speechd-el/speech-dispatcher punctuation mode

[Pierre Lorenzon]

Hi,

I reccursively ask this question but it is not yet solved. When
mapping faces to voices all voices parameters are not set and
in particular punctuation-mode. I get following message in
speech-dispatcher log file :

    >>>  -- Message 

    [Tue Aug 14 15:17:36 2012 : 241349] speechd:     19:REPLY:|332 ERR NOT 
ALLOWED INSIDE BLOCK


    >>>  -- End Message

The fact is that speechd-el tries to set certain parameters
after a block has been open. In fact it should not because
`speechd--block-commands' constant inhibit trying to set
punctuation inside a block. However I modified this constant to
see what the problem was. In fact when trying to send such a
command inside block speech-dispatcher returns an error.


So my question is: Are there not too many block levels in
speechd-el or restricting commands to be allowed to be send
inside block in speech-dispatcher is maybe abusive?

For the latter I wonder why it is allowed to set the pitch and
other parameters like that and not punctuation mode and cap
recognition.


Anyway I wonder as well if so many nested blocks are necessary
in speechd-el.

Finally I should explain (what I always done in the past) why I
am requiring such features. My goal is to have LaTeX code
spoken in the most convenient way. One can distinguish two sets
of text in a LaTeX code. The text itself and formulas. It is
reasonable to speak text in a literary way whereas formulas
must be spoeken with all sines. Auctex distinguishes between
these two parts of buffer content by fontifying differently
text and formulas. So it seems very easy to make speechd-el to
speak them differently by mapping faces to different
voices. Except that voice speaking text should have
punctuation-mode set to some and voice speaking formulas should
have punctuation-mode set to all. Hence switching between these
two punctuation-modes should be possible and it is not ! 

Regards 

Pierre