csv-blobs improvements: windows binary, more speed, dramatically less memory

Comments are moderated. It may take a few minutes before your comment appears.
Markdown is supported in your comments.

After some profiling, the best-case input (a blank page) runs twice as fast and the worst-case input (100% static) runs 25 times faster. There still remains room for improvement. It is single threaded but parts of the algorithm could be run in parallel. If you only need to track blobs of a single color it could be made several times faster and lighter.

Mail: (not shown)

Please type this: