Fix log.error where log is None

Closes #231

.docker/Dockerfile

@@ -3,28 +3,49 @@
 FROM      ubuntu:18.04
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
+  build-essential autoconf automake libtool \
+  libleptonica-dev \
+  zlib1g-dev \
+  libexempi3 \
   ocrmypdf \
-  unpaper \
+  pngquant \
+  python3-pip \
+  python3-venv \
+  tesseract-ocr \
+  tesseract-ocr-chi-sim \
+  tesseract-ocr-deu \
   tesseract-ocr-eng \
   tesseract-ocr-fra \
-  tesseract-ocr-deu \
-  tesseract-ocr-spa \
   tesseract-ocr-por \
-  tesseract-ocr-chi-sim \
-  python3-venv \
-  python3-pip
+  tesseract-ocr-spa \
+  unpaper \
+  wget
+
 
 ENV LANG=C.UTF-8
 
+# 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 - | \
+      tar xz -C jbig2 --strip-components=1 \
+  && cd jbig2 \
+  && ./autogen.sh && ./configure && make && make install \
+  && cd .. \
+  && rm -rf jbig2
+
+RUN apt-get remove -y autoconf automake libtool
+
 RUN python3 -m venv --system-site-packages /appenv
 
 # This installs the latest binary wheel instead of the code in the current
 # folder. Installing from source will fail, apparently because cffi needs
-# build-essentials (gcc) to do a source installation 
+# build-essentials (gcc) to do a source installation
 # (i.e. "pip install ."). It's unclear to me why this is the case.
 RUN . /appenv/bin/activate; \
   pip install --upgrade pip \
-  && pip install ocrmypdf[fitz]
+  && pip install --upgrade ocrmypdf
 
 # Now copy the application in, mainly to get the test suite.
 # Do this now to make the best use of Docker cache.
@@ -35,6 +56,7 @@ RUN . /appenv/bin/activate; \
 # Remove the junk, including the source version of application since it was
 # already installed
 RUN rm -rf /tmp/* /var/tmp/* /root/* /application/ocrmypdf \
+  && apt-get remove -y build-essential \
   && apt-get autoremove -y \
   && apt-get autoclean -y
 

.gitignore

@@ -1,47 +1,40 @@
 # Development environment
+.bash_history
+.pylintrc
+.pytest_cache/
+.ruffus_history.sqlite
+.venv/
 *.pyc
 *.sublime-*
-venv*/
-.venv/
-pyvenv.cfg
-tasks.py
-.bash_history
-.ruffus_history.sqlite
-.idea/
-.pytest_cache/
-.pylintrc
 
 # Package building
-*.egg-info/
-.cache/
 .eggs/
+*.egg-info/
 build/
 dist/
 wheelhouse/
 
 # Automatically generated files
-ocrmypdf/lib/_*.py
-ocrmypdf/version.py
 docs/_build/
 docs/_static/
 docs/_templates/
 docs/Makefile
+ocrmypdf/lib/_*.py
 
 # Code coverage
 .coverage
 htmlcov/
 
 # Testing
-log/
+.ipynb_checkpoints/
+.vscode/
+*.ipynb
+*.profile
 /*.pdf
 /*.qdf
-*.ipynb
-.ipynb_checkpoints/
+/scratch.py
+IDEAS
+log/
 tests/output/
 tests/resources/private/
 tmp/
-pdfbox-app*.jar
-.vscode/
-IDEAS
-_Dockerfile.local
-/scratch.py

.travis.yml

@@ -1,50 +1,70 @@
-dist: trusty
-language: python
 cache:
   pip: true
   directories:
   - $HOME/Library/Caches/Homebrew
 
-addons:
-  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
-
-env:
-  global:
-    - secure: "hsf6MT+n2x3OiDM2fQyJZdV0/PWYmv81LdVqC6cfnHBE/8N3DloJRqQ7WfO14TxhiK9PEC7MpyCj0lSabUHEO7gSH6Vks6I1asoSkt8S9/bSMlhT4hei+pwVpeGEiU5xHVATNjY+D919VC3IFvc3XmjT74h/2SLhaZ+jhEmDggM="  # HOMEBREW_OCRMYPDF_TOKEN
-
 matrix:
   include:
     - os: linux
+      dist: trusty
       sudo: required
       language: python
       python: "3.5"
+      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: trusty
       sudo: required
       language: python
       python: "3.6"
+      env:
+        - DIST=trusty
+      addons: *trusty_apt
     - os: linux
+      dist: xenial
       sudo: required
       language: python
-      python: "3.7-dev"
+      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: xcode8
       language: generic
@@ -57,9 +77,11 @@ before_install: |
   if [[ "$TRAVIS_OS_NAME" == "linux" ]]; then
     pip3 install --upgrade pip
     pip3 install --upgrade wheel
-    mkdir -p packages
-    wget -q 'https://www.dropbox.com/s/vaq0kbwi6e6au80/unpaper_6.1-1.deb?raw=1' -O packages/unpaper_6.1-1.deb
-    sudo dpkg -i packages/unpaper_6.1-1.deb
+    if [[ "$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
     brew update && brew bundle --file=.travis/Brewfile
     pip3 install --upgrade pip
@@ -94,18 +116,3 @@ deploy:
     tags: true
     condition: $TRAVIS_PYTHON_VERSION == "3.6" && $TRAVIS_OS_NAME == "linux"
   skip_upload_docs: true
-
-  # null deploy for osx
-  # we really just want to run after_deploy *after* pypi upload is done, but
-  # after_deploy on runs if a given box deployed
-- provider: script
-  script: /usr/bin/true
-  on:
-    branch: master
-    tags: true
-    condition: $TRAVIS_OS_NAME == "osx"
-
-after_deploy: |
-  if [[ "$TRAVIS_OS_NAME" == "osx" ]]; then
-    bash .travis/osx_brew.sh
-  fi

.travis/Brewfile

@@ -2,10 +2,8 @@ brew 'python'
 
 brew 'exempi'
 brew 'ghostscript'
-brew 'jbig2dec'
 brew 'jbig2enc'
 brew 'leptonica'
-brew 'libffi'
 brew 'libtiff'
 brew 'openjpeg'
 brew 'pngquant'

.travis/autobrew.py

@@ -1,93 +0,0 @@
-#!/usr/bin/env python3
-# © 2017-18 James R. Barlow: github.com/jbarlow83
-
-from string import Template
-from subprocess import run, PIPE
-import re
-
-recipe_template = Template("""
-class Ocrmypdf < Formula
-  include Language::Python::Virtualenv
-
-  desc "Adds an OCR text layer to scanned PDF files"
-  homepage "https://github.com/jbarlow83/OCRmyPDF"
-  ${ocrmypdf_url}
-  ${ocrmypdf_sha256}
-
-  depends_on "pkg-config" => :build
-  depends_on "mupdf-tools" => :build  # statically links libmupdf.a
-  depends_on "freetype"
-  depends_on "ghostscript"
-  depends_on "jpeg"
-  depends_on "libpng"
-  depends_on "python"
-  depends_on "qpdf"
-  depends_on "tesseract"
-  depends_on "unpaper"
-
-${resources}
-  def install
-    venv = virtualenv_create(libexec, "python3")
-
-    resource("Pillow").stage do
-      inreplace "setup.py" do |s|
-        sdkprefix = MacOS::CLT.installed? ? "" : MacOS.sdk_path
-        s.gsub! "openjpeg.h", "probably_not_a_header_called_this_eh.h"
-        s.gsub! "ZLIB_ROOT = None", "ZLIB_ROOT = ('#{sdkprefix}/usr/lib', '#{sdkprefix}/usr/include')"
-        s.gsub! "JPEG_ROOT = None", "JPEG_ROOT = ('#{Formula["jpeg"].opt_prefix}/lib', '#{Formula["jpeg"].opt_prefix}/include')"
-        s.gsub! "FREETYPE_ROOT = None", "FREETYPE_ROOT = ('#{Formula["freetype"].opt_prefix}/lib', '#{Formula["freetype"].opt_prefix}/include')"
-      end
-
-      # avoid triggering "helpful" distutils code that doesn't recognize Xcode 7 .tbd stubs
-      ENV.append "CFLAGS", "-I#{MacOS.sdk_path}/System/Library/Frameworks/Tk.framework/Versions/8.5/Headers" unless MacOS::CLT.installed?
-      venv.pip_install Pathname.pwd
-    end
-
-    res = resources.map(&:name).to_set - ["Pillow"]
-
-    res.each do |r|
-      venv.pip_install resource(r)
-    end
-
-    venv.pip_install_and_link buildpath
-  end
-
-  test do
-    # Since we use Python 3, we require a UTF-8 locale
-    ENV["LC_ALL"] = "en_US.UTF-8"
-
-    system "#{bin}/ocrmypdf", "-f", "-q", "--deskew",
-                              test_fixtures("test.pdf"), "ocr.pdf"
-    assert_predicate testpath/"ocr.pdf", :exist?
-  end
-end
-""")
-
-def main():
-    p = run(['poet', '--single', 'ocrmypdf'],
-            encoding='utf-8', stdout=PIPE, check=True)
-
-    ocrmypdf_lines = p.stdout.splitlines()
-    ocrmypdf_url = ocrmypdf_lines[1].strip()
-    ocrmypdf_sha256 = ocrmypdf_lines[2].strip()
-
-    ocrmypdf_version = re.search(
-        r'ocrmypdf-(.+)\.tar.*', ocrmypdf_url).group(1)
-    print(f"Autobrewing {ocrmypdf_version}")
-
-    p = run(['poet', '--resources', 'ocrmypdf'],
-            encoding='utf-8', stdout=PIPE, check=True)
-
-    poet_resources = p.stdout
-
-    # Remove the duplicate "ocrmypdf" resource block
-    all_resources = poet_resources.split('resource')
-    kept_resources = [block for block in all_resources if 'ocrmypdf' not in block]
-    resources = 'resource'.join(kept_resources)
-
-    with open('ocrmypdf.rb', 'w') as out:
-        out.write(recipe_template.substitute(**locals()))
-
-
-if __name__ == '__main__':
-    main()

.travis/osx_brew.sh

@@ -1,23 +0,0 @@
-#!/bin/bash
-# © 2017 James R. Barlow: github.com/jbarlow83
-set -uo pipefail
-set -x
-
-pip3 install homebrew-pypi-poet
-python3 .travis/autobrew.py
-cat ocrmypdf.rb
-
-# brew audit crashes Travis
-#brew audit ocrmypdf.rb
-
-# Important: disable debug output so token is hidden
-set +x
-git clone https://$HOMEBREW_OCRMYPDF_TOKEN@github.com/jbarlow83/homebrew-ocrmypdf.git
-set -x
-
-pushd homebrew-ocrmypdf
-cp ../ocrmypdf.rb Formula/ocrmypdf.rb
-git add Formula/ocrmypdf.rb
-git commit -m "homebrew-ocrmypdf: automatic release $TRAVIS_BUILD_NUMBER $TRAVIS_TAG"
-git push origin master
-popd

README.md

@@ -0,0 +1,132 @@
+OCRmyPDF
+========
+
+![image](https://travis-ci.org/jbarlow83/OCRmyPDF.svg?branch=master%0A%20:target:%20https://travis-ci.org/jbarlow83/OCRmyPDF)
+
+![image](https://img.shields.io/pypi/v/ocrmypdf.svg%0A%20:target:%20https://pypi.org/project/ocrmypdf/)
+
+![image](https://img.shields.io/homebrew/v/ocrmypdf.svg%0A%20:alt:%20homebrew%0A%20:target:%20http://brewformulas.org/Ocrmypdf)
+
+OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to be searched or copy-pasted.
+
+```bash
+ocrmypdf                      # it's a scriptable command line program
+   -l eng+fra                 # it supports multiple languages
+   --rotate-pages             # it can fix pages that are misrotated
+   --deskew                   # it can deskew crooked PDFs!
+   --title "My PDF"           # it can change output metadata
+   --jobs 4                   # it uses multiple cores by default
+   --output-type pdfa         # it produces PDF/A by default
+   input_scanned.pdf          # takes PDF input (or images)
+   output_searchable.pdf      # produces validated PDF output
+```
+
+Main features
+-------------
+
+-   Generates a searchable [PDF/A](https://en.wikipedia.org/?title=PDF/A) file from a regular PDF
+-   Places OCR text accurately below the image to ease copy / paste
+-   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
+-   Validates input and output files
+-   Distributes work across all available CPU cores
+-   Uses [Tesseract OCR](https://github.com/tesseract-ocr/tesseract) engine
+-   Supports more than [100 languages](https://github.com/tesseract-ocr/tessdata) recognized by Tesseract
+-   Battle-tested on thousands of PDFs, a test suite and continuous integration
+
+For details: please consult the [documentation](https://ocrmypdf.readthedocs.io/en/latest/).
+
+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.
+
+-   Either they produced PDF files with misplaced text under the image (making copy/paste impossible)
+-   Or they did not handle accents and multilingual characters
+-   Or they changed the resolution of the embedded images
+-   Or they generated ridiculously large PDF files
+-   Or they crashed when trying to OCR
+-   Or they did not produce valid PDF files
+-   On top of that none of them produced PDF/A files (format dedicated for long time storage)
+
+...so I decided to develop my own tool.
+
+Installation
+------------
+
+Linux, UNIX, and macOS are supported. Windows is not directly supported but there is a Docker image available that runs on Windows.
+
+Users of Debian 9 or later or Ubuntu 16.10 or later may simply
+
+```bash
+apt-get install ocrmypdf
+```
+
+and macOS users with Homebrew may simply
+
+```bash
+brew install ocrmypdf
+```
+
+For everyone else, [see our documentation](https://ocrmypdf.readthedocs.io/en/latest/installation.html) for installation steps.
+
+Languages
+---------
+
+OCRmyPDF uses Tesseract for OCR, and relies on its language packs. For Linux users, you can often find packages that provide language packs:
+
+```bash
+# Display a list of all Tesseract language packs
+apt-cache search tesseract-ocr
+
+# Debian/Ubuntu users
+apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language back
+```
+
+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.
+
+Documentation and support
+-------------------------
+
+Once ocrmypdf is installed, the built-in help which explains the command syntax and options can be accessed via:
+
+```bash
+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)
+
+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.
+
+Press & Media
+-------------
+
+-   [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)
+
+License
+-------
+
+The OCRmyPDF software is licensed under the GNU GPLv3. Certain files are covered by other licenses, as noted in their source files.
+
+The license for each test file varies, and is noted in tests/resources/README.rst. The documentation is licensed under Creative Commons Attribution-ShareAlike 4.0 (CC-BY-SA 4.0).
+
+OCRmyPDF versions prior to 6.0 were licensed under the MIT License.
+
+Disclaimer
+----------
+
+The software is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

README.rst

@@ -1,150 +0,0 @@
-OCRmyPDF
-========
-
-.. image:: https://travis-ci.org/jbarlow83/OCRmyPDF.svg?branch=master
-    :target: https://travis-ci.org/jbarlow83/OCRmyPDF
-
-.. image:: https://img.shields.io/pypi/v/ocrmypdf.svg
-    :target: https://pypi.org/project/ocrmypdf/
-
-.. image:: https://img.shields.io/homebrew/v/ocrmypdf.svg
-    :alt: homebrew
-    :target: http://brewformulas.org/Ocrmypdf
-
-
-OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to
-be searched or copy-pasted.
-
-.. code-block:: bash
-
-   ocrmypdf                      # it's a scriptable command line program
-      -l eng+fra                 # it supports multiple languages
-      --rotate-pages             # it can fix pages that are misrotated
-      --deskew                   # it can deskew crooked PDFs!
-      --title "My PDF"           # it can change output metadata
-      --jobs 4                   # it uses multiple cores by default
-      --output-type pdfa         # it produces PDF/A by default
-      input_scanned.pdf          # takes PDF input (or images)
-      output_searchable.pdf      # produces validated PDF output
-
-
-Main features
--------------
-
--  Generates a searchable
-   `PDF/A <https://en.wikipedia.org/?title=PDF/A>`_ file from a regular PDF
--  Places OCR text accurately below the image to ease copy / paste
--  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
--  Validates input and output files
--  Distributes work across all available CPU cores
--  Uses `Tesseract OCR <https://github.com/tesseract-ocr/tesseract>`_ engine
--  Supports more than `100 languages <https://github.com/tesseract-ocr/tessdata>`_ recognized by Tesseract
--  Battle-tested on thousands of PDFs, a test suite and continuous integration
-
-For details: please consult the `documentation <https://ocrmypdf.readthedocs.io/en/latest/>`_.
-
-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.
-
--  Either they produced PDF files with misplaced text under the image (making copy/paste impossible)
--  Or they did not handle accents and multilingual characters
--  Or they changed the resolution of the embedded images
--  Or they generated ridiculously large PDF files
--  Or they crashed when trying to OCR
--  Or they did not produce valid PDF files
--  On top of that none of them produced PDF/A files (format dedicated for long time storage)
-
-...so I decided to develop my own tool.
-
-Installation
-------------
-
-Linux, UNIX, and macOS are supported. Windows is not directly supported but there is a Docker image available that runs on Windows.
-
-Users of Debian 9 or later or Ubuntu 16.10 or later may simply
-
-.. code-block:: bash
-
-   apt-get install ocrmypdf
-
-and macOS users with Homebrew may simply
-
-.. code-block:: bash
-
-   brew install ocrmypdf
-
-For everyone else, `see our documentation <https://ocrmypdf.readthedocs.io/en/latest/installation.html>`_ for installation steps.
-
-Languages
----------
-
-OCRmyPDF uses Tesseract for OCR, and relies on its language packs. For Linux users,
-you can often find packages that provide language packs:
-
-.. code-block:: bash
-
-   # Display a list of all Tesseract language packs
-   apt-cache search tesseract-ocr
-
-   # Debian/Ubuntu users
-   apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language back
-
-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.
-
-Documentation and support
--------------------------
-
-Once ocrmypdf is installed, the built-in help which explains the command syntax and options can be accessed via:
-
-.. code-block:: bash
-
-   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)
-
-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.
-
-Press & Media
--------------
-
--  `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>`_
-
-License
--------
-
-The OCRmyPDF software is licensed under the GNU GPLv3. Certain files are covered by other licenses, as noted in their source files.
-
-The license for each test file varies, and is noted in tests/resources/README.rst. The documentation is licensed under Creative Commons Attribution-ShareAlike 4.0 (CC-BY-SA 4.0).
-
-OCRmyPDF versions prior to 6.0 were licensed under the MIT License.
-
-Disclaimer
-----------
-
-The software is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
-CONDITIONS OF ANY KIND, either express or implied.

docs/conf.py

@@ -60,6 +60,24 @@ author = 'James R. Barlow'
 # built documents.
 #
 # The short X.Y version.
+
+import os
+on_rtd = os.environ.get('READTHEDOCS') == 'True'
+
+if on_rtd:
+    # Help ReadTheDocs avoid having to install any binary extension modules
+    import sys
+    from unittest.mock import MagicMock
+
+    class Mock(MagicMock):
+        @classmethod
+        def __getattr__(cls, name):
+            return MagicMock()
+
+    MOCK_MODULES = ['pikepdf', 'libxmp', 'libxmp.utils']
+    sys.modules.update((mod_name, Mock()) for mod_name in MOCK_MODULES)
+
+
 from ocrmypdf import __version__ as OCRMYPDF_VERSION
 
 _version_parts = OCRMYPDF_VERSION.split('.')

docs/cookbook.rst

@@ -61,7 +61,7 @@ If the page is "just a little off horizontal", like a crooked picture, then you
 OCR languages other than English
 """"""""""""""""""""""""""""""""
 
-By default OCRmyPDF assumes the document is English. 
+By default OCRmyPDF assumes the document is English.
 
 .. code-block:: bash
 
@@ -127,7 +127,7 @@ OCRmyPDF perform some image processing on each page of a PDF, if desired.  The s
 * ``--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.
@@ -176,7 +176,7 @@ To redo OCR on a file OCRed with other OCR software or a previous version of OCR
 
     ocrmypdf --force-ocr input.pdf output.pdf
 
-Note that the method above will force rasterization of all pages, potentially reducing quality or losing vector content. 
+Note that the method above will force rasterization of all pages, potentially reducing quality or losing vector content.
 
 To ensure quality is preserved, one could extract all of the images and rebuild the PDF for a lossless transformation. This recipe does not work when PDFs contain multiple images per page, as many do in practice. It will also lose any page rotation information.
 
@@ -198,3 +198,19 @@ The `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.
 
 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.
+
+The ``--optimize N`` (short form ``-O``) argument controls optimization, where ``N`` ranges from 0 to 3. ``--optimize 0`` disables optimizations. ``1`` enables lossless optimizations that can be performed safely with no quality loss. ``2`` enables lossy optimizations such as image color quantizations. ``3`` enables more aggressive optimizations and targets a lower JPEG 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 will not be optimized.
+
+Currently optimization attempts to find more efficient encodings for images. The types of optimization available may expand over time. By default, OCRmyPDF compresses data streams inside PDFs, and will change inefficient encodings 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
+
+    ocrmypdf --optimize 3 in.pdf out.pdf  # Make it as small as possible

docs/index.rst

@@ -18,6 +18,7 @@ PDF is the best format for storing and exchanging scanned documents.  Unfortunat
    release_notes
    installation
    languages
+   jbig2
 
 .. toctree::
    :caption: Usage
@@ -36,4 +37,3 @@ Indices and tables
 * :ref:`genindex`
 * :ref:`modindex`
 * :ref:`search`
-

docs/installation.rst

@@ -1,16 +1,53 @@
 Installation
 ============
 
+.. |latest| image:: https://img.shields.io/pypi/v/ocrmypdf.svg
+    :alt: OCRmyPDF latest released version on PyPI
+
+|latest|
+
 The easiest way to install OCRmyPDF to follow the steps for your operating system/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.
 
 .. contents:: Platform-specific steps
-    :depth: 1
+    :depth: 2
     :local:
 
-Installing on Debian and Ubuntu 16.10 or newer
-----------------------------------------------
+Installing on Linux
+-------------------
+
+Debian and Ubuntu 16.10 or newer
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+.. |deb-stable| image:: https://repology.org/badge/version-for-repo/debian_stable/ocrmypdf.svg
+    :alt: Debian 9 stable ("stretch")
+
+.. |deb-testing| image:: https://repology.org/badge/version-for-repo/debian_testing/ocrmypdf.svg
+    :alt: Debian 10 testing ("buster")
+
+.. |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
+
+
++-------------------------------------------+
+| **OCRmyPDF version**                      |
++-------------------------------------------+
+| |latest|                                  |
++-------------------------------------------+
+| |deb-stable| |deb-testing| |deb-unstable| |
++-------------------------------------------+
+| |ubu-1710| |ubu-1804| |ubu-1810|          |
++-------------------------------------------+
 
 Users of Debian 9 ("stretch") or later or Ubuntu 16.10 or later may simply
 
@@ -18,10 +55,178 @@ Users of Debian 9 ("stretch") or later or Ubuntu 16.10 or later may simply
 
     apt-get install ocrmypdf
 
-To see what versions are available, check the `Debian Package Tracker <https://tracker.debian.org/pkg/ocrmypdf>`_ or `Ubuntu launchpad.net <https://launchpad.net/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`_.
 
-Installing on macOS with Homebrew
----------------------------------
+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`.
+
+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:
+
+.. code-block:: bash
+
+    sudo apt-get update
+    sudo apt-get install \
+        ocrmypdf \
+        python3-pip
+
+There are a few dependency changes between ocrmypdf 6.1.2 and 7.x. Let's get
+these, too.
+
+.. code-block:: bash
+
+    sudo apt-get install \
+        libexempi3 \
+        pngquant
+
+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
+
+To add JBIG2 encoding, see :ref:`jbig2`.
+
+Ubuntu 16.04 LTS
+^^^^^^^^^^^^^^^^
+
+No package is currently available for Ubuntu 16.04, but you can install the dependencies manually:
+
+.. code-block:: bash
+
+    sudo apt-get update
+    sudo apt-get install \
+        ghostscript \
+        libexempi3 \
+        pngquant \
+        python3-cffi \
+        python3-pip \
+        qpdf \
+        tesseract-ocr \
+        unpaper
+
+If you wish install OCRmyPDF for the current user, and ensure that the ``PATH``
+environment variable contains ``$HOME/.local/bin``.
+
+.. code-block:: bash
+
+    export PATH=$HOME/.local/bin:$PATH
+    pip3 install --user ocrmypdf
+
+Alternately, you can install ocrmypdf system-wide. (Not recommended.)
+
+.. code-block:: bash
+
+    sudo pip3 install ocrmypdf
+
+At your option, you may upgrade Ubuntu 16.04 LTS to Tesseract 4.0 for improved OCR results.
+
+.. code-block:: bash
+
+    sudo apt-get install -y software-properties-common python-software-properties
+    sudo add-apt-repository ppa:alex-p/tesseract-ocr -y
+    sudo apt-get update
+    sudo apt-get upgrade tesseract-ocr
+
+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 it is older and does not provide ``pip``.
+
+Update apt-get:
+
+.. code-block:: bash
+
+    sudo apt-get update
+
+Install system dependencies:
+
+.. code-block:: bash
+
+    sudo apt-get install \
+        software-properties-common python-software-properties \
+        zlib1g-dev \
+        libexempi3 \
+        libjpeg-dev \
+        libffi-dev \
+        pngquant \
+        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.
+
+If you prefer to not modify your system in this matter, consider using a Docker container.
+
+.. code-block:: bash
+
+    sudo add-apt-repository ppa:vshn/ghostscript -y
+    sudo add-apt-repository ppa:heyarje/libav-11 -y
+    sudo add-apt-repository ppa:alex-p/tesseract-ocr -y
+    sudo add-apt-repository ppa:jonathonf/python-3.6 -y
+
+    sudo apt-get update
+
+    sudo apt-get install \
+        python3.6-dev \
+        ghostscript \
+        tesseract-ocr \
+        tesseract-ocr-eng \
+        libavformat56 libavcodec56 libavutil54 \
+        wget
+
+Now we need to install ``pip`` and let it install ocrmypdf:
+
+.. code-block:: bash
+
+    curl https://bootstrap.pypa.io/ez_setup.py -o - | python3.6 && python3.6 -m easy_install pip
+    pip3.6 install ocrmypdf
+
+The ``wget`` command will download a program and run it.
+
+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
+
+To add JBIG2 encoding, see :ref:`jbig2`.
+
+ArchLinux
+^^^^^^^^^
+
+.. image:: https://repology.org/badge/version-for-repo/aur/ocrmypdf.svg
+    :alt: ArchLinux
+    :target: https://repology.org/metapackage/ocrmypdf
+
+The author is aware of an `ArchLinux User Repository package for ocrmypdf <https://aur.archlinux.org/packages/ocrmypdf/>`_. You can use the following command.
+
+.. code-block:: bash
+
+    yaourt -S ocrmypdf
+
+If you have any difficulties with installation, check the repository package page.
+
+Other Linux packages
+^^^^^^^^^^^^^^^^^^^^
+
+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.
+
+Installing on macOS
+-------------------
+
+Homebrew
+^^^^^^^^
 
 .. image:: https://img.shields.io/homebrew/v/ocrmypdf.svg
     :alt: homebrew
@@ -41,12 +246,64 @@ OCRmyPDF is now a standard `Homebrew <https://brew.sh>`_ formula. To install on
 
     Users who previously installed OCRmyPDF from the private tap should switch to the mainline version (``brew untap jbarlow83/ocrmypdf``) and install from there.
 
-.. _Docker-install:
+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/>`_.
+
+Update Homebrew:
+
+.. code-block:: bash
+
+    brew update
+
+Install or upgrade the required Homebrew packages, if any are missing. To do this, download the ``Brewfile`` that lists all of the dependencies to the current directory, and run ``brew bundle`` to process them (installing or upgrading as needed). ``Brewfile`` is a plain text file.
+
+.. code-block:: bash
+
+    wget https://github.com/jbarlow83/OCRmyPDF/raw/master/.travis/Brewfile
+    brew bundle
+
+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
+
+    brew install tesseract --with-all-languages  # Option 2: for all language packs
+
+Update the homebrew pip:
+
+.. code-block:: bash
+
+    pip3 install --upgrade pip
+
+You can then install OCRmyPDF from PyPI, for the current user:
+
+.. code-block:: bash
+
+    pip3 install --user ocrmypdf
+
+or system-wide:
+
+.. code-block:: bash
+
+    pip3 install ocrmypdf
+
+The command line program should now be available:
+
+.. code-block:: bash
+
+    ocrmypdf --help
+
+.. _docker-install:
 
 Installing the Docker image
 ---------------------------
 
-For many 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. For Windows, it is the only option.
 
 If you have `Docker <https://docs.docker.com/>`_ installed on your system, you can install a Docker image of the latest release.
 
@@ -142,201 +399,12 @@ Or in the wonderful `fish shell <https://fishshell.com/>`_:
 
 .. note::
 
-    The ocrmypdf Docker containers are designed to be used for a single OCR job. The ``docker run --rm`` argument tells Docker to delete temporary storage associated with container when it is done executing.
-
-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/>`_.
-
-Update Homebrew:
-
-.. code-block:: bash
-
-    brew update
-
-Install or upgrade the required Homebrew packages, if any are missing:
-
-.. code-block:: bash
-
-    brew install libpng openjpeg jbig2dec libtiff     # image libraries
-    brew install qpdf
-    brew install ghostscript
-    brew install python3
-    brew install libxml2 libffi leptonica
-    brew install unpaper   # optional
-
-Python 3.5, 3.6 and 3.7 are supported.
-
-Install the required Tesseract OCR engine with the language packs you plan to use:
-
-.. code-block:: bash
-
-    brew install tesseract                       # Option 1: for English, French, German, Spanish
-
-.. _macos-all-languages:
-
-.. code-block:: bash
-
-    brew install tesseract --with-all-languages  # Option 2: for all language packs
-
-Update the homebrew pip:
-
-.. code-block:: bash
-
-    pip3 install --upgrade pip
-
-You can then install OCRmyPDF from PyPI, for the current user:
-
-.. code-block:: bash
-
-    pip3 install --user ocrmypdf
-
-or system-wide:
-
-.. code-block:: bash
-
-    pip3 install ocrmypdf
-
-The command line program should now be available:
-
-.. code-block:: bash
-
-    ocrmypdf --help
-
-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 all the dependencies:
-
-.. code-block:: bash
-
-    sudo apt-get update
-    sudo apt-get install \
-        ocrmypdf \
-        python3-pip
-
-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
-
-
-Installing on Ubuntu 16.04 LTS
-------------------------------
-
-No package is currently available for Ubuntu 16.04, but you can install the dependencies manually:
-
-.. code-block:: bash
-
-    sudo apt-get update
-    sudo apt-get install \
-        unpaper \
-        ghostscript \
-        tesseract-ocr \
-        qpdf \
-        python3-pip \
-        python3-cffi
-
-If you wish install OCRmyPDF for the current user:
-
-.. code-block:: bash
-
-    pip3 install --user ocrmypdf
-
-Alternately, system-wide. Note that this may modify the system Python environment:
-
-.. code-block:: bash
-
-    sudo pip3 install ocrmypdf
-
-If you wish to install OCRmyPDF to a virtual environment to isolate the system Python, you can follow these steps.
-
-.. code-block:: bash
-
-    python3 -m venv venv-ocrmypdf
-    source venv-ocrmypdf/bin/activate
-    pip3 install ocrmypdf
-
-Installing on Ubuntu 14.04 LTS
-------------------------------
-
-Installing on Ubuntu 14.04 LTS (trusty) is more difficult than some other options, because it is older and does not provide ``pip``.
-
-Update apt-get:
-
-.. code-block:: bash
-
-    sudo apt-get update
-
-Install system dependencies:
-
-.. code-block:: bash
-
-    sudo apt-get install \
-        software-properties-common python-software-properties \
-        zlib1g-dev \
-        libjpeg-dev \
-        libffi-dev \
-        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.
-
-If you prefer to not modify your system in this matter, consider using a Docker container.
-
-.. code-block:: bash
-
-    sudo add-apt-repository ppa:vshn/ghostscript -y
-    sudo add-apt-repository ppa:heyarje/libav-11 -y
-    sudo add-apt-repository ppa:alex-p/tesseract-ocr -y
-    sudo add-apt-repository ppa:jonathonf/python-3.6 -y
-
-    sudo apt-get update
-
-    sudo apt-get install \
-        python3.6-dev \
-        ghostscript \
-        tesseract-ocr \
-        tesseract-ocr-eng \
-        libavformat56 libavcodec56 libavutil54 \
-        wget
-
-Now we need to install ``pip`` and let it install ocrmypdf:
-
-.. code-block:: bash
-
-    curl https://bootstrap.pypa.io/ez_setup.py -o - | python3.6 && python3.6 -m easy_install pip
-    pip3.6 install ocrmypdf
-
-The ``wget`` command will download a program and run it.
-
-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
-
-
-Installing on ArchLinux
------------------------
-
-The author is aware of an `ArchLinux package for ocrmypdf <https://aur.archlinux.org/packages/ocrmypdf/>`_. It seems like the following command might work.
-
-.. code-block:: bash
-
-    pacman -S ocrmypdf
-
+    The ocrmypdf Docker images are designed for application delivery, to enable use of OCRmyPDF without fussing with dependencies. ``docker run --rm`` argument tells Docker to delete the container after it runs, because each container is only good for a single job. The Docker image is not designed for use as a persistent web service or for use on Amazon EC2 Container Service (AWS ECS).
 
 Installing on Windows
 ---------------------
 
-Direct installation on Windows is not possible.  Install the _`Docker` container as described above.  Ensure that your command prompt can run the docker "hello world" container.
+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.
 
 It would probably not be too difficult to run 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.
 
@@ -358,7 +426,11 @@ where /c/Users/sampleuser is a Unix representation of the Windows path C:\\Users
 Installing with Python pip
 --------------------------
 
-First, install `your platform's version <https://repology.org/metapackage/ocrmypdf/versions>`_ of ``ocrmypdf``, if available, as a way of ensuring that external dependencies are (mostly) satisified, even though the platform version may be out of date. Use ``ocrmypdf --version`` to confirm what version was installed.
+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.
+
+Use ``ocrmypdf --version`` to confirm what version was installed.
 
 Then you can install the latest OCRmyPDF from the Python wheels. First try:
 
@@ -377,27 +449,36 @@ Since ``pip3 install --user`` does not work correctly on some platforms, notably
 Requirements for pip and HEAD install
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
-OCRmyPDF currently requires these external programs to be installed:
+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 3.5 or newer
-- Tesseract 3.04 or newer
 - Ghostscript 9.15 or newer
-- qpdf 7.0.0 or newer
+- libexempi3 2.2.0 or newer
+- qpdf 8.1.0 or newer
+- Tesseract 3.04 or newer
 
-The following dependencies are recommended:
+As of ocrmypdf 7.0.0, the following versions are recommended:
 
-- Python 3.6
-- Tesseract 4.00 or newer
-- Ghostscript 9.22 or newer
-- qpdf 8.0.2 or newer
+- Python 3.7
+- Ghostscript 9.23
+- libexempi3 2.4.5
+- qpdf 8.1.0
+- Tesseract 4.0.0-beta1
+- jbig2enc 0.29 or newer
+- pngquant 2.5 or newer
 - unpaper 6.1
-- PyMuPDF 1.12.5 or newer
+
+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`.
+
+**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.
 
-Python 3.6 and Tesseract 4.0.0-beta.1 are recommended for best OCR results and best performance.
-
-The library PyMuPDF is not widely available in platform distributions, and it improves OCRmyPDF in certain conditions. Consider installing OCRmyPDF from the Python binary wheels, which include a precompiled version of this library.
+Python 3.7 and Tesseract 4.0.0-beta.1 are recommended for best OCR results and best performance.
 
 
 Installing HEAD revision from sources
@@ -405,6 +486,8 @@ Installing HEAD revision from sources
 
 If you have ``git`` and Python 3.5 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.
+
 To install the HEAD revision from sources in the current Python 3 environment:
 
 .. code-block:: bash
@@ -427,8 +510,8 @@ 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 after
-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:
 
@@ -441,8 +524,18 @@ need to be installed. The script requires specific versions of the
 dependencies. Older version than the ones mentioned in the release notes
 are likely not to be compatible to OCRmyPDF.
 
+For development
+~~~~~~~~~~~~~~~
 
-Other Linux packages
---------------------
+To install all of the development and test requirements:
 
-See the `Repology <https://repology.org/metapackage/ocrmypdf/versions>`_ page.
+.. code-block:: bash
+
+    git clone -b master https://github.com/jbarlow83/OCRmyPDF.git
+    python3 -m venv
+    source venv/bin/activate
+    cd OCRmyPDF
+    pip install -e .
+    pip install -r dev_requirements.txt -r test_requirements.txt
+
+To add JBIG2 encoding, see :ref:`jbig2`.

docs/introduction.rst

@@ -1,9 +1,7 @@
 Introduction
 ============
-
 OCRmyPDF is a Python 3 package that adds OCR layers to PDFs.
 
-
 About OCR
 ---------
 
@@ -66,6 +64,8 @@ In the case of a PDF that is nothing other than a container of images (no rotati
 
 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. 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
 -----------
@@ -78,7 +78,6 @@ OCRmyPDF is limited by the Tesseract OCR engine.  As such it experiences these l
 * 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 join text across the columns.
 * Poor quality scans may produce poor quality OCR. Garbage in, garbage out.
-* PDFs that use transparent layers are not currently checked in the test suite, so they may not work correctly.
 * It does not expose information about what font family text belongs to.
 
 OCRmyPDF is also limited by the PDF specification:
@@ -91,9 +90,11 @@ 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.
-  
-OCRmyPDF is currently not designed to be used as a Python API; it is designed to be run as a command line tool. ``import ocrmypf`` currently attempts to process the command line on ``sys.argv`` at import time so it has side effects that will interfere with its use as a package. The API it presents should not be considered stable.
 
+Regarding OCRmyPDF itself:
+
+* PDFs that use transparency are not currently represented in the test suite
+* The Python API exported by ``import ocrmypdf`` is design to help scripts that use OCRmyPDF but is not currently capable of running OCRmyPDF jobs due to limitations in an underlying library.
 
 Similar programs
 ----------------
@@ -112,4 +113,3 @@ Web front-ends
 * `OCRmyPDF-web <https://github.com/sseemayer/OCRmyPDF-web>`_, a micro web-frontend for OCRmyPDF (third-party, not actively maintained)
 
 Bear in mind that OCRmyPDF is not designed to be secure against malware-bearing PDFs (see `Using OCRmyPDF online`_).
-

docs/jbig2.rst

@@ -0,0 +1,22 @@
+.. _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.
+
+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.
+
+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

docs/release_notes.rst

@@ -13,12 +13,72 @@ Note that it is licensed under GPLv3, so scripts that ``import ocrmypdf`` and ar
    find:    [^`]\#([0-9]{1,3})[^0-9]
    replace: `#$1 <https://github.com/jbarlow83/OCRmyPDF/issues/$1>`_
 
+v7.1.0
+------
+
+-   Improve the performance of initial text extraction, which is done to determine if a file contains existing text of some kind or not. On large files, this initial processing is now about 20x times faster. (`#299 <https://github.com/jbarlow83/OCRmyPDF/issues/299>`_)
+
+-   pikepdf 0.3.3 is now required.
+
+-   Fixed issue `#231 <https://github.com/jbarlow83/OCRmyPDF/issues/231>`_, a problem with JPEG2000 images where image metadata was only available inside the JPEG2000 file.
+
+-   Fixed some additional Ghostscript 9.25 compatibility issues.
+
+-   Improved handling of KeyboardInterrupt error messages. (`#301 <https://github.com/jbarlow83/OCRmyPDF/issues/301>`_)
+
+-   README.md is now served in GitHub markdown instead of reStructuredText.
+
+v7.0.6
+------
+
+-   Blacklist Ghostscript 9.24, now that 9.25 is available and fixes many regressions in 9.24.
+
+
+v7.0.5
+------
+
+-   Improve capability with Ghostscript 9.24, and enable the JPEG passthrough feature when this version in installed.
+
+-   Ghostscript 9.24 lost the ability to set PDF title, author, subject and keyword metadata to Unicode strings. OCRmyPDF will set ASCII strings and warn when Unicode is suppressed. Other software may be used to update metadata. This is a short term work around.
+
+-   PDFs generated by Kodak Capture Desktop, or generally PDFs that contain indirect references to null objects in their table of contents, would have an invalid table of contents after processing by OCRmyPDF that might interfere with other viewers. This has been fixed.
+
+-   Detect PDFs generated by Adobe LiveCycle, which can only be displayed in Adobe Acrobat and Reader currently. When these are encountered, exit with an error instead of performing OCR on the "Please wait" error message page.
+
+v7.0.4
+------
+
+-   Fix exception thrown when trying to optimize a certain type of PNG embedded in a PDF with the ``-O2``
+
+-   Update to pikepdf 0.3.2, to gain support for optimizing some additional image types that were previously excluded from optimization (CMYK and grayscale). Fixes `#285 <https://github.com/jbarlow83/OCRmyPDF/issues/285>`_.
+
+v7.0.3
+------
+
+-   Fix issue `#284 <https://github.com/jbarlow83/OCRmyPDF/issues/284>`_, an error when parsing inline images that have are also image masks, by upgrading pikepdf to 0.3.1
+
+v7.0.2
+------
+
+-   Fix a regression with ``--rotate-pages`` on pages that already had rotations applied. (`#279 <https://github.com/jbarlow83/OCRmyPDF/issues/279>`_)
+
+-   Improve quality of page rotation in some cases by rasterizing a higher quality preview image. (`#281 <https://github.com/jbarlow83/OCRmyPDF/issues/281>`_)
+
+v7.0.1
+------
+
+-   Fix compatibility with img2pdf >= 0.3.0 by rejecting input images that have an alpha channel
+
+-   Add forward compatibility for pikepdf 0.3.0 (unrelated to img2pdf)
+
+-   Various documentation updates for v7.0.0 changes
+
 v7.0.0
 ------
 
 -   The core algorithm for combining OCR layers with existing PDF pages has been rewritten and improved considerably.  PDFs are no longer split into single page PDFs for processing; instead, images are rendered and the OCR results are grafted onto the input PDF.  The new algorithm uses less temporary disk space and is much more performant especially for large files.
 
--   New dependency: `pikepdf <https://github.com/pikepdf/pikepdf>`_. pikepdf is a  powerful new Python PDF library driving the latest OCRmyPDF features, built on the QPDF C++ library (libqpdf).
+-   New dependency: `pikepdf <https://github.com/pikepdf/pikepdf>`_. pikepdf is a powerful new Python PDF library driving the latest OCRmyPDF features, built on the QPDF C++ library (libqpdf).
 
 -   New feature: PDF optimization with ``-O`` or ``--optimize``.  After OCR, OCRmyPDF will perform image optimizations relevant to OCR PDFs.
 
@@ -48,8 +108,7 @@ v7.0.0
 
     +   ``PyMuPDF``
 
--   The ``sandwich`` PDF renderer can be used with all supported versions of Tesseract, including that those prior to v3.05 which don't support
-``-c textonly``. (Tesseract v4.0.0 is recommended and more efficient.)
+-   The ``sandwich`` PDF renderer can be used with all supported versions of Tesseract, including that those prior to v3.05 which don't support ``-c textonly``. (Tesseract v4.0.0 is recommended and more efficient.)
 
 -   ``--pdf-renderer auto`` option and the diagnostics used to select a PDF renderer now work better with old versions, but may make different decisions than past versions.
 
@@ -61,6 +120,26 @@ v7.0.0
 
     +   It may be necessary to separately ``pip install pycparser`` to avoid `another Python 3.7 issue <https://github.com/eliben/pycparser/pull/135>`_.
 
+v6.2.4
+------
+
+-   Backport Ghostscript 9.25 compatibility fixes, which removes support for setting Unicode metadata
+-   Backport blacklisting Ghostscript 9.24
+-   Older versions of Ghostscript are still supported
+
+v6.2.3
+------
+
+-   Fix compatibility with img2pdf >= 0.3.0 by rejecting input images that have an alpha channel
+-   This version will be included in Ubuntu 18.10
+
+v6.2.2
+------
+
+-   Backport compatibility fixes for Python 3.7 and ruffus 2.7.0 from v7.0.0
+-   Backport fix to ignore masks when deciding what colors are on a page
+-   Backport some minor improvements from v7.0.0: better argument validation and warnings about the Tesseract 4.0.0 ``--user-words`` regression
+
 v6.2.1
 ------
 
@@ -176,7 +255,7 @@ v6.0.0
 v5.7.0
 ------
 
--   Fixed an issue that caused poor CPU utilization on machines more than 4 cores when running Tesseract 4. (Related to issue `#217 <https://github.com/jbarlow83/OCRmyPDF/issues/217>`_.)
+-   Fixed an issue that caused poor CPU utilization on machines with more than 4 cores when running Tesseract 4. (Related to issue `#217 <https://github.com/jbarlow83/OCRmyPDF/issues/217>`_.)
 
 -   The 'hocr' renderer has been improved. The 'sandwich' and 'tesseract' renderers are still better for most use cases, but 'hocr' may be useful for people who work with the PDF.js renderer in English/ASCII languages. (`#225 <https://github.com/jbarlow83/OCRmyPDF/issues/225>`_)
 
@@ -315,7 +394,6 @@ v5.0.1
 
 -   Fixed issue `#169 <https://github.com/jbarlow83/OCRmyPDF/issues/169>`_, exception due to failure to create sidecar text files on some versions of Tesseract 3.04, including the jbarlow83/ocrmypdf Docker image
 
-
 v5.0
 ----
 
@@ -331,7 +409,6 @@ v5.0
 -   Fixed issue `#164 <https://github.com/jbarlow83/OCRmyPDF/issues/164>`_, a typo
 -   Removed the command line parameters ``-n`` and ``--just-print`` since they have not worked for some time (reported as Ubuntu bug `#1687308 <https://bugs.launchpad.net/ubuntu/+source/ocrmypdf/+bug/1687308>`_)
 
-
 v4.5.6
 ------
 
@@ -339,21 +416,18 @@ v4.5.6
 -   Fixed issue `#158 <https://github.com/jbarlow83/OCRmyPDF/issues/158>`_, ocrmypdf now stops and terminates if Ghostscript fails on an intermediate step, as it is not possible to proceed.
 -   Fixed issue `#160 <https://github.com/jbarlow83/OCRmyPDF/issues/160>`_, exception thrown on certain invalid arguments instead of error message
 
-
 v4.5.5
 ------
 
 -   Automated update of macOS homebrew tap
 -   Fixed issue `#154 <https://github.com/jbarlow83/OCRmyPDF/issues/154>`_, KeyError '/Contents' when searching for text on blank pages that have no /Contents record.  Note: incomplete fix for this issue.
 
-
 v4.5.4
 ------
 
 -   Fix ``--skip-big`` raising an exception if a page contains no images (`#152 <https://github.com/jbarlow83/OCRmyPDF/issues/152>`_) (thanks to @TomRaz)
 -   Fix an issue where pages with no images might trigger "cannot write mode P as JPEG" (`#151 <https://github.com/jbarlow83/OCRmyPDF/issues/151>`_)
 
-
 v4.5.3
 ------
 
@@ -362,7 +436,6 @@ v4.5.3
 -   Fixed an issue where the ``tess4`` renderer would duplicate content onto output pages if tesseract failed or timed out
 -   Fixed ``tess4`` renderer not recognized when lossless reconstruction is possible
 
-
 v4.5.2
 ------
 
@@ -370,21 +443,18 @@ v4.5.2
 -   Make "using Tesseract 4.0" warning less ominous
 -   Set up machinery for homebrew OCRmyPDF tap
 
-
 v4.5.1
 ------
 
 -   Fix issue `#137 <https://github.com/jbarlow83/OCRmyPDF/issues/137>`_, proportions of images with a non-square pixel aspect ratio would be distorted in output for ``--force-ocr`` and some other combinations of flags
 
-
 v4.5
 ----
 
--   Exotic PDFs containing "Form XObjects" are now supported (issue `#134 <https://github.com/jbarlow83/OCRmyPDF/issues/134>`_; PDF reference manual 8.10), and images they contain are taken into account when determining the resolution for rasterizing
+-   PDFs containing "Form XObjects" are now supported (issue `#134 <https://github.com/jbarlow83/OCRmyPDF/issues/134>`_; PDF reference manual 8.10), and images they contain are taken into account when determining the resolution for rasterizing
 -   The Tesseract 4 Docker image no longer includes all languages, because it took so long to build something would tend to fail
 -   OCRmyPDF now warns about using ``--pdf-renderer tesseract`` with Tesseract 3.04 or lower due to issues with Ghostscript corrupting the OCR text in these cases
 
-
 v4.4.2
 ------
 
@@ -395,7 +465,6 @@ v4.4.2
 
 -   OCRmyPDF now prevents running the Tesseract 4 renderer with Tesseract 3.04, which was permitted in v4.4 and v4.4.1 but will not work
 
-
 v4.4.1
 ------
 
@@ -422,32 +491,27 @@ v4.4
 
     + However, OCRmyPDF's dependency "ruffus" is not re-entrant, so no Python API is available. Scripts should continue to use the command line interface.
 
-
 v4.3.5
 ------
 
 -   Update documentation to confirm Python 3.6.0 compatibility. No code changes were needed, so many earlier versions are likely supported.
 
-
 v4.3.4
 ------
 
 -   Fixed "decimal.InvalidOperation: quantize result has too many digits" for high DPI images
 
-
 v4.3.3
 ------
 
 -   Fixed PDF/A creation with Ghostscript 9.20 properly
 -   Fixed an exception on inline stencil masks with a missing optional parameter
 
-
 v4.3.2
 ------
 
 -   Fixed a PDF/A creation issue with Ghostscript 9.20 (note: this fix did not actually work)
 
-
 v4.3.1
 ------
 
@@ -458,7 +522,6 @@ v4.3.1
 -   Cleaned up debug output
 -   Tesseract caching in test cases is now more cautious about false cache hits and reproducing exact output, not that any problems were observed
 
-
 v4.3
 ----
 
@@ -476,7 +539,6 @@ v4.2.5
 -   Fixed an issue (`#100 <https://github.com/jbarlow83/OCRmyPDF/issues/100>`_) with PDFs that omit the optional /BitsPerComponent parameter on images
 -   Removed non-free file milk.pdf
 
-
 v4.2.4
 ------
 
@@ -496,20 +558,17 @@ v4.2.3
 -   Added a test case for indirect objects and linearized PDFs
 -   Deprecated the OCRmyPDF.sh shell script
 
-
 v4.2.2
 ------
 
 -   Improvements to documentation
 
-
 v4.2.1
 ------
 
 -   Fixed an issue where PDF pages that contained stencil masks would report an incorrect DPI and cause Ghostscript to abort
 -   Implemented stdin streaming
 
-
 v4.2
 ----
 
@@ -548,7 +607,6 @@ v4.1.4
 
 -   Bug fix: monochrome images with an ICC profile attached were incorrectly converted to full color images if lossless reconstruction was not possible due to other settings; consequence was increased file size for these images
 
-
 v4.1.3
 ------
 
@@ -557,14 +615,12 @@ v4.1.3
 -   Fix order of operations for matrix multiplication (no effect on most users)
 -   Add a few leptonica wrapper functions (no effect on most users)
 
-
 v4.1.2
 ------
 
 -   Replace IEC sRGB ICC profile with Debian's sRGB (from icc-profiles-free) which is more compatible with the MIT license
 -   More helpful error message for an error related to certain types of malformed PDFs
 
-
 v4.1
 ----
 
@@ -572,20 +628,17 @@ v4.1
 -   Fixed problems in error checking if ``unpaper`` is uninstalled or missing at run-time
 -   Fixed problems with "RethrownJobError" errors during error handling that suppressed the useful error messages
 
-
 v4.0.7
 ------
 
 -   Minor correction to Ghostscript output settings
 
-
 v4.0.6
 ------
 
 -   Update install instructions
 -   Provide a sRGB profile instead of using Ghostscript's
 
-
 v4.0.5
 ------
 
@@ -599,18 +652,14 @@ v4.0.4
 
 Released with verbose debug message turned on. Do not use. Skip to v4.0.5.
 
-
 v4.0.3
 ------
 
 New features
-^^^^^^^^^^^^
 
 -   Page orientations detected are now reported in a summary comment
 
-
 Fixes
-^^^^^
 
 -   Show stack trace if unexpected errors occur
 -   Treat "too few characters" error message from Tesseract as a reason to skip that page rather than
@@ -622,7 +671,7 @@ v4.0.2
 ------
 
 Fixes
-^^^^^
+
 
 -   Fixed compatibility with Tesseract 3.04.01 release, particularly its different way of outputting
     orientation information
@@ -634,7 +683,7 @@ v4.0.1
 ------
 
 Fixes
-^^^^^
+
 
 -   Fixed a KeyError if tesseract fails to find page orientation information
 
@@ -643,7 +692,6 @@ v4.0
 ----
 
 New features
-^^^^^^^^^^^^
 
 -   Automatic page rotation (``-r``) is now available. It uses ignores any prior rotation information
     on PDFs and sets rotation based on the dominant orientation of detectable text. This feature is
@@ -653,7 +701,6 @@ New features
 
 
 Fixes
-^^^^^
 
 -   Fixed an issue where lossless reconstruction could cause some pages to be appear incorrectly
     if the page was rotated by the user in Acrobat after being scanned (specifically if it a /Rotate tag)
@@ -662,7 +709,6 @@ Fixes
 
 
 Changes
-^^^^^^^
 
 -   Logging output is now much easier to read
 -   ``--deskew`` is now performed by Leptonica instead of unpaper (`#25 <https://github.com/jbarlow83/OCRmyPDF/issues/25>`_)
@@ -678,7 +724,6 @@ v3.2.1
 ------
 
 Changes
-^^^^^^^
 
 -   Fixed issue `#47 <https://github.com/jbarlow83/OCRmyPDF/issues/47>`_ "convert() got and unexpected keyword argument 'dpi'" by upgrading to img2pdf 0.2
 -   Tweaked the Dockerfiles
@@ -688,7 +733,6 @@ v3.2
 ----
 
 New features
-^^^^^^^^^^^^
 
 -   Lossless reconstruction: when possible, OCRmyPDF will inject text layers without
     otherwise manipulating the content and layout of a PDF page. For example, a PDF containing a mix
@@ -700,7 +744,6 @@ New features
     for the polyglots among us. It is much larger.
 
 Changes
-^^^^^^^
 
 -   JPEG transcoding quality is now 95 instead of the default 75. Bigger file sizes for less degradation.
 
@@ -710,7 +753,6 @@ v3.1.1
 ------
 
 Changes
-^^^^^^^
 
 -   Fixed bug that caused incorrect page size and DPI calculations on documents with mixed page sizes
 
@@ -718,7 +760,6 @@ v3.1
 ----
 
 Changes
-^^^^^^^
 
 -   Default output format is now PDF/A-2b instead of PDF/A-1b
 -   Python 3.5 and macOS El Capitan are now supported platforms - no changes were
@@ -736,7 +777,6 @@ v3.0
 ----
 
 New features
-^^^^^^^^^^^^
 
 -   Easier installation with a Docker container or Python's ``pip`` package manager
 -   Eliminated many external dependencies, so it's easier to setup
@@ -759,7 +799,6 @@ New features
 -   Multiple images on the same PDF page are now supported
 
 Changes
-^^^^^^^
 
 -   New, robust rewrite in Python 3.4+ with ruffus_ pipelines
 -   Now uses Ghostscript 9.14's improved color conversion model to preserve PDF colors
@@ -796,8 +835,7 @@ Changes
 .. _Unpaper: https://github.com/Flameeyes/unpaper
 .. _JHOVE: http://jhove.sourceforge.net/
 
-Release candidates
-^^^^^^^^^^^^^^^^^^
+Release candidates^
 
 -   rc9:
 
@@ -865,12 +903,11 @@ where ``settings.txt`` contains *one argument per line*, for example:
 
 
 Fixes
-^^^^^
+
 
 -   Handling of filenames containing spaces: fixed
 
 Notes and known issues
-^^^^^^^^^^^^^^^^^^^^^^
 
 -   Some dependencies may work with lower versions than tested, so try
     overriding dependencies if they are "in the way" to see if they work.

requirements.txt

@@ -2,10 +2,10 @@
 # setup.py lists a separate set of requirements that are looser to simplify
 # installation
 cffi == 1.11.5
-img2pdf == 0.2.4
-pikepdf == 0.2.2
+img2pdf == 0.3.0
+pikepdf == 0.3.3
 Pillow >= 5.0.0, != 5.1.0 ; sys_platform == "darwin"
 pycparser == 2.18
 python-xmp-toolkit == 2.0.1
 reportlab == 3.4.0
-ruffus == 2.7
+ruffus == 2.7.0

setup.py

@@ -205,13 +205,14 @@ tests_require = open('test_requirements.txt').read().splitlines()
 
 
 def readme():
-    with open('README.rst') as f:
+    with open('README.md') as f:
         return f.read()
 
 setup(
     name='ocrmypdf',
     description='OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to be searched',
     long_description=readme(),
+    long_description_content_type='text/markdown',
     url='https://github.com/jbarlow83/OCRmyPDF',
     author='James R. Barlow',
     author_email='jim@purplerock.ca',
@@ -241,6 +242,7 @@ setup(
         'cffi >= 1.9.1',        # to build the leptonica module
         'pytest-runner',        # to enable python setup.py test
         'setuptools_scm',       # so that version will work
+        'setuptools_scm_git_archive'    # enable version from github tarballs
     ],
     use_scm_version={'version_scheme': 'post-release'},
     cffi_modules=[
@@ -248,14 +250,14 @@ setup(
     ],
     install_requires=[
         'cffi >= 1.9.1',          # must be a setup and install requirement
-        'img2pdf >= 0.2.4',       # pure Python, so track HEAD closely
-        'pikepdf >= 0.2.2, < 0.3',
+        'img2pdf >= 0.2.4, < 0.4',       # pure Python, so track HEAD closely
+        'pikepdf >= 0.3.3, < 0.4',
         '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
         'python-xmp-toolkit >= 2, < 3',
         'reportlab >= 3.3.0',     # oldest released version with sane image handling
-        'ruffus >= 2.7',
+        'ruffus >= 2.7.0',
     ],
     extras_require={
         'fitz': [],  # Backward compatibility
@@ -268,4 +270,10 @@ setup(
     },
     package_data={'ocrmypdf': ['data/sRGB.icc']},
     include_package_data=True,
-    zip_safe=False)
+    zip_safe=False,
+    project_urls={
+        'Documentation': 'https://ocrmypdf.readthedocs.io/',
+        'Source': 'https://github.com/jbarlow83/ocrmypdf',
+        'Tracker': 'https://github.com/jbarlow83/ocrmypdf/issues'
+    }
+)

src/ocrmypdf/__main__.py

@@ -17,7 +17,6 @@
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
 from tempfile import mkdtemp
-from collections.abc import Sequence
 from pathlib import Path
 import sys
 import os
@@ -36,7 +35,7 @@ import ruffus.proxy_logger as proxy_logger
 from ._jobcontext import JobContext, JobContextManager, cleanup_working_files
 from ._pipeline import build_pipeline
 from .pdfa import file_claims_pdfa
-from .helpers import is_iterable_notstr, re_symlink, is_file_writable, \
+from .helpers import re_symlink, is_file_writable, \
     available_cpu_count
 from .exec import tesseract, qpdf, ghostscript
 from . import PROGRAM_NAME, VERSION
@@ -598,18 +597,24 @@ def do_ruffus_exception(ruffus_five_tuple, options, log):
     description of the error message that occurred."""
     exit_code = None
 
-    task_name, job_name, exc_name, exc_value, exc_stack = ruffus_five_tuple
-    task_name = task_name  # unused
-    job_name = job_name  # unused
-    if exc_name == 'builtins.SystemExit':
+    _task_name, _job_name, exc_name, exc_value, exc_stack = ruffus_five_tuple
+
+    if isinstance(exc_name, type):
+        # ruffus is full of mystery... sometimes (probably when the process
+        # group leader is killed) exc_name is the class object of the exception,
+        # rather than a str. So reach into the object and get its name.
+        exc_name = exc_name.__name__
+
+    if exc_name in ('builtins.SystemExit', 'SystemExit'):
         match = re.search(r"\.(.+?)\)", exc_value)
         exit_code_name = match.groups()[0]
         exit_code = getattr(ExitCode, exit_code_name, 'other_error')
     elif exc_name == 'ruffus.ruffus_exceptions.MissingInputFileError':
         log.error(cleanup_ruffus_error_message(exc_value))
         exit_code = ExitCode.input_file
-    elif exc_name == 'builtins.KeyboardInterrupt':
-        log.error("Interrupted by user")
+    elif exc_name in ('builtins.KeyboardInterrupt', 'KeyboardInterrupt'):
+        # We have to print in this case because the log daemon might be toast
+        print("Interrupted by user", file=sys.stderr)
         exit_code = ExitCode.ctrl_c
     elif exc_name == 'subprocess.CalledProcessError':
         # It's up to the subprocess handler to report something useful
@@ -751,6 +756,7 @@ def preamble(_log):
     _log.debug('ocrmypdf ' + VERSION)
     _log.debug('tesseract ' + tesseract.version())
     _log.debug('qpdf ' + qpdf.version())
+    _log.debug('gs ' + ghostscript.version())
 
 
 def check_environ(options, _log):
@@ -835,9 +841,11 @@ def report_output_file_size(options, _log, input_file, output_file):
         """.format(ratio, explanation)))
 
 
-def run_pipeline():
-    options = parser.parse_args()
+def run_pipeline(args=None):
+    options = parser.parse_args(args=args)
     options.verbose_abbreviated_path = 1
+    if os.environ.get('_OCRMYPDF_THREADS'):
+        options.use_threads = True
 
     if not check_closed_streams(options):
         return ExitCode.bad_args
@@ -858,6 +866,14 @@ def run_pipeline():
             "security vulnerabilities with certain malformed PDFs. Consider "
             "upgrading to version 7.0.0 or newer.".format(qpdf.version()))
 
+    if ghostscript.version() == '9.24':
+        complain(
+            "Ghostscript 9.24 contains serious regressions and is not "
+            "supported. Please upgrade to Ghostscript 9.25 or use an older "
+            "version."
+        )
+        return ExitCode.missing_dependency
+
     # Any changes to options will not take effect for options that are already
     # bound to function parameters in the pipeline. (For example
     # options.input_file, options.pdf_renderer are already bound.)

src/ocrmypdf/_pipeline.py

@@ -19,7 +19,6 @@ from contextlib import suppress
 from shutil import copyfileobj, copyfile
 from pathlib import Path
 from datetime import datetime, timezone
-
 import sys
 import os
 import re
@@ -85,6 +84,13 @@ def triage_image_file(input_file, output_file, log, options):
                 "image was scanned and specify it using --image-dpi.")
             raise DpiError()
 
+        if im.mode in ('RGBA', 'LA'):
+            log.error(
+                "The input image has an alpha channel. Remove the alpha "
+                "channel first."
+            )
+            raise UnsupportedImageFormatError()
+
         if 'iccprofile' not in im.info:
             if im.mode == 'RGB':
                 log.info('Input image has no ICC profile, assuming sRGB')
@@ -157,13 +163,20 @@ def repair_and_parse_pdf(
     copyfile(input_file, output_file)
 
     try:
-        pdfinfo = PdfInfo(output_file)
+        pdfinfo = PdfInfo(output_file, log=log)
     except pikepdf.PasswordError as e:
         raise EncryptedPdfError()
     except pikepdf.PdfError as e:
         log.error(e)
         raise InputFileError()
 
+    if pdfinfo.needs_rendering:
+        log.error(
+            "This PDF contains dynamic XFA forms created by Adobe LiveCycle "
+            "Designer and can only be read by Adobe Acrobat or Adobe Reader."
+        )
+        raise InputFileError()
+
     if pdfinfo.has_userunit and options.output_type.startswith('pdfa'):
         log.error(
             "This input file uses a PDF feature that is not supported "
@@ -312,16 +325,6 @@ def marker_pages(
         page.symlink_to(input_file)  # pylint: disable=E1101
 
 
-def split_page(
-        placeholder_file,
-        output_file,
-        log,
-        context):
-    pageno = page_number(placeholder_file) - 1
-    input_pdf = context.get_pdfinfo().filename
-    qpdf.extract_page(input_pdf, output_file, pageno)
-
-
 def ocr_or_skip(
         input_files,
         output_files,
@@ -353,13 +356,13 @@ def rasterize_preview(
         context):
     pageinfo = get_pageinfo(input_file, context)
     options = context.get_options()
-    canvas_dpi = get_canvas_square_dpi(pageinfo, options) / 2
-    page_dpi = get_page_square_dpi(pageinfo, options) / 2
+    canvas_dpi = get_canvas_square_dpi(pageinfo, options)
+    page_dpi = get_page_square_dpi(pageinfo, options)
 
     ghostscript.rasterize_pdf(
         input_file, output_file, xres=canvas_dpi, yres=canvas_dpi,
         raster_device='jpeggray', log=log, page_dpi=(page_dpi, page_dpi),
-        pageno=page_number(input_file), rotation=pageinfo.rotation)
+        pageno=page_number(input_file))
 
 
 def orient_page(
@@ -367,6 +370,22 @@ def orient_page(
         output_file,
         log,
         context):
+    """
+    Work out orientation correct for each page.
+
+    We ask Ghostscript to draw a preview page, which will rasterize with the
+    current /Rotate applied, and then ask Tesseract which way the page is
+    oriented. If the value of /Rotate is correct (e.g., a user already
+    manually fixed rotation), then Tesseract will say the page is pointing
+    up and the correction is zero. Otherwise, the orientation found by
+    Tesseract represents the clockwise rotation, or the counterclockwise
+    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
+    the image and text layers.
+
+    """
 
     options = context.get_options()
     page_pdf = next(ii for ii in infiles if ii.endswith('.page.pdf'))
@@ -389,33 +408,43 @@ def orient_page(
         270: '⇦'
     }
 
+    pageno = page_number(page_pdf) - 1
+    pdfinfo = context.get_pdfinfo()
+    existing_rotation = pdfinfo[pageno].rotation
+
+    correction = orient_conf.angle % 360
+
     apply_correction = False
-    description = ''
+    action = ''
     if orient_conf.confidence >= options.rotate_pages_threshold:
-        if orient_conf.angle != 0:
+        if correction != 0:
             apply_correction = True
-            description = ' - will rotate'
+            action = ' - will rotate'
         else:
-            description = ' - rotation appears correct'
+            action = ' - rotation appears correct'
     else:
-        if orient_conf.angle != 0:
-            description = ' - confidence too low to rotate'
+        if correction != 0:
+            action = ' - confidence too low to rotate'
         else:
-            description = ' - no change'
+            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, '?'))
 
     log.info(
-        '{0:4d}: page is facing {1}, confidence {2:.2f}{3}'.format(
-            page_number(preview),
-            direction.get(orient_conf.angle, '?'),
-            orient_conf.confidence,
-            description)
+        '{pagenum:4d}: {facing}, confidence {conf:.2f}{action}'.format(
+            pagenum=page_number(preview),
+            facing=facing,
+            conf=orient_conf.confidence,
+            action=action)
     )
 
     re_symlink(page_pdf, output_file, log)
     if apply_correction:
-        pageno = page_number(page_pdf) - 1
-        pdfinfo = context.get_pdfinfo()
-        correction = (orient_conf.angle - pdfinfo[pageno].rotation) % 360
         context.set_rotation(pageno, correction)
 
 
@@ -751,7 +780,20 @@ def generate_postscript_stub(
     options = context.get_options()
     pdf = pikepdf.open(input_file)
     pdfmark = get_pdfmark(pdf, options)
-    generate_pdfa_ps(output_file, pdfmark)
+
+    ascii_docinfo = False
+    if ghostscript.version() >= '9.24':
+        ascii_docinfo = True
+        try:
+            for v in pdfmark.values():
+                v.encode('ascii', errors='strict')
+        except UnicodeEncodeError:
+            log.warning(
+                "Ghostscript 9.24+ does not support Unicode strings in "
+                " metadata. These will be converted to ASCII if possible."
+            )
+
+    generate_pdfa_ps(output_file, pdfmark, ascii_docinfo=ascii_docinfo)
 
 
 def metadata_fixup(

src/ocrmypdf/_weave.py

@@ -137,14 +137,9 @@ def _find_font(text, pdf_base):
     return font, font_key
 
 
-def _fix_toc(pdf_base, pageref_remap, log):
-    """Repair the table of contents
-
-    Whenever we replace a page wholesale, it gets assigned a new objgen number
-    and other references to it within the PDF become invalid, most notably in
-    the table of contents (/Outlines in PDF-speak).  In weave_layers we collect
-    pageref_remap, a mapping that describes the new objgen number given an old
-    one.  (objgen is a tuple, and the gen is almost always zero.)
+def _traverse_toc(pdf_base, visitor_fn, log):
+    """
+    Walk the table of contents, calling visitor_fn() at each node
 
     The /Outlines data structure is a messy data structure, but rather than
     navigating hierarchically we just track unique nodes.  Enqueue nodes when
@@ -152,8 +147,6 @@ def _fix_toc(pdf_base, pageref_remap, log):
     the two types of object in the table of contents that can be page bookmarks
     and update the page entry.
 
-    It may ultimately be better to find a way to rebuild a page in place.
-
     """
 
     visited = set()
@@ -162,10 +155,54 @@ def _fix_toc(pdf_base, pageref_remap, log):
 
     if not '/Outlines' in pdf_base.root:
         return
+
+    queue.add(pdf_base.root.Outlines.objgen)
+    while queue:
+        objgen = queue.pop()
+        visited.add(objgen)
+        node = pdf_base.get_object(objgen)
+        log.debug('fix toc: exploring outline entries at %r', objgen)
+
+        # Enumerate other nodes we could visit from here
+        for key in link_keys:
+            if key not in node:
+                continue
+            item = node[key]
+            if not item.is_indirect:
+                #  or not isinstance(item, pikepdf.Dictionary):
+                # # If there is garbage data, replace the key with an indirect
+                # # ref to None. Kodak Capture Desktop produces keys like these.
+                # log.error('Removing invalid reference from TOC: %s', repr(item))
+                # node[key] = pdf_base.make_indirect(None)
+                continue
+            objgen = item.objgen
+            if objgen not in visited:
+                queue.add(objgen)
+
+        visitor_fn(pdf_base, node, log)
+
+
+def _fix_toc(pdf_base, pageref_remap, log):
+    """Repair the table of contents
+
+    Whenever we replace a page wholesale, it gets assigned a new objgen number
+    and other references to it within the PDF become invalid, most notably in
+    the table of contents (/Outlines in PDF-speak).  In weave_layers we collect
+    pageref_remap, a mapping that describes the new objgen number given an old
+    one.  (objgen is a tuple, and the gen is almost always zero.)
+
+    It may ultimately be better to find a way to rebuild a page in place.
+
+    """
+
     if not pageref_remap:
         return
 
     def remap_dest(dest_node):
+        """
+        Inner helper function: change the objgen for any page from the old we
+        invalidated to its new one.
+        """
         if not isinstance(dest_node, pikepdf.Array):
             return
         pageref = dest_node[0]
@@ -174,30 +211,24 @@ def _fix_toc(pdf_base, pageref_remap, log):
             new_objgen = pageref_remap[pageref.objgen]
             dest_node[0] = pdf_base.get_object(new_objgen)
 
-    queue.add(pdf_base.root.Outlines.objgen)
-    while queue:
-        objgen = queue.pop()
-        visited.add(objgen)
-        node = pdf_base.get_object(objgen)
-        log.debug('fix toc: visiting %r', objgen)
-
-        # Enumerate other nodes we could visit from here
-        for key in link_keys:
-            if key not in node:
-                continue
-            item = node[key]
-            if not item.is_indirect:
-                continue
-            objgen = item.objgen
-            if objgen not in visited:
-                queue.add(objgen)
+    def visit_remap_dest(pdf_base, node, log):
+        """
+        Visitor function to fix ToC entries
 
+        Test for the two types of references to pages that can occur in ToCs.
+        Both types have the same final format (an indirect reference to the
+        target page).
+        """
         if '/Dest' in node:
+            # /Dest reference to another page (old method)
             remap_dest(node['/Dest'])
         elif '/A' in node:
+            # /A (action) command set to "GoTo" (newer method)
             if '/S' in node['/A'] and node['/A']['/S'] == '/GoTo':
                 remap_dest(node['/A']['/D'])
 
+    _traverse_toc(pdf_base, visit_remap_dest, log)
+
 
 def weave_layers(
         infiles,
@@ -247,6 +278,12 @@ def weave_layers(
     pdfinfo = context.get_pdfinfo()
     pagerefs = {}
 
+    # Walk the table of contents first, to trigger pikepdf/qpdf to resolve all
+    # page references in the table of contents. Some PDF generators put invalid
+    # references in the ToC, so we want to resolve them to null before we
+    # create any references, or the ToC will be corrupted
+    _traverse_toc(pdf_base, lambda *args: None, log)
+
     procset = pdf_base.make_indirect(
         pikepdf.Object.parse(b'[ /PDF /Text /ImageB /ImageC /ImageI ]'))
 

src/ocrmypdf/exec/ghostscript.py

@@ -32,15 +32,20 @@ def version():
 
 
 def jpeg_passthrough_available():
-    """
-    Ghostscript 9.23 introduced JPEG passthrough but it seems to corrupt the
-    last two bytes of certain images, for now we disable it for 9.23 and
-    do not mention it for < 9.23.
+    """Returns True if the installed version of Ghostscript supports JPEG passthru
 
+    Prior to 9.23, Ghostscript decode and re-encoded JPEGs internally. In 9.23
+    it gained the ability to keep JPEGs unmodified. However, the 9.23
+    implementation was buggy and would deletes the last two bytes of images in
+    some cases, as reported here.
     https://bugs.ghostscript.com/show_bug.cgi?id=699216
 
+    The issue was fixed for 9.24, hence that is the first version we consider
+    the feature available. (However, we don't use 9.24 at all, so the first
+    version that allows JPEG passthrough is 9.25.
+
     """
-    return False
+    return version() >= '9.24'
 
 
 def _gs_error_reported(stream):
@@ -48,8 +53,7 @@ def _gs_error_reported(stream):
 
 
 def extract_text(input_file, pageno=1):
-    """
-    Use the txtwrite device to get text layout information out
+    """Use the txtwrite device to get text layout information out
 
     For details on options of -dTextFormat see
     https://www.ghostscript.com/doc/current/VectorDevices.htm#TXT
@@ -60,10 +64,18 @@ def extract_text(input_file, pageno=1):
     <span bbox="left top right bottom" font="..." size="...">
     <char bbox="...." c="X"/>
 
+    :param pageno: number of page to extract, or all pages if None
     :return: XML-ish text representation in bytes
-
     """
 
+    if pageno is not None:
+        pages = [
+            '-dFirstPage=%i' % pageno,
+            '-dLastPage=%i' % pageno
+        ]
+    else:
+        pages = []
+
     args_gs = [
         'gs',
         '-dQUIET',
@@ -72,10 +84,9 @@ def extract_text(input_file, pageno=1):
         '-dNOPAUSE',
         '-sDEVICE=txtwrite',
         '-dTextFormat=0',
-        '-dFirstPage=%i' % pageno,
-        '-dLastPage=%i' % pageno,
+    ] + pages + [
         '-o', '-',
-        input_file
+        fspath(input_file)
     ]
 
     p = run(args_gs, stdout=PIPE, stderr=PIPE)
@@ -91,8 +102,7 @@ def extract_text(input_file, pageno=1):
 
 def rasterize_pdf(input_file, output_file, xres, yres, raster_device, log,
                   pageno=1, page_dpi=None, rotation=None):
-    """
-    Rasterize one page of a PDF at resolution (xres, yres) in canvas units.
+    """Rasterize one page of a PDF at resolution (xres, yres) in canvas units.
 
     The image is sized to match the integer pixels dimensions implied by
     (xres, yres) even if those numbers are noninteger. The image's DPI will
@@ -113,8 +123,6 @@ def rasterize_pdf(input_file, output_file, xres, yres, raster_device, log,
     if not page_dpi:
         page_dpi = res
 
-    autorotate = '/PageByPage' if rotation is None else '/None'
-
     with NamedTemporaryFile(delete=True) as tmp:
         args_gs = [
             'gs',
@@ -127,7 +135,7 @@ def rasterize_pdf(input_file, output_file, xres, yres, raster_device, log,
             '-dLastPage=%i' % pageno,
             '-r{0}x{1}'.format(str(int_res[0]), str(int_res[1])),
             '-o', tmp.name,
-            '-dAutoRotatePages=%s' % autorotate,
+            '-dAutoRotatePages=/None',  # Probably has no effect on raster
             '-f',
             fspath(input_file)
         ]
@@ -176,6 +184,23 @@ def rasterize_pdf(input_file, output_file, xres, yres, raster_device, log,
 
 def generate_pdfa(pdf_pages, output_file, compression, log,
                   threads=1, pdf_version='1.5', pdfa_part='2'):
+    """Generate a PDF/A.
+
+    The pdf_pages, a list files, will be merged into output_file. One or more
+    PDF files may be merged. One of the files in this list must be a pdfmark
+    file that provides Ghostscript with details on how to perform the PDF/A
+    conversion. By default with we pick PDF/A-2b, but this works for 1 or 3.
+
+    compression can be 'jpeg', 'lossless', or an empty string. In 'jpeg',
+    Ghostscript is instructed to convert color and grayscale images to DCT
+    (JPEG encoding). In 'lossless' Ghostscript is told to convert images to
+    Flate (lossless/PNG). If the parameter is omitted Ghostscript is left to
+    make its own decisions about how to encode images; it appears to use a
+    heuristic to decide how to encode images. As of Ghostscript 9.25, we
+    support passthrough JPEG which allows Ghostscript to avoid transcoding
+    images entirely. (The feature was added in 9.23 but broken, and the 9.24
+    release of Ghostscript had regressions, so we don't support it until 9.25.)
+    """
     compression_args = []
     if compression == 'jpeg':
         compression_args = [
@@ -209,6 +234,9 @@ def generate_pdfa(pdf_pages, output_file, compression, log,
         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",
@@ -218,8 +246,7 @@ def generate_pdfa(pdf_pages, output_file, compression, log,
             "-dNumRenderingThreads=" + str(threads),
             "-sDEVICE=pdfwrite",
             "-dAutoRotatePages=/None",
-            "-sColorConversionStrategy=" + strategy,
-            "-sProcessColorModel=DeviceRGB"
+            "-sColorConversionStrategy=" + strategy
         ] + compression_args + [
             "-dJPEGQ=95",
             "-dPDFA=" + pdfa_part,

src/ocrmypdf/leptonica.py

@@ -32,7 +32,7 @@ from functools import lru_cache
 from .lib._leptonica import ffi
 from .helpers import fspath
 
-# pylint: disable=w0212
+# pylint: disable=protected-access
 
 lept = ffi.dlopen(find_library('lept'))
 
@@ -45,12 +45,13 @@ def stderr(*objs):
 
 
 class _LeptonicaErrorTrap:
-    """Context manager to trap errors reported by Leptonica.
+    """
+    Context manager to trap errors reported by Leptonica.
 
-    Leptonica's error return codes are unreliable to the point of being
-    almost useless.  It does, however, write errors to stderr provided that is
-    not disabled at its compile time.  Fortunately this is done using error
-    macros so it is very self-consistent.
+    Leptonica's error return codes don't provide much informatino about what
+    went wrong. Leptonica does, however, write more detailed errors to stderr
+    (provided this is not disabled at compile time). The Leptonica source
+    code is very consistent in its use of macros to generate errors.
 
     This context manager redirects stderr to a temporary file which is then
     read and parsed for error messages.  As a side benefit, debug messages
@@ -116,7 +117,8 @@ class LeptonicaIOError(LeptonicaError):
 
 
 class Pix:
-    """Wrapper around leptonica's PIX object.
+    """
+    Wrapper around leptonica's PIX object.
 
     Leptonica uses referencing counting on PIX objects. Also, many Leptonica
     functions return the original object with an increased reference count

src/ocrmypdf/optimize.py

@@ -32,7 +32,7 @@ from .exec import pngquant, jbig2enc
 
 PAGE_GROUP_SIZE = 10
 DEFAULT_JPEG_QUALITY = 75
-DEFAULT_PNG_QUALITY = (65, 75)
+DEFAULT_PNG_QUALITY = 70
 
 
 def img_name(root, xref, ext):
@@ -240,8 +240,7 @@ def transcode_jpegs(pike, jpegs, root, log, options):
         compdata = leptonica.CompressedData.open(opt_jpg)
         im_obj = pike.get_object(xref, 0)
         im_obj.write(
-            compdata.read(), pikepdf.Name('/DCTDecode'),
-            pikepdf.Null()
+            compdata.read(), filter=pikepdf.Name('/DCTDecode')
         )
 
 
@@ -281,7 +280,7 @@ def transcode_pngs(pike, pngs, root, log, options):
         if len(compdata) > int(im_obj.stream_dict.Length):
             continue  # If we produced a larger image, don't use
 
-        predictor = pikepdf.Null()
+        predictor = None
         if compdata.predictor > 0:
             predictor = pikepdf.Dictionary({'/Predictor': compdata.predictor})
 
@@ -304,7 +303,10 @@ def transcode_pngs(pike, pngs, root, log, options):
             elif compdata.spp == 4:
                 cs = pikepdf.Name('/DeviceCMYK')
         im_obj.ColorSpace = cs
-        im_obj.write(compdata.read(), pikepdf.Name('/FlateDecode'), predictor)
+        im_obj.write(
+            compdata.read(),
+            filter=pikepdf.Name('/FlateDecode'), decode_parms=predictor
+        )
 
 
 def optimize(

src/ocrmypdf/pdfa.py

@@ -15,7 +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/>.
 
-# Generate a PDFA_def.ps file for Ghostscript >= 9.14
+"""
+Generate a PDFMARK file for Ghostscript >= 9.14, for PDF/A conversion
+
+pdfmark is an extension to the Postscript language that describes some PDF
+features like bookmarks and annotations. It was originally specified Adobe
+Distiller, for Postscript to PDF conversion:
+https://www.adobe.com/content/dam/acom/en/devnet/acrobat/pdfs/pdfmark_reference.pdf
+
+Ghostscript uses pdfmark for PDF to PDF/A conversion as well. To use Ghostscript
+to create a PDF/A, we need to create a pdfmark file with the necessary metadata.
+
+This takes care of the many version-specific bugs and pecularities in
+Ghostscript's handling of pdfmark.
+
+"""
 
 from binascii import hexlify
 from datetime import datetime
@@ -78,7 +92,8 @@ def
 
 
 def encode_text_string(s: str) -> str:
-    '''Encode text string to hex string for use in a PDF
+    """
+    Encode text string to hex string for use in a PDF
 
     From PDF 32000-1:2008 a string object may be included in hexademical form
     if it is enclosed in angle brackets.  For general Unicode the string should
@@ -86,7 +101,7 @@ def encode_text_string(s: str) -> str:
     ASCII strings could be encoded as PdfDocEncoding literals provided
     that certain Postscript sequences are escaped.  But it's far simpler to
     encode everything as UTF-16.
-    '''
+    """
 
     # Sometimes lazy C programmers leave their NULs at the end of strings they
     # insert into PDFs
@@ -102,8 +117,28 @@ def encode_text_string(s: str) -> str:
     return ascii_hex_str
 
 
+def _encode_ascii(s: str) -> str:
+    """
+    Aggressively strip non-ASCII and PDF escape sequences
+
+    Ghostscript 9.24+ lost support for UTF-16BE in pdfmark files for reasons
+    given in GhostPDL commit e997c683. Our temporary workaround is use ASCII
+    and drop all non-ASCII characters. A slightly improved alternative would
+    be to implement PdfDocEncoding in pikepdf and encode to that, or handle
+    metadata there.
+    """
+    trans = str.maketrans({
+        '(': '',
+        ')': '',
+        '\\': '',
+        '\0': ''
+    })
+    return s.translate(trans).encode('ascii', errors='replace').decode()
+
+
 def encode_pdf_date(d: datetime) -> str:
-    """Encode Python datetime object as PDF date string
+    """
+    Encode Python datetime object as PDF date string
 
     From Adobe pdfmark manual:
     (D:YYYYMMDDHHmmSSOHH'mm')
@@ -137,6 +172,13 @@ def encode_pdf_date(d: datetime) -> str:
 
 
 def decode_pdf_date(s: str) -> datetime:
+    """
+    Decode a pdfmark date to a Python datetime object
+
+    A pdfmark date is a string in a paritcular format. See the pdfmark
+    Reference for the specification.
+
+    """
     if s.startswith('D:'):
         s = s[2:]
 
@@ -153,11 +195,13 @@ def decode_pdf_date(s: str) -> datetime:
 
 
 def _get_pdfmark_dates(pdfmark):
-    """Encode dates for pdfmark Postscript.  The best way to deal with a
-    missing date entry is set it to null, because if the key is omitted
-    Ghostscript will set it to now - we do not want to erase the fact that
-    the value was unknown.  Setting to an empty string breaks Ghostscript
-    9.22 as reported here:
+    """
+    Encode dates in the expected format for pdfmark Postscript
+
+    The best way to deal with amissing date entry is set it to null, because if
+    the key is omitted Ghostscript will set it to now - we do not want to erase
+    the fact that the value was unknown.  Setting to an empty string breaks
+    Ghostscript 9.22 as reported here:
     https://bugs.ghostscript.com/show_bug.cgi?id=699182
     """
 
@@ -178,7 +222,7 @@ def _get_pdfmark_dates(pdfmark):
             yield '  {} null'.format(key)
 
 
-def _get_pdfa_def(icc_profile, icc_identifier, pdfmark):
+def _get_pdfa_def(icc_profile, icc_identifier, pdfmark, ascii_docinfo=False):
     """
     Create a Postscript pdfmark file for Ghostscript.
 
@@ -188,6 +232,9 @@ def _get_pdfa_def(icc_profile, icc_identifier, pdfmark):
     :param icc_profile: filename of the ICC profile to include in pdfmark
     :param icc_identifier: ICC identifier such as 'sRGB'
     :param pdfmark: a dictionary containing keys to include the pdfmark
+    :param ascii_docinfo: if True, the docinfo block must be encoded in pure
+        ASCII and may not contain UTF-16BE-BOM-hex encoded strings, as
+        required for Ghostscript 9.24+
 
     :returns: a string containing the entire pdfmark
 
@@ -198,14 +245,19 @@ def _get_pdfa_def(icc_profile, icc_identifier, pdfmark):
     # https://bugs.ghostscript.com/show_bug.cgi?id=697684
     # Work around this by only adding keys that have a nontrivial value
     docinfo_keys = ('/Title', '/Author', '/Subject', '/Creator', '/Keywords')
-    docinfo_line_template = '  {key} <{value}>'
 
     def docinfo_gen():
+        if not ascii_docinfo:
+            docinfo_line_template = '  {key} <{value}>'
+            encode = encode_text_string
+        else:
+            docinfo_line_template = '  {key} ({value})'
+            encode = _encode_ascii
         yield from _get_pdfmark_dates(pdfmark)
         for key in docinfo_keys:
             if key in pdfmark and pdfmark[key].strip() != '':
                 line = docinfo_line_template.format(
-                    key=key, value=encode_text_string(pdfmark[key]))
+                    key=key, value=encode(pdfmark[key]))
                 yield line
     docinfo = '\n'.join(docinfo_gen())
 
@@ -216,7 +268,7 @@ def _get_pdfa_def(icc_profile, icc_identifier, pdfmark):
     return result
 
 
-def generate_pdfa_ps(target_filename, pdfmark, icc='sRGB'):
+def generate_pdfa_ps(target_filename, pdfmark, icc='sRGB', ascii_docinfo=False):
     if icc == 'sRGB':
         icc_profile = SRGB_ICC_PROFILE
     else:
@@ -233,7 +285,7 @@ def generate_pdfa_ps(target_filename, pdfmark, icc='sRGB'):
     hex_icc_profile = hexlify(bytes_icc_profile)
     icc_profile = '<' + hex_icc_profile.decode('ascii') + '>'
 
-    ps = _get_pdfa_def(icc_profile, icc, pdfmark)
+    ps = _get_pdfa_def(icc_profile, icc, pdfmark, ascii_docinfo=ascii_docinfo)
 
     # We should have encoded everything to pure ASCII by this point, and
     # to be safe, only allow ASCII in PostScript

src/ocrmypdf/pdfinfo.py

@@ -16,12 +16,14 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from decimal import Decimal
-from math import hypot, isclose
-import re
 from collections import namedtuple
-from pathlib import Path
+from decimal import Decimal
 from enum import Enum
+from math import hypot, isclose
+from pathlib import Path
+from unittest.mock import Mock
+import re
+import xml.etree.ElementTree as ET
 
 from .exec import ghostscript
 from .helpers import fspath
@@ -90,28 +92,6 @@ FRIENDLY_COMP = {
 
 UNIT_SQUARE = (1.0, 0.0, 0.0, 1.0, 0.0, 0.0)
 
-
-def _matrix_from_shorthand(shorthand):
-    """Convert from PDF matrix shorthand to full matrix
-
-    PDF 1.7 spec defines a shorthand for describing the entries of a matrix
-    since the last column is always (0, 0, 1).
-    """
-
-    a, b, c, d, e, f = map(float, shorthand)
-    return ((a, b, 0),
-            (c, d, 0),
-            (e, f, 1))
-
-
-def _shorthand_from_matrix(matrix):
-    """Convert from transformation matrix to PDF shorthand."""
-    a, b = matrix[0][0], matrix[0][1]
-    c, d = matrix[1][0], matrix[1][1]
-    e, f = matrix[2][0], matrix[2][1]
-    return tuple(map(float, (a, b, c, d, e, f)))
-
-
 def _is_unit_square(shorthand):
     values = map(float, shorthand)
     pairwise = zip(values, UNIT_SQUARE)
@@ -508,27 +488,13 @@ def _find_images(*, pdf, container, shorthand=None):
     yield from _find_form_xobject_images(pdf, container, contentsinfo)
 
 
-def _page_get_textblocks(infile, pageno):
+def _page_get_textblocks(infile, pageno, xmltext):
     """Smarter text detection"""
-    import xml.etree.ElementTree as ET
 
-    gstext = ghostscript.extract_text(infile, pageno+1)
-
-    # Remove all <char /> tags, because they might contain invalid XML entities
-    # like <char bbox="348 596 348 596" c="&#x1;"/> which chokes on the
-    # inclusion of U+0001. Understandably.
-    # Just remove the whole <char /> tag since we don't use it at all, and they
-    # are only generated as innermost self-closing tags.
-    gstext = regex_remove_char_tags.sub(b' ', gstext)
-
-    if gstext.strip() == '':
+    root = xmltext
+    if not hasattr(xmltext, 'findall'):
         return []
 
-    try:
-        root = ET.fromstring(gstext)
-    except ET.ParseError as e:
-        return []  # If we can't parse, assume none...
-
     def blocks():
         for span in root.findall('.//span'):
             bbox_str = span.attrib['bbox']
@@ -587,14 +553,15 @@ def _page_has_text(text_blocks, page_width, page_height):
     return has_text
 
 
-def _pdf_get_pageinfo(pdf, pageno: int, infile):
+def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
     pageinfo = {}
     pageinfo['pageno'] = pageno
     pageinfo['images'] = []
 
     page = pdf.pages[pageno]
 
-    pageinfo['textinfo'] = _page_get_textblocks(fspath(infile), pageno)
+    pageinfo['textinfo'] = _page_get_textblocks(
+        fspath(infile), pageno, xmltext=xmltext)
 
     mediabox = [Decimal(d) for d in page.MediaBox.as_list()]
     width_pt = mediabox[2] - mediabox[0]
@@ -631,16 +598,47 @@ def _pdf_get_pageinfo(pdf, pageno: int, infile):
     return pageinfo
 
 
-def _pdf_get_all_pageinfo(infile):
+def _pdf_get_all_pageinfo(infile, log=None):
+    if not log:
+        log = Mock()
+
     pdf = pikepdf.open(infile)
-    return [PageInfo(pdf, n, infile) for n in range(len(pdf.pages))]
+
+    existing_text = ghostscript.extract_text(infile, pageno=None)
+    existing_text = regex_remove_char_tags.sub(b' ', existing_text)
+
+    try:
+        root = ET.fromstringlist([
+            b'<document>\n', existing_text, b'</document>\n'
+        ])
+        page_xml = root.findall('page')
+    except ET.ParseError as e:
+        log.error(
+            "An error occurred while attempting to retrieve existing text in "
+            "the input file. Will attempt to continue assuming that there is "
+            "no existing text in the file. The error was:")
+        log.error(e)
+        page_xml = [None] * len(pdf.pages)
+
+    page_count_difference = len(pdf.pages) - len(page_xml)
+    if page_count_difference != 0:
+        log.error("The number of pages in the input file is inconsistent.")
+        if page_count_difference > 0:
+            page_xml.extend([None] * page_count_difference)
+
+    pages = []
+    for n in range(len(pdf.pages)):
+        page = PageInfo(pdf, n, infile, page_xml[n])
+        pages.append(page)
+
+    return pages, pdf
 
 
 class PageInfo:
-    def __init__(self, pdf, pageno, infile):
+    def __init__(self, pdf, pageno, infile, xmltext):
         self._pageno = pageno
         self._infile = infile
-        self._pageinfo = _pdf_get_pageinfo(pdf, pageno, infile)
+        self._pageinfo = _pdf_get_pageinfo(pdf, pageno, infile, xmltext)
 
     @property
     def pageno(self):
@@ -717,9 +715,10 @@ class PdfInfo:
     """Get summary information about a PDF
 
     """
-    def __init__(self, infile):
+    def __init__(self, infile, log=None):
         self._infile = infile
-        self._pages = _pdf_get_all_pageinfo(infile)
+        self._pages, pdf = _pdf_get_all_pageinfo(infile, log=log)
+        self._needs_rendering = pdf.root.get('/NeedsRendering', False)
 
     @property
     def pages(self):
@@ -740,6 +739,10 @@ class PdfInfo:
             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]
 
@@ -749,15 +752,6 @@ class PdfInfo:
     def __repr__(self):
         return "<PdfInfo('...'), page count={}>".format(len(self))
 
-    # def __getstate__(self):
-    #     state = {'_infile': self._infile}
-    #     return state
-    #
-    # def __setstate__(self, state):
-    #     self._infile = state['_infile']
-    #     self._pages = _pdf_get_all_pageinfo(self._infile)
-
-
 def main():
     import argparse
     parser = argparse.ArgumentParser()

tests/conftest.py

@@ -145,7 +145,8 @@ def check_ocrmypdf(input_file, output_file, *args, env=None):
 
 
 @pytest.helpers.register
-def run_ocrmypdf(input_file, output_file, *args, env=None):
+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:
@@ -155,7 +156,7 @@ def run_ocrmypdf(input_file, output_file, *args, env=None):
              [str(input_file), str(output_file)]
     p = Popen(
         p_args, close_fds=True, stdout=PIPE, stderr=PIPE,
-        universal_newlines=True, env=env)
+        universal_newlines=universal_newlines, env=env)
     out, err = p.communicate()
     #print(err)
 
@@ -168,3 +169,21 @@ def first_page_dimensions(pdf):
     info = pdfinfo.PdfInfo(pdf)
     page0 = info[0]
     return (page0.width_inches, page0.height_inches)
+
+
+def pytest_addoption(parser):
+    parser.addoption(
+        "--runslow", action="store_true", default=False,
+        help=("run slow tests only useful for development (unlikely to be "
+              "useful for downstream packagers)")
+    )
+
+
+def pytest_collection_modifyitems(config, items):
+    if config.getoption("--runslow"):
+        # --runslow given in cli: do not skip slow tests
+        return
+    skip_slow = pytest.mark.skip(reason="need --runslow option to run")
+    for item in items:
+        if "slow" in item.keywords:
+            item.add_marker(skip_slow)

tests/resources/README.rst

@@ -104,6 +104,16 @@ licensed under the specified license.
         - @jbarlow83
         - @jbarlow83
         - CC-BY-SA 4.0
+    *   - kcs.pdf
+        - PDF file generated by Kodak Capture Desktop Software 1.2; has invalid table of contents
+        - @jbarlow83
+        - @jbarlow83
+        - CC-BY-SA 4.0
+    *   - livecycle.pdf
+        - a minimal PDF that claims to use dynamic XFA forms
+        - @jbarlow83
+        - @jbarlow83
+        - CC-BY-SA 4.0
     *   - masks.pdf
         - file containing explicit masks and a stencil mask drawn without a proper transformation matrix; printout of a German Wikipedia article (CC-BY-SA)
         - @supergrobi
@@ -146,8 +156,9 @@ Assemblies
 
 These test resources are assemblies or derivatives from other previously mentioned files, released under the same license terms as their input files.
 
-- baiona_gray.png (from baiona.png)
-- baiona_colormapped.png (from baiona.png)
+- baiona_gray.png (from baiona.png, grayscale version)
+- baiona_colormapped.png (from baiona.png, palette version)
+- baiona_alpha.png (from baiona.png, RGB+A version)
 - cardinal.pdf (four cardinal directions, baked-in rotated copies of linn.png)
 - ccitt.pdf (linn.png, converted to CCITT encoding)
 - encrypted_algo4.pdf (congress.jpg, encrypted with algorithm 4 - not supported by PyPDF2)

tests/resources/livecycle.pdf

@@ -0,0 +1,28 @@
+%PDF-1.7
+%¿÷¢þ
+1 0 obj
+<< /NeedsRendering true /Pages 2 0 R /Type /Catalog >>
+endobj
+2 0 obj
+<< /Count 1 /Kids [ 3 0 R ] /Type /Pages >>
+endobj
+3 0 obj
+<< /Contents 4 0 R /MediaBox [ 0 0 300 300 ] /Parent 2 0 R /Resources << >> /Type /Page >>
+endobj
+4 0 obj
+<< /Length 0 >>
+stream
+
+endstream
+endobj
+xref
+0 5
+0000000000 65535 f 
+0000000015 00000 n 
+0000000085 00000 n 
+0000000144 00000 n 
+0000000250 00000 n 
+trailer << /Root 1 0 R /Size 5 /ID [<8088c23e0edc07ef4fbed3daa55f52ed><9d36eee510d20004f32255f5800b6fd7>] >>
+startxref
+299
+%%EOF

tests/spoof/gs_render_failure.py

@@ -20,11 +20,11 @@
 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
+"""Replicate Ghostscript render failure while allowing rasterizing"""
 
 import sys
 import os
 
-"""Replicate Ghostscript render failure while allowing rasterizing"""
 
 
 def real_ghostscript(argv):

tests/test_main.py

@@ -691,6 +691,7 @@ def test_no_contents(spoof_tesseract_noop, resources, outpdf):
 @pytest.mark.parametrize('image', [
     'baiona.png',
     'baiona_gray.png',
+    'baiona_alpha.png',
     'congress.jpg'
     ])
 def test_compression_preserved(spoof_tesseract_noop, ocrmypdf_exec,
@@ -699,7 +700,6 @@ def test_compression_preserved(spoof_tesseract_noop, ocrmypdf_exec,
     output_file = str(outpdf)
 
     im = Image.open(input_file)
-
     # Runs: ocrmypdf - output.pdf < testfile
     with open(input_file, 'rb') as input_stream:
         p_args = ocrmypdf_exec + [
@@ -710,7 +710,12 @@ def test_compression_preserved(spoof_tesseract_noop, ocrmypdf_exec,
             stdin=input_stream, env=spoof_tesseract_noop)
         out, err = p.communicate()
 
-        assert p.returncode == ExitCode.ok
+        if im.mode in ('RGBA', 'LA'):
+            # If alpha image is input, expect an error
+            assert p.returncode != ExitCode.ok and b'alpha' in err
+            return
+
+        assert p.returncode == ExitCode.ok, err.decode('utf-8')
 
     pdfinfo = PdfInfo(output_file)
 
@@ -754,7 +759,7 @@ def test_compression_changed(spoof_tesseract_noop, ocrmypdf_exec,
             stdin=input_stream, env=spoof_tesseract_noop)
         out, err = p.communicate()
 
-        assert p.returncode == ExitCode.ok
+        assert p.returncode == ExitCode.ok, err
 
     pdfinfo = PdfInfo(output_file)
 
@@ -920,3 +925,12 @@ def test_output_is_symlink(spoof_tesseract_noop, resources, outdir):
     )
     assert p.returncode == ExitCode.ok, err
     assert (outdir / 'out.pdf').stat().st_size > 0, 'target file not created'
+
+
+def test_livecycle(resources, no_outpdf):
+    p, _, err = run_ocrmypdf(
+        resources / 'livecycle.pdf',
+        no_outpdf
+    )
+
+    assert p.returncode == ExitCode.input_file, err

tests/test_metadata.py

@@ -17,7 +17,6 @@
 
 
 import pytest
-import PyPDF2 as pypdf
 
 from datetime import timezone
 from pathlib import Path
@@ -25,12 +24,15 @@ from shutil import copyfile
 from unittest.mock import patch
 import datetime
 
+import pikepdf
+
 from ocrmypdf.exceptions import ExitCode
 from ocrmypdf.helpers import fspath
 from ocrmypdf.pdfa import (
     file_claims_pdfa, encode_pdf_date, decode_pdf_date, generate_pdfa_ps,
     SRGB_ICC_PROFILE
 )
+from ocrmypdf.exec import ghostscript
 
 try:
     import fitz
@@ -51,17 +53,17 @@ spoof = pytest.helpers.spoof
     ])
 def test_preserve_metadata(spoof_tesseract_noop, output_type,
                            resources, outpdf):
-    pdf_before = pypdf.PdfFileReader(str(resources / 'graph.pdf'))
+    pdf_before = pikepdf.open(resources / 'graph.pdf')
 
     output = check_ocrmypdf(
             resources / 'graph.pdf', outpdf,
             '--output-type', output_type,
             env=spoof_tesseract_noop)
 
-    pdf_after = pypdf.PdfFileReader(str(output))
+    pdf_after = pikepdf.open(output)
 
     for key in ('/Title', '/Author'):
-        assert pdf_before.documentInfo[key] == pdf_after.documentInfo[key]
+        assert pdf_before.metadata[key] == pdf_after.metadata[key]
 
     pdfa_info = file_claims_pdfa(str(output))
     assert pdfa_info['output'] == output_type
@@ -85,15 +87,18 @@ def test_override_metadata(spoof_tesseract_noop, output_type, resources,
 
     assert p.returncode == ExitCode.ok, err
 
-    before = pypdf.PdfFileReader(str(input_file))
-    after = pypdf.PdfFileReader(outpdf)
+    before = pikepdf.open(input_file)
+    after = pikepdf.open(outpdf)
 
-    assert after.documentInfo['/Title'] == german
-    assert after.documentInfo['/Author'] == chinese
-    assert after.documentInfo.get('/Keywords', '') == ''
+    if ghostscript.version() >= '9.24':
+        pytest.xfail('Ghostscript 9.24+ does not support Unicode DOCINFO')
 
-    before_date = decode_pdf_date(before.documentInfo['/CreationDate'])
-    after_date = decode_pdf_date(after.documentInfo['/CreationDate'])
+    assert after.metadata.Title == german, after.metadata
+    assert after.metadata.Author == chinese, after.metadata
+    assert after.metadata.get('/Keywords', '') == ''
+
+    before_date = decode_pdf_date(str(before.metadata.CreationDate))
+    after_date = decode_pdf_date(str(after.metadata.CreationDate))
     assert before_date == after_date
 
     pdfa_info = file_claims_pdfa(outpdf)
@@ -146,26 +151,30 @@ def test_creation_date_preserved(spoof_tesseract_noop, output_type, resources,
                                  infile, outpdf):
     input_file = resources / infile
 
-    before = pypdf.PdfFileReader(str(input_file)).getDocumentInfo()
     check_ocrmypdf(
         input_file, outpdf, '--output-type', output_type,
         env=spoof_tesseract_noop)
-    after = pypdf.PdfFileReader(str(outpdf)).getDocumentInfo()
+
+    pdf_before = pikepdf.open(input_file)
+    pdf_after = pikepdf.open(outpdf)
+
+    before = pdf_before.trailer.get('/Info', {})
+    after = pdf_after.trailer.get('/Info', {})
 
     if not before:
         # If there was input creation date, none should be output
         # because of Ghostscript quirks we set it to null
         # This test would be better if we had a test file with /DocumentInfo but
         # no /CreationDate, which we don't
-        assert not after.get('/CreationDate')
+        assert after.get('/CreationDate', '') == ''
     else:
         # We expect that the creation date stayed the same
-        date_before = decode_pdf_date(before['/CreationDate'])
-        date_after = decode_pdf_date(after['/CreationDate'])
+        date_before = decode_pdf_date(str(before['/CreationDate']))
+        date_after = decode_pdf_date(str(after['/CreationDate']))
         assert seconds_between_dates(date_before, date_after) < 1000
 
     # We expect that the modified date is quite recent
-    date_after = decode_pdf_date(after['/ModDate'])
+    date_after = decode_pdf_date(str(after['/ModDate']))
     assert seconds_between_dates(
         date_after, datetime.datetime.now(timezone.utc)) < 1000
 
@@ -260,3 +269,15 @@ def test_srgb_in_unicode_path(tmpdir):
 
     with patch('ocrmypdf.pdfa.SRGB_ICC_PROFILE', new=str(dst)):
         generate_pdfa_ps(dstdir / 'out.ps', {})
+
+
+def test_kodak_toc(resources, outpdf, spoof_tesseract_noop):
+    output = check_ocrmypdf(
+        resources / 'kcs.pdf', outpdf,
+        '--output-type', 'pdf',
+        env=spoof_tesseract_noop)
+
+    p = pikepdf.open(outpdf)
+
+    if pikepdf.Name.First in p.root.Outlines:
+        assert isinstance(p.root.Outlines.First, pikepdf.Dictionary)

tests/test_rotation.py

@@ -16,12 +16,17 @@
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
 import logging
+from io import BytesIO
 
+from PIL import Image
 import pytest
+import img2pdf
+import pikepdf
 
 from ocrmypdf import leptonica
 from ocrmypdf.pdfinfo import PdfInfo
 from ocrmypdf.exec import ghostscript
+from ocrmypdf.helpers import fspath
 
 
 # pytest.helpers is dynamic
@@ -51,7 +56,8 @@ def check_monochrome_correlation(
             return
         ghostscript.rasterize_pdf(
             pdf, png, xres=100, yres=100,
-            raster_device='pngmono', log=gslog, pageno=pageno)
+            raster_device='pngmono', log=gslog, pageno=pageno,
+            rotation=0)
 
     rasterize(reference_pdf, reference_pageno, reference_png)
     rasterize(test_pdf, test_pageno, test_png)
@@ -83,6 +89,7 @@ def test_monochrome_correlation(resources, outdir):
     assert corr > 0.90
 
 
+@pytest.mark.slow
 @pytest.mark.parametrize('renderer', RENDERERS)
 def test_autorotate(spoof_tesseract_cache, renderer, resources, outdir):
     # cardinal.pdf contains four copies of an image rotated in each cardinal
@@ -171,3 +178,46 @@ def test_rotate_deskew_timeout(resources, outdir):
 
     # Confirm that the page still got deskewed
     assert correlation > 0.50
+
+
+@pytest.mark.slow
+@pytest.mark.parametrize('page_angle', (0, 90, 180, 270))
+@pytest.mark.parametrize('image_angle', (0, 90, 180, 270))
+def test_rotate_page_level(image_angle, page_angle, resources, outdir):
+
+    def make_rotate_test(prefix, image_angle, page_angle):
+        im = Image.open(fspath(resources / 'typewriter.png'))
+        if image_angle != 0:
+            ccw_angle = -image_angle % 360
+            im = im.transpose(getattr(Image, 'ROTATE_{}'.format(ccw_angle)))
+        memimg = BytesIO()
+        im.save(memimg, format='PNG')
+        memimg.seek(0)
+        mempdf = BytesIO()
+        img2pdf.convert(
+            memimg.read(),
+            layout_fun=img2pdf.get_fixed_dpi_layout_fun((200, 200)),
+            outputstream=mempdf
+        )
+        mempdf.seek(0)
+        pike = pikepdf.open(mempdf)
+        pike.pages[0].Rotate = page_angle
+        target = outdir / '{}_{}_{}.pdf'.format(prefix, image_angle, page_angle)
+        pike.save(target)
+        return target
+
+    reference = make_rotate_test('ref', 0, 0)
+    test = make_rotate_test('test', image_angle, page_angle)
+    out = test.with_suffix('.out.pdf')
+
+    p, _, err = run_ocrmypdf(
+        test, out,
+        '-O0',
+        '--rotate-pages',
+        '--rotate-pages-threshold', '0.001',
+        universal_newlines=False
+    )
+    err = err.decode('utf-8', errors='replace')
+    assert p.returncode == 0, err
+
+    assert check_monochrome_correlation(outdir, reference, 1, out, 1) > 0.2