Adjusted `xml2rfc` compatibility path implementation about bibxml-service HOT 3 CLOSED

Just to clarify that there are 2 types of mappings:

1. A path pattern mapping where a legacy filename path maps to a new citation document ID. This works for file paths that have a strong path pattern consistency, i.e. RFCs and I-Ds.
2. A manual file path mapping (which is technically the singular case for 1), where the legacy file path is not based on a consistent underlying path pattern, and that we need to make a one-to-one mapping from legacy file path to document ID. e.g. this will be the case for the ieee, misc datasets

And the most important thing is that the "legacy path mappings" should be considered configuration and changeable without needing code changes.

from bibxml-service.

Just to clarify that there are 2 types of mappings:

Yes, to use the original terminology, there is a manual mapping, and there is an automated path=>docid transformation attempted in absence of the mapping. We could call it automated mapping, I guess.

docid obtained as a result of automatic transformation may not yield an indexed citation, in which case we fall back to original xml2rfc tools website source contents.

And the most important thing is that the "legacy path mappings" should be considered configuration and changeable without needing code changes.

This depends on how we provide legacy path mappings. Originally I intended to provide a web GUI for that, but you mentioned some data analysis was being performed and I thought we might get a pre-made map as a result of that effort, in which case it’s probably not going to change.

from bibxml-service.

New implementation uses fetcher functions, mapped to relevant bibxml directories in URL configuration.

from bibxml-service.