Typo in organization name (“Istitute” of Electrical and Electronics Engineers)

[strogonoff]

https://github.com/ietf-ribose/relaton-data-misc/blob/1f9465d443de097c8be6dac02f1db2b74eddace0/data/reference.IEEE.802.1Qcc-2018.yaml#L29-L40

And the same situation as with #2, “publisher” is specified as IETF, although the document is IEEE. If Relaton data says who published the document itself, not who published a resource with bibliographic item about that document, then saying an IEEE document is published by IETF is an error.

[ronaldtse]

Thanks @strogonoff , there are several problems in this file (and this dataset in general)

1. The publisher is not IETF.
2. There is a typo in the organization name anyway
3. The IEEE documents are available in the relaton-data-ieee data set.
4. This misc dataset should be considered static now and no longer mirroring the legacy bibxml copy. We should now edit this dataset manually to correct errors.

[ronaldtse]

My earlier comment wasn't actionable...

I can resolve 4 by removing the GHA workflow from this repository.

Can 3 be resolved using the new mapping system? If that is done then 1/2 are no longer issues.

[strogonoff]

@ronaldtse Yes, https://dev.bibxml.org/management/xml2rfc-compat/ allows manually mapping xml2rfc paths and verifying XML responses. (It might not work reliably for bibxml3 on our demo infrastructure, I get error 500 sometimes, but it should work well for bibxml2.)

If we map some paths, we should export mapped path JSON in case DB is wiped at some point. See also docs. We could pre-map some path before delivering to IETF and provide an importable JSON file as part of deliverable.

Who’s the best person to do the mapping? As was established in the issue about IEEE.802.11 2012, I may lack subject domain knowledge to confirm that resulting XML represents correct documents.

[strogonoff]

@ronaldtse Alternatively, instead of mapping manually, we can just remove documents from bibxml-misc and adjust automatic logic for resolving bibxml-misc XML paths so that it does smarter parsing for filenames and has a higher chance of automatically finding IEEE references we may have indexed from other sources.

It should be confirmed by hand though, some of those IEEE XML filenames can be weird and not resolve automatically, requiring a manual map.

[ronaldtse]

@strogonoff right, but there is an additional consideration that there are only 94 IEEE entries, so a manual map is easier to do than having to tweak a translation pattern 😉

So for IEEE documents, we will:

1. Remove the IEEE documents from the misc data set
2. Create manual mappings for those documents to the ieee data set

OK?

[ronaldtse]

@strogonoff can I also remove all the files that are prefixed with _ now in the misc data set?

[strogonoff]

@strogonoff can I also remove all the files that are prefixed with _ now in the misc data set?

I don’t see any such paths here: https://github.com/ietf-ribose/relaton-data-misc/tree/main/data

[strogonoff]

If you mean bibxml-data-archive, if you remove prefix variations it will break fallback behavior (if path didn’t resolve) and xml2rfc path resolution checker tool (if IETF staff wants to confirm that _-prefixed paths resolve, they will not be shown). So no, do not remove them. The purpose of that archive is to have all paths for fallback and for path resolution validation. We are currently actually missing paths there in bibxml3 directory, and we should definitely not remove any paths.

[ronaldtse]

I don’t see any such paths here: https://github.com/ietf-ribose/relaton-data-misc/tree/main/data

Ah okay.

[strogonoff]

I checked with the XML file that seems to be the source for this, and there is no typo there (the abbreviation is not decoded at all).

Ping @ronaldtse @andrew2net

[andrew2net]

@strogonoff I'll fix it in the next relaton-bib release