Summary:	ASTERISK-03063: Incorrect pronunciation of certain numbers in Spanish
Reporter:	Enrique Zamudio López (ezamudio)	Labels:
Date Opened:	2004-12-21 18:58:00.000-0600	Date Closed:	2008-01-15 15:24:21.000-0600
Priority:	Minor	Regression?	No
Status:	Closed/Complete	Components:	Core/Internationalization
Versions:		Frequency of Occurrence
Related Issues:
Environment:		Attachments:	( 0) say.c ( 1) say-1.0.5.patch
Description:	The numbers from 101 to 199, as well as from 1000 to 1999, are incorrectly pronounced. For example 120 is pronounced "cien veinte" when it should be "ciento veinte". 1250 should be "mil doscientos cincuenta" but it says "uno mil doscientos cincuenta". ****** ADDITIONAL INFORMATION ****** I'm submitting a modified say.c with the corrections. This file is from Asterisk release 1.0.3. The changes are from lines 850 to 861. The sound files were already in the package I found in voip-info.org for Spanish, so the only thing missing was a condition for (num >100 && num < 200) and for (num >= 1000 && num < 2000) so that 101 is "ciento uno" and "1500" is "mil quinientos".
Comments:	By: Mark Spencer (markster) 2004-12-21 19:13:16.000-0600 Please supply a "cvs diff -u" off of CVS head, rather than the entire "say.c" file. By: Clod Patry (junky) 2004-12-21 22:22:02.000-0600 And when you'll submit the patch, confirm us you have send disclaimer to digium too. By: Clod Patry (junky) 2004-12-27 20:59:32.000-0600 ezamudio: any update here? can we see your patch? have you send your disclaimer to digium? Thanks. By: alric (alric) 2005-01-03 22:11:55.000-0600 Reminder sent to ezamudio Any update here?  Status of disclaimer? -- Housekeeping By: Mark Spencer (markster) 2005-01-09 03:56:16.000-0600 User appears to have lost interest in this patch. By: Enrique Zamudio López (ezamudio) 2005-02-03 16:54:01.000-0600 It's not that I have lost interest. I just need some help sending the disclaimer... if someone can point me to a sample file... I just downloaded version 1.0.5 and corrected some more stuff. It has to do with the sound files. Numbers from 31 to 99 were being assembled using three audio files ("treinta" "y" "dos", for instance, for the number 32) and it just doesn't sound right. So I got someone to record all the numbers again for me and modified the routine in spanish so that any number from 0 to 100 just uses its sound file, instead of trying to assemble the number like that. It sounds a lot better. I can diff my recent modifications against the stable 1.0.5 I just downloaded and tested; I'll post the patch but I need someone to guide me baby steps to send the disclaimer to digium. By: Enrique Zamudio López (ezamudio) 2005-02-03 16:58:56.000-0600 I also would like to note where can I upload these sound files, and what's the best format for it. I have the original AIFF files, so I can encode them as U-law or GSM... By: Mark Spencer (markster) 2005-02-04 01:11:41.000-0600 Fixed in CVS head with the generous assistance of David Corredor.  We used the files linked from the wiki for the Spanish sounds, but they seem to be recorded somewhat unevenly.  We did not change the code to use enumerated 30-99.  Please feel free to post your updated sounds at the Wiki.  Thanks! By: Russell Bryant (russell) 2005-02-13 17:36:06.000-0600 This is one of those tough calls to make... I think that since the patch that went in changes filenames in the code, I'm just going to leave this as one of the "known issues" in 1.0 to keep from breaking it for the people that have it working as it is.  The next version isn't too far off. By: Digium Subversion (svnbot) 2008-01-15 15:24:21.000-0600 Repository: asterisk Revision: 4964 U   trunk/say.c ------------------------------------------------------------------------ r4964 | markster | 2008-01-15 15:24:20 -0600 (Tue, 15 Jan 2008) | 2 lines Fix spanish pronunciation (bug ASTERISK-3063, thanks to David Corredor) ------------------------------------------------------------------------ http://svn.digium.com/view/asterisk?view=rev&revision=4964