As a content editor, I want to search for tags with or without diacritics and get the same results so that I can more easily find the correct tag.

[mrustow]

testing notes

• Try the below search and confirm it works as expected
• try adding a new tag with diacritics — tag should be saved with the diacritics removed
• confirm existing tags have had diacritics removed
• confirm tag variants with and without diacritics have been merged correctly

deploy notes

• solr reindex

---

On the admin interface > taggit, I typed hujra (no dot under the H) into the search bar and received this result:

However, I knew that #ḥujra is associated with 4 documents. So then I searched for ḥujra and received this result:

Lots of our tags don't use proper diacritics, so it would be better for us to be diacritic-agnostic.

Which we do seem to be already in the document view in the admin interface, and in searching the public site.

Thanks! 🥇

[rlskoeser]

@richmanrachel is it ok for tags to not allow diacritics? wondering if we can tackle this in tandem with #499 — collapse tags that are the same except for diacritics, and then add logic so that in future diacritics are ignored when creating tags.

[richmanrachel]

@rlskoeser - yes! Especially since Marina already requested this, I'm happy with this solution

[richmanrachel]

This didn't quite work either. The slug correctly took out diacritics, but the search doesn't include the slugs I guess? no results for mualim but returning correct result with diacritics:

[blms]

Ah—this is a problem with the "add tag" functionality in the Tags section of the admin, rather than adding a tag to a document. Looks like our diacritic-stripping code is not run when you create a tag separately!

[richmanrachel]

I didn't read through all the tags to ensure that all of the diacritics were properly dealt with, but I did look over about 400. The only potential error I saw was this one:

But if I only found 1/400 with errors, I think that's still enough to close as this gets us to a much easier place for tag cleanup. Do you agree, @rlskoeser ?

[rlskoeser]

@richmanrachel I agree with you — the near-duplicates I saw were variations like plural/singular, and I think those are ones that I think would be better to handle manually when we get the tag merge function implemented. Thanks for reviewing so carefully.

[richmanrachel]

@rlskoeser - perfect, closing now!