search

xbmc / metadata.themoviedb.org.python

Other
46 stars 41 forks source link

bump to version 0.7.0 - candidate for release #20

Closed rmrector closed 4 years ago

rmrector commented 4 years ago

With the other three PRs this will be ready for final evaluation for submitting to the official add-on repo.

razzeee commented 4 years ago

Have you done a performance comparison with the old scraper? Can you share that data?

On Thu, 2 Jan 2020, 01:20 Ryan Rector, notifications@github.com wrote:

With the other three PRs this will be ready for final evaluation for submitting to the official add-on repo.

You can view, comment on, or merge this pull request online at:

https://github.com/xbmc/metadata.themoviedb.org.python/pull/20 Commit Summary

  • bump to version 0.7.0 - candidate for release

File Changes

Patch Links:

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/xbmc/metadata.themoviedb.org.python/pull/20?email_source=notifications&email_token=ABNLEZFTI35MZQDLECBMGDLQ3UXOPA5CNFSM4KB5OSO2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IDTDTLQ, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABNLEZHMJBODFB4E26X4NCTQ3UXOPANCNFSM4KB5OSOQ .

rmrector commented 4 years ago

Nothing concrete yet, but clearly slower on Windows. I'll put together a small test case and come up with a better answer soon.

rmrector commented 4 years ago

Huh. So the "clearly slower on Windows" had IMDB ratings disabled for the original XML scraper -- with that enabled like the Python scraper the Python scraper is notably faster... on Windows.

Quick time trials on Pi3 and Pi4 do show the expected slowdown, though. The setup is a single movie library with 50 items, scanned into a fresh library with each scraper 5 times, then I compared the average times.

Something else is probably up with Windows, the Python scraper shouldn't be faster. Trakt ratings will be responsible for a bit of the slowdown in the Python scraper - I didn't think to disable that for these tests, but I may try on one of the Pis.

rmrector commented 4 years ago

Commenting out the trakt ratings didn't significantly improve the Python scraper speed vs the XML.

However considering the speed improvement for the XML scraper when I disable the IMDB rating, it is probably a good idea to put in a switch to disable IMDB rating info for the Python scraper, too. And a switch for trakt since we're there.

KarellenX commented 4 years ago

While you are working on ratings... don't suppose you have had a change of mind for RottenTomatoes and Metacritic ratings?

Out of curiosity will you be working on the tvshow version of the scraper?

rmrector commented 4 years ago

Not really a change of mind - I'd prefer they go on a more complex scraper like the "Universal" scrapers to keep this one simple and focused, but since such a thing doesn't exist then if someone were to implement a PR for those cleanly (like @Razzeee did for Trakt) I wouldn't be against merging it here for now.

I have no plans for a TV scraper. I'm more interested in getting the above "more complex scraper" implemented for movies so I can see how much of Artwork Beef is still needed.

KarellenX commented 4 years ago

ok, thanks