Manual Intervention

Information and Tutorials on features in Mylar and how to use it
Post Reply
Aphixx
Posts: 7
Joined: Wed Nov 29, 2017 3:28 pm

Manual Intervention

Post by Aphixx »

Hi,

Working through a new Mylar setup on a small Windows Server Essentials setup. Have been able to navigate through the setup issues and the key functions appear to be working now.

After scanning my current library the Import Results page gives me a series of entries with the status of 'Manual Intervention'. Can we capture what I should actually do in this case? Is this something I can take action on within the Mylar interface? Or do I need to change the filename? Or something else and re-run the scan?

I'm all in on manually intervening when needed - just need to know whats appropriate. Thanks.
User avatar
evilhero
Site Admin
Posts: 2883
Joined: Sat Apr 20, 2013 3:43 pm
Contact:

Re: Manual Intervention

Post by evilhero »

If you see the 'Manual Intervention' status listed beside the series, you do need to use the 'select' option located to the far right under Options. This will bring up an additional screen whereby you will be presented with the options that Mylar has narrowed it down to - which is based off of metadata first (if available), or the filename itself and the information presented therein (ie. as in Volume, Series Year, etc).

When the Manual Intervention appears, there were too many choices for it to auto-determine what was a match vs. what wasn't. Usually this is due to the naming of the file being too minimal (ie. something like 001.cbz would be a bad thing). Proper naming convention should included, at a minimum the series title and issue number. To help Mylar make better decisions and lower the chance of a Manual Intervention, including the Volume Year (ie. V3, or V2014) and/or the Issue Year will help significantly. If some of the files aren't getting scanned in, or are scanned in improperly (cause sometimes CV is a pain to match up to), you might have to manually fix the filenames, but unless your files are named like 001.cbz this shouldn't be required.

There is also a chance that the results being returned are due to the same title being present in different publishers - there are some marvel/dc publications that fall under the European publishers name and thus will be considered a different set. You can blacklist these publishers from ever appearing in a search/import result fairly easily. You do have to edit the config.ini file (with Mylar not running), and find the blacklisted_publishers field and then just enter in the publications to ignore, comma separated (ie. mine looks like this: blacklisted_publishers = panini comics, deagostini, Editorial Televisa, Editora Trama)

hope that helps!
Aphixx
Posts: 7
Joined: Wed Nov 29, 2017 3:28 pm

Re: Manual Intervention

Post by Aphixx »

It got me started yes. Thank you.

So now that the search providers is working Im trying to start importing my current collection. I'm trying to use the following file format:

The Uncanny X-Men (1981) #315.cbr

But mylar cant import. I get the 'there is more than one result that might be valid' message in the log. Is there a way I can provide additional detail to the filename to help? The file has also been scraped and populated with metadata by comicrack.

I'm happy to adhere to whatever detailed file name gives me the best chance of finding matches. Just need to know what that is. :D

Thanks again for all your hard work!
User avatar
evilhero
Site Admin
Posts: 2883
Joined: Sat Apr 20, 2013 3:43 pm
Contact:

Re: Manual Intervention

Post by evilhero »

Well Mylar won't read metadata that's not in a cbz container just due to the way the specifications are/were. Maybe at one point I'll add in that functionality, but the only proven method thus far is the cbz container just due to the fact that you can't add a zip comment into a rar (cbr), and the zip comment is what ComicBookLover metadata uses. ComicRack tho probably would work just fine with metadata in cbr, but tbh the development on that has stifled over the last little while (a few months shy of 2yrs without any dev).

Your filenames are fine, albeit I think I can see where the problem might lie with the series. You're using (1981) as the series year, but normally series years are preceded by the V to indicate a volume label (ie. V1981) as having 2 years in a filename can be confusing (if you did say The Uncanny X-Men (1981) #315 (1994).cbr).

Plus, in that series there's like 400+ issues that span over the course of a decade+, so when Mylar sees the (1981) it assumes that's the issue year, which would be causing the problems. If you wanted both series year and issue year in the title (as many do), you could do something like The Uncanny X-Men V1981 #315 (1994).cbr which would work just fine as well. Again having the Volume label designated is a clear indicator to Mylar which series on CV it will match to, and improve matches quite significantly.
Aphixx
Posts: 7
Joined: Wed Nov 29, 2017 3:28 pm

Re: Manual Intervention

Post by Aphixx »

This is great. Thanks for helping!

I'm building my workflow to get my current comic files organized properly so I can do a larger import, which is why Im trying to work out these steps. So for this particular batch example I have converted them all to cbz files (comicrack does that) and used the library mgmt addin for comicrack to re-organize the filenames based on the metadata. BTW, the lack of development on ComicRack is exactly why Im trying to pull out of it and get this new way of organizing going.

So now I have a batch of files with this new filename format and file container:

The Uncanny X-Men V1981 #313 (1994).cbz

Where the first date is the starting (volume) date of the series and the second year is the issue publish year. The first step of the import properly parses the '1981' into the Year column and the Comic Name comes in 'The Uncanny X-Men'. I still get the 'there is more than one result' message in the log and the status flips to Manual Intervention.

Other ideas? Can I somehow embed the actual comicid into the files or filename?
User avatar
evilhero
Site Admin
Posts: 2883
Joined: Sat Apr 20, 2013 3:43 pm
Contact:

Re: Manual Intervention

Post by evilhero »

Well with .cbz files if you have the 'Use Existing Metadata' option enabled, it will scan each cbz file for metadata and use that information instead of relying solely on name parsing. The problem is in this particular case is the general number of files you're importing of the series. The series spans over 10 years so by Mylar's current definition it will think anything between those year ranges would be a possible match - as it doesn't look at the issue numbering itself as a constraint to the matching. It's not uncommon to get Manual Intervention status for some series, for some it just can't be avoided, for others it should never happen. In this case, based on the size of the import for the series, I'd imagine that it couldn't be avoided without more code being added to account for issue numbering and comparing sequences, which would be a task in itself to implement/test properly without it messing things up ;)
Aphixx
Posts: 7
Joined: Wed Nov 29, 2017 3:28 pm

Re: Manual Intervention

Post by Aphixx »

Thanks. Im certainly not looking for specific development here - just trying to optimize whats already there based on my legacy catalog of files.

I will say I have found some success while using the manual post-processing vs the 'scan library'. In the example I mentioned above I was using scan library. But then tried to post-process the files and mylar was able to identify them, rename them to the mylar spec and move them to the right directory. If I have the 'move files' option enabled in scan library I'm making the leap that essentially there is no big difference between scan and post-process.

It may be obvious to you but from a user standpoint I find it interesting that post-processing did what I was trying to do but scan library did not.

Thanks for all your help though, its really appreciated.
User avatar
evilhero
Site Admin
Posts: 2883
Joined: Sat Apr 20, 2013 3:43 pm
Contact:

Re: Manual Intervention

Post by evilhero »

Aphixx wrote: Sun Dec 10, 2017 6:56 pm Thanks. Im certainly not looking for specific development here - just trying to optimize whats already there based on my legacy catalog of files.

I will say I have found some success while using the manual post-processing vs the 'scan library'. In the example I mentioned above I was using scan library. But then tried to post-process the files and mylar was able to identify them, rename them to the mylar spec and move them to the right directory. If I have the 'move files' option enabled in scan library I'm making the leap that essentially there is no big difference between scan and post-process.

It may be obvious to you but from a user standpoint I find it interesting that post-processing did what I was trying to do but scan library did not.

Thanks for all your help though, its really appreciated.
Well the thing is that Post-Processing will only post-process series that are on your watchlist, it will not add them and then post-process. Thus the reason why you would need to first do the import, get the respective series on your watchlist, and then any post-processing actions done thereafter can be targeted towards the watchlisted series.

Not having the move option enabled basically means Mylar will import the given series, and mark each issue that it can find a file for as Archived, but not touch the location of the original files. The other caveat is that with the import, meta-tagging is not performed at all - just renaming/moving (if the move option is enabled). Metatagging will only occur during post-processing.

Again, the post-processing of issues during an import is something that needs to get implemented, as well as a 'copy' option instead of just move (this was brought up just today as those that use torrents and seed stuff, need the copy option for seeding purposes).
Post Reply