Hash checking for extractor.*.skip

[panhartstuff]

It would be nice to have hash checking options for the extractor.*.skip configuration
Very useful in a scenario where a module gets updated with a better downloader (i.e. grabs better image quality)

Some suggestions:

• "hashoverwrite": Downloads the file again, then compare the hash with the older version of the file. If they mismatch, overwrite the older version
• "hashenumerate": Downloads the file again, then compare the hash with the older version of the file. If they mismatch, add an enumeration index to the beginning of the filename extension (file.1.ext, file.2.ext, etc.)

[mikf]

This functionality is now implemented as a post processor module: 254f7c3
(It's a lot easier to do it like that)

It requires extractor.skip to be set to false, so it re-downloads the file even though it was downloaded before, and downloader.part set to true (the default), so it writes the downloaded file to a temporary location first.

It will compare old and new versions of a file by comparing their size and then their content, and replace/enumerate them on mismatch. No hashing, because that's at least twice as slow than a direct comparison.

Is that (roughly) what you had in mind?