When looking for the ending of a string ("blah") or character literal ('\0xFEED'), TCSyntaxParser does limited handling of any preceding escape character (backslash), which may escape the ending quote (which in turn means it is not the end of the literal). To do this properly, an arbitrary amount of backtracking is needed. In particular, note that an odd number of preceding backslashes leaves the last one to escape the quote, while an even number of backslashes does not (in which case it is really the end of the literal).
The following screenshot shows the issue: bug on the right, and fix on the left. Note that in the window on the right, some of the "s" suffixes on the strings and character literals listed at the top of the file show up in red as part of the string, while these all show correctly in normal black text in the window on the left.

Although this issue was seen in TCSyntaxParser, both for string and character literals, it may also affect parsers for other formats, since this code is heavily copy-pasted.
(Note that the incorrect rendering of whitespace seen on the right is a separate issue [bugs:#581].)
Bugs: #284
Bugs: #581
Feature Requests: #252
Wiki: OWLNext_Roadmap_and_Prereleases
This issue was fixed in the overhaul of TCSyntaxParser [r7154]. See [feature-requests:#252].
Related
Commit: [r7154]
Feature Requests: #252