What I need to do, at the command line, is:

1. Extract the contents of a a ZIP archive.
2. Change an arbitrary number of the extracted files.
3. Repeat step 1, but because it is a huge archive, only extract the archived copies of the files changed in step 2 which is much faster.

Ideally the extraction in step 3 would do something like a checksum on the files on disk and only extract those where the file in the archive has a different checksum. Or maybe compare the date changed stamp on the disk file.

At the minute I use pkzipc.exe which is the command-line version of PkZip. I can't see a way to do it with this though. You can extract files from the archive that are newer than the disk files, but what I want is the opposite of that in a sense.