Implement URI or CURIE functionality

[cthoyt]

The Converter.expand function turns a CURIE into a URI. This PR implements Converter.expand_or_standardize, which works like the normal Converter.expand, but if a URI is given, then it standardizes it and returns it. The other mechanics for "strict" and "passthrough" work the same.

This PR also implements Converter.compress_or_standardize as a counterpart for Converter.compress

Demo

The expansion and compression demos use a very simple extended prefix map:

from curies import Converter, Record

converter = Converter.from_extended_prefix_map([
    Record(
        prefix="CHEBI",
        prefix_synonyms=["chebi"],
        uri_prefix="http://purl.obolibrary.org/obo/CHEBI_",
        uri_prefix_synonyms=["https://identifiers.org/chebi:"],
    ),
])

Expansion

# Expand CURIEs
>>> converter.expand_or_standardize("CHEBI:138488")
'http://purl.obolibrary.org/obo/CHEBI_138488'
>>> converter.expand_or_standardize("chebi:138488")
'http://purl.obolibrary.org/obo/CHEBI_138488'

# Standardize URIs
>>> converter.expand_or_standardize("http://purl.obolibrary.org/obo/CHEBI_138488")
'http://purl.obolibrary.org/obo/CHEBI_138488'
>>> converter.expand_or_standardize("https://identifiers.org/chebi:138488")
'http://purl.obolibrary.org/obo/CHEBI_138488'

# Handle cases that aren't valid w.r.t. the converter
>>> converter.expand_or_standardize("missing:0000000")
>>> converter.expand_or_standardize("https://example.com/missing:0000000")

Compression

# Compress URIs
>>> converter.compress_or_standardize("http://purl.obolibrary.org/obo/CHEBI_138488")
'CHEBI:138488'
>>> converter.compress_or_standardize("https://identifiers.org/chebi:138488")
'CHEBI:138488'

# Standardize CURIEs
>>> converter.compress_or_standardize("CHEBI:138488")
'CHEBI:138488'
>>> converter.compress_or_standardize("chebi:138488")
'CHEBI:138488'

# Handle cases that aren't valid w.r.t. the converter
>>> converter.compress_or_standardize("missing:0000000")
>>> converter.compress_or_standardize("https://example.com/missing:0000000")

Known Use Cases

• Projects like LinkML that do not often make distinctions between URIs and CURIEs being in the same field would be able to use this functionality nicely
• SSSOM-py also implements its own version of this function https://github.com/mapping-commons/sssom-py/blob/ad70a83fb513e3d84a186bd500017471c4c9f7be/src/sssom/util.py#L1488-L1504 that can be replaced.
• Data science workflows that have started relying on "passthrough=True" could better use these updated functions to be more careful about keeping content that is valid w.r.t. the prefix map and also standardized.

TODO

• [x] I am looking for a better name for this before merging it.
• [x] Add docs for bulk operations
• [x] Add corresponding compress operation (after picking a better nomenclature)