View Single Post
Old 12-13-2015, 12:19 PM   #239
trying
Member
trying doesn't littertrying doesn't litter
 
Posts: 21
Karma: 104
Join Date: Oct 2013
Device: none
Quote:
Originally Posted by davidfor View Post
I doubt adding the dash to the end of the URL will really help. I think it is more likely that when the blocking is happening, the Goodreads site thinks this is different URL. I would expect that it could get blocked and then the URL without the dash would work. Or you would need two dashes. You could automate this, try it and if there was a 403, add the dash and try that. I don't like that as there is reason that Goodreads has blocked the URL and we probably should not sidestep that

But, are people still seeing this problem? At the time it happened, there was another problem with getting related books. I was wondering if both were caused by a bad update to Goodreads.
I got the 403 errors two days ago after I had downloaded metadata using the v1.1.7 version of the plugin, noticed that I was missing Comments, updated to v1.1.10, and then redownloaded metadata. Once goodreads was blocking a bare book url I would, as Krazykiwi mentions, get 403 errors from a browser for a particular bare url and 200 okay when adding the dash. I could repeatedly get a 403 or 200 for that url depending only on whether a dash was added.

To test your theory, I just downloaded metadata for 288 books that already had a "goodreads:" id using my patch. It took 17m:34s (3.7s per book) with no 403 errors but I forget to turn off the Amazon metadata plugin. Redoing again with just the Goodreads plugin took 6m:24s (1.3s per book) but failed for 2 books but they were "No matches found with query" errors.

I then removed my patch, and downloaded the metadata for the same 288 books. It took 3m:38s (0.76s per book) but only successfully downloaded metadata for 57 books, and failed for 231 (with 229 "httperror_seek_wrapper: HTTP Error 403: Forbidden" errors). This matches the initial behavior that caused me to investigate the problem in the first place.

Just to be sure I put my patch back in, redownloaded (6m:20s, 1.3s per book), and again successfully got metadata for all the books for which metadata exists. So it seems your theory is wrong?
trying is offline   Reply With Quote