I am trying to run concurrent threads which each do their own TMDB lookups, however it appears that the python module is not thread safe and I can't find a way of creating a new instance of the import without actually copying the entire package or using multiprocessing which dupes the entire process space.
Has anyone else run into this problem/knows how to solve it?
Thanks!
No trobeu una pel·lícula o una sèrie? Inicieu la sessió per a crear-la.
Desitgeu valorar o afegir aquest element a una llista?
No en sou membre?
Resposta per Travis Bell
el 8 i de 2017 a les 6:58 P.M.
Hi @quietearth We don't maintain the libraries so you'll probably have better luck contacting the developer directly.
Unless someone else comes along who has experience with it of course...
Resposta per quietearth
el 12 i de 2017 a les 4:31 A.M.
For those searching from the future:
I cut out the tmdb3 caching backend (as it used a single file) and directly cached every entry by modifying put() and get() which is where all lookups are fed through (cache.py). Also, all tmdb3 lookups, eg searchMovie, require locks. This includes instance shortcuts such as title.imdb and title.alternative_titles which also performed additional lookups when accessed.