describe bug, was: Re: [Maxima] First step toward a 5.9.2 rel ease



>>>>> "Vadim" == Vadim V Zhytnikov  writes:

    Vadim> As far as I understand Maxima and GCL info doesn't bound
    Vadim> to particular words in  - it just expects that first
    Vadim> character of  is [a-zA-Z].  So it should work well
    Vadim> for any Latin-alphabet language but not for Russian, Greek
    Vadim> etc.

Yes, you're right.  I thought the search was more specific.  In the
translated info files, does it change the "Function and Variable
Index" to something else?  If so, we won't be able to find the index
of functions and variables.

    Vadim> We can derive locale information from LANG environment
    Vadim> variable and if specific locale is present info
    Vadim> should try to use localized *.info files (if present) and
    Vadim> replace [a-zA-Z] by other pattern.

    Vadim> I'm also a bit surprised that info doesn't work with
    Vadim> Russian on sbcl and clisp - on Lisps with unicode
    Vadim> support but works more or less OK  on gcl and cmucl.

I suspect that nregex basically only understands ASCII.  So A-Z might
not match what you would expect in other languages.  That it works in
gcl and cmucl is probably just happy coincidence.

And with unicode sbcl and clisp, don't you have to open the files with
the appropriate external format to get it read correctly?  I doubt
maxima does that.

But I know very little about unicode, so I am probably all wrong.

Ray