PATCH: zsh/pcre module docs

PATCH: zsh/pcre module docs

[Clint Adams]

This describes the behavior introduced by 15228.
Perhaps there should be a switch to disable
the clobbering of pparams, though I don't believe
perl has a similar feature.

Index: Doc/Zsh/mod_pcre.yo
===================================================================
RCS file: /cvsroot/zsh/zsh/Doc/Zsh/mod_pcre.yo,v
retrieving revision 1.1
diff -u -r1.1 mod_pcre.yo
--- Doc/Zsh/mod_pcre.yo	2001/07/02 19:39:35	1.1
+++ Doc/Zsh/mod_pcre.yo	2001/07/03 16:48:15
@@ -18,5 +18,9 @@
 item(tt(pcre_match) var(string))(
 Returns successfully if tt(string) matches the previously-compiled
 PCRE.
+
+If the expression captures substrings within parentheses,
+tt(pcre_match) will set the positional parameters to
+those substrings, starting with $1 for the first.
 )
 enditem()

Re: PATCH: zsh/pcre module docs

[Bart Schaefer]

On Jul 3, 12:51pm, Clint Adams wrote:
} Subject: PATCH: zsh/pcre module docs
}
} This describes the behavior introduced by 15228.
} Perhaps there should be a switch to disable
} the clobbering of pparams, though I don't believe
} perl has a similar feature.

Wouldn't it have been better to put these in the $match array (or into
some other named array) the way that backrefs in glob patterns do?

$1 etc. in Perl are not positional parameters -- you don't lose access
to @ARGV when you do a pattern match.  I think clobbering the positionals
is more harmful than helpful; if you want to be able to reference $1 etc.
you can always do

	pcre_match ...
	set - "$match[@]"	# or whatever it is named

For comlete generality, pcre_match could take the name of the array to set
as an argument; then the above could be accomplished with

	pcre_match -a argv ...

(where I've chosen -A somewhat arbitrarily, and another character might be
better).

-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com

Zsh: http://www.zsh.org | PHPerl Project: http://phperl.sourceforge.net   

Re: PATCH: zsh/pcre module docs

[Clint Adams]

> Wouldn't it have been better to put these in the $match array (or into
> some other named array) the way that backrefs in glob patterns do?

Yes.

> 	pcre_match ...
> 	set - "$match[@]"	# or whatever it is named
> 
> For comlete generality, pcre_match could take the name of the array to set
> as an argument; then the above could be accomplished with
> 
> 	pcre_match -a argv ...

Both are implemented, though now you can pass a second argument to pcre_match
that will be ignored.

Index: Doc/Zsh/mod_pcre.yo
===================================================================
RCS file: /cvsroot/zsh/zsh/Doc/Zsh/mod_pcre.yo,v
retrieving revision 1.2
diff -u -r1.2 mod_pcre.yo
--- Doc/Zsh/mod_pcre.yo	2001/07/03 16:52:35	1.2
+++ Doc/Zsh/mod_pcre.yo	2001/07/03 17:30:05
@@ -15,12 +15,13 @@
 matching.
 )
 findex(pcre_match)
-item(tt(pcre_match) var(string))(
+item(tt(pcre_match) [ tt(-a) var(arr) ] var(string))(
 Returns successfully if tt(string) matches the previously-compiled
 PCRE.
 
 If the expression captures substrings within parentheses,
-tt(pcre_match) will set the positional parameters to
-those substrings, starting with $1 for the first.
+tt(pcre_match) will set the array var($match) to those
+substrings, unless the tt(-a) option is given, in which
+case it will set the array var(arr).
 )
 enditem()
Index: Src/Modules/pcre.c
===================================================================
RCS file: /cvsroot/zsh/zsh/Src/Modules/pcre.c,v
retrieving revision 1.2
diff -u -r1.2 pcre.c
--- Src/Modules/pcre.c	2001/07/03 16:42:36	1.2
+++ Src/Modules/pcre.c	2001/07/03 17:30:05
@@ -90,8 +90,16 @@
 bin_pcre_match(char *nam, char **args, char *ops, int func)
 {
     int ret, capcount, *ovec, ovecsize;
-    char **captures;
+    char **captures, **matches, *receptacle = NULL;
     
+    if(ops['a']) {
+	receptacle = *args++;
+	if(!*args) {
+	    zwarnnam(nam, "not enough arguments", NULL, 0);
+	    return 1;
+	}
+    }
+    
     if (pcre_fullinfo(pcre_pattern, pcre_hints, PCRE_INFO_CAPTURECOUNT, &capcount))
     {
 	zwarnnam(nam, "error in fullinfo", NULL, 0);
@@ -107,9 +115,12 @@
     else if (ret==PCRE_ERROR_NOMATCH) return 1; /* no match */
     else if (ret>0) {
 	if(!pcre_get_substring_list(*args, ovec, ret, (const char ***)&captures)) {
-
-	    freearray(pparams);
-	    pparams = zarrdup(&captures[1]); /* first one would be entire string */
+	    
+	    matches = zarrdup(&captures[1]); /* first one would be entire string */
+	    if (receptacle == NULL)
+		setaparam("match", matches);
+	    else
+		setaparam(receptacle, matches);
 	    
 	    pcre_free_substring_list((const char **)captures);
 	    return 0;
@@ -128,7 +139,7 @@
 #ifdef HAVE_PCRE_STUDY
     BUILTIN("pcre_study",   0, bin_pcre_study,   0, 0, 0, NULL,    NULL),
 #endif
-    BUILTIN("pcre_match",   0, bin_pcre_match,   1, 1, 0, NULL,    NULL)
+    BUILTIN("pcre_match",   0, bin_pcre_match,   1, 2, 0, "a",    NULL)
 };
 
 /**/