On Wed, Aug 19, 2009 at 4:12 PM, Francesco Romani<from...@gmail.com> wrote: > On Wed, 2009-08-19 at 23:08 +0200, Francesco Romani wrote: >> Fellows pythonistas beware! The code deserves a massive amount of PEP8 >> love and still a lot of polishment. > > and a lot of polish is still pending. >
yay python! I plan on making the following test thingy: 1. use PIL (python image lib) to generate an image with some text on it. 2. use something (maybe http://code.google.com/p/pyffmpeg) to make various movie files. 3. for test in [lots of test cases]: 4. transcode this.foo to that.bar 5. more magic to extract a frame from that.bar 6. use gocr or http://code.google.com/p/pytesser/ to try and read the text. I am guessing many of the tests will crash-n-burn before they ever hit 6, but I also am hoping that I will be able to trap those cases and log the failure. btw - I just wrote some code to ocr a .dv file. it works good enough. -- Carl K
