I can look into the linter errors/warnings later, but I would prefer to first know that these changes are fine.

(
The test is not running the correct command, but rather it is running:

teslint "Primo" "Normalized" "XML.js"

instead of

teslint "Primo Normalized XML.js"

)

@retorquere -- I think you built that part of CI? Seems to struggle with filenames with spaces in it (see zuphilip's post above and travis details)

@adam3smith I'm looking into it

#1953

What's the rationale for the change in keywords/tags from search to display?

Currently, we add for example tags for all name forms of Germany which are part of the search section, e.g. https://primo-49man.hosted.exlibrisgroup.com/permalink/f/19ojnqi/MAN_ALMA21126560510002561 . It makes sense to have several name forms saved in the search section, because then all of them can be search in primo, but the normalized form is what you see in the display section.

Relatedly, do we generally import these long hierarchical tags and does that makes sense as a practice?

Do you mean the "Schlagwortketten" like Deutschland / Gerichtsberichterstattung / Elektronische Medien / Verbot / Verfassungsmäßigkeit? We could split this up into individual tags, as I would say is more what Zotero does. Should I do that?

@retorquere thanks!

@zuphilip Thanks for the explanation; makes sense on search vs. display and OK on the missing subtitle delimiter. Yes, I think splitting up the Schlagwortkette makes a lot of sense (and then fix any remaining linting issues)

Okay, I did split up the tags as you suggested. Please have a look. Linting is also done. I can squash later on.

Yes, I think that works better -- thanks! You can squash the PNX and Primo linting into one or leave them as two commits -- I don't have a preference on that either way (the two non-linting PNX ones need to get squashed for sure, though)

Squashed the firsts two commits together.

Thanks!