VoIP Mailing List Archives

[murf at parsetree.com]

I'm not going to bore you with all the stuff I've done since

November here. I put it, and some examples, in the file

update1.txt in the git archive. To read it,


do a git clone of https://github.com/WyoMurf/SaySentence.git


I a nutshell, I've upgraded the SayScript grammar to handle

expressions in the file names, upgraded the current en, fr,

it, hu, and some others, to use the same approach as say.conf.
Upgraded the test suite. Finished converting Asterisk trunk
(mostly) to use SaySentence internally.


I'll make the Asterisk Community an offer. Somebody out there,

I hope, would like to have Asterisk in his/her language, which

is not yet coded into Asterisk. I offer to help you, whoever you

are, to develop a soundpack for your language. Here's the steps:


1. Get the en.template file in my github project. Copy it to the

   translation/<your lang_locale> example: rw_RW

2. Get the scripts that go with the trunk version for the sound files.

   (That's easy, just download any english core sound file set for
    Asterisk. It's in there, with the ending of ".txt".)

3. Translate the scripts for the various files. These are really

   a mixture of full sentences, and phrases (parts of a sentence).

4. Now, go to the translation file, with your scripts. Look for

   the non-trivial sentences, which usually involve stuff like

   '%' variable references, and more than one sound file. See if

   the sentence parts fit together acceptably. If not, rearrange

   under the [format] header for that utterance. Plan new files

   with corresponding sentence parts if necessary.

5. Now, Plan out the way your numbers, ordinal and cardinal, are

   spoken, and how the money for your locale is spoken. I will help

   you (for free) to generate a set of SayScripts for your language.

   We'll have to deal with issues like gender, or who-knows-what-else!

6. Run a bunch of tests to make sure the numbers, etc are generated

   properly.

7. Find a friend with a nice voice, and hand them the 400+ file

   scripts. If you can't find one, or can't afford one, find a quiet

   room, and record them yourself, as best you can. Edit the recording

   into separate files. Clean them the best you can. Get them in the

   best, highest quality format you can. 44+ kiloherz wav pcm files

   if possible. I'll get them into a good format; Asterisk trunk

   can handle sln32 or higher file formats. My intention is that

   the sound files in soundpacks will be a single best choice format.

   When installing sound packs, the files can be converted into any

   combination of formats you wish. We can use Asterisk itself to

   convert the files, or sox. All I know is that the higher quality

   the source sounds are, the better off you'll be converting them.

8. We'll form your sound files, scripts, and SayScripts into a single

   Sound Pack. Possibly the first one ever. I'll have to get busy

   and write some scripts to build sound-packs, and some scripts to

   install them.


I'd like to help with at least one new language, if possible, to see

what can be done to smooth the process.



Interested? Write me!



murf


--

Steve Murphy
ParseTree Corporation
57 Lane 17
Cody, WY 82414
✉  murf at parsetree dot com
☎ [url=tel:307-899-5535]307-899-5535[/url]