Orange Tools

Handy utilities used in development of Orange data mining software or its documentation.

trim.py

Trims the input PNG file. The upper-left pixel of the image is considered on the border and is used to mark the border color. Border is clipped-out, border color is considered a background and is transparent.

% python trim.py examples/paint-data.png
.. processing /Users/username/Desktop/paint-data.png, size (1264, 858)
   -> trimming: paint-data.png, size (1241, 834)
   -> thumb: paint-data-thumb.png, size (180, 120)

To construct a appropriate screenshot file on Windows 8, go to Control Panel, type "shadows" in the search box, and go to System, then there disable "Show shadows under windows". Choose a weird enough color for the background.

todo.py

Manage list of stickers for Orange. Prepares a list of html pages for sticker printout. The following call will print out all stickers that are marked not printed and are of priority 3.

% python todo.py -p 3

Stickers are defined on google sheet that can be accessed by biolab members.

stamper.py

Run it on a screenshot to place labels (circled numbers from 1 to 10). These augmented screenshots are then used in the widget documentation.

widget_icon_export.py

Run ./widget_icon_export.py [<format>] [<export-icon-size>] to export all currently installed widgets' icons into the current directory. If no arguments are given, format defaults to 'png', export-icon-size defaults to 100.

convert-to-indexed.sh

Run bash convert-to-indexed.sh <path-to-folder>. The script creates a new folder with indexed images. It requires ImageMagick first. On OSX, install with brew install imagemagick. See ImageMagick for alternatives.

tinify_images.py

This uses TinyPNG API to compress images in a given directory.

Example usage: python tinify_images.py -k <your_api_key> -p <path_to_dir> optionaly use -e parameter to speficy file extension (default are .png and .jpg).