speechd-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

sd_generic - a little idea


From: Bohdan R . Rau
Subject: sd_generic - a little idea
Date: Mon, 23 Aug 2010 11:27:50 +0200

On Mon, 23 Aug 2010 09:43:02 +0200, Hynek Hanke <hanke at brailcom.org>
wrote:
> On 21.8.2010 13:20, Chris Brannon wrote:
>>> GenericExecuteSynth should expect text on it's standard input, not as
>>> variable. And sd_generic module should use something like
(simplified):
>>>      
>> Yes, this is simple and safe.  I like it; a patch is forthcoming.
>> Can anyone think of a situation in which this would not work?
>>    
> 
> The current approach was more generic as it allowed
> you to pass the text as argument in case pipe mode is not
> supported by the synthesizer or create chains of commands.

Hmm...

some_strange_speech_command --text="`cat`"

won't work?

As for me the generic module needs some fresh look to be fully functional.

For example: why not use internal audio modules instead of external play
command?
Of course I know there are hardware synthesizer which uses it's own audio,
but in case of software synthesizer we wouldn't care which command shoud be
use to play audio (paplay, aplay, play, mynewsuperaudiosystemplay or
something).

Lets imagine something like:

GenericExecuteSynth "espeak --stdout"
GenericInputMethod "pipe"
GenericOutputMethod="pipe;format=wav"

Isn't is simpler?

BTW. using "system" to execute GenericExecuteSynth with pseudovariables
replaced with current values is _not_ good idea. Why not use execve
function and put these values into environment?

ethanak
-- 
http://milena.polip.com/ - Pa pa, Ivonko!



reply via email to

[Prev in Thread] Current Thread [Next in Thread]