Re: Spanish vowels and accent table

[Carl Distefano cld@xxxxxxxx]

Reply to note from Correo Fenda  Fri, 16
Sep 2005 13:18:21 +0200

Manuel:

> But, if I attempt to verify a text written this way, XyWrite
> apparently reads acute vowels as separators. For example:
> "automático" is read as "autom" by the speller and it ignores
> the remaining letters. No problem with n tilde (lowercase or
> uppercase).

That's because, with no ACcent table loaded, func S1 encodes the
acute a, i, o and u as 3-byte (Speedo) characters, two of which (a
and i) have a separator as one of their constituent bytes. When the
spell-checker encounters the internal "separator", it's fooled and
thinks the word has ended when it hasn't. In your "automático"
example, as you noted, the 2nd byte of the Speedo acute a is hex 2E
= decimal 46 = "." The spell-check sees a full stop and thinks the
last word in the sentence is "autom".

The solution, as you saw, is to load an AC table that pairs the
unaccented vowels with their 1-byte accented counterparts. Loading
a SEparator table alone won't address the bug in the spell-checker
that causes 3-byte chars to be misinterpreted.

--
Carl Distefano
cld@xxxxxxxx