Closes#254625

This feature switches fromu-mode (unicode) to the more powerfulv-mode (unicodeSets) regular expressions for regexp search/find.

To test manually: regexp search/find for[[a-z]--[f-v]] within a document containing the text "hello, world!". The letterse,w, andd will be matched:

RegExpOptions'sunicode is now a ternarytrue = u, "unicodeSets" = v, false/nullish = none. Hope that's acceptable as an interface, I could alternatively addunicodeSets as another boolean option (suboptimal asu andv are mutually exclusive) or change it to a union of string values (but then more changes to calling code would be needed).

This PR also aliases the existingescapeRegExpCharacters to the newRegExp.escape built-in API (available in Electron 37). This is necessary to ensure the escaping logic works correctly for all possible inputs in all contexts inv-mode as well as other modes:

constchars=' !"#$%&\'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~'for(constflagsof['','u','v']){for(constcharofchars){assert(newRegExp(`[${escapeRegExpCharacters(chars)}]`,flags).test(char))}}