Rotten Tomatoes doesnt seem to work (1 Viewer)

[Kotik]

As the tittle says i dont think that rotten tomatoes score is working.

So i am adding a movie over moving pictures, then i am forcing imdb+ over it and the score doesnt change.

I tried manually changing the score and then re-running the imdb+ scrapper and it doesnt affect the score.

In other words, score always remain the same and its value is determined by the 1st scrapper used when u import a movie into the moving pictures DB.

I am using themoviedb.org as my primary scrapper and as so all my scores are from there.

Any suggestions?


EDIT: Really strange behaviour, now it is actually using imdb+ scraper to import movies BUT it imports a score of 0!!!!!



Edit 2: If you remove the themoviedb.org and imdb scrapers from the list then it will import a score of 0. If you dont remove them then it will always import the score from the themoviedb.org scraper.

This is really strange as it is showing me that it is indeed using the scraper from imdb+ when it actually trying to import the movie.

Edit 3: Something major is going on. I went ahead and removed only themoviedb.org scraper to see which score it will import this time, guess what...... this time it is using the imdb score.... since the 2 scrapers left on the list are imdb+ and imdb............ and if i remove the imdb scraper it is just importing a score 0!! I know which score it is importing cause i have 3 tabs open on my IE with the same movie over imdb, themoviedb.org and rotten tomatoes.


EUREKA!!!!!!!

11-Mar-2015 00:07:13 Warn [ WebGrabber]: Connection failed: Reached retry limit of 5. URL=http://www.rottentomatoes.com/alias?type=imdbid&s=0103772

How do i fix this one?

Edit 4: Bad news guys the IMDB ID feature seems to not work anymore due to RT removing? the feature

So right now u cannot link to RT while using IMDB IDs and here i am, wasting 4 hours to find out what is wrong.

http://developer.rottentomatoes.com/forum/read/110751/2

 

[RoChess]

They switched to a JSON model now it looks like.

Requested an API key, and added some new code to v4.9.42 to test it out.

Stay tuned...

 

[Kotik]

They switched to a JSON model now it looks like.

Requested an API key, and added some new code to v4.9.42 to test it out.

Stay tuned...

RoChess thank you for your hard efforts on this great plugin. Loving it so far.

Could you confirm that IMDB+ is reverting to the next available scraper when some data is not available on the default scraper?

Also another question, with themoviedb.org scraper most of my movies are being detected spot on. Now with IMDB+ there is an issue with most of my movies being detected totally wrong and i have to manually select the right choice. To give you some more information, what i noticed is that this is happening due to the fact that i am from Greece and as so IMDB+ is parsing a huge list of relevant movie names and within this list one of them is the Greek name of the movie, which 100% of the time is the right match. The plugin ofc doesn's know this ( and i dont expect it to know it) BUT how come themoviedb.org scraper is able to pick up the right one straight away?

It might have to do with the fact that the right one is never available for the IMDB+ since the right one is the Greek one.

Any thoughts?

 

[RoChess]

It might have to do with the fact that the right one is never available for the IMDB+ since the right one is the Greek one.

IMDb.com kinda sucks on many levels with all the changes they keep doing. Since I keep an eye on it when I import an English movie, it is easy for me to notice when that side breaks. What IMDb.com does for Greek users I don't know until somebody like yourself tells me it fails.

If you can supply me with the right information then I can have IMDb+ get 100% detection rate for you as well.

To assist in that there are some hidden options in IMDb+ config. Full instructions at: http://code.google.com/p/imdbplus/wiki/DebugIMDb

Try to (re)import just *ONE* movie that you know fails in IMDb+, as the reason that one movie fails is probably the same for all the others. That way the log file will be easier to read.

PS: Moving-Pictures automatically moves to the next scraper-script if the primary one can not find anything (or not all). This does make hunting down errors sometimes hard, and that is why on my own systems I got all the other scapers disabled. That way the moment IMDb+ fails I instantly know (for English info that is).

 

[Kotik]

Will supply you with logs as soon as i get some free time during the weekend.

 

[Kotik]

So here it is.

I removed all the other movies from the DB, enabled the debug options as per your instructions, disabled all the other scrapers and tried to import one movie.

The movie is Contact (1997) imdb id is tt0118884 and Greek name for the movie is "Επαφή".

Now to the point. I dont want the Greek name i want the English name of the movie and the result is that IMDB+ is defaulting to a wrong Contact movie with a release date of 1992.

 

[RoChess]

Yeah, to give you a backstory on the problem. IMDb.com supports localized titles. Meaning it knows your computer supports Greek language. So any movie that has an AKA title entry in your language will be shown as such. The problem is that nowhere does IMDb.com then let me know they are doing that, so from a technical perspective I have no idea which movies are English-title-without-Greek-title, Greek-title-shown-instead-of-English, Foreign-title-shown-as-is, Foreign-title-translated-to-English, or Foreign-title-translated-to-Greek.

As you can see from the "Contact (1997)" movie AKA page, there are two Greek title entries -- http://www.imdb.com/title/tt0118884/releaseinfo#akas

That is why when IMDb.com detects your system wants Greek titles, it will shown them as such, and IMDb+ will have to do its best to make sense of it.

Thought I had actually fixed this, but IMDb.com must have changed something again. Will now have to analyze the log your provided to see if it holds the answer for me to do just that.

Stay tuned...

 

[RoChess]

@Kotik, blimey, forgot the problem is in the search node, and I did not implement any IMDb+ debug code in there.

Please get me the raw HTML results (the actual notepad/file contents) that you get from: http://akas.imdb.com/find?s=tt&q=contact

and: http://www.imdb.com/title/tt0118884/releaseinfo

Ensure that your browser acts the same way that Moving-Pictures engine does, meaning that when you open http://www.imdb.com/title/tt0118884/ it shows "Επαφή" as title.

You are giving me the toughest example to fix, because there are a ton of movies called "Contact", and your filename does not contain IMDb tt-ID info (would have been instant correct match otherwise), or the correct year. Now the most popular movie called "Contact" is still shown at the top, but it is using the Greek title. That is why Moving-Pictures fails to auto-match with the import settings you are using, but it does find a match to the 2nd movie listed, which is "Contact (1992)".

Hopefully somewhere in those HTML sources lies the key that will allow me to find why my AKA insertion for 1st match found failed (which should have fixed this import), or maybe I can somehow detect a difference between the way IMDb.com shows the same HTML source to me as a USA-user.

I do love puzzles

 

[Kotik]

Ok if you love puzzles then how would we explain the fact that when i open the imdb link it gives me the English name?

Cause i am confused right now. I double checked it, it will use the English name for the link so somehow only IMDB+ is parsing the Greek one.

Edit 1: Did some more testing, with IMDB+ scraper as a match result i get Contact 1992, BUT with the default IMDB scraper i get Contact 1997 as the match, which is the right result. So how is the imdb scraper able to get the right one?

So i cannot actually post you html files since i cannot get the page to open in Greek language. Now that i double checked it my windows regional settings are all set to English (United States) so the only way for imdb page to detect my location would be my ip address. Even then how is imdb default scraper able to get the Enlish name?

Edit 2: IMDB+ scraper has way more non perfect matches when it is parsing the data and something important to notice, it doesnt have the English title nowhere in the list, so if you want to get the english title you have to scroll down the list, find the localised name of the movie, select it and import it and then it will actually use the English title.

The default IMDB scraper comes up with around 5 results, one of them is the right Enlish title and it picks that one correctly. Something important to notice here is that the default IMDB scraper never actually lists any localised movie name, so in my example it never comes up with "Επαφή" in its results list.

Hope that makes any sense.

PS. Can get you teamviewer access to my pc so you can perform the tests that you require over a Greek ip.

 

[RoChess]

If I had teamviewer access, this would make things a lot easier indeed. I can somewhat replicate the behaviour by adjusting the HTTP_ACCEPT_LANGUAGE tag in my browser, but trying to do the same via IMDb+ led to disasterous results before. IMDb.com uses like a combination of that setting, with other stuff. IMDb+ relies on Moving-Pictures to do the actual fetching of the HTTP data via .NET framework methods, and I guess it uses different OS based settings then what your browser is configured as.

Having direct remote access to a system not based in the USA might allow me to find a proper workaround for the time being. A full rewrite is in order, but that is just so time-consuming that I am forced to keep putting it off.

I will need to install "SQLite Expert Personal" program, and switch MediaPortal into debug mode, and Moving-Pictures into scraper-debug mode. That way I can make 'live' changes to IMDb+ to test out different theories to get the results from IMDb.com that are needed.