BMT File Name Scraping

[wrems]

In reviewing your help doc I was going over the proper naming conventions BMT works best with. For TV scraping you have an example for SxxExx format and you provide some examples:

House s01e01.avi
House - S01E01 - Episode Title.avi
Lie to Me s02e03.avi
Lie to Me - s01e01 - Pilot Episode.avi

I've been formatting my archived content as a variation of your 2nd example:

House S01E01 Episode Title.xxx

Does BMT interpret the two as the same or should I be putting in the dashes between show name and then again between season-episode/title?

If you had to recommend a single convention to adapt to, which would it be for greatest hit ratio?

[stuckless]

Originally Posted by wrems

In reviewing your help doc I was going over the proper naming conventions BMT works best with. For TV scraping you have an example for SxxExx format and you provide some examples:

House s01e01.avi
House - S01E01 - Episode Title.avi
Lie to Me s02e03.avi
Lie to Me - s01e01 - Pilot Episode.avi

I've been formatting my archived content as a variation of your 2nd example:

House S01E01 Episode Title.xxx

Does BMT interpret the two as the same or should I be putting in the dashes between show name and then again between season-episode/title?

If you had to recommend a single convention to adapt to, which would it be for greatest hit ratio?

For TV files... the basic process is to look for S##E## and if I find that, then take everything on the left of that as the Series, and ignore everything on the right of it. The dashes don't matter at all... since after I parse the series, I then trim off all whilespace and non alhpanumeric charaters. So, a series like "House - S01E01 - Epsidoe Name" is really the same as "House S01E01 Epsidoe Name" as far as bmt is converned.

And on a side note... This is similar for movies as well. ie, if you have "Movie Name (Year)" that that is most optimal filename for bmt, since it will use the year in determining the best result. But you can also have "Movie (Year) Other Information" and bmt just ignores anything after the (Year). So if you have a director's cut edition of a movie you can do something like "Terminator 2 (1994) Director's Cut" and BMT will use "Terminator 2" as the movie, 1994 as the year, and ignore everything else.

One thing that I don't have, but I've been meaning to add is, adding a year to the filename scrapers for TV. So I could have, "Hawaii Five-0 (2010) S01E01" and then I'd know the "year" to use when narrowing down matches.

[wrems]

I figured the dashes were insignificant.

You know for TV naming what about the inclusion of the TVDB IDCode? Would that increase the likelihood of a perfect match?

Say something like this:

House (73255) SxxExx - Episode Title.xxx

Or, would the inclusion of the year instead of the IDCode yield the same result?

[stuckless]

Originally Posted by wrems

I figured the dashes were insignificant.

You know for TV naming what about the inclusion of the TVDB IDCode? Would that increase the likelihood of a perfect match?

Say something like this:

House (73255) SxxExx - Episode Title.xxx

Or, would the inclusion of the year instead of the IDCode yield the same result?

I can certainly consider something like that... but right now the scrapers don't currently support a GetID()... but to ensure that I don't confuse it with the year... i'd probably do something like tvdb#73255 and it might go in the "unused area", ie, "House S##E## - Episode Title (tvdb#73255)". It probably wouldn't be too hard to add something like that.

[razrsharpe]

Originally Posted by stuckless

i'd probably do something like tvdb#73255 and it might go in the "unused area", ie, "House S##E## - Episode Title (tvdb#73255)". It probably wouldn't be too hard to add something like that.

But why? with the mediatitles.xml file you can almost gaurantee a correct match with filename matching... I guess i dont see the point of dirtying up the filename with some information that you already have access to in the mediatitles.xml file...

[stuckless]

Originally Posted by razrsharpe

But why? with the mediatitles.xml file you can almost gaurantee a correct match with filename matching... I guess i dont see the point of dirtying up the filename with some information that you already have access to in the mediatitles.xml file...

I guess part of it... is that either I need to add a mediatitles editor to bmt to make it easier for people to manage it, or find some other ways to reduce the support. Right now, mediatitles is probably one of my biggest support headaches

But, yeah, I agree. MediaTitles does this now.

I have started to add better support for the mediatitles... I recently added the "Add Matcher" to the metadata editor what simplifies adding a new mediatilte entry (ie, a matcher) to the MediaTitles.xml file.

[wrems]

I’m having some issues... No surprises, right?

In regards to BMT 4.9.13.

I recently did some house cleaning on my server. Nuked everything. Wiz.bin and even the existing fanart, re-installed Sage, it was a completely fresh start. During this process I restructured my imports and ensured that my media files were properly named so that BMT could handle them on the first manual scan. That didn’t happen as I expected. The recordings that I moved are fine, as I followed the FAQ to a T.

Here are the issues I encountered with my imported/archived TV content:

Most if not all of my imported TV content was partially identified by BMT. For example, the show House. My naming convention for the file is “House S07E05 Unplanned Parenthood”. I would think this would be a pretty good match. This is just a random episode I chose. Here is how BMT imported the metadata:

Fanart Title - House
Show Title - Imported TV/House/House
Episode Name - House
Season # - 7
Episode # - 5

As you can see BMT scraped most of the file properly, but it hung up on the show title and episode name.

If I click Find Metadata and then click search I get an error that says No Results. I have to click Discover Defaults and then search and then choose the appropriate House from the list. Then it will be corrected.

The other method I’ve found to deal with these, because there are many, is to use the Batch Update and clear the fields for Episode Name and Show Title. Then re-scan that folder.

While the method works fine, I’m at a loss as to why I should have to manipulate any of the fields? I would think my naming convention should have sufficed and BMT would have matched everything automatically...

Am I doing something wrong?

[stuckless]

Originally Posted by wrems

I’m having some issues... No surprises, right?

In regards to BMT 4.9.13.

I recently did some house cleaning on my server. Nuked everything. Wiz.bin and even the existing fanart, re-installed Sage, it was a completely fresh start. During this process I restructured my imports and ensured that my media files were properly named so that BMT could handle them on the first manual scan. That didn’t happen as I expected. The recordings that I moved are fine, as I followed the FAQ to a T.

Here are the issues I encountered with my imported/archived TV content:

Most if not all of my imported TV content was partially identified by BMT. For example, the show House. My naming convention for the file is “House S07E05 Unplanned Parenthood”. I would think this would be a pretty good match. This is just a random episode I chose. Here is how BMT imported the metadata:

Fanart Title - House
Show Title - Imported TV/House/House
Episode Name - House
Season # - 7
Episode # - 5

As you can see BMT scraped most of the file properly, but it hung up on the show title and episode name.

If I click Find Metadata and then click search I get an error that says No Results. I have to click Discover Defaults and then search and then choose the appropriate House from the list. Then it will be corrected.

The other method I’ve found to deal with these, because there are many, is to use the Batch Update and clear the fields for Episode Name and Show Title. Then re-scan that folder.

While the method works fine, I’m at a loss as to why I should have to manipulate any of the fields? I would think my naming convention should have sufficed and BMT would have matched everything automatically...

Am I doing something wrong?

I am willing to bet that your House episodes had existing .properties in their directories. And if so, then bmt did not scan them. You can always verify if bmt is automatically scanning your items by openings the phoenix-metadata.log file and search for your filename. If you see SKIP in there, then bmt knew about the file, but decided to SKIP it. Typically files are skipped if the are accompanied by an existing .properties file. The reason for this, to not have BMT process files that have existing metadata (perhaps from mymovies, or if someone manually creates a .properties file).

The very fact you see a relative path in the show title is an indication to me that BMT didn't automatically scan the file, since BMT doesn't use relative paths in titles.

[wrems]

There would have been no .properties files to begin with. I had that option turned off from the BMT config area. However, I did forget to set that option after the re-install of everything. When I did realize that the properties files were created I made sure to change that option in BMT so that Sage would stop creating them, and then followed it by your batch operatoin to delete .properties files.

Having done those steps, shouldn't the next manual scan fix them? It didnt. The only way that I've found so far is to clear those fields and then do another scan.

Good to know though. Is it possible for the default BMT to be shipped with the option for .properties files creation to be turned off? Not a big deal just wondering since they aren't necessary anymore anyway.

[stuckless]

Originally Posted by wrems

There would have been no .properties files to begin with. I had that option turned off from the BMT config area. However, I did forget to set that option after the re-install of everything. When I did realize that the properties files were created I made sure to change that option in BMT so that Sage would stop creating them, and then followed it by your batch operatoin to delete .properties files.

Having done those steps, shouldn't the next manual scan fix them? It didnt. The only way that I've found so far is to clear those fields and then do another scan.

Good to know though. Is it possible for the default BMT to be shipped with the option for .properties files creation to be turned off? Not a big deal just wondering since they aren't necessary anymore anyway.

That .properties creation is not a bmt setting, but a core sagetv setting. I allow you to change it from the web ui, but I won't be changing it automatically during an install.

I would still check to see if you are seeing SKIP showing up in the phoenix-metadata.log, since whenever you see relative paths in titles, that's a good indication that bmt didn't scan the file. Also, if you manually scan the file, and it magically fixes that file, then that's another good indication that BMT didn't scan the file.

Once a file has been imported, bmt will never automatically scan that file again. If you do want it rescanned, then you'll need to either manually update the metadata using the web ui, or fire off a scan against the folder using the web ui, or delete and re-import the file.

[MacOCD]

Is there anywhere in the naming convention to allow "double episodes"?

I can't work out how to handle TV Episodes where 2 are run as a double length show? EG. ShowTitle.s01e01-e02 ?

From Time to Time I get this on my TV show DVDs. theTVBD lists them as separate episodes, but my DVD has them as a double length show. I'm happy to manually adjust it with BMT, I just don't know how to name them so it's clear in the Imported TV Listings in SageTV.

I do remember a previous thread on the forums about naming conventions but I don't seem to be able to find it now.

Mark.

[stuckless]

Originally Posted by MacOCD

Is there anywhere in the naming convention to allow "double episodes"?

I can't work out how to handle TV Episodes where 2 are run as a double length show? EG. ShowTitle.s01e01-e02 ?

From Time to Time I get this on my TV show DVDs. theTVBD lists them as separate episodes, but my DVD has them as a double length show. I'm happy to manually adjust it with BMT, I just don't know how to name them so it's clear in the Imported TV Listings in SageTV.

I do remember a previous thread on the forums about naming conventions but I don't seem to be able to find it now.

Mark.

bmt doesn't support multiple episodes. you'll need manually update the metadata to include both episodes, but you'll only be able set a single episode number.