Re: Pronunciation in emacs-w3m mode

["Lukas Loehrer" <listaddr1@xxxxxxxxxxx>]

Ok, a version of the unicode patch that works with today's emacspeak svn is linked from here:

http://homepage.hispeed.ch/loehrer/emacspeak.html

I cannot post a direct link because my ISP no longer allows them. The
above page also links to a test file that contains some special
characters.

The patch works with emacs 22 and possibly with emacs 23 before the
unicode branch merge. I will produce a patch for the new unicode
support in emacs 23 once it is more stable.

The patch requires the unicode data file. See documentation of the
variable describe-char-unicodedata-file after loading the library
descr-text.

The patch provides two new commands:

dtk-unicode-customize-char
dtk-unicode-uncustomize-char

They are ment for redefining the names of characters with long names
to something short. The replacement is only used with dtk-speak. The
long name will still be read when reading individual characters.
Unfortunatlly, the customizations are not saved automatically, because
I could not find out how to programatically do what the "Save for
future sessions" option in a customization buffer does.

There is another replacement facility provided by the variable
dtk-unicode-name-transformation-rules-alist. It is based on regexps.
For example it will cause names like "greek small letter XXX" to be
pronounced as XXX or "XXX symbol" as XXX.

I mostly tested this patch with the outloud speech-server. I did some
preliminary testing with eflite and espeak too.

Best regards, Lukas

T. V. Raman writes ("Re: Pronunciation in emacs-w3m mode"):
> post that patch anyway, if nothing else it should at least be
> checked into the contrib area so it doesn't get lost.

-----------------------------------------------------------------------------
To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
subject of "unsubscribe" or "help"