Reported by shadyar on 2013-04-09 10:23
in Farsi is usually used Arabic numbers,
as I choose espeak from the list of synthesizers and Farsi language and I try to navigate on Arabic numbers, character by character, NVDA does not read them correct, it adds a prefix "letter 6" to any digit. but
while I choose SAPI5 and choose Farsi language, NVDA reads them correctly.
I found out if "use spelling functionality" in preferences\Voices be not check, this problem is resolved. but why is there this problem in Farsi language on the Arabic numbers?
Comment 1 by jteh on 2013-04-10 01:53
This seems like another case of the eSpeak issue fixed for Chinese (ticket:3106#comment:7). This SSML reproduces it: <say-as interpret-as="characters">١</say-as>
Comment 2 by jteh on 2013-04-10 01:54
Changed title from "in Farsi language NVDA does not read Arabic numbers character by character correctly" to "eSpeak: in Farsi language NVDA does not read Arabic numbers character by character correctly"
Comment 3 by jonsd on 2013-04-10 23:47
Fixed now in eSpeak 1.47.05a at
This bug affected only non-ascii digits, as used in Arabic and Indian alphabets.
eSpeak will now speak the same name as the corresponding ascii digit. Is that correct, or do we need to distinguish between them?
I notice that Farsi can use 3 different symbols for a digit. For example, the number one can be ascii '1' or U+0661 or U+06F1.
Comment 4 by jteh on 2013-04-18 01:32
Milestone changed from None to 2013.1
Comment 5 by James Teh <jamie@... on 2013-04-18 02:10
```CommitTicketReference repository="" revision="98e0b9b90214021776c4336a14c259f989b4e990"
Update eSpeak to 1.47.06. Fixes #3132 and #3143.
Attachment app.php added by Slavon on 2014-01-13 04:51
Attachment func.php added by Slavon on 2014-01-13 04:53
Attachment main.inc.php added by Slavon on 2014-01-13 04:53
Attachment apoteksgruppen.zip added by Slavon on 2014-05-25 15:52
Attachment home-icon.gif added by Slavon on 2015-05-03 10:53