a real-time screen-scraping and primitive ocr toolkit
chatscraper.py
: monitor and dump on-screen, scrolling text windowsgui/capture.py
: filter and select windows by title, then click button to take screenshotsgui/fonts.py
: browse collected font imagesgui/pngscraper.py
: python shell + image browsergui/WindowSelector
: select and scrape a currently open window
config.py
: vocabulary for config file setupconvert.py
: misc conversion routinesfontdata.py
: class that maps images to charactersGlyph.py
: represents a glyph on-screenmi_spike
: helpers for anti-aliased fontsRectangle.py
: basic rectangle classRegion.py
: classes for processing fixed regions of the screenscrape.py
: algorithms for scraping text from imagestools.py
: various recognier tools that you can hook to a regionwindows.py
: a wrapper around the win32 api for finding windowsgui/builder.py
: class for building scrape config files
test.sh
: simple / dumb test runnertests/
: the actual tests. lots of commentary.
- glint
- glyph integer. basically a bitmap packed into an int
- glyph
- an image to recognize (usually single character)