Recurring mistranslation: 'Sort' → 'Solr' (fuzzy) across 7 locales

[openlibrary-bot]

What happened

While translating untranslated strings in a 23-language batch (ar, as, bn, cs, de, es, fr, hi, hr, id, it, ja, ko, pl, pt, ro, ru, sc, te, tl, tr, uk, zh), I noticed the source string "Sort" (from merge_request_table/table_header.html) is translated as the literal word "Solr" — apparently confused with the Solr search engine name rather than translated as "sort" — in multiple locale files. The entries are marked #, fuzzy, so they aren't served, but they sit as latent bad data:

locale/bn/messages.po
locale/fr/messages.po
locale/ru/messages.po
locale/hi/messages.po
locale/pt/messages.po
locale/uk/messages.po
locale/zh/messages.po

Example from fr/messages.po:

#: merge_request_table/table_header.html
#, fuzzy
msgid "Sort"
msgstr "Solr"

The identical wrong output ("Solr") across 7 unrelated languages strongly suggests a shared bug in a prior bulk-translation pass (e.g. the translator conflated "Sort" with "Solr" as a proper noun and echoed it untranslated in every language), rather than independent per-language error.

Impact

Not currently user-facing (fuzzy entries are excluded from compiled .mo files), but the bad data persists silently until some future run clears/retranslates fuzzy entries.

Suggested fix

• When a future batch clears fuzzy entries for these locales, translate "Sort" properly instead of trusting the existing fuzzy string as a hint.
• Consider a one-off audit for other proper-noun-confusion patterns (e.g. other product/library names accidentally echoed verbatim) across locales, since this one was only caught by chance while sampling context strings.