Imoprter imports movie File twice (1 Viewer)

[bta489]

Hi,

MovingPictures is acting a little weird on my machine. When I Import a movie from within the Configuration It often (not always) imports it twice, so that the movie consists of the same video twice in a row. I guess it occurs when I adjust the search parameters (searching title, year and IMDB-ID) while it is still searching, but I'm not sure how and why exactly.
This is very annoying since the duplicate file can't be removed. Additionally the Runtime in GUI is wrong, playback won't stop after the movie and it won't be recognised as watched.

I tried removing the duplicate files manually in the db (which kinda works, but surely isn't the best way) and tried to ignore the entire movie and import it once again.

I'm using MP 1.8.0, MovPic 1.7.2.1582 and FilmInfo+ as preferred Scraper.

 

[RoChess]

Just wait till it is done searching, and then fix the ones that failed, or adjust the filename instead so that you do not have to go through this in the future on a fresh import.

There is an advanced setting that controls what the runtime is, either from the mediafile itself, or via scraper-script. MediaInfo is used to obtain the runtime (as well as the other tech info) from the mediafile, but it is really bad on concurrent access, so if you force MovPic into a race condition with two imports on the same file (by default is allows 5 concurrent threads on importing), then MediaInfo might fail/crash. You might also need to update the MediaInfo DLL to match new container formats to read them correctly.

 

[bta489]

Thanks for the quick reply!

I usually have to correct every movie, since I like to append the language to the filename, which freaks the scraper out. The correct title is rarely recognised without adjusting the search-string. My patience is not always that big to sit through the more or less useless first search. I'm also waiting to adjust the language field manually (for language-logos).

The runtime-field in the database is usually correct, but the GUI seems to use another method of determining the duration. But that's a minor inconvenience. I simply divide suspicious runtimes (> 3 hours) by two and thats it

 

[RoChess]

If you can give me some examples of files that fail to auto-approve, I can help you modify the RegularExpression based MovPic noise filter, so you do not have to do anything manual anymore. If a computer is not making your life easier, you are using it wrong

Verify the bad runtime ones, more then likely you have grouping of files occuring then.

 

[bta489]

If a computer is not making your life easier, you are using it wrong

Amen to that! To be honest I'm pretty confused by the whole RegExp-thing. I read several tutorials and managed to get similar results to what I was hoping for on some occasaions, but I barely knew what I was doing. A little help would be highly appreciated.

I usually name movies only with the title and append abbreviations of the language in square brackets, like this:
"Some Random Title [DE]" or
"A movie with 2 languages [DE, EN]"

Should be pretty easy for someone who understands the maze of brackets and special characters.
Regarding the runtimes: My DB is pretty clean at the moment, so I don't have any wrong runtimes in there right now.

Thanks for the help. Another big step towards a perfect HTPC-Experience.

 

[RoChess]

That is easy to filter out then, just append the following to your noise filter:

|\s*\[[A-Z\s,]+\]

This will filter out any [EN], [EN, DE] or [AB, CD, EF] you might have, but it will leave alone [2014] or [tt1234567]

 

[bta489]

Cool.
The more I see working RegExp, the more I understand it
This seems to work perfectly. I re-imported some movies (which had missing information, so I had to reimport anyway) and they were recognised instantly. Thanks for the support!

 

[RoChess]

RegExp is actually pretty easy once you break it up into small groups. Expresso is an awesome builder/tester of expressions, and it also auto-breaks the long expressions up.

(...) capture group
(?<name>....) named capture group
(?:....) capture group, but excluded from results

As to the one for you:

|\s*\[[A-Z\s,]+\]

| = like an 'or', so A|B means look for A or B, so it can be added to the existing one
\s = space, \s* means look for 0+ spaces, if you want to force 1+, then it would be \s+
\[ = the actual '[' char, because [ is used in Regular Expressions you have to 'escape' it.
[...]+ = match all the characters within at least 1+, including ranges A-Z = A to Z, \s = space, "," = comma
\] the actual ']' char, because ] is used in Regular Expressions you have to 'escape' it.