NVDA-IBMTTS-Driver
NVDA-IBMTTS-Driver copied to clipboard
davidacm
Dates are read using month date year format regardless of how they're written.
Running the latest release version(22.07.3) As an example, if today's date is written as '25 July 2022', it is read as July 25th 2022. This is of course an issue if you are proofreading something and want to make sure the format is correct. It only happens if the first letter of the month is capitalised I'm not sure if this is a driver issue or with IBM TTS itself. Ideally I'd like to see options to turn off text processing.
This happens in Eloquence too. I think it's a library issue, rather a driver issue. I think that this could be solved, but I don't have the time these days... I agree with you, is a very annoying behavior
That's coming from Eloquence itself. It has its uses, but I can see it being annoying. It could probably be worked around by turning the months into lowercase with a regular expression. IBMTTS doesn't do that, but it has its own issues.
I wonder if we can actually disable some of the built-in filters, including that one? I know when I installed the IBMTTS binaries from IBM's FTP server I ran across something that was very obviously a mail filter for a few languages.
As far as I know, no, you can't. Those filters are an IBM thing that SpeechWorks never got.
On 8/1/2022 12:08 AM, Colton Hill wrote:
I wonder if we can actually disable some of the built-in filters, including that one? I know when I installed the IBMTTS binaries from IBM's FTP server I ran across something that was very obviously a mail filter for a few languages.
— Reply to this email directly, view it on GitHub https://github.com/davidacm/NVDA-IBMTTS-Driver/issues/73#issuecomment-1200673215, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADY4AYAIQ7ELEI3L2E54L63VW5EUXANCNFSM54T42UUQ. You are receiving this because you commented.Message ID: @.***>
I just ran a demo of JAWS to refresh my memory of how Eloquence included with that reads dates. Day month format e.g. 14 July is read in the correct order. I don't know if this is down to some text processing JFW does or the Eloquence version included with it. iOS eloquence does the same thing as NVDA, so I'm guessing it's text processing on JFW's end. Just thought I'd point this out in case it's something that could be implemented.
You might be able to get around that by adding a number sign (#) to the beginning of line 45 in ibmeci.py in the add-on directory, usually %appdata%\nvda\addons\ibmtts\synthDrivers. Open ibmeci.py in Notepad or another text editor, go to line 45, and add a number sign to the beginning of that line.