[TIKA-4033] Improve metadata for incremental updates, take 2

• Type: Task
• Status: Resolved
• Priority: Trivial
• Resolution: Fixed
• Fix Version/s: 2.8.0
• Assignee: Unassigned
• Reporter: Tim Allison
• Created: 09/May/23 12:59
• Updated: 10/May/23 15:48
• Resolved: 09/May/23 16:50

Description

We're currently generating a "resourceName" in the PDFParser for incremental updates. The following isn't well documented (I don't think?), but we try to reserve "resourceName" for embedded files to be the actual name that the container document has for that embedded file.

Now, we need some kind of name for the embedded resource path in RecursiveParserWrapper, so we generate something based on the resourceName or, if that doesn't exist, the the relationship id, and if that doesn't exist we create /embedded-NUM.

But that's a separate issue.

We should use another option so that RecursiveParserWrapper knows to name the path /version-number-0 or similar. We should not misuse "resourceName".

Issue Links

• links to: GitHub Pull Request #1121