in JavaScript/ECMAscript, nothing needs to precede a regex literal for it to be valid.
eg: /a|b/i is a valid regex and /a|b/i.test("baby") is valid use of it.
currently the HTML lexer differs from the JS lexer in the way they interpret regex literals, even having different preceding character requirements, it should be made uniform. i'm not sure who maintains each of these lexers, but if i have time i will try to submit a patch. might be a faster fix for someone more familiar with the codebase,
perhaps related to http://sourceforge.net/tracker/?func=detail&aid=1777016&group_id=2439&atid=102439
Leon
Bug fixes are generally contributed by people that have sufficient need to provide them. I won't be working on this myself.
following test case can be fixed by adding
>intosetOKBeforeJSRE.Marked as fixed by https://github.com/ScintillaOrg/lexilla/issues/250