v0.23.0

Update dependencies for Python 3.13 support

• numpy >=2.1.0
• scikit-image >=0.25.0rc1 (as 0.25.0 stable with Python 3.13 has not been released yet)

v0.22.0

Features

• Improve processing speed of ps and pcl files
  • Call pcl6 and gs (ghostscript) via subprocess to render images instead of using wand.image
• Option to embed template screenshots in Image Should Contain Template in original size

Thank you @wanghui-00 and @simonmeggle for your contributions! 👍

v0.21.0

Features

• Image Should Contain Template has option to crop template image before comparison by using arguments tpl_crop_x1, tpl_crop_y1, tpl_crop_x2, tpl_crop_x2,
• Default threshold in Image Should Contain Template has been adjusted to 0.0 to only consider exact matches
• Template image can be logged as screenshot in Image Should Contain Template
• Names of screenshot files are added to Log

Fixes

• Image Should Contain Template now throws correct AssertionError in case the template was not found in image

Thank you @simonmeggle for your contribution 🦸‍♂️

v0.20.0

Features

• Python 3.12 support
• Add Keyword Image Should Contain Template
  • This will find a small image template in a bigger image

v0.19.0

Features:

• Support URLs (in addition to local files) for PDF Content and Print Job Keywords

v0.18.1

Lock PyMuPDF version to <1.23.0.
The new 1.23.0 brought some breaking changes, so maximum version of PyMuPDF for now is 1.22.5 until I fix the problems.

v0.18.0

Remove ThreadPool Executor in VisualTest

• Reference and Candidate Images were rendered in parallel using ThreadPool Executor
• It was now removed because
  • There was no improved speed
  • It made debugging tricky
  • Logging from those parallel Threads was tricky
  • It caused some strange problems when an exception was raised in one of the Threads.

Speed improvement when using OCR for Masks

• In case of multiple masks using OCR, the text recognition was triggered for each mask separately
• Now OCR is only triggered once and the identified texts and positions are stored and re-used by the succeeding mask

v0.16.0

Features

• Get Text From Document
  • Add ocr_confidence argument

v0.15.0

Features

• Get Text From Document
  • Add ocr_config, ocr_lang and increase_resolution arguments

v0.14.0

Features

• Add embed_screenshots argument on library import
  Screenshots will be embedded as base64 into log.html