Yep, that's just it, the comma is used to separate different filter criteria. For example "*test1,*test2" will filter everything ending in either test1 or test2. The way of getting round this is (as ccowgill states) prefixing any control character with a reverse qoute (`). From the developer's help these are the control codes used in wcmatch (which are the same as that used in ssget:
Code:
Wild-card characters
Character Definition
#(pound) Matches any single numeric digit.
@(at) Matches any single alphabetic character.
.(period) Matches any single nonalphanumeric character.
*(asterisk) Matches any character sequence, including an empty one, and it can be used anywhere in the search pattern: at the beginning, middle, or end.
?(question mark) Matches any single character.
~(tilde) If it is the first character in the pattern, it matches anything except the pattern.
[...] Matches any one of the characters enclosed.
[~...] Matches any single character not enclosed.
-(hyphen) Used inside brackets to specify a range for a single character.
,(comma) Separates two patterns.
`(reverse quote) Escapes special characters (reads next character literally).