pytesseract

• Add support for Python 3.12

• New functionality run_and_get_multiple_output #511

• Fix for config parsing on Windows (#356 and #501)
• Fixes for the default hocr and boxing configs (#106 and #454)
• get_tesseract_version caching is optional and disabled by default for the user (#411)
• Various CI upgrades

• Fix image_to_osd regression (reported by @klavdijS)

• Support only python 3.7+ (3.6 is End of Life)
• Switch from deprecated distutils version parsing to packaging.version
• Add build system requirements in order for pip to properly install Pillow

Bugfix release:

• Header handling in file_to_dict function by @igor-ma
• Proper raising behavior in case of timeout by @amenezes
• __version__ attribute support by @amenezes
• Couple of bug fixes

[breaking change] Python 2 and 3.5 support dropped - reason, End of Life.
Migration to GitHub Actions CI workflow and pre-commit.ci for linting.
Added new function get_languages.
A couple of bug fixes.

Added support for ALTO XML output (by @eivtho)
Minor bugfix with TSV parsing.
More graceful timeout termination under Unix-like environments.

Removed pre-processing conversion - Pillow's image mode matrix doesn't support all use cases.
Fix for the output data from tesseract
Fix for dynamic I/O error handling

Support for WebP images
Support for python 3.8 (CI testing)
Improved cli error reporting
Fix for temp files under Windows

Bug fixes for:

• temp file read access on Windows
• duplicate format argument upon image save

Bug fix for proper cleaning of the temp files

• License changed from GPLv3 to Apache 2.0 ( for more info see #237 )
• Additional documentation and examples
• Extended config support for pandas
• pre-commit config for the project
• Bug fixes and refactoring

• Fix issue with leaking file descriptors.

• Additional documentation and exported new method run_and_get_output.
• Improve image format handling, whenever using Pillow library manipulation.
• Fixed issue with numpy support.
• Fix timeout issues with Travis CI.

• Various minor fixes and refactoring
• Introduce the new timeout functionality
• Improved and updated documentation with additional examples

• Documentation improvements and new examples for batch processing
• Various fixes and code refactoring for the child process routines
• No more support for python 2.6, only python 2.7+

• Support for the pandas DataFrame format
• Code refactoring and fixes for the subprocess routines

• Support for searchable pdf and hocr output format
• Cleanup of deprecated options
• Documentation adjustments

• Refactoring and raw image support (by passing relative or absolute image path)
• Fix psm/osd/tsv mode handling logic
• Documentation adjustments