[prev in list] [next in list] [prev in thread] [next in thread] 

List:       sr-dev
Subject:    Re: [Serdev] regexp stuff
From:       Andrei Pelinescu-Onciul <pelinescu-onciul () fokus ! fraunhofer ! de>
Date:       2003-08-25 14:18:08
Message-ID: 20030825141808.GB10706 () fokus ! fraunhofer ! de
[Download RAW message or body]

On Aug 25, 2003 at 14:39, Jiri Kuthan <jiri@iptel.org> wrote:
> Juha,
> 
> Andrei was kind enough to introduce support for such re-based
> replacement to the textops module in the development branch.
> The actions are called "subst" and "subst_uri".
> 
> -jiri
> 
> 
> At 07:21 PM 5/4/2003, Juha Heinanen wrote:
> >>i needed to write some regexp functions for the new version of enum
> >>module that i coudn't find in any c libraries.  basically the php
> >>function below is now done is c and it part of ser.
> >>
> >>i think it could be quite easily used to allow regexp replacement also
> >>in ser.cfg scripts.  may be someone can implement that in the next next
> >>release.


The difference between your code and mine is I do "pre-compile" both the 
 REs and the replacement. It's better for script use: I do it only once
 when the script is parsed and then I don't have to go char by char
 through the replacement string again. The results of subst_run is
  also "optimized" for easy conversion to struct lump.

Quick code usage example:
 se=subst_parser(expr); /* where expr.s="/ab(c+)x/\1\u\n/ig"
 rl=subst_run(se, input, msg);
 /* convert rl to lumps */

or to get a str with input after running the RE on it:
se=subst_parser(expr);
s=subst_str(input, msg, se);

Texops script usage example:


subst('#^From:#f#ig'); /* on the entire message except the first line
(uri) */


subst_uri('/(sip:)([0-9]+)@/\1+\2@/i'); /* add a + in front of numbers*/

Supported escapes:
 \n, \r
 \u - message uri (any other suggestions?)

Andrei


[prev in list] [next in list] [prev in thread] [next in thread]