v10.1.0 notes

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

.coveragerc

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

.github/ISSUE_TEMPLATE/bug_report.md

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

.github/ISSUE_TEMPLATE/feature_request.md

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

.travis.yml

@@ -1,159 +0,0 @@
-branches:
-  except:
-    - azure
-
-cache:
-  pip: true
-  directories:
-    - $HOME/Library/Caches/Homebrew
-
-matrix:
-  include:
-    - os: linux
-      dist: trusty
-      sudo: required
-      language: python
-      python: "3.6"
-      env:
-        - DIST=trusty
-        - MINIMAL=true
-      addons:
-        apt:
-          update: true
-          sources:
-            - sourceline: "ppa:alex-p/tesseract-ocr"
-            - sourceline: "ppa:vshn/ghostscript"
-          packages:
-            - ghostscript
-            - libffi-dev
-            - tesseract-ocr
-            - tesseract-ocr-deu
-            - tesseract-ocr-eng
-            - tesseract-ocr-fra
-      before_install: |
-        pip3 install --upgrade pip
-        pip3 install --upgrade wheel
-    - os: linux
-      dist: trusty
-      sudo: required
-      language: python
-      python: "3.6"
-      env:
-        - DIST=trusty
-      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
-            - libffi-dev
-            - tesseract-ocr
-            - tesseract-ocr-deu
-            - tesseract-ocr-eng
-            - tesseract-ocr-fra
-            - libexempi3 # --- optional extras from here ---
-            - pngquant
-            - poppler-utils
-      before_install: |
-        mkdir -p bin packages
-        pip3 install --upgrade pip
-        pip3 install --upgrade wheel
-    - os: linux
-      dist: xenial
-      sudo: required
-      language: python
-      python: "3.7"
-      env:
-        - DIST=xenial
-      addons:
-        apt:
-          update: true
-          sources:
-            - sourceline: "ppa:alex-p/tesseract-ocr"
-          packages:
-            - ghostscript
-            - libexempi3
-            - libffi-dev
-            - pngquant
-            - poppler-utils
-            - tesseract-ocr
-            - tesseract-ocr-deu
-            - tesseract-ocr-eng
-            - tesseract-ocr-fra
-            - unpaper
-    - os: linux
-      dist: bionic
-      sudo: required
-      language: python
-      python: "3.8"
-      env:
-        - DIST=bionic
-      addons:
-        apt:
-          update: true
-          sources:
-            - sourceline: "ppa:alex-p/tesseract-ocr"
-          packages:
-            - ghostscript
-            - libexempi3
-            - libffi-dev
-            - pngquant
-            - poppler-utils
-            - tesseract-ocr
-            - tesseract-ocr-deu
-            - tesseract-ocr-eng
-            - tesseract-ocr-fra
-            - unpaper
-    - os: osx
-      language: generic
-      addons:
-        homebrew:
-          update: true
-          packages:
-            - exempi
-            - ghostscript
-            - jbig2enc
-            - leptonica
-            - openjpeg
-            - pngquant
-            - python
-            - qpdf
-            - tesseract
-            - unpaper
-      before_install: |
-        pip3 install --upgrade pip
-        pip3 install wheel
-
-before_cache:
-  - rm -f $HOME/.cache/pip/log/debug.log
-
-install:
-  - mkdir -p bin
-  - export PATH=$PWD/bin:$PATH
-  - pip3 install -r requirements/main.txt -r requirements/test.txt .
-
-script:
-  - tesseract --version
-  - pytest -n auto
-# deploy:
-#   # release for main pypi
-#   # 3.7 is considered the build leader and does the deploy, otherwise there is
-#   # a race and all versions will try to deploy
-#   # OTOH if we ever need separate binary wheels then each version needs its
-#   # own deploy
-#   - provider: pypi
-#     user: ocrmypdf-travis
-#     password:
-#       secure: "DTFOmmNL6olA0+yXvp4u9jXZlZeqrJsJ0526jzqf4a3gZ6jnGTq5UI6WzRsslSyoMMfXKtHQebqHM6ogSgCZinyZ3ufHJo8fn9brxbEc2gsiWkbj5o3bGwdWMT1vNNE7XW0VCpw87rZ1EEwjl4FJHFudMlPR1yfU5+uq0k0PACo="
-#     distributions: "sdist bdist_wheel"
-#     on:
-#       branch: master
-#       tags: true
-#       condition: $TRAVIS_PYTHON_VERSION == "3.7" && $TRAVIS_OS_NAME == "linux"
-#     skip_upload_docs: true

README.md

@@ -38,7 +38,7 @@ Main features
 - Keeps the exact resolution of the original embedded images
 - When possible, inserts OCR information as a "lossless" operation without disrupting any other content
 - Optimizes PDF images, often producing files smaller than the input file
-- If requested deskews and/or cleans the image before performing OCR
+- If requested, deskews and/or cleans the image before performing OCR
 - Validates input and output files
 - Distributes work across all available CPU cores
 - Uses [Tesseract OCR](https://github.com/tesseract-ocr/tesseract) engine to recognize more than [100 languages](https://github.com/tesseract-ocr/tessdata)
@@ -50,7 +50,7 @@ For details: please consult the [documentation](https://ocrmypdf.readthedocs.io/
 Motivation
 ----------
 
-I searched the web for a free command line tool to OCR PDF files: I found many, but none of them were really satisfying.
+I searched the web for a free command line tool to OCR PDF files: I found many, but none of them were really satisfying:
 
 - Either they produced PDF files with misplaced text under the image (making copy/paste impossible)
 - Or they did not handle accents and multilingual characters
@@ -97,7 +97,10 @@ OCRmyPDF uses Tesseract for OCR, and relies on its language packs. For Linux use
 apt-cache search tesseract-ocr
 
 # Debian/Ubuntu users
-apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language back
+apt-get install tesseract-ocr-chi-sim  # Example: Install Chinese Simplified language pack
+
+# Arch Linux users
+pacman -S tesseract-data-eng tesseract-data-deu # Example: Install the English and German language packs
 ```
 
 You can then pass the `-l LANG` argument to OCRmyPDF to give a hint as to what languages it should search for. Multiple languages can be requested.
@@ -105,7 +108,7 @@ You can then pass the `-l LANG` argument to OCRmyPDF to give a hint as to what l
 Documentation and support
 -------------------------
 
-Once ocrmypdf is installed, the built-in help which explains the command syntax and options can be accessed via:
+Once OCRmyPDF is installed, the built-in help which explains the command syntax and options can be accessed via:
 
 ```bash
 ocrmypdf --help
@@ -118,20 +121,21 @@ Please report issues on our [GitHub issues](https://github.com/jbarlow83/OCRmyPD
 Requirements
 ------------
 
-In addition to the required Python version (3.6+), OCRmyPDF requires external program installations of Ghostscript, Tesseract OCR, QPDF, and Leptonica. ocrmypdf is pure Python, but uses CFFI to portably generate library bindings. OCRmyPDF works on pretty much  everything: Linux, macOS, Windows and FreeBSD.
+In addition to the required Python version (3.6+), OCRmyPDF requires external program installations of Ghostscript, Tesseract OCR, QPDF, and Leptonica. OCRmyPDF is pure Python, but uses CFFI to portably generate library bindings. OCRmyPDF works on pretty much everything: Linux, macOS, Windows and FreeBSD.
 
 Press & Media
 -------------
 
 - [Going paperless with OCRmyPDF](https://medium.com/@ikirichenko/going-paperless-with-ocrmypdf-e2f36143f46a)
 - [Converting a scanned document into a compressed searchable PDF with redactions](https://medium.com/@treyharris/converting-a-scanned-document-into-a-compressed-searchable-pdf-with-redactions-63f61c34fe4c)
-- [c't 1-2014, page 59](http://heise.de/-2279695): Detailed presentation of OCRmyPDF v1.0 in the leading German IT magazine c't
-- [heise Open Source, 09/2014: Texterkennung mit OCRmyPDF](http://heise.de/-2356670)
+- [c't 1-2014, page 59](https://heise.de/-2279695): Detailed presentation of OCRmyPDF v1.0 in the leading German IT magazine c't
+- [heise Open Source, 09/2014: Texterkennung mit OCRmyPDF](https://heise.de/-2356670)
+- [heise Durchsuchbare PDF-Dokumente mit OCRmyPDF erstellen](https://www.heise.de/ratgeber/Durchsuchbare-PDF-Dokumente-mit-OCRmyPDF-erstellen-4607592.html)
 
 Business enquiries
 ------------------
 
-OCRmyPDF would not be the software that it is today is without companies and users choosing to provide support for feature development and consulting enquiries. We are happy to discuss all enquiries, whether for extending the existing feature set, or integrating OCRmyPDF into a larger system.
+OCRmyPDF would not be the software that it is today without companies and users choosing to provide support for feature development and consulting enquiries. We are happy to discuss all enquiries, whether for extending the existing feature set, or integrating OCRmyPDF into a larger system.
 
 License
 -------

azure-pipelines.yml

@@ -153,12 +153,13 @@ stages:
           matrix:
             Python37:
               python.version: "3.7"
-            Python38:
-              python.version: "3.8"
+            # Python38:
+            #   python.version: "3.8"
         steps:
-          - task: UsePythonVersion@0
-            inputs:
-              versionSpec: "$(python.version)"
+          # https://github.com/actions/virtual-environments/issues/664
+          # - task: UsePythonVersion@0
+          #   inputs:
+          #     versionSpec: "$(python.version)"
           - bash: |
               brew update
               brew unlink python@2
@@ -169,6 +170,7 @@ stages:
                 leptonica \
                 openjpeg \
                 pngquant \
+                python \
                 tesseract \
                 unpaper
             displayName: "Install system packages"

debian/copyright

@@ -95,12 +95,6 @@ Files: tests/resources/vector.pdf
 Copyright: (C) 2018 Catscratch
 License: Expat
 
-Files: test/resources/enron*.pdf
-Copyright: EnronData.org
-License: CC-BY-3.0
- See: https://enrondata.readthedocs.io/en/latest/data/edo-enron-email-pst-dataset/
-Comment: Unprocessed.
-
 Files: src/ocrmypdf/data/sRGB.icc
 Copyright:  Kai-Uwe Behrmann <www.behrmann.name>
             Marti Maria <www.littlecms.com>

docs/advanced.rst

@@ -148,7 +148,8 @@ In addition to tesseract, OCRmyPDF uses the following external binaries:
 
 -  ``gs`` (Ghostscript)
 -  ``unpaper``
--  ``qpdf``
+-  ``pngquant``
+-  ``jbig2``
 
 In each case OCRmyPDF will search the ``PATH`` environment variable to
 locate the binaries.

docs/api.rst

@@ -51,6 +51,14 @@ Forking a child process to call ``ocrmypdf.ocr()`` is suggested. That
 way your application will survive and remain interactive even if
 OCRmyPDF does not.
 
+.. warning::
+
+    On Windows, the script that calls ``ocrmypdf.ocr()`` must be protected
+    by an "ifmain" guard (``if __name__ == '__main__'``) or you must use
+    ``ocrmypdf.ocr(...use_threads=True)``. If you do not take at least one
+    of these steps, Windows process semantics will prevent OCRmyPDF from working
+    correctly.
+
 Logging
 -------
 
@@ -97,8 +105,8 @@ Reference
     :members:
     :undoc-members:
 
-.. autoclass:: ocrmypdf.ExitCode
+.. autofunction:: ocrmypdf.configure_logging
+
+.. automodule:: ocrmypdf.exceptions
     :members:
     :undoc-members:
-
-.. autofunction:: ocrmypdf.configure_logging

docs/batch.rst

@@ -69,53 +69,8 @@ Sample script
 This user contributed script also provides an example of batch
 processing.
 
-.. code-block:: python
-
-   #!/usr/bin/env python3
-   # Walk through directory tree, replacing all files with OCR'd version
-   # Original version by DeliciousPickle@github; modified
-
-   import logging
-   import os
-   import subprocess
-   import sys
-
-   import ocrmypdf
-
-   script_dir = os.path.dirname(os.path.realpath(__file__))
-   print(script_dir + '/ocr-tree.py: Start')
-
-   if len(sys.argv) > 1:
-       start_dir = sys.argv[1]
-   else:
-       start_dir = '.'
-
-   if len(sys.argv) > 2:
-       log_file = sys.argv[2]
-   else:
-       log_file = script_dir + '/ocr-tree.log'
-
-   logging.basicConfig(
-           level=logging.INFO, format='%(asctime)s %(message)s',
-           filename=log_file, filemode='w')
-
-   ocrmypdf.configure_logging(ocrmypdf.Verbosity.default)
-
-   for dir_name, subdirs, file_list in os.walk(start_dir):
-       logging.info('\n')
-       logging.info(dir_name + '\n')
-       os.chdir(dir_name)
-       for filename in file_list:
-           file_ext = os.path.splitext(filename)[1]
-           if file_ext == '.pdf':
-               full_path = dir_name + '/' + filename
-               print(full_path)
-               result = ocrmypdf.ocr(filename, filename, deskew=True)
-               if result == ocrmypdf.ExitCode.already_done_ocr:
-                   print("Skipped document because it already contained text")
-               elif result == ocrmypdf.ExitCode.ok:
-                   print("OCR complete")
-               logging.info(result)
+.. literalinclude:: ../misc/batch.py
+    :caption: misc/batch.py
 
 Synology DiskStations
 ---------------------
@@ -131,62 +86,8 @@ products use ARM or Power processors and do not support Docker. Further
 adjustments might be needed to deal with the Synology's relatively
 limited CPU and RAM.
 
-.. code-block:: python
-
-   #!/bin/env python3
-   # Contributed by github.com/Enantiomerie
-
-   # script needs 2 arguments
-   # 1. source dir with *.pdf - default is location of script
-   # 2. move dir where *.pdf and *_OCR.pdf are moved to
-
-   import logging
-   import os
-   import subprocess
-   import sys
-   import time
-   import shutil
-
-   script_dir = os.path.dirname(os.path.realpath(__file__))
-   timestamp = time.strftime("%Y-%m-%d-%H%M_")
-   log_file = script_dir + '/' + timestamp + 'ocrmypdf.log'
-   logging.basicConfig(level=logging.INFO, format='%(asctime)s %(message)s', filename=log_file, filemode='w')
-
-   if len(sys.argv) > 1:
-       start_dir = sys.argv[1]
-   else:
-       start_dir = '.'
-
-   for dir_name, subdirs, file_list in os.walk(start_dir):
-       logging.info('\n')
-       logging.info(dir_name + '\n')
-       os.chdir(dir_name)
-       for filename in file_list:
-           file_ext = os.path.splitext(filename)[1]
-           if file_ext == '.pdf':
-               full_path = dir_name + '/' + filename
-               file_noext = os.path.splitext(filename)[0]
-               timestamp_OCR = time.strftime("%Y-%m-%d-%H%M_OCR_")
-               filename_OCR = timestamp_OCR + file_noext + '.pdf'
-               docker_mount = dir_name + ':/home/docker'
-   # create string for pdf processing
-   # diskstation needs a user:group docker:docker. find uid:gid of your diskstation docker:docker with id docker.
-   # use this uid:gid in -u flag
-   # rw rights for docker:docker at source dir are also necessary
-   # the script is processed as root user via chron
-               cmd = ['docker', 'run', '--rm', '-v', docker_mount, '-u=1030:65538', 'jbarlow83/ocrmypdf', , '--deskew' , filename, filename_OCR]
-               logging.info(cmd)
-               proc = subprocess.run(cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
-               result = proc.stdout.read()
-               logging.info(result)
-               full_path_OCR = dir_name + '/' + filename_OCR
-               os.chmod(full_path_OCR, 0o666)
-               os.chmod(full_path, 0o666)
-               full_path_OCR_archive = sys.argv[2]
-               full_path_archive = sys.argv[2] + '/no_ocr'
-               shutil.move(full_path_OCR,full_path_OCR_archive)
-               shutil.move(full_path, full_path_archive)
-   logging.info('Finished.\n')
+.. literalinclude:: ../misc/synology.py
+    :caption: misc/synology.py - Sample script for Synology DiskStations
 
 Huge batch jobs
 ---------------
@@ -198,11 +99,45 @@ and all inquiries are appreciated.
 Hot (watched) folders
 =====================
 
+Watched folders with watcher.py
+-------------------------------
+
+OCRmyPDF has a folder watcher called watcher.py, which is currently included in source
+distributions but not part of the main program. It may be used natively or may run
+in a Docker container. Native instances tend to give better performance. watcher.py
+works on all platforms.
+
+Users may need to customize the script to meet their requirements.
+
+.. code-block:: bash
+
+    pip3 install -r requirements/watcher.txt
+
+    env OCR_INPUT_DIRECTORY=/mnt/input-pdfs \
+        OCR_OUTPUT_DIRECTORY=/mnt/output-pdfs \
+        OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1 \
+        python3 watcher.py
+
+.. csv-table:: watcher.py environment variables
+    :header: "Environment variable", "Description"
+    :widths: 50, 50
+
+    "OCR_INPUT_DIRECTORY", "Set input directory to monitor (recursive)"
+    "OCR_OUTPUT_DIRECTORY", "Set output directory (should not be under input)"
+    "OCR_ON_SUCCESS_DELETE", "This will delete the input file if the exit code is 0 (OK)"
+    "OCR_OUTPUT_DIRECTORY_YEAR_MONTH", "This will place files in the output in ``{output}/{year}/{month}/{filename}``"
+    "OCR_DESKEW", "Apply deskew to crooked input PDFs"
+    "OCR_JSON_SETTINGS", "A JSON string specifying any other arguments for ``ocrmypdf.ocr``"
+    "OCR_POLL_NEW_FILE_SECONDS", "Polling interval"
+    "OCR_LOGLEVEL", "Level of log messages to report"
+
+One could configure a networked scanner or scanning computer to drop files in the
+watched folder.
+
 Watched folders with Docker
 ---------------------------
 
-The OCRmyPDF Docker image includes a watcher service. This service can
-be launched as follows:
+The watcher service is included in the OCRmyPDF Docker image. To run it:
 
 .. code-block:: bash
 
@@ -210,6 +145,9 @@ be launched as follows:
         -v <path to files to convert>:/input \
         -v <path to store results>:/output \
         -e OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1 \
+        -e OCR_ON_SUCCESS_DELETE=1 \
+        -e OCR_DESKEW=1 \
+        -e PYTHONUNBUFFERED=1 \
         -it --entrypoint python3 \
         jbarlow83/ocrmypdf \
         watcher.py
@@ -223,61 +161,21 @@ convert it to a OCRed PDF in ``/output/``. The parameters to this image are:
 
     "``-v <path to files to convert>:/input``", "Files placed in this location will be OCRed"
     "``-v <path to store results>:/output``", "This is where OCRed files will be stored"
-    "``-e OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1``", "This will place files in the output in {output}/{year}/{month}/{filename}"
+    "``-e OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1``", "Define environment variable OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1"
+    "``-e OCR_ON_SUCCESS_DELETE=1``", "Define environment variable"
+    "``-e OCR_DESKEW=1``", "Define environment variable"
+    "``-e PYTHONBUFFERED=1``", "This will force STDOUT to be unbuffered and allow you to see messages in docker logs"
 
 This service relies on polling to check for changes to the filesystem. It
 may not be suitable for some environments, such as filesystems shared on a
 slow network.
 
-Watched folders with watcher.py
--------------------------------
+A configuration manager such as Docker Compose could be used to ensure that the
+service is always available.
 
-The watcher service may also be run natively.
-
-.. code-block:: bash
-
-    pip3 install -r reqs/watcher.txt
-
-    env OCR_INPUT_DIRECTORY=/mnt/input-pdfs \
-        OCR_OUTPUT_DIRECTORY=/mnt/output-pdfs \
-        OCR_OUTPUT_DIRECTORY_YEAR_MONTH=1 \
-        python3 watcher.py
-
-Watched folders with CLI
-------------------------
-
-To set up a "hot folder" that will trigger OCR for every file inserted,
-use a program like Python
-`watchdog <https://pypi.python.org/pypi/watchdog>`__ (supports all major
-OS).
-
-One could then configure a scanner to automatically place scanned files
-in a hot folder, so that they will be queued for OCR and copied to the
-destination.
-
-.. code-block:: bash
-
-   pip install watchdog
-
-watchdog installs the command line program ``watchmedo``, which can be
-told to run ``ocrmypdf`` on any .pdf added to the current directory
-(``.``) and place the result in the previously created ``out/`` folder.
-
-.. code-block:: bash
-
-   cd hot-folder
-   mkdir out
-   watchmedo shell-command \
-       --patterns="*.pdf" \
-       --ignore-directories \
-       --command='ocrmypdf "${watch_src_path}" "out/${watch_src_path}" ' \
-       .  # don't forget the final dot
-
-On file servers, you could configure watchmedo as a system service so it
-will run all the time.
-
-For more complex behavior you can write a Python script around to use
-the watchdog API. You can refer to the watcher.py script as an example.
+.. literalinclude:: ../misc/docker-compose.example.yml
+    :language: yaml
+    :caption: misc/docker-compose.example.yml
 
 Caveats
 -------
@@ -304,6 +202,13 @@ Alternatives
 -  `Watchman <https://facebook.github.io/watchman/>`__ is a more
    powerful alternative to ``watchmedo``.
 
+AWS Lambda is not viable
+------------------------
+
+AWS Lambda and its equivalents have low limits on execution time and payload
+size, relative to OCRmyPDF's needs. As of this writing, the request/response
+payload for AWS Lambda was 6 MB, which means many PDFs will not fit.
+
 macOS Automator
 ===============
 
@@ -313,8 +218,7 @@ of Automator, the ``PATH`` may be set differently your Terminal's
 ``PATH``; you may need to explicitly set the PATH to include
 ``ocrmypdf``. The following example may serve as a starting point:
 
-|Example macOS Automator script|
+.. figure:: images/macos-workflow.png
+    :alt: Example macOS Automator workflow
 
 You may customize the command sent to ocrmypdf.
-
-.. |Example macOS Automator script| image:: images/macos-workflow.png

docs/cookbook.rst

@@ -56,7 +56,11 @@ portrait pages.
     ocrmypdf --rotate-pages myfile.pdf myfile.pdf
 
 You can increase (decrease) the parameter ``--rotate-pages-threshold``
-to make page rotation more (less) aggressive.
+to make page rotation more (less) aggressive. The threshold number is the ratio
+of how confidence the OCR engine is that the document image should be changed,
+compared to kept the same. A value of ``15.0`` is the default, and is fairly
+conservative. A value of ``2.0`` will produce more rotations, and more false
+positives.
 
 If the page is "just a little off horizontal", like a crooked picture,
 then you want ``--deskew``. ``--rotate-pages`` is for when the cardinal
@@ -89,6 +93,18 @@ This produces a file named "output.pdf" and a companion text file named
 
     ocrmypdf --sidecar output.txt input.pdf output.pdf
 
+.. note::
+
+    The sidecar file contains the **OCR text** found by OCRmyPDF. If the document
+    contains pages that already have text, that text will not appear in the
+    sidecar. If the option ``--pages`` is used, only those pages on which OCR
+    was performed will be included in the sidecar. If certain pages were skipped
+    because of options like ``--skip-big`` or ``--tesseract-timeout``, those pages
+    will not be in the sidecar.
+
+    To extract all text from a PDF, whether generated from OCR or otherwise,
+    use a program like Poppler's ``pdftotext`` or ``pdfgrep``.
+
 OCR images, not PDFs
 --------------------
 
@@ -225,6 +241,32 @@ You can also optimize all images without performing any OCR:
 
     ocrmypdf --tesseract-timeout=0 --optimize 3 --skip-text input.pdf output.pdf
 
+Perform OCR only certain pages
+------------------------------
+
+You can ask OCRmyPDF to only apply OCR to certain pages.
+
+.. code-block:: bash
+
+    ocrmypdf --pages 2,3,13-17 input.pdf output.pdf
+
+Hyphens denote a range of pages and commas separate page numbers. If you prefer
+to use spaces, quote all of the page numbers: ``--pages '2, 3, 5, 7'``.
+
+OCRmyPDF will warn if your list of page numbers contains duplicates or
+overlap pages. OCRmyPDF does not currently account for document page numbers,
+such as an introduction section of a book that uses Roman numerals. It simply
+counts the number of virtual pieces of paper since the start.
+
+Regardless of the argument to ``--pages``, OCRmyPDF will optimize all pages in
+the file and convert it to PDF/A, unless you disable those options. In this
+example, we want to OCR only the title and otherwise change the PDF as little
+as possible:
+
+.. code-block:: bash
+
+    ocrmypdf --pages 1 --output-type pdf --optimize 0 input.pdf output.pdf
+
 Redo existing OCR
 =================
 

docs/docker.rst

@@ -56,52 +56,52 @@ See the Docker documentation for
 Using the Docker image on the command line
 ==========================================
 
-**Unlike typical Docker containers**, in this mode we are using the
-OCRmyPDF Docker container is intended to be emphemeral – it runs for one
-OCR job and then terminates, just like a command line program. We are
-using Docker as a way of delivering an application, not a server.
+**Unlike typical Docker containers**, in this section the OCRmyPDF Docker
+container is emphemeral – it runs for one OCR job and terminates, just like a
+command line program. We are using Docker to deliver an application (as opposed
+to the more conventional case, where a Docker container runs as a server).
 
 To start a Docker container (instance of the image):
 
 .. code-block:: bash
 
    docker tag jbarlow83/ocrmypdf ocrmypdf
-   docker run --rm -i ocrmypdf (... all other arguments here...)
+   docker run --rm -i ocrmypdf (... all other arguments here...) - -
 
 For convenience, create a shell alias to hide the Docker command. It is
-easier to send the input file to file stdin and read the output from
-stdout – this avoids the occasionally messy permission issues with
-Docker entirely.
+easier to send the input file as stdin and read the output from
+stdout – **this avoids the messy permission issues with Docker entirely**.
 
 .. code-block:: bash
 
-   alias ocrmypdf='docker run --rm -i ocrmypdf'
-   ocrmypdf --version  # runs docker version
-   ocrmypdf <input.pdf >output.pdf
+   alias docker_ocrmypdf='docker run --rm -i ocrmypdf'
+   docker_ocrmypdf --version  # runs docker version
+   docker_ocrmypdf - - <input.pdf >output.pdf
 
 Or in the wonderful `fish shell <https://fishshell.com/>`__:
 
 .. code-block:: fish
 
-   alias ocrmypdf 'docker run --rm ocrmypdf'
-   funcsave ocrmypdf
+   alias docker_ocrmypdf 'docker run --rm ocrmypdf'
+   funcsave docker_ocrmypdf
 
 Alternately, you could mount the local current working directory as a
 Docker volume:
 
 .. code-block:: bash
 
-   docker run --rm -v $(pwd):/data ocrmypdf /data/input.pdf /data/output.pdf
+   alias docker_ocrmypdf='docker run --rm  -i --user "$(id -u):$(id -g)" --workdir /data -v "$PWD:/data" ocrmypdf'
+   docker_ocrmypdf /data/input.pdf /data/output.pdf
 
 .. _docker-lang-packs:
 
 Adding languages to the Docker image
 ====================================
 
-By default the Docker image includes English, German and Simplified
-Chinese, the most popular languages for OCRmyPDF users based on
-feedback. You may add other languages by creating a new Dockerfile based
-on the public one:
+By default the Docker image includes English, German, Simplified Chinese,
+French, Portuguese and Spanish, the most popular languages for OCRmyPDF
+users based on feedback. You may add other languages by creating a new
+Dockerfile based on the public one:
 
 .. code-block:: dockerfile
 

docs/errors.rst

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

docs/index.rst

@@ -1,8 +1,8 @@
 OCRmyPDF documentation
 ======================
 
-OCRmyPDF adds an OCR text layer to scanned PDF files, allowing them to
-be searched.
+OCRmyPDF adds an optical charcter recognition (OCR) text layer to scanned PDF
+files, allowing them to be searched.
 
 PDF is the best format for storing and exchanging scanned documents.
 Unfortunately, PDFs can be difficult to modify. OCRmyPDF makes it easy to apply
@@ -26,7 +26,8 @@ image processing and OCR to existing PDFs.
    docker
    advanced
    batch
-   security
+   performance
+   pdfsecurity
    errors
 
 .. toctree::
@@ -34,6 +35,7 @@ image processing and OCR to existing PDFs.
    :maxdepth: 2
 
    api
+   plugins
    contributing
 
 Indices and tables

docs/installation.rst

@@ -45,14 +45,8 @@ Debian and Ubuntu 18.04 or newer
 .. |ubu-1804| image:: https://repology.org/badge/version-for-repo/ubuntu_18_04/ocrmypdf.svg
     :alt: Ubuntu 18.04 LTS
 
-.. |ubu-1810| image:: https://repology.org/badge/version-for-repo/ubuntu_18_10/ocrmypdf.svg
-    :alt: Ubuntu 18.10
-
-.. |ubu-1904| image:: https://repology.org/badge/version-for-repo/ubuntu_19_04/ocrmypdf.svg
-    :alt: Ubuntu 19.04
-
-.. |ubu-1910| image:: https://repology.org/badge/version-for-repo/ubuntu_19_10/ocrmypdf.svg
-    :alt: Ubuntu 19.10
+.. |ubu-2004| image:: https://repology.org/badge/version-for-repo/ubuntu_20_04/ocrmypdf.svg
+    :alt: Ubuntu 20.04 LTS
 
 +-----------------------------------------------+
 | **OCRmyPDF versions in Debian & Ubuntu**      |
@@ -61,7 +55,7 @@ Debian and Ubuntu 18.04 or newer
 +-----------------------------------------------+
 | |deb-stable| |deb-testing| |deb-unstable|     |
 +-----------------------------------------------+
-| |ubu-1804| |ubu-1810| |ubu-1904| |ubu-1910|   |
+| |ubu-1804| |ubu-2004|                         |
 +-----------------------------------------------+
 
 Users of Debian 9 ("stretch") or later or Ubuntu 18.04 or later, including users
@@ -110,7 +104,7 @@ Fedora 29 or newer
 | |fedora-29| |fedora-30| |fedora-rawhide|      |
 +-----------------------------------------------+
 
-Users of Fedora 29 later may simply
+Users of Fedora 29 or later may simply
 
 .. code-block:: bash
 
@@ -133,26 +127,64 @@ from sources <#installing-head-revision-from-sources>`__.
 
 .. _ubuntu-lts-latest:
 
-Installing the latest version on Ubuntu 18.04 LTS
+Installing the latest version on Ubuntu 20.04 LTS
 -------------------------------------------------
 
-Ubuntu 18.04 includes ocrmypdf 6.1.2. To install a more recent version,
-first install the system version to get most of the dependencies:
+Ubuntu 20.04 includes ocrmypdf 9.6.0 - you can install that with ``apt``. To
+install a more recent version, uninstall the system-provided version of
+ocrmypdf, and install the following dependencies:
 
 .. code-block:: bash
 
-    sudo apt-get update
-    sudo apt-get install \
-        ocrmypdf
-
-There are a few system dependency changes since ocrmypdf 6.1.2. Let's
-get these, too.
-
-.. code-block:: bash
-
-    sudo apt-get install \
+    sudo apt-get -y remove ocrmypdf  # remove system ocrmypdf, if installed
+    sudo apt-get -y update
+    sudo apt-get -y install \
+        ghostscript \
+        icc-profiles-free \
+        liblept5 \
         libxml2 \
-        pngquant
+        pngquant \
+        python3-pip \
+        tesseract-ocr \
+        zlib1g
+
+To install ocrmypdf for the system:
+
+.. code-block:: bash
+
+    sudo pip3 install ocrmypdf
+
+To install for the current user only:
+
+.. code-block:: bash
+
+    export PATH=$HOME/.local/bin:$PATH
+    pip3 install --user ocrmypdf
+
+Ubuntu 18.04 LTS
+-------------------------------------------------
+
+Ubuntu 18.04 includes ocrmypdf 6.1.2 - you can install that with ``apt``, but
+it is quite old now. To install a more recent version, uninstall the old version
+of ocrmypdf, and install the following dependencies:
+
+.. code-block:: bash
+
+    sudo apt-get -y remove ocrmypdf
+    sudo apt-get -y update
+    sudo apt-get -y install \
+        ghostscript \
+        icc-profiles-free \
+        liblept5 \
+        libxml2 \
+        pngquant \
+        python3-cffi \
+        python3-distutils \
+        python3-pkg-resources \
+        python3-reportlab \
+        qpdf \
+        tesseract-ocr \
+        zlib1g
 
 We will need a newer version of ``pip`` then was available for Ubuntu 18.04:
 
@@ -214,7 +246,8 @@ of ``pip`` at ``/usr/local/bin/pip``.
 **Install OCRmyPDF**
 
 OCRmyPDF requires the locale to be set for UTF-8. **On some minimal
-Ubuntu installations systems**, it may be necessary to set the locale.
+Ubuntu installations**, such as the Ubuntu 16.04 Docker images it may be
+necessary to set the locale.
 
 .. code-block:: bash
 
@@ -228,7 +261,7 @@ environment variable contains ``$HOME/.local/bin``.
 .. code-block:: bash
 
     export PATH=$HOME/.local/bin:$PATH
-    pip3 install --user ocrmypdf
+    pip3.6 install --user ocrmypdf
 
 To add JBIG2 encoding, see :ref:`jbig2`.
 
@@ -294,23 +327,72 @@ compiled by hand.
 
 To add JBIG2 encoding, see :ref:`jbig2`.
 
-ArchLinux (AUR)
----------------
+Arch Linux (AUR)
+----------------
 
 .. image:: https://repology.org/badge/version-for-repo/aur/ocrmypdf.svg
     :alt: ArchLinux
     :target: https://repology.org/metapackage/ocrmypdf
 
-There is an `ArchLinux User Repository package for
-ocrmypdf <https://aur.archlinux.org/packages/ocrmypdf/>`__. You can use
-the following command.
+There is an `Arch User Repository (AUR) package for OCRmyPDF
+<https://aur.archlinux.org/packages/ocrmypdf/>`__.
+
+Installing AUR packages as root is not allowed, so you must first `setup a
+non-root user
+<https://wiki.archlinux.org/index.php/Users_and_groups#User_management>`__ and
+`configure sudo <https://wiki.archlinux.org/index.php/Sudo#Configuration>`__.
+The standard Docker image, ``archlinux/base:latest``, does **not** have a
+non-root user configured, so users of that image must follow these guides. If
+you are using a VM image, such as `the official Vagrant image
+<https://app.vagrantup.com/archlinux/boxes/archlinux>`__, this work may already
+be completed for you.
+
+Next you should install the `base-devel package group
+<https://www.archlinux.org/groups/x86_64/base-devel/>`__. This includes the
+standard tooling needed to build packages, such as a compiler and binary tools.
 
 .. code-block:: bash
 
-    yaourt -S ocrmypdf
+   sudo pacman -S base-devel
 
-If you have any difficulties with installation, check the repository
-package page.
+Now you are ready to install the OCRmyPDF package.
+
+.. code-block:: bash
+
+   curl -O https://aur.archlinux.org/cgit/aur.git/snapshot/ocrmypdf.tar.gz
+   tar xvzf ocrmypdf.tar.gz
+   cd ocrmypdf
+   makepkg -sri
+
+At this point you will have a working install of OCRmyPDF, but the Tesseract
+install won’t include any OCR language data. You can install `the
+tesseract-data package group
+<https://www.archlinux.org/groups/any/tesseract-data/>`__ to add all supported
+languages, or use that package listing to identify the appropriate package for
+your desired language.
+
+.. code-block:: bash
+
+   sudo pacman -S tesseract-data-eng
+
+As an alternative to this manual procedure, consider using an `AUR helper
+<https://wiki.archlinux.org/index.php/AUR_helpers>`__. Such a tool will
+automatically fetch, build and install the AUR package, resolve dependencies
+(including dependencies on AUR packages), and ease the upgrade procedure.
+
+If you have any difficulties with installation, check the repository package
+page.
+
+.. note::
+
+    The OCRmyPDF AUR package currently omits the JBIG2 encoder. OCRmyPDF works
+    fine without it but will produce larger output files. The encoder is
+    available from `the jbig2enc-git AUR package
+    <https://aur.archlinux.org/packages/jbig2enc-git/>`__ and may be installed
+    using the same series of steps as for the installation OCRmyPDF AUR
+    package. Alternatively, it may be built manually from source following the
+    instructions in `Installing the JBIG2 encoder <jbig2>`__.  If JBIG2 is
+    installed, OCRmyPDF 7.0.0 and later will automatically detect it.
 
 Alpine Linux
 ------------
@@ -375,7 +457,10 @@ languages you can optionally install them all:
 Manual installation on macOS
 ----------------------------
 
-These instructions probably work on all macOS supported by Homebrew.
+These instructions probably work on all macOS supported by Homebrew, and are
+for installing a more current version of OCRmyPDF than is available from
+Homebrew. Note that the Homebrew versions usually track the release versions
+fairly closely.
 
 If it's not already present, `install Homebrew <http://brew.sh/>`__.
 
@@ -386,14 +471,8 @@ Update Homebrew:
     brew update
 
 Install or upgrade the required Homebrew packages, if any are missing.
-To do this, download the ``Brewfile`` that lists all of the dependencies
-to the current directory, and run ``brew bundle`` to process them
-(installing or upgrading as needed). ``Brewfile`` is a plain text file.
-
-.. code-block:: bash
-
-    wget https://github.com/jbarlow83/OCRmyPDF/raw/master/.travis/Brewfile
-    brew bundle
+To do this, use ``brew edit ocrmypdf`` to obtain a recent list of Homebrew
+dependencies. You could also check the ``azure-pipelines.yml``.
 
 This will include the English, French, German and Spanish language
 packs. If you need other languages you can optionally install them all:
@@ -435,12 +514,12 @@ Installing on FreeBSD
     :alt: FreeBSD
     :target: https://repology.org/project/python:ocrmypdf/versions
 
-FreeBSD 11.2, 11.3, 12.0-RELEASE and 13.0-CURRENT are supported. Other
+FreeBSD 11.3, 12.0, 12.1-RELEASE and 13.0-CURRENT are supported. Other
 versions likely work but have not been tested.
 
 .. code-block:: bash
 
-    pkg install py36-ocrmypdf
+    pkg install py37-ocrmypdf
 
 To install a more recent version, you could attempt to first install the system
 version with ``pkg``, then use ``pip install --user ocrmypdf``.
@@ -507,7 +586,7 @@ Installing on Windows Subsystem for Linux
 
 .. code-block:: powershell
 
-    wsl sudo ln -s  /home/user/.local/bin/ocrmypdf /usr/local/bin/ocrmypdf
+    wsl sudo ln -s  /home/$USER/.local/bin/ocrmypdf /usr/local/bin/ocrmypdf
 
 Then confirm that the expected version from PyPI (|latest|) is installed:
 
@@ -524,6 +603,53 @@ Docker
 You can also :ref:`Install the Docker <docker-install>` container on Windows. Ensure that
 your command prompt can run the docker "hello world" container.
 
+Installing on Cygwin64 under Windows
+====================================
+
+First install the the following prerequisite Cygwin packages using ``setup-x86_64.exe``::
+
+    python36 (or later)
+    python3?-devel
+    python3?-pip
+    python3?-lxml
+    python3?-imaging
+
+       (where 3? means match the version of python3 you installed)
+
+    gcc-g++
+    ghostscript (<=9.50 or >=9.52-2 see note below)
+    libexempi3
+    libexempi-devel
+    libffi6
+    libffi-devel
+    pngquant
+    qpdf
+    libqpdf-devel
+    tesseract-ocr
+    tesseract-ocr-devel
+
+.. note::
+
+    The Cygwin package for Ghostscript in versions 9.52 and
+    9.52-1 contained a bug that caused an exception to occur when
+    ocrmypdf invoked gs.  Make sure you have either 9.50 (or earlier)
+    or 9.52-2 (or later).
+
+Then open a Cygwin terminal (i.e. ``mintty``), run the following commands. Note
+that if you are using the version of ``pip`` that was installed with the Cygwin
+Python package, the command name will be ``pip3``.  If you have since updated
+``pip`` (with, for instance ``pip3 install --upgrade pip``) the the command is
+likely just ``pip`` instead of ``pip3``:
+
+.. code-block:: bash
+
+    pip3 install wheel
+    pip3 install ocrmypdf
+
+The optional dependency "unpaper" that is currently not available under Cygwin.
+Without it, certain options such as ``--clean`` will produce an error message.
+However, the OCR-to-text-layer functionality is available.
+
 Installing with Python pip
 ==========================
 

docs/introduction.rst

@@ -10,7 +10,7 @@ About OCR
 `Optical character
 recognition <https://en.wikipedia.org/wiki/Optical_character_recognition>`__
 is technology that converts images of typed or handwritten text, such as
-in a scanned document, to computer text that can be searched and copied.
+in a scanned document, to computer text that can be selected, searched and copied.
 
 OCRmyPDF uses
 `Tesseract <https://github.com/tesseract-ocr/tesseract>`__, the best

docs/languages.rst

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

docs/pdfsecurity.rst

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

docs/performance.rst

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

docs/plugins.rst

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

docs/release_notes.rst

@@ -5,14 +5,187 @@ Release notes
 OCRmyPDF uses `semantic versioning <http://semver.org/>`__ for its
 command line interface and its public API.
 
-The ``ocrmypdf`` package may now be imported. The public API may be
-useful in scripts that launch OCRmyPDF processes or that wish to use
-some of its features for working with PDFs.
+OCRmyPDF's output messages are not considered part of the stable interface -
+that is, output messages may be improved at any release level, so parsing them
+may be unreliable. Use the API to depend on precise behavior.
+
+The public API may be useful in scripts that launch OCRmyPDF processes or that
+wish to use some of its features for working with PDFs.
 
 Note that it is licensed under GPLv3, so scripts that
 ``import ocrmypdf`` and are released publicly should probably also be
 licensed under GPLv3.
 
+v10.1.0
+=======
+
+-  Previously, we ``--clean-final`` would cause an unpaper-cleaned page image to
+   be produced twice, which was necessary in some cases but not in general. We
+   now take this optimization opportunity and reuse the image if possible.
+-  We now provide PNG files as input to unpaper, since it accepts them, instead
+   of generating PPM files which can be very large. This can improve performance
+   and temporary disk usage.
+-  Documentation updated for plugins.
+
+v10.0.1
+=======
+
+-  Fix regression when ``-l lang1+lang2`` is used from command line.
+
+v10.0.0
+=======
+
+**Breaking changes**
+
+-  Support for pdfminer.six version 20181108 has been dropped, along with a
+   monkeypatch that made this version work.
+-  Output messages are now displayed in color (when supported by the terminal)
+   and prefixes describing the severity of the message are removed. As such
+   programs that parse OCRmyPDF's log message will need to be revised. (Please
+   consider using OCRmyPDF as a library instead.)
+-  The minimum version for certain dependencies has increased.
+-  Many API changes; see developer changes.
+-  The Python libraries pluggy and coloredlogs are now required.
+
+**New features and improvements**
+
+-  PDF page scanning is now parallelized across CPUs, speeding up this phase
+   dramatically for files with a high page counts.
+-  PDF page scanning is optimized, addressing some performance regressions.
+-  PDF page scanning is no longer run on pages that are not selected when the
+   ``--pages`` argument is used.
+-  PDF page scanning is now independent of Ghostscript, ending our past reliance
+   on this occasionally unstable feature in Ghostscript.
+-  A plugin architecture has been added, currently allowing one to more easily
+   use a different OCR engine or PDF renderer from Tesseract and Ghostscript,
+   respectively. A plugin can also override some decisions, such changing
+   the OCR settings after initial scanning.
+-  Colored log messages.
+
+**Developer changes**
+
+-  The test spoofing mechanism, used to test correct handling of failures in
+   Tesseract and Ghostscript, has been removed in favor of using plugins for
+   testing. The spoofing mechanism was fairly complex and required many special
+   hacks for Windows.
+-  Code describing the resolution in DPI of images was refactored into a
+   ``ocrmypdf.helpers.Resolution`` class.
+-  The module ``ocrmypdf._exec`` is now private to OCRmyPDF.
+-  The ``ocrmypdf.hocrtransform`` module has been updated to follow PEP8 naming
+   conventions.
+-  Ghostscript is no longer used for finding the location of text in PDFs, and
+   APIs related to this feature have been removed.
+-  Lots of internal reorganization to support plugins.
+
+v9.8.2
+======
+
+-  Fixed an issue where OCRmyPDF would ignore text inside Form XObject when
+   making certain decisions about whether a document already had text.
+-  Fixed file size increase warning to take overhead of small files into account.
+-  Added instructions for installing on Cygwin.
+
+v9.8.1
+======
+
+-  Fixed an issue where unexpected files in the ``%PROGRAMFILES%\gs`` directory
+   (Windows) caused an exception.
+-  Mark pdfminer.six 20200517 as supported.
+-  If jbig2enc is missing and optimization is requested, a warning is issued
+   instead of an error, which was the intended behavior.
+-  Documentation updates.
+
+v9.8.0
+======
+
+-  Fixed issue where only the first PNG (FlateDecode) image in a file would be
+   considered for optimization. File sizes should be improved from here on.
+-  Fixed a startup crash when the chosen language was Japanese (#543).
+-  Added options to configure polling and log level to watcher.py.
+
+v9.7.2
+======
+
+-  Fixed an issue with ``ocrmypdf.ocr(...language=)`` not accepting a list of
+   languages as documented.
+-  Updated setup.py to confirm that pdfminer.six version 20200402 is supported.
+
+v9.7.1
+======
+
+-  Fixed version check failing when used with qpdf 10.0.0.
+-  Added some missing type annotations.
+-  Updated documentation to warn about need for "ifmain" guard and Windows.
+
+v9.7.0
+======
+
+-  Fixed an error in watcher.py if ``OCR_JSON_SETTINGS`` was not defined.
+-  Ghostscript 9.51 is now blacklisted, due to numerous problems with this version.
+-  Added a workaround for a problem with "txtwrite" in Ghostscript 9.52.
+-  Fixed an issue where the incorrect number of threads used was shown when
+   ``OMP_THREAD_LIMIT`` was manipulated.
+-  Removed a possible performance bottlenecks for files that use hundreds to
+   thousands of images on the same page.
+-  Documentation improvements.
+-  Optimization will now be applied to some monochrome images that have a color
+   profile defined instead of only black and white.
+-  ICC profiles are consulted when determining the simplified colorspace of an
+   image.
+
+v9.6.1
+======
+
+-  Documentation improvements - thanks to many users for their contributions!
+
+      - Fixed installation instructions for ArchLinux (@pigmonkey)
+      - Updated installation instructions for FreeBSD and other OSes (@knobix)
+      - Added instructions for using Docker Compose with watchdog (@ianalexander,
+        @deisi)
+      - Other miscellany (@mb720, @toy, @caiofacchinato)
+      - Some scripts provided in the documentation have been migrated out so that
+        they can be copied out as whole files, and to ensure syntax checking
+        is maintained.
+
+-  Fixed an error that caused bash completions to fail on macOS. (#502, #504;
+   @AlexanderWillner)
+-  Fixed a rare case where OCRmyPDF threw an exception while processing a PDF
+   with the wrong object type in its ``/Trailer /Info``. The error is now logged
+   and incorrect object is ignored. (#497)
+-  Removed potentially non-free file ``enron1.pdf`` and simplified the test that
+   used it.
+-  Removed potentially non-free file ``misc/media/logo.afdesign``.
+
+v9.6.0
+======
+
+-  Fixed a regression with transferring metadata from the input PDF to the output
+   PDF in certain situations.
+-  pdfminer.six is now supported up to version 2020-01-24.
+-  Messages are explaining page rotation decisions are now shown at the standard
+   verbosity level again when ``--rotate-pages``. In some previous version they
+   were set to debug level messages that only appeared with the parameter ``-v1``.
+-  Improvements to ``misc/watcher.py``. Thanks to @ianalexander and @svenihoney.
+-  Documentation improvements.
+
+v9.5.0
+======
+
+-  Added API functions to measure OCR quality.
+-  Modest improvements to handling PDFs with difficult/non compliant metadata.
+
+v9.4.0
+======
+
+-  Updated recommended dependency versions.
+-  Improvements to test coverage and changes to facilitate better measurement of
+   test coverage, such as when tests run in subprocesses.
+-  Improvements to error messages when Leptonica is not installed correctly.
+-  Fixed use of pytest "session scope" that may have caused some intermittent
+   CI failures.
+-  When the argument ``--keep-temporary-files`` or verbosity is set to ``-v1``,
+   a debug log file is generated in the working temporary folder.
+
 v9.3.0
 ======
 

misc/batch.py

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

misc/completion/ocrmypdf.bash

@@ -5,7 +5,33 @@ set -o errexit
 _ocrmypdf()
 {
     local cur prev cword words split
-    _init_completion -s || return
+
+    # Homebrew on Macs have version 1.3 of bash-completion which doesn't include - see #502
+    if declare -F _init_completions >/dev/null 2>&1; then
+        _init_completion -s || return
+    else
+        COMPREPLY=()
+        _get_comp_words_by_ref cur prev words cword
+    fi
+
+    if [[ $cur == -* ]]; then
+        COMPREPLY=( $( compgen -W '--language --image-dpi --output-type
+            --sidecar --version --jobs --quiet --verbose --title --author
+            --subject --keywords --rotate-pages --remove-background --deskew
+            --clean --clean-final --unpaper-args --oversample --remove-vectors
+            --threshold --force-ocr --skip-text --redo-ocr
+            --skip-big --jpeg-quality --png-quality --jbig2-lossy
+            --max-image-mpixels --tesseract-config --tesseract-pagesegmode
+            --help --tesseract-oem --pdf-renderer --tesseract-timeout
+            --rotate-pages-threshold --pdfa-image-compression --user-words
+            --user-patterns --keep-temporary-files --output-type
+            --no-progress-bar --pages --fast-web-view' \
+            --  "$cur" ) )
+        return
+    else
+        _filedir
+        return
+    fi
 
     case $prev in
         --version|-h|--help)
@@ -65,25 +91,6 @@ _ocrmypdf()
     esac
 
     $split && return
-
-    if [[ $cur == -* ]]; then
-        COMPREPLY=( $( compgen -W '--language --image-dpi --output-type
-            --sidecar --version --jobs --quiet --verbose --title --author
-            --subject --keywords --rotate-pages --remove-background --deskew
-            --clean --clean-final --unpaper-args --oversample --remove-vectors
-            --threshold --force-ocr --skip-text --redo-ocr
-            --skip-big --jpeg-quality --png-quality --jbig2-lossy
-            --max-image-mpixels --tesseract-config --tesseract-pagesegmode
-            --help --tesseract-oem --pdf-renderer --tesseract-timeout
-            --rotate-pages-threshold --pdfa-image-compression --user-words
-            --user-patterns --keep-temporary-files --output-type
-            --no-progress-bar --pages --fast-web-view' \
-            --  "$cur" ) )
-        return
-    else
-        _filedir
-        return
-    fi
 } &&
 complete -F _ocrmypdf ocrmypdf
 

misc/docker-compose.example.yml

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

misc/example_plugin.py

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

misc/synology.py

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

misc/watcher.py

@@ -1,4 +1,5 @@
 # Copyright (C) 2019 Ian Alexander: https://github.com/ianalexander
+# Copyright (C) 2020 James R Barlow: https://github.com/jbarlow83
 #
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
@@ -13,53 +14,134 @@
 # You should have received a copy of the GNU General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
+import json
+import logging
 import os
+import sys
 import time
 from datetime import datetime
 from pathlib import Path
 
+import pikepdf
 from watchdog.events import PatternMatchingEventHandler
 from watchdog.observers import Observer
+from watchdog.observers.polling import PollingObserver
 
 import ocrmypdf
 
+# pylint: disable=logging-format-interpolation
+
 INPUT_DIRECTORY = os.getenv('OCR_INPUT_DIRECTORY', '/input')
 OUTPUT_DIRECTORY = os.getenv('OCR_OUTPUT_DIRECTORY', '/output')
-OUTPUT_DIRECTORY_YEAR_MONTH = bool(os.getenv('OCR_OUTPUT_DIRECTORY_YEAR_MONTH', False))
+OUTPUT_DIRECTORY_YEAR_MONTH = bool(os.getenv('OCR_OUTPUT_DIRECTORY_YEAR_MONTH', ''))
+ON_SUCCESS_DELETE = bool(os.getenv('OCR_ON_SUCCESS_DELETE', ''))
+DESKEW = bool(os.getenv('OCR_DESKEW', ''))
+OCR_JSON_SETTINGS = json.loads(os.getenv('OCR_JSON_SETTINGS', '{}'))
+POLL_NEW_FILE_SECONDS = int(os.getenv('OCR_POLL_NEW_FILE_SECONDS', '1'))
+USE_POLLING = bool(os.getenv('OCR_USE_POLLING', ''))
+LOGLEVEL = os.getenv('OCR_LOGLEVEL', 'INFO').upper()
 PATTERNS = ['*.pdf']
 
+log = logging.getLogger('ocrmypdf-watcher')
 
-def execute_ocrmypdf(file_path):
-    filename = Path(file_path).name
+
+def get_output_dir(root, basename):
     if OUTPUT_DIRECTORY_YEAR_MONTH:
         today = datetime.today()
-        output_directory_year_month = Path(
-            f'{OUTPUT_DIRECTORY}/{today.year}/{today.month}'
+        output_directory_year_month = (
+            Path(root) / str(today.year) / f'{today.month:02d}'
         )
         if not output_directory_year_month.exists():
             output_directory_year_month.mkdir(parents=True, exist_ok=True)
-        output_path = Path(output_directory_year_month) / filename
+        output_path = Path(output_directory_year_month) / basename
     else:
-        output_path = Path(OUTPUT_DIRECTORY) / filename
-    print(f'New file: {file_path}.\nAttempting to OCRmyPDF to: {output_path}')
-    ocrmypdf.ocr(file_path, output_path)
+        output_path = Path(OUTPUT_DIRECTORY) / basename
+    return output_path
+
+
+def wait_for_file_ready(file_path):
+    # This loop waits to make sure that the file is completely loaded on
+    # disk before attempting to read. Docker sometimes will publish the
+    # watchdog event before the file is actually fully on disk, causing
+    # pikepdf to fail.
+
+    retries = 5
+    while retries:
+        try:
+            pdf = pikepdf.open(file_path)
+        except (FileNotFoundError, pikepdf.PdfError) as e:
+            log.info(f"File {file_path} is not ready yet")
+            log.debug("Exception was", exc_info=e)
+            time.sleep(POLL_NEW_FILE_SECONDS)
+            retries -= 1
+        else:
+            pdf.close()
+            return True
+
+    return False
+
+
+def execute_ocrmypdf(file_path):
+    file_path = Path(file_path)
+    output_path = get_output_dir(OUTPUT_DIRECTORY, file_path.name)
+
+    log.info("-" * 20)
+    log.info(f'New file: {file_path}. Waiting until fully loaded...')
+    if not wait_for_file_ready(file_path):
+        log.info(f"Gave up waiting for {file_path} to become ready")
+        return
+    log.info(f'Attempting to OCRmyPDF to: {output_path}')
+    exit_code = ocrmypdf.ocr(
+        input_file=file_path,
+        output_file=output_path,
+        deskew=DESKEW,
+        **OCR_JSON_SETTINGS,
+    )
+    if exit_code == 0 and ON_SUCCESS_DELETE:
+        log.info(f'OCR is done. Deleting: {file_path}')
+        file_path.unlink()
+    else:
+        log.info('OCR is done')
 
 
 class HandleObserverEvent(PatternMatchingEventHandler):
     def on_any_event(self, event):
-        if event.event_type in ['created', 'modified']:
+        if event.event_type in ['created']:
             execute_ocrmypdf(event.src_path)
 
 
-if __name__ == "__main__":
-    print(
+def main():
+    ocrmypdf.configure_logging(
+        verbosity=ocrmypdf.Verbosity.default, manage_root_logger=True
+    )
+    log.setLevel(LOGLEVEL)
+    log.info(
         f"Starting OCRmyPDF watcher with config:\n"
         f"Input Directory: {INPUT_DIRECTORY}\n"
         f"Output Directory: {OUTPUT_DIRECTORY}\n"
         f"Output Directory Year & Month: {OUTPUT_DIRECTORY_YEAR_MONTH}"
     )
+    log.debug(
+        f"INPUT_DIRECTORY: {INPUT_DIRECTORY}\n"
+        f"OUTPUT_DIRECTORY: {OUTPUT_DIRECTORY}\n"
+        f"OUTPUT_DIRECTORY_YEAR_MONTH: {OUTPUT_DIRECTORY_YEAR_MONTH}\n"
+        f"ON_SUCCESS_DELETE: {ON_SUCCESS_DELETE}\n"
+        f"DESKEW: {DESKEW}\n"
+        f"ARGS: {OCR_JSON_SETTINGS}\n"
+        f"POLL_NEW_FILE_SECONDS: {POLL_NEW_FILE_SECONDS}\n"
+        f"USE_POLLING: {USE_POLLING}\n"
+        f"LOGLEVEL: {LOGLEVEL}\n"
+    )
+
+    if 'input_file' in OCR_JSON_SETTINGS or 'output_file' in OCR_JSON_SETTINGS:
+        log.error('OCR_JSON_SETTINGS should not specify input file or output file')
+        sys.exit(1)
+
     handler = HandleObserverEvent(patterns=PATTERNS)
-    observer = Observer()
+    if USE_POLLING:
+        observer = PollingObserver()
+    else:
+        observer = Observer()
     observer.schedule(handler, INPUT_DIRECTORY, recursive=True)
     observer.start()
     try:
@@ -68,3 +150,7 @@ if __name__ == "__main__":
     except KeyboardInterrupt:
         observer.stop()
     observer.join()
+
+
+if __name__ == "__main__":
+    main()

requirements/dev.txt

@@ -1,2 +0,0 @@
-twine >= 1.8.1
-coverage >= 4.5

requirements/main.txt

@@ -1,10 +1,12 @@
 # requirements.txt can be used to replicate the developer's build environment
 # setup.py lists a separate set of requirements that are looser to simplify
 # installation
-cffi == 1.13.2
-img2pdf == 0.3.3
-pdfminer.six == 20191110
-pikepdf == 1.8.1
-Pillow >= 6.2.0
-reportlab == 3.5.32
-tqdm == 4.37.0
+cffi == 1.14.0
+coloredlogs == 14.0  # technically optional
+img2pdf == 0.3.4
+pdfminer.six == 20200517
+pikepdf == 1.14.0
+pluggy == 0.13.1
+Pillow == 7.1.1
+reportlab == 3.5.34
+tqdm == 4.45.0

requirements/test.txt

@@ -1,8 +1,7 @@
 pytest >= 5.0.0
 pytest-helpers-namespace >= 2019.1.8
-pytest-xdist >= 1.29.0  # For DumpError fix
-pytest-cov >= 2.6.1
+pytest-xdist >= 1.31.0
+pytest-cov >= 2.9.0
 python-xmp-toolkit == 2.0.1  # requires apt-get install libexempi3
                      # or brew install exempi
-PyPDF2 >= 1.26.0
 #PyMuPDF == 1.13.4    # optional

setup.cfg

@@ -23,7 +23,7 @@ force_grid_wrap=0
 use_parentheses=True
 line_length=88
 known_first_party = ocrmypdf
-known_third_party = PIL,PyPDF2,_cffi_backend,cffi,flask,gs,img2pdf,pdfminer,pikepdf,pkg_resources,pytest,reportlab,setuptools,sphinx_rtd_theme,tqdm,watchdog,werkzeug
+known_third_party = PIL,_cffi_backend,cffi,flask,img2pdf,pdfminer,pikepdf,pkg_resources,pluggy,pytest,reportlab,setuptools,sphinx_rtd_theme,tqdm,watchdog,werkzeug
 
 [metadata]
 license_file = LICENSE

setup.py

@@ -21,27 +21,12 @@ from __future__ import print_function, unicode_literals
 
 import sys
 
+from setuptools import find_packages, setup
+
 if sys.version_info < (3, 6):
     print("Python 3.6 or newer is required", file=sys.stderr)
     sys.exit(1)
 
-from setuptools import setup, find_packages
-
-# pylint: disable=w0613
-
-
-command = next((arg for arg in sys.argv[1:] if not arg.startswith('-')), '')
-if command.startswith('install') or command in [
-    'check',
-    'test',
-    'nosetests',
-    'easy_install',
-]:
-    forced = '--force' in sys.argv
-    if forced:
-        print("The argument --force is deprecated. Please discontinue use.")
-
-
 if 'upload' in sys.argv[1:]:
     print('Use twine to upload the package - setup.py upload is insecure')
     sys.exit(1)
@@ -61,7 +46,7 @@ setup(
     long_description_content_type='text/markdown',
     url='https://github.com/jbarlow83/OCRmyPDF',
     author='James R. Barlow',
-    author_email='jim@purplerock.ca',
+    author_email='james@purplerock.ca',
     packages=find_packages('src', exclude=["tests", "tests.*"]),
     package_dir={'': 'src'},
     keywords=['PDF', 'OCR', 'optical character recognition', 'PDF/A', 'scanning'],
@@ -94,12 +79,13 @@ setup(
     use_scm_version={'version_scheme': 'post-release'},
     cffi_modules=['src/ocrmypdf/lib/compile_leptonica.py:ffibuilder'],
     install_requires=[
-        'chardet >= 3.0.4, < 4',  # unlisted requirement of pdfminer.six 20181108
         'cffi >= 1.9.1',  # must be a setup and install requirement
+        'coloredlogs >= 14.0',  # strictly optional
         'img2pdf >= 0.3.0, < 0.4',  # pure Python, so track HEAD closely
-        'pdfminer.six >= 20181108, <= 20191110',
-        'pikepdf >= 1.8.1, < 2',
-        'Pillow >= 6.2.0',
+        'pdfminer.six >= 20191110, <= 20200517',
+        'pikepdf >= 1.14.0, < 2',
+        'Pillow >= 7.0.0',
+        'pluggy >= 0.13.0',
         'reportlab >= 3.3.0',  # oldest released version with sane image handling
         'tqdm >= 4',
     ],

src/ocrmypdf/__init__.py

@@ -15,14 +15,18 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from . import helpers, hocrtransform, leptonica, pdfa, pdfinfo
-from ._version import PROGRAM_NAME, __version__
-from .api import Verbosity, configure_logging, ocr
-from .exceptions import (
+
+from pluggy import HookimplMarker as _HookimplMarker
+
+from ocrmypdf import helpers, hocrtransform, leptonica, pdfa, pdfinfo
+from ocrmypdf._version import PROGRAM_NAME, __version__
+from ocrmypdf.api import Verbosity, configure_logging, ocr
+from ocrmypdf.exceptions import (
     BadArgsError,
     DpiError,
     EncryptedPdfError,
     ExitCode,
+    ExitCodeException,
     InputFileError,
     MissingDependencyError,
     OutputFileAccessError,
@@ -32,3 +36,6 @@ from .exceptions import (
     TesseractConfigError,
     UnsupportedImageFormatError,
 )
+from ocrmypdf.pluginspec import OcrEngine, OrientationConfidence
+
+hookimpl = _HookimplMarker('ocrmypdf')

src/ocrmypdf/__main__.py

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

src/ocrmypdf/_concurrent.py

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

src/ocrmypdf/_exec/__init__.py

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

src/ocrmypdf/_exec/ghostscript.py

@@ -20,9 +20,6 @@
 import logging
 import os
 import re
-import warnings
-from contextlib import suppress
-from functools import lru_cache
 from io import BytesIO
 from os import fspath
 from pathlib import Path
@@ -31,10 +28,11 @@ from subprocess import PIPE, CalledProcessError
 
 from PIL import Image
 
-from ..exceptions import MissingDependencyError, SubprocessOutputError
-from . import get_version, run
+from ocrmypdf.exceptions import MissingDependencyError, SubprocessOutputError
+from ocrmypdf.helpers import Resolution
+from ocrmypdf.subprocess import get_version, run
 
-gslog = logging.getLogger()
+log = logging.getLogger(__name__)
 
 GS = 'gs'
 if os.name == 'nt':
@@ -57,12 +55,11 @@ if os.name == 'nt':
     GS = Path(GS).stem
 
 
-@lru_cache(maxsize=1)
 def version():
     return get_version(GS)
 
 
-def jpeg_passthrough_available():
+def jpeg_passthrough_available() -> bool:
     """Returns True if the installed version of Ghostscript supports JPEG passthru
 
     Prior to 9.23, Ghostscript decode and re-encoded JPEGs internally. In 9.23
@@ -79,94 +76,25 @@ def jpeg_passthrough_available():
     return version() >= '9.24'
 
 
-def _gs_error_reported(stream):
+def _gs_error_reported(stream) -> bool:
     return re.search(r'error', stream, flags=re.IGNORECASE)
 
 
-def extract_text(input_file, pageno=1):
-    """Use the txtwrite device to get text layout information out
-
-    For details on options of -dTextFormat see
-    https://www.ghostscript.com/doc/current/VectorDevices.htm#TXT
-
-    Format is like
-    <page>
-    <line>
-    <span bbox="left top right bottom" font="..." size="...">
-    <char bbox="...." c="X"/>
-
-    :param pageno: number of page to extract, or all pages if None
-    :return: XML-ish text representation in bytes
-    """
-
-    if pageno is not None:
-        pages = ['-dFirstPage=%i' % pageno, '-dLastPage=%i' % pageno]
-    else:
-        pages = []
-
-    # Note due to bug https://bugs.ghostscript.com/show_bug.cgi?id=701971
-    # Ghostscript <= 9.50 will truncate output unless we write to stdout, so
-    # don't write to a file.
-    args_gs = (
-        [
-            GS,
-            '-dQUIET',
-            '-dSAFER',
-            '-dBATCH',
-            '-dNOPAUSE',
-            '-sDEVICE=txtwrite',
-            '-dTextFormat=0',
-        ]
-        + pages
-        + ['-o', '-', fspath(input_file), "-sstdout=%stderr"]
-    )
-
-    try:
-        p = run(args_gs, stdout=PIPE, stderr=PIPE, check=True)
-    except CalledProcessError as e:
-        raise SubprocessOutputError(
-            'Ghostscript text extraction failed\n%s\n%s'
-            % (input_file, e.stderr.decode(errors='replace'))
-        )
-
-    return p.stdout
-
-
 def rasterize_pdf(
-    input_file,
-    output_file,
-    xres,
-    yres,
-    raster_device,
-    log,
-    pageno=1,
-    page_dpi=None,
-    rotation=None,
-    filter_vector=False,
+    input_file: os.PathLike,
+    output_file: os.PathLike,
+    *,
+    raster_device: str,
+    raster_dpi: Resolution,
+    pageno: int = 1,
+    page_dpi: Resolution = None,
+    rotation: int = None,
+    filter_vector: bool = False,
 ):
-    """Rasterize one page of a PDF at resolution (xres, yres) in canvas units.
-
-    The image is sized to match the integer pixels dimensions implied by
-    (xres, yres) even if those numbers are noninteger. The image's DPI will
-     be overridden with the values in page_dpi.
-
-    :param input_file: pathlike
-    :param output_file: pathlike
-    :param xres: resolution at which to rasterize page
-    :param yres:
-    :param raster_device:
-    :param log:
-    :param pageno: page number to rasterize (beginning at page 1)
-    :param page_dpi: resolution tuple (x, y) overriding output image DPI
-    :param rotation: 0, 90, 180, 270: clockwise angle to rotate page
-    :param filter_vector: if True, remove vector graphics objects
-    :return:
-    """
-    res = round(xres, 6), round(yres, 6)
+    """Rasterize one page of a PDF at resolution raster_dpi in canvas units."""
+    raster_dpi = raster_dpi.round(6)
     if not page_dpi:
-        page_dpi = res
-    if not log:
-        log = gslog
+        page_dpi = raster_dpi
 
     args_gs = (
         [
@@ -178,7 +106,7 @@ def rasterize_pdf(
             f'-sDEVICE={raster_device}',
             f'-dFirstPage={pageno}',
             f'-dLastPage={pageno}',
-            f'-r{res[0]:f}x{res[1]:f}',
+            f'-r{raster_dpi.x:f}x{raster_dpi.y:f}',
         ]
         + (['-dFILTERVECTOR'] if filter_vector else [])
         + [
@@ -191,12 +119,9 @@ def rasterize_pdf(
         ]
     )
 
-    log.debug(args_gs)
     try:
         p = run(args_gs, stdout=PIPE, stderr=PIPE, check=True)
     except CalledProcessError as e:
-        with suppress(OSError):
-            Path(output_file).unlink()  # no unfinished files
         log.error(e.stderr.decode(errors='replace'))
         raise SubprocessOutputError('Ghostscript rasterizing failed')
     else:
@@ -218,43 +143,17 @@ def rasterize_pdf(
             elif rotation == 270:
                 im = im.transpose(Image.ROTATE_270)
             if rotation % 180 == 90:
-                page_dpi = page_dpi[1], page_dpi[0]
+                page_dpi = page_dpi.flip_axis()
         im.save(fspath(output_file), dpi=page_dpi)
 
 
 def generate_pdfa(
     pdf_pages,
-    output_file,
-    compression,
-    log,
-    threads=None,  # deprecated parameter
-    pdf_version='1.5',
-    pdfa_part='2',
+    output_file: os.PathLike,
+    compression: str,
+    pdf_version: str = '1.5',
+    pdfa_part: str = '2',
 ):
-    """Generate a PDF/A.
-
-    The pdf_pages, a list files, will be merged into output_file. One or more
-    PDF files may be merged. One of the files in this list must be a pdfmark
-    file that provides Ghostscript with details on how to perform the PDF/A
-    conversion. By default with we pick PDF/A-2b, but this works for 1 or 3.
-
-    compression can be 'jpeg', 'lossless', or an empty string. In 'jpeg',
-    Ghostscript is instructed to convert color and grayscale images to DCT
-    (JPEG encoding). In 'lossless' Ghostscript is told to convert images to
-    Flate (lossless/PNG). If the parameter is omitted Ghostscript is left to
-    make its own decisions about how to encode images; it appears to use a
-    heuristic to decide how to encode images. As of Ghostscript 9.25, we
-    support passthrough JPEG which allows Ghostscript to avoid transcoding
-    images entirely. (The feature was added in 9.23 but broken, and the 9.24
-    release of Ghostscript had regressions, so we don't support it until 9.25.)
-    """
-    if not log:
-        log = gslog
-    if threads is not None:
-        warnings.warn(
-            "use of deprecated parameter 'threads'", category=DeprecationWarning
-        )
-
     compression_args = []
     if compression == 'jpeg':
         compression_args = [
@@ -313,15 +212,12 @@ def generate_pdfa(
         ]
     )
     args_gs.extend(fspath(s) for s in pdf_pages)  # Stringify Path objs
-    log.debug(args_gs)
     try:
         with Path(output_file).open('wb') as output:
             p = run(args_gs, stdout=output, stderr=PIPE, check=True)
     except CalledProcessError as e:
         # Ghostscript does not change return code when it fails to create
         # PDF/A - check PDF/A status elsewhere
-        with suppress(OSError):
-            Path(output_file).unlink()
         log.error(e.stderr.decode(errors='replace'))
         raise SubprocessOutputError('Ghostscript PDF/A rendering failed')
     else:
@@ -346,5 +242,3 @@ def generate_pdfa(
                 "Ghostscript had to remove PDF 'overprinting' from the "
                 "input file to complete PDF/A conversion. "
             )
-        else:
-            log.debug(stderr)

src/ocrmypdf/_exec/jbig2enc.py

@@ -17,14 +17,12 @@
 
 """Interface to jbig2 executable"""
 
-from functools import lru_cache
 from subprocess import PIPE
 
-from ..exceptions import MissingDependencyError
-from . import get_version, run
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.subprocess import get_version, run
 
 
-@lru_cache(maxsize=1)
 def version():
     return get_version('jbig2', regex=r'jbig2enc (\d+(\.\d+)*).*')
 

src/ocrmypdf/_exec/pngquant.py

@@ -17,17 +17,14 @@
 
 """Interface to pngquant executable"""
 
-from functools import lru_cache
-from subprocess import run
 from tempfile import NamedTemporaryFile
 
 from PIL import Image
 
-from ..exceptions import MissingDependencyError
-from . import get_version
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.subprocess import get_version, run
 
 
-@lru_cache(maxsize=1)
 def version():
     return get_version('pngquant', regex=r'(\d+(\.\d+)*).*')
 

src/ocrmypdf/_exec/tesseract.py

@@ -21,17 +21,21 @@ import logging
 import os
 import shutil
 from collections import namedtuple
-from contextlib import suppress
 from os import fspath
+from pathlib import Path
 from subprocess import PIPE, STDOUT, CalledProcessError, TimeoutExpired
+from typing import List
 
-from ..exceptions import (
+from PIL import Image
+
+from ocrmypdf.exceptions import (
     MissingDependencyError,
     SubprocessOutputError,
     TesseractConfigError,
 )
-from ..helpers import page_number, safe_symlink
-from . import get_version, run
+from ocrmypdf.subprocess import get_version, run
+
+log = logging.getLogger(__name__)
 
 OrientationConfidence = namedtuple('OrientationConfidence', ('angle', 'confidence'))
 
@@ -59,16 +63,11 @@ class TesseractLoggerAdapter(logging.LoggerAdapter):
         return '[tesseract] %s' % (msg), kwargs
 
 
-def version(tesseract_env=None):
-    return get_version('tesseract', regex=r'tesseract\s(.+)', env=tesseract_env)
+def version():
+    return get_version('tesseract', regex=r'tesseract\s(.+)')
 
 
-def v4(tesseract_env=None):
-    "Is this Tesseract v4.0?"
-    return version(tesseract_env) >= '4'
-
-
-def has_textonly_pdf(tesseract_env=None, langs=None):
+def has_textonly_pdf(langs=None):
     """Does Tesseract have textonly_pdf capability?
 
     Available in v4.00.00alpha since January 2017. Best to
@@ -77,34 +76,28 @@ def has_textonly_pdf(tesseract_env=None, langs=None):
     args_tess = tess_base_args(langs, engine_mode=None) + ['--print-parameters', 'pdf']
     params = ''
     try:
-        proc = run(
-            args_tess,
-            check=True,
-            universal_newlines=True,
-            stdout=PIPE,
-            stderr=STDOUT,
-            env=tesseract_env,
-        )
+        proc = run(args_tess, check=True, stdout=PIPE, stderr=STDOUT)
         params = proc.stdout
     except CalledProcessError as e:
         raise MissingDependencyError(
-            "Could not --print-parameters from tesseract"
+            "Could not --print-parameters from tesseract. This can happen if the "
+            "TESSDATA_PREFIX environment is not set to a valid tessdata folder. "
         ) from e
-    if 'textonly_pdf' in params:
+    if b'textonly_pdf' in params:
         return True
     return False
 
 
-def has_user_words(tesseract_env=None):
+def has_user_words():
     """Does Tesseract have --user-words capability?
 
     Not available in 4.0, but available in 4.1. Also available in 3.x, but
     we no longer support 3.x.
     """
-    return version(tesseract_env) >= '4.1'
+    return version() >= '4.1'
 
 
-def languages(tesseract_env=None):
+def get_languages():
     def lang_error(output):
         msg = (
             "Tesseract failed to report available languages.\n"
@@ -117,12 +110,7 @@ def languages(tesseract_env=None):
     args_tess = ['tesseract', '--list-langs']
     try:
         proc = run(
-            args_tess,
-            universal_newlines=True,
-            stdout=PIPE,
-            stderr=STDOUT,
-            check=True,
-            env=tesseract_env,
+            args_tess, universal_newlines=True, stdout=PIPE, stderr=STDOUT, check=True
         )
         output = proc.stdout
     except CalledProcessError as e:
@@ -131,11 +119,11 @@ def languages(tesseract_env=None):
     for line in output.splitlines():
         if line.startswith('Error'):
             raise MissingDependencyError(lang_error(output))
-    header, *rest = output.splitlines()
+    _header, *rest = output.splitlines()
     return set(lang.strip() for lang in rest)
 
 
-def tess_base_args(langs, engine_mode):
+def tess_base_args(langs: List[str], engine_mode) -> List[str]:
     args = ['tesseract']
     if langs:
         args.extend(['-l', '+'.join(langs)])
@@ -144,7 +132,7 @@ def tess_base_args(langs, engine_mode):
     return args
 
 
-def get_orientation(input_file, engine_mode, timeout: float, log, tesseract_env=None):
+def get_orientation(input_file: Path, engine_mode, timeout: float):
     args_tesseract = tess_base_args(['osd'], engine_mode) + [
         '--psm',
         '0',
@@ -153,19 +141,13 @@ def get_orientation(input_file, engine_mode, timeout: float, log, tesseract_env=
     ]
 
     try:
-        p = run(
-            args_tesseract,
-            stdout=PIPE,
-            stderr=STDOUT,
-            timeout=timeout,
-            check=True,
-            env=tesseract_env,
-        )
+        p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
         stdout = p.stdout
     except TimeoutExpired:
         return OrientationConfidence(angle=0, confidence=0.0)
     except CalledProcessError as e:
-        tesseract_log_output(log, e.output, input_file)
+        tesseract_log_output(e.stdout)
+        tesseract_log_output(e.stderr)
         if (
             b'Too few characters. Skipping this page' in e.output
             or b'Image too large' in e.output
@@ -187,15 +169,17 @@ def get_orientation(input_file, engine_mode, timeout: float, log, tesseract_env=
         return oc
 
 
-def tesseract_log_output(mainlog, stdout, input_file):
-    log = TesseractLoggerAdapter(
-        mainlog, extra=mainlog.extra if hasattr(mainlog, 'extra') else None
+def tesseract_log_output(stream):
+    tlog = TesseractLoggerAdapter(
+        log, extra=log.extra if hasattr(log, 'extra') else None
     )
 
+    if not stream:
+        return
     try:
-        text = stdout.decode()
+        text = stream.decode()
     except UnicodeDecodeError:
-        text = stdout.decode('utf-8', 'ignore')
+        text = stream.decode('utf-8', 'ignore')
 
     lines = text.splitlines()
     for line in lines:
@@ -204,67 +188,58 @@ def tesseract_log_output(mainlog, stdout, input_file):
         elif line.startswith("Warning in pixReadMem"):
             continue
         elif 'diacritics' in line:
-            log.warning("lots of diacritics - possibly poor OCR")
+            tlog.warning("lots of diacritics - possibly poor OCR")
         elif line.startswith('OSD: Weak margin'):
-            log.warning("unsure about page orientation")
+            tlog.warning("unsure about page orientation")
         elif 'Error in pixScanForForeground' in line:
             pass  # Appears to be spurious/problem with nonwhite borders
         elif 'Error in boxClipToRectangle' in line:
             pass  # Always appears with pixScanForForeground message
         elif 'parameter not found: ' in line.lower():
-            log.error(line.strip())
+            tlog.error(line.strip())
             problem = line.split('found: ')[1]
             raise TesseractConfigError(problem)
         elif 'error' in line.lower() or 'exception' in line.lower():
-            log.error(line.strip())
+            tlog.error(line.strip())
         elif 'warning' in line.lower():
-            log.warning(line.strip())
+            tlog.warning(line.strip())
         elif 'read_params_file' in line.lower():
-            log.error(line.strip())
+            tlog.error(line.strip())
         else:
-            log.info(line.strip())
+            tlog.info(line.strip())
 
 
-def page_timedout(log, input_file, timeout):
+def page_timedout(timeout):
     if timeout == 0:
         return
-    prefix = f"{(page_number(input_file)):4d}: [tesseract] "
-    log.warning(prefix + " took too long to OCR - skipping")
+    log.warning("[tesseract] took too long to OCR - skipping")
 
 
-def _generate_null_hocr(output_hocr, output_sidecar, image):
+def _generate_null_hocr(output_hocr, output_text, image):
     """Produce a .hocr file that reports no text detected on a page that is
     the same size as the input image."""
-    from PIL import Image
-
     with Image.open(image) as im:
         w, h = im.size
 
-    with open(output_hocr, 'w', encoding="utf-8") as f:
-        f.write(HOCR_TEMPLATE.format(w, h))
-    with open(output_sidecar, 'w', encoding='utf-8') as f:
-        f.write('[skipped page]')
+    output_hocr.write_text(HOCR_TEMPLATE.format(w, h), encoding='utf-8')
+    output_text.write_text('[skipped page]', encoding='utf-8')
 
 
 def generate_hocr(
-    input_file,
-    output_files,
-    language: list,
+    input_file: Path,
+    output_hocr: Path,
+    output_text: Path,
+    languages: list,
     engine_mode,
     tessconfig: list,
     timeout: float,
     pagesegmode: int,
     user_words,
     user_patterns,
-    tesseract_env,
-    log,
 ):
+    prefix = output_hocr.with_suffix('')
 
-    output_hocr = next(o for o in output_files if o.endswith('.hocr'))
-    output_sidecar = next(o for o in output_files if o.endswith('.txt'))
-    prefix = os.path.splitext(output_hocr)[0]
-
-    args_tesseract = tess_base_args(language, engine_mode)
+    args_tesseract = tess_base_args(languages, engine_mode)
 
     if pagesegmode is not None:
         args_tesseract.extend(['--psm', str(pagesegmode)])
@@ -275,98 +250,70 @@ def generate_hocr(
     if user_patterns:
         args_tesseract.extend(['--user-patterns', user_patterns])
 
-    if user_words or user_patterns:
-        args_tesseract.extend(['-c', 'lstm_use_matrix=1'])
-
-    # Reminder: test suite tesseract spoofers will break after any changes
+    # Reminder: test suite tesseract test plugins will break after any changes
     # to the number of order parameters here
     args_tesseract.extend([input_file, prefix, 'hocr', 'txt'] + tessconfig)
     try:
-        log.debug(args_tesseract)
-        p = run(
-            args_tesseract,
-            stdout=PIPE,
-            stderr=STDOUT,
-            timeout=timeout,
-            check=True,
-            env=tesseract_env,
-        )
+        p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
         stdout = p.stdout
     except TimeoutExpired:
         # Generate a HOCR file with no recognized text if tesseract times out
         # Temporary workaround to hocrTransform not being able to function if
         # it does not have a valid hOCR file.
-        page_timedout(log, input_file, timeout)
-        _generate_null_hocr(output_hocr, output_sidecar, input_file)
+        page_timedout(timeout)
+        _generate_null_hocr(output_hocr, output_text, input_file)
     except CalledProcessError as e:
-        tesseract_log_output(log, e.output, input_file)
+        tesseract_log_output(e.output)
         if b'Image too large' in e.output:
-            _generate_null_hocr(output_hocr, output_sidecar, input_file)
+            _generate_null_hocr(output_hocr, output_text, input_file)
             return
 
         raise SubprocessOutputError() from e
     else:
-        tesseract_log_output(log, stdout, input_file)
+        tesseract_log_output(stdout)
         # The sidecar text file will get the suffix .txt; rename it to
         # whatever caller wants it named
-        if os.path.exists(prefix + '.txt'):
-            shutil.move(prefix + '.txt', output_sidecar)
+        if prefix.with_suffix('.txt').exists():
+            shutil.move(prefix.with_suffix('.txt'), output_text)
 
 
-def use_skip_page(text_only, skip_pdf, output_pdf, output_text):
-    with open(output_text, 'w') as f:
-        f.write('[skipped page]')
+def use_skip_page(output_pdf, output_text):
+    output_text.write_text('[skipped page]', encoding='utf-8')
 
-    if skip_pdf and not text_only:
-        # Substitute a "skipped page"
-        with suppress(FileNotFoundError):
-            os.remove(output_pdf)  # In case it was partially created
-        safe_symlink(skip_pdf, output_pdf)
-        return
-
-    # Or normally, just write a 0 byte file to the output to indicate a skip
-    with open(output_pdf, 'wb') as out:
-        out.write(b'')
+    # A 0 byte file to the output to indicate a skip
+    output_pdf.write_bytes(b'')
 
 
 def generate_pdf(
     *,
-    input_image,
-    skip_pdf=None,
-    output_pdf,
-    output_text,
-    language: list,
+    input_file: Path,
+    output_pdf: Path,
+    output_text: Path,
+    languages: List[str],
     engine_mode,
-    text_only: bool,
-    tessconfig: list,
+    tessconfig: List[str],
     timeout: float,
     pagesegmode: int,
     user_words,
     user_patterns,
-    tesseract_env,
-    log,
 ):
     """Use Tesseract to render a PDF.
 
-    input_image -- image to analyze
-    skip_pdf -- if we time out, use this file as output
+    input_file -- image to analyze
     output_pdf -- file to generate
     output_text -- OCR text file
-    language -- list of languages to consider
+    languages -- list of languages to consider
     engine_mode -- engine mode argument for tess v4
-    text_only -- enable tesseract text only mode?
     tessconfig -- tesseract configuration
     timeout -- timeout (seconds)
-    log -- logger object
     """
 
-    args_tesseract = tess_base_args(language, engine_mode)
+    args_tesseract = tess_base_args(languages, engine_mode)
 
     if pagesegmode is not None:
         args_tesseract.extend(['--psm', str(pagesegmode)])
 
-    if text_only and has_textonly_pdf(tesseract_env, language):
-        args_tesseract.extend(['-c', 'textonly_pdf=1'])
+    args_tesseract.extend(['-c', 'textonly_pdf=1'])
 
     if user_words:
         args_tesseract.extend(['--user-words', user_words])
@@ -376,31 +323,23 @@ def generate_pdf(
 
     prefix = os.path.splitext(output_pdf)[0]  # Tesseract appends suffixes
 
-    # Reminder: test suite tesseract spoofers might break after any changes
+    # Reminder: test suite tesseract test plugins might break after any changes
     # to the number of order parameters here
 
-    args_tesseract.extend([input_image, prefix, 'pdf', 'txt'] + tessconfig)
+    args_tesseract.extend([input_file, prefix, 'pdf', 'txt'] + tessconfig)
     try:
-        log.debug(args_tesseract)
-        p = run(
-            args_tesseract,
-            stdout=PIPE,
-            stderr=STDOUT,
-            timeout=timeout,
-            check=True,
-            env=tesseract_env,
-        )
+        p = run(args_tesseract, stdout=PIPE, stderr=STDOUT, timeout=timeout, check=True)
         stdout = p.stdout
         if os.path.exists(prefix + '.txt'):
             shutil.move(prefix + '.txt', output_text)
     except TimeoutExpired:
-        page_timedout(log, input_image, timeout)
-        use_skip_page(text_only, skip_pdf, output_pdf, output_text)
+        page_timedout(timeout)
+        use_skip_page(output_pdf, output_text)
     except CalledProcessError as e:
-        tesseract_log_output(log, e.output, input_image)
+        tesseract_log_output(e.output)
         if b'Image too large' in e.output:
-            use_skip_page(text_only, skip_pdf, output_pdf, output_text)
+            use_skip_page(output_pdf, output_text)
             return
         raise SubprocessOutputError() from e
     else:
-        tesseract_log_output(log, stdout, input_image)
+        tesseract_log_output(stdout)

src/ocrmypdf/_exec/unpaper.py

@@ -20,31 +20,32 @@
 
 """Interface to unpaper executable"""
 
+import logging
 import os
 import shlex
-from functools import lru_cache
+from pathlib import Path
 from subprocess import PIPE, STDOUT, CalledProcessError
 from tempfile import TemporaryDirectory
+from typing import Tuple
 
 from PIL import Image
 
-from ..exceptions import MissingDependencyError, SubprocessOutputError
-from . import get_version
-from . import run as external_run
+from ocrmypdf.exceptions import MissingDependencyError, SubprocessOutputError
+from ocrmypdf.subprocess import get_version
+from ocrmypdf.subprocess import run as external_run
+
+log = logging.getLogger(__name__)
 
 
-@lru_cache(maxsize=1)
 def version():
     return get_version('unpaper')
 
 
-def run(input_file, output_file, dpi, log, mode_args):
-    args_unpaper = ['unpaper', '-v', '--dpi', str(dpi)] + mode_args
-
+def _setup_unpaper_io(tmpdir: Path, input_file: Path) -> Tuple[Path, Path]:
     SUFFIXES = {'1': '.pbm', 'L': '.pgm', 'RGB': '.ppm'}
-
-    with TemporaryDirectory() as tmpdir, Image.open(input_file) as im:
-        if im.mode not in SUFFIXES.keys():
+    with Image.open(input_file) as im:
+        im_modified = False
+        if im.mode not in SUFFIXES:
             log.info("Converting image to other colorspace")
             try:
                 if im.mode == 'P' and len(im.getcolors()) == 2:
@@ -52,11 +53,11 @@ def run(input_file, output_file, dpi, log, mode_args):
                 else:
                     im = im.convert(mode='RGB')
             except IOError as e:
-                im.close()
                 raise MissingDependencyError(
                     "Could not convert image with type " + im.mode
                 ) from e
-
+            else:
+                im_modified = True
         try:
             suffix = SUFFIXES[im.mode]
         except KeyError:
@@ -64,9 +65,21 @@ def run(input_file, output_file, dpi, log, mode_args):
                 "Failed to convert image to a supported format."
             ) from e
 
-        input_pnm = os.path.join(tmpdir, f'input{suffix}')
-        output_pnm = os.path.join(tmpdir, f'output{suffix}')
-        im.save(input_pnm, format='PPM')
+        if im_modified or input_file.suffix != '.png':
+            input_png = tmpdir / 'input.png'
+            im.save(input_png, format='PNG', compress_level=1)
+        else:
+            # No changes, PNG input, just use the file we already have
+            input_png = input_file
+        output_pnm = tmpdir / f'output{suffix}'
+    return input_png, output_pnm
+
+
+def run(input_file, output_file, dpi, mode_args):
+    args_unpaper = ['unpaper', '-v', '--dpi', str(dpi)] + mode_args
+
+    with TemporaryDirectory() as tmpdir:
+        input_png, output_pnm = _setup_unpaper_io(Path(tmpdir), input_file)
 
         # To prevent any shenanigans from accepting arbitrary parameters in
         # --unpaper-args, we:
@@ -75,23 +88,22 @@ def run(input_file, output_file, dpi, log, mode_args):
         # 3) append absolute paths for the input and output file
         # This should ensure that a user cannot clobber some other file with
         # their unpaper arguments (whether intentionally or otherwise)
-        args_unpaper.extend([input_pnm, output_pnm])
+        args_unpaper.extend([os.fspath(input_png), os.fspath(output_pnm)])
         try:
             proc = external_run(
                 args_unpaper,
                 check=True,
                 close_fds=True,
                 universal_newlines=True,
-                stderr=STDOUT,
-                cwd=tmpdir,
+                stderr=STDOUT,  # unpaper writes logging output to stdout and stderr
+                cwd=tmpdir,  # and cannot send file output to stdout
                 stdout=PIPE,
             )
         except CalledProcessError as e:
-            log.debug(e.output)
+            log.debug(e.stderr)
             raise e from e
         else:
-            log.debug(proc.stdout)
-            # unpaper sets dpi to 72; fix this
+            log.debug(proc.stderr)
             try:
                 with Image.open(output_pnm) as imout:
                     imout.save(output_file, dpi=(dpi, dpi))
@@ -110,7 +122,7 @@ def validate_custom_args(args: str):
     return unpaper_args
 
 
-def clean(input_file, output_file, dpi, log, unpaper_args=None):
+def clean(input_file, output_file, dpi, unpaper_args=None):
     default_args = [
         '--layout',
         'none',
@@ -124,4 +136,4 @@ def clean(input_file, output_file, dpi, log, unpaper_args=None):
     ]
     if not unpaper_args:
         unpaper_args = default_args
-    run(input_file, output_file, dpi, log, unpaper_args)
+    run(input_file, output_file, dpi, unpaper_args)

src/ocrmypdf/_graft.py

@@ -15,12 +15,14 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import os
+import logging
 from contextlib import suppress
 from pathlib import Path
+from typing import Optional
 
 import pikepdf
 
+log = logging.getLogger(__name__)
 MAX_REPLACE_PAGES = 100
 
 
@@ -88,99 +90,10 @@ def strip_invisible_text(pdf, page):
     page.Contents = pikepdf.Stream(pdf, content_stream)
 
 
-def _graft_text_layer(
-    *, pdf_base, page_num, text, font, font_key, procset, rotation, strip_old_text, log
-):
-    """Insert the text layer from text page 0 on to pdf_base at page_num"""
-
-    log.debug("Grafting")
-    if Path(text).stat().st_size == 0:
-        return
-
-    # This is a pointer indicating a specific page in the base file
-    pdf_text = pikepdf.open(text)
-    pdf_text_contents = pdf_text.pages[0].Contents.read_bytes()
-
-    base_page = pdf_base.pages.p(page_num)
-
-    # The text page always will be oriented up by this stage but the original
-    # content may have a rotation applied. Wrap the text stream with a rotation
-    # so it will be oriented the same way as the rest of the page content.
-    # (Previous versions OCRmyPDF rotated the content layer to match the text.)
-    mediabox = [float(pdf_text.pages[0].MediaBox[v]) for v in range(4)]
-    wt, ht = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
-
-    mediabox = [float(base_page.MediaBox[v]) for v in range(4)]
-    wp, hp = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
-
-    translate = pikepdf.PdfMatrix().translated(-wt / 2, -ht / 2)
-    untranslate = pikepdf.PdfMatrix().translated(wp / 2, hp / 2)
-    corner = pikepdf.PdfMatrix().translated(mediabox[0], mediabox[1])
-    # -rotation because the input is a clockwise angle and this formula
-    # uses CCW
-    rotation = -rotation % 360
-    rotate = pikepdf.PdfMatrix().rotated(rotation)
-
-    # Because of rounding of DPI, we might get a text layer that is not
-    # identically sized to the target page. Scale to adjust. Normally this
-    # is within 0.998.
-    if rotation in (90, 270):
-        wt, ht = ht, wt
-    scale_x = wp / wt
-    scale_y = hp / ht
-
-    # log.debug('%r', scale_x, scale_y)
-    scale = pikepdf.PdfMatrix().scaled(scale_x, scale_y)
-
-    # Translate the text so it is centered at (0, 0), rotate it there, adjust
-    # for a size different between initial and text PDF, then untranslate, and
-    # finally move the lower left corner to match the mediabox
-    ctm = translate @ rotate @ scale @ untranslate @ corner
-
-    pdf_text_contents = b'q %s cm\n' % ctm.encode() + pdf_text_contents + b'\nQ\n'
-
-    new_text_layer = pikepdf.Stream(pdf_base, pdf_text_contents)
-
-    if strip_old_text:
-        strip_invisible_text(pdf_base, base_page)
-
-    base_page.page_contents_add(new_text_layer, prepend=True)
-
-    _update_page_resources(
-        page=base_page, font=font, font_key=font_key, procset=procset
-    )
-    pdf_text.close()
-
-
-def _find_font(text, pdf_base):
-    """Copy a font from the filename text into pdf_base"""
-
-    font, font_key = None, None
-    possible_font_names = ('/f-0-0', '/F1')
-    try:
-        with pikepdf.open(text) as pdf_text:
-            try:
-                pdf_text_fonts = pdf_text.pages[0].Resources.get('/Font', {})
-            except (AttributeError, IndexError, KeyError):
-                return None, None
-            for f in possible_font_names:
-                pdf_text_font = pdf_text_fonts.get(f, None)
-                if pdf_text_font is not None:
-                    font_key = f
-                    break
-            if pdf_text_font:
-                font = pdf_base.copy_foreign(pdf_text_font)
-            return font, font_key
-    except (FileNotFoundError, pikepdf.PdfError):
-        # PdfError occurs if a 0-length file is written e.g. due to OCR timeout
-        return None, None
-
-
 class OcrGrafter:
     def __init__(self, context):
         self.context = context
-        self.log = context.log
-        self.path_base = Path(context.origin).resolve()
+        self.path_base = context.origin
 
         self.pdf_base = pikepdf.open(self.path_base)
         self.font, self.font_key = None, None
@@ -195,10 +108,16 @@ class OcrGrafter:
         self.emplacements = 1
         self.interim_count = 0
 
-    def graft_page(self, page_result):
-        pageno, image, text, _sidecar, autorotate_correction = page_result
-        if text and not self.font:
-            self.font, self.font_key = _find_font(text, self.pdf_base)
+    def graft_page(
+        self,
+        *,
+        pageno: int,
+        image: Optional[Path],
+        textpdf: Optional[Path],
+        autorotate_correction: int,
+    ):
+        if textpdf and not self.font:
+            self.font, self.font_key = self._find_font(textpdf)
 
         emplaced_page = False
         content_rotation = self.pdfinfo[pageno].rotation
@@ -206,7 +125,7 @@ class OcrGrafter:
         if path_image is not None and path_image != self.path_base:
             # We are updating the old page with a rasterized PDF of the new
             # page (without changing objgen, to preserve references)
-            self.log.debug("Emplacement update")
+            log.debug("Emplacement update")
             with pikepdf.open(image) as pdf_image:
                 self.emplacements += 1
                 foreign_image_page = pdf_image.pages[0]
@@ -220,25 +139,23 @@ class OcrGrafter:
             content_rotation = autorotate_correction
         text_rotation = autorotate_correction
         text_misaligned = (text_rotation - content_rotation) % 360
-        self.log.debug(
+        log.debug(
             f"Rotations for page {pageno}: [text, auto, misalign, content] = "
             f"{text_rotation}, {autorotate_correction}, "
             f"{text_misaligned}, {content_rotation}"
         )
 
-        if text and self.font:
+        if textpdf and self.font:
             # Graft the text layer onto this page, whether new or old
             strip_old = self.context.options.redo_ocr
-            _graft_text_layer(
-                pdf_base=self.pdf_base,
+            self._graft_text_layer(
                 page_num=pageno + 1,
-                text=text,
+                textpdf=textpdf,
                 font=self.font,
                 font_key=self.font_key,
                 rotation=text_misaligned,
                 procset=self.procset,
                 strip_old_text=strip_old,
-                log=self.log,
             )
 
         # Correct the rotation if applicable
@@ -250,10 +167,13 @@ class OcrGrafter:
             self.save_and_reload()
 
     def save_and_reload(self):
-        # Periodically save and reload the Pdf object. This will keep a
-        # lid on our memory usage for very large files. Attach the font to
-        # page 1 even if page 1 doesn't use it, so we have a way to get it
-        # back.
+        """Save and reload the Pdf.
+
+        This will keep a lid on our memory usage for very large files. Attach
+        the font to page 1 even if page 1 doesn't use it, so we have a way to get it
+        back.
+        """
+
         page0 = self.pdf_base.pages[0]
         _update_page_resources(
             page=page0, font=self.font, font_key=self.font_key, procset=self.procset
@@ -264,12 +184,14 @@ class OcrGrafter:
         # {interim_count} is the opened file we were updateing
         # {interim_count - 1} can be deleted
         # {interim_count + 1} is the new file will produce and open
-        old_file = self.output_file + f'_working{self.interim_count - 1}.pdf'
+        old_file = self.output_file.with_suffix(f'.working{self.interim_count - 1}.pdf')
         if not self.context.options.keep_temporary_files:
             with suppress(FileNotFoundError):
-                os.unlink(old_file)
+                old_file.unlink()
 
-        next_file = self.output_file + f'_working{self.interim_count + 1}.pdf'
+        next_file = self.output_file.with_suffix(
+            f'.working{self.interim_count + 1}.pdf'
+        )
         self.pdf_base.save(next_file)
         self.pdf_base.close()
 
@@ -282,3 +204,98 @@ class OcrGrafter:
         self.pdf_base.save(self.output_file)
         self.pdf_base.close()
         return self.output_file
+
+    def _find_font(self, text):
+        """Copy a font from the filename text into pdf_base"""
+
+        font, font_key = None, None
+        possible_font_names = ('/f-0-0', '/F1')
+        try:
+            with pikepdf.open(text) as pdf_text:
+                try:
+                    pdf_text_fonts = pdf_text.pages[0].Resources.get('/Font', {})
+                except (AttributeError, IndexError, KeyError):
+                    return None, None
+                for f in possible_font_names:
+                    pdf_text_font = pdf_text_fonts.get(f, None)
+                    if pdf_text_font is not None:
+                        font_key = f
+                        break
+                if pdf_text_font:
+                    font = self.pdf_base.copy_foreign(pdf_text_font)
+                return font, font_key
+        except (FileNotFoundError, pikepdf.PdfError):
+            # PdfError occurs if a 0-length file is written e.g. due to OCR timeout
+            return None, None
+
+    def _graft_text_layer(
+        self,
+        *,
+        page_num: int,
+        textpdf: Path,
+        font: pikepdf.Object,
+        font_key: pikepdf.Object,
+        procset: pikepdf.Object,
+        rotation: int,
+        strip_old_text: bool,
+    ):
+        """Insert the text layer from text page 0 on to pdf_base at page_num"""
+
+        log.debug("Grafting")
+        if Path(textpdf).stat().st_size == 0:
+            return
+
+        # This is a pointer indicating a specific page in the base file
+        with pikepdf.open(textpdf) as pdf_text:
+            pdf_text_contents = pdf_text.pages[0].Contents.read_bytes()
+
+            base_page = self.pdf_base.pages.p(page_num)
+
+            # The text page always will be oriented up by this stage but the original
+            # content may have a rotation applied. Wrap the text stream with a rotation
+            # so it will be oriented the same way as the rest of the page content.
+            # (Previous versions OCRmyPDF rotated the content layer to match the text.)
+            mediabox = [float(pdf_text.pages[0].MediaBox[v]) for v in range(4)]
+            wt, ht = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
+
+            mediabox = [float(base_page.MediaBox[v]) for v in range(4)]
+            wp, hp = mediabox[2] - mediabox[0], mediabox[3] - mediabox[1]
+
+            translate = pikepdf.PdfMatrix().translated(-wt / 2, -ht / 2)
+            untranslate = pikepdf.PdfMatrix().translated(wp / 2, hp / 2)
+            corner = pikepdf.PdfMatrix().translated(mediabox[0], mediabox[1])
+            # -rotation because the input is a clockwise angle and this formula
+            # uses CCW
+            rotation = -rotation % 360
+            rotate = pikepdf.PdfMatrix().rotated(rotation)
+
+            # Because of rounding of DPI, we might get a text layer that is not
+            # identically sized to the target page. Scale to adjust. Normally this
+            # is within 0.998.
+            if rotation in (90, 270):
+                wt, ht = ht, wt
+            scale_x = wp / wt
+            scale_y = hp / ht
+
+            # log.debug('%r', scale_x, scale_y)
+            scale = pikepdf.PdfMatrix().scaled(scale_x, scale_y)
+
+            # Translate the text so it is centered at (0, 0), rotate it there, adjust
+            # for a size different between initial and text PDF, then untranslate, and
+            # finally move the lower left corner to match the mediabox
+            ctm = translate @ rotate @ scale @ untranslate @ corner
+
+            pdf_text_contents = (
+                b'q %s cm\n' % ctm.encode() + pdf_text_contents + b'\nQ\n'
+            )
+
+            new_text_layer = pikepdf.Stream(self.pdf_base, pdf_text_contents)
+
+            if strip_old_text:
+                strip_invisible_text(self.pdf_base, base_page)
+
+            base_page.page_contents_add(new_text_layer, prepend=True)
+
+            _update_page_resources(
+                page=base_page, font=font, font_key=font_key, procset=procset
+            )

src/ocrmypdf/_jobcontext.py

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

src/ocrmypdf/_logging.py

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

src/ocrmypdf/_pipeline.py

@@ -15,52 +15,58 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
+import logging
 import os
 import re
 import sys
 from datetime import datetime, timezone
+from pathlib import Path
 from shutil import copyfileobj
+from typing import Dict, Iterable, Optional
 
 import img2pdf
 import pikepdf
 from pikepdf.models.metadata import encode_pdf_date
-from PIL import Image
+from PIL import Image, ImageColor, ImageDraw
 
-from . import leptonica
-from ._version import PROGRAM_NAME
-from ._version import __version__ as VERSION
-from .exceptions import (
+from ocrmypdf import leptonica
+from ocrmypdf._exec import unpaper
+from ocrmypdf._jobcontext import PageContext, PdfContext
+from ocrmypdf._version import PROGRAM_NAME
+from ocrmypdf._version import __version__ as VERSION
+from ocrmypdf.exceptions import (
     DpiError,
     EncryptedPdfError,
     InputFileError,
     PriorOcrFoundError,
     UnsupportedImageFormatError,
 )
-from .exec import ghostscript, tesseract
-from .helpers import safe_symlink
-from .hocrtransform import HocrTransform
-from .optimize import optimize
-from .pdfa import generate_pdfa_ps
-from .pdfinfo import Colorspace, Encoding, PdfInfo
+from ocrmypdf.helpers import Resolution, safe_symlink
+from ocrmypdf.hocrtransform import HocrTransform
+from ocrmypdf.optimize import optimize
+from ocrmypdf.pdfa import generate_pdfa_ps
+from ocrmypdf.pdfinfo import Colorspace, Encoding, PdfInfo
+
+log = logging.getLogger(__name__)
 
 VECTOR_PAGE_DPI = 400
 
 
-def triage_image_file(input_file, output_file, options, log):
+def triage_image_file(input_file, output_file, options):
     log.info("Input file is not a PDF, checking if it is an image...")
     try:
         im = Image.open(input_file)
     except EnvironmentError as e:
         # Recover the original filename
-        log.error(str(e).replace(input_file, options.input_file))
+        log.error(str(e).replace(str(input_file), str(options.input_file)))
         raise UnsupportedImageFormatError() from e
 
     with im:
         log.info("Input file is an image")
         if 'dpi' in im.info:
             if im.info['dpi'] <= (96, 96) and not options.image_dpi:
-                log.info("Image size: (%d, %d)" % im.size)
-                log.info("Image resolution: (%d, %d)" % im.info['dpi'])
+                log.info("Image size: (%d, %d)", *im.size)
+                log.info("Image resolution: (%d, %d)", *im.info['dpi'])
                 log.error(
                     "Input file is an image, but the resolution (DPI) is "
                     "not credible.  Estimate the resolution at which the "
@@ -68,7 +74,7 @@ def triage_image_file(input_file, output_file, options, log):
                 )
                 raise DpiError()
         elif not options.image_dpi:
-            log.info("Image size: (%d, %d)" % im.size)
+            log.info("Image size: (%d, %d)", *im.size)
             log.error(
                 "Input file is an image, but has no resolution (DPI) "
                 "in its metadata.  Estimate the resolution at which "
@@ -95,11 +101,14 @@ def triage_image_file(input_file, output_file, options, log):
         layout_fun = img2pdf.default_layout_fun
         if options.image_dpi:
             layout_fun = img2pdf.get_fixed_dpi_layout_fun(
-                (options.image_dpi, options.image_dpi)
+                Resolution(options.image_dpi, options.image_dpi)
             )
         with open(output_file, 'wb') as outf:
             img2pdf.convert(
-                input_file, layout_fun=layout_fun, with_pdfrw=False, outputstream=outf
+                os.fspath(input_file),
+                layout_fun=layout_fun,
+                with_pdfrw=False,
+                outputstream=outf,
             )
         log.info("Successfully converted to PDF, processing...")
     except img2pdf.ImageOpenError as e:
@@ -123,7 +132,7 @@ def _pdf_guess_version(input_file, search_window=1024):
     return ''
 
 
-def triage(input_file, output_file, options, log):
+def triage(original_filename, input_file, output_file, options):
     try:
         if _pdf_guess_version(input_file):
             if options.image_dpi:
@@ -135,17 +144,28 @@ def triage(input_file, output_file, options, log):
             safe_symlink(input_file, output_file)
             return output_file
     except EnvironmentError as e:
-        log.error(e)
-        raise InputFileError() from e
+        log.debug(f"Temporary file was at: {input_file}")
+        msg = str(e).replace(str(input_file), original_filename)
+        raise InputFileError(msg) from e
 
-    triage_image_file(input_file, output_file, options, log)
+    triage_image_file(input_file, output_file, options)
     return output_file
 
 
-def get_pdfinfo(input_file, detailed_page_analysis=False, progbar=False):
+def get_pdfinfo(
+    input_file,
+    detailed_analysis=False,
+    progbar=False,
+    max_workers=None,
+    check_pages=None,
+):
     try:
         return PdfInfo(
-            input_file, detailed_page_analysis=detailed_page_analysis, progbar=progbar
+            input_file,
+            detailed_analysis=detailed_analysis,
+            progbar=progbar,
+            max_workers=max_workers,
+            check_pages=check_pages,
         )
     except pikepdf.PasswordError:
         raise EncryptedPdfError()
@@ -153,8 +173,7 @@ def get_pdfinfo(input_file, detailed_page_analysis=False, progbar=False):
         raise InputFileError()
 
 
-def validate_pdfinfo_options(context):
-    log = context.log
+def validate_pdfinfo_options(context: PdfContext):
     pdfinfo = context.pdfinfo
     options = context.options
 
@@ -181,7 +200,7 @@ def validate_pdfinfo_options(context):
             )
             raise InputFileError()
         else:
-            log.warn(
+            log.warning(
                 "This PDF has a fillable form. "
                 "Chances are it is a pure digital "
                 "document that does not need OCR."
@@ -192,54 +211,56 @@ def validate_pdfinfo_options(context):
                     "form and all filled form fields. The output PDF will be "
                     "'flattened' and will no longer be fillable."
                 )
+    context.plugin_manager.hook.validate(pdfinfo=pdfinfo, options=options)
 
 
 def get_page_dpi(pageinfo, options):
     "Get the DPI when nonsquare DPI is tolerable"
     xres = max(
-        pageinfo.xres or VECTOR_PAGE_DPI,
-        options.oversample or 0,
-        VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
+        pageinfo.dpi.x or VECTOR_PAGE_DPI,
+        options.oversample or 0.0,
+        VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
     )
     yres = max(
-        pageinfo.yres or VECTOR_PAGE_DPI,
+        pageinfo.dpi.y or VECTOR_PAGE_DPI,
         options.oversample or 0,
-        VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
+        VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
     )
-    return (float(xres), float(yres))
+    return Resolution(float(xres), float(yres))
 
 
-def get_page_square_dpi(pageinfo, options):
+def get_page_square_dpi(pageinfo, options) -> Resolution:
     "Get the DPI when we require xres == yres, scaled to physical units"
-    xres = pageinfo.xres or 0
-    yres = pageinfo.yres or 0
-    userunit = pageinfo.userunit or 1
-    return float(
+    xres = pageinfo.dpi.x or 0.0
+    yres = pageinfo.dpi.y or 0.0
+    userunit = float(pageinfo.userunit) or 1.0
+    units = float(
         max(
             (xres * userunit) or VECTOR_PAGE_DPI,
             (yres * userunit) or VECTOR_PAGE_DPI,
-            VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
-            options.oversample or 0,
+            VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
+            options.oversample or 0.0,
         )
     )
+    return Resolution(units, units)
 
 
-def get_canvas_square_dpi(pageinfo, options):
+def get_canvas_square_dpi(pageinfo, options) -> Resolution:
     """Get the DPI when we require xres == yres, in Postscript units"""
-    return float(
+    units = float(
         max(
-            (pageinfo.xres) or VECTOR_PAGE_DPI,
-            (pageinfo.yres) or VECTOR_PAGE_DPI,
-            VECTOR_PAGE_DPI if pageinfo.has_vector else 0,
-            options.oversample or 0,
+            (pageinfo.dpi.x) or VECTOR_PAGE_DPI,
+            (pageinfo.dpi.y) or VECTOR_PAGE_DPI,
+            VECTOR_PAGE_DPI if pageinfo.has_vector else 0.0,
+            options.oversample or 0.0,
         )
     )
+    return Resolution(units, units)
 
 
-def is_ocr_required(page_context):
+def is_ocr_required(page_context: PageContext):
     pageinfo = page_context.pageinfo
     options = page_context.options
-    log = page_context.log
 
     ocr_required = True
 
@@ -249,14 +270,15 @@ def is_ocr_required(page_context):
     elif pageinfo.has_text:
         if not options.force_ocr and not (options.skip_text or options.redo_ocr):
             raise PriorOcrFoundError(
-                "page already has text! - aborting (use --force-ocr to force OCR)"
+                "page already has text! - aborting (use --force-ocr to force OCR; "
+                " see also help for the arguments --skip-text and --redo-ocr"
             )
         elif options.force_ocr:
             log.info("page already has text! - rasterizing text and running OCR anyway")
             ocr_required = True
         elif options.redo_ocr:
             if pageinfo.has_corrupt_text:
-                log.warn(
+                log.warning(
                     "some text on this page cannot be mapped to characters: "
                     "consider using --force-ocr instead"
                 )
@@ -283,7 +305,7 @@ def is_ocr_required(page_context):
             )
         elif options.force_ocr:
             # Warn the user they might not want to do this
-            log.warn(
+            log.warning(
                 "page has no images - "
                 "all vector content will be "
                 f"rasterized at {VECTOR_PAGE_DPI} DPI, losing some resolution and likely "
@@ -303,99 +325,94 @@ def is_ocr_required(page_context):
         pixel_count = pageinfo.width_pixels * pageinfo.height_pixels
         if pixel_count > (options.skip_big * 1_000_000):
             ocr_required = False
-            log.warn(
+            log.warning(
                 "page too big, skipping OCR "
                 f"({(pixel_count / 1_000_000):.1f} MPixels > {options.skip_big:.1f} MPixels --skip-big)"
             )
     return ocr_required
 
 
-def rasterize_preview(input_file, page_context):
+def rasterize_preview(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('rasterize_preview.jpg')
     canvas_dpi = get_canvas_square_dpi(page_context.pageinfo, page_context.options)
     page_dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    ghostscript.rasterize_pdf(
-        input_file,
-        output_file,
-        xres=canvas_dpi,
-        yres=canvas_dpi,
+    page_context.plugin_manager.hook.rasterize_pdf_page(
+        input_file=input_file,
+        output_file=output_file,
         raster_device='jpeggray',
-        log=page_context.log,
-        page_dpi=(page_dpi, page_dpi),
+        raster_dpi=canvas_dpi,
+        page_dpi=page_dpi,
         pageno=page_context.pageinfo.pageno + 1,
     )
     return output_file
 
 
-def get_orientation_correction(preview, page_context):
+def describe_rotation(page_context: PageContext, orient_conf, correction: int):
     """
-    Work out orientation correct for each page.
+    Describe the page rotation we are going to perform.
+    """
+    direction = {0: '⇧', 90: '⇨', 180: '⇩', 270: '⇦'}
+    turns = {0: ' ', 90: '⬏', 180: '↻', 270: '⬑'}
+
+    existing_rotation = page_context.pageinfo.rotation
+    action = ''
+    if orient_conf.confidence >= page_context.options.rotate_pages_threshold:
+        if correction != 0:
+            action = 'will rotate ' + turns[correction]
+        else:
+            action = 'rotation appears correct'
+    else:
+        if correction != 0:
+            action = 'confidence too low to rotate'
+        else:
+            action = 'no change'
+
+    facing = ''
+
+    if existing_rotation != 0:
+        facing = f"with existing rotation {direction.get(existing_rotation, '?')}, "
+    facing += f"page is facing {direction.get(orient_conf.angle, '?')}"
+
+    return f"{facing}, confidence {orient_conf.confidence:.2f} - {action}"
+
+
+def get_orientation_correction(preview: Path, page_context: PageContext):
+    """Work out orientation correct for each page.
 
     We ask Ghostscript to draw a preview page, which will rasterize with the
-    current /Rotate applied, and then ask Tesseract which way the page is
+    current /Rotate applied, and then ask OCR which way the page is
     oriented. If the value of /Rotate is correct (e.g., a user already
-    manually fixed rotation), then Tesseract will say the page is pointing
+    manually fixed rotation), then OCR will say the page is pointing
     up and the correction is zero. Otherwise, the orientation found by
-    Tesseract represents the clockwise rotation, or the counterclockwise
+    OCR represents the clockwise rotation, or the counterclockwise
     correction to rotation.
 
     When we draw the real page for OCR, we rotate it by the CCW correction,
     which points it (hopefully) upright. _graft.py takes care of the orienting
     the image and text layers.
-
     """
 
-    orient_conf = tesseract.get_orientation(
-        preview,
-        engine_mode=page_context.options.tesseract_oem,
-        timeout=page_context.options.tesseract_timeout,
-        log=page_context.log,
-        tesseract_env=page_context.options.tesseract_env,
+    orient_conf = page_context.plugin_manager.hook.get_ocr_engine().get_orientation(
+        preview, page_context.options
     )
 
-    direction = {0: '⇧', 90: '⇨', 180: '⇩', 270: '⇦'}
-
-    existing_rotation = page_context.pageinfo.rotation
-
     correction = orient_conf.angle % 360
-
-    apply_correction = False
-    action = ''
-    if orient_conf.confidence >= page_context.options.rotate_pages_threshold:
-        if correction != 0:
-            apply_correction = True
-            action = ' - will rotate'
-        else:
-            action = ' - rotation appears correct'
-    else:
-        if correction != 0:
-            action = ' - confidence too low to rotate'
-        else:
-            action = ' - no change'
-
-    facing = ''
-    if existing_rotation != 0:
-        facing = 'with existing rotation {}, '.format(
-            direction.get(existing_rotation, '?')
-        )
-    facing += 'page is facing {}'.format(direction.get(orient_conf.angle, '?'))
-
-    page_context.log.debug(
-        '{pagenum:4d}: {facing}, confidence {conf:.2f}{action}'.format(
-            pagenum=page_context.pageinfo.pageno,
-            facing=facing,
-            conf=orient_conf.confidence,
-            action=action,
-        )
-    )
-
-    if apply_correction:
+    log.info(describe_rotation(page_context, orient_conf, correction))
+    if (
+        orient_conf.confidence >= page_context.options.rotate_pages_threshold
+        and correction != 0
+    ):
         return correction
+
     return 0
 
 
 def rasterize(
-    input_file, page_context, correction=0, output_tag='', remove_vectors=None
+    input_file: Path,
+    page_context: PageContext,
+    correction: int = 0,
+    output_tag: str = '',
+    remove_vectors=None,
 ):
     colorspaces = ['pngmono', 'pnggray', 'png256', 'png16m']
     device_idx = 0
@@ -425,21 +442,19 @@ def rasterize(
 
     device = colorspaces[device_idx]
 
-    page_context.log.debug(f"Rasterize with {device}")
+    log.debug(f"Rasterize with {device}")
 
     # Produce the page image with square resolution or else deskew and OCR
     # will not work properly.
     canvas_dpi = get_canvas_square_dpi(pageinfo, page_context.options)
     page_dpi = get_page_square_dpi(pageinfo, page_context.options)
 
-    ghostscript.rasterize_pdf(
-        input_file,
-        output_file,
-        xres=canvas_dpi,
-        yres=canvas_dpi,
+    page_context.plugin_manager.hook.rasterize_pdf_page(
+        input_file=input_file,
+        output_file=output_file,
         raster_device=device,
-        log=page_context.log,
-        page_dpi=(page_dpi, page_dpi),
+        raster_dpi=canvas_dpi,
+        page_dpi=page_dpi,
         pageno=pageinfo.pageno + 1,
         rotation=correction,
         filter_vector=remove_vectors,
@@ -447,39 +462,31 @@ def rasterize(
     return output_file
 
 
-def preprocess_remove_background(input_file, page_context):
+def preprocess_remove_background(input_file: Path, page_context: PageContext):
     if any(image.bpc > 1 for image in page_context.pageinfo.images):
         output_file = page_context.get_path('pp_rm_bg.png')
         leptonica.remove_background(input_file, output_file)
         return output_file
     else:
-        page_context.log.info("background removal skipped on mono page")
+        log.info("background removal skipped on mono page")
         return input_file
 
 
-def preprocess_deskew(input_file, page_context):
+def preprocess_deskew(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('pp_deskew.png')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    leptonica.deskew(input_file, output_file, dpi)
+    leptonica.deskew(input_file, output_file, dpi.x)
     return output_file
 
 
-def preprocess_clean(input_file, page_context):
-    from .exec import unpaper
-
+def preprocess_clean(input_file: Path, page_context: PageContext):
     output_file = page_context.get_path('pp_clean.png')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    unpaper.clean(
-        input_file,
-        output_file,
-        dpi,
-        page_context.log,
-        page_context.options.unpaper_args,
-    )
+    unpaper.clean(input_file, output_file, dpi.x, page_context.options.unpaper_args)
     return output_file
 
 
-def create_ocr_image(image, page_context):
+def create_ocr_image(image: Path, page_context: PageContext):
     """Create the image we send for OCR. May not be the same as the display
     image depending on preprocessing. This image will never be shown to the
     user."""
@@ -487,15 +494,11 @@ def create_ocr_image(image, page_context):
     output_file = page_context.get_path('ocr.png')
     options = page_context.options
     with Image.open(image) as im:
-        from PIL import ImageColor
-        from PIL import ImageDraw
-
         white = ImageColor.getcolor('#ffffff', im.mode)
         # pink = ImageColor.getcolor('#ff0080', im.mode)
         draw = ImageDraw.ImageDraw(im)
 
-        xres, yres = im.info['dpi']
-        page_context.log.debug('resolution %r %r' % (xres, yres))
+        log.debug('resolution %r', im.info['dpi'])
 
         if not options.force_ocr:
             # Do not mask text areas when forcing OCR, because we need to OCR
@@ -511,15 +514,15 @@ def create_ocr_image(image, page_context):
                 # without regard whatever resolution is in pageinfo (may differ or
                 # be None)
                 bbox = [float(v) for v in textarea]
-                xscale, yscale = float(xres) / 72.0, float(yres) / 72.0
+                xyscale = tuple(float(coord) / 72.0 for coord in im.info['dpi'])
                 pixcoords = [
-                    bbox[0] * xscale,
-                    im.height - bbox[3] * yscale,
-                    bbox[2] * xscale,
-                    im.height - bbox[1] * yscale,
+                    bbox[0] * xyscale[0],
+                    im.height - bbox[3] * xyscale[1],
+                    bbox[2] * xyscale[0],
+                    im.height - bbox[1] * xyscale[1],
                 ]
                 pixcoords = [int(round(c)) for c in pixcoords]
-                page_context.log.debug('blanking %r', pixcoords)
+                log.debug('blanking %r', pixcoords)
                 draw.rectangle(pixcoords, fill=white)
                 # draw.rectangle(pixcoords, outline=pink)
 
@@ -529,28 +532,30 @@ def create_ocr_image(image, page_context):
             im = pix.topil()
 
         del draw
+
+        filter_im = page_context.plugin_manager.hook.filter_ocr_image(
+            page=page_context, image=im
+        )
+        if filter_im is not None:
+            im = filter_im
+
         # Pillow requires integer DPI
-        dpi = round(xres), round(yres)
+        dpi = tuple(round(coord) for coord in im.info['dpi'])
         im.save(output_file, dpi=dpi)
     return output_file
 
 
-def ocr_tesseract_hocr(input_file, page_context):
+def ocr_engine_hocr(input_file: Path, page_context: PageContext):
     hocr_out = page_context.get_path('ocr_hocr.hocr')
     hocr_text_out = page_context.get_path('ocr_hocr.txt')
     options = page_context.options
-    tesseract.generate_hocr(
+
+    ocr_engine = page_context.plugin_manager.hook.get_ocr_engine()
+    ocr_engine.generate_hocr(
         input_file=input_file,
-        output_files=[hocr_out, hocr_text_out],
-        language=options.language,
-        engine_mode=options.tesseract_oem,
-        tessconfig=options.tesseract_config,
-        timeout=options.tesseract_timeout,
-        pagesegmode=options.tesseract_pagesegmode,
-        user_words=options.user_words,
-        user_patterns=options.user_patterns,
-        tesseract_env=options.tesseract_env,
-        log=page_context.log,
+        output_hocr=hocr_out,
+        output_text=hocr_text_out,
+        options=options,
     )
     return (hocr_out, hocr_text_out)
 
@@ -560,24 +565,26 @@ def should_visible_page_image_use_jpg(pageinfo):
     return pageinfo.images and all(im.enc == Encoding.jpeg for im in pageinfo.images)
 
 
-def create_visible_page_jpg(image, page_context):
+def create_visible_page_jpg(image: Path, page_context: PageContext) -> Path:
     output_file = page_context.get_path('visible.jpg')
     with Image.open(image) as im:
         # At this point the image should be a .png, but deskew, unpaper
         # might have removed the DPI information. In this case, fall back to
         # square DPI used to rasterize. When the preview image was
         # rasterized, it was also converted to square resolution, which is
-        # what we want to give tesseract, so keep it square.
-        fallback_dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-        dpi = im.info.get('dpi', (fallback_dpi, fallback_dpi))
+        # what we want to give to the OCR engine, so keep it square.
+        if 'dpi' in im.info:
+            dpi = Resolution(*im.info['dpi'])
+        else:
+            # Fallback to page-implied DPI
+            dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
 
         # Pillow requires integer DPI
-        dpi = round(dpi[0]), round(dpi[1])
-        im.save(output_file, format='JPEG', dpi=dpi)
+        im.save(output_file, format='JPEG', dpi=dpi.to_int())
     return output_file
 
 
-def create_pdf_page_from_image(image, page_context):
+def create_pdf_page_from_image(image: Path, page_context: PageContext):
     # We rasterize a square DPI version of each page because most image
     # processing tools don't support rectangular DPI. Use the square DPI as it
     # accurately describes the image. It would be possible to resample the image
@@ -586,56 +593,50 @@ def create_pdf_page_from_image(image, page_context):
     # sandwich renderer would be fine.
     output_file = page_context.get_path('visible.pdf')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    layout_fun = img2pdf.get_fixed_dpi_layout_fun((dpi, dpi))
+    layout_fun = img2pdf.get_fixed_dpi_layout_fun(dpi)
 
     # This create a single page PDF
     with open(image, 'rb') as imfile, open(output_file, 'wb') as pdf:
-        page_context.log.debug('convert')
+        log.debug('convert')
         img2pdf.convert(
             imfile, with_pdfrw=False, layout_fun=layout_fun, outputstream=pdf
         )
-        page_context.log.debug('convert done')
+        log.debug('convert done')
     return output_file
 
 
-def render_hocr_page(hocr, page_context):
+def render_hocr_page(hocr: Path, page_context: PageContext):
     output_file = page_context.get_path('ocr_hocr.pdf')
     dpi = get_page_square_dpi(page_context.pageinfo, page_context.options)
-    hocrtransform = HocrTransform(hocr, dpi)
+    hocrtransform = HocrTransform(hocr, dpi.x)  # square
     hocrtransform.to_pdf(
         output_file,
-        imageFileName=None,
-        showBoundingboxes=False,
-        invisibleText=True,
-        interwordSpaces=True,
+        image_filename=None,
+        show_bounding_boxes=False,
+        invisible_text=True,
+        interword_spaces=True,
     )
     return output_file
 
 
-def ocr_tesseract_textonly_pdf(input_image, page_context):
+def ocr_engine_textonly_pdf(input_image: Path, page_context: PageContext):
     output_pdf = page_context.get_path('ocr_tess.pdf')
     output_text = page_context.get_path('ocr_tess.txt')
     options = page_context.options
-    tesseract.generate_pdf(
-        input_image=input_image,
-        skip_pdf=None,
+
+    ocr_engine = page_context.plugin_manager.hook.get_ocr_engine()
+    ocr_engine.generate_pdf(
+        input_file=input_image,
         output_pdf=output_pdf,
         output_text=output_text,
-        language=options.language,
-        engine_mode=options.tesseract_oem,
-        text_only=True,
-        tessconfig=options.tesseract_config,
-        timeout=options.tesseract_timeout,
-        pagesegmode=options.tesseract_pagesegmode,
-        user_words=options.user_words,
-        user_patterns=options.user_patterns,
-        tesseract_env=options.tesseract_env,
-        log=page_context.log,
+        options=options,
     )
     return (output_pdf, output_text)
 
 
-def get_docinfo(base_pdf, options):
+def get_docinfo(base_pdf: pikepdf.Pdf, context: PdfContext) -> Dict[str, str]:
+    options = context.options
+
     def from_document_info(key):
         try:
             s = base_pdf.docinfo[key]
@@ -647,7 +648,6 @@ def get_docinfo(base_pdf, options):
         k: from_document_info(k)
         for k in ('/Title', '/Author', '/Keywords', '/Subject', '/CreationDate')
     }
-    renderer_tag = 'OCR'
     if options is not None:
         if options.title:
             pdfmark['/Title'] = options.title
@@ -658,12 +658,9 @@ def get_docinfo(base_pdf, options):
         if options.subject:
             pdfmark['/Subject'] = options.subject
 
-        if options.pdf_renderer == 'sandwich':
-            renderer_tag = 'OCR-PDF'
+    creator_tag = context.plugin_manager.hook.get_ocr_engine().creator_tag(options)
 
-    pdfmark['/Creator'] = (
-        f'{PROGRAM_NAME} {VERSION} / ' f'Tesseract {renderer_tag} {tesseract.version()}'
-    )
+    pdfmark['/Creator'] = f'{PROGRAM_NAME} {VERSION} / {creator_tag}'
     pdfmark['/Producer'] = f'pikepdf {pikepdf.__version__}'
     if 'OCRMYPDF_CREATOR' in os.environ:
         pdfmark['/Creator'] = os.environ['OCRMYPDF_CREATOR']
@@ -674,13 +671,13 @@ def get_docinfo(base_pdf, options):
     return pdfmark
 
 
-def generate_postscript_stub(context):
+def generate_postscript_stub(context: PdfContext):
     output_file = context.get_path('pdfa.ps')
     generate_pdfa_ps(output_file)
     return output_file
 
 
-def convert_to_pdfa(input_pdf, input_ps_stub, context):
+def convert_to_pdfa(input_pdf: Path, input_ps_stub: Path, context: PdfContext):
     options = context.options
     input_pdfinfo = context.pdfinfo
     fix_docinfo_file = context.get_path('fix_docinfo.pdf')
@@ -693,83 +690,92 @@ def convert_to_pdfa(input_pdf, input_ps_stub, context):
     # stamping them out as soon as possible.
     modified = False
     with pikepdf.open(input_pdf) as pdf_file:
-        if pdf_file.docinfo:
-            for k, v in pdf_file.docinfo.items():
-                if b'\x00' in bytes(v):
-                    pdf_file.docinfo[k] = bytes(v).replace(b'\x00', b'')
-                    modified = True
+        try:
+            len(pdf_file.docinfo)
+        except TypeError:
+            log.error(
+                "File contains a malformed DocumentInfo block - continuing anyway"
+            )
+        else:
+            if pdf_file.docinfo:
+                for k, v in pdf_file.docinfo.items():
+                    if b'\x00' in bytes(v):
+                        pdf_file.docinfo[k] = bytes(v).replace(b'\x00', b'')
+                        modified = True
         if modified:
             pdf_file.save(fix_docinfo_file)
         else:
             safe_symlink(input_pdf, fix_docinfo_file)
 
-    ghostscript.generate_pdfa(
+    context.plugin_manager.hook.generate_pdfa(
         pdf_version=input_pdfinfo.min_version,
-        pdf_pages=[fix_docinfo_file, input_ps_stub],
+        pdf_pages=[fix_docinfo_file],
+        pdfmark=input_ps_stub,
         output_file=output_file,
         compression=options.pdfa_image_compression,
-        log=context.log,
         pdfa_part=options.output_type[-1],  # is pdfa-1, pdfa-2, or pdfa-3
     )
 
     return output_file
 
 
-def should_linearize(working_file, context):
+def should_linearize(working_file: Path, context: PdfContext):
     filesize = os.stat(working_file).st_size
     if filesize > (context.options.fast_web_view * 1_000_000):
         return True
     return False
 
 
-def metadata_fixup(working_file, context):
+def metadata_fixup(working_file: Path, context: PdfContext):
     output_file = context.get_path('metafix.pdf')
     options = context.options
-    original = pikepdf.open(context.origin)
-    docinfo = get_docinfo(original, options)
-    pdf = pikepdf.open(working_file)
-    with pdf.open_metadata() as meta:
-        meta.load_from_docinfo(docinfo, delete_missing=False)
-        # If xmp:CreateDate is missing, set it to the modify date to
-        # match Ghostscript, for consistency
-        if 'xmp:CreateDate' not in meta:
-            meta['xmp:CreateDate'] = meta.get('xmp:ModifyDate', '')
 
-        meta_original = original.open_metadata()
-        not_copied = set(meta_original.keys()) - set(meta.keys())
-        if not_copied:
-            if options.output_type.startswith('pdfa'):
-                context.log.warning(
-                    "Some input metadata could not be copied because it is not "
-                    "permitted in PDF/A. You may wish to examine the output "
-                    "PDF's XMP metadata."
-                )
-                context.log.debug(
-                    "The following metadata fields were not copied: %r", not_copied
-                )
-            else:
-                context.log.error(
-                    "Some input metadata could not be copied."
-                    "You may wish to examine the output PDF's XMP metadata."
-                )
-                context.log.info(
-                    "The following metadata fields were not copied: %r", not_copied
-                )
-    pdf.save(
-        output_file,
-        compress_streams=True,
-        preserve_pdfa=True,
-        object_stream_mode=pikepdf.ObjectStreamMode.generate,
-        linearize=(  # Don't linearize if optimize() will be linearizing too
-            should_linearize(working_file, context) if options.optimize == 0 else False
-        ),
-    )
-    original.close()
-    pdf.close()
+    def report_on_metadata(missing):
+        if not missing:
+            return
+        if options.output_type.startswith('pdfa'):
+            log.warning(
+                "Some input metadata could not be copied because it is not "
+                "permitted in PDF/A. You may wish to examine the output "
+                "PDF's XMP metadata."
+            )
+            log.debug("The following metadata fields were not copied: %r", missing)
+        else:
+            log.error(
+                "Some input metadata could not be copied."
+                "You may wish to examine the output PDF's XMP metadata."
+            )
+            log.info("The following metadata fields were not copied: %r", missing)
+
+    with pikepdf.open(context.origin) as original, pikepdf.open(working_file) as pdf:
+        docinfo = get_docinfo(original, context)
+        with pdf.open_metadata() as meta:
+            meta.load_from_docinfo(docinfo, delete_missing=False, raise_failure=False)
+            # If xmp:CreateDate is missing, set it to the modify date to
+            # match Ghostscript, for consistency
+            if 'xmp:CreateDate' not in meta:
+                meta['xmp:CreateDate'] = meta.get('xmp:ModifyDate', '')
+
+            meta_original = original.open_metadata()
+            missing = set(meta_original.keys()) - set(meta.keys())
+            report_on_metadata(missing)
+
+        pdf.save(
+            output_file,
+            compress_streams=True,
+            preserve_pdfa=True,
+            object_stream_mode=pikepdf.ObjectStreamMode.generate,
+            linearize=(  # Don't linearize if optimize() will be linearizing too
+                should_linearize(working_file, context)
+                if options.optimize == 0
+                else False
+            ),
+        )
+
     return output_file
 
 
-def optimize_pdf(input_file, context):
+def optimize_pdf(input_file: Path, context: PdfContext):
     output_file = context.get_path('optimize.pdf')
     save_settings = dict(
         compress_streams=True,
@@ -781,7 +787,7 @@ def optimize_pdf(input_file, context):
     return output_file
 
 
-def merge_sidecars(txt_files, context):
+def merge_sidecars(txt_files: Iterable[Optional[Path]], context: PdfContext):
     output_file = context.get_path('sidecar.txt')
     with open(output_file, 'w', encoding="utf-8") as stream:
         for page_num, txt_file in enumerate(txt_files):
@@ -790,11 +796,9 @@ def merge_sidecars(txt_files, context):
             if txt_file:
                 with open(txt_file, 'r', encoding="utf-8") as in_:
                     txt = in_.read()
-                    # Tesseract v4 alpha started adding form feeds in
-                    # commit aa6eb6b
-                    # No obvious way to detect what binaries will do this, so
-                    # for consistency just ignore its form feeds and insert our
-                    # own
+                    # Some OCR engines (e.g. Tesseract v4 alpha) add form feeds
+                    # between pages, and some do not. For consistency, we ignore
+                    # any added by the OCR engine and them on our own.
                     if txt.endswith('\f'):
                         stream.write(txt[:-1])
                     else:
@@ -804,8 +808,8 @@ def merge_sidecars(txt_files, context):
     return output_file
 
 
-def copy_final(input_file, output_file, context):
-    context.log.debug('%s -> %s', input_file, output_file)
+def copy_final(input_file: Path, output_file: Path, _context: PdfContext):
+    log.debug('%s -> %s', input_file, output_file)
     with open(input_file, 'rb') as input_stream:
         if output_file == '-':
             copyfileobj(input_stream, sys.stdout.buffer)

src/ocrmypdf/_plugin_manager.py

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

src/ocrmypdf/_sync.py

@@ -17,20 +17,20 @@
 
 import logging
 import logging.handlers
-import multiprocessing
 import os
-import signal
 import sys
 import threading
-from collections import namedtuple
+from functools import partial
+from pathlib import Path
 from tempfile import mkdtemp
+from typing import List, NamedTuple, Optional, Tuple
 
 import PIL
-from tqdm import tqdm
 
-from ._graft import OcrGrafter
-from ._jobcontext import PDFContext, cleanup_working_files, make_logger
-from ._pipeline import (
+from ocrmypdf._concurrent import exec_progress_pool
+from ocrmypdf._graft import OcrGrafter
+from ocrmypdf._jobcontext import PageContext, PdfContext, cleanup_working_files
+from ocrmypdf._pipeline import (
     convert_to_pdfa,
     copy_final,
     create_ocr_image,
@@ -42,8 +42,8 @@ from ._pipeline import (
     is_ocr_required,
     merge_sidecars,
     metadata_fixup,
-    ocr_tesseract_hocr,
-    ocr_tesseract_textonly_pdf,
+    ocr_engine_hocr,
+    ocr_engine_textonly_pdf,
     optimize_pdf,
     preprocess_clean,
     preprocess_deskew,
@@ -55,22 +55,51 @@ from ._pipeline import (
     triage,
     validate_pdfinfo_options,
 )
-from ._validation import (
+from ocrmypdf._plugin_manager import get_plugin_manager
+from ocrmypdf._validation import (
     check_requested_output_file,
     create_input_file,
     report_output_file_size,
 )
-from .exceptions import ExitCode, ExitCodeException
-from .exec import qpdf
-from .helpers import available_cpu_count
-from .pdfa import file_claims_pdfa
+from ocrmypdf.exceptions import ExitCode, ExitCodeException
+from ocrmypdf.helpers import available_cpu_count, check_pdf, samefile
+from ocrmypdf.pdfa import file_claims_pdfa
 
-PageResult = namedtuple(
-    'PageResult', 'pageno, pdf_page_from_image, ocr, text, orientation_correction'
-)
+log = logging.getLogger(__name__)
 
 
-def preprocess(page_context, image, remove_background, deskew, clean):
+class PageResult(NamedTuple):
+    pageno: int
+    pdf_page_from_image: Optional[Path]
+    ocr: Optional[Path]
+    text: Optional[Path]
+    orientation_correction: int
+
+
+tls = threading.local()
+tls.pageno = None
+
+
+old_factory = logging.getLogRecordFactory()
+
+
+def record_factory(*args, **kwargs):
+    record = old_factory(*args, **kwargs)
+    if hasattr(tls, 'pageno'):
+        record.pageno = tls.pageno
+    return record
+
+
+logging.setLogRecordFactory(record_factory)
+
+
+def preprocess(
+    page_context: PageContext,
+    image: Path,
+    remove_background: bool,
+    deskew: bool,
+    clean: bool,
+) -> Path:
     if remove_background:
         image = preprocess_remove_background(image, page_context)
     if deskew:
@@ -80,54 +109,55 @@ def preprocess(page_context, image, remove_background, deskew, clean):
     return image
 
 
-def exec_page_sync(page_context):
+def make_intermediate_images(
+    page_context: PageContext, orientation_correction: int
+) -> Tuple[Path, Optional[Path]]:
     options = page_context.options
-    orientation_correction = 0
-    pdf_page_from_image_out = None
-    ocr_out = None
-    text_out = None
-    if is_ocr_required(page_context):
-        if options.rotate_pages:
-            # Rasterize
-            rasterize_preview_out = rasterize_preview(page_context.origin, page_context)
-            orientation_correction = get_orientation_correction(
-                rasterize_preview_out, page_context
-            )
 
-        rasterize_out = rasterize(
-            page_context.origin,
+    ocr_image = preprocess_out = None
+    rasterize_out = rasterize(
+        page_context.origin,
+        page_context,
+        correction=orientation_correction,
+        remove_vectors=False,
+    )
+
+    if not any([options.clean, options.clean_final, options.remove_vectors]):
+        ocr_image = preprocess_out = preprocess(
             page_context,
-            correction=orientation_correction,
-            remove_vectors=False,
+            rasterize_out,
+            options.remove_background,
+            options.deskew,
+            clean=False,
         )
-
-        if not any([options.clean, options.clean_final, options.remove_vectors]):
-            ocr_image = preprocess_out = preprocess(
+    else:
+        if not options.lossless_reconstruction:
+            preprocess_out = preprocess(
                 page_context,
                 rasterize_out,
                 options.remove_background,
                 options.deskew,
-                clean=False,
+                clean=options.clean_final,
+            )
+        if options.remove_vectors:
+            rasterize_ocr_out = rasterize(
+                page_context.origin,
+                page_context,
+                correction=orientation_correction,
+                remove_vectors=True,
+                output_tag='_ocr',
             )
         else:
-            if not options.lossless_reconstruction:
-                preprocess_out = preprocess(
-                    page_context,
-                    rasterize_out,
-                    options.remove_background,
-                    options.deskew,
-                    clean=options.clean_final,
-                )
-            if options.remove_vectors:
-                rasterize_ocr_out = rasterize(
-                    page_context.origin,
-                    page_context,
-                    correction=orientation_correction,
-                    remove_vectors=True,
-                    output_tag='_ocr',
-                )
-            else:
-                rasterize_ocr_out = rasterize_out
+            rasterize_ocr_out = rasterize_out
+
+        if (
+            preprocess_out
+            and rasterize_ocr_out == rasterize_out
+            and options.clean == options.clean_final
+        ):
+            # Optimization: image for OCR is identical to presentation image
+            ocr_image = preprocess_out
+        else:
             ocr_image = preprocess(
                 page_context,
                 rasterize_ocr_out,
@@ -135,28 +165,56 @@ def exec_page_sync(page_context):
                 options.deskew,
                 clean=options.clean,
             )
+    return ocr_image, preprocess_out
 
-        ocr_image_out = create_ocr_image(ocr_image, page_context)
 
-        pdf_page_from_image_out = None
-        if not options.lossless_reconstruction:
-            visible_image_out = preprocess_out
-            if should_visible_page_image_use_jpg(page_context.pageinfo):
-                visible_image_out = create_visible_page_jpg(
-                    visible_image_out, page_context
-                )
-            pdf_page_from_image_out = create_pdf_page_from_image(
-                visible_image_out, page_context
-            )
+def exec_page_sync(page_context: PageContext):
+    options = page_context.options
+    tls.pageno = page_context.pageno + 1
 
-        if options.pdf_renderer == 'hocr':
-            (hocr_out, text_out) = ocr_tesseract_hocr(ocr_image_out, page_context)
-            ocr_out = render_hocr_page(hocr_out, page_context)
+    if not is_ocr_required(page_context):
+        return PageResult(
+            pageno=page_context.pageno,
+            pdf_page_from_image=None,
+            ocr=None,
+            text=None,
+            orientation_correction=0,
+        )
 
-        if options.pdf_renderer == 'sandwich':
-            (ocr_out, text_out) = ocr_tesseract_textonly_pdf(
-                ocr_image_out, page_context
-            )
+    orientation_correction = 0
+    if options.rotate_pages:
+        # Rasterize
+        rasterize_preview_out = rasterize_preview(page_context.origin, page_context)
+        orientation_correction = get_orientation_correction(
+            rasterize_preview_out, page_context
+        )
+
+    ocr_image, preprocess_out = make_intermediate_images(
+        page_context, orientation_correction
+    )
+    ocr_image_out = create_ocr_image(ocr_image, page_context)
+
+    pdf_page_from_image_out = None
+    if not options.lossless_reconstruction:
+        assert preprocess_out
+        visible_image_out = preprocess_out
+        if should_visible_page_image_use_jpg(page_context.pageinfo):
+            visible_image_out = create_visible_page_jpg(visible_image_out, page_context)
+        filtered_image = page_context.plugin_manager.hook.filter_page_image(
+            page=page_context, image_filename=visible_image_out
+        )
+        if filtered_image:
+            visible_image_out = filtered_image
+        pdf_page_from_image_out = create_pdf_page_from_image(
+            visible_image_out, page_context
+        )
+
+    if options.pdf_renderer == 'hocr':
+        (hocr_out, text_out) = ocr_engine_hocr(ocr_image_out, page_context)
+        ocr_out = render_hocr_page(hocr_out, page_context)
+
+    if options.pdf_renderer == 'sandwich':
+        (ocr_out, text_out) = ocr_engine_textonly_pdf(ocr_image_out, page_context)
 
     return PageResult(
         pageno=page_context.pageno,
@@ -167,7 +225,7 @@ def exec_page_sync(page_context):
     )
 
 
-def post_process(pdf_file, context):
+def post_process(pdf_file, context: PdfContext):
     pdf_out = pdf_file
     if context.options.output_type.startswith('pdfa'):
         ps_stub_out = generate_postscript_stub(context)
@@ -177,119 +235,50 @@ def post_process(pdf_file, context):
     return optimize_pdf(pdf_out, context)
 
 
-def worker_init(queue, max_pixels):
-    """Initialize a process pool worker"""
-
-    # Ignore SIGINT (our parent process will kill us gracefully)
-    signal.signal(signal.SIGINT, signal.SIG_IGN)
-
-    # Reconfigure the root logger for this process to send all messages to a queue
-    h = logging.handlers.QueueHandler(queue)
-    root = logging.getLogger()
-    root.handlers = []
-    root.addHandler(h)
-
+def worker_init(max_pixels: int):
     # In Windows, child process will not inherit our change to this value in
-    # the parent process, so ensure workers get it set
+    # the parent process, so ensure workers get it set. Not needed when running
+    # threaded, but harmless to set again.
     PIL.Image.MAX_IMAGE_PIXELS = max_pixels
 
 
-def worker_thread_init(_queue, max_pixels):
-    # This is probably not needed since threads should all see the same memory,
-    # but done for consistency.
-    PIL.Image.MAX_IMAGE_PIXELS = max_pixels
-
-
-def log_listener(queue):
-    """Listen to the worker processes and forward the messages to logging
-
-    For simplicity this is a thread rather than a process. Only one process
-    should actually write to sys.stderr or whatever we're using, so if this is
-    made into a process the main application needs to be directed to it.
-
-    See https://docs.python.org/3/howto/logging-cookbook.html#logging-to-a-single-file-from-multiple-processes
-    """
-
-    while True:
-        try:
-            record = queue.get()
-            if record is None:
-                break
-            logger = logging.getLogger(record.name)
-            logger.handle(record)
-        except Exception:
-            import traceback
-
-            print("Logging problem", file=sys.stderr)
-            traceback.print_exc(file=sys.stderr)
-
-
-def exec_concurrent(context):
+def exec_concurrent(context: PdfContext):
     """Execute the pipeline concurrently"""
 
     # Run exec_page_sync on every page context
     max_workers = min(len(context.pdfinfo), context.options.jobs)
     if max_workers > 1:
-        context.log.info("Start processing %d pages concurrent", max_workers)
+        log.info("Start processing %d pages concurrently", max_workers)
 
-    # Tesseract 4.x can be multithreaded, and we also run multiple workers. We want
-    # to manage how many threads it uses to avoid creating total threads than cores.
-    # Performance testing shows we're better off
-    # parallelizing ocrmypdf and forcing Tesseract to be single threaded, which we
-    # get by setting the envvar OMP_THREAD_LIMIT to 1. But if the page count of the
-    # input file is small, then we allow Tesseract to use threads, subject to the
-    # constraint: (ocrmypdf workers) * (tesseract threads) <= max_workers.
-    # As of Tesseract 4.1, 3 threads is the most effective on a 4 core/8 thread system.
-    tess_threads = min(3, context.options.jobs // max_workers)
-    if context.options.tesseract_env is None:
-        context.options.tesseract_env = os.environ.copy()
-    context.options.tesseract_env.setdefault('OMP_THREAD_LIMIT', str(tess_threads))
-    if tess_threads > 1:
-        context.log.info("Using Tesseract OpenMP thread limit %d", tess_threads)
-
-    if context.options.use_threads:
-        from multiprocessing.dummy import Pool
-
-        initializer = worker_thread_init
-    else:
-        Pool = multiprocessing.Pool
-        initializer = worker_init
-
-    sidecars = [None] * len(context.pdfinfo)
+    sidecars: List[Optional[Path]] = [None] * len(context.pdfinfo)
     ocrgraft = OcrGrafter(context)
 
-    log_queue = multiprocessing.Queue(-1)
-    listener = threading.Thread(target=log_listener, args=(log_queue,))
-    listener.start()
-    with tqdm(
-        total=(2 * len(context.pdfinfo)),
-        desc='OCR',
-        unit='page',
-        unit_scale=0.5,
-        disable=not context.options.progress_bar,
-    ) as pbar, Pool(
-        processes=max_workers,
-        initializer=initializer,
-        initargs=(log_queue, PIL.Image.MAX_IMAGE_PIXELS),
-    ) as pool:
-        results = pool.imap_unordered(exec_page_sync, context.get_page_contexts())
-        while True:
-            try:
-                page_result = results.next()
-                sidecars[page_result.pageno] = page_result.text
-                pbar.update()
-                ocrgraft.graft_page(page_result)
-                pbar.update()
-            except StopIteration:
-                break
-            except (Exception, KeyboardInterrupt):
-                pool.terminate()
-                log_queue.put_nowait(None)  # Terminate log listener
-                # Don't try listener.join() here, will deadlock
-                raise
+    def update_page(result: PageResult, pbar):
+        sidecars[result.pageno] = result.text
+        pbar.update()
+        ocrgraft.graft_page(
+            pageno=result.pageno,
+            image=result.pdf_page_from_image,
+            textpdf=result.ocr,
+            autorotate_correction=result.orientation_correction,
+        )
+        pbar.update()
 
-    log_queue.put_nowait(None)
-    listener.join()
+    exec_progress_pool(
+        use_threads=context.options.use_threads,
+        max_workers=max_workers,
+        tqdm_kwargs=dict(
+            total=(2 * len(context.pdfinfo)),
+            desc='OCR',
+            unit='page',
+            unit_scale=0.5,
+            disable=not context.options.progress_bar,
+        ),
+        task_initializer=partial(worker_init, PIL.Image.MAX_IMAGE_PIXELS),
+        task=exec_page_sync,
+        task_arguments=context.get_page_contexts(),
+        task_finished=update_page,
+    )
 
     # Output sidecar text
     if context.options.sidecar:
@@ -313,39 +302,52 @@ class NeverRaise(Exception):
     pass  # pylint: disable=unnecessary-pass
 
 
-def samefile(f1, f2):
-    if os.name == 'nt':
-        return f1 == f2
-    else:
-        return os.path.samefile(f1, f2)
+def configure_debug_logging(log_filename, prefix=''):
+    log_file_handler = logging.FileHandler(log_filename, delay=True)
+    log_file_handler.setLevel(logging.DEBUG)
+    formatter = logging.Formatter(
+        '[%(asctime)s] - %(name)s - %(levelname)7s -%(pageno)s %(message)s'
+    )
+    log_file_handler.setFormatter(formatter)
+    logging.getLogger(prefix).addHandler(log_file_handler)
+    return log_file_handler
 
 
-def run_pipeline(options, api=False):
-    log = make_logger(options, __name__)
-
+def run_pipeline(options, *, plugin_manager, api=False):
     # Any changes to options will not take effect for options that are already
     # bound to function parameters in the pipeline. (For example
     # options.input_file, options.pdf_renderer are already bound.)
     if not options.jobs:
         options.jobs = available_cpu_count()
+    if not plugin_manager:
+        plugin_manager = get_plugin_manager(options.plugins)
+
+    work_folder = Path(mkdtemp(prefix="com.github.ocrmypdf."))
+    debug_log_handler = None
+    if (options.keep_temporary_files or options.verbose >= 1) and not os.environ.get(
+        'PYTEST_CURRENT_TEST', ''
+    ):
+        debug_log_handler = configure_debug_logging(Path(work_folder) / "debug.log")
 
-    work_folder = mkdtemp(prefix="com.github.ocrmypdf.")
     try:
         check_requested_output_file(options)
-        start_input_file = create_input_file(options, work_folder)
+        start_input_file, original_filename = create_input_file(options, work_folder)
 
         # Triage image or pdf
         origin_pdf = triage(
-            start_input_file, os.path.join(work_folder, 'origin.pdf'), options, log
+            original_filename, start_input_file, work_folder / 'origin.pdf', options
         )
 
         # Gather pdfinfo and create context
         pdfinfo = get_pdfinfo(
             origin_pdf,
-            detailed_page_analysis=options.redo_ocr,
+            detailed_analysis=options.redo_ocr,
             progbar=options.progress_bar,
+            max_workers=options.jobs if not options.use_threads else 1,  # To help debug
+            check_pages=options.pages,
         )
-        context = PDFContext(options, work_folder, origin_pdf, pdfinfo)
+
+        context = PdfContext(options, work_folder, origin_pdf, pdfinfo, plugin_manager)
 
         # Validate options are okay for this pdf
         validate_pdfinfo_options(context)
@@ -370,7 +372,7 @@ def run_pipeline(options, api=False):
                         pdfa_info['conformance'],
                     )
                     return ExitCode.pdfa_conversion_failed
-            if not qpdf.check(options.output_file, log):
+            if not check_pdf(options.output_file):
                 log.warning('Output file: The generated PDF is INVALID')
                 return ExitCode.invalid_output_pdf
             report_output_file_size(options, start_input_file, options.output_file)
@@ -387,10 +389,16 @@ def run_pipeline(options, api=False):
         else:
             log.error(type(e).__name__)
         return e.exit_code
-    except (Exception if not api else NeverRaise) as e:
+    except (Exception if not api else NeverRaise) as e:  # pylint: disable=broad-except
         log.exception("An exception occurred while executing the pipeline")
         return ExitCode.other_error
     finally:
+        if debug_log_handler:
+            try:
+                debug_log_handler.close()
+                log.removeHandler(debug_log_handler)
+            except EnvironmentError as e:
+                print(e, file=sys.stderr)
         cleanup_working_files(work_folder, options)
 
     return ExitCode.ok

src/ocrmypdf/_validation.py

@@ -21,28 +21,29 @@ import locale
 import logging
 import os
 import sys
+import unicodedata
 from pathlib import Path
 from shutil import copyfileobj
+from typing import Tuple
 
+import pikepdf
 import PIL
 
-from ._unicodefun import verify_python3_env
-from .exceptions import (
+from ocrmypdf._exec import jbig2enc, pngquant, unpaper
+from ocrmypdf._unicodefun import verify_python3_env
+from ocrmypdf.exceptions import (
     BadArgsError,
     InputFileError,
     MissingDependencyError,
     OutputFileAccessError,
 )
-from .exec import (
-    check_external_program,
-    ghostscript,
-    jbig2enc,
-    pngquant,
-    qpdf,
-    tesseract,
-    unpaper,
+from ocrmypdf.helpers import (
+    is_file_writable,
+    is_iterable_notstr,
+    monotonic,
+    safe_symlink,
 )
-from .helpers import is_file_writable, is_iterable_notstr, monotonic, safe_symlink
+from ocrmypdf.subprocess import check_external_program
 
 # -------------
 # External dependencies
@@ -67,35 +68,26 @@ def check_platform():
         )
 
 
-def check_options_languages(options):
-    if not options.language:
-        options.language = [DEFAULT_LANGUAGE]
+def check_options_languages(options, ocr_engine_languages):
+    if not options.languages:
+        options.languages = {DEFAULT_LANGUAGE}
         system_lang = locale.getlocale()[0]
         if system_lang and not system_lang.startswith('en'):
             log.debug("No language specified; assuming --language %s", DEFAULT_LANGUAGE)
-
-    # Support v2.x "eng+deu" language syntax
-    if '+' in options.language[0]:
-        options.language = options.language[0].split('+')
-
-    languages = set(options.language)
-    if not languages.issubset(tesseract.languages()):
+    if not ocr_engine_languages:
+        return
+    if not options.languages.issubset(ocr_engine_languages):
         msg = (
-            "The installed version of tesseract does not have language "
-            "data for the following requested languages: \n"
+            f"OCR engine does not have language data for the following "
+            "requested languages: \n"
         )
-        for lang in languages - tesseract.languages():
+        for lang in options.languages - ocr_engine_languages:
             msg += lang + '\n'
         raise MissingDependencyError(msg)
 
 
 def check_options_output(options):
-    # We have these constraints to check for.
-    # 1. Ghostscript < 9.20 mangles multibyte Unicode
-    # 2. hocr doesn't work on non-Latin languages (so don't select it)
-
-    languages = set(options.language)
-    is_latin = languages.issubset(HOCR_OK_LANGS)
+    is_latin = options.languages.issubset(HOCR_OK_LANGS)
 
     if options.pdf_renderer == 'hocr' and not is_latin:
         msg = (
@@ -105,37 +97,6 @@ def check_options_output(options):
         )
         log.warning(msg)
 
-    if ghostscript.version() < '9.20' and options.output_type != 'pdf' and not is_latin:
-        # https://bugs.ghostscript.com/show_bug.cgi?id=696874
-        # Ghostscript < 9.20 fails to encode multibyte characters properly
-        msg = (
-            "The installed version of Ghostscript does not work correctly "
-            "with the OCR languages you specified. Use --output-type pdf or "
-            "upgrade to Ghostscript 9.20 or later to avoid this issue."
-        )
-        msg += f"Found Ghostscript {ghostscript.version()}"
-        log.warning(msg)
-
-    # Decide on what renderer to use
-    if options.pdf_renderer == 'auto':
-        options.pdf_renderer = 'sandwich'
-
-    if options.pdf_renderer == 'sandwich' and not tesseract.has_textonly_pdf(
-        options.tesseract_env, languages
-    ):
-        raise MissingDependencyError(
-            "You are using an alpha version of Tesseract 4.0 that does not support "
-            "the textonly_pdf parameter. We don't support versions this old."
-        )
-
-    if options.output_type == 'pdfa':
-        options.output_type = 'pdfa-2'
-
-    if options.output_type == 'pdfa-3' and ghostscript.version() < '9.19':
-        raise MissingDependencyError(
-            "--output-type pdfa-3 requires Ghostscript 9.19 or later"
-        )
-
     lossless_reconstruction = False
     if not any(
         (
@@ -270,18 +231,9 @@ def check_options_advanced(options):
             "--pdfa-image-compression argument has no effect when "
             "--output-type is not 'pdfa', 'pdfa-1', or 'pdfa-2'"
         )
-    if not tesseract.has_user_words(options.tesseract_env) and (
-        options.user_words or options.user_patterns
-    ):
-        log.warning(
-            "Tesseract 4.0 ignores --user-words and --user-patterns, so these "
-            "arguments have no effect."
-        )
 
 
 def check_options_metadata(options):
-    import unicodedata
-
     docinfo = [options.title, options.author, options.keywords, options.subject]
     for s in (m for m in docinfo if m):
         for c in s:
@@ -300,9 +252,9 @@ def check_options_pillow(options):
         PIL.Image.MAX_IMAGE_PIXELS = None
 
 
-def check_options(options):
+def _check_options(options, plugin_manager, ocr_engine_languages):
     check_platform()
-    check_options_languages(options)
+    check_options_languages(options, ocr_engine_languages)
     check_options_metadata(options)
     check_options_output(options)
     check_options_sidecar(options)
@@ -311,10 +263,15 @@ def check_options(options):
     check_options_optimizing(options)
     check_options_advanced(options)
     check_options_pillow(options)
-    check_dependency_versions(options)
+    plugin_manager.hook.check_options(options=options)
 
 
-def check_closed_streams(options):
+def check_options(options, plugin_manager):
+    ocr_engine_languages = plugin_manager.hook.get_ocr_engine().languages(options)
+    _check_options(options, plugin_manager, ocr_engine_languages)
+
+
+def check_closed_streams(options):  # pragma: no cover
     """Work around Python issue with multiprocessing forking on closed streams
 
     https://bugs.python.org/issue28326
@@ -373,19 +330,19 @@ def log_page_orientations(pdfinfo):
         log.info('Page orientations detected: %s', ' '.join(orientations))
 
 
-def create_input_file(options, work_folder):
+def create_input_file(options, work_folder: Path) -> Tuple[Path, str]:
     if options.input_file == '-':
         # stdin
         log.info('reading file from standard input')
-        target = os.path.join(work_folder, 'stdin')
+        target = work_folder / 'stdin'
         with open(target, 'wb') as stream_buffer:
             copyfileobj(sys.stdin.buffer, stream_buffer)
-        return target
+        return target, "<stdin>"
     else:
         try:
-            target = os.path.join(work_folder, 'origin')
+            target = work_folder / 'origin'
             safe_symlink(options.input_file, target)
-            return target
+            return target, os.fspath(options.input_file)
         except FileNotFoundError:
             raise InputFileError(f"File not found - {options.input_file}")
 
@@ -410,8 +367,15 @@ def report_output_file_size(options, input_file, output_file):
         input_size = Path(input_file).stat().st_size
     except FileNotFoundError:
         return  # Outputting to stream or something
+    with pikepdf.open(output_file) as p:
+        # Overhead constants obtained by estimating amount of data added by OCR
+        # PDF/A conversion, and possible XMP metadata addition, with compression
+        FILE_OVERHEAD = 4000
+        OCR_PER_PAGE_OVERHEAD = 3000
+        reasonable_overhead = FILE_OVERHEAD + OCR_PER_PAGE_OVERHEAD * len(p.pages)
     ratio = output_size / input_size
-    if ratio < 1.35 or input_size < 25000:
+    reasonable_ratio = output_size / (input_size + reasonable_overhead)
+    if reasonable_ratio < 1.35 or input_size < 25000:
         return  # Seems fine
 
     reasons = []
@@ -451,30 +415,3 @@ def report_output_file_size(options, input_file, output_file):
         f"The output file size is {ratio:.2f}× larger than the input file.\n"
         f"{explanation}"
     )
-
-
-def check_dependency_versions(options):
-    check_external_program(
-        program='tesseract',
-        package={'linux': 'tesseract-ocr'},
-        version_checker=tesseract.version,
-        need_version='4.0.0',  # using backport for Travis CI
-    )
-    check_external_program(
-        program='gs',
-        package='ghostscript',
-        version_checker=ghostscript.version,
-        need_version='9.15',  # limited by Travis CI / Ubuntu 14.04 backports
-    )
-    if ghostscript.version() == '9.24':
-        raise MissingDependencyError(
-            "Ghostscript 9.24 contains serious regressions and is not "
-            "supported. Please upgrade to Ghostscript 9.25 or use an older "
-            "version."
-        )
-    check_external_program(
-        program='qpdf',
-        package='qpdf',
-        version_checker=qpdf.version,
-        need_version='8.0.2',
-    )

src/ocrmypdf/api.py

@@ -18,36 +18,21 @@
 import logging
 import os
 import sys
-import warnings
+from argparse import ArgumentParser
 from enum import IntEnum
 from pathlib import Path
-from typing import Dict, List, Optional
+from typing import Iterable
 
-from tqdm import tqdm
+from ocrmypdf._logging import PageNumberFilter, TqdmConsole
+from ocrmypdf._plugin_manager import get_plugin_manager
+from ocrmypdf._sync import run_pipeline
+from ocrmypdf._validation import check_options
+from ocrmypdf.cli import get_parser
 
-from ._sync import run_pipeline
-from ._validation import check_options
-from .cli import parser
-
-
-class TqdmConsole:
-    """Wrapper to log messages in a way that is compatible with tqdm progress bar"""
-
-    def __init__(self, file):
-        self.file = file
-        self.py36 = sys.version_info[0:2] == (3, 6)
-
-    def write(self, msg):
-        # When no progress bar is active, tqdm.write() routes to print()
-        if self.py36:
-            if msg.strip() != '':
-                tqdm.write(msg.rstrip(), end='\n', file=self.file)
-        else:
-            tqdm.write(msg.rstrip(), end='\n', file=self.file)
-
-    def flush(self):
-        if hasattr(self.file, "flush"):
-            self.file.flush()
+try:
+    import coloredlogs
+except ModuleNotFoundError:
+    coloredlogs = None
 
 
 class Verbosity(IntEnum):
@@ -59,7 +44,11 @@ class Verbosity(IntEnum):
     debug_all = 2  #: More detailed debugging from ocrmypdf and dependent modules
 
 
-def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=False):
+def configure_logging(
+    verbosity: Verbosity,
+    progress_bar_friendly: bool = True,
+    manage_root_logger: bool = False,
+):
     """Set up logging.
 
     Library users may wish to use this function if they want their log output to be
@@ -80,11 +69,15 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
             overwrite the progress bar
         manage_root_logger (bool): Configure the process's root logger, to ensure
             all log output is sent through
+
+    Returns:
+        The toplevel logger for ocrmypdf (or the root logger, if we are managing it).
     """
 
     prefix = '' if manage_root_logger else 'ocrmypdf'
+
     log = logging.getLogger(prefix)
-    log.setLevel(logging.INFO)
+    log.setLevel(logging.DEBUG)
 
     if progress_bar_friendly:
         console = logging.StreamHandler(stream=TqdmConsole(sys.stderr))
@@ -98,11 +91,25 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
     else:
         console.setLevel(logging.INFO)
 
-    formatter = logging.Formatter('%(levelname)7s - %(message)s')
-    if verbosity >= 1:
-        log.setLevel(logging.DEBUG)
+    console.addFilter(PageNumberFilter())
+
     if verbosity >= 2:
-        formatter = logging.Formatter('%(name)s - %(levelname)7s - %(message)s')
+        fmt = '%(levelname)7s %(name)s -%(pageno)s %(message)s'
+    else:
+        fmt = '%(pageno)s%(message)s'
+
+    use_colors = progress_bar_friendly
+    if not coloredlogs:
+        use_colors = False
+    if use_colors:
+        if os.name == 'nt':
+            use_colors = coloredlogs.enable_ansi_support()
+        if use_colors:
+            use_colors = coloredlogs.terminal_supports_colors()
+    if use_colors:
+        formatter = coloredlogs.ColoredFormatter(fmt=fmt)
+    else:
+        formatter = logging.Formatter(fmt=fmt)
 
     console.setFormatter(formatter)
     log.addHandler(console)
@@ -116,8 +123,16 @@ def configure_logging(verbosity, progress_bar_friendly=True, manage_root_logger=
     if manage_root_logger:
         logging.captureWarnings(True)
 
+    return log
 
-def create_options(*, input_file, output_file, **kwargs):
+
+def create_options(
+    *,
+    input_file: os.PathLike,
+    output_file: os.PathLike,
+    parser: ArgumentParser,
+    **kwargs,
+):
     cmdline = []
     deferred = []
 
@@ -127,7 +142,7 @@ def create_options(*, input_file, output_file, **kwargs):
 
         # These arguments with special handling for which we bypass
         # argparse
-        if arg in {'tesseract_env', 'progress_bar'}:
+        if arg in {'progress_bar', 'plugins'}:
             deferred.append((arg, val))
             continue
 
@@ -139,6 +154,12 @@ def create_options(*, input_file, output_file, **kwargs):
                 cmdline.append(f"--{cmd_style_arg}")
             continue
 
+        if isinstance(val, Iterable) and not isinstance(val, str):
+            for elem in val:
+                cmdline.append(f"--{cmd_style_arg}")
+                cmdline.append(elem)
+            continue
+
         # We have a parameter
         cmdline.append(f"--{cmd_style_arg}")
         if isinstance(val, (int, float)):
@@ -153,15 +174,10 @@ def create_options(*, input_file, output_file, **kwargs):
     cmdline.append(str(input_file))
     cmdline.append(str(output_file))
 
-    parser.api_mode = True
+    parser._api_mode = True
     options = parser.parse_args(cmdline)
     for keyword, val in deferred:
         setattr(options, keyword, val)
-
-    # If we are running a Tesseract spoof, ensure it knows what the input file is
-    if os.environ.get('PYTEST_CURRENT_TEST') and options.tesseract_env:
-        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(input_file)
-
     return options
 
 
@@ -169,7 +185,7 @@ def ocr(  # pylint: disable=unused-argument
     input_file: os.PathLike,
     output_file: os.PathLike,
     *,
-    language: List[str] = None,
+    language: Iterable[str] = None,
     image_dpi: int = None,
     output_type=None,
     sidecar: os.PathLike = None,
@@ -199,7 +215,7 @@ def ocr(  # pylint: disable=unused-argument
     jbig2_page_group_size: int = None,
     pages: str = None,
     max_image_mpixels: float = None,
-    tesseract_config: List[str] = None,
+    tesseract_config: Iterable[str] = None,
     tesseract_pagesegmode: int = None,
     tesseract_oem: int = None,
     pdf_renderer=None,
@@ -209,9 +225,10 @@ def ocr(  # pylint: disable=unused-argument
     user_words: os.PathLike = None,
     user_patterns: os.PathLike = None,
     fast_web_view: float = None,
+    plugins: Iterable[str] = None,
     keep_temporary_files: bool = None,
     progress_bar: bool = None,
-    tesseract_env: Dict[str, str] = None,
+    **kwargs,
 ):
     """Run OCRmyPDF on one PDF or image.
 
@@ -222,7 +239,6 @@ def ocr(  # pylint: disable=unused-argument
         use_threads (bool): Use worker threads instead of processes. This reduces
             performance but may make debugging easier since it is easier to set
             breakpoints.
-        tesseract_env (dict): Override environment variables for Tesseract
     Raises:
         ocrmypdf.PdfMergeFailedError: If the input PDF is malformed, preventing merging
             with the OCR layer.
@@ -246,7 +262,20 @@ def ocr(  # pylint: disable=unused-argument
     Returns:
         :class:`ocrmypdf.ExitCode`
     """
+    if not plugins:
+        plugins = []
+    else:
+        plugins = list(plugins)
 
-    options = create_options(**locals())
-    check_options(options)
-    return run_pipeline(options, api=True)
+    parser = get_parser()
+    _plugin_manager = get_plugin_manager(plugins)
+    _plugin_manager.hook.add_options(parser=parser)  # pylint: disable=no-member
+
+    create_options_kwargs = {
+        k: v for k, v in locals().items() if not k.startswith('_') and k != 'kwargs'
+    }
+    create_options_kwargs.update(kwargs)
+
+    options = create_options(**create_options_kwargs)
+    check_options(options, _plugin_manager)
+    return run_pipeline(options=options, plugin_manager=_plugin_manager, api=True)

src/ocrmypdf/builtin_plugins/__init__.py

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

src/ocrmypdf/builtin_plugins/ghostscript.py

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

src/ocrmypdf/builtin_plugins/tesseract_ocr.py

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

src/ocrmypdf/cli.py

@@ -17,8 +17,8 @@
 
 import argparse
 
-from ._version import PROGRAM_NAME as _PROGRAM_NAME
-from ._version import __version__ as _VERSION
+from ocrmypdf._version import PROGRAM_NAME as _PROGRAM_NAME
+from ocrmypdf._version import __version__ as _VERSION
 
 
 def numeric(basetype, min_=None, max_=None):
@@ -45,27 +45,43 @@ class ArgumentParser(argparse.ArgumentParser):
 
     def __init__(self, *args, **kwargs):
         super().__init__(*args, **kwargs)
-        self.api_mode = False
+        self._api_mode = False
 
     def error(self, message):
-        if not self.api_mode:
+        if not self._api_mode:
             super().error(message)
             return
         raise ValueError(message)
 
 
-parser = ArgumentParser(
-    prog=_PROGRAM_NAME,
-    fromfile_prefix_chars='@',
-    formatter_class=argparse.RawDescriptionHelpFormatter,
-    description="""\
+class LanguageSetAction(argparse.Action):
+    def __init__(self, option_strings, dest, default=None, **kwargs):
+        if default is None:
+            default = set()
+        super().__init__(option_strings, dest, default=default, **kwargs)
+
+    def __call__(self, parser, namespace, values, option_string=None):
+        dest = getattr(namespace, self.dest)
+        if '+' in values:
+            dest.update(lang for lang in values.split('+'))
+        else:
+            dest.add(values)
+
+
+def get_parser():
+    parser = ArgumentParser(
+        prog=_PROGRAM_NAME,
+        allow_abbrev=True,
+        fromfile_prefix_chars='@',
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        description="""\
 Generates a searchable PDF or PDF/A from a regular PDF.
 
 OCRmyPDF rasterizes each page of the input PDF, optionally corrects page
 rotation and performs image processing, runs the Tesseract OCR engine on the
 image, and then creates a PDF from the OCR information.
 """,
-    epilog="""\
+        epilog="""\
 OCRmyPDF attempts to keep the output file at about the same size.  If a file
 contains losslessly compressed images, and output file will be losslessly
 compressed as well.
@@ -106,386 +122,368 @@ Online documentation is located at:
     https://ocrmypdf.readthedocs.io/en/latest/introduction.html
 
 """,
-)
+    )
 
-parser.add_argument(
-    'input_file',
-    metavar="input_pdf_or_image",
-    help="PDF file containing the images to be OCRed (or '-' to read from "
-    "standard input)",
+    parser.add_argument(
+        'input_file',
+        metavar="input_pdf_or_image",
+        help="PDF file containing the images to be OCRed (or '-' to read from "
+        "standard input)",
+    )
+    parser.add_argument(
+        'output_file',
+        metavar="output_pdf",
+        help="Output searchable PDF file (or '-' to write to standard output). "
+        "Existing files will be ovewritten. If same as input file, the "
+        "input file will be updated only if processing is successful.",
+    )
+    parser.add_argument(
+        '-l',
+        '--language',
+        dest='languages',
+        action=LanguageSetAction,
+        help="Language(s) of the file to be OCRed (see tesseract --list-langs for "
+        "all language packs installed in your system). Use -l eng+deu for "
+        "multiple languages.",
+    )
+    parser.add_argument(
+        '--image-dpi',
+        metavar='DPI',
+        type=int,
+        help="For input image instead of PDF, use this DPI instead of file's.",
+    )
+    parser.add_argument(
+        '--output-type',
+        choices=['pdfa', 'pdf', 'pdfa-1', 'pdfa-2', 'pdfa-3'],
+        default='pdfa',
+        help="Choose output type. 'pdfa' creates a PDF/A-2b compliant file for "
+        "long term archiving (default, recommended) but may not suitable "
+        "for users who want their file altered as little as possible. 'pdfa' "
+        "also has problems with full Unicode text. 'pdf' attempts to "
+        "preserve file contents as much as possible. 'pdf-a1' creates a "
+        "PDF/A1-b file. 'pdf-a2' is equivalent to 'pdfa'. 'pdf-a3' creates a "
+        "PDF/A3-b file.",
+    )
+
+    # Use null string '\0' as sentinel to indicate the user supplied no argument,
+    # since that is the only invalid character for filepaths on all platforms
+    # bool('\0') is True in Python
+    parser.add_argument(
+        '--sidecar',
+        nargs='?',
+        const='\0',
+        default=None,
+        metavar='FILE',
+        help="Generate sidecar text files that contain the same text recognized "
+        "by Tesseract. This may be useful for building a OCR text database. "
+        "If FILE is omitted, the sidecar file be named {output_file}.txt "
+        "If FILE is set to '-', the sidecar is written to stdout (a "
+        "convenient way to preview OCR quality). The output file and sidecar "
+        "may not both use stdout at the same time.",
+    )
+
+    parser.add_argument(
+        '--version',
+        action='version',
+        version=_VERSION,
+        help="Print program version and exit",
+    )
+
+    jobcontrol = parser.add_argument_group("Job control options")
+    jobcontrol.add_argument(
+        '-j',
+        '--jobs',
+        metavar='N',
+        type=numeric(int, 0, 256),
+        help="Use up to N CPU cores simultaneously (default: use all).",
+    )
+    jobcontrol.add_argument(
+        '-q', '--quiet', action='store_true', help="Suppress INFO messages"
+    )
+    jobcontrol.add_argument(
+        '-v',
+        '--verbose',
+        type=numeric(int, 0, 2),
+        default=0,
+        const=1,
+        nargs='?',
+        help="Print more verbose messages for each additional verbose level. Use "
+        "`-v 1` typically for much more detailed logging. Higher numbers "
+        "are probably only useful in debugging.",
+    )
+    jobcontrol.add_argument(
+        '--no-progress-bar',
+        action='store_false',
+        dest='progress_bar',
+        help=argparse.SUPPRESS,
+    )
+    jobcontrol.add_argument(
+        '--use-threads', action='store_true', help=argparse.SUPPRESS
+    )
+
+    metadata = parser.add_argument_group(
+        "Metadata options",
+        "Set output PDF/A metadata (default: copy input document's metadata)",
+    )
+    metadata.add_argument(
+        '--title', type=str, help="Set document title (place multiple words in quotes)"
+    )
+    metadata.add_argument('--author', type=str, help="Set document author")
+    metadata.add_argument(
+        '--subject', type=str, help="Set document subject description"
+    )
+    metadata.add_argument('--keywords', type=str, help="Set document keywords")
+
+    preprocessing = parser.add_argument_group(
+        "Image preprocessing options",
+        "Options to improve the quality of the final PDF and OCR",
+    )
+    preprocessing.add_argument(
+        '-r',
+        '--rotate-pages',
+        action='store_true',
+        help="Automatically rotate pages based on detected text orientation",
+    )
+    preprocessing.add_argument(
+        '--remove-background',
+        action='store_true',
+        help="Attempt to remove background from gray or color pages, setting it "
+        "to white ",
+    )
+    preprocessing.add_argument(
+        '-d',
+        '--deskew',
+        action='store_true',
+        help="Deskew each page before performing OCR",
+    )
+    preprocessing.add_argument(
+        '-c',
+        '--clean',
+        action='store_true',
+        help="Clean pages from scanning artifacts before performing OCR, and send "
+        "the cleaned page to OCR, but do not include the cleaned page in "
+        "the output",
+    )
+    preprocessing.add_argument(
+        '-i',
+        '--clean-final',
+        action='store_true',
+        help="Clean page as above, and incorporate the cleaned image in the final "
+        "PDF.  Might remove desired content.",
+    )
+    preprocessing.add_argument(
+        '--unpaper-args',
+        type=str,
+        default=None,
+        help="A quoted string of arguments to pass to unpaper. Requires --clean. "
+        "Example: --unpaper-args '--layout double'.",
+    )
+    preprocessing.add_argument(
+        '--oversample',
+        metavar='DPI',
+        type=numeric(int, 0, 5000),
+        default=0,
+        help="Oversample images to at least the specified DPI, to improve OCR "
+        "results slightly",
+    )
+    preprocessing.add_argument(
+        '--remove-vectors',
+        action='store_true',
+        help="EXPERIMENTAL. Mask out any vector objects in the PDF so that they "
+        "will not be included in OCR. This can eliminate false characters.",
+    )
+    preprocessing.add_argument(
+        '--threshold',
+        action='store_true',
+        help=(
+            "EXPERIMENTAL. Threshold image to 1bpp before sending it to Tesseract "
+            "for OCR. Can improve OCR quality compared to Tesseract's thresholder."
+        ),
+    )
+
+    ocrsettings = parser.add_argument_group("OCR options", "Control how OCR is applied")
+    ocrsettings.add_argument(
+        '-f',
+        '--force-ocr',
+        action='store_true',
+        help="Rasterize any text or vector objects on each page, apply OCR, and "
+        "save the rastered output (this rewrites the PDF)",
+    )
+    ocrsettings.add_argument(
+        '-s',
+        '--skip-text',
+        action='store_true',
+        help="Skip OCR on any pages that already contain text, but include the "
+        "page in final output; useful for PDFs that contain a mix of "
+        "images, text pages, and/or previously OCRed pages",
+    )
+    ocrsettings.add_argument(
+        '--redo-ocr',
+        action='store_true',
+        help="Attempt to detect and remove the hidden OCR layer from files that "
+        "were previously OCRed with OCRmyPDF or another program. Apply OCR "
+        "to text found in raster images. Existing visible text objects will "
+        "not be changed. If there is no existing OCR, OCR will be added.",
+    )
+    ocrsettings.add_argument(
+        '--skip-big',
+        type=numeric(float, 0, 5000),
+        metavar='MPixels',
+        help="Skip OCR on pages larger than the specified amount of megapixels, "
+        "but include skipped pages in final output",
+    )
+
+    optimizing = parser.add_argument_group(
+        "Optimization options", "Control how the PDF is optimized after OCR"
+    )
+    optimizing.add_argument(
+        '-O',
+        '--optimize',
+        type=int,
+        choices=range(0, 4),
+        default=1,
+        help=(
+            "Control how PDF is optimized after processing:"
+            "0 - do not optimize; "
+            "1 - do safe, lossless optimizations (default); "
+            "2 - do some lossy optimizations; "
+            "3 - do aggressive lossy optimizations (including lossy JBIG2)"
+        ),
+    )
+    optimizing.add_argument(
+        '--jpeg-quality',
+        type=numeric(int, 0, 100),
+        default=0,
+        metavar='Q',
+        help=(
+            "Adjust JPEG quality level for JPEG optimization. "
+            "100 is best quality and largest output size; "
+            "1 is lowest quality and smallest output; "
+            "0 uses the default."
+        ),
+    )
+    optimizing.add_argument(
+        '--jpg-quality',
+        type=numeric(int, 0, 100),
+        default=0,
+        metavar='Q',
+        dest='jpeg_quality',
+        help=argparse.SUPPRESS,  # Alias for --jpeg-quality
+    )
+    optimizing.add_argument(
+        '--png-quality',
+        type=numeric(int, 0, 100),
+        default=0,
+        metavar='Q',
+        help=(
+            "Adjust PNG quality level to use when quantizing PNGs. "
+            "Values have same meaning as with --jpeg-quality"
+        ),
+    )
+    optimizing.add_argument(
+        '--jbig2-lossy',
+        action='store_true',
+        help=(
+            "Enable JBIG2 lossy mode (better compression, not suitable for some "
+            "use cases - see documentation)."
+        ),
+    )
+    optimizing.add_argument(
+        '--jbig2-page-group-size',
+        type=numeric(int, 1, 10000),
+        default=0,
+        metavar='N',
+        # Adjust number of pages to consider at once for JBIG2 compression
+        help=argparse.SUPPRESS,
+    )
+
+    advanced = parser.add_argument_group(
+        "Advanced", "Advanced options to control OCRmyPDF"
+    )
+    advanced.add_argument(
+        '--pages',
+        type=str,
+        help=(
+            "Limit OCR to the specified pages (ranges or comma separated), "
+            "skipping others"
+        ),
+    )
+    advanced.add_argument(
+        '--max-image-mpixels',
+        action='store',
+        type=numeric(float, 0),
+        metavar='MPixels',
+        help="Set maximum number of pixels to unpack before treating an image as a "
+        "decompression bomb",
+        default=128.0,
+    )
+    advanced.add_argument(
+        '--pdf-renderer',
+        choices=['auto', 'hocr', 'sandwich'],
+        default='auto',
+        help="Choose OCR PDF renderer - the default option is to let OCRmyPDF "
+        "choose.  See documentation for discussion.",
+    )
+    advanced.add_argument(
+        '--rotate-pages-threshold',
+        default=14.0,
+        type=numeric(float, 0, 1000),
+        metavar='CONFIDENCE',
+        help="Only rotate pages when confidence is above this value (arbitrary "
+        "units reported by tesseract)",
+    )
+    advanced.add_argument(
+        '--pdfa-image-compression',
+        choices=['auto', 'jpeg', 'lossless'],
+        default='auto',
+        help="Specify how to compress images in the output PDF/A. 'auto' lets "
+        "OCRmyPDF decide.  'jpeg' changes all grayscale and color images to "
+        "JPEG compression.  'lossless' uses PNG-style lossless compression "
+        "for all images.  Monochrome images are always compressed using a "
+        "lossless codec.  Compression settings "
+        "are applied to all pages, including those for which OCR was "
+        "skipped.  Not supported for --output-type=pdf ; that setting "
+        "preserves the original compression of all images.",
+    )
+    advanced.add_argument(
+        '--fast-web-view',
+        type=numeric(float, 0),
+        default=1.0,
+        metavar="MEGABYTES",
+        help="If the size of file is more than this threshold (in MB), then "
+        "linearize the PDF for fast web viewing. This allows the PDF to be "
+        "displayed before it is fully downloaded in web browsers, but increases "
+        "the space required slightly. By default we skip this for small files "
+        "which do not benefit. If the threshold is 0 it will be apply to all files. "
+        "Set the threshold very high to disable.",
+    )
+    advanced.add_argument(
+        '--plugin',
+        dest='plugins',
+        action='append',
+        default=[],
+        help="Name of plugin to import.",
+    )
+
+    debugging = parser.add_argument_group(
+        "Debugging", "Arguments to help with troubleshooting and debugging"
+    )
+    debugging.add_argument(
+        '-k',
+        '--keep-temporary-files',
+        action='store_true',
+        help="Keep temporary files (helpful for debugging)",
+    )
+    return parser
+
+
+plugins_only_parser = ArgumentParser(
+    prog=_PROGRAM_NAME, fromfile_prefix_chars='@', add_help=False, allow_abbrev=False
 )
-parser.add_argument(
-    'output_file',
-    metavar="output_pdf",
-    help="Output searchable PDF file (or '-' to write to standard output). "
-    "Existing files will be ovewritten. If same as input file, the "
-    "input file will be updated only if processing is successful.",
-)
-parser.add_argument(
-    '-l',
-    '--language',
+plugins_only_parser.add_argument(
+    '--plugin',
+    dest='plugins',
     action='append',
-    help="Language(s) of the file to be OCRed (see tesseract --list-langs for "
-    "all language packs installed in your system). Use -l eng+deu for "
-    "multiple languages.",
-)
-parser.add_argument(
-    '--image-dpi',
-    metavar='DPI',
-    type=int,
-    help="For input image instead of PDF, use this DPI instead of file's.",
-)
-parser.add_argument(
-    '--output-type',
-    choices=['pdfa', 'pdf', 'pdfa-1', 'pdfa-2', 'pdfa-3'],
-    default='pdfa',
-    help="Choose output type. 'pdfa' creates a PDF/A-2b compliant file for "
-    "long term archiving (default, recommended) but may not suitable "
-    "for users who want their file altered as little as possible. 'pdfa' "
-    "also has problems with full Unicode text. 'pdf' attempts to "
-    "preserve file contents as much as possible. 'pdf-a1' creates a "
-    "PDF/A1-b file. 'pdf-a2' is equivalent to 'pdfa'. 'pdf-a3' creates a "
-    "PDF/A3-b file.",
-)
-
-# Use null string '\0' as sentinel to indicate the user supplied no argument,
-# since that is the only invalid character for filepaths on all platforms
-# bool('\0') is True in Python
-parser.add_argument(
-    '--sidecar',
-    nargs='?',
-    const='\0',
-    default=None,
-    metavar='FILE',
-    help="Generate sidecar text files that contain the same text recognized "
-    "by Tesseract. This may be useful for building a OCR text database. "
-    "If FILE is omitted, the sidecar file be named {output_file}.txt "
-    "If FILE is set to '-', the sidecar is written to stdout (a "
-    "convenient way to preview OCR quality). The output file and sidecar "
-    "may not both use stdout at the same time.",
-)
-
-parser.add_argument(
-    '--version',
-    action='version',
-    version=_VERSION,
-    help="Print program version and exit",
-)
-
-jobcontrol = parser.add_argument_group("Job control options")
-jobcontrol.add_argument(
-    '-j',
-    '--jobs',
-    metavar='N',
-    type=numeric(int, 0, 256),
-    help="Use up to N CPU cores simultaneously (default: use all).",
-)
-jobcontrol.add_argument(
-    '-q', '--quiet', action='store_true', help="Suppress INFO messages"
-)
-jobcontrol.add_argument(
-    '-v',
-    '--verbose',
-    type=numeric(int, 0, 2),
-    default=0,
-    const=1,
-    nargs='?',
-    help="Print more verbose messages for each additional verbose level. Use "
-    "`-v 1` typically for much more detailed logging. Higher numbers "
-    "are probably only useful in debugging.",
-)
-jobcontrol.add_argument(
-    '--no-progress-bar',
-    action='store_false',
-    dest='progress_bar',
-    help=argparse.SUPPRESS,
-)
-jobcontrol.add_argument('--use-threads', action='store_true', help=argparse.SUPPRESS)
-
-metadata = parser.add_argument_group(
-    "Metadata options",
-    "Set output PDF/A metadata (default: copy input document's metadata)",
-)
-metadata.add_argument(
-    '--title', type=str, help="Set document title (place multiple words in quotes)"
-)
-metadata.add_argument('--author', type=str, help="Set document author")
-metadata.add_argument('--subject', type=str, help="Set document subject description")
-metadata.add_argument('--keywords', type=str, help="Set document keywords")
-
-preprocessing = parser.add_argument_group(
-    "Image preprocessing options",
-    "Options to improve the quality of the final PDF and OCR",
-)
-preprocessing.add_argument(
-    '-r',
-    '--rotate-pages',
-    action='store_true',
-    help="Automatically rotate pages based on detected text orientation",
-)
-preprocessing.add_argument(
-    '--remove-background',
-    action='store_true',
-    help="Attempt to remove background from gray or color pages, setting it "
-    "to white ",
-)
-preprocessing.add_argument(
-    '-d', '--deskew', action='store_true', help="Deskew each page before performing OCR"
-)
-preprocessing.add_argument(
-    '-c',
-    '--clean',
-    action='store_true',
-    help="Clean pages from scanning artifacts before performing OCR, and send "
-    "the cleaned page to OCR, but do not include the cleaned page in "
-    "the output",
-)
-preprocessing.add_argument(
-    '-i',
-    '--clean-final',
-    action='store_true',
-    help="Clean page as above, and incorporate the cleaned image in the final "
-    "PDF.  Might remove desired content.",
-)
-preprocessing.add_argument(
-    '--unpaper-args',
-    type=str,
-    default=None,
-    help="A quoted string of arguments to pass to unpaper. Requires --clean. "
-    "Example: --unpaper-args '--layout double'.",
-)
-preprocessing.add_argument(
-    '--oversample',
-    metavar='DPI',
-    type=numeric(int, 0, 5000),
-    default=0,
-    help="Oversample images to at least the specified DPI, to improve OCR "
-    "results slightly",
-)
-preprocessing.add_argument(
-    '--remove-vectors',
-    action='store_true',
-    help="EXPERIMENTAL. Mask out any vector objects in the PDF so that they "
-    "will not be included in OCR. This can eliminate false characters.",
-)
-preprocessing.add_argument(
-    '--threshold',
-    action='store_true',
-    help="EXPERIMENTAL. Threshold image to 1bpp before sending it to Tesseract for OCR. Can "
-    "improve OCR quality compared to Tesseract's thresholder.",
-)
-
-ocrsettings = parser.add_argument_group("OCR options", "Control how OCR is applied")
-ocrsettings.add_argument(
-    '-f',
-    '--force-ocr',
-    action='store_true',
-    help="Rasterize any text or vector objects on each page, apply OCR, and "
-    "save the rastered output (this rewrites the PDF)",
-)
-ocrsettings.add_argument(
-    '-s',
-    '--skip-text',
-    action='store_true',
-    help="Skip OCR on any pages that already contain text, but include the "
-    "page in final output; useful for PDFs that contain a mix of "
-    "images, text pages, and/or previously OCRed pages",
-)
-ocrsettings.add_argument(
-    '--redo-ocr',
-    action='store_true',
-    help="Attempt to detect and remove the hidden OCR layer from files that "
-    "were previously OCRed with OCRmyPDF or another program. Apply OCR "
-    "to text found in raster images. Existing visible text objects will "
-    "not be changed. If there is no existing OCR, OCR will be added.",
-)
-ocrsettings.add_argument(
-    '--skip-big',
-    type=numeric(float, 0, 5000),
-    metavar='MPixels',
-    help="Skip OCR on pages larger than the specified amount of megapixels, "
-    "but include skipped pages in final output",
-)
-
-optimizing = parser.add_argument_group(
-    "Optimization options", "Control how the PDF is optimized after OCR"
-)
-optimizing.add_argument(
-    '-O',
-    '--optimize',
-    type=int,
-    choices=range(0, 4),
-    default=1,
-    help=(
-        "Control how PDF is optimized after processing:"
-        "0 - do not optimize; "
-        "1 - do safe, lossless optimizations (default); "
-        "2 - do some lossy optimizations; "
-        "3 - do aggressive lossy optimizations (including lossy JBIG2)"
-    ),
-)
-optimizing.add_argument(
-    '--jpeg-quality',
-    type=numeric(int, 0, 100),
-    default=0,
-    metavar='Q',
-    help=(
-        "Adjust JPEG quality level for JPEG optimization. "
-        "100 is best quality and largest output size; "
-        "1 is lowest quality and smallest output; "
-        "0 uses the default."
-    ),
-)
-optimizing.add_argument(
-    '--jpg-quality',
-    type=numeric(int, 0, 100),
-    default=0,
-    metavar='Q',
-    dest='jpeg_quality',
-    help=argparse.SUPPRESS,  # Alias for --jpeg-quality
-)
-optimizing.add_argument(
-    '--png-quality',
-    type=numeric(int, 0, 100),
-    default=0,
-    metavar='Q',
-    help=(
-        "Adjust PNG quality level to use when quantizing PNGs. "
-        "Values have same meaning as with --jpeg-quality"
-    ),
-)
-optimizing.add_argument(
-    '--jbig2-lossy',
-    action='store_true',
-    help=(
-        "Enable JBIG2 lossy mode (better compression, not suitable for some "
-        "use cases - see documentation)."
-    ),
-)
-optimizing.add_argument(
-    '--jbig2-page-group-size',
-    type=numeric(int, 1, 10000),
-    default=0,
-    metavar='N',
-    # Adjust number of pages to consider at once for JBIG2 compression
-    help=argparse.SUPPRESS,
-)
-
-advanced = parser.add_argument_group(
-    "Advanced", "Advanced options to control Tesseract's OCR behavior"
-)
-advanced.add_argument(
-    '--pages',
-    type=str,
-    help="Limit OCR to the specified pages (ranges or comma separated), skipping others",
-)
-advanced.add_argument(
-    '--max-image-mpixels',
-    action='store',
-    type=numeric(float, 0),
-    metavar='MPixels',
-    help="Set maximum number of pixels to unpack before treating an image as a "
-    "decompression bomb",
-    default=128.0,
-)
-advanced.add_argument(
-    '--tesseract-config',
-    action='append',
-    metavar='CFG',
     default=[],
-    help="Additional Tesseract configuration files -- see documentation",
+    help="Name of plugin to import.",
 )
-advanced.add_argument(
-    '--tesseract-pagesegmode',
-    action='store',
-    type=int,
-    metavar='PSM',
-    choices=range(0, 14),
-    help="Set Tesseract page segmentation mode (see tesseract --help)",
-)
-advanced.add_argument(
-    '--tesseract-oem',
-    action='store',
-    type=int,
-    metavar='MODE',
-    choices=range(0, 4),
-    help=(
-        "Set Tesseract 4.0 OCR engine mode: "
-        "0 - original Tesseract only; "
-        "1 - neural nets LSTM only; "
-        "2 - Tesseract + LSTM; "
-        "3 - default."
-    ),
-)
-advanced.add_argument(
-    '--pdf-renderer',
-    choices=['auto', 'hocr', 'sandwich'],
-    default='auto',
-    help="Choose OCR PDF renderer - the default option is to let OCRmyPDF "
-    "choose.  See documentation for discussion.",
-)
-advanced.add_argument(
-    '--tesseract-timeout',
-    default=180.0,
-    type=numeric(float, 0),
-    metavar='SECONDS',
-    help='Give up on OCR after the timeout, but copy the preprocessed page '
-    'into the final output',
-)
-advanced.add_argument(
-    '--rotate-pages-threshold',
-    default=14.0,
-    type=numeric(float, 0, 1000),
-    metavar='CONFIDENCE',
-    help="Only rotate pages when confidence is above this value (arbitrary "
-    "units reported by tesseract)",
-)
-advanced.add_argument(
-    '--pdfa-image-compression',
-    choices=['auto', 'jpeg', 'lossless'],
-    default='auto',
-    help="Specify how to compress images in the output PDF/A. 'auto' lets "
-    "OCRmyPDF decide.  'jpeg' changes all grayscale and color images to "
-    "JPEG compression.  'lossless' uses PNG-style lossless compression "
-    "for all images.  Monochrome images are always compressed using a "
-    "lossless codec.  Compression settings "
-    "are applied to all pages, including those for which OCR was "
-    "skipped.  Not supported for --output-type=pdf ; that setting "
-    "preserves the original compression of all images.",
-)
-advanced.add_argument(
-    '--user-words',
-    metavar='FILE',
-    help="Specify the location of the Tesseract user words file. This is a "
-    "list of words Tesseract should consider while performing OCR in "
-    "addition to its standard language dictionaries. This can improve "
-    "OCR quality especially for specialized and technical documents.",
-)
-advanced.add_argument(
-    '--user-patterns',
-    metavar='FILE',
-    help="Specify the location of the Tesseract user patterns file.",
-)
-advanced.add_argument(
-    '--fast-web-view',
-    type=numeric(float, 0),
-    default=1.0,
-    metavar="MEGABYTES",
-    help="If the size of file is more than this threshold (in MB), then "
-    "linearize the PDF for fast web viewing. This allows the PDF to be "
-    "displayed before it is fully downloaded in web browsers, but increases "
-    "the space required slightly. By default we skip this for small files "
-    "which do not benefit. If the threshold is 0 it will be apply to all files. "
-    "Set the threshold very high to disable.",
-)
-
-debugging = parser.add_argument_group(
-    "Debugging", "Arguments to help with troubleshooting and debugging"
-)
-debugging.add_argument(
-    '-k',
-    '--keep-temporary-files',
-    action='store_true',
-    help="Keep temporary files (helpful for debugging)",
-)
-debugging.add_argument('--tesseract-env', type=str, help=argparse.SUPPRESS)

src/ocrmypdf/exec/qpdf.py

@@ -1,61 +0,0 @@
-# © 2017 James R. Barlow: github.com/jbarlow83
-#
-# This file is part of OCRmyPDF.
-#
-# OCRmyPDF is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# OCRmyPDF is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
-
-"""Interface to qpdf executable"""
-
-from io import StringIO
-
-import pikepdf
-
-
-def version():
-    return pikepdf.__libqpdf_version__
-
-
-def check(input_file, log=None):
-    pdf = None
-    try:
-        pdf = pikepdf.open(input_file)
-    except pikepdf.PdfError as e:
-        if log:
-            log.error(e)
-        return False
-    else:
-        messages = pdf.check()
-        for msg in messages:
-            if 'error' in msg.lower():
-                log.error(msg)
-            else:
-                log.warning(msg)
-
-        sio = StringIO()
-        linearize = None
-        try:
-            pdf.check_linearization(sio)
-        except RuntimeError:
-            pass
-        else:
-            linearize = sio.getvalue()
-            if linearize:
-                log.warning(linearize)
-
-        if not messages and not linearize:
-            return True
-        return False
-    finally:
-        if pdf:
-            pdf.close()

src/ocrmypdf/helpers.py

@@ -20,23 +20,56 @@ import multiprocessing
 import os
 import shutil
 import warnings
+from collections import namedtuple
 from collections.abc import Iterable
 from contextlib import suppress
 from functools import wraps
+from io import StringIO
+from math import isclose
 from pathlib import Path
+from typing import Any, Sequence
+
+import pikepdf
 
 log = logging.getLogger(__name__)
 
 
-def safe_symlink(input_file, soft_link_name, *args, **kwargs):
+class Resolution(namedtuple('Resolution', ('x', 'y'))):
+    __slots__ = ()
+
+    def round(self, ndigits):
+        return Resolution(round(self.x, ndigits), round(self.y, ndigits))
+
+    def to_int(self):
+        return Resolution(int(round(self.x)), int(round(self.y)))
+
+    @property
+    def is_square(self):
+        return isclose(self.x, self.y, rel_tol=1e-3)
+
+    def take_max(self, vals, yvals=None):
+        if yvals is not None:
+            return Resolution(max(self.x, *vals), max(self.y, *yvals))
+        max_x, max_y = self.x, self.y
+        for x, y in vals:
+            max_x = max(x, max_x)
+            max_y = max(y, max_y)
+        return Resolution(max_x, max_y)
+
+    def flip_axis(self):
+        return Resolution(self.y, self.x)
+
+    def __str__(self):
+        return f"{self.x:f}x{self.y:f}"
+
+    def __repr__(self):
+        return f"Resolution({self.x}x{self.y} dpi)"
+
+
+def safe_symlink(input_file: os.PathLike, soft_link_name: os.PathLike):
     """
     Helper function: relinks soft symbolic link if necessary
     """
-    if len(args) == 1 and isinstance(args[0], logging.Logger):
-        log.warning("Deprecated: safe_symlink(,log)")
-    if 'log' in kwargs:
-        log.warning('Deprecated: safe_symlink(...log=)')
-
     input_file = os.fspath(input_file)
     soft_link_name = os.fspath(soft_link_name)
 
@@ -72,21 +105,28 @@ def safe_symlink(input_file, soft_link_name, *args, **kwargs):
     os.symlink(os.path.abspath(input_file), soft_link_name)
 
 
-def is_iterable_notstr(thing):
+def samefile(f1: os.PathLike, f2: os.PathLike):
+    if os.name == 'nt':
+        return f1 == f2
+    else:
+        return os.path.samefile(f1, f2)
+
+
+def is_iterable_notstr(thing: Any) -> bool:
     return isinstance(thing, Iterable) and not isinstance(thing, str)
 
 
-def monotonic(L):
+def monotonic(L: Sequence) -> bool:
     """Does list increase monotonically?"""
     return all(b > a for a, b in zip(L, L[1:]))
 
 
-def page_number(input_file):
+def page_number(input_file: os.PathLike) -> int:
     """Get one-based page number implied by filename (000002.pdf -> 2)"""
     return int(os.path.basename(os.fspath(input_file))[0:6])
 
 
-def available_cpu_count():
+def available_cpu_count() -> int:
     try:
         return multiprocessing.cpu_count()
     except NotImplementedError:
@@ -97,35 +137,73 @@ def available_cpu_count():
     return 1
 
 
-def is_file_writable(test_file):
+def is_file_writable(test_file: os.PathLike) -> bool:
     """Intentionally racy test if target is writable.
 
     We intend to write to the output file if and only if we succeed and
     can replace it atomically. Before doing the OCR work, make sure
     the location is writable.
     """
-    p = Path(test_file)
+    try:
+        p = Path(test_file)
+        if p.is_symlink():
+            p = p.resolve(strict=False)
 
-    if p.is_symlink():
-        p = p.resolve(strict=False)
-
-    # p.is_file() throws an exception in some cases
-    if p.exists() and p.is_file():
-        return os.access(
-            os.fspath(p),
-            os.W_OK,
-            effective_ids=(os.access in os.supports_effective_ids),
-        )
-    else:
-        try:
-            fp = p.open('wb')
-        except OSError:
-            return False
+        # p.is_file() throws an exception in some cases
+        if p.exists() and p.is_file():
+            return os.access(
+                os.fspath(p),
+                os.W_OK,
+                effective_ids=(os.access in os.supports_effective_ids),
+            )
         else:
-            fp.close()
-            with suppress(OSError):
-                p.unlink()
-        return True
+            try:
+                fp = p.open('wb')
+            except OSError:
+                return False
+            else:
+                fp.close()
+                with suppress(OSError):
+                    p.unlink()
+            return True
+    except (EnvironmentError, RuntimeError) as e:
+        log.debug(e)
+        log.error(str(e))
+        return False
+
+
+def check_pdf(input_file: Path) -> bool:
+    pdf = None
+    try:
+        pdf = pikepdf.open(input_file)
+    except pikepdf.PdfError as e:
+        log.error(e)
+        return False
+    else:
+        messages = pdf.check()
+        for msg in messages:
+            if 'error' in msg.lower():
+                log.error(msg)
+            else:
+                log.warning(msg)
+
+        sio = StringIO()
+        linearize = None
+        try:
+            pdf.check_linearization(sio)
+        except RuntimeError:
+            pass
+        else:
+            linearize = sio.getvalue()
+            if linearize:
+                log.warning(linearize)
+
+        if not messages and not linearize:
+            return True
+        return False
+    finally:
+        if pdf:
+            pdf.close()
 
 
 def deprecated(func):

src/ocrmypdf/hocrtransform.py

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

src/ocrmypdf/leptonica.py

@@ -29,13 +29,13 @@ from collections.abc import Sequence
 from contextlib import suppress
 from ctypes.util import find_library
 from functools import lru_cache
-from io import BytesIO
+from io import BytesIO, UnsupportedOperation
 from os import fspath
 from tempfile import TemporaryFile
 
-from .exceptions import MissingDependencyError
-from .exec import shim_paths_with_program_files
-from .lib._leptonica import ffi
+from ocrmypdf.exceptions import MissingDependencyError
+from ocrmypdf.lib._leptonica import ffi
+from ocrmypdf.subprocess import shim_paths_with_program_files
 
 # pylint: disable=protected-access
 
@@ -47,21 +47,32 @@ if os.name == 'nt':
 else:
     libname = 'lept'
 _libpath = find_library(libname)
-if not _libpath and os.name == 'nt':
+if not _libpath:
     raise MissingDependencyError(
         """
         ---------------------------------------------------------------------
-        This error normally occurs when ocrmypdf can't find a file named
-        liblept-5.dll (Leptonica).  Please ensure Tesseract-OCR is installed
-        and its location is added to the system PATH environment variable.
+        This error normally occurs when ocrmypdf can't find the Leptonica
+        library, which is usually installed with Tesseract OCR. It could be that
+        Tesseract is not installed properly, we can't find the installation
+        on your system PATH environment variable.
 
-        For details see:
+        The library we are looking for is usually called:
+            liblept-5.dll   (Windows)
+            liblept*.dylib  (macOS)
+            liblept*.so     (Linux/BSD)
+
+        Please review our installation procedures to find a solution:
             https://ocrmypdf.readthedocs.io/en/latest/installation.html
         ---------------------------------------------------------------------
         """
     )
-lept = ffi.dlopen(_libpath)
-lept.setMsgSeverity(lept.L_SEVERITY_WARNING)
+try:
+    lept = ffi.dlopen(_libpath)
+    lept.setMsgSeverity(lept.L_SEVERITY_WARNING)
+except ffi.error as e:
+    raise MissingDependencyError(
+        f"Leptonica library found at {_libpath}, but we could not access it"
+    ) from e
 
 
 class _LeptonicaErrorTrap:
@@ -85,7 +96,6 @@ class _LeptonicaErrorTrap:
         self.no_stderr = False
 
     def __enter__(self):
-        from io import UnsupportedOperation
 
         self.tmpfile = TemporaryFile()
 
@@ -340,7 +350,7 @@ class Pix(LeptonicaObject):
             py_file.write(buffer)
 
     @classmethod
-    def frompil(self, pillow_image):
+    def frompil(cls, pillow_image):
         """Create a copy of a PIL.Image from this Pix"""
         bio = BytesIO()
         pillow_image.save(bio, format='png', compress_level=1)
@@ -352,7 +362,7 @@ class Pix(LeptonicaObject):
 
     def topil(self):
         """Returns a PIL.Image version of this Pix"""
-        from PIL import Image
+        from PIL import Image  # pylint: disable=import-outside-toplevel
 
         # Leptonica manages data in words, so it implicitly does an endian
         # swap.  Tell Pillow about this when it reads the data.
@@ -523,16 +533,7 @@ class Pix(LeptonicaObject):
             )
             return Pix(thresh_pix)
 
-    def crop_to_foreground(
-        self,
-        threshold=128,
-        mindist=70,
-        erasedist=30,
-        pagenum=0,
-        showmorph=0,
-        display=0,
-        pdfdir=ffi.NULL,
-    ):
+    def crop_to_foreground(self, threshold=128, mindist=70, erasedist=30, showmorph=0):
         if get_leptonica_version() < 'leptonica-1.76':
             # Leptonica 1.76 changed the API for pixFindPageForeground; we don't
             # support the old version

src/ocrmypdf/lib/compile_leptonica.py

@@ -443,6 +443,12 @@ pixReadBarcodes(PIXA     *pixa,
                 SARRAY  **psaw,
                 l_int32 debugflag);
 
+PIX *
+pixGenHalftoneMask(PIX      *pixs,
+                        PIX     **ppixtext,
+                        l_int32  *phtfound,
+                        PIXA     *pixadb);
+
 l_int32
 l_generateCIDataForPdf(const char *fname,
                        PIX *pix,

src/ocrmypdf/optimize.py

@@ -15,10 +15,11 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import concurrent.futures
+import logging
 import sys
 import tempfile
 from collections import defaultdict
+from functools import partial
 from os import fspath
 from pathlib import Path
 
@@ -27,11 +28,14 @@ from pikepdf import Dictionary, Name
 from PIL import Image
 from tqdm import tqdm
 
-from . import leptonica
-from ._jobcontext import PDFContext
-from .exceptions import OutputFileAccessError
-from .exec import jbig2enc, pngquant
-from .helpers import safe_symlink
+from ocrmypdf import leptonica
+from ocrmypdf._concurrent import exec_progress_pool
+from ocrmypdf._exec import jbig2enc, pngquant
+from ocrmypdf._jobcontext import PdfContext
+from ocrmypdf.exceptions import OutputFileAccessError
+from ocrmypdf.helpers import safe_symlink
+
+log = logging.getLogger(__name__)
 
 DEFAULT_JPEG_QUALITY = 75
 DEFAULT_PNG_QUALITY = 70
@@ -53,7 +57,7 @@ def tif_name(root, xref):
     return img_name(root, xref, '.tif')
 
 
-def extract_image_filter(pike, root, log, image, xref):
+def extract_image_filter(pike, root, image, xref):
     if image.Subtype != Name.Image:
         return None
     if image.Length < 100:
@@ -79,8 +83,8 @@ def extract_image_filter(pike, root, log, image, xref):
     return pim, filtdp
 
 
-def extract_image_jbig2(*, pike, root, log, image, xref, options):
-    result = extract_image_filter(pike, root, log, image, xref)
+def extract_image_jbig2(*, pike, root, image, xref, options):
+    result = extract_image_filter(pike, root, image, xref)
     if result is None:
         return None
     pim, filtdp = result
@@ -101,8 +105,8 @@ def extract_image_jbig2(*, pike, root, log, image, xref, options):
     return None
 
 
-def extract_image_generic(*, pike, root, log, image, xref, options):
-    result = extract_image_filter(pike, root, log, image, xref)
+def extract_image_generic(*, pike, root, image, xref, options):
+    result = extract_image_filter(pike, root, image, xref)
     if result is None:
         return None
     pim, filtdp = result
@@ -155,11 +159,22 @@ def extract_image_generic(*, pike, root, log, image, xref, options):
         # generating a PNG from compressed data
         pim.as_pil_image().save(png_name(root, xref))
         return xref, '.png'
+    elif (
+        not pim.indexed
+        and pim.colorspace == Name.ICCBased
+        and pim.bits_per_component == 1
+        and not options.jbig2_lossy
+    ):
+        # We can losslessly optimize 1-bit images to CCITT or JBIG2 without
+        # paying any attention to the ICC profile, provided we're not doing
+        # lossy JBIG2
+        pim.as_pil_image().save(png_name(root, xref))
+        return xref, '.png'
 
     return None
 
 
-def extract_images(pike, root, log, options, extract_fn):
+def extract_images(pike, root, options, extract_fn):
     """Extract image using extract_fn
 
     Enumerate images on each page, lookup their xref/ID number in the PDF.
@@ -201,9 +216,9 @@ def extract_images(pike, root, log, options, extract_fn):
         image = pike.get_object((xref, 0))
         try:
             result = extract_fn(
-                pike=pike, root=root, log=log, image=image, xref=xref, options=options
+                pike=pike, root=root, image=image, xref=xref, options=options
             )
-        except Exception as e:
+        except Exception as e:  # pylint: disable=broad-except
             log.debug("Image xref %s, error %s", xref, repr(e))
             errors += 1
         else:
@@ -212,12 +227,12 @@ def extract_images(pike, root, log, options, extract_fn):
                 yield pageno_for_xref[xref], xref, ext
 
 
-def extract_images_generic(pike, root, log, options):
+def extract_images_generic(pike, root, options):
     """Extract any >=2bpp image we think we can improve"""
 
     jpegs = []
     pngs = []
-    for _, xref, ext in extract_images(pike, root, log, options, extract_image_generic):
+    for _, xref, ext in extract_images(pike, root, options, extract_image_generic):
         log.debug('xref = %s ext = %s', xref, ext)
         if ext == '.png':
             pngs.append(xref)
@@ -227,13 +242,11 @@ def extract_images_generic(pike, root, log, options):
     return jpegs, pngs
 
 
-def extract_images_jbig2(pike, root, log, options):
+def extract_images_jbig2(pike, root, options):
     """Extract any bitonal image that we think we can improve as JBIG2"""
 
     jbig2_groups = defaultdict(list)
-    for pageno, xref, ext in extract_images(
-        pike, root, log, options, extract_image_jbig2
-    ):
+    for pageno, xref, ext in extract_images(pike, root, options, extract_image_jbig2):
         group = pageno // options.jbig2_page_group_size
         jbig2_groups[group].append((xref, ext))
 
@@ -245,55 +258,55 @@ def extract_images_jbig2(pike, root, log, options):
     return jbig2_groups
 
 
-def _produce_jbig2_images(jbig2_groups, root, log, options):
+def _produce_jbig2_images(jbig2_groups, root, options):
     """Produce JBIG2 images from their groups"""
 
-    def jbig2_group_futures(executor, root, groups):
+    def jbig2_group_args(root, groups):
         for group, xref_exts in groups.items():
             prefix = f'group{group:08d}'
-            future = executor.submit(
-                jbig2enc.convert_group,
+            yield dict(
                 cwd=fspath(root),
                 infiles=(img_name(root, xref, ext) for xref, ext in xref_exts),
                 out_prefix=prefix,
             )
-            yield future
 
-    def jbig2_single_futures(executor, root, groups):
+    def jbig2_single_args(root, groups):
         for group, xref_exts in groups.items():
             prefix = f'group{group:08d}'
             # Second loop is to ensure multiple images per page are unpacked
             for n, xref_ext in enumerate(xref_exts):
                 xref, ext = xref_ext
-                future = executor.submit(
-                    jbig2enc.convert_single,
+                yield dict(
                     cwd=fspath(root),
                     infile=img_name(root, xref, ext),
                     outfile=root / f'{prefix}.{n:04d}',
                 )
-                yield future
+
+    def convert_generic(fn, kwargs_dict):
+        return fn(**kwargs_dict)
 
     if options.jbig2_page_group_size > 1:
-        jbig2_futures = jbig2_group_futures
+        jbig2_args = jbig2_group_args
+        jbig2_convert = partial(convert_generic, jbig2enc.convert_group)
     else:
-        jbig2_futures = jbig2_single_futures
+        jbig2_args = jbig2_single_args
+        jbig2_convert = partial(convert_generic, jbig2enc.convert_single)
 
-    with concurrent.futures.ThreadPoolExecutor(max_workers=options.jobs) as executor:
-        futures = jbig2_futures(executor, root, jbig2_groups)
-        with tqdm(
+    exec_progress_pool(
+        use_threads=True,
+        max_workers=options.jobs,
+        tqdm_kwargs=dict(
             total=len(jbig2_groups),
             desc="JBIG2",
             unit='item',
             disable=not options.progress_bar,
-        ) as pbar:
-            for future in concurrent.futures.as_completed(futures):
-                proc = future.result()
-                if proc.stderr:
-                    log.debug(proc.stderr.decode())
-                pbar.update()
+        ),
+        task=jbig2_convert,
+        task_arguments=jbig2_args(root, jbig2_groups),
+    )
 
 
-def convert_to_jbig2(pike, jbig2_groups, root, log, options):
+def convert_to_jbig2(pike, jbig2_groups, root, options):
     """Convert images to JBIG2 and insert into PDF.
 
     When the JBIG2 page group size is > 1 we do several JBIG2 images at once
@@ -307,7 +320,7 @@ def convert_to_jbig2(pike, jbig2_groups, root, log, options):
     and needs no dictionary. Currently this must be lossless JBIG2.
     """
 
-    _produce_jbig2_images(jbig2_groups, root, log, options)
+    _produce_jbig2_images(jbig2_groups, root, options)
 
     for group, xref_exts in jbig2_groups.items():
         prefix = f'group{group:08d}'
@@ -331,7 +344,7 @@ def convert_to_jbig2(pike, jbig2_groups, root, log, options):
             )
 
 
-def transcode_jpegs(pike, jpegs, root, log, options):
+def transcode_jpegs(pike, jpegs, root, options):
     for xref in tqdm(
         jpegs, desc="JPEGs", unit='image', disable=not options.progress_bar
     ):
@@ -354,37 +367,40 @@ def transcode_jpegs(pike, jpegs, root, log, options):
         im_obj.write(compdata.read(), filter=Name.DCTDecode)
 
 
-def transcode_pngs(pike, images, image_name_fn, root, log, options):
+def transcode_pngs(pike, images, image_name_fn, root, options):
     modified = set()
     if options.optimize >= 2:
         png_quality = (
             max(10, options.png_quality - 10),
             min(100, options.png_quality + 10),
         )
-        with concurrent.futures.ThreadPoolExecutor(
-            max_workers=options.jobs
-        ) as executor:
-            futures = []
+
+        def pngquant_args():
             for xref in images:
                 log.debug(image_name_fn(root, xref))
-                futures.append(
-                    executor.submit(
-                        pngquant.quantize,
-                        image_name_fn(root, xref),
-                        png_name(root, xref),
-                        png_quality[0],
-                        png_quality[1],
-                    )
+                yield (
+                    image_name_fn(root, xref),
+                    png_name(root, xref),
+                    png_quality[0],
+                    png_quality[1],
                 )
                 modified.add(xref)
-            with tqdm(
+
+        def pngquant_fn(args):
+            pngquant.quantize(*args)
+
+        exec_progress_pool(
+            use_threads=True,
+            max_workers=options.jobs,
+            tqdm_kwargs=dict(
                 desc="PNGs",
-                total=len(futures),
+                total=len(images),
                 unit='image',
                 disable=not options.progress_bar,
-            ) as pbar:
-                for _future in concurrent.futures.as_completed(futures):
-                    pbar.update()
+            ),
+            task=pngquant_fn,
+            task_arguments=pngquant_args(),
+        )
 
     for xref in modified:
         im_obj = pike.get_object(xref, 0)
@@ -410,12 +426,12 @@ def transcode_pngs(pike, images, image_name_fn, root, log, options):
             )
             continue
         if compdata.type == leptonica.lept.L_FLATE_ENCODE:
-            return rewrite_png(pike, im_obj, compdata, log)
+            rewrite_png(pike, im_obj, compdata)
         elif compdata.type == leptonica.lept.L_G4_ENCODE:
-            return rewrite_png_as_g4(pike, im_obj, compdata, log)
+            rewrite_png_as_g4(pike, im_obj, compdata)
 
 
-def rewrite_png_as_g4(pike, im_obj, compdata, log):
+def rewrite_png_as_g4(pike, im_obj, compdata):
     im_obj.BitsPerComponent = 1
     im_obj.Width = compdata.w
     im_obj.Height = compdata.h
@@ -435,7 +451,7 @@ def rewrite_png_as_g4(pike, im_obj, compdata, log):
     return
 
 
-def rewrite_png(pike, im_obj, compdata, log):
+def rewrite_png(pike, im_obj, compdata):
     # When a PNG is inserted into a PDF, we more or less copy the IDAT section from
     # the PDF and transfer the rest of the PNG headers to PDF image metadata.
     # One thing we have to do is tell the PDF reader whether a predictor was used
@@ -489,7 +505,6 @@ def rewrite_png(pike, im_obj, compdata, log):
 
 
 def optimize(input_file, output_file, context, save_settings):
-    log = context.log
     options = context.options
     if options.optimize == 0:
         safe_symlink(input_file, output_file)
@@ -506,15 +521,15 @@ def optimize(input_file, output_file, context, save_settings):
         root = Path(output_file).parent / 'images'
         root.mkdir(exist_ok=True)
 
-        jpegs, pngs = extract_images_generic(pike, root, log, options)
-        transcode_jpegs(pike, jpegs, root, log, options)
+        jpegs, pngs = extract_images_generic(pike, root, options)
+        transcode_jpegs(pike, jpegs, root, options)
         # if options.optimize >= 2:
         # Try pngifying the jpegs
-        #    transcode_pngs(pike, jpegs, jpg_name, root, log, options)
-        transcode_pngs(pike, pngs, png_name, root, log, options)
+        #    transcode_pngs(pike, jpegs, jpg_name, root, options)
+        transcode_pngs(pike, pngs, png_name, root, options)
 
-        jbig2_groups = extract_images_jbig2(pike, root, log, options)
-        convert_to_jbig2(pike, jbig2_groups, root, log, options)
+        jbig2_groups = extract_images_jbig2(pike, root, options)
+        convert_to_jbig2(pike, jbig2_groups, root, options)
 
         target_file = Path(output_file).with_suffix('.opt.pdf')
         pike.remove_unreferenced_resources()
@@ -542,8 +557,8 @@ def optimize(input_file, output_file, context, save_settings):
 
 
 def main(infile, outfile, level, jobs=1):
-    from tempfile import TemporaryDirectory
-    from shutil import copy
+    from tempfile import TemporaryDirectory  # pylint: disable=import-outside-toplevel
+    from shutil import copy  # pylint: disable=import-outside-toplevel
 
     class OptimizeOptions:
         """Emulate ocrmypdf's options"""
@@ -571,7 +586,7 @@ def main(infile, outfile, level, jobs=1):
     )
 
     with TemporaryDirectory() as td:
-        context = PDFContext(options, td, infile, None)
+        context = PdfContext(options, td, infile, None, None)
         tmpout = Path(td) / 'out.pdf'
         optimize(
             infile,

src/ocrmypdf/pdfinfo/__init__.py

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

src/ocrmypdf/pdfinfo/ghosttext.py

@@ -1,102 +0,0 @@
-# © 2018 James R. Barlow: github.com/jbarlow83
-#
-# This file is part of OCRmyPDF.
-#
-# OCRmyPDF is free software: you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation, either version 3 of the License, or
-# (at your option) any later version.
-#
-# OCRmyPDF is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
-
-import logging
-import re
-import xml.etree.ElementTree as ET
-
-from ..exec import ghostscript
-
-gslog = logging.getLogger()
-
-# Forgive me for I have sinned
-# I am using regular expressions to parse XML. However the XML in this case,
-# generated by Ghostscript, is self-consistent enough to be parseable.
-regex_remove_char_tags = re.compile(
-    br"""
-    <char\b
-    (?:   [^>]   # anything single character but >
-        | \">\"  # special case: trap ">"
-    )*
-    />           # terminate with '/>'
-""",
-    re.VERBOSE,
-)
-
-
-def page_get_textblocks(infile, pageno, xmltext, height):
-    """Get text boxes out of Ghostscript txtwrite xml"""
-
-    root = xmltext
-    if not hasattr(xmltext, 'findall'):
-        return []
-
-    def blocks():
-        for span in root.findall('.//span'):
-            bbox_str = span.attrib['bbox']
-            font_size = span.attrib['size']
-            pts = [int(pt) for pt in bbox_str.split()]
-            pts[1] = pts[1] - int(float(font_size) + 0.5)
-            bbox_topdown = tuple(pts)
-            bb = bbox_topdown
-            bbox_bottomup = (bb[0], height - bb[3], bb[2], height - bb[1])
-            yield bbox_bottomup
-
-    def joined_blocks():
-        prev = None
-        for bbox in blocks():
-            if prev is None:
-                prev = bbox
-            if bbox[1] == prev[1] and bbox[3] == prev[3]:
-                gap = prev[2] - bbox[0]
-                height = abs(bbox[3] - bbox[1])
-                if gap < height:
-                    # Join boxes
-                    prev = (prev[0], prev[1], bbox[2], bbox[3])
-                    continue
-            # yield previously joined bboxes and start anew
-            yield prev
-            prev = bbox
-        if prev is not None:
-            yield prev
-
-    return [block for block in joined_blocks()]
-
-
-def extract_text_xml(infile, pdf, pageno=None, log=gslog):
-    existing_text = ghostscript.extract_text(infile, pageno=None)
-    existing_text = regex_remove_char_tags.sub(b' ', existing_text)
-
-    try:
-        root = ET.fromstringlist([b'<document>\n', existing_text, b'</document>\n'])
-        page_xml = root.findall('page')
-    except ET.ParseError as e:
-        log.error(
-            "An error occurred while attempting to retrieve existing text in "
-            "the input file. Will attempt to continue assuming that there is "
-            "no existing text in the file. The error was:"
-        )
-        log.error(e)
-        page_xml = [None] * len(pdf.pages)
-
-    page_count_difference = len(pdf.pages) - len(page_xml)
-    if page_count_difference != 0:
-        log.error("The number of pages in the input file is inconsistent.")
-        log.error(f"Expected {len(pdf.pages)}, txtwrite says {len(page_xml)}")
-        if page_count_difference > 0:
-            page_xml.extend([None] * page_count_difference)
-    return page_xml

src/ocrmypdf/pdfinfo/info.py

@@ -18,29 +18,29 @@
 
 import logging
 import re
-from collections import namedtuple
+from collections import defaultdict, namedtuple
 from decimal import Decimal
 from enum import Enum
+from functools import partial
 from math import hypot, isclose
-from os import fspath
+from os import PathLike
 from pathlib import Path
 from warnings import warn
 
 import pikepdf
 from pikepdf import PdfMatrix
-from tqdm import tqdm
 
-from ocrmypdf.exceptions import EncryptedPdfError, MissingDependencyError
-
-from . import ghosttext
-from .layout import get_page_analysis, get_text_boxes
+from ocrmypdf._concurrent import exec_progress_pool
+from ocrmypdf.exceptions import EncryptedPdfError
+from ocrmypdf.helpers import Resolution, available_cpu_count
+from ocrmypdf.pdfinfo.layout import get_page_analysis, get_text_boxes
 
 logger = logging.getLogger()
 
 Colorspace = Enum('Colorspace', 'gray rgb cmyk lab icc index sep devn pattern jpeg2000')
 
 Encoding = Enum(
-    'Encoding', 'ccitt jpeg jpeg2000 jbig2 asciihex ascii85 lzw flate ' + 'runlength'
+    'Encoding', 'ccitt jpeg jpeg2000 jbig2 asciihex ascii85 lzw flate runlength'
 )
 
 FRIENDLY_COLORSPACE = {
@@ -98,15 +98,19 @@ XobjectSettings = namedtuple('XobjectSettings', ['name', 'shorthand', 'stack_dep
 InlineSettings = namedtuple('InlineSettings', ['iimage', 'shorthand', 'stack_depth'])
 
 ContentsInfo = namedtuple(
-    'ContentsInfo', ['xobject_settings', 'inline_images', 'found_vector']
+    'ContentsInfo',
+    ['xobject_settings', 'inline_images', 'found_vector', 'found_text', 'name_index'],
 )
 
 TextboxInfo = namedtuple('TextboxInfo', ['bbox', 'is_visible', 'is_corrupt'])
 
 
-class VectorInfo:
-    def __init__(self):
-        pass
+class VectorMarker:
+    pass
+
+
+class TextMarker:
+    pass
 
 
 def _normalize_stack(graphobjs):
@@ -151,10 +155,13 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
     ctm = PdfMatrix(initial_shorthand)
     xobject_settings = []
     inline_images = []
+    name_index = defaultdict(lambda: [])
     found_vector = False
+    found_text = False
     vector_ops = set('S s f F f* B B* b b*'.split())
+    text_showing_ops = set("""TJ Tj " '""".split())
     image_ops = set('BI ID EI q Q Do cm'.split())
-    operator_whitelist = ' '.join(vector_ops | image_ops)
+    operator_whitelist = ' '.join(vector_ops | text_showing_ops | image_ops)
 
     for n, graphobj in enumerate(
         _normalize_stack(
@@ -185,6 +192,7 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
                 name=image_name, shorthand=ctm.shorthand, stack_depth=len(stack)
             )
             xobject_settings.append(settings)
+            name_index[image_name].append(settings)
         elif operator == 'INLINE IMAGE':  # BI/ID/EI are grouped into this
             iimage = operands[0]
             inline = InlineSettings(
@@ -193,11 +201,15 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
             inline_images.append(inline)
         elif operator in vector_ops:
             found_vector = True
+        elif operator in text_showing_ops:
+            found_text = True
 
     return ContentsInfo(
         xobject_settings=xobject_settings,
         inline_images=inline_images,
         found_vector=found_vector,
+        found_text=found_text,
+        name_index=name_index,
     )
 
 
@@ -262,7 +274,7 @@ def _get_dpi(ctm_shorthand, image_size):
     dpi_w = scale_w * 72.0
     dpi_h = scale_h * 72.0
 
-    return dpi_w, dpi_h
+    return Resolution(dpi_w, dpi_h)
 
 
 class ImageInfo:
@@ -303,14 +315,22 @@ class ImageInfo:
         if self._enc == Encoding.jpeg2000:
             self._color = Colorspace.jpeg2000
 
-        self._comp = FRIENDLY_COMP.get(self._color, '?')
+        if self._color == Colorspace.icc:
+            # Check the ICC profile to determine actual colorspace
+            pim_icc = pim.icc
+            if pim_icc.profile.xcolor_space == 'GRAY':
+                self._comp = 1
+            elif pim_icc.profile.xcolor_space == 'CMYK':
+                self._comp = 4
+            else:
+                self._comp = 3
+        else:
+            self._comp = FRIENDLY_COMP.get(self._color, '?')
 
-        # Bit of a hack... infer grayscale if component count is uncertain
-        # but encoding must be monochrome. This happens if a monochrome image
-        # has an ICC profile attached. Better solution would be to examine
-        # the ICC profile.
-        if self._comp == '?' and self._enc in (Encoding.ccitt, Encoding.jbig2):
-            self._comp = FRIENDLY_COMP[Colorspace.gray]
+            # Bit of a hack... infer grayscale if component count is uncertain
+            # but encoding only supports monochrome.
+            if self._comp == '?' and self._enc in (Encoding.ccitt, Encoding.jbig2):
+                self._comp = FRIENDLY_COMP[Colorspace.gray]
 
     @property
     def name(self):
@@ -345,12 +365,8 @@ class ImageInfo:
         return self._enc
 
     @property
-    def xres(self):
-        return _get_dpi(self._shorthand, (self._width, self._height))[0]
-
-    @property
-    def yres(self):
-        return _get_dpi(self._shorthand, (self._width, self._height))[1]
+    def dpi(self):
+        return _get_dpi(self._shorthand, (self._width, self._height))
 
     def __repr__(self):
         class_locals = {
@@ -360,7 +376,7 @@ class ImageInfo:
         }
         return (
             "<ImageInfo '{name}' {type_} {width}x{height} {color} "
-            "{comp} {bpc} {enc} {xres}x{yres}>"
+            "{comp} {bpc} {enc} {dpi}>"
         ).format(**class_locals)
 
 
@@ -411,13 +427,9 @@ def _find_regular_images(container, contentsinfo):
     """
 
     for pdfimage, xobj in _image_xobjects(container):
-
-        # For each image that is drawn on this, check if we drawing the
-        # current image - yes this is O(n^2), but n == 1 almost always
-        for draw in contentsinfo.xobject_settings:
-            if draw.name != xobj:
-                continue
-
+        if xobj not in contentsinfo.name_index:
+            continue
+        for draw in contentsinfo.name_index[xobj]:
             if draw.stack_depth == 0 and _is_unit_square(draw.shorthand):
                 # At least one PDF in the wild (and test suite) draws an image
                 # when the graphics stack depth is 0, meaning that the image
@@ -502,7 +514,9 @@ def _process_content_streams(*, pdf, container, shorthand=None):
     contentsinfo = _interpret_contents(container, initial_shorthand)
 
     if contentsinfo.found_vector:
-        yield VectorInfo()
+        yield VectorMarker()
+    if contentsinfo.found_text:
+        yield TextMarker()
     yield from _find_inline_images(contentsinfo)
     yield from _find_regular_images(container, contentsinfo)
     yield from _find_form_xobject_images(pdf, container, contentsinfo)
@@ -551,7 +565,9 @@ def simplify_textboxes(miner, textbox_getter):
         yield TextboxInfo(box.bbox, visible, corrupt)
 
 
-def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
+def _pdf_get_pageinfo(
+    pdf, pageno: int, infile: PathLike, check_pages, detailed_analysis
+):
     pageinfo = {}
     pageinfo['pageno'] = pageno
     pageinfo['images'] = []
@@ -561,18 +577,18 @@ def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
     width_pt = mediabox[2] - mediabox[0]
     height_pt = mediabox[3] - mediabox[1]
 
-    if xmltext is not None:
-        bboxes = ghosttext.page_get_textblocks(
-            fspath(infile), pageno, xmltext=xmltext, height=height_pt
-        )
-        pageinfo['bboxes'] = bboxes
-    else:
+    check_this_page = pageno in check_pages
+
+    if check_this_page and detailed_analysis:
         pscript5_mode = str(pdf.docinfo.get('/Creator')).startswith('PScript5')
         miner = get_page_analysis(infile, pageno, pscript5_mode)
         pageinfo['textboxes'] = list(simplify_textboxes(miner, get_text_boxes))
         bboxes = (box.bbox for box in pageinfo['textboxes'])
 
-    pageinfo['has_text'] = _page_has_text(bboxes, width_pt, height_pt)
+        pageinfo['has_text'] = _page_has_text(bboxes, width_pt, height_pt)
+    else:
+        pageinfo['textboxes'] = []
+        pageinfo['has_text'] = None  # i.e. "no information"
 
     userunit = page.get('/UserUnit', Decimal(1.0))
     if not isinstance(userunit, Decimal):
@@ -587,59 +603,96 @@ def _pdf_get_pageinfo(pdf, pageno: int, infile, xmltext):
         pageinfo['rotate'] = 0
 
     userunit_shorthand = (userunit, 0, 0, userunit, 0, 0)
-    contentsinfo = [
-        ci
+
+    if check_this_page:
+        pageinfo['has_vector'] = False
+        pageinfo['has_text'] = False
+        pageinfo['images'] = []
         for ci in _process_content_streams(
             pdf=pdf, container=page, shorthand=userunit_shorthand
-        )
-    ]
+        ):
+            if isinstance(ci, VectorMarker):
+                pageinfo['has_vector'] = True
+            elif isinstance(ci, TextMarker):
+                pageinfo['has_text'] = True
+            elif isinstance(ci, ImageInfo):
+                pageinfo['images'].append(ci)
+            else:
+                raise NotImplementedError()
+    else:
+        pageinfo['has_vector'] = None  # i.e. "no information"
+        pageinfo['has_text'] = None
+        pageinfo['images'] = None
 
-    pageinfo['has_vector'] = False
-    if any(isinstance(ci, VectorInfo) for ci in contentsinfo):
-        pageinfo['has_vector'] = True
-
-    pageinfo['images'] = [im for im in contentsinfo if isinstance(im, ImageInfo)]
     if pageinfo['images']:
-        xres = Decimal(max(image.xres for image in pageinfo['images']))
-        yres = Decimal(max(image.yres for image in pageinfo['images']))
-        pageinfo['xres'], pageinfo['yres'] = xres, yres
-        pageinfo['width_pixels'] = int(round(xres * pageinfo['width_inches']))
-        pageinfo['height_pixels'] = int(round(yres * pageinfo['height_inches']))
+        dpi = Resolution(0.0, 0.0).take_max(image.dpi for image in pageinfo['images'])
+        pageinfo['dpi'] = dpi
+        pageinfo['width_pixels'] = int(round(dpi.x * float(pageinfo['width_inches'])))
+        pageinfo['height_pixels'] = int(round(dpi.y * float(pageinfo['height_inches'])))
 
     return pageinfo
 
 
-def _pdf_get_all_pageinfo(infile, detailed_analysis=False, log=None, progbar=False):
-    pdf = pikepdf.open(infile)  # Do not close in this function
-    if pdf.is_encrypted:
-        pdf.close()
-        raise EncryptedPdfError()  # Triggered by encryption with empty passwd
-    if detailed_analysis:
-        pages_xml = None
-    else:
-        pages_xml = ghosttext.extract_text_xml(infile, pdf, pageno=None, log=log)
+worker_pdf = None
 
-    pages = []
-    for n, _ in tqdm(
-        enumerate(pdf.pages),
-        total=len(pdf.pages),
-        desc="Scan",
-        unit='page',
-        disable=not progbar,
-    ):
-        page_xml = pages_xml[n] if pages_xml else None
-        page = PageInfo(pdf, n, infile, page_xml, detailed_analysis)
-        pages.append(page)
 
-    return pages, pdf
+def _pdf_pageinfo_sync_init(infile):
+    global worker_pdf  # pylint: disable=global-statement
+    worker_pdf = pikepdf.open(infile)
+
+
+def _pdf_pageinfo_sync(args):
+    global worker_pdf  # pylint: disable=global-statement
+    pageno, infile, check_pages, detailed_analysis = args
+    page = PageInfo(worker_pdf, pageno, infile, check_pages, detailed_analysis)
+    return page
+
+
+def _pdf_pageinfo_concurrent(
+    pdf, infile, progbar, max_workers, check_pages, detailed_analysis=False
+):
+    pages = [None] * len(pdf.pages)
+
+    def update_pageinfo(result, pbar):
+        page = result
+        pages[page.pageno] = page
+        pbar.update()
+
+    if max_workers is None:
+        max_workers = available_cpu_count()
+
+    total = len(pdf.pages)
+    contexts = ((n, infile, check_pages, detailed_analysis) for n in range(total))
+
+    use_threads = False  # No performance gain if threaded due to GIL
+    n_workers = min(1 + len(pages) // 4, max_workers)
+    if n_workers == 1:
+        # But if we decided on only one worker, there is no point in using
+        # a separate process.
+        use_threads = True
+
+    exec_progress_pool(
+        use_threads=use_threads,
+        max_workers=n_workers,
+        tqdm_kwargs=dict(
+            total=total, desc="Scanning contents", unit='page', disable=not progbar
+        ),
+        task_initializer=partial(_pdf_pageinfo_sync_init, infile),
+        task=_pdf_pageinfo_sync,
+        task_arguments=contexts,
+        task_finished=update_pageinfo,
+    )
+    return pages
 
 
 class PageInfo:
-    def __init__(self, pdf, pageno, infile, xmltext, detailed_analysis=False):
+    def __init__(self, pdf, pageno, infile, check_pages, detailed_analysis=False):
         self._pageno = pageno
         self._infile = infile
-        self._pageinfo = _pdf_get_pageinfo(pdf, pageno, infile, xmltext)
         self._detailed_analysis = detailed_analysis
+        self._pageinfo = _pdf_get_pageinfo(
+            pdf, pageno, infile, check_pages, detailed_analysis
+        )
 
     @property
     def pageno(self):
@@ -669,11 +722,11 @@ class PageInfo:
 
     @property
     def width_pixels(self):
-        return int(round(self.width_inches * self.xres))
+        return int(round(float(self.width_inches) * self.dpi.x))
 
     @property
     def height_pixels(self):
-        return int(round(self.height_inches * self.yres))
+        return int(round(float(self.height_inches) * self.dpi.y))
 
     @property
     def rotation(self):
@@ -703,7 +756,7 @@ class PageInfo:
 
         if 'textboxes' not in self._pageinfo:
             if visible is not None and corrupt is not None:
-                raise NotImplementedError('Ghostscript textboxes cannot be classified')
+                raise NotImplementedError('Incomplete information on textboxes')
             return self._pageinfo['bboxes']
 
         return (
@@ -713,12 +766,8 @@ class PageInfo:
         )
 
     @property
-    def xres(self):
-        return self._pageinfo.get('xres', None)
-
-    @property
-    def yres(self):
-        return self._pageinfo.get('yres', None)
+    def dpi(self):
+        return self._pageinfo.get('dpi', Resolution(0.0, 0.0))
 
     @property
     def userunit(self):
@@ -733,34 +782,45 @@ class PageInfo:
 
     def __repr__(self):
         return (
-            '<PageInfo ' 'pageno={} {}"x{}" rotation={} res={}x{} has_text={}>'
-        ).format(
-            self.pageno,
-            self.width_inches,
-            self.height_inches,
-            self.rotation,
-            self.xres,
-            self.yres,
-            self.has_text,
+            f'<PageInfo '
+            f'pageno={self.pageno} {self.width_inches}"x{self.height_inches}" '
+            f'rotation={self.rotation} dpi={self.dpi} has_text={self.has_text}>'
         )
 
 
 class PdfInfo:
     """Get summary information about a PDF"""
 
-    def __init__(self, infile, detailed_page_analysis=False, log=logger, progbar=False):
+    def __init__(
+        self,
+        infile,
+        detailed_analysis=False,
+        progbar=False,
+        max_workers=None,
+        check_pages=None,
+    ):
         self._infile = infile
-        self._pages, pdf = _pdf_get_all_pageinfo(
-            infile, detailed_page_analysis, log=log, progbar=progbar
-        )
-        self._needs_rendering = pdf.root.get('/NeedsRendering', False)
-        self._has_acroform = False
-        if '/AcroForm' in pdf.root:
-            if len(pdf.root.AcroForm.get('/Fields', [])) > 0:
-                self._has_acroform = True
-            elif '/XFA' in pdf.root.AcroForm:
-                self._has_acroform = True
-        pdf.close()
+        if check_pages is None:
+            check_pages = range(0, 1_000_000_000)
+
+        with pikepdf.open(infile) as pdf:
+            if pdf.is_encrypted:
+                raise EncryptedPdfError()  # Triggered by encryption with empty passwd
+            self._pages = _pdf_pageinfo_concurrent(
+                pdf,
+                infile,
+                progbar,
+                max_workers,
+                check_pages=check_pages,
+                detailed_analysis=detailed_analysis,
+            )
+            self._needs_rendering = pdf.root.get('/NeedsRendering', False)
+            self._has_acroform = False
+            if '/AcroForm' in pdf.root:
+                if len(pdf.root.AcroForm.get('/Fields', [])) > 0:
+                    self._has_acroform = True
+                elif '/XFA' in pdf.root.AcroForm:
+                    self._has_acroform = True
 
     @property
     def pages(self):
@@ -800,15 +860,19 @@ class PdfInfo:
 
 
 def main():
-    import argparse
+    import argparse  # pylint: disable=import-outside-toplevel
+    from pprint import pprint  # pylint: disable=import-outside-toplevel
 
     parser = argparse.ArgumentParser()
     parser.add_argument('infile')
     args = parser.parse_args()
-    info = _pdf_get_all_pageinfo(args.infile)
-    from pprint import pprint
+    pdfinfo = PdfInfo(args.infile)
 
-    pprint(info)
+    pprint(pdfinfo)
+    for page in pdfinfo.pages:
+        pprint(page)
+        for im in page.images:
+            pprint(im)
 
 
 if __name__ == '__main__':

src/ocrmypdf/pdfinfo/layout.py

@@ -25,65 +25,16 @@ import pdfminer.encodingdb
 import pdfminer.pdfdevice
 import pdfminer.pdfinterp
 from pdfminer.converter import PDFLayoutAnalyzer
-from pdfminer.glyphlist import glyphname2unicode
 from pdfminer.layout import LAParams, LTChar, LTPage, LTTextBox
 from pdfminer.pdfdocument import PDFTextExtractionNotAllowed
-from pdfminer.pdffont import PDFFont, PDFSimpleFont, PDFUnicodeNotDefined
+from pdfminer.pdffont import PDFSimpleFont, PDFUnicodeNotDefined
 from pdfminer.pdfpage import PDFPage
 from pdfminer.utils import bbox2str, matrix2str
 
-from ..exceptions import EncryptedPdfError
+from ocrmypdf.exceptions import EncryptedPdfError
 
 STRIP_NAME = re.compile(r'[0-9]+')
 
-#
-#  pdfminer 20181108 patches
-#
-
-if pdfminer.__version__ == '20181108':
-
-    def name2unicode(name):
-        """Fix pdfminer's name2unicode function
-
-        Font cids that are mapped to names of the form /g123 seem to be, by convention
-        characters with no corresponding Unicode entry. These can be subsetted fonts
-        or symbolic fonts. There seems to be no way to map /g123 fonts to Unicode,
-        barring a ToUnicode data structure.
-        """
-        if name in glyphname2unicode:
-            return glyphname2unicode[name]
-        if name.startswith('g') or name.startswith('a'):
-            raise KeyError(name)
-        if name.startswith('uni'):
-            try:
-                return chr(int(name[3:], 16))
-            except ValueError:  # Not hexadecimal
-                raise KeyError(name)
-        m = STRIP_NAME.search(name)
-        if not m:
-            raise KeyError(name)
-        return chr(int(m.group(0)))
-
-    pdfminer.encodingdb.name2unicode = name2unicode
-
-    original_PDFFont_init = PDFFont.__init__
-
-    def PDFFont__init__(self, descriptor, widths, default_width=None):
-        original_PDFFont_init(self, descriptor, widths, default_width)
-        # PDF spec says descent should be negative
-        # A font with a positive descent implies it floats entirely above the
-        # baseline, i.e. it's not really a baseline anymore. I have fonts that
-        # claim a positive descent, but treating descent as positive always seems
-        # to misposition text.
-        if self.descent > 0:
-            self.descent = -self.descent
-
-    PDFFont.__init__ = PDFFont__init__
-
-#
-# end of pdfminer 20181108 patches
-#
-
 
 original_PDFSimpleFont_init = PDFSimpleFont.__init__
 
@@ -269,7 +220,9 @@ class TextPositionTracker(PDFLayoutAnalyzer):
 
 def get_page_analysis(infile, pageno, pscript5_mode):
     rman = pdfminer.pdfinterp.PDFResourceManager(caching=True)
-    dev = TextPositionTracker(rman, laparams=LAParams())
+    dev = TextPositionTracker(
+        rman, laparams=LAParams(all_texts=True, detect_vertical=True)
+    )
     interp = pdfminer.pdfinterp.PDFPageInterpreter(rman, dev)
 
     if pscript5_mode:

src/ocrmypdf/pluginspec.py

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

src/ocrmypdf/quality.py

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

src/ocrmypdf/subprocess.py

@@ -1,4 +1,4 @@
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # This file is part of OCRmyPDF.
 #
@@ -23,29 +23,22 @@ import re
 import shutil
 import sys
 from collections.abc import Mapping
+from contextlib import suppress
+from distutils.version import LooseVersion
 from functools import lru_cache
+from pathlib import Path
 from subprocess import PIPE, STDOUT, CalledProcessError
 from subprocess import run as subprocess_run
 
-from ..exceptions import ExitCode, MissingDependencyError
+from ocrmypdf.exceptions import MissingDependencyError
 
-log = logging.Logger(__name__)
-
-
-def _get_program(args, env=None):
-    program = args[0]
-    test_path = env.get('_OCRMYPDF_TEST_PATH', '')
-    if test_path:
-        program = shutil.which(program, path=test_path)
-    return program
+log = logging.getLogger(__name__)
 
 
 def run(args, *, env=None, **kwargs):
     """Wrapper around subprocess.run()
 
-    The main purpose of this wrapper is to allow us to substitute the main program
-    for a spoof in the test suite. The hidden variable _OCRMYPDF_TEST_PATH replaces
-    the main PATH as a location to check for programs to run.
+    The main purpose of this wrapper is to log subprocess output.
 
     Secondly we have to account for behavioral differences in Windows in particular.
     Creating symbolic links in Windows requires administrator privileges and
@@ -60,31 +53,59 @@ def run(args, *, env=None, **kwargs):
         env = os.environ
 
     # Search in spoof path if necessary
-    program = _get_program(args, env)
-
-    # If we are running a .py on Windows, ensure we call it with this Python
-    # (to support test suite shims)
-    if os.name == 'nt' and program.lower().endswith('.py'):
-        args = [sys.executable, program] + args[1:]
-    else:
-        args = [program] + args[1:]
+    program = args[0]
 
     if os.name == 'nt':
-        paths = os.pathsep.join(os.get_exec_path(env))
-        if not shutil.which(args[0], path=paths):
-            shimmed_path = shim_paths_with_program_files(env)
-            new_args0 = shutil.which(args[0], path=shimmed_path)
-            if new_args0:
-                args[0] = new_args0
+        args = _fix_windows_args(program, args, env)
 
-    log.debug(args)
+    log.debug("Running: %s", args)
+    process_log = log.getChild('subprocess.' + os.path.basename(program))
     if sys.version_info < (3, 7) and os.name == 'nt':
         # Can't use close_fds=True on Windows with Python 3.6 or older
         # https://bugs.python.org/issue19575, etc.
         kwargs['close_fds'] = False
-    return subprocess_run(args, env=env, **kwargs)
+
+    stderr = None
+    try:
+        proc = subprocess_run(args, env=env, **kwargs)
+    except CalledProcessError as e:
+        stderr = getattr(e, 'stderr', None)
+        raise
+    else:
+        stderr = getattr(proc, 'stderr', None)
+    finally:
+        if process_log.isEnabledFor(logging.DEBUG) and stderr:
+            with suppress(AttributeError, UnicodeDecodeError):
+                stderr = stderr.decode('utf-8', 'replace')
+            process_log.debug("stderr = %s", stderr)
+    return proc
 
 
+def _fix_windows_args(program, args, env):
+    """Adjust our desired program and command line arguments for use on Windows"""
+
+    if sys.version_info < (3, 8):
+        # bpo-33617 - Windows needs manual Path -> str conversion
+        args = [os.fspath(arg) for arg in args]
+        program = os.fspath(program)
+
+    # If we are running a .py on Windows, ensure we call it with this Python
+    # (to support test suite shims)
+    if program.lower().endswith('.py'):
+        args = [sys.executable] + args
+
+    paths = os.pathsep.join(os.get_exec_path(env))
+    if not shutil.which(args[0], path=paths):
+        # If the program we want is not on the PATH, add some interesting
+        # locations in %PROGRAMFILES% to the PATH and try again
+        shimmed_path = shim_paths_with_program_files(env)
+        new_args0 = shutil.which(args[0], path=shimmed_path)
+        if new_args0:
+            args[0] = new_args0
+    return args
+
+
+@lru_cache(maxsize=None)
 def get_version(program, *, version_arg='--version', regex=r'(\d+(\.\d+)*)', env=None):
     """Get the version of the specified program"""
     args_prog = [program, version_arg]
@@ -128,24 +149,25 @@ def shim_paths_with_program_files(env=None):
     program_files = env.get('PROGRAMFILES', '')
     if not program_files:
         return env.get('PATH', '')
-    paths = []
-    try:
-        for dirname in os.listdir(program_files):
-            if dirname.lower() == 'tesseract-ocr':
-                paths.append(os.path.join(program_files, dirname))
-            elif dirname.lower() == 'gs':
-                try:
-                    latest_gs = max(
-                        os.listdir(os.path.join(program_files, dirname)),
-                        key=lambda d: float(d[2:]),
-                    )
-                except (FileNotFoundError, NotADirectoryError):
-                    continue
-                paths.append(os.path.join(program_files, dirname, latest_gs, 'bin'))
-    except EnvironmentError:
-        pass
-    paths.extend(path for path in os.get_exec_path(env) if path not in set(paths))
-    return os.pathsep.join(paths)
+
+    def path_walker():
+        for path in Path(program_files).iterdir():
+            if not path.is_dir():
+                continue
+            if path.name.lower() == 'tesseract-ocr':
+                yield path
+            elif path.name.lower() == 'gs':
+                yield from (p for p in path.glob('**/bin') if p.is_dir())
+
+    paths = sorted(
+        (p for p in path_walker()), key=lambda p: (p.name, p.parent.name), reverse=True
+    )
+    paths.extend(
+        Path(str_path)
+        for str_path in os.get_exec_path(env)
+        if Path(str_path) not in set(paths)
+    )
+    return os.pathsep.join(str(p) for p in paths)
 
 
 missing_program = '''
@@ -223,10 +245,10 @@ def _error_trailer(program, package, **kwargs):
 
 
 def _error_missing_program(program, package, required_for, recommended):
-    if required_for:
+    if recommended:
+        log.warning(missing_recommend_program.format(**locals()))
+    elif required_for:
         log.error(missing_optional_program.format(**locals()))
-    elif recommended:
-        log.info(missing_recommend_program.format(**locals()))
     else:
         log.error(missing_program.format(**locals()))
     _error_trailer(**locals())
@@ -248,20 +270,27 @@ def check_external_program(
     need_version,
     required_for=None,
     recommended=False,
-    **kwargs,  # To consume log parameter
 ):
-    if kwargs:
-        if not 'log' in kwargs:
-            log.warning('check_external_program(log=...) is deprecated')
     try:
-        found_version = version_checker()
+        if callable(version_checker):
+            found_version = version_checker()
+        else:
+            found_version = version_checker
     except (CalledProcessError, FileNotFoundError, MissingDependencyError):
         _error_missing_program(program, package, required_for, recommended)
         if not recommended:
             raise MissingDependencyError()
         return
 
-    if found_version < need_version:
+    def remove_leading_v(s):
+        if s.startswith('v'):
+            return s[1:]
+        return s
+
+    found_version = remove_leading_v(found_version)
+    need_version = remove_leading_v(need_version)
+
+    if found_version and LooseVersion(found_version) < LooseVersion(need_version):
         _error_old_version(program, package, need_version, found_version, required_for)
         if not recommended:
             raise MissingDependencyError()

tests/cache/manifest.jsonl

@@ -69,3 +69,4 @@
 {"tesseract_version": "tesseract 4.1.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.5", "argv_slug": "__-l__eng__000002_ocr.png__000002_ocr_tess__pdf__txt", "sourcefile": "resources/3small.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000002_ocr.png", "$TMPDIR/000002_ocr_tess", "pdf", "txt"]}
 {"tesseract_version": "tesseract 4.1.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.5", "argv_slug": "__-l__eng__000001_ocr.png__000001_ocr_tess__pdf__txt", "sourcefile": "resources/3small.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000001_ocr.png", "$TMPDIR/000001_ocr_tess", "pdf", "txt"]}
 {"tesseract_version": "tesseract 4.1.0  leptonica-1.78.0   libgif 5.1.4 : libjpeg 9c : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.0.3 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.5", "argv_slug": "__-l__eng__000003_ocr.png__000003_ocr_tess__pdf__txt", "sourcefile": "resources/3small.pdf", "args": ["-l", "eng", "-c", "textonly_pdf=1", "$TMPDIR/000003_ocr.png", "$TMPDIR/000003_ocr_tess", "pdf", "txt"]}
+{"tesseract_version": "tesseract 4.1.1  leptonica-1.79.0   libgif 5.2.1 : libjpeg 9d : libpng 1.6.37 : libtiff 4.1.0 : zlib 1.2.11 : libwebp 1.1.0 : libopenjp2 2.3.1  Found AVX2  Found AVX  Found FMA  Found SSE ", "platform": "Darwin-18.7.0-x86_64-i386-64bit", "python": "3.7.7", "argv_slug": "__-l__eng__000002_ocr.png__000002_ocr_hocr__hocr__txt", "sourcefile": "resources/multipage.pdf", "args": ["-l", "eng", "$TMPDIR/000002_ocr.png", "$TMPDIR/000002_ocr_hocr", "hocr", "txt"]}

tests/cache/multipage/__-l__eng__000002_ocr.png__000002_ocr_hocr__hocr__txt/hocr.bin

@@ -0,0 +1,30 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
+    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+ <head>
+  <title></title>
+  <meta http-equiv="Content-Type" content="text/html;charset=utf-8"/>
+  <meta name='ocr-system' content='tesseract 4.1.1' />
+  <meta name='ocr-capabilities' content='ocr_page ocr_carea ocr_par ocr_line ocrx_word ocrp_wconf'/>
+ </head>
+ <body>
+  <div class='ocr_page' id='page_1' title='image "/var/folders/2s/7t022mgj0h5cprbq0dtb1ksm0000gn/T/com.github.ocrmypdf.626pjb52/000002_ocr.png"; bbox 0 0 12000 12000; ppageno 0'>
+   <div class='ocr_carea' id='block_1_1' title="bbox 1055 986 5888 1952">
+    <p class='ocr_par' id='par_1_1' lang='eng' title="bbox 1055 986 5888 1952">
+     <span class='ocr_line' id='line_1_1' title="bbox 1056 986 5888 1347; baseline -0.001 -70; x_size 367.5; x_descenders 78.5; x_ascenders 70.5">
+      <span class='ocrx_word' id='word_1_1' title='bbox 1056 991 2984 1277; x_wconf 11'>YOOOxXYOO0O</span>
+      <span class='ocrx_word' id='word_1_2' title='bbox 3132 986 4023 1347; x_wconf 91'>pixels</span>
+      <span class='ocrx_word' id='word_1_3' title='bbox 4160 1001 4447 1277; x_wconf 96'>at</span>
+      <span class='ocrx_word' id='word_1_4' title='bbox 4576 991 5219 1277; x_wconf 93'>GOO</span>
+      <span class='ocrx_word' id='word_1_5' title='bbox 5365 986 5888 1273; x_wconf 92'>DPI</span>
+     </span>
+     <span class='ocr_line' id='line_1_2' title="bbox 1055 1587 3452 1952; baseline 0 -73; x_size 366; x_descenders 78; x_ascenders 70">
+      <span class='ocrx_word' id='word_1_6' title='bbox 1055 1592 1394 1879; x_wconf 37'>oO]</span>
+      <span class='ocrx_word' id='word_1_7' title='bbox 1617 1587 3452 1952; x_wconf 62'>megapixels</span>
+     </span>
+    </p>
+   </div>
+  </div>
+ </body>
+</html>

tests/cache/multipage/__-l__eng__000002_ocr.png__000002_ocr_hocr__hocr__txt/stderr.bin

@@ -0,0 +1 @@
+Tesseract Open Source OCR Engine v4.1.1 with Leptonica

tests/cache/multipage/__-l__eng__000002_ocr.png__000002_ocr_hocr__hocr__txt/txt.bin

@@ -0,0 +1,3 @@
+YOOOxXYOO0O pixels at GOO DPI
+oO] megapixels
+

tests/conftest.py

@@ -15,7 +15,6 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import ast
 import os
 import platform
 import sys
@@ -24,16 +23,12 @@ from subprocess import PIPE, run
 
 import pytest
 
-from ocrmypdf import api, cli
+from ocrmypdf import api, pdfinfo
+from ocrmypdf._exec import unpaper
+from ocrmypdf._plugin_manager import get_parser_options_plugins
 
 pytest_plugins = ['helpers_namespace']
 
-try:
-    from pytest_cov.embed import cleanup_on_sigterm
-except ImportError:
-    pass
-else:
-    cleanup_on_sigterm()
 
 # pylint: disable=E1101
 # pytest.helpers is dynamic so it confuses pylint
@@ -57,7 +52,7 @@ def is_macos():
 def running_in_docker():
     # Docker creates a file named /.dockerenv (newer versions) or
     # /.dockerinit (older) -- this is undocumented, not an offical test
-    return os.path.exists('/.dockerenv') or os.path.exists('/.dockerinit')
+    return Path('/.dockerenv').exists() or Path('/.dockerinit').exists()
 
 
 @pytest.helpers.register
@@ -68,87 +63,17 @@ def running_in_travis():
 @pytest.helpers.register
 def have_unpaper():
     try:
-        from ocrmypdf.exec import unpaper
-
         unpaper.version()
-    except Exception:
+    except Exception:  # pylint: disable=broad-except
         return False
     return True
 
 
-TESTS_ROOT = os.path.abspath(os.path.dirname(__file__))
-SPOOF_PATH = os.path.join(TESTS_ROOT, 'spoof')
-PROJECT_ROOT = os.path.dirname(TESTS_ROOT)
+TESTS_ROOT = Path(__file__).parent.resolve()
+PROJECT_ROOT = TESTS_ROOT
 OCRMYPDF = [sys.executable, '-m', 'ocrmypdf']
 
 
-WINDOWS_SHIM_TEMPLATE = """
-# This is a shim for Windows that has the same effect as a symlink to the target .py
-# file
-import os
-import subprocess
-import sys
-
-args = [sys.executable, {spoofer}, *sys.argv[1:]]
-p = subprocess.run(args, check=False, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
-sys.stdout.buffer.write(p.stdout)
-sys.stderr.buffer.write(p.stderr)
-sys.exit(p.returncode)
-"""
-
-assert ast.parse(WINDOWS_SHIM_TEMPLATE.format(spoofer=repr(r"C:\\Temp\\file.py")))
-
-
-@pytest.helpers.register
-def spoof(tmp_path_factory, **kwargs):
-    """Modify PATH to override subprocess executables
-
-    spoof(program1='replacement', ...)
-
-    Creates temporary directory with symlinks to targets.
-
-    """
-    env = os.environ.copy()
-    slug = '-'.join(v.replace('.py', '') for v in sorted(kwargs.values()))
-    spoofer_base = tmp_path_factory.mktemp('spoofers')
-    tmpdir = Path(spoofer_base / slug)
-    tmpdir.mkdir(parents=True)
-
-    for replace_program, with_spoof in kwargs.items():
-        spoofer = Path(SPOOF_PATH) / with_spoof
-        if os.name != 'nt':
-            spoofer.chmod(0o755)
-            (tmpdir / replace_program).symlink_to(spoofer)
-        else:
-            py_file = WINDOWS_SHIM_TEMPLATE.format(
-                spoofer=repr(os.fspath(spoofer.absolute()))
-            )
-            if replace_program == 'gs':
-                programs = ['gswin64c', 'gswin32c']
-            else:
-                programs = [replace_program]
-            for prog in programs:
-                (tmpdir / f'{prog}.py').write_text(py_file, encoding='utf-8')
-
-    env['_OCRMYPDF_TEST_PATH'] = str(tmpdir) + os.pathsep + env['PATH']
-    if os.name == 'nt':
-        if '.py' not in env['PATHEXT'].lower():
-            raise EnvironmentError("PATHEXT is not configured to support .py")
-    return env
-
-
-@pytest.fixture(scope='session')
-def spoof_tesseract_noop(tmp_path_factory):
-    return spoof(tmp_path_factory, tesseract='tesseract_noop.py')
-
-
-@pytest.fixture(scope='session')
-def spoof_tesseract_cache(tmp_path_factory):
-    if running_in_docker():
-        return os.environ.copy()
-    return spoof(tmp_path_factory, tesseract="tesseract_cache.py")
-
-
 @pytest.fixture
 def resources():
     return Path(TESTS_ROOT) / 'resources'
@@ -179,58 +104,64 @@ def no_outpdf(tmp_path):
 
 
 @pytest.helpers.register
-def check_ocrmypdf(input_file, output_file, *args, env=None):
+def check_ocrmypdf(input_file, output_file, *args):
     """Run ocrmypdf and confirmed that a valid file was created"""
+    args = [str(input_file), str(output_file)] + [
+        str(arg) for arg in args if arg is not None
+    ]
 
-    options = cli.parser.parse_args(
-        [str(input_file), str(output_file)]
-        + [str(arg) for arg in args if arg is not None]
-    )
-    api.check_options(options)
-    if env:
-        options.tesseract_env = env
-        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(input_file)
-    result = api.run_pipeline(options, api=True)
+    _parser, options, plugin_manager = get_parser_options_plugins(args=args)
+    api.check_options(options, plugin_manager)
+    result = api.run_pipeline(options, plugin_manager=plugin_manager, api=True)
 
     assert result == 0
-    assert os.path.exists(str(output_file)), "Output file not created"
-    assert os.stat(str(output_file)).st_size > 100, "PDF too small or empty"
+    assert output_file.exists(), "Output file not created"
+    assert output_file.stat().st_size > 100, "PDF too small or empty"
 
     return output_file
 
 
 @pytest.helpers.register
-def run_ocrmypdf_api(input_file, output_file, *args, env=None):
-    "Run ocrmypdf and let caller deal with results"
+def run_ocrmypdf_api(input_file, output_file, *args):
+    """Run ocrmypdf via API and let caller deal with results
 
-    options = cli.parser.parse_args(
-        [str(input_file), str(output_file)]
-        + [str(arg) for arg in args if arg is not None]
-    )
-    api.check_options(options)
-    if env:
-        options.tesseract_env = env.copy()
-        options.tesseract_env['_OCRMYPDF_TEST_INFILE'] = os.fspath(input_file)
-    if options.tesseract_env:
-        assert all(isinstance(v, (str, bytes)) for v in options.tesseract_env.values())
+    Does not currently have a way to manipulate the PATH except for Tesseract.
+    """
 
-    return api.run_pipeline(options, api=False)
+    args = [str(input_file), str(output_file)] + [
+        str(arg) for arg in args if arg is not None
+    ]
+    _parser, options, plugin_manager = get_parser_options_plugins(args=args)
+
+    api.check_options(options, plugin_manager)
+    return api.run_pipeline(options, plugin_manager=None, api=False)
 
 
 @pytest.helpers.register
-def run_ocrmypdf(input_file, output_file, *args, env=None, universal_newlines=True):
+def run_ocrmypdf(input_file, output_file, *args, universal_newlines=True):
     "Run ocrmypdf and let caller deal with results"
 
-    if env is None:
-        env = os.environ
-
     p_args = (
         OCRMYPDF
         + [str(arg) for arg in args if arg is not None]
         + [str(input_file), str(output_file)]
     )
+
+    # Tell subprocess where to find coverage.py configuration
+    # This has no effect except when coverage is running
+    # Details: https://coverage.readthedocs.io/en/coverage-5.0/subprocess.html
+    coverage_rc = Path(__file__).parent.parent / '.coveragerc'
+    assert coverage_rc.exists()
+    env = os.environ.copy()
+    env['COVERAGE_PROCESS_START'] = os.fspath(coverage_rc)
+
     p = run(
-        p_args, stdout=PIPE, stderr=PIPE, universal_newlines=universal_newlines, env=env
+        p_args,
+        stdout=PIPE,
+        stderr=PIPE,
+        universal_newlines=universal_newlines,
+        env=env,
+        check=False,
     )
     # print(p.stderr)
     return p, p.stdout, p.stderr
@@ -238,8 +169,6 @@ def run_ocrmypdf(input_file, output_file, *args, env=None, universal_newlines=Tr
 
 @pytest.helpers.register
 def first_page_dimensions(pdf):
-    from ocrmypdf import pdfinfo
-
     info = pdfinfo.PdfInfo(pdf)
     page0 = info[0]
     return (page0.width_inches, page0.height_inches)

tests/plugins/gs_feature_elision.py

@@ -1,5 +1,4 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # Permission is hereby granted, free of charge, to any person obtaining a
 # copy of this software and associated documentation files (the
@@ -20,34 +19,31 @@
 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
+from unittest.mock import patch
 
-import os
-import sys
-from subprocess import check_call
-
-from gs import real_ghostscript
-
-"""Replicate one type of Ghostscript feature elision warning during
-PDF/A creation."""
-
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins import ghostscript
+from ocrmypdf.subprocess import run
 
 elision_warning = """GPL Ghostscript 9.20: Setting Overprint Mode to 1
 not permitted in PDF/A-2, overprint mode not set"""
 
 
-def main():
-    if '--version' in sys.argv:
-        print('9.20')
-        print('SPOOFED: ' + os.path.basename(__file__))
-        sys.exit(0)
-    gs_args = ['gs'] + sys.argv[1:]
-    check_call(gs_args)
-
-    if '-sDEVICE=pdfwrite' in sys.argv[1:]:
-        print(elision_warning)
-
-    sys.exit(0)
+def run_append_stderr(*args, **kwargs):
+    proc = run(*args, **kwargs)
+    proc.stderr = b'\n'.join([proc.stderr, elision_warning.encode('utf-8')])
+    return proc
 
 
-if __name__ == '__main__':
-    main()
+@hookimpl
+def generate_pdfa(pdf_pages, pdfmark, output_file, compression, pdf_version, pdfa_part):
+    with patch('ocrmypdf._exec.ghostscript.run', new=run_append_stderr):
+        ghostscript.generate_pdfa(
+            pdf_pages=pdf_pages,
+            pdfmark=pdfmark,
+            output_file=output_file,
+            compression=compression,
+            pdf_version=pdf_version,
+            pdfa_part=pdfa_part,
+        )
+        return output_file

tests/plugins/gs_pdfa_failure.py

@@ -1,5 +1,4 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # Permission is hereby granted, free of charge, to any person obtaining a
 # copy of this software and associated documentation files (the
@@ -20,41 +19,33 @@
 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
-import os
-import sys
+from unittest.mock import patch
 
-from gs import real_ghostscript
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins import ghostscript
+from ocrmypdf.subprocess import run
 
 
-"""Replicate Ghostscript PDF/A conversion failure by suppressing some
-arguments"""
-
-
-def main():
-    if '--version' in sys.argv:
-        print('9.20')
-        print('SPOOFED: ' + os.path.basename(__file__))
-        sys.exit(0)
-
-    # Unless some argument is calling for PDFA generation, forward to
-    # real ghostscript
-    if not any(arg.startswith('-dPDFA') for arg in sys.argv):
-        real_ghostscript(sys.argv)
-        return
-
+def run_rig_args(args, **kwargs):
     # Remove the two arguments that tell ghostscript to create a PDF/A
     # Does not remove the Postscript definition file - not necessary
     # to cause PDF/A creation failure
-    argv = []
-    for arg in sys.argv:
-        if arg.startswith('-dPDFA'):
-            continue
-        elif arg.startswith('-dPDFACompatibilityPolicy'):
-            continue
-        argv.append(arg)
-
-    real_ghostscript(argv)
+    new_args = [
+        arg for arg in args if not arg.startswith('-dPDFA') and not arg.endswith('.ps')
+    ]
+    proc = run(new_args, **kwargs)
+    return proc
 
 
-if __name__ == '__main__':
-    main()
+@hookimpl
+def generate_pdfa(pdf_pages, pdfmark, output_file, compression, pdf_version, pdfa_part):
+    with patch('ocrmypdf._exec.ghostscript.run', new=run_rig_args):
+        ghostscript.generate_pdfa(
+            pdf_pages=pdf_pages,
+            pdfmark=pdfmark,
+            output_file=output_file,
+            compression=compression,
+            pdf_version=pdf_version,
+            pdfa_part=pdfa_part,
+        )
+        return output_file

tests/plugins/gs_raster_failure.py

@@ -1,5 +1,4 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # Permission is hereby granted, free of charge, to any person obtaining a
 # copy of this software and associated documentation files (the
@@ -20,30 +19,40 @@
 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
+from pathlib import Path
+from subprocess import CalledProcessError
+from unittest.mock import patch
 
-import os
-import sys
-
-from gs import real_ghostscript
-
-"""Replicate Ghostscript raster failure while allowing rendering"""
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins import ghostscript
 
 
-def main():
-    if '--version' in sys.argv:
-        print('9.20')
-        print('SPOOFED: ' + os.path.basename(__file__))
-        sys.exit(0)
-
-    # For any rendering calls (device == pdfwrite) call real ghostscript
-    if '-sDEVICE=pdfwrite' in sys.argv:
-        real_ghostscript(sys.argv)
-        return
-
-    # Fail
-    print("ERROR: Ghost story archive not found")
-    sys.exit(1)
+def raise_gs_fail(*args, **kwargs):
+    raise CalledProcessError(
+        1, 'gs', output=b"", stderr=b"ERROR: Ghost story archive not found"
+    )
 
 
-if __name__ == '__main__':
-    main()
+@hookimpl
+def rasterize_pdf_page(
+    input_file,
+    output_file,
+    raster_device,
+    raster_dpi,
+    pageno,
+    page_dpi=None,
+    rotation=None,
+    filter_vector=False,
+) -> Path:
+    with patch('ocrmypdf._exec.ghostscript.run', new=raise_gs_fail):
+        ghostscript.rasterize_pdf_page(
+            input_file=input_file,
+            output_file=output_file,
+            raster_device=raster_device,
+            raster_dpi=raster_dpi,
+            pageno=pageno,
+            page_dpi=page_dpi,
+            rotation=rotation,
+            filter_vector=filter_vector,
+        )
+        return output_file

tests/plugins/gs_render_failure.py

@@ -1,5 +1,4 @@
-#!/usr/bin/env python3
-# © 2016-18 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # Permission is hereby granted, free of charge, to any person obtaining a
 # copy of this software and associated documentation files (the
@@ -20,29 +19,28 @@
 # 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"""
+from subprocess import CalledProcessError
+from unittest.mock import patch
 
-import os
-import sys
-
-from gs import real_ghostscript
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins import ghostscript
 
 
-def main():
-    if '--version' in sys.argv:
-        print('9.20')
-        print('SPOOFED: ' + os.path.basename(__file__))
-        sys.exit(0)
-
-    # For any rasterize calls (device != pdfwrite) call real ghostscript
-    if '-sDEVICE=pdfwrite' not in sys.argv:
-        real_ghostscript(sys.argv)
-        return
-
-    # Fail
-    print("ERROR: Casper is not a friendly ghost")
-    sys.exit(1)
+def raise_gs_fail(*args, **kwargs):
+    raise CalledProcessError(
+        1, 'gs', output=b"", stderr=b"ERROR: Casper is not a friendly ghost"
+    )
 
 
-if __name__ == '__main__':
-    main()
+@hookimpl
+def generate_pdfa(pdf_pages, pdfmark, output_file, compression, pdf_version, pdfa_part):
+    with patch('ocrmypdf._exec.ghostscript.run', new=raise_gs_fail):
+        ghostscript.generate_pdfa(
+            pdf_pages=pdf_pages,
+            pdfmark=pdfmark,
+            output_file=output_file,
+            compression=compression,
+            pdf_version=pdf_version,
+            pdfa_part=pdfa_part,
+        )
+        return output_file

tests/plugins/tesseract_badutf8.py

@@ -0,0 +1,63 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# Permission is hereby granted, free of charge, to any person obtaining a
+# copy of this software and associated documentation files (the
+# "Software"), to deal in the Software without restriction, including
+# without limitation the rights to use, copy, modify, merge, publish,
+# distribute, sublicense, and/or sell copies of the Software, and to
+# permit persons to whom the Software is furnished to do so, subject to
+# the following conditions:
+#
+# The above copyright notice and this permission notice shall be included
+# in all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+"""Tesseract bad utf8
+
+In some cases, some versions of Tesseract can output binary gibberish or data
+that is not UTF-8 compatible, so we are forced to check that we can convert it
+and present it to the user.
+"""
+
+from subprocess import CalledProcessError
+from unittest.mock import patch
+
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins.tesseract_ocr import TesseractOcrEngine
+
+
+def bad_utf8(*args, **kwargs):
+    raise CalledProcessError(
+        1,
+        'tesseract',
+        output=b'\x96\xb3\x8c\xf8\x82\xc8UTF-8\x0a',  # "Invalid UTF-8" in Shift JIS
+        stderr=b"",
+    )
+
+
+class BadUtf8OcrEngine(TesseractOcrEngine):
+    @staticmethod
+    def generate_hocr(input_file, output_hocr, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=bad_utf8):
+            TesseractOcrEngine.generate_hocr(
+                input_file, output_hocr, output_text, options
+            )
+
+    @staticmethod
+    def generate_pdf(input_file, output_pdf, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=bad_utf8):
+            TesseractOcrEngine.generate_pdf(
+                input_file, output_pdf, output_text, options
+            )
+
+
+@hookimpl
+def get_ocr_engine():
+    return BadUtf8OcrEngine()

tests/plugins/tesseract_big_image_error.py

@@ -0,0 +1,61 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# Permission is hereby granted, free of charge, to any person obtaining a
+# copy of this software and associated documentation files (the
+# "Software"), to deal in the Software without restriction, including
+# without limitation the rights to use, copy, modify, merge, publish,
+# distribute, sublicense, and/or sell copies of the Software, and to
+# permit persons to whom the Software is furnished to do so, subject to
+# the following conditions:
+#
+# The above copyright notice and this permission notice shall be included
+# in all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+from subprocess import CalledProcessError
+from unittest.mock import patch
+
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins.tesseract_ocr import TesseractOcrEngine
+
+
+def raise_size_exception(*args, **kwargs):
+    raise CalledProcessError(
+        1,
+        'tesseract',
+        output=b"Image too large: (33830, 14959)\nError during processing.",
+        stderr=b"",
+    )
+
+
+class BigImageErrorOcrEngine(TesseractOcrEngine):
+    @staticmethod
+    def get_orientation(input_file, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=raise_size_exception):
+            return TesseractOcrEngine.get_orientation(input_file, options)
+
+    @staticmethod
+    def generate_hocr(input_file, output_hocr, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=raise_size_exception):
+            TesseractOcrEngine.generate_hocr(
+                input_file, output_hocr, output_text, options
+            )
+
+    @staticmethod
+    def generate_pdf(input_file, output_pdf, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=raise_size_exception):
+            TesseractOcrEngine.generate_pdf(
+                input_file, output_pdf, output_text, options
+            )
+
+
+@hookimpl
+def get_ocr_engine():
+    return BigImageErrorOcrEngine()

tests/plugins/tesseract_cache.py

@@ -1,5 +1,4 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # Permission is hereby granted, free of charge, to any person obtaining a
 # copy of this software and associated documentation files (the
@@ -22,11 +21,10 @@
 
 """Cache output of tesseract to speed up test suite
 
-The cache is keyed by an environment variable that slips the input test file
-from tests/resources/ to us.  The input arguments are slugged into a hideous
-filename that more or less represents them literally.  Joined together, this
-becomes the name of the cache folder.  A few name files like stdout, stderr,
-hocr, pdf, describe the output to reproduce.
+The cache is keyed by by the input test file The input arguments are slugged
+into a hideous filename that more or less represents them literally.  Joined
+together, this becomes the name of the cache folder.  A few name files like
+stdout, stderr, hocr, pdf, describe the output to reproduce.
 
 Changes to tests/resources/ or image processing algorithms don't trigger a
 cache miss.  By design, an input image that varies according to platform
@@ -40,10 +38,7 @@ information about the system that produced the results used when cache was
 generated.  This mainly a log to answer questions about how the files
 were produced.
 
-For performance reasons, especially the slow performance of Tesseract on
-machines with AVX2, the cache is now bundled.
-
-Certain operations are not cached and routed to tesseract directly.
+Certain operations are not cached and routed to Tesseract OCR directly.
 
 Assumes Tesseract 4.0.0-alpha or higher.
 
@@ -51,17 +46,23 @@ Assumes Tesseract 4.0.0-alpha or higher.
 
 import argparse
 import json
-import os
+import logging
 import platform
 import re
 import shutil
-import subprocess
-import sys
+from functools import partial
 from pathlib import Path
+from subprocess import PIPE, CalledProcessError, CompletedProcess
+from unittest.mock import patch
 
-__version__ = subprocess.check_output(
-    ['tesseract', '--version'], stderr=subprocess.STDOUT
-).decode()
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins.tesseract_ocr import TesseractOcrEngine
+from ocrmypdf.subprocess import run
+
+log = logging.getLogger(__name__)
+
+TESTS_ROOT = Path(__file__).resolve().parent.parent
+CACHE_ROOT = TESTS_ROOT / 'cache'
 
 
 parser = argparse.ArgumentParser(
@@ -77,43 +78,15 @@ parser.add_argument('-c', action='append')
 parser.add_argument('--psm', type=int)
 parser.add_argument('--oem', type=int)
 
-TESTS_ROOT = Path(__file__).resolve().parent.parent
-CACHE_ROOT = TESTS_ROOT / 'cache'
-
-
-def real_tesseract():
-    tess_args = ['tesseract'] + sys.argv[1:]
-    os.execvp("tesseract", tess_args)
-    return  # Not reachable
-
-
-def main():
-    if any(
-        opt in sys.argv[1:]
-        for opt in ('--print-parameters', '--list-langs', '--version')
-    ):
-        real_tesseract()  # jump into real tesseract, replacing this process
-
-    # Convert non-standard but supported -psm to --psm
-    sys.argv = ['--psm' if arg == '-psm' else arg for arg in sys.argv]
-
-    if '_OCRMYPDF_TEST_INFILE' not in os.environ:
-        real_tesseract()  # test not properly set up
-    source = os.environ['_OCRMYPDF_TEST_INFILE']  # required
-    args = parser.parse_args()
-
-    cache_disabled = os.environ.get('_OCRMYPDF_CACHE_DISABLED', False)
-
-    if args.imagename == 'stdin':
-        real_tesseract()
 
+def get_cache_folder(source_pdf, run_args, parsed_args):
     def slugs():
         yield ''  # so we don't start with a '-' which makes rm difficult
-        for arg in sys.argv[1:]:
-            if arg == args.imagename:
-                yield Path(args.imagename).name
-            elif arg == args.outputbase:
-                yield Path(args.outputbase).name
+        for arg in run_args[1:]:
+            if arg == parsed_args.imagename:
+                yield Path(parsed_args.imagename).name
+            elif arg == parsed_args.outputbase:
+                yield Path(parsed_args.outputbase).name
             elif arg == '-c' or arg.startswith('textonly'):
                 pass
             else:
@@ -122,18 +95,26 @@ def main():
     argv_slug = '__'.join(slugs())
     argv_slug = argv_slug.replace('/', '___')
 
-    cache_folder = Path(CACHE_ROOT) / Path(source).stem / argv_slug
+    return Path(CACHE_ROOT) / Path(source_pdf).stem / argv_slug
+
+
+def cached_run(options, run_args, **run_kwargs):
+    run_args = [str(arg) for arg in run_args]  # flatten PosixPaths
+    args = parser.parse_args(run_args[1:])
+
+    if args.imagename in ('stdin', '-'):
+        return run(run_args, **run_kwargs)
+
+    source_file = options.input_file
+    cache_folder = get_cache_folder(source_file, run_args, args)
     cache_folder.mkdir(parents=True, exist_ok=True)
 
-    print(f"Tesseract cache folder {cache_folder} - ", end='', file=sys.stderr)
+    log.debug("Using Tesseract cache {cache_folder}")
 
-    if (cache_folder / 'stderr.bin').exists() and not cache_disabled:
-        # Cache hit
-        print("HIT", file=sys.stderr)
+    if (cache_folder / 'stderr.bin').exists():
+        log.debug("Cache HIT")
 
         # Replicate stdout/err
-        sys.stdout.buffer.write((cache_folder / 'stdout.bin').read_bytes())
-        sys.stderr.buffer.write((cache_folder / 'stderr.bin').read_bytes())
         if args.outputbase != 'stdout':
             if not args.configfiles:
                 args.configfiles.append('txt')
@@ -141,25 +122,28 @@ def main():
                 # cp cache -> output
                 tessfile = args.outputbase + '.' + configfile
                 shutil.copy(str(cache_folder / configfile) + '.bin', tessfile)
-        sys.exit(0)
+        return CompletedProcess(
+            args=run_args,
+            returncode=0,
+            stdout=(cache_folder / 'stdout.bin').read_bytes(),
+            stderr=(cache_folder / 'stderr.bin').read_bytes(),
+        )
 
-    # Cache miss
-    print("MISS", file=sys.stderr)
+    log.debug("Cache MISS")
 
-    # Call tesseract
-    print(sys.argv[1:])
-    p = subprocess.run(
-        ['tesseract'] + sys.argv[1:], stdout=subprocess.PIPE, stderr=subprocess.PIPE
-    )
-    sys.stdout.buffer.write(p.stdout)
-    sys.stderr.buffer.write(p.stderr)
-
-    if p.returncode != 0:
-        # Do not cache errors or crashes
-        print("Tesseract error", file=sys.stderr)
-        return p.returncode
+    cache_kwargs = {
+        k: v for k, v in run_kwargs.items() if k not in ('stdout', 'stderr')
+    }
+    assert cache_kwargs['check']
+    try:
+        p = run(run_args, stdout=PIPE, stderr=PIPE, **cache_kwargs)
+    except CalledProcessError as e:
+        log.exception(e)
+        raise  # Pass exception onward
 
+    # Update cache
     (cache_folder / 'stdout.bin').write_bytes(p.stdout)
+    (cache_folder / 'stderr.bin').write_bytes(p.stderr)
 
     if args.outputbase != 'stdout':
         if not args.configfiles:
@@ -172,27 +156,46 @@ def main():
             tessfile = args.outputbase + '.' + configfile
             shutil.copy(tessfile, str(cache_folder / configfile) + '.bin')
 
-    (cache_folder / 'stderr.bin').write_bytes(p.stderr)
-
     manifest = {}
-    manifest['tesseract_version'] = __version__.replace('\n', ' ')
+    manifest['tesseract_version'] = TesseractOcrEngine.version().replace('\n', ' ')
     manifest['platform'] = platform.platform()
     manifest['python'] = platform.python_version()
-    manifest['argv_slug'] = argv_slug
-    manifest['sourcefile'] = str(Path(source).relative_to(TESTS_ROOT))
+    manifest['argv_slug'] = cache_folder.name
+    manifest['sourcefile'] = str(Path(source_file).relative_to(TESTS_ROOT))
 
     def clean_sys_argv():
-        for arg in sys.argv[1:]:
+        for arg in run_args[1:]:
             yield re.sub(r'.*/com.github.ocrmypdf[^/]+[/](.*)', r'$TMPDIR/\1', arg)
 
     manifest['args'] = list(clean_sys_argv())
-
-    # pylint: disable=E1101
     with (Path(CACHE_ROOT) / 'manifest.jsonl').open('a') as f:
         json.dump(manifest, f)
         f.write('\n')
         f.flush()
+    return p
 
 
-if __name__ == '__main__':
-    main()
+class CacheOcrEngine(TesseractOcrEngine):
+    @staticmethod
+    def get_orientation(input_file, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=partial(cached_run, options)):
+            return TesseractOcrEngine.get_orientation(input_file, options)
+
+    @staticmethod
+    def generate_hocr(input_file, output_hocr, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=partial(cached_run, options)):
+            TesseractOcrEngine.generate_hocr(
+                input_file, output_hocr, output_text, options
+            )
+
+    @staticmethod
+    def generate_pdf(input_file, output_pdf, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=partial(cached_run, options)):
+            TesseractOcrEngine.generate_pdf(
+                input_file, output_pdf, output_text, options
+            )
+
+
+@hookimpl
+def get_ocr_engine():
+    return CacheOcrEngine()

tests/plugins/tesseract_crash.py

@@ -0,0 +1,63 @@
+# © 2020 James R. Barlow: github.com/jbarlow83
+#
+# Permission is hereby granted, free of charge, to any person obtaining a
+# copy of this software and associated documentation files (the
+# "Software"), to deal in the Software without restriction, including
+# without limitation the rights to use, copy, modify, merge, publish,
+# distribute, sublicense, and/or sell copies of the Software, and to
+# permit persons to whom the Software is furnished to do so, subject to
+# the following conditions:
+#
+# The above copyright notice and this permission notice shall be included
+# in all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+import signal
+from subprocess import CalledProcessError
+from unittest.mock import patch
+
+from ocrmypdf import hookimpl
+from ocrmypdf.builtin_plugins.tesseract_ocr import TesseractOcrEngine
+
+
+def raise_crash(*args, **kwargs):
+    raise CalledProcessError(
+        128 + signal.SIGABRT,
+        'tesseract',
+        output=b"",
+        stderr=b"libc++abi.dylib: terminating with uncaught exception of type "
+        + b"std::bad_alloc: std::bad_alloc",
+    )
+
+
+class CrashOcrEngine(TesseractOcrEngine):
+    @staticmethod
+    def get_orientation(input_file, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=raise_crash):
+            return TesseractOcrEngine.get_orientation(input_file, options)
+
+    @staticmethod
+    def generate_hocr(input_file, output_hocr, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=raise_crash):
+            TesseractOcrEngine.generate_hocr(
+                input_file, output_hocr, output_text, options
+            )
+
+    @staticmethod
+    def generate_pdf(input_file, output_pdf, output_text, options):
+        with patch('ocrmypdf._exec.tesseract.run', new=raise_crash):
+            TesseractOcrEngine.generate_pdf(
+                input_file, output_pdf, output_text, options
+            )
+
+
+@hookimpl
+def get_ocr_engine():
+    return CrashOcrEngine()

tests/plugins/tesseract_noop.py

@@ -1,5 +1,4 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
+# © 2020 James R. Barlow: github.com/jbarlow83
 #
 # Permission is hereby granted, free of charge, to any person obtaining a
 # copy of this software and associated documentation files (the
@@ -20,7 +19,7 @@
 # TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
 # SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
-"""Tesseract no-op spoof
+"""Tesseract no-op plugin
 
 To quickly run tests where getting OCR output is not necessary.
 
@@ -31,20 +30,10 @@ In 'pdf' mode, convert the image to PDF using another program.
 In orientation check mode, report the orientation is upright.
 """
 
-import sys
-
-import img2pdf
-import PyPDF2 as pypdf
+import pikepdf
 from PIL import Image
 
-VERSION_STRING = '''tesseract 4.0.0
- leptonica-1.77.0
-  libjpeg 9c : libpng 1.6.35 : libtiff 4.0.10 : zlib 1.2.11 : libopenjp2 2.3.0
- Found AVX2
- Found AVX
- Found SSE
-SPOOFED
-'''
+from ocrmypdf import OcrEngine, OrientationConfidence, hookimpl
 
 HOCR_TEMPLATE = '''<?xml version="1.0" encoding="UTF-8"?>
 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
@@ -69,67 +58,49 @@ HOCR_TEMPLATE = '''<?xml version="1.0" encoding="UTF-8"?>
 </html>'''
 
 
-def main():
-    if sys.argv[1] == '--version':
-        print(VERSION_STRING, file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[1] == '--list-langs':
-        print('List of available languages (1):\neng', file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[-2] == '--print-parameters':
-        print("Some parameters", file=sys.stderr)
-        print("textonly_pdf\t1\tSome help text")
-        sys.exit(0)
-    elif sys.argv[-2] == 'hocr':
-        inputf = sys.argv[-4]
-        output = sys.argv[-3]
-        with Image.open(inputf) as im, open(
-            output + '.hocr', 'w', encoding='utf-8'
+class NoopOcrEngine(OcrEngine):
+    @staticmethod
+    def version():
+        return '4.0.0'
+
+    @staticmethod
+    def creator_tag(options):
+        tag = '-PDF' if options.pdf_renderer == 'sandwich' else ''
+        return f"NO-OP {tag} {NoopOcrEngine.version()}"
+
+    def __str__(self):
+        return f"NO-OP {NoopOcrEngine.version()}"
+
+    @staticmethod
+    def languages(options):
+        return {'eng'}
+
+    @staticmethod
+    def get_orientation(input_file, options):
+        return OrientationConfidence(angle=0, confidence=0.0)
+
+    @staticmethod
+    def generate_hocr(input_file, output_hocr, output_text, options):
+        with Image.open(input_file) as im, open(
+            output_hocr, 'w', encoding='utf-8'
         ) as f:
             w, h = im.size
             f.write(HOCR_TEMPLATE.format(str(w), str(h)))
-        with open(output + '.txt', 'w') as f:
+        with open(output_text, 'w') as f:
             f.write('')
-    elif sys.argv[-2] == 'pdf':
-        if 'textonly_pdf=1' in sys.argv:
-            inputf = sys.argv[-4]
-            output = sys.argv[-3]
-            with Image.open(inputf) as im:
-                dpi = im.info['dpi']
-                pagesize = im.size[0] / dpi[0], im.size[1] / dpi[1]
-                ptsize = pagesize[0] * 72, pagesize[1] * 72
 
-            pdf_out = pypdf.PdfFileWriter()
-            pdf_out.addBlankPage(ptsize[0], ptsize[1])
-            with open(output + '.pdf', 'wb') as f:
-                pdf_out.write(f)
-            with open(output + '.txt', 'w') as f:
-                f.write('')
-        else:
-            inputf = sys.argv[-4]
-            output = sys.argv[-3]
-            pdf_bytes = img2pdf.convert([inputf], dpi=300)
-            with open(output + '.pdf', 'wb') as f:
-                f.write(pdf_bytes)
-            with open(output + '.txt', 'w') as f:
-                f.write('')
-    elif sys.argv[-1] == 'stdout':
-        inputf = sys.argv[-2]
-        print(
-            """Orientation: 0
-Orientation in degrees: 0
-Orientation confidence: 100.00
-Script: 1
-Script confidence: 100.00""",
-            file=sys.stderr,
-        )
-    else:
-        print("Spoof doesn't understand arguments", file=sys.stderr)
-        print(sys.argv, file=sys.stderr)
-        sys.exit(1)
-
-    sys.exit(0)
+    @staticmethod
+    def generate_pdf(input_file, output_pdf, output_text, options):
+        with Image.open(input_file) as im:
+            dpi = im.info['dpi']
+            pagesize = im.size[0] / dpi[0], im.size[1] / dpi[1]
+        ptsize = pagesize[0] * 72, pagesize[1] * 72
+        pdf = pikepdf.new()
+        pdf.add_blank_page(page_size=ptsize)
+        pdf.save(output_pdf, static_id=True)
+        output_text.write_text('')
 
 
-if __name__ == '__main__':
-    main()
+@hookimpl
+def get_ocr_engine():
+    return NoopOcrEngine()

tests/spoof/tesseract_badutf8.py

@@ -1,80 +0,0 @@
-#!/usr/bin/env python3
-# © 2017 James R. Barlow: github.com/jbarlow83
-#
-# Permission is hereby granted, free of charge, to any person obtaining a
-# copy of this software and associated documentation files (the
-# "Software"), to deal in the Software without restriction, including
-# without limitation the rights to use, copy, modify, merge, publish,
-# distribute, sublicense, and/or sell copies of the Software, and to
-# permit persons to whom the Software is furnished to do so, subject to
-# the following conditions:
-#
-# The above copyright notice and this permission notice shall be included
-# in all copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
-# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
-# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
-# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
-# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
-# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
-# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
-import sys
-
-"""Tesseract bad utf8 spoof
-
-In 'hocr' mode or 'pdf' mode, return error code 1 and some non-Unicode
-text because tesseract seems to do that in some cases related to
-language pack version mismatches
-
-"""
-
-
-VERSION_STRING = '''tesseract 4.0.0
- leptonica-1.77.0
-  libjpeg 9c : libpng 1.6.35 : libtiff 4.0.10 : zlib 1.2.11 : libopenjp2 2.3.0
- Found AVX2
- Found AVX
- Found SSE
-SPOOFED
-'''
-
-# Japanese "Invalid UTF-8" encoded in Shift JIS
-BAD_UTF8 = b'\x96\xb3\x8c\xf8\x82\xc8UTF-8\x0a'
-
-
-def main():
-    if sys.argv[1] == '--version':
-        print(VERSION_STRING, file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[1] == '--list-langs':
-        print('List of available languages (1):\neng', file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[-2] == '--print-parameters':
-        print("Some parameters", file=sys.stderr)
-        print("textonly_pdf\t1\tSome help text")
-        sys.exit(0)
-    elif sys.argv[-2] in ('hocr', 'pdf'):
-        sys.stdout.buffer.write(BAD_UTF8)
-        sys.exit(1)
-    elif sys.argv[-1] == 'stdout':
-        # input file is at sys.argv[-2] but we don't look at it
-        print(
-            """Orientation: 0
-Orientation in degrees: 0
-Orientation confidence: 100.00
-Script: 1
-Script confidence: 100.00""",
-            file=sys.stderr,
-        )
-    else:
-        print("Spoof doesn't understand arguments", file=sys.stderr)
-        print(sys.argv, file=sys.stderr)
-        sys.exit(1)
-
-    sys.exit(0)
-
-
-if __name__ == '__main__':
-    main()

tests/spoof/tesseract_big_image_error.py

@@ -1,77 +0,0 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
-#
-# Permission is hereby granted, free of charge, to any person obtaining a
-# copy of this software and associated documentation files (the
-# "Software"), to deal in the Software without restriction, including
-# without limitation the rights to use, copy, modify, merge, publish,
-# distribute, sublicense, and/or sell copies of the Software, and to
-# permit persons to whom the Software is furnished to do so, subject to
-# the following conditions:
-#
-# The above copyright notice and this permission notice shall be included
-# in all copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
-# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
-# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
-# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
-# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
-# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
-# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
-
-import sys
-
-VERSION_STRING = '''tesseract 4.0.0
- leptonica-1.77.0
-  libjpeg 9c : libpng 1.6.35 : libtiff 4.0.10 : zlib 1.2.11 : libopenjp2 2.3.0
- Found AVX2
- Found AVX
- Found SSE
-SPOOFED: return error claiming image too big
-'''
-
-"""Simulates an error of Tesseract failing on attempts to process large images
-
-"""
-
-
-def main():
-    if sys.argv[1] == '--version':
-        print(VERSION_STRING, file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[1] == '--list-langs':
-        print('List of available languages (1):\neng\n', file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[-2] == '--print-parameters':
-        print('A parameter list would go here\ntextonly_pdf 0\n', file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[-2] == 'hocr':
-        print(
-            "Image too large: (33830, 14959)\n" "Error during processing.",
-            file=sys.stderr,
-        )
-        sys.exit(1)
-    elif sys.argv[-2] == 'pdf':
-        print(
-            "Image too large: (33830, 14959)\n" "Error during processing.",
-            file=sys.stderr,
-        )
-        sys.exit(1)
-    elif sys.argv[-1] == 'stdout':
-        print(
-            "Image too large: (33830, 14959)\n" "Error during processing.",
-            file=sys.stderr,
-        )
-        sys.exit(1)
-    else:
-        print("Spoof doesn't understand arguments", file=sys.stderr)
-        print(sys.argv, file=sys.stderr)
-        sys.exit(1)
-
-    sys.exit(0)
-
-
-if __name__ == '__main__':
-    main()

tests/spoof/tesseract_crash.py

@@ -1,78 +0,0 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
-#
-# Permission is hereby granted, free of charge, to any person obtaining a
-# copy of this software and associated documentation files (the
-# "Software"), to deal in the Software without restriction, including
-# without limitation the rights to use, copy, modify, merge, publish,
-# distribute, sublicense, and/or sell copies of the Software, and to
-# permit persons to whom the Software is furnished to do so, subject to
-# the following conditions:
-#
-# The above copyright notice and this permission notice shall be included
-# in all copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
-# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
-# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
-# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
-# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
-# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
-# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
-import signal
-import sys
-
-VERSION_STRING = '''tesseract 4.0.0
- leptonica-1.77.0
-  libjpeg 9c : libpng 1.6.35 : libtiff 4.0.10 : zlib 1.2.11 : libopenjp2 2.3.0
- Found AVX2
- Found AVX
- Found SSE
-SPOOFED: CRASH ON OCR or --psm 0
-'''
-
-"""Simulates a Tesseract crash when asked to run OCR
-
-It isn't strictly necessary to crash the process and that has unwanted
-side effects like triggering core dumps or error reporting, logging and such.
-It's enough to dump some text to stderr and return an error code.
-
-Follows the POSIX(?) convention of returning 128 + signal number.
-
-"""
-
-
-def main():
-    if sys.argv[1] == '--version':
-        print(VERSION_STRING, file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[1] == '--list-langs':
-        print('List of available languages (1):\neng', file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[-2] == '--print-parameters':
-        print('A parameter list would go here\ntextonly_pdf 0\n', file=sys.stderr)
-        sys.exit(0)
-    elif sys.argv[-2] == 'hocr':
-        print("KABOOM! Tesseract failed for some reason", file=sys.stderr)
-        sys.exit(128 + signal.SIGSEGV)
-    elif sys.argv[-2] == 'pdf':
-        print("KABOOM! Tesseract failed for some reason", file=sys.stderr)
-        sys.exit(128 + signal.SIGSEGV)
-    elif sys.argv[-1] == 'stdout':
-        print(
-            "libc++abi.dylib: terminating with uncaught exception of type "
-            "std::bad_alloc: std::bad_alloc",
-            file=sys.stderr,
-        )
-        sys.exit(128 + signal.SIGABRT)
-    else:
-        print("Spoof doesn't understand arguments", file=sys.stderr)
-        print(sys.argv, file=sys.stderr)
-        sys.exit(1)
-
-    sys.exit(0)
-
-
-if __name__ == '__main__':
-    main()

tests/spoof/unpaper_oldversion.py

@@ -1,37 +0,0 @@
-#!/usr/bin/env python3
-# © 2016 James R. Barlow: github.com/jbarlow83
-#
-# Permission is hereby granted, free of charge, to any person obtaining a
-# copy of this software and associated documentation files (the
-# "Software"), to deal in the Software without restriction, including
-# without limitation the rights to use, copy, modify, merge, publish,
-# distribute, sublicense, and/or sell copies of the Software, and to
-# permit persons to whom the Software is furnished to do so, subject to
-# the following conditions:
-#
-# The above copyright notice and this permission notice shall be included
-# in all copies or substantial portions of the Software.
-#
-# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
-# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
-# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
-# IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
-# CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
-# TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
-# SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
-
-
-import sys
-
-
-def main():
-    if sys.argv[1] == '--version':
-        print('0.5')
-        sys.exit(0)
-
-    print("Only supports --version")
-    sys.exit(1)
-
-
-if __name__ == '__main__':
-    main()

tests/test_acroform.py

@@ -15,6 +15,8 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
+import logging
+
 import pytest
 
 import ocrmypdf
@@ -30,4 +32,11 @@ def acroform(resources):
 def test_acroform_and_redo(acroform, caplog, no_outpdf):
     with pytest.raises(ocrmypdf.exceptions.InputFileError):
         check_ocrmypdf(acroform, no_outpdf, '--redo-ocr')
-        assert '--redo-ocr is not currently possible' in caplog.text
+    assert '--redo-ocr is not currently possible' in caplog.text
+
+
+def test_acroform_message(acroform, caplog, outpdf):
+    caplog.set_level(logging.INFO)
+    check_ocrmypdf(acroform, outpdf, '--plugin', 'tests/plugins/tesseract_noop.py')
+    assert 'fillable form' in caplog.text
+    assert '--force-ocr' in caplog.text

tests/test_api.py

@@ -0,0 +1,68 @@
+# © 2019 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+from io import StringIO
+
+import pytest
+from tqdm import tqdm
+
+import ocrmypdf
+
+
+def test_raw_console():
+    bio = StringIO()
+    tqconsole = ocrmypdf.api.TqdmConsole(file=bio)
+    tqconsole.write("Test")
+    tqconsole.flush()
+    assert "Test" in bio.getvalue()
+
+
+def test_tqdm_console():
+    log = logging.getLogger()
+    log.setLevel(logging.INFO)
+
+    formatter = logging.Formatter('%(message)s')
+
+    bio = StringIO()
+    console = logging.StreamHandler(ocrmypdf.api.TqdmConsole(file=bio))
+    console.setFormatter(formatter)
+
+    log.addHandler(console)
+
+    def before_pbar(message):
+        # Ensure that log messages appear before the progress bar, even when
+        # printed after the progress bar updates.
+        v = bio.getvalue()
+        pbar_start_marker = '|#'
+        return v.index(message) < v.index(pbar_start_marker)
+
+    with tqdm(total=2, file=bio, disable=False) as pbar:
+        pbar.update()
+        msg = "1/2 above progress bar"
+        log.info(msg)
+        assert before_pbar(msg)
+
+    log.info("done")
+    assert not before_pbar("done")
+
+
+def test_language_list():
+    with pytest.raises(
+        (ocrmypdf.exceptions.InputFileError, ocrmypdf.exceptions.MissingDependencyError)
+    ):
+        ocrmypdf.ocr('doesnotexist.pdf', '_.pdf', language=['eng', 'deu'])

tests/test_check_pdf.py

@@ -17,9 +17,9 @@
 
 import pytest
 
-import ocrmypdf.exec.qpdf as qpdf
+from ocrmypdf.helpers import check_pdf
 
 
-def test_qpdf_error(resources):
-    assert qpdf.check(resources / 'blank.pdf')
-    assert not qpdf.check(__file__)
+def test_pdf_error(resources):
+    assert check_pdf(resources / 'blank.pdf')
+    assert not check_pdf(__file__)

tests/test_ghostscript.py

@@ -22,123 +22,117 @@ import pikepdf
 import pytest
 from PIL import Image
 
+from ocrmypdf._exec.ghostscript import rasterize_pdf
 from ocrmypdf.exceptions import ExitCode
-from ocrmypdf.exec.ghostscript import rasterize_pdf
+from ocrmypdf.helpers import Resolution
 
-check_ocrmypdf = pytest.helpers.check_ocrmypdf
-run_ocrmypdf = pytest.helpers.run_ocrmypdf
-run_ocrmypdf_api = pytest.helpers.run_ocrmypdf_api
-spoof = pytest.helpers.spoof
+check_ocrmypdf = pytest.helpers.check_ocrmypdf  # pylint: disable=no-member
+run_ocrmypdf = pytest.helpers.run_ocrmypdf  # pylint: disable=no-member
+run_ocrmypdf_api = pytest.helpers.run_ocrmypdf_api  # pylint: disable=no-member
 
-
-@pytest.fixture(scope='session')
-def spoof_no_tess_gs_render_fail(tmp_path_factory):
-    return spoof(
-        tmp_path_factory, tesseract='tesseract_noop.py', gs='gs_render_failure.py'
-    )
-
-
-@pytest.fixture(scope='session')
-def spoof_no_tess_gs_raster_fail(tmp_path_factory):
-    return spoof(
-        tmp_path_factory, tesseract='tesseract_noop.py', gs='gs_raster_failure.py'
-    )
-
-
-@pytest.fixture(scope='session')
-def spoof_no_tess_no_pdfa(tmp_path_factory):
-    return spoof(
-        tmp_path_factory, tesseract='tesseract_noop.py', gs='gs_pdfa_failure.py'
-    )
-
-
-@pytest.fixture(scope='session')
-def spoof_no_tess_pdfa_warning(tmp_path_factory):
-    return spoof(
-        tmp_path_factory, tesseract='tesseract_noop.py', gs='gs_feature_elision.py'
-    )
+# pylint: disable=redefined-outer-name
 
 
 @pytest.fixture
-def linn(resources):
-    path = resources / 'linn.pdf'
+def francais(resources):
+    path = resources / 'francais.pdf'
     return path, pikepdf.open(path)
 
 
-def test_rasterize_size(linn, outdir, caplog):
-    path, pdf = linn
+def test_rasterize_size(francais, outdir):
+    path, pdf = francais
     page_size_pts = (pdf.pages[0].MediaBox[2], pdf.pages[0].MediaBox[3])
     assert pdf.pages[0].MediaBox[0] == pdf.pages[0].MediaBox[1] == 0
     page_size = (page_size_pts[0] / Decimal(72), page_size_pts[1] / Decimal(72))
-    target_size = Decimal('200.0'), Decimal('150.0')
-    target_dpi = 42.0, 4242.0
+    target_size = Decimal('50.0'), Decimal('30.0')
+    forced_dpi = Resolution(42.0, 4242.0)
 
-    log = logging.getLogger()
     rasterize_pdf(
         path,
         outdir / 'out.png',
-        target_size[0] / page_size[0],
-        target_size[1] / page_size[1],
         raster_device='pngmono',
-        log=log,
-        page_dpi=target_dpi,
+        raster_dpi=Resolution(
+            target_size[0] / page_size[0], target_size[1] / page_size[1]
+        ),
+        page_dpi=forced_dpi,
     )
 
     with Image.open(outdir / 'out.png') as im:
         assert im.size == target_size
-        assert im.info['dpi'] == target_dpi
+        assert im.info['dpi'] == forced_dpi
 
 
-def test_rasterize_rotated(linn, outdir, caplog):
-    path, pdf = linn
+def test_rasterize_rotated(francais, outdir, caplog):
+    path, pdf = francais
     page_size_pts = (pdf.pages[0].MediaBox[2], pdf.pages[0].MediaBox[3])
     assert pdf.pages[0].MediaBox[0] == pdf.pages[0].MediaBox[1] == 0
     page_size = (page_size_pts[0] / Decimal(72), page_size_pts[1] / Decimal(72))
-    target_size = Decimal('200.0'), Decimal('150.0')
-    target_dpi = 42.0, 4242.0
+    target_size = Decimal('50.0'), Decimal('30.0')
+    forced_dpi = Resolution(42.0, 4242.0)
 
-    log = logging.getLogger()
     caplog.set_level(logging.DEBUG)
     rasterize_pdf(
         path,
         outdir / 'out.png',
-        target_size[0] / page_size[0],
-        target_size[1] / page_size[1],
         raster_device='pngmono',
-        log=log,
-        page_dpi=target_dpi,
+        raster_dpi=Resolution(
+            target_size[0] / page_size[0], target_size[1] / page_size[1]
+        ),
+        page_dpi=forced_dpi,
         rotation=90,
     )
 
     with Image.open(outdir / 'out.png') as im:
         assert im.size == (target_size[1], target_size[0])
-        assert im.info['dpi'] == (target_dpi[1], target_dpi[0])
+        assert im.info['dpi'] == (forced_dpi[1], forced_dpi[0])
 
 
-def test_gs_render_failure(spoof_no_tess_gs_render_fail, resources, outpdf):
-    p, out, err = run_ocrmypdf(
-        resources / 'blank.pdf', outpdf, env=spoof_no_tess_gs_render_fail
+def test_gs_render_failure(resources, outpdf):
+    p, _out, err = run_ocrmypdf(
+        resources / 'blank.pdf',
+        outpdf,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
+        '--plugin',
+        'tests/plugins/gs_render_failure.py',
     )
-    print(err)
+    assert 'Casper is not a friendly ghost' in err
     assert p.returncode == ExitCode.child_process_error
 
 
-def test_gs_raster_failure(spoof_no_tess_gs_raster_fail, resources, outpdf):
-    p, out, err = run_ocrmypdf(
-        resources / 'ccitt.pdf', outpdf, env=spoof_no_tess_gs_raster_fail
+def test_gs_raster_failure(resources, outpdf):
+    p, _out, err = run_ocrmypdf(
+        resources / 'francais.pdf',
+        outpdf,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
+        '--plugin',
+        'tests/plugins/gs_raster_failure.py',
     )
-    print(err)
+    assert 'Ghost story archive not found' in err
     assert p.returncode == ExitCode.child_process_error
 
 
-def test_ghostscript_pdfa_failure(spoof_no_tess_no_pdfa, resources, outpdf):
-    p, out, err = run_ocrmypdf(
-        resources / 'ccitt.pdf', outpdf, env=spoof_no_tess_no_pdfa
+def test_ghostscript_pdfa_failure(resources, outpdf):
+    p, _out, _err = run_ocrmypdf(
+        resources / 'francais.pdf',
+        outpdf,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
+        '--plugin',
+        'tests/plugins/gs_pdfa_failure.py',
     )
     assert (
         p.returncode == ExitCode.pdfa_conversion_failed
     ), "Unexpected return when PDF/A fails"
 
 
-def test_ghostscript_feature_elision(spoof_no_tess_pdfa_warning, resources, outpdf):
-    check_ocrmypdf(resources / 'ccitt.pdf', outpdf, env=spoof_no_tess_pdfa_warning)
+def test_ghostscript_feature_elision(resources, outpdf):
+    check_ocrmypdf(
+        resources / 'francais.pdf',
+        outpdf,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
+        '--plugin',
+        'tests/plugins/gs_feature_elision.py',
+    )

tests/test_graft.py

@@ -15,7 +15,6 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import os
 from unittest.mock import patch
 
 import pikepdf
@@ -25,25 +24,30 @@ import ocrmypdf
 
 
 def test_no_glyphless_graft(resources, outdir):
-    pdf = pikepdf.open(resources / 'francais.pdf')
-    pdf_aspect = pikepdf.open(resources / 'aspect.pdf')
-    pdf_cmyk = pikepdf.open(resources / 'cmyk.pdf')
-    pdf.pages.extend(pdf_aspect.pages)
-    pdf.pages.extend(pdf_cmyk.pages)
-    pdf.save(outdir / 'test.pdf')
+    with pikepdf.open(resources / 'francais.pdf') as pdf, pikepdf.open(
+        resources / 'aspect.pdf'
+    ) as pdf_aspect, pikepdf.open(resources / 'cmyk.pdf') as pdf_cmyk:
+        pdf.pages.extend(pdf_aspect.pages)
+        pdf.pages.extend(pdf_cmyk.pages)
+        pdf.save(outdir / 'test.pdf')
 
     with patch('ocrmypdf._graft.MAX_REPLACE_PAGES', 2):
         ocrmypdf.ocr(
-            outdir / 'test.pdf', outdir / 'out.pdf', deskew=True, tesseract_timeout=0
+            outdir / 'test.pdf',
+            outdir / 'out.pdf',
+            deskew=True,
+            tesseract_timeout=0,
+            force_ocr=True,
         )
+    # This test needs asserts
 
 
 def test_links(resources, outpdf):
     ocrmypdf.ocr(
         resources / 'link.pdf', outpdf, redo_ocr=True, oversample=200, output_type='pdf'
     )
-    pdf = pikepdf.open(outpdf)
-    p1 = pdf.pages[0]
-    p2 = pdf.pages[1]
-    assert p1.Annots[0].A.D[0].objgen == p2.objgen
-    assert p2.Annots[0].A.D[0].objgen == p1.objgen
+    with pikepdf.open(outpdf) as pdf:
+        p1 = pdf.pages[0]
+        p2 = pdf.pages[1]
+        assert p1.Annots[0].A.D[0].objgen == p2.objgen
+        assert p2.Annots[0].A.D[0].objgen == p1.objgen

tests/test_helpers.py

@@ -0,0 +1,115 @@
+# © 2019 James R. Barlow: github.com/jbarlow83
+#
+# This file is part of OCRmyPDF.
+#
+# OCRmyPDF is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# OCRmyPDF is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
+
+import logging
+import multiprocessing
+import os
+from unittest.mock import MagicMock
+
+import pytest
+
+import ocrmypdf.helpers as helpers
+from ocrmypdf.subprocess import shim_paths_with_program_files
+
+
+class TestSafeSymlink:
+    def test_safe_symlink_link_self(self, tmp_path, caplog):
+        helpers.safe_symlink(tmp_path / 'self', tmp_path / 'self')
+        assert caplog.record_tuples[0][1] == logging.WARNING
+
+    def test_safe_symlink_overwrite(self, tmp_path):
+        (tmp_path / 'regular_file').touch()
+        with pytest.raises(FileExistsError):
+            helpers.safe_symlink(tmp_path / 'input', tmp_path / 'regular_file')
+
+    def test_safe_symlink_relink(self, tmp_path):
+        (tmp_path / 'regular_file_a').touch()
+        (tmp_path / 'regular_file_b').write_bytes(b'ABC')
+        (tmp_path / 'link').symlink_to(tmp_path / 'regular_file_a')
+        helpers.safe_symlink(tmp_path / 'regular_file_b', tmp_path / 'link')
+        assert (tmp_path / 'link').samefile(tmp_path / 'regular_file_b') or (
+            tmp_path / 'link'
+        ).read_bytes() == b'ABC'
+
+
+def test_no_cpu_count(monkeypatch):
+    def cpu_count_raises():
+        raise NotImplementedError()
+
+    monkeypatch.setattr(multiprocessing, 'cpu_count', cpu_count_raises)
+    with pytest.warns(expected_warning=UserWarning):
+        assert helpers.available_cpu_count() == 1
+
+
+def test_deprecated():
+    @helpers.deprecated
+    def old_function():
+        return 42
+
+    with pytest.deprecated_call():
+        assert old_function() == 42
+
+
+class TestFileIsWritable:
+    @pytest.fixture
+    def non_existent(self, tmp_path):
+        return tmp_path / 'nofile'
+
+    @pytest.fixture
+    def basic_file(self, tmp_path):
+        basic = tmp_path / 'basic'
+        basic.touch()
+        return basic
+
+    def test_plain(self, non_existent):
+        assert helpers.is_file_writable(non_existent)
+
+    def test_symlink_loop(self, tmp_path):
+        loop = tmp_path / 'loop'
+        loop.symlink_to(loop)
+        assert not helpers.is_file_writable(loop)
+
+    def test_chmod(self, basic_file):
+        assert helpers.is_file_writable(basic_file)
+        basic_file.chmod(0o400)
+        assert not helpers.is_file_writable(basic_file)
+        basic_file.chmod(0o000)
+        assert not helpers.is_file_writable(basic_file)
+
+    def test_permission_error(self, basic_file):
+        pathmock = MagicMock(spec_set=basic_file)
+        pathmock.is_symlink.return_value = False
+        pathmock.exists.return_value = True
+        pathmock.is_file.side_effect = PermissionError
+        assert not helpers.is_file_writable(pathmock)
+
+
+def test_shim_paths(tmp_path):
+    progfiles = tmp_path / 'Program Files'
+    progfiles.mkdir()
+    (progfiles / 'tesseract-ocr').mkdir()
+    (progfiles / 'gs' / '9.51' / 'bin').mkdir(parents=True)
+    (progfiles / 'gs' / '9.52' / 'bin').mkdir(parents=True)
+    syspath = tmp_path / 'bin'
+    env = {'PROGRAMFILES': str(progfiles), 'PATH': str(syspath)}
+
+    result_str = shim_paths_with_program_files(env=env)
+    results = result_str.split(os.pathsep)
+    assert results[0].endswith('tesseract-ocr')
+    assert results[1].endswith(os.path.join('gs', '9.52', 'bin'))
+    assert results[2].endswith(os.path.join('gs', '9.51', 'bin'))
+    assert results[3] == str(syspath)

tests/test_hocrtransform.py

@@ -15,14 +15,12 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-from pathlib import Path
-
 import pytest
 from PIL import Image
 
 from ocrmypdf import hocrtransform
-from ocrmypdf.exec import qpdf
-from ocrmypdf.exec.tesseract import HOCR_TEMPLATE
+from ocrmypdf._exec.tesseract import HOCR_TEMPLATE
+from ocrmypdf.helpers import check_pdf
 
 # pylint: disable=redefined-outer-name
 
@@ -41,6 +39,6 @@ def test_mono_image(blank_hocr, outdir):
     im.save(outdir / 'mono.tif', format='TIFF')
 
     hocr = hocrtransform.HocrTransform(str(blank_hocr), 300)
-    hocr.to_pdf(str(outdir / 'mono.pdf'), imageFileName=str(outdir / 'mono.tif'))
+    hocr.to_pdf(str(outdir / 'mono.pdf'), image_filename=str(outdir / 'mono.tif'))
 
-    qpdf.check(str(outdir / 'mono.pdf'))
+    check_pdf(str(outdir / 'mono.pdf'))

tests/test_image_input.py

@@ -33,9 +33,14 @@ def baiona(resources):
     return Image.open(resources / 'baiona_gray.png')
 
 
-def test_image_to_pdf(spoof_tesseract_noop, resources, outpdf):
+def test_image_to_pdf(resources, outpdf):
     check_ocrmypdf(
-        resources / 'crom.png', outpdf, '--image-dpi', '200', env=spoof_tesseract_noop
+        resources / 'crom.png',
+        outpdf,
+        '--image-dpi',
+        '200',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
 
@@ -77,7 +82,7 @@ def test_img2pdf_fails(resources, no_outpdf):
         assert rc == ocrmypdf.ExitCode.input_file
 
 
-def test_jpeg_in_jpeg_out(resources, outpdf, spoof_tesseract_noop):
+def test_jpeg_in_jpeg_out(resources, outpdf):
     check_ocrmypdf(
         resources / 'congress.jpg',
         outpdf,
@@ -86,7 +91,8 @@ def test_jpeg_in_jpeg_out(resources, outpdf, spoof_tesseract_noop):
         '--output-type',
         'pdf',  # specifically check pdf because Ghostscript may convert to JPEG
         '--remove-background',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     with pikepdf.open(outpdf) as pdf:
         assert next(pdf.pages[0].images.values()).Filter == pikepdf.Name.DCTDecode

tests/test_main.py

@@ -28,10 +28,11 @@ import pytest
 from PIL import Image
 
 import ocrmypdf
+from ocrmypdf._exec import ghostscript, tesseract
 from ocrmypdf.exceptions import ExitCode, MissingDependencyError
-from ocrmypdf.exec import ghostscript, qpdf, tesseract
 from ocrmypdf.pdfa import file_claims_pdfa
 from ocrmypdf.pdfinfo import Colorspace, Encoding, PdfInfo
+from ocrmypdf.subprocess import get_version
 
 # pytest.helpers is dynamic
 # pylint: disable=no-member,redefined-outer-name
@@ -39,28 +40,19 @@ from ocrmypdf.pdfinfo import Colorspace, Encoding, PdfInfo
 check_ocrmypdf = pytest.helpers.check_ocrmypdf
 run_ocrmypdf = pytest.helpers.run_ocrmypdf
 run_ocrmypdf_api = pytest.helpers.run_ocrmypdf_api
-spoof = pytest.helpers.spoof
 
 
 RENDERERS = ['hocr', 'sandwich']
 
 
-@pytest.fixture(scope='session')
-def spoof_tesseract_crash(tmp_path_factory):
-    return spoof(tmp_path_factory, tesseract='tesseract_crash.py')
-
-
-@pytest.fixture(scope='session')
-def spoof_tesseract_big_image_error(tmp_path_factory):
-    return spoof(tmp_path_factory, tesseract='tesseract_big_image_error.py')
-
-
-def test_quick(spoof_tesseract_cache, resources, outpdf):
-    check_ocrmypdf(resources / 'ccitt.pdf', outpdf, env=spoof_tesseract_cache)
+def test_quick(resources, outpdf):
+    check_ocrmypdf(
+        resources / 'ccitt.pdf', outpdf, '--plugin', 'tests/plugins/tesseract_cache.py'
+    )
 
 
 @pytest.mark.parametrize('renderer', RENDERERS)
-def test_oversample(spoof_tesseract_cache, renderer, resources, outpdf):
+def test_oversample(renderer, resources, outpdf):
     oversampled_pdf = check_ocrmypdf(
         resources / 'skew.pdf',
         outpdf,
@@ -69,13 +61,14 @@ def test_oversample(spoof_tesseract_cache, renderer, resources, outpdf):
         '-f',
         '--pdf-renderer',
         renderer,
-        env=spoof_tesseract_cache,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
 
     pdfinfo = PdfInfo(oversampled_pdf)
 
-    print(pdfinfo[0].xres)
-    assert abs(pdfinfo[0].xres - 350) < 1
+    print(pdfinfo[0].dpi.x)
+    assert abs(pdfinfo[0].dpi.x - 350) < 1
 
 
 def test_repeat_ocr(resources, no_outpdf):
@@ -83,17 +76,25 @@ def test_repeat_ocr(resources, no_outpdf):
     assert result == ExitCode.already_done_ocr
 
 
-def test_force_ocr(spoof_tesseract_cache, resources, outpdf):
+def test_force_ocr(resources, outpdf):
     out = check_ocrmypdf(
-        resources / 'graph_ocred.pdf', outpdf, '-f', env=spoof_tesseract_cache
+        resources / 'graph_ocred.pdf',
+        outpdf,
+        '-f',
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
     pdfinfo = PdfInfo(out)
     assert pdfinfo[0].has_text
 
 
-def test_skip_ocr(spoof_tesseract_cache, resources, outpdf):
+def test_skip_ocr(resources, outpdf):
     out = check_ocrmypdf(
-        resources / 'graph_ocred.pdf', outpdf, '-s', env=spoof_tesseract_cache
+        resources / 'graph_ocred.pdf',
+        outpdf,
+        '-s',
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
     pdfinfo = PdfInfo(out)
     assert pdfinfo[0].has_text
@@ -101,17 +102,17 @@ def test_skip_ocr(spoof_tesseract_cache, resources, outpdf):
 
 def test_redo_ocr(resources, outpdf):
     in_ = resources / 'graph_ocred.pdf'
-    before = PdfInfo(in_, detailed_page_analysis=True)
+    before = PdfInfo(in_)
     out = outpdf
     out = check_ocrmypdf(in_, out, '--redo-ocr')
-    after = PdfInfo(out, detailed_page_analysis=True)
+    after = PdfInfo(out)
     assert before[0].has_text and after[0].has_text
     assert (
         before[0].get_textareas() != after[0].get_textareas()
     ), "Expected text to be different after re-OCR"
 
 
-def test_argsfile(spoof_tesseract_noop, resources, outdir):
+def test_argsfile(resources, outdir):
     path_argsfile = outdir / 'test_argsfile.txt'
     with open(str(path_argsfile), 'w') as argsfile:
         print(
@@ -119,15 +120,14 @@ def test_argsfile(spoof_tesseract_noop, resources, outdir):
             'ArgsFile Test',
             '--author',
             'Test Cases',
+            '--plugin',
+            'tests/plugins/tesseract_noop.py',
             sep='\n',
             end='\n',
             file=argsfile,
         )
     check_ocrmypdf(
-        resources / 'graph.pdf',
-        path_argsfile,
-        '@' + str(outdir / 'test_argsfile.txt'),
-        env=spoof_tesseract_noop,
+        resources / 'graph.pdf', path_argsfile, '@' + str(outdir / 'test_argsfile.txt')
     )
 
 
@@ -145,9 +145,14 @@ def test_ocr_timeout(renderer, resources, outpdf):
     assert not pdfinfo[0].has_text
 
 
-def test_skip_big(spoof_tesseract_cache, resources, outpdf):
+def test_skip_big(resources, outpdf):
     out = check_ocrmypdf(
-        resources / 'jbig2.pdf', outpdf, '--skip-big', '1', env=spoof_tesseract_cache
+        resources / 'jbig2.pdf',
+        outpdf,
+        '--skip-big',
+        '1',
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
     pdfinfo = PdfInfo(out)
     assert not pdfinfo[0].has_text
@@ -155,9 +160,7 @@ def test_skip_big(spoof_tesseract_cache, resources, outpdf):
 
 @pytest.mark.parametrize('renderer', RENDERERS)
 @pytest.mark.parametrize('output_type', ['pdf', 'pdfa'])
-def test_maximum_options(
-    spoof_tesseract_cache, renderer, output_type, resources, outpdf
-):
+def test_maximum_options(renderer, output_type, resources, outpdf):
     check_ocrmypdf(
         resources / 'multipage.pdf',
         outpdf,
@@ -178,18 +181,15 @@ def test_maximum_options(
         renderer,
         '--output-type',
         output_type,
-        env=spoof_tesseract_cache,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
 
 
-def test_tesseract_missing_tessdata(resources, no_outpdf, tmpdir):
-    env = os.environ.copy()
-    env['TESSDATA_PREFIX'] = os.fspath(tmpdir)
-
-    returncode = run_ocrmypdf_api(
-        resources / 'graph.pdf', no_outpdf, '-v', '1', '--skip-text', env=env
-    )
-    assert returncode == ExitCode.missing_dependency
+def test_tesseract_missing_tessdata(monkeypatch, resources, no_outpdf, tmpdir):
+    monkeypatch.setenv("TESSDATA_PREFIX", os.fspath(tmpdir))
+    with pytest.raises(MissingDependencyError):
+        run_ocrmypdf_api(resources / 'graph.pdf', no_outpdf, '-v', '1', '--skip-text')
 
 
 def test_invalid_input_pdf(resources, no_outpdf):
@@ -202,22 +202,26 @@ def test_blank_input_pdf(resources, outpdf):
     assert result == ExitCode.ok
 
 
-def test_force_ocr_on_pdf_with_no_images(spoof_tesseract_crash, resources, no_outpdf):
+def test_force_ocr_on_pdf_with_no_images(resources, no_outpdf):
     # As a correctness test, make sure that --force-ocr on a PDF with no
     # content still triggers tesseract. If tesseract crashes, then it was
     # called.
     p, _, _ = run_ocrmypdf(
-        resources / 'blank.pdf', no_outpdf, '--force-ocr', env=spoof_tesseract_crash
+        resources / 'blank.pdf',
+        no_outpdf,
+        '--force-ocr',
+        '--plugin',
+        'tests/plugins/tesseract_crash.py',
     )
     assert p.returncode == ExitCode.child_process_error
-    assert not os.path.exists(no_outpdf)
+    assert not no_outpdf.exists()
 
 
 @pytest.mark.skipif(
     pytest.helpers.is_macos() and pytest.helpers.running_in_travis(),
     reason="takes too long to install language packs in Travis macOS homebrew",
 )
-def test_german(spoof_tesseract_cache, resources, outdir):
+def test_german(resources, outdir):
     # Produce a sidecar too - implicit test that system locale is set up
     # properly. It is fine that we are testing -l deu on a French file because
     # we are exercising the functionality not going for accuracy.
@@ -230,10 +234,11 @@ def test_german(spoof_tesseract_cache, resources, outdir):
             'deu',  # more commonly installed
             '--sidecar',
             sidecar,
-            env=spoof_tesseract_cache,
+            '--plugin',
+            'tests/plugins/tesseract_cache.py',
         )
     except MissingDependencyError:
-        if 'deu' not in tesseract.languages():
+        if 'deu' not in tesseract.get_languages():
             pytest.xfail(reason="tesseract-deu language pack not installed")
         raise
 
@@ -243,23 +248,27 @@ def test_klingon(resources, outpdf):
     assert p.returncode == ExitCode.missing_dependency
 
 
-def test_missing_docinfo(spoof_tesseract_noop, resources, outpdf):
+def test_missing_docinfo(resources, outpdf):
     result = run_ocrmypdf_api(
         resources / 'missing_docinfo.pdf',
         outpdf,
         '-l',
         'eng',
         '--skip-text',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     assert result == ExitCode.ok
 
 
-def test_uppercase_extension(spoof_tesseract_noop, resources, outdir):
+def test_uppercase_extension(resources, outdir):
     shutil.copy(str(resources / "skew.pdf"), str(outdir / "UPPERCASE.PDF"))
 
     check_ocrmypdf(
-        outdir / "UPPERCASE.PDF", outdir / "UPPERCASE_OUT.PDF", env=spoof_tesseract_noop
+        outdir / "UPPERCASE.PDF",
+        outdir / "UPPERCASE_OUT.PDF",
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
 
@@ -270,6 +279,16 @@ def test_input_file_not_found(caplog, no_outpdf):
     assert input_file in caplog.text
 
 
+@pytest.mark.skipif(os.name == 'nt', reason="chmod")
+def test_input_file_not_readable(caplog, resources, outdir, no_outpdf):
+    input_file = outdir / 'trivial.pdf'
+    shutil.copy(resources / 'trivial.pdf', input_file)
+    input_file.chmod(0o000)
+    result = run_ocrmypdf_api(input_file, no_outpdf)
+    assert result == ExitCode.input_file
+    assert str(input_file) in caplog.text
+
+
 def test_input_file_not_a_pdf(caplog, no_outpdf):
     input_file = __file__  # Try to OCR this file
     result = run_ocrmypdf_api(input_file, no_outpdf)
@@ -285,7 +304,7 @@ def test_encrypted(resources, caplog, no_outpdf):
 
 
 @pytest.mark.parametrize('renderer', RENDERERS)
-def test_pagesegmode(renderer, spoof_tesseract_cache, resources, outpdf):
+def test_pagesegmode(renderer, resources, outpdf):
     check_ocrmypdf(
         resources / 'skew.pdf',
         outpdf,
@@ -295,12 +314,13 @@ def test_pagesegmode(renderer, spoof_tesseract_cache, resources, outpdf):
         '1',
         '--pdf-renderer',
         renderer,
-        env=spoof_tesseract_cache,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
 
 
 @pytest.mark.parametrize('renderer', RENDERERS)
-def test_tesseract_crash(renderer, spoof_tesseract_crash, resources, no_outpdf, caplog):
+def test_tesseract_crash(renderer, resources, no_outpdf):
     p, _, err = run_ocrmypdf(
         resources / 'ccitt.pdf',
         no_outpdf,
@@ -308,29 +328,32 @@ def test_tesseract_crash(renderer, spoof_tesseract_crash, resources, no_outpdf,
         '1',
         '--pdf-renderer',
         renderer,
-        env=spoof_tesseract_crash,
+        '--plugin',
+        'tests/plugins/tesseract_crash.py',
     )
     assert p.returncode == ExitCode.child_process_error
-    assert not os.path.exists(no_outpdf)
+    assert not no_outpdf.exists()
     assert "SubprocessOutputError" in err
 
 
-def test_tesseract_crash_autorotate(spoof_tesseract_crash, resources, no_outpdf):
+def test_tesseract_crash_autorotate(resources, no_outpdf):
     p, out, err = run_ocrmypdf(
-        resources / 'ccitt.pdf', no_outpdf, '-r', env=spoof_tesseract_crash
+        resources / 'ccitt.pdf',
+        no_outpdf,
+        '-r',
+        '--plugin',
+        'tests/plugins/tesseract_crash.py',
     )
     assert p.returncode == ExitCode.child_process_error
-    assert not os.path.exists(no_outpdf)
-    assert "ERROR" in err
+    assert not no_outpdf.exists()
+    assert "uncaught exception" in err
     print(out)
     print(err)
 
 
 @pytest.mark.parametrize('renderer', RENDERERS)
 @pytest.mark.slow
-def test_tesseract_image_too_big(
-    renderer, spoof_tesseract_big_image_error, resources, outpdf
-):
+def test_tesseract_image_too_big(renderer, resources, outpdf):
     check_ocrmypdf(
         resources / 'hugemono.pdf',
         outpdf,
@@ -339,18 +362,22 @@ def test_tesseract_image_too_big(
         renderer,
         '--max-image-mpixels',
         '0',
-        env=spoof_tesseract_big_image_error,
+        '--plugin',
+        'tests/plugins/tesseract_big_image_error.py',
     )
 
 
-def test_algo4(resources, spoof_tesseract_noop, outpdf):
+def test_algo4(resources, outpdf):
     p, _, _ = run_ocrmypdf(
-        resources / 'encrypted_algo4.pdf', outpdf, env=spoof_tesseract_noop
+        resources / 'encrypted_algo4.pdf',
+        outpdf,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     assert p.returncode == ExitCode.encrypted_pdf
 
 
-def test_jbig2_passthrough(spoof_tesseract_cache, resources, outpdf):
+def test_jbig2_passthrough(resources, outpdf):
     out = check_ocrmypdf(
         resources / 'jbig2.pdf',
         outpdf,
@@ -358,49 +385,64 @@ def test_jbig2_passthrough(spoof_tesseract_cache, resources, outpdf):
         'pdf',
         '--pdf-renderer',
         'hocr',
-        env=spoof_tesseract_cache,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
     out_pageinfo = PdfInfo(out)
     assert out_pageinfo[0].images[0].enc == Encoding.jbig2
 
 
-def test_masks(spoof_tesseract_noop, resources, outpdf):
+def test_masks(resources, outpdf):
     assert (
         ocrmypdf.ocr(
-            resources / 'masks.pdf', outpdf, tesseract_env=spoof_tesseract_noop
+            resources / 'masks.pdf', outpdf, plugins=['tests/plugins/tesseract_noop.py']
         )
         == ExitCode.ok
     )
 
 
-def test_linearized_pdf_and_indirect_object(spoof_tesseract_noop, resources, outpdf):
-    check_ocrmypdf(resources / 'epson.pdf', outpdf, env=spoof_tesseract_noop)
-
-
-def test_very_high_dpi(spoof_tesseract_cache, resources, outpdf):
-    "Checks for a Decimal quantize error with high DPI, etc"
-    check_ocrmypdf(resources / '2400dpi.pdf', outpdf, env=spoof_tesseract_cache)
-    pdfinfo = PdfInfo(outpdf)
-
-    image = pdfinfo[0].images[0]
-    assert isclose(image.xres, image.yres)
-    assert isclose(image.xres, 2400)
-
-
-def test_overlay(spoof_tesseract_noop, resources, outpdf):
+def test_linearized_pdf_and_indirect_object(resources, outpdf):
     check_ocrmypdf(
-        resources / 'overlay.pdf', outpdf, '--skip-text', env=spoof_tesseract_noop
+        resources / 'epson.pdf', outpdf, '--plugin', 'tests/plugins/tesseract_noop.py'
     )
 
 
-def test_destination_not_writable(spoof_tesseract_noop, resources, outdir):
+def test_very_high_dpi(resources, outpdf):
+    "Checks for a Decimal quantize error with high DPI, etc"
+    check_ocrmypdf(
+        resources / '2400dpi.pdf',
+        outpdf,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
+    )
+    pdfinfo = PdfInfo(outpdf)
+
+    image = pdfinfo[0].images[0]
+    assert isclose(image.dpi.x, image.dpi.y)
+    assert isclose(image.dpi.x, 2400)
+
+
+def test_overlay(resources, outpdf):
+    check_ocrmypdf(
+        resources / 'overlay.pdf',
+        outpdf,
+        '--skip-text',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
+    )
+
+
+def test_destination_not_writable(resources, outdir):
     if os.name != 'nt' and (os.getuid() == 0 or os.geteuid() == 0):
         pytest.xfail(reason="root can write to anything")
     protected_file = outdir / 'protected.pdf'
     protected_file.touch()
     protected_file.chmod(0o400)  # Read-only
-    p, out, err = run_ocrmypdf(
-        resources / 'jbig2.pdf', protected_file, env=spoof_tesseract_noop
+    p, _out, _err = run_ocrmypdf(
+        resources / 'jbig2.pdf',
+        protected_file,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     assert p.returncode == ExitCode.file_access_error, "Expected error"
 
@@ -437,7 +479,7 @@ THIS FILE IS INVALID
 '''
         )
 
-    p, out, err = run_ocrmypdf(
+    p, _out, err = run_ocrmypdf(
         resources / 'ccitt.pdf',
         outdir / 'out.pdf',
         '--pdf-renderer',
@@ -473,9 +515,13 @@ def test_user_words_ocr(resources, outdir):
     )
 
 
-def test_form_xobject(spoof_tesseract_noop, resources, outpdf):
+def test_form_xobject(resources, outpdf):
     check_ocrmypdf(
-        resources / 'formxobject.pdf', outpdf, '--force-ocr', env=spoof_tesseract_noop
+        resources / 'formxobject.pdf',
+        outpdf,
+        '--force-ocr',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
 
@@ -507,33 +553,36 @@ def test_pagesize_consistency(renderer, resources, outpdf):
     assert isclose(before_dims[1], after_dims[1], rel_tol=1e-4)
 
 
-def test_skip_big_with_no_images(spoof_tesseract_noop, resources, outpdf):
+def test_skip_big_with_no_images(resources, outpdf):
     check_ocrmypdf(
         resources / 'blank.pdf',
         outpdf,
         '--skip-big',
         '5',
         '--force-ocr',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
 
 @pytest.mark.skipif(
-    '8.0.0' <= qpdf.version() <= '8.0.1',
-    reason="qpdf regression on pages with no contents",
+    '8.0.0' <= pikepdf.__libqpdf_version__ <= '8.0.1',
+    reason="libqpdf regression on pages with no contents",
 )
-def test_no_contents(spoof_tesseract_noop, resources, outpdf):
+def test_no_contents(resources, outpdf):
     check_ocrmypdf(
-        resources / 'no_contents.pdf', outpdf, '--force-ocr', env=spoof_tesseract_noop
+        resources / 'no_contents.pdf',
+        outpdf,
+        '--force-ocr',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
 
 @pytest.mark.parametrize(
     'image', ['baiona.png', 'baiona_gray.png', 'baiona_alpha.png', 'congress.jpg']
 )
-def test_compression_preserved(
-    spoof_tesseract_noop, ocrmypdf_exec, resources, image, outpdf
-):
+def test_compression_preserved(ocrmypdf_exec, resources, image, outpdf):
     input_file = str(resources / image)
     output_file = str(outpdf)
 
@@ -547,6 +596,8 @@ def test_compression_preserved(
             '150',
             '--output-type',
             'pdf',
+            '--plugin',
+            'tests/plugins/tesseract_noop.py',
             '-',
             output_file,
         ]
@@ -556,7 +607,7 @@ def test_compression_preserved(
             stderr=PIPE,
             stdin=input_stream,
             universal_newlines=True,
-            env=spoof_tesseract_noop,
+            check=False,
         )
 
         if im.mode in ('RGBA', 'LA'):
@@ -589,9 +640,7 @@ def test_compression_preserved(
         ('congress.jpg', 'lossless'),
     ],
 )
-def test_compression_changed(
-    spoof_tesseract_noop, ocrmypdf_exec, resources, image, compression, outpdf
-):
+def test_compression_changed(ocrmypdf_exec, resources, image, compression, outpdf):
     input_file = str(resources / image)
     output_file = str(outpdf)
 
@@ -608,6 +657,8 @@ def test_compression_changed(
             '0',
             '--pdfa-image-compression',
             compression,
+            '--plugin',
+            'tests/plugins/tesseract_noop.py',
             '-',
             output_file,
         ]
@@ -617,7 +668,7 @@ def test_compression_changed(
             stderr=PIPE,
             stdin=input_stream,
             universal_newlines=True,
-            env=spoof_tesseract_noop,
+            check=False,
         )
         assert p.returncode == ExitCode.ok, p.stderr
 
@@ -643,7 +694,7 @@ def test_compression_changed(
     im.close()
 
 
-def test_sidecar_pagecount(spoof_tesseract_cache, resources, outpdf):
+def test_sidecar_pagecount(resources, outpdf):
     sidecar = outpdf.with_suffix('.txt')
     check_ocrmypdf(
         resources / '3small.pdf',
@@ -651,7 +702,8 @@ def test_sidecar_pagecount(spoof_tesseract_cache, resources, outpdf):
         '--skip-text',
         '--sidecar',
         sidecar,
-        env=spoof_tesseract_cache,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
 
     pdfinfo = PdfInfo(resources / '3small.pdf')
@@ -667,10 +719,15 @@ def test_sidecar_pagecount(spoof_tesseract_cache, resources, outpdf):
     ), "Sidecar page count does not match PDF page count"
 
 
-def test_sidecar_nonempty(spoof_tesseract_cache, resources, outpdf):
+def test_sidecar_nonempty(resources, outpdf):
     sidecar = outpdf.with_suffix('.txt')
     check_ocrmypdf(
-        resources / 'ccitt.pdf', outpdf, '--sidecar', sidecar, env=spoof_tesseract_cache
+        resources / 'ccitt.pdf',
+        outpdf,
+        '--sidecar',
+        sidecar,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
 
     with open(sidecar, 'r', encoding='utf-8') as f:
@@ -679,7 +736,7 @@ def test_sidecar_nonempty(spoof_tesseract_cache, resources, outpdf):
 
 
 @pytest.mark.parametrize('pdfa_level', ['1', '2', '3'])
-def test_pdfa_n(spoof_tesseract_cache, pdfa_level, resources, outpdf):
+def test_pdfa_n(pdfa_level, resources, outpdf):
     if pdfa_level == '3' and ghostscript.version() < '9.19':
         pytest.xfail(reason='Ghostscript >= 9.19 required')
 
@@ -688,7 +745,8 @@ def test_pdfa_n(spoof_tesseract_cache, pdfa_level, resources, outpdf):
         outpdf,
         '--output-type',
         'pdfa-' + pdfa_level,
-        env=spoof_tesseract_cache,
+        '--plugin',
+        'tests/plugins/tesseract_cache.py',
     )
 
     pdfa_info = file_claims_pdfa(outpdf)
@@ -700,44 +758,61 @@ def test_pdfa_n(spoof_tesseract_cache, pdfa_level, resources, outpdf):
 )
 @pytest.mark.slow
 def test_decompression_bomb(resources, outpdf):
-    p, out, err = run_ocrmypdf(resources / 'hugemono.pdf', outpdf)
+    p, _out, err = run_ocrmypdf(resources / 'hugemono.pdf', outpdf)
     assert 'decompression bomb' in err
 
-    p, out, err = run_ocrmypdf(
+    p, _out, err = run_ocrmypdf(
         resources / 'hugemono.pdf', outpdf, '--max-image-mpixels', '2000'
     )
     assert p.returncode == 0
 
 
-def test_text_curves(spoof_tesseract_noop, resources, outpdf):
+def test_text_curves(resources, outpdf):
     with patch('ocrmypdf._pipeline.VECTOR_PAGE_DPI', 100):
-        check_ocrmypdf(resources / 'vector.pdf', outpdf, env=spoof_tesseract_noop)
+        check_ocrmypdf(
+            resources / 'vector.pdf',
+            outpdf,
+            '--plugin',
+            'tests/plugins/tesseract_noop.py',
+        )
 
         info = PdfInfo(outpdf)
         assert len(info.pages[0].images) == 0, "added images to the vector PDF"
 
         check_ocrmypdf(
-            resources / 'vector.pdf', outpdf, '--force-ocr', env=spoof_tesseract_noop
+            resources / 'vector.pdf',
+            outpdf,
+            '--force-ocr',
+            '--plugin',
+            'tests/plugins/tesseract_noop.py',
         )
 
         info = PdfInfo(outpdf)
         assert len(info.pages[0].images) != 0, "force did not rasterize"
 
 
-def test_output_is_dir(spoof_tesseract_noop, resources, outdir):
-    p, out, err = run_ocrmypdf(
-        resources / 'trivial.pdf', outdir, '--force-ocr', env=spoof_tesseract_noop
+def test_output_is_dir(resources, outdir):
+    p, _out, err = run_ocrmypdf(
+        resources / 'trivial.pdf',
+        outdir,
+        '--force-ocr',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     assert p.returncode == ExitCode.file_access_error
     assert 'is not a writable file' in err
 
 
 @pytest.mark.skipif(os.name == 'nt', reason="symlink needs admin permissions")
-def test_output_is_symlink(spoof_tesseract_noop, resources, outdir):
+def test_output_is_symlink(resources, outdir):
     sym = Path(outdir / 'this_is_a_symlink')
     sym.symlink_to(outdir / 'out.pdf')
-    p, out, err = run_ocrmypdf(
-        resources / 'trivial.pdf', sym, '--force-ocr', env=spoof_tesseract_noop
+    p, _out, err = run_ocrmypdf(
+        resources / 'trivial.pdf',
+        sym,
+        '--force-ocr',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     assert p.returncode == ExitCode.ok, err
     assert (outdir / 'out.pdf').stat().st_size > 0, 'target file not created'
@@ -750,8 +825,6 @@ def test_livecycle(resources, no_outpdf):
 
 
 def test_version_check():
-    from ocrmypdf.exec import get_version
-
     with pytest.raises(MissingDependencyError):
         get_version('NOT_FOUND_UNLIKELY_ON_PATH')
 
@@ -775,9 +848,7 @@ def test_version_check():
         [0.0, 1, 'pdf', True],
     ],
 )
-def test_fast_web_view(
-    spoof_tesseract_noop, resources, outpdf, threshold, optimize, output_type, expected
-):
+def test_fast_web_view(resources, outpdf, threshold, optimize, output_type, expected):
     check_ocrmypdf(
         resources / 'trivial.pdf',
         outpdf,
@@ -787,18 +858,20 @@ def test_fast_web_view(
         optimize,
         '--output-type',
         output_type,
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     with pikepdf.open(outpdf) as pdf:
         assert pdf.is_linearized == expected
 
 
-def test_image_dpi_not_image(caplog, spoof_tesseract_noop, resources, outpdf):
+def test_image_dpi_not_image(caplog, resources, outpdf):
     check_ocrmypdf(
         resources / 'trivial.pdf',
         outpdf,
         '--image-dpi',
         '100',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
     assert '--image-dpi is being ignored' in caplog.text

tests/test_metadata.py

@@ -17,22 +17,23 @@
 
 
 import datetime
-import logging
 import mmap
-import os
 from datetime import timezone
 from os import fspath
-from pathlib import Path
-from shutil import copyfile, move
-from unittest.mock import MagicMock, patch
+from shutil import copyfile
+from unittest.mock import patch
 
 import pikepdf
 import pytest
 from pikepdf.models.metadata import decode_pdf_date
 
-from ocrmypdf._jobcontext import PDFContext
+from ocrmypdf._jobcontext import PdfContext
+from ocrmypdf._pipeline import convert_to_pdfa, metadata_fixup
+from ocrmypdf._plugin_manager import get_plugin_manager
+from ocrmypdf.cli import get_parser
 from ocrmypdf.exceptions import ExitCode
 from ocrmypdf.pdfa import SRGB_ICC_PROFILE, file_claims_pdfa, generate_pdfa_ps
+from ocrmypdf.pdfinfo import PdfInfo
 
 try:
     import fitz
@@ -41,17 +42,15 @@ except ImportError:
 
 # pytest.helpers is dynamic
 # pylint: disable=no-member
-# pylint: disable=w0612
 
 pytestmark = pytest.mark.filterwarnings('ignore:.*XMLParser.*:DeprecationWarning')
 
 check_ocrmypdf = pytest.helpers.check_ocrmypdf
 run_ocrmypdf = pytest.helpers.run_ocrmypdf
-spoof = pytest.helpers.spoof
 
 
 @pytest.mark.parametrize("output_type", ['pdfa', 'pdf'])
-def test_preserve_metadata(spoof_tesseract_noop, output_type, resources, outpdf):
+def test_preserve_metadata(output_type, resources, outpdf):
     pdf_before = pikepdf.open(resources / 'graph.pdf')
 
     output = check_ocrmypdf(
@@ -59,7 +58,8 @@ def test_preserve_metadata(spoof_tesseract_noop, output_type, resources, outpdf)
         outpdf,
         '--output-type',
         output_type,
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     pdf_after = pikepdf.open(output)
@@ -72,12 +72,12 @@ def test_preserve_metadata(spoof_tesseract_noop, output_type, resources, outpdf)
 
 
 @pytest.mark.parametrize("output_type", ['pdfa', 'pdf'])
-def test_override_metadata(spoof_tesseract_noop, output_type, resources, outpdf):
+def test_override_metadata(output_type, resources, outpdf):
     input_file = resources / 'c02-22.pdf'
     german = 'Du siehst den Wald vor lauter Bäumen nicht.'
     chinese = '孔子'
 
-    p, out, err = run_ocrmypdf(
+    p, _out, err = run_ocrmypdf(
         input_file,
         outpdf,
         '--title',
@@ -86,7 +86,8 @@ def test_override_metadata(spoof_tesseract_noop, output_type, resources, outpdf)
         chinese,
         '--output-type',
         output_type,
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     assert p.returncode == ExitCode.ok, err
@@ -106,21 +107,22 @@ def test_override_metadata(spoof_tesseract_noop, output_type, resources, outpdf)
     assert pdfa_info['output'] == output_type
 
 
-def test_high_unicode(spoof_tesseract_noop, resources, no_outpdf):
+def test_high_unicode(resources, no_outpdf):
 
     # Ghostscript doesn't support high Unicode, so neither do we, to be
     # safe
     input_file = resources / 'c02-22.pdf'
     high_unicode = 'U+1030C is: 𐌌'
 
-    p, out, err = run_ocrmypdf(
+    p, _out, err = run_ocrmypdf(
         input_file,
         no_outpdf,
         '--subject',
         high_unicode,
         '--output-type',
         'pdfa',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     assert p.returncode == ExitCode.bad_args, err
@@ -129,9 +131,7 @@ def test_high_unicode(spoof_tesseract_noop, resources, no_outpdf):
 @pytest.mark.skipif(not fitz, reason="test uses fitz")
 @pytest.mark.parametrize('ocr_option', ['--skip-text', '--force-ocr'])
 @pytest.mark.parametrize('output_type', ['pdf', 'pdfa'])
-def test_bookmarks_preserved(
-    spoof_tesseract_noop, output_type, ocr_option, resources, outpdf
-):
+def test_bookmarks_preserved(output_type, ocr_option, resources, outpdf):
     input_file = resources / 'toc.pdf'
     before_toc = fitz.Document(str(input_file)).getToC()
 
@@ -141,7 +141,8 @@ def test_bookmarks_preserved(
         ocr_option,
         '--output-type',
         output_type,
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     after_toc = fitz.Document(str(outpdf)).getToC()
@@ -156,13 +157,16 @@ def seconds_between_dates(date1, date2):
 
 @pytest.mark.parametrize('infile', ['trivial.pdf', 'jbig2.pdf'])
 @pytest.mark.parametrize('output_type', ['pdf', 'pdfa'])
-def test_creation_date_preserved(
-    spoof_tesseract_noop, output_type, resources, infile, outpdf
-):
+def test_creation_date_preserved(output_type, resources, infile, outpdf):
     input_file = resources / infile
 
     check_ocrmypdf(
-        input_file, outpdf, '--output-type', output_type, env=spoof_tesseract_noop
+        input_file,
+        outpdf,
+        '--output-type',
+        output_type,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     pdf_before = pikepdf.open(input_file)
@@ -185,19 +189,23 @@ def test_creation_date_preserved(
 
 
 @pytest.mark.parametrize('output_type', ['pdf', 'pdfa'])
-def test_xml_metadata_preserved(spoof_tesseract_noop, output_type, resources, outpdf):
+def test_xml_metadata_preserved(output_type, resources, outpdf):
     input_file = resources / 'graph.pdf'
 
     try:
-        from libxmp import consts
-        from libxmp.utils import file_to_dict
-    except Exception:
+        from libxmp.utils import file_to_dict  # pylint: disable=import-outside-toplevel
+    except Exception:  # pylint: disable=broad-except
         pytest.skip("libxmp not available or libexempi3 not installed")
 
     before = file_to_dict(str(input_file))
 
     check_ocrmypdf(
-        input_file, outpdf, '--output-type', output_type, env=spoof_tesseract_noop
+        input_file,
+        outpdf,
+        '--output-type',
+        output_type,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     after = file_to_dict(str(outpdf))
@@ -275,9 +283,14 @@ def test_srgb_in_unicode_path(tmp_path):
         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
+def test_kodak_toc(resources, outpdf):
+    _output = check_ocrmypdf(
+        resources / 'kcs.pdf',
+        outpdf,
+        '--output-type',
+        'pdf',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     p = pikepdf.open(outpdf)
@@ -287,16 +300,15 @@ def test_kodak_toc(resources, outpdf, spoof_tesseract_noop):
 
 
 def test_metadata_fixup_warning(resources, outdir, caplog):
-    from ocrmypdf.__main__ import parser
-    from ocrmypdf._pipeline import metadata_fixup
-
-    options = parser.parse_args(
+    options = get_parser().parse_args(
         args=['--output-type', 'pdfa-2', 'graph.pdf', 'out.pdf']
     )
 
     copyfile(resources / 'graph.pdf', outdir / 'graph.pdf')
 
-    context = PDFContext(options, outdir, outdir / 'graph.pdf', None)
+    context = PdfContext(
+        options, outdir, outdir / 'graph.pdf', None, get_plugin_manager([])
+    )
     metadata_fixup(working_file=outdir / 'graph.pdf', context=context)
     for record in caplog.records:
         assert record.levelname != 'WARNING'
@@ -307,25 +319,31 @@ def test_metadata_fixup_warning(resources, outdir, caplog):
         meta['prism2:publicationName'] = 'OCRmyPDF Test'
     graph.save(outdir / 'graph_mod.pdf')
 
-    context = PDFContext(options, outdir, outdir / 'graph_mod.pdf', None)
+    context = PdfContext(
+        options, outdir, outdir / 'graph_mod.pdf', None, get_plugin_manager([])
+    )
     metadata_fixup(working_file=outdir / 'graph.pdf', context=context)
     assert any(record.levelname == 'WARNING' for record in caplog.records)
 
 
 def test_prevent_gs_invalid_xml(resources, outdir):
-    from ocrmypdf.__main__ import parser
-    from ocrmypdf._pipeline import convert_to_pdfa
-    from ocrmypdf.pdfa import generate_pdfa_ps
-    from ocrmypdf.pdfinfo import PdfInfo
-
     generate_pdfa_ps(outdir / 'pdfa.ps')
-    copyfile(resources / 'enron1.pdf', outdir / 'layers.rendered.pdf')
+    copyfile(resources / 'trivial.pdf', outdir / 'layers.rendered.pdf')
 
-    options = parser.parse_args(
+    # Inject a string with a trailing nul character into the DocumentInfo
+    # dictionary of this PDF, as often occurs in practice.
+    with pikepdf.open(outdir / 'layers.rendered.pdf') as pike:
+        pike.Root.DocumentInfo = pikepdf.Dictionary(
+            Title=b'String with trailing nul\x00'
+        )
+
+    options = get_parser().parse_args(
         args=['-j', '1', '--output-type', 'pdfa-2', 'a.pdf', 'b.pdf']
     )
-    pdfinfo = PdfInfo(resources / 'enron1.pdf')
-    context = PDFContext(options, outdir, resources / 'enron1.pdf', pdfinfo)
+    pdfinfo = PdfInfo(outdir / 'layers.rendered.pdf')
+    context = PdfContext(
+        options, outdir, outdir / 'layers.rendered.pdf', pdfinfo, get_plugin_manager([])
+    )
 
     convert_to_pdfa(
         str(outdir / 'layers.rendered.pdf'), str(outdir / 'pdfa.ps'), context
@@ -339,5 +357,32 @@ def test_prevent_gs_invalid_xml(resources, outdir):
             xmp_start = mm.find(XMP_MAGIC)
             xmp_end = mm.rfind(b'<?xpacket end', xmp_start)
             assert 0 < xmp_start < xmp_end
+            # Ensure we did not carry the nul forward.
             assert mm.find(b'&#0;', xmp_start, xmp_end) == -1, "found escaped nul"
             assert mm.find(b'\x00', xmp_start, xmp_end) == -1
+
+
+def test_malformed_docinfo(caplog, resources, outdir):
+    generate_pdfa_ps(outdir / 'pdfa.ps')
+    # copyfile(resources / 'trivial.pdf', outdir / 'layers.rendered.pdf')
+
+    with pikepdf.open(resources / 'trivial.pdf') as pike:
+        pike.trailer.Info = pikepdf.Stream(pike, b"<xml></xml>")
+        pike.save(outdir / 'layers.rendered.pdf', fix_metadata_version=False)
+
+    options = get_parser().parse_args(
+        args=['-j', '1', '--output-type', 'pdfa-2', 'a.pdf', 'b.pdf']
+    )
+    pdfinfo = PdfInfo(outdir / 'layers.rendered.pdf')
+    context = PdfContext(
+        options, outdir, outdir / 'layers.rendered.pdf', pdfinfo, get_plugin_manager([])
+    )
+
+    convert_to_pdfa(
+        str(outdir / 'layers.rendered.pdf'), str(outdir / 'pdfa.ps'), context
+    )
+
+    print(caplog.records)
+    assert any(
+        'malformed DocumentInfo block' in record.message for record in caplog.records
+    )

tests/test_optimize.py

@@ -15,17 +15,19 @@
 # You should have received a copy of the GNU General Public License
 # along with OCRmyPDF.  If not, see <http://www.gnu.org/licenses/>.
 
-import logging
 from os import fspath
 from pathlib import Path
+from unittest.mock import patch
 
+import img2pdf
 import pikepdf
 import pytest
-from PIL import Image
+from PIL import Image, ImageDraw
 
 from ocrmypdf import optimize as opt
-from ocrmypdf.exec import jbig2enc, pngquant
-from ocrmypdf.exec.ghostscript import rasterize_pdf
+from ocrmypdf._exec import jbig2enc, pngquant
+from ocrmypdf._exec.ghostscript import rasterize_pdf
+from ocrmypdf.helpers import Resolution
 
 check_ocrmypdf = pytest.helpers.check_ocrmypdf  # pylint: disable=e1101
 
@@ -45,10 +47,8 @@ def test_mono_not_inverted(resources, outdir):
     rasterize_pdf(
         outdir / 'out.pdf',
         outdir / 'im.png',
-        xres=10,
-        yres=10,
         raster_device='pnggray',
-        log=logging.getLogger(name='test_mono_not_inverted'),
+        raster_dpi=Resolution(10, 10),
     )
 
     with Image.open(fspath(outdir / 'im.png')) as im:
@@ -56,7 +56,7 @@ def test_mono_not_inverted(resources, outdir):
 
 
 @pytest.mark.skipif(not pngquant.available(), reason='need pngquant')
-def test_jpg_png_params(resources, outpdf, spoof_tesseract_noop):
+def test_jpg_png_params(resources, outpdf):
     check_ocrmypdf(
         resources / 'crom.png',
         outpdf,
@@ -68,13 +68,14 @@ def test_jpg_png_params(resources, outpdf, spoof_tesseract_noop):
         '50',
         '--png-quality',
         '20',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
 
 @pytest.mark.skipif(not jbig2enc.available(), reason='need jbig2enc')
 @pytest.mark.parametrize('lossy', [False, True])
-def test_jbig2_lossy(lossy, resources, outpdf, spoof_tesseract_noop):
+def test_jbig2_lossy(lossy, resources, outpdf):
     args = [
         resources / 'ccitt.pdf',
         outpdf,
@@ -86,11 +87,13 @@ def test_jbig2_lossy(lossy, resources, outpdf, spoof_tesseract_noop):
         '50',
         '--png-quality',
         '20',
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     ]
     if lossy:
         args.append('--jbig2-lossy')
 
-    check_ocrmypdf(*args, env=spoof_tesseract_noop)
+    check_ocrmypdf(*args)
 
     pdf = pikepdf.open(outpdf)
     pim = pikepdf.PdfImage(next(iter(pdf.pages[0].images.values())))
@@ -106,7 +109,7 @@ def test_jbig2_lossy(lossy, resources, outpdf, spoof_tesseract_noop):
     not jbig2enc.available() or not pngquant.available(),
     reason='need jbig2enc and pngquant',
 )
-def test_flate_to_jbig2(resources, outdir, spoof_tesseract_noop):
+def test_flate_to_jbig2(resources, outdir):
     # This test requires an image that pngquant is capable of converting to
     # to 1bpp - so use an existing 1bpp image, convert up, confirm it can
     # convert down
@@ -124,9 +127,49 @@ def test_flate_to_jbig2(resources, outdir, spoof_tesseract_noop):
         '50',
         '--optimize',
         '3',
-        env=spoof_tesseract_noop,
+        '--plugin',
+        'tests/plugins/tesseract_noop.py',
     )
 
     pdf = pikepdf.open(outdir / 'out.pdf')
     pim = pikepdf.PdfImage(next(iter(pdf.pages[0].images.values())))
     assert pim.filters[0] == '/JBIG2Decode'
+
+
+def test_multiple_pngs(resources, outdir):
+    with Path.open(outdir / 'in.pdf', 'wb') as inpdf:
+        img2pdf.convert(
+            fspath(resources / 'baiona_colormapped.png'),
+            fspath(resources / 'baiona_gray.png'),
+            with_pdfrw=False,
+            outputstream=inpdf,
+        )
+
+    def mockquant(input_file, output_file, _quality_min, _quality_max):
+        with Image.open(input_file) as im:
+            draw = ImageDraw.Draw(im)
+            draw.rectangle((0, 0, im.width, im.height), fill=128)
+            im.save(output_file)
+
+    with patch('ocrmypdf.optimize.pngquant.quantize', new=mockquant):
+        check_ocrmypdf(
+            outdir / 'in.pdf',
+            outdir / 'out.pdf',
+            '--optimize',
+            '3',
+            '--jobs',
+            '1',
+            '--use-threads',
+            '--output-type',
+            'pdf',
+            '--plugin',
+            'tests/plugins/tesseract_noop.py',
+        )
+
+    with pikepdf.open(outdir / 'in.pdf') as inpdf, pikepdf.open(
+        outdir / 'out.pdf'
+    ) as outpdf:
+        for n in range(len(inpdf.pages)):
+            inim = next(iter(inpdf.pages[n].images.values()))
+            outim = next(iter(outpdf.pages[n].images.values()))
+            assert len(outim.read_raw_bytes()) < len(inim.read_raw_bytes()), n