Summary:	ASTERISK-07121: [branch] bugs in AEL code
Reporter:	Steve Murphy (murf)	Labels:
Date Opened:	2006-06-07 18:31:43	Date Closed:	2006-06-18 16:36:39
Priority:	Minor	Regression?	No
Status:	Closed/Complete	Components:	PBX/pbx_ael
Versions:		Frequency of Occurrence
Related Issues:
Environment:		Attachments:
Description:	1. cidmatch spec in extension name (eg - 12223334444/8095551212 ) isn't working. 2. switch in switch emits bad code. 3. current ael.y file does not compile under bison-2.1a or higher. These probs are fixed in the teams/murf/AEL-trunk-fixesonly branch. ****** ADDITIONAL INFORMATION ****** Cleared up some problems-- 1. cidmatch in extension name ( 12223334444/8095551212 ) wasn't working--   this is because the code didn't put the cidmatch part in the add_extension2() call's   callerid arg. This is fixed. The stuff after the '/' goes into that arg.   To fix this, the ael_structs.h file was modified to add the cidmatch field in the   extension struct. 2. switch in switch wasn't emitting the right instructions. This was because   a variable (control_statement_count) was being incremented by a recursive   call to gen_prios(). This was happening only in switch() statements. A temp   local copy of this var is now set and used after the recursive call to gen_prios().   This fixed the problem. 3. The ael.y file wasn't 'compilable' by bison, because of an unused rule in the grammar.   I removed the unused rule. Some previous optimizations had introduced two more shift/reduce   conflicts. While I still have to investigate these, for now I have changed the %expect   directive to 7 instead of 5, and these changes now allow bison to process the ael.y file. 4. I added a regression test that covers both the cidmatch and the switch-in-switch   problems, and updated the rest to pass the checks.
Comments:	By: Steve Murphy (murf) 2006-06-07 18:59:33 Note- The problems with cidmatch and switch-in-switch were reported in the users mailing list last week. Many thanks for bringing these bugs forward!!! (Namely: Julian Lyndon-Smith,  Wednesday, May 31, 2006 1:21 PM, plus some personal correspondence) The prob with ael.y is something I discovered when I went to fix the bugs. By: Serge Vecher (serge-v) 2006-06-12 16:15:49 murf: could you please ask people from the asterisk-users list to test the code and report on it (I'm not subscribed to it). Thanks By: Russell Bryant (russell) 2006-06-18 16:36:38 merged into the trunk in revision 34665, thanks!