encodings and cross references

Hi,

This is still about html cross references. First a proposal and then 
questions.

The current idea is that things like @ubaraccent{r} and the corresponding 
encoded character, whatever the encoding, lead to the same representation,
like _1E5E (_ followed by the corresponding unicode representation).

This implies that @ubaraccent{r} leads to a different cross reference than
r_, although currently makeinfo while doing info don't see that
@node @ubaraccent{r}
and 
@node r_
are different. 

This seems logical to me for 2 reasons. 

- some ascii representations of characters may make sense and thus 
shouldn't lead to the same node than the character (be it encoded or 
represented with @-commands). (for example 
@node o.
and 
@node @dotaccent{o}
should really lead to different nodes)

- the encoded characters and the same character represented with @-commands
should lead to the same cross reference. Thus it makes sens to represent them
with the same representation, namely a _ followed by the unicode representation
in hex.

You can comment on this, but if you agree there are some questions remaining:

1) It is possible to combine letters with diacritical marks to make any 
accented letter. Should this be used ? For example there is no direct 
representation of @H{a} in unicode. However it could be represented as a 
'a' followed by the representation of the double acute. 

2) Letters with two accents, can be represented in texinfo as, say
@'{@dotaccent{o}} for a o with a dot and an acute accent. Should those 
constructs lead to the corresponding character when it exists ? Or to a
combination with diacritical marks.

3) There are more than one representation for a given character in unicode.
For example there is the representation with diacritical combination, the
use of a given unicode representation, and they may be more than one
representation. It should be better to use just one representation, such 
that when there is an encoded character we know for sure what it will be 
in the resulting cross reference whatever process is used to convert it.
Should we specify such a thing, and then what representation should we use ?

Pat
_______________________________________________
Texinfo home page: http://www.gnu.org/software/texinfo/
[email protected]
http://ff0.org/mailman/listinfo/texinfo-pretest