[Scythe42]

Logs show the following with 3.8.6

Example refreshing "1408 (2007).mkv"

Code:

21:12:27 T:3912   DEBUG: CScraperUrl::Get: Using "UTF-8" charset for "http://api.tmdb.org/3/search/movie?api_key=XXX&query=1408&year=2007&language=en"
...
21:12:27 T:3912   DEBUG: scraper: GetSearchResults returned <results><entity><title>1408</title><id>3021</id><year>2007</year><url cache="tmdb-en-2007.json">http://api.tmdb.org/3/movie/2007?api_key=XXX&amp;language=en</url></entity><entity><title>1408</title><id>3021</id><year>2007</year><url cache="tmdb-en-2007.json">http://api.tmdb.org/3/movie/2007?api_key=XXX&amp;language=en</url></entity></results>
...
21:12:30 T:11712   DEBUG: CGUIWindowVideoBase::ShowIMDB: user selected movie '1408 (2007)' with URL 'http://api.tmdb.org/3/movie/2007?api_key=57983e31fb435df4df77afb854740ea9&language=en'

So basically instead of the movie id it queries the movie with the ID 2007 in this example. And all the IDs in that 2000-2020 range are from the year 2007. That explains why ppl. see the year 2007, because only movies from 2007 like in this ID range.

It should have taken "3021" as the ID and not the year "2007".

So the the problem lies in GetSearchResults.

Given that the change was the order of items, the regexp has been updated, but not the reference in the URL to the new position of the ID. It sill uses 1, which of course now with the new order is the year.

Let's take a look at the current code, and yes, release date is at the 1st position and ID at the 2nd, but the URLs are still constructed with $1 instead of $2.

url cache="tmdb-$INFO[language]-\1.json">http://api.tmdb.org/3/movie/\1?api_key=

So replacing the 1 with a 2 in the three URL constructions in the GetSearchResults section, we should have a fix.

Quick test on my end shows that this seems to work and now uses the ID istead of the release date.

Code:

22:22:19 T:21452   DEBUG: CScraperUrl::Get: Using "UTF-8" charset for "http://api.tmdb.org/3/search/movie?api_key=XXX&query=1408&year=2007&language=en"
...
22:22:19 T:21452   DEBUG: scraper: GetSearchResults returned <results><entity><title>1408</title><id>3021</id><year>2007</year><url cache="tmdb-en-3021.json">http://api.tmdb.org/3/movie/3021?api_key=XXX&amp;language=en</url></entity><entity><title>1408</title><id>3021</id><year>2007</year><url cache="tmdb-en-3021.json">http://api.tmdb.org/3/movie/3021?api_key=XXX&amp;language=en</url></entity></results>
...
22:22:20 T:21396   DEBUG: CGUIWindowVideoBase::ShowIMDB: user selected movie '1408 (2007)' with URL 'http://api.tmdb.org/3/movie/3021?api_key=XXX&language=en'

Can some one try this fix and confirm it works for other as well?

So the devs can release a 3.8.7 as the 3.8.6 fix doesn't work properly, in case my assumptions (1st look at a scraper ever) is correct.