color
+
+    @property
+    def comp(self):
+        return self._comp
+
+    @property
+    def enc(self):
+        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 __repr__(self):
+        class_locals = {
+            attr: getattr(self, attr, None)
+            for attr in dir(self)
+            if not attr.startswith('_')
+        }
+        return (
+            "<ImageInfo '{name}' {type_} {width}x{height} {color} "
+            "{comp} {bpc} {enc} {xres}x{yres}>"
+        ).format(**class_locals)
+
+
+def _find_inline_images(contentsinfo):
+    "Find inline images in the contentstream"
+
+    for n, inline in enumerate(contentsinfo.inline_images):
+        yield ImageInfo(
+            name='inline-%02d' % n, shorthand=inline.shorthand, inline=inline
+        )
+
+
+def _image_xobjects(container):
+    """Search for all XObject-based images in the container
+
+    Usually the container is a page, but it could also be a Form XObject
+    that contains images. Filter out the Form XObjects which are dealt with
+    elsewhere.
+
+    Generate a sequence of tuples (image, xobj container), where container,
+    where xobj is the name of the object and image is the object itself,
+    since the object does not know its own name.
+
+    """
+
+    if '/Resources' not in container:
+        return
+    resources = container['/Resources']
+    if '/XObject' not in resources:
+        return
+    xobjs = resources['/XObject'].as_dict()
+    for xobj in xobjs:
+        candidate = xobjs[xobj]
+        if not '/Subtype' in candidate:
+            continue
+        if candidate['/Subtype'] == '/Image':
+            pdfimage = candidate
+            yield (pdfimage, xobj)
+
+
+def _find_regular_images(container, contentsinfo):
+    """Find images stored in the container's /Resources /XObject
+
+    Usually the container is a page, but it could also be a Form XObject
+    that contains images.
+
+    Generates images with their DPI at time of drawing.
+    """
+
+    for pdfimage, xobj in _image_xobjects(container):
+        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
+                # gets drawn into a square of 1x1 PDF units (or 1/72",
+                # or 0.35 mm).  The equivalent DPI will be >100,000.  Exclude
+                # these from our DPI calculation for the page.
+                continue
+
+            yield ImageInfo(name=draw.name, pdfimage=pdfimage, shorthand=draw.shorthand)
+
+
+def _find_form_xobject_images(pdf, container, contentsinfo):
+    """Find any images that are in Form XObjects in the container
+
+    The container may be a page, or a parent Form XObject.
+
+    """
+    if '/Resources' not in container:
+        return
+    resources = container['/Resources']
+    if '/XObject' not in resources:
+        return
+    xobjs = resources['/XObject'].as_dict()
+    for xobj in xobjs:
+        candidate = xobjs[xobj]
+        if candidate['/Subtype'] != '/Form':
+            continue
+
+        form_xobject = candidate
+        for settings in contentsinfo.xobject_settings:
+            if settings.name != xobj:
+                continue
+
+            # Find images once for each time this Form XObject is drawn.
+            # This could be optimized to cache the multiple drawing events
+            # but in practice both Form XObjects and multiple drawing of the
+            # same object are both very rare.
+            ctm_shorthand = settings.shorthand
+            yield from _process_content_streams(
+                pdf=pdf, container=form_xobject, shorthand=ctm_shorthand
+            )
+
+
+def _process_content_streams(*, pdf, container, shorthand=None):
+    """Find all individual instances of images drawn in the container
+
+    Usually the container is a page, but it may also be a Form XObject.
+
+    On a typical page images are stored inline or as regular images
+    in an XObject.
+
+    Form XObjects may include inline images, XObject images,
+    and recursively, other Form XObjects; and also vector graphic objects.
+
+    Every instance of an image being drawn somewhere is flattened and
+    treated as a unique image, since if the same image is drawn multiple times
+    on one page it may be drawn at differing resolutions, and our objective
+    is to find the resolution at which the page can be rastered without
+    downsampling.
+
+    """
+
+    if container.get('/Type') == '/Page' and '/Contents' in container:
+        initial_shorthand = shorthand or UNIT_SQUARE
+    elif container.get('/Type') == '/XObject' and container['/Subtype'] == '/Form':
+        # Set the CTM to the state it was when the "Do" operator was
+        # encountered that is drawing this instance of the Form XObject
+        ctm = PdfMatrix(shorthand) if shorthand else PdfMatrix.identity()
+
+        # A Form XObject may provide its own matrix to map form space into
+        # user space. Get this if one exists
+        form_shorthand = container.get('/Matrix', PdfMatrix.identity())
+        form_matrix = PdfMatrix(form_shorthand)
+
+        # Concatenate form matrix with CTM to ensure CTM is correct for
+        # drawing this instance of the XObject
+        ctm = form_matrix @ ctm
+        initial_shorthand = ctm.shorthand
+    else:
+        return
+
+    contentsinfo = _interpret_contents(container, initial_shorthand)
+
+    if contentsinfo.found_vector:
+        yield VectorInfo()
+    yield from _find_inline_images(contentsinfo)
+    yield from _find_regular_images(container, contentsinfo)
+    yield from _find_form_xobject_images(pdf, container, contentsinfo)
+
+
+def _page_has_text(text_blocks, page_width, page_height):
+    """Smarter text detection that ignores text in margins"""
+
+    pw, ph = float(page_width), float(page_height)
+
+    margin_ratio = 0.125
+    interior_bbox = (
+        margin_ratio * pw,  # left
+        (1 - margin_ratio) * ph,  # top
+        (1 - margin_ratio) * pw,  # right
+        margin_ratio * ph,  # bottom  (first quadrant: bottom < top)
+    )
+
+    def rects_intersect(a, b):
+        """
+        Where (a,b) are 4-tuple rects (left-0, top-1, right-2, bottom-3)
+        https://stackoverflow.com/questions/306316/determine-if-two-rectangles-overlap-each-other
+        Formula assumes all boxes are in first quadrant
+        """
+        return a[0] < b[2] and a[2] > b[0] and a[1] > b[3] and a[3] < b[1]
+
+    has_text = False
+    for bbox in text_blocks:
+        if rects_intersect(bbox, interior_bbox):
+            has_text = True
+            break
+    return has_text
+
+
+def simplify_textboxes(miner, textbox_getter):
+    """Extract only limited content from text boxes
+
+    We do this to save memory and ensure that our objects are pickleable.
+    """
+    for box in textbox_getter(miner):
+        first_line = box._objs[0]
+        first_char = first_line._objs[0]
+
+        visible = first_char.rendermode != 3
+        corrupt = first_char.get_text() == '\ufffd'
+        yield TextboxInfo(box.bbox, visible, corrupt)
+
+
+def _pdf_get_pageinfo(pdf, pageno: int, infile: PathLike, xmltext: str):
+    pageinfo = {}
+    pageinfo['pageno'] = pageno
+    pageinfo['images'] = []
+
+    page = pdf.pages[pageno]
+    mediabox = [Decimal(d) for d in page.MediaBox.as_list()]
+    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:
+        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)
+
+    userunit = page.get('/UserUnit', Decimal(1.0))
+    if not isinstance(userunit, Decimal):
+        userunit = Decimal(userunit)
+    pageinfo['userunit'] = userunit
+    pageinfo['width_inches'] = width_pt * userunit / Decimal(72.0)
+    pageinfo['height_inches'] = height_pt * userunit / Decimal(72.0)
+
+    try:
+        pageinfo['rotate'] = int(page['/Rotate'])
+    except KeyError:
+        pageinfo['rotate'] = 0
+
+    userunit_shorthand = (userunit, 0, 0, userunit, 0, 0)
+    contentsinfo = [
+        ci
+        for ci in _process_content_streams(
+            pdf=pdf, container=page, shorthand=userunit_shorthand
+        )
+    ]
+
+    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']))
+
+    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
+    try:
+        if pdf.is_encrypted:
+            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)
+
+        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)
+    except Exception:
+        pdf.close()
+        raise
+
+    return pages, pdf
+
+
+class PageInfo:
+    def __init__(self, pdf, pageno, infile, xmltext, detailed_analysis=False):
+        self._pageno = pageno
+        self._infile = infile
+        self._pageinfo = _pdf_get_pageinfo(pdf, pageno, infile, xmltext)
+        self._detailed_analysis = detailed_analysis
+
+    @property
+    def pageno(self):
+        return self._pageno
+
+    @property
+    def has_text(self):
+        return self._pageinfo['has_text']
+
+    @property
+    def has_corrupt_text(self):
+        if not self._detailed_analysis:
+            raise NotImplementedError('Did not do detailed analysis')
+        return any(tbox.is_corrupt for tbox in self._pageinfo['textboxes'])
+
+    @property
+    def has_vector(self):
+        return self._pageinfo['has_vector']
+
+    @property
+    def width_inches(self):
+        return self._pageinfo['width_inches']
+
+    @property
+    def height_inches(self):
+        return self._pageinfo['height_inches']
+
+    @property
+    def width_pixels(self):
+        return int(round(self.width_inches * self.xres))
+
+    @property
+    def height_pixels(self):
+        return int(round(self.height_inches * self.yres))
+
+    @property
+    def rotation(self):
+        return self._pageinfo.get('rotate', None)
+
+    @rotation.setter
+    def rotation(self, value):
+        if value in (0, 90, 180, 270, 360, -90, -180, -270):
+            self._pageinfo['rotate'] = value
+        else:
+            raise ValueError("rotation must be a cardinal angle")
+
+    @property
+    def images(self):
+        return self._pageinfo['images']
+
+    def get_textareas(self, visible=None, corrupt=None):
+        def predicate(obj, want_visible, want_corrupt):
+            result = True
+            if want_visible is not None:
+                if obj.is_visible != want_visible:
+                    result = False
+            if want_corrupt is not None:
+                if obj.is_corrupt != want_corrupt:
+                    result = False
+            return result
+
+        if 'textboxes' not in self._pageinfo:
+            if visible is not None and corrupt is not None:
+                raise NotImplementedError('Ghostscript textboxes cannot be classified')
+            return self._pageinfo['bboxes']
+
+        return (
+            obj.bbox
+            for obj in self._pageinfo['textboxes']
+            if predicate(obj, visible, corrupt)
+        )
+
+    @property
+    def xres(self):
+        return self._pageinfo.get('xres', None)
+
+    @property
+    def yres(self):
+        return self._pageinfo.get('yres', None)
+
+    @property
+    def userunit(self):
+        return self._pageinfo.get('userunit', None)
+
+    @property
+    def min_version(self):
+        if self.userunit is not None:
+            return '1.6'
+        else:
+            return '1.5'
+
+    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,
+        )
+
+
+class PdfInfo:
+    """Get summary information about a PDF"""
+
+    def __init__(self, infile, detailed_page_analysis=False, log=logger, progbar=False):
+        self._infile = infile
+        if ghostscript.version() in ('9.52',):
+            detailed_page_analysis = True  # txtwrite doesn't work in these versions
+        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()
+
+    @property
+    def pages(self):
+        return self._pages
+
+    @property
+    def min_version(self):
+        # The minimum PDF is the maximum version that any particular page needs
+        return max(page.min_version for page in self.pages)
+
+    @property
+    def has_userunit(self):
+        return any(page.userunit != 1.0 for page in self.pages)
+
+    @property
+    def has_acroform(self):
+        return self._has_acroform
+
+    @property
+    def filename(self):
+        if not isinstance(self._infile, (str, Path)):
+            raise NotImplementedError("can't get filename from stream")
+        return self._infile
+
+    @property
+    def needs_rendering(self):
+        return self._needs_rendering
+
+    def __getitem__(self, item):
+        return self._pages[item]
+
+    def __len__(self):
+        return len(self._pages)
+
+    def __repr__(self):
+        return f"<PdfInfo('...'), page count={len(self)}>"
+
+
+def main():
+    import argparse
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument('infile')
+    args = parser.parse_args()
+    pagesinfo, pdfinfo = _pdf_get_all_pageinfo(args.infile)
+    from pprint import pprint
+
+    pprint(pdfinfo)
+    for page in pagesinfo:
+        pprint(page)
+        for im in page.images:
+            pprint(im)
+
+
+if __name__ == '__main__':
+    main()

src/ocrmypdf/pdfinfo/layout.py

@@ -20,6 +20,7 @@ from math import copysign
 from pathlib import Path
 from unittest.mock import patch
 
+import pdfminer
 import pdfminer.encodingdb
 import pdfminer.pdfdevice
 import pdfminer.pdfinterp
@@ -36,51 +37,54 @@ from ..exceptions import EncryptedPdfError
 STRIP_NAME = re.compile(r'[0-9]+')
 
 #
-# Unconditional pdfminer patches
+#  pdfminer 20181108 patches
 #
 
+if pdfminer.__version__ == '20181108':
 
-def name2unicode(name):
-    """Fix pdfminer's name2unicode function
+    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
+        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)
-    m = STRIP_NAME.search(name)
-    if not m:
-        raise KeyError(name)
-    return chr(int(m.group(0)))
+        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
 
-pdfminer.encodingdb.name2unicode = name2unicode
+    original_PDFFont_init = PDFFont.__init__
 
-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__
 
-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
+#
+# end of pdfminer 20181108 patches
+#
 
 
-PDFFont.__init__ = PDFFont__init__
-
 original_PDFSimpleFont_init = PDFSimpleFont.__init__
 
 
@@ -97,6 +101,7 @@ def PDFSimpleFont__init__(self, descriptor, widths, spec):
 
 
 PDFSimpleFont.__init__ = PDFSimpleFont__init__
+
 #
 # pdfminer patches when creator is PScript5.dll
 #
@@ -207,6 +212,7 @@ 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)

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

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.
+

tests/cache/graph_ocred/__-l__deu__000001_ocr.png__000001_ocr_tess__pdf__txt/stderr.bin

@@ -0,0 +1 @@
+Tesseract Open Source OCR Engine v4.0.0 with Leptonica

tests/cache/graph_ocred/__-l__deu__000001_ocr.png__000001_ocr_tess__pdf__txt/txt.bin

@@ -0,0 +1,13 @@
+Portez ce vieux whisky au juge
+blond qui fume sur son Ile
+interieure, a cöte de l'alcöve
+ovoide, oU les büches se
+consument dans l'ätre, ce qui
+lui permet de penser & la
+caenogenese de |'etre dont il
+est question dans la cause
+ambigu& entendue a MoY, dans
+un capharnaüm qui, pense-t-il,
+diminue ca et la la qualite de son
+ceuvre.
+

tests/cache/manifest.jsonl

@@ -65,3 +65,7 @@
 {"tesseract_version": "tesseract 4.0.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.0.10 : zlib 1.2.11 : libwebp 1.0.2 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.5.0-x86_64-i386-64bit", "python": "3.7.3", "argv_slug": "__-l__eng__000001_ocr.png__000001_ocr_tess__pdf__txt", "sourcefile": "resources/poster.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000001_ocr.png", "$TMPDIR/000001_ocr_tess", "pdf", "txt"]}
 {"tesseract_version": "tesseract 4.0.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.0.10 : zlib 1.2.11 : libwebp 1.0.2 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.6.0-x86_64-i386-64bit", "python": "3.7.3", "argv_slug": "__-l__deu__000001_ocr.png__000001_ocr_tess__pdf__txt", "sourcefile": "resources/cmyk.pdf", "args": ["-l", "deu", "-c", "textonly_pdf=1", "$TMPDIR/000001_ocr.png", "$TMPDIR/000001_ocr_tess", "pdf", "txt"]}
 {"tesseract_version": "tesseract 4.0.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.0.10 : zlib 1.2.11 : libwebp 1.0.2 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.6.0-x86_64-i386-64bit", "python": "3.7.3", "argv_slug": "__-l__deu__000001_ocr.png__000001_ocr_tess__pdf__txt", "sourcefile": "resources/lichtenstein.pdf", "args": ["-l", "deu", "-c", "textonly_pdf=1", "$TMPDIR/000001_ocr.png", "$TMPDIR/000001_ocr_tess", "pdf", "txt"]}
+{"tesseract_version": "tesseract 4.0.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.0.10 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.6.0-x86_64-i386-64bit", "python": "3.7.4", "argv_slug": "__-l__deu__000001_ocr.png__000001_ocr_tess__pdf__txt", "sourcefile": "resources/graph_ocred.pdf", "args": ["-l", "deu", "-c", "textonly_pdf=1", "$TMPDIR/000001_ocr.png", "$TMPDIR/000001_ocr_tess", "pdf", "txt"]}
+{"tesseract_version": "tesseract 4.1.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.5", "argv_slug": "__-l__eng__000002_ocr.png__000002_ocr_tess__pdf__txt", "sourcefile": "resources/3small.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000002_ocr.png", "$TMPDIR/000002_ocr_tess", "pdf", "txt"]}
+{"tesseract_version": "tesseract 4.1.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.5", "argv_slug": "__-l__eng__000001_ocr.png__000001_ocr_tess__pdf__txt", "sourcefile": "resources/3small.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000001_ocr.png", "$TMPDIR/000001_ocr_tess", "pdf", "txt"]}
+{"tesseract_version": "tesseract 4.1.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.5", "argv_slug": "__-l__eng__000003_ocr.png__000003_ocr_tess__pdf__txt", "sourcefile": "resources/3small.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000003_ocr.png", "$TMPDIR/000003_ocr_tess", "pdf", "txt"]}

tests/conftest.py

@@ -15,24 +15,19 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
+import ast
 import os
 import platform
 import sys
-from contextlib import contextmanager
 from pathlib import Path
 from subprocess import PIPE, run
-from ocrmypdf import api, cli
 
 import pytest
 
+from ocrmypdf import api, cli
+
 pytest_plugins = ['helpers_namespace']
 
-try:
-    from pytest_cov.embed import cleanup_on_sigterm
-except ImportError:
-    pass
-else:
-    cleanup_on_sigterm()
 
 # pylint: disable=E1101
 # pytest.helpers is dynamic so it confuses pylint
@@ -64,16 +59,6 @@ def running_in_travis():
     return os.environ.get('TRAVIS') == 'true'
 
 
-@pytest.helpers.register
-def needs_pdfminer(fn):
-    try:
-        import pdfminer
-    except ImportError:
-        skip = pytest.mark.skipif(True, reason="pdfminer not available")
-        return skip(fn)
-    return fn
-
-
 @pytest.helpers.register
 def have_unpaper():
     try:
@@ -91,14 +76,69 @@ PROJECT_ROOT = os.path.dirname(TESTS_ROOT)
 OCRMYPDF = [sys.executable, '-m', 'ocrmypdf']
 
 
+WINDOWS_SHIM_TEMPLATE = """
+# This is a shim for Windows that has the same effect as a symlink to the target .py
+# file
+import os
+import subprocess
+import sys
+
+args = [sys.executable, {spoofer}, *sys.argv[1:]]
+p = subprocess.run(args, check=False, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+sys.stdout.buffer.write(p.stdout)
+sys.stderr.buffer.write(p.stderr)
+sys.exit(p.returncode)
+"""
+
+assert ast.parse(WINDOWS_SHIM_TEMPLATE.format(spoofer=repr(r"C:\\Temp\\file.py")))
+
+
 @pytest.helpers.register
 def spoof(tmp_path_factory, **kwargs):
     """Modify PATH to override subprocess executables
 
-    spoof(program1='replacement', ...)
+    spoof(tmp_path_factory, program1='replacement', ...)
 
-    Creates temporary directory with symlinks to targets.
+    For the test suite we need a way override executables, so that we can
+    substitute desired results such as errors or just speed up OCR.
 
+    On POSIXish platforms we create a temporary folder with overrides that
+    are symlinks to the executables we want to run. We do not actually override
+    PATH. We also set an environment variable _OCRMYPDF_TEST_PATH, which
+    OCRmyPDF's subprocess wrapper will check before they use regular PATH. The
+    output is a folder full of executables we are overriding. We can override
+    multiple executables. The end result is a folder we can use in a PATH-style
+    lookup to override some executables:
+
+        /tmp/abcxyz/tesseract -> ocrmypdf/tests/resources/spoof/tesseract_crash.py
+        /tmp/abcxyz/gs -> ocrmypdf/tests/resources/spoof/gs_backflip.py
+
+    Windows needs extra help from us because usually, only the Administrator
+    can create symlinks. Instead we create small Python scripts that call
+    the programs we want, implementing the effect of a symlink. This is cleaner
+    than creating Windows executables or trying to use non-Python scripts.
+    The temporary folder generated for Windows could like:
+
+        %TEMP%\abcxyz\tesseract.py:
+            (script that runs ocrmypdf/tests/resources/spoof/tesseract_crash.py)
+        %TEMP%\abcxyz\gswin32c.py:
+            (script that runs ocrmypdf/tests/resources/spoof/gs_backflip.py)
+        %TEMP%\abcxyz\gswin64c.py:
+            (script that runs ocrmypdf/tests/resources/spoof/gs_backflip.py)
+
+    We also address one quirk here, that Ghostscript may be known as gswin32c
+    or gswin64c, depending on what the user installed (regardless of Windows
+    itself). On POSIX, Ghostscript is just 'gs'.  We handle the special case here
+    too.
+
+    All of this is intimately dependent on the machinery in ocrmypdf.exec.run().
+    In particular, for Windows, that code has to know that if there is a .py
+    file, it needs to run it with Python, since Windows does not like being
+    asked to execute files.
+
+    We don't overload PATH directly because we have some tests where we call
+    ocrmypdf as a subprocess (to exercise the command line interface) and some
+    tests where we call it as an API.
     """
     env = os.environ.copy()
     slug = '-'.join(v.replace('.py', '') for v in sorted(kwargs.values()))
@@ -108,45 +148,33 @@ def spoof(tmp_path_factory, **kwargs):
 
     for replace_program, with_spoof in kwargs.items():
         spoofer = Path(SPOOF_PATH) / with_spoof
-        spoofer.chmod(0o755)
-        (tmpdir / replace_program).symlink_to(spoofer)
-
-    env['_OCRMYPDF_SAVE_PATH'] = env['PATH']
-    env['PATH'] = str(tmpdir) + ":" + env['PATH']
+        if os.name != 'nt':
+            spoofer.chmod(0o755)
+            (tmpdir / replace_program).symlink_to(spoofer)
+        else:
+            py_file = WINDOWS_SHIM_TEMPLATE.format(
+                spoofer=repr(os.fspath(spoofer.absolute()))
+            )
+            if replace_program == 'gs':
+                programs = ['gswin64c', 'gswin32c']
+            else:
+                programs = [replace_program]
+            for prog in programs:
+                (tmpdir / f'{prog}.py').write_text(py_file, encoding='utf-8')
 
+    env['_OCRMYPDF_TEST_PATH'] = str(tmpdir) + os.pathsep + env['PATH']
+    if os.name == 'nt':
+        if '.py' not in env['PATHEXT'].lower():
+            raise EnvironmentError("PATHEXT is not configured to support .py")
     return env
 
 
-@pytest.helpers.register
-@contextmanager
-def os_environ(new_env):
-    old_env = os.environ.copy()
-    if new_env is None:
-        new_env = {}
-
-    for k, v in new_env.items():
-        if k != 'PYTEST_CURRENT_TEST':
-            os.environ[k] = v
-    yield
-    new_keys = set(os.environ.copy()) - set(old_env)
-    for k in new_keys:
-        if k != 'PYTEST_CURRENT_TEST':
-            del os.environ[k]
-    for k in old_env:
-        if k != 'PYTEST_CURRENT_TEST':
-            os.environ[k] = old_env[k]
-
-    for k, v in os.environ.copy().items():
-        if k != 'PYTEST_CURRENT_TEST':
-            assert v == old_env[k]
-
-
-@pytest.fixture(scope='session')
+@pytest.fixture
 def spoof_tesseract_noop(tmp_path_factory):
     return spoof(tmp_path_factory, tesseract='tesseract_noop.py')
 
 
-@pytest.fixture(scope='session')
+@pytest.fixture
 def spoof_tesseract_cache(tmp_path_factory):
     if running_in_docker():
         return os.environ.copy()
@@ -186,15 +214,14 @@ def no_outpdf(tmp_path):
 def check_ocrmypdf(input_file, output_file, *args, env=None):
     """Run ocrmypdf and confirmed that a valid file was created"""
 
-    # p, out, err = run_ocrmypdf(input_file, output_file, *args, env=env)
-
     options = cli.parser.parse_args(
-        [str(input_file), str(output_file)] + [str(arg) for arg in args]
+        [str(input_file), str(output_file)]
+        + [str(arg) for arg in args if arg is not None]
     )
     api.check_options(options)
     if env:
         options.tesseract_env = env
-        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = input_file
+        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(input_file)
     result = api.run_pipeline(options, api=True)
 
     assert result == 0
@@ -204,18 +231,51 @@ def check_ocrmypdf(input_file, output_file, *args, env=None):
     return output_file
 
 
+@pytest.helpers.register
+def run_ocrmypdf_api(input_file, output_file, *args, env=None):
+    """Run ocrmypdf via API and let caller deal with results
+
+    Does not currently have a way to manipulate the PATH except for Tesseract.
+    """
+
+    options = cli.parser.parse_args(
+        [str(input_file), str(output_file)]
+        + [str(arg) for arg in args if arg is not None]
+    )
+    api.check_options(options)
+    if env:
+        options.tesseract_env = env.copy()
+        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(input_file)
+        first_path = env.get('_OCRMYPDF_TEST_PATH', '').split(os.pathsep)[0]
+        if 'spoof' in first_path:
+            assert 'gs' not in first_path, "use run_ocrmypdf() for gs"
+            assert 'tesseract' in first_path
+    if options.tesseract_env:
+        assert all(isinstance(v, (str, bytes)) for v in options.tesseract_env.values())
+
+    return api.run_pipeline(options, api=False)
+
+
 @pytest.helpers.register
 def run_ocrmypdf(input_file, output_file, *args, env=None, universal_newlines=True):
     "Run ocrmypdf and let caller deal with results"
 
     if env is None:
-        env = os.environ
+        env = os.environ.copy()
 
     p_args = (
         OCRMYPDF
         + [str(arg) for arg in args if arg is not None]
         + [str(input_file), str(output_file)]
     )
+
+    # Tell subprocess where to find coverage.py configuration
+    # This has no effect except when coverage is running
+    # Details: https://coverage.readthedocs.io/en/coverage-5.0/subprocess.html
+    coverage_rc = Path(__file__).parent.parent / '.coveragerc'
+    assert coverage_rc.exists()
+    env['COVERAGE_PROCESS_START'] = os.fspath(coverage_rc)
+
     p = run(
         p_args, stdout=PIPE, stderr=PIPE, universal_newlines=universal_newlines, env=env
     )