You can edit them if you open the files with a text editor.
Here is a explanation of the commands: https://docs.mp3tag.de/tag-sources
If you want to go into details, you also need some knowledge about regular expressions: https://docs.mp3tag.de/actions/replace-regexp
Since my script has 14 different versions, it's a good idea to use an advanced text editor where you can do replacements of the same text in different files in one go. Like this one: http://notepad-plus-plus.org/ .
For this editor, Notepad++, exists an very usefull language extension for the mp3tag web sources script language which highligts the comands of the script for a better overview: Web Sources Framework: Help needed
But you can also use simple notepad or wordpad. Edit one scirpt and copy and paste the whole [ParserScriptAlbum]=... part of the script to the others.
For your desired editing:
It's the outputto "..." command which determines which text goes to which tag field. So if you replace every outputto "Mixartist" with outputto "Composer", you get what was written to MIXARTIST before into COMPOSER now.
For the Credits part it's a bit trickier.
You have to split Credits and Notes, which are combined in on Field (UNSYNCEDLYRICS) by my script. But I predpared my script for this option:
replace
# Discogs Credits & Notes to Tag-Field UnsyncedLyrics
outputto "unsyncedlyrics"
with
# Discogs Credits & Notes to Tag-Field UnsyncedLyrics
#outputto "unsyncedlyrics"
and
replace
# Credits
#outputto "Discogs_Credits"
with
# Credits
outputto "Composer"
and
replace
# Notes
#outputto "Discogs_Notes"
with
# Notes
outputto "Unsyncedlyrics
(or whatever field you like here)
The next thing is to shift the Credits from a multline field to a oneline field. For this:
replace
regexpreplace "\s*<br />\s*" "\r\n"
with
regexpreplace "\s*<br />\s*" "; "
and
replace
say "Credits:"
sayNewline
with
#say "Credits:"
#SayNewline
(or delete these two lines)
Hm, hm, ...
after testing my instructions, I see there seems to be a problem with combining data from the tracks loop (a loop runing in the script which writes the same fields with different values for every track) and data from normal fields (which have the same value for every track) into one field. To Overcome this problem, I would have to integrate the Credits part into the Trackloop. Please tell me if it's worth the work, because the information given in the Credtis part at discogs.com is often much more as you want. E.g. infrmation about the designer or photographer of the cover artwork or similar stuff.
A different approach would be looking just for the composer instead of every person credit given on the page. You could only for certain expressions like "written by" "composed by" and write only these persons to the COMPOSER field. I don't know how much these expressions are standardized at discogs. But you can make a list of expressions which you want to include.
The problem of combining the information from the track loop and the credits part at the end would be still there.