@sarahboyce I'm sorry, but I don't have time to review many proposals these days.

However, I have already tried to explain in the issues you linked the problem related to the search terms that are also stop words.

I'll try to add a some considerations that may help.

PostgreSQL full-text search is heavily based on the dictionaries of the language in which the search is performed, so switching to the “simple” dictionary makes you lose many of the features, not only the removal of stop words. See the PostgreSQL documentation.

Dictionaries allow fine-grained control over how tokens are normalized. With appropriate dictionaries, you can:

• Define stop words that should not be indexed.
• Map synonyms to a single word using Ispell.
• Map phrases to a single word using a thesaurus.
• Map different variations of a word to a canonical form using an Ispell dictionary.
• Map different variations of a word to a canonical form using Snowball stemmer rules.

https://www.postgresql.org/docs/current/textsearch-intro.html

BTW, the issue you're trying to solve in this PR only apply for searches in English, where a term in English is the same as the one used in the code. But English is only one of the various supported by the documentation, and using the “simple” dictionary could improve things for English but worsen them for other languages.

In fact, if you search for the word “through” in French you will get correct results, and we want to maintain this behavior for all languages ​​other than English.
https://docs.djangoproject.com/fr/5.1/search/?q=through

I'm not sure using the “simple” dictionary for all languages is the best solution.

Another solution I can think of to solve the problem with the English language and leave things unchanged for the other languages of the documentation is to create a custom dictionary for English. This involves removing from the stop words those that match names used in the code.

See https://www.postgresql.org/docs/current/textsearch-dictionaries.html

Here I had reported an example of creating a custom dictionary for French as an example, the procedure should be similar.

https://stackoverflow.com/a/47248109/755343