* Nils Ohlmeier nils@iptel.org [061026 18:42]:
Hi Atle, On Thursday 26 October 2006 18:31, Atle Samuelsen wrote:
Very cool!
Thanks
Looking forward to tst it!
Feedback before testing it would to maybe make it also print the "most" logical request-uri changes.
(like, rewritehostport, lookup(".."))
You are right. That would make sense for sure. But the problem with that is that the script then somehow would need to know "statically" which functions might change the RURI. But I do not like such "static" approaches very much, because someone has to maintain them. But if someone comes up with such a solution that would be a big step in the direction to an SER GDB :-)
Hmm.. good point.. Would it be possible to maybe use a include file where this could be put in as a linebreak seperated list. Yes somebody would need to maintain it (if you want I could do it, it's really not that much work.. not much changes related to new functionames that rewrite the RURI)
BTW before solving the RURI problem, the first problem for route_graph is, that module functions might call a route. This is not solved yet.
Hmm.. What about putting them into the same file as I'm thinking of above. something like:
/*functionname;functiontype */ rewritehostport;0 functioncallingroute(x);1
- Atle
Nils
-Atle
- Nils Ohlmeier nils@iptel.org [061026 18:22]:
Hello,
in case you are not following the CVS commit log messages: I just checked in a small python script which should be able to print a dependency graph in ASCII of the route blocks of a SER config file.
When you try to understand more complex configuration files this tool might be helpfull. But be warned: this is still alpha code! It will not eat your config file, but the output might not meet your expectations :-)