FilmInfo+ - A german movie details scraper with auto grouping (5 Viewers)

[RoChess]

@badboyxx, lol, I just realized that MovPic does a conversion step that must take place before it processes the noise filter, and that is replace all (dots) to (spaces).

So adjust \. for \s

Filter out any optional spaces = \s*
At least one space, but include more = \s+
Specifically capture a number of spaces (two in the example) = \s{2}
Capture a range of spaces from 2 to 7 = \s{2,7}

 

[badboyxx]

@badboyxx, lol, I just realized that MovPic does a conversion step that must take place before it processes the noise filter, and that is replace all (dots) to (spaces).

So adjust \. for \s

Filter out any optional spaces = \s*
At least one space, but include more = \s+
Specifically capture a number of spaces (two in the example) = \s{2}
Capture a range of spaces from 2 to 7 = \s{2,7}

How do you mean it exactly? Can you give me an example please.

 

[RoChess]

Now that you provided examples you could simplify things a lot actually.

Filter out:

everything after "year", not including year.
WebRip and everything after it

That solves the issue for all the movies you listed.

RegExp needed = (existing-expression)|\sWEBRip\s.+|(?<=\s\d{4}\s).+

Problem solved

 

[badboyxx]

Now that you provided examples you could simplify things a lot actually.

Filter out:

everything after "year", not including year.
WebRip and everything after it

That solves the issue for all the movies you listed.

RegExp needed = (existing-expression)|\sWEBRip\s.+|(?<=\s\d{4}\s).+

Problem solved

Thank you Rochess, I will try it when I'm back at home.

Is it possible to redirect "und" that the importer is searching with "und" and "&" and "ae, ue, oe" as "ae, ue, oe" and "ä,ü,ö"? I hope you understand what I mean.

 

[RoChess]

Sorry, no idea what you mean with that last part.

 

[badboyxx]

In my collection e. g. the movie Arlo & Spot is stored with the filename Arlo und Spot... The importer can't find a match because of the difference "und" and "&". The importer is awaiting the filename as Arlo & Spot. Yes I could edit the filename but such case happens time and time again.
The same is with the umlauts ä, ü and ö. They can be stored in the filename as ae, ue and oe and then you have the problem to find a match.

So my question is, is it possible to tell the importer that it should search in such cases with the correct word. I hope I could explain it a little bit better now.

 

[RoChess]

Moving-Pictures already accounts for some of that itself, then the scraper-script (FilmInfo+ in this case) can add additional logic to do more intelligent searches, but in most cases this can all be solved by the original data source itself. The IMDb.com search results for example account for a lot of those situations automatically for me, but it sounds like the FilmInfo+ data source(s) do not have a smart search algorithm in place, so then it falls all on the scraper-script itself to do this.

That is a major undertaking though, but it is possible, and for all I know some logic already exists to account for this that has to be adjusted/expanded.

In short FilmInfo+ would have to do a search for title as-is, and if no results return to then search for "title with adaptive replacement #1", and if no results returned then to do a search with adaptive replacements #2, etc, etc... The risk is that this could add an enormous amount of extra retrieve calls that can cause issues and mass delays. It is therefore much easier if the data source handles this with a proper search method.

 

[Brudertac]

@RoChess
Thanks for your explaining Details regarding the Search Functions.
FilmInfo+ currently uses 4-5 Servers to get Movie Informations. As we have learned in the last Days, this makes it slow and vulnerable for Servererrors. (like Timeouts etc). I also don´t think that it would be a good Idea to let it also Search for alternate Titles.... (It does this already for some Functions, but only because it have to look on different Servers...)

 

[RoChess]

Yes, ideally it supports ID lookup, to where you rely on providing an ID alike to IMDb tt-ID for IMDb.com based scrapers that bypasses the whole search function, and goes straight for the correct movie.

I just don't know if FilmInfo+ data sources support that direct, or indirect. For example I use reverse IMDb tt-ID lookup at RottenTomatoes, moviemeter.nl (for Dutch language support), nyheter24.se (Swedish), and kvikmyndir.is (Icelandic).

That way I can rely on user providing IMDb tt-ID via filename, foldername, NFO file, or fall back on the excellent IMDb AKA search system to first locate the movie (even with foreign filename titles), and then locate any additional data on those alternative language sites.

So if such a reverse IMDb tt-ID lookup exists at any German movie info site, then I can add support for German language in about 10 minutes to IMDb+ (well re-add, because I had support initially via imdb.de itself).

 

[Brudertac]

FilmInfo+ also uses the IMDB AKA Servers. There is also a reverse IMDB tt-ID Search on ODFB.de available, BUT, they have very limited Movie Summarys. Thats where the other Backup Functions of Filminfo+ comes to work.

The thing is, currently i am not for 100% sure if there are non working Functions... I just began to correct the Code that is aged and does not work on some places. So, right now, the search works, but maybe i can make it "more stable" in future.