Display the entire string for name subjects in browse list

[Kathy Lussier]

Bug 1662541 changed the subject browse indexes so that the browse list displays the entire subject screen rather than the individual components of the string. However, the change only applies to topic, temporal, and geographic subject indexes.

Similar changes were not done for name subjects because of punctuation issues. For example, without further development, a name subject for $aNapoleon$bI,$cEmperor of the French,$d1769-1821$xFiction would display as:

Napoleon -- I -- Emperor of the French -- 1769-1821 -- Fiction

instead of the expected display of

Napoleon I, Emperor of the French, 1769-1821 -- Fiction

We would like to see the code improved so that we can get name subjects to display in one string without the punctuation errors.

[Mackenzie Johnson (mtjohnsonupei)]

I don't have an exact solution, but from a MODS/XPath sense, if a <subject> element's first child is a <name> element, then when this subject node is parsed through biblio.extract_metabib_fields, we want the <namePart> children within <name> to be extracted and concatenated as part of the full heading string as usual, but for the "raw_text := raw_text || joiner" portion, there needs to be a test or trigger somewhere in the code such that either the concatenation ignores joiners between <namePart> elements, or that the concatenation only triggers on following-siblings of <name> and not children. Whether that can all be done through edits to biblio.extract_metabib_fields, or if it'd be better to create an entire new function for this bit of unique parsing.

Relatedly, there need to be changes to prevent relator terms in the $e field (which are held in <role><roleTerm> nodes in MODS) from being extracted to browse lists as well (though this has been addressed to some extent in https://bugs.launchpad.net/evergreen/+bug/1772981).