Handling title `keywords`

[DaftGorillaz]

Is there a way to grab keywords? I don't know if this project uses the IMDB Data Dumps or scrapes from the website.

[tveronesi]

check pivot keywords is that you mean?

[DaftGorillaz]

Something similar when searching for films used in the Cinemagoer project

Used by the advanced search in IMDB

[tveronesi]

give me an example. a movie ID and the keywords you expect to find in the response please

[DaftGorillaz]

Case 1: I want to find the keywords associated with the film Superman (2025) get_keywords(tt5950044) it will return a list of keywords associated with this film [action hero, superhero, ...].

Case 2: I want to find films given a keyword search_keyword('super hero') this will a return a list of ids or titles ['tt5950044', 'tt20969586', ... ]

I would like to contribute to this project to add this feature.

[tveronesi]

thank you for your feedback
now it is clear !

[tveronesi]

so, as I mentioned earlier, the most relevant keywords (5) are already included in the response under storyline_keywords.
For example, in get_movie(tt5950044) we have:

...
  "storyline_keywords": [
    "superhero",
    "dc universe",
    "clark kent character",
    "superman character",
    "hero"
  ],
...

For now, I’ll keep it this way. Having more than 5 keywords isn’t very useful, since the additional ones are usually only voted up by a small number of people.

As for advanced search by keywords also , I’ll look into it later—it seems like a worthwhile enhancement.

For the advanced search i will look into it, as it is a good enhancement after all

[MiM-MiM]

Something similar when searching for films used in the Cinemagoer project

Used by the advanced search in IMDB

If you really want the advanced search features this should be using the actual graphql search and overall queries.
This is both going to be faster and more feature rich. You can filter a lot via the API compared to the web interface that this uses.
https://github.com/MiM-MiM/MyMovieGraphQLPy/blob/master/MyMovieGraphQL/Search.py
The one I made implements it completely. Keywords can be fully grabbed just calling update on the keywords until no more are found if you want all from a given title. Then searching it supports doing a search for keywords, even telling it that a title has to match all/any/exclude. The abstracted search doesn't allow you to do more than one type, but the actual graphql API would and you could still call the search directly with that fairly easily. I just didn't feel the extra arguments made sense to have with such a niche use case.

Personally I'd suggest a rewrite using the API over half and half.