Re: Regexp Question

The following message is a courtesy copy of an article that has been posted to comp.emacs.xemacs as well. Hello George, taking a closer look I think you found a bug in XEmacs. Here is an enhanced test-case with my results included. Testers: Just keep on eval-last-sexp to insert your results on various platforms. (progn (setq reg1 "a\\(b\\(c\\)d\\|bc\\)" reg2 "a\\(\\(bc\\)d\\|bc\\)") (save-excursion (re-search-forward reg1) (goto-char (point-max)) (insert ";" (emacs-version) "\n") (insert (format ";reg1\\1:%s\n" (match-string 1))) (insert (format ";reg1\\2:%s\n" (match-string 2)))) (save-excursion (re-search-forward reg2) (goto-char (point-max)) (insert (format ";reg2\\1:%s\n" (match-string 1))) (insert (format ";reg2\\2:%s\n" (match-string 2))))) abc ;XEmacs 21.1 (patch 14) "Cuyahoga Valley" [Lucid] (i386-pc-win32) of Wed Mar 14 2001 on D5DC120J ;reg1\1:bc ;reg1\2:c ;reg2\1:bc ;reg2\2:nil ;XEmacs 21.4 (beta1) "Copyleft" [Lucid] (i586-pc-win32) of Tue Apr 24 2001 on D5DC120J ;reg1\1:bc ;reg1\2:c ;reg2\1:bc ;reg2\2:nil Best regards, Adrian George> I have this in a buffer (XEmacs 21.1.14): George> ---------- Buffer: foo ---------- George> (save-excursion George> (setq reg1 "a\\(b\\(c\\)d\\|bc\\)" George> reg2 "a\\(\\(bc\\)d\\|bc\\)") George> (re-search-forward reg1) George> (match-string 2))-!- George> abc George> ---------- Buffer: foo ---------- George> If I eval-last-sexp, it returns "c", but if I use reg2 George> instead of reg1, it returns nil. Is this how it should George> be? The left side of \| wasn't used in either match, was George> it? And so (match-string 2) should be nil for both? APA> Hi George, APA> to determine the matching subexpression, you have to count the number APA> of \( from the left, starting at one, match-string 0 returns the whole APA> match. APA> So, the \| is not what's relevant here, it just specifies an APA> alternative match inside the same subexpression. APA> Hope this helps, APA> Adrian APA> -- APA> Adrian Aichner APA> mailto:adrian＠xemacs.org APA> http://www.xemacs.org/ -- Adrian Aichner mailto:adrian＠xemacs.org http://www.xemacs.org/