Comic Rack Updating From Bottom Up

[SteveHutchinsASBFA]

Hello,

I'm not sure whether to label this as a bug report or feature request because I'm not sure how the program is supposed to be functioning. Please feel free to label this appropriately. This may also be two separate issues, but they seem like the same basic thing to me.

I had two instances recently where I tried to add information to ComicRack outside of the program itself and it failed.

1. The AI wrote me some Python scripts to inject custom field values "Width" and "Height" with appropriate information into the ComicDB.xml file. All of them failed. We wrote a script to determine the information was actually added to the ComicDB.xml file and it was. We also wrote scripts to validate the format of the XML information we injected and it was valid. Basically, all evidence suggests this should have worked, but the information will not show in the program.

Disclaimer: I don't know Python, so when I say "we", I don't really mean "we". It was all the AI (but it was a great learning experience for me, so, I guess there's a "we" there). I have a good, concise explanation from the AI as to our journey and I still have most, if not all, of the scripts, if you need any of that.

2. I've also tried to update ComicInfo.xml files via the Comic Tagger app. In ComicRack, I've clicked "refresh", made sure "allow writing of book info into files" and "book files are updated automatically" were engaged and even rebooted the program. None of these added the information to the program itself. The only thing that worked is deleting and re-adding the book. This isn't ideal, though, as any metadata outside of the ComicDB.xml would be lost.

So, basically, it seems like ComicRack has some issues importing data from the ground up.

Thanks,

--Steve

[maforget]

@SteveHutchinsASBFA I've converted this to a discussion thread instead.

1. I don't know what the AI gave you, but you can't simply add a Width element to the XML and expect it to magically appear in the program.
   • The program uses serialization to read XML files. Serialization is like a snapshot of an object in memory as a file in this case XML, but it can be JSON files. When it deserialize it reads the files and matches the relevant properties to the XML, if your Width & Height field don't exists it will just discard them (well not completely anymore, because I've added a special properties to catch all these extra values in case there are extra values from other software).
   • Custom values aren't simply added to the XML like a new element, they exists in the CustomValuesStore, search your DB for those they will be in the format of <CustomValuesStore>,comicvine_issue=133331,comicvine_volume=18867</CustomValuesStore>. This what you want to modify to add Custom values.
2. This has already been added in 90ad22d. Read the description for that commit. The program will keep a storage of files even for those not in the library and only in the folder tab. So that explains why you need to rename the file or close the program. With the addition I pointed to above you can now Refresh while pressing CTRL, this will force a refresh of the metadata from outside sources. I added the CTRL modifier so the users doesn't refresh by mistake and lose uncommitted changes.
   
   Refresh doesn't do anything for files in the folder tab (I've just changed this for next release to at least refresh the cover). For the rest it updates the cover and should re-scan the files, but it just outputs the file from the cached storage unless the files are named starting with id: with a GUID from the DB, not sure what the use is because it is impossible in Windows to have a colon in the filename. So it would work only on files in a container (aka the library) but not in the storage (Library). Might be used when connecting to a remote server, because otherwise it would always stop before it could actually refresh some info but even then seems like the remote library doesn't have the permission to Rescan, so it wouldn't hit that part of the code either. So unsure what the use case is where it would actually refresh the file (without the modifier).

[SteveHutchinsASBFA]

You are both brilliant and wonderful.

CTRL+Refresh works exactly as advertised. It pulls up and populates the data from the book file and is exactly what I was looking for. All I would suggest is maybe incorporating that into the menu(s) (even as a separate function, like "Refresh From CSV" or something, if you're afraid of changing the behavior of a current menu function).

For my second issue, your response allowed me to finally get a script written that will update the custom values with information from "ScanInformation". This will save me so much time, it's amazing.

Thank you so much for sharing your knowledge and your help!

--Steve

[maforget]

Wouldn't it be better to use Data Manager if your goal is to save Scan Information to custom values.

[SteveHutchinsASBFA]

Yes, you're right, and that was actually the first thing I tried. I guess there are currently issues with that plug-in and RegExVarReplace. It just didn't work no matter what i tried.

Now I found out that smart lists don't do number calculations very well or recognizing null values. So my current script populates the width/height custom fields, calculates an entire area and assigns a quality resolution based on that data (high/medium/low). Then I just set up the smart lists to pick that field up.

Not exactly a straightforward path, but I learn far more that way than the easy way, so it's all good.