I have database (csv) of the albums, with %index% as the unique record locator for each album. The same index is added to music collection with record locator %index%.
Now, in the database (csv) as I have %index%;%year%;%composer%.
I want to batch process the the entire collection to update %year% and %composer%, using %index% as the record locator in the music collection, instead of %_filename_ext% as the record locator. (music collection already has the %index%)
I can do work around, I can output the entire collection with %_filename_ext%;%index% to csv and merge (vlookup in excel) this file with main database to obtain %_filename_ext%;%index%;%year%;%composer%
But this would add additional step everytime, there is change in input database and additional record.
If there is a way to use %index% as the record locator, then it will easy solution.
If you succeed to get the files in the same order as the %index% in the database, then it should work.
E.g. you could sort the files in MP3tag by %index%
and do the same in the database file.
Ideally, the order should be the same in the database file and in files list of MP3tag.
That is the problem: you hardly have any other field/property that allows a unique relation between the file and the record. Even in your construction it does not work in the same way as the filename which has to be unique (if you consider a fully qualified filename).
You can mimic the converter "Textfile - Tag" by an action "Import text file" and some actions additionally.
This way, a relation between a user-defined tag field and the matching physical file can be done by a destinated group of actions.
In the first step the complete file has to be read into a temporary tag-field, having the list of all files in question. In the second step apply actions and fetch out data, which fits to the current physical file in work.
The content from the tagfield INDEX should build the unique relation to the related data within the file.
Search the Mp3tag forum messages, such requests have been solved in the past repeatedly.