Conversation
kba
left a comment
There was a problem hiding this comment.
LGTM. This changes the fragment URL to point to headings though, so I'll add (invisible) anchors for the previous URLs.
| they changed the METS. This information is mainly for human consumption to get | ||
| an overview of the software agents involved in the METS file's creation. More | ||
| detailed or machine-actionable provenance information is outside the scope of | ||
| the processor. |
There was a problem hiding this comment.
...outside the scope of the processor, or just the spec?
The wording about "machine-actionable" should be revisited along with #108 BTW.
There was a problem hiding this comment.
...outside the scope of the processor, or just the spec?
This goes way back to 2018 and then meant "the module projects do not have to implement this".
These days I tend to run processors or ocrd process with |tee <name-of-processor>.log to retain the logs. Not actionable but more complete than either the mets:agent we add to METS or the pc:processingSteps we add to PAGE-XML, the latter being pretty close to being machine-actionable in the sense that they have the complete set of parameters.
We could add at least the processingStep mechanism to the specifications.
There was a problem hiding this comment.
We could add at least the processingStep mechanism to the specifications.
+1
Co-authored-by: Robert Sachunsky <[email protected]>
tboenig
left a comment
There was a problem hiding this comment.
see the typo with pdf ->tei,
otherwise I think it's good 👍
Co-authored-by: Robert Sachunsky <[email protected]>
Co-authored-by: Robert Sachunsky <[email protected]>
This is meant as a replacement for/supersedes 154 and 155 since the changes and discussion there became a bit fragmented and difficult to review.
It integrates all additions from both PRs into a new
mets.mdwith this structure:I hope I have not missed anything and that this will allow us to soon integrate these changes.