We have a space that has a rule that extracts common metadata fields when content is added to a folder or created.
The problem is that when we do an import of zip file, the metadata is wrong. Some of the files grab metadata from another file in the zip archive and uses that for its metadata. Its usually the headline and caption fields that get messed up in the import.
Ideally this should never be happen as during bulk import first the contents are created in your repository and then your rule of extracting metadata will be triggered on newly created contents to extract comman metadata from it. Are you using any custom metadata extractor? Is this happening with specific mimtype files?
I think this may have to do with corrupted metadata or a format of metadata that may be incompatible and causing problems.
I can't remember when the last time I edited the metadata on the files that I am using, but when I modify the metadata in in Bridge CS2 and then import the zip file the metadata extracts fine.
When I don't edit the metadata in Bridge CS2, I run into problems.
I am going to run some more tests to see if this is the problem.