Unicode has defined a few characters that may appear as commas, depending on the font used to display the glyphs.
Some accent characters can appear to be commas when using some fonts. A stand-alone cedilla, as on the c in façade, may look like a comma. There are also some other glyphs that might be mistaken for commas, but I forget most of them: they are rare for an American to see.
Unicode also has invisible characters. U+2063, an invisible separator, can be used to prevent this forum's software translating a word to BBS code or inserting smilies where you don't want them. Here's an example line where it's used to prevent the interpretation of code tags.
[code]These are not code tags.[/code]
I wonder what the original encoding was. A code editor with more capabilities than Notepad should be able to determine what is the mystery character and be able to search for it and delete it. I hesitate to recommend that, because most code editors I'm familiar with have a steep learning curve.
