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

Bug in emacspeak?



Precisely.

One yardstick to use when judging what to do about such irritants
is:

" how much time will it take me to program around this, 
what is that time-factor's pain equivalent, and finally, is the
pain caused by the irritant greater than that other pain" -- it's
a measure that leads to keeping software  avoid unnecessary bloat 
    Tim> If changing the regexp would result in degraded
    Tim> performance, I think it would be better to leave things
    Tim> as they are - its not a big issue really. I just thought
    Tim> we should have consistent behavior, but not at any cost.
    Tim> 
    Tim> Tim
    Tim> 
    Tim> T. V. Raman writes:
    >> 
    >> Here is what happens:
    >> 
    >> If there are no alphanumeric chars, it gets caught by the
    >> code in emacspeak-speak-line which matches various
    >> "decoration" rules. If you stick an alphanumeric in there,
    >> those patterns wont match, the text will make it to
    >> through to the speech server, and the processing defined
    >> in the server -- cleanup + tts processing happens.
    >> 
    >> I suppose it might make sense to make the decoration rule
    >> patterns smarter, but that will come at a cost, since
    >> those are regexps that are getting checked all the time,
    >> so keeping it simple, where it works in the majority of
    >> cases is the right thing from the big picture point of
    >> view.Tim> T. V. Raman writes:
    >> >> 
    >> >> I copied your lines of code into a foo.rb file so that
    >> it >> was in ruby mode, and yes, the third line of special
    >> chars >> is not spoken with punctuations set to some.
    >> >> 
    >> >> On the Dectalk Express it does produce a tone; I need
    >> to >> verify if/whether it produces a tone on the
    >> Viavoice.
    >> >> 
    >> >> Personally I dont think this is a bug --- which chars
    >> did >> you expect to be spoken out of the line: >>
    >> "#{$`}<<#{$&}>>#{$'}"
    Tim> 
    Tim> The reason I thought it may be a bug is because the
    Tim> behavior is inconsistent. If you add a alphanumeric
    Tim> character to that line, the dollar characters are spoken
    Tim> as well as the alphanumeric character.  Without the
    Tim> alphanumeric character nothing is spoken. I would have
    Tim> assumed that with punctuation set to some, the same
    Tim> punctuation characters should be spoken regardless of
    Tim> whether therre is an alphanumeric character as well.
    Tim> 
    >> >> 
 >  > Note that punctuation some in TTS is purely to preserve
 >  > intonation structure, and is not necessarily very clever about
 >  > selectively speaking subsets of chars out of a string like
 >  > above. 
 >  > 
 >  > But the bigger question: why would you program with punctuations
 >  > set to some? Anyone who does that will surely get what they deserve:-)
 >  >
 > 
 > Actually, this occured while reading a web page about ruby i.e. I was
 > not programming, but surfing the web. 
 > 
 > Tim
 > 
 >  
 >  > >>>>> "Tim" == Tim Cross <tcross@rapttech.com.au> writes:
 >  >     Tim> Dear all,
 >  >     Tim> 
 >  >     Tim> for a while, I've observed occasions when emacspeak
 >  >     Tim> refuses to speak a line which contains punctuation
 >  >     Tim> characters, but no alphanumberic characters even when
 >  >     Tim> punctuation mode is set to none.. I've not looked into
 >  >     Tim> this at all, but wanted to see if anyone else
 >  >     Tim> experiences similar behavior. I suspect its possibly a
 >  >     Tim> bug in the cleanup patterns used in the tcl script, but
 >  >     Tim> this is just a wild arse guess. I'm wondering if users
 >  >     Tim> with other synthesises experience the same problem (I'm
 >  >     Tim> using ViaVoice). this may help me track down what the
 >  >     Tim> error is due to.
 >  >     Tim> 
 >  >     Tim> Below is an example using a piece of ruby code. In this
 >  >     Tim> example, the 3rd line is not spoken unless punctuation
 >  >     Tim> is set to all. However, I think it should still speak
 >  >     Tim> some of the characters if punctuation is set to some,
 >  >     Tim> but this doesn't happen unless you also add some
 >  >     Tim> non-punctuation characters.
 >  >     Tim> 
 >  >     Tim> def showRE(a,re) if a =~ re "#{$`}<<#{$&}>>#{$'}" else
 >  >     Tim> "no match" end end
 >  >     Tim> 
 >  >     Tim> The reason I suspect this is a bug rather than a feature
 >  >     Tim> is because its behavior is inconsistent. For example, if
 >  >     Tim> you add an alphernumeric character to the end of the
 >  >     Tim> line - for example 'f', then the line is spoken with all
 >  >     Tim> the $ characters and the f. Once you remove the f,
 >  >     Tim> nothing is spoken, but I would have expected the dollar
 >  >     Tim> signs should still have been spoken (this is with
 >  >     Tim> punctuation set to some). I've also noticed that if you
 >  >     Tim> get this page spoken using page up and then page down,
 >  >     Tim> the dollar signs do get spoken, but if you move through
 >  >     Tim> the page using the down arrow or C-n, the line is
 >  >     Tim> blank. You also don't hear the indent spoken.
 >  >     Tim> 
 >  >     Tim> I'm hoping someone using a different synthesizer which
 >  >     Tim> uses a tcl script does not observe this inconsistency as
 >  >     Tim> this will give me something to compare and assist in
 >  >     Tim> debugging these reasonably complex regular expressions.
 >  >     Tim> 
 >  >     Tim> -- Tim Cross tcross@rapttech.com.au
 >  >     Tim> 
 >  >     Tim> There are two types of people in IT - those who do not
 >  >     Tim> manage what they understand and those who do not
 >  >     Tim> understand what they manage.
 >  >     Tim> 
 >  >     Tim> -----------------------------------------------------------------------------
 >  >     Tim> To unsubscribe from the emacspeak list or change your
 >  >     Tim> address on the emacspeak list send mail to
 >  >     Tim> "emacspeak-request@cs.vassar.edu" with a subject of
 >  >     Tim> "unsubscribe" or "help"
 >  > 
 >  > -- 
 >  > Best Regards,
 >  > --raman
 >  > 
 >  >       
 >  > Email:  raman@users.sf.net
 >  > WWW:    http://emacspeak.sf.net/raman/
 >  > AIM:    emacspeak       GTalk: tv.raman.tv@gmail.com
 >  > PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
 >  > Google: tv+raman 
 >  > IRC:    irc://irc.freenode.net/#emacs
 > 
 > -- 
 > Tim Cross
 > tcross@rapttech.com.au
 > 
 > There are two types of people in IT - those who do not manage what they 
 > understand and those who do not understand what they manage.
 > 
 > -----------------------------------------------------------------------------
 > To unsubscribe from the emacspeak list or change your address on the
 > emacspeak list send mail to "emacspeak-request@cs.vassar.edu" with a
 > subject of "unsubscribe" or "help"
 > 
 > -- 
 > Best Regards,
 > --raman
 > 
 >       
 > Email:  raman@users.sf.net
 > WWW:    http://emacspeak.sf.net/raman/
 > AIM:    emacspeak       GTalk: tv.raman.tv@gmail.com
 > PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
 > Google: tv+raman 
 > IRC:    irc://irc.freenode.net/#emacs

-- 
Tim Cross
tcross@rapttech.com.au

There are two types of people in IT - those who do not manage what they 
understand and those who do not understand what they manage.

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

-- 
Best Regards,
--raman

      
Email:  raman@users.sf.net
WWW:    http://emacspeak.sf.net/raman/
AIM:    emacspeak       GTalk: tv.raman.tv@gmail.com
PGP:    http://emacspeak.sf.net/raman/raman-almaden.asc
Google: tv+raman 
IRC:    irc://irc.freenode.net/#emacs

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


Emacspeak Files | Subscribe | Unsubscribe | Search