Summary: | ASTERISK-07894: [patch] say_alpha does not break when DTMF digit is keyed | ||
Reporter: | Patrick Baker (patbaker82) | Labels: | |
Date Opened: | 2006-10-07 00:58:12 | Date Closed: | 2007-07-11 19:59:23 |
Priority: | Major | Regression? | No |
Status: | Closed/Complete | Components: | Core/General |
Versions: | Frequency of Occurrence | ||
Related Issues: | |||
Environment: | Attachments: | ( 0) patch.diff ( 1) patch2.diff | |
Description: | 'SAY ALPHA thisisatest #' say_alpha does not break when DTMF digit is keyed. In this example when say_alpha is spelling thisisatest, it does not break and return when # is keyed. ****** ADDITIONAL INFORMATION ****** Asterisk SVN-branch-1.2-r43509M | ||
Comments: | By: Patrick Baker (patbaker82) 2006-10-07 01:26:21 tested with Asterisk SVN-branch-1.2-r44580M as well By: James Golovich (jamesgolovich) 2006-10-07 02:32:56 I was just going to open a bug on this but saw you had. Not sure if I have commit access anymore and I don't even have svn installed anywhere easily accesible so I can't generate a diff. The problem is in say.c all of the functions that do a ast_waitstream_full in a loop need to do: while (str[num] && !res) instead of just while (str[num]) the code is never breaking out of the loop when res is set, so it just drops your dtmf. By: Patrick Baker (patbaker82) 2006-10-07 11:11:01 Thanks James, I've made a simple patch for this and tested. By: James Golovich (jamesgolovich) 2006-10-07 11:14:23 There is more than one case that this needs to be applied to in say.c. I believe the original say digits one is ok, but when I looked I think I saw at least 2 functions that needed to be patched. I'm about to head out the door and won't be back till monday, but I wanted to get this comment in before I left. By: Patrick Baker (patbaker82) 2006-10-07 11:33:30 Yes you're correct, say_phonetic requires this as well. I have attached another patch that I've tested. By: jmls (jmls) 2006-11-06 05:05:22.000-0600 patbaker82, if you supply a patch you must have a disclaimer on file with Digium. Have you sent in a disclaimer ? If not, please do so and let me know. Thanks. By: Jason Parker (jparker) 2006-12-04 14:45:12.000-0600 Do we really want to change existing behavior with this? By: James Golovich (jamesgolovich) 2006-12-04 14:56:08.000-0600 The behavior was intended to be there, it just wasn't working correctly. Otherwise ints wouldn't be passed to ast_waitstream_full. The original say_ functions work this way, but the recently added ones do not follow the same behavior. By: Russell Bryant (russell) 2007-06-07 18:20:11 Fixed in 1.2, 1.4, and trunk in revisions 68351, 68354, and 68359. Thanks! |