On 2013-08-20, Jean-Marc Lasgouttes wrote:
> 20/08/2013 01:03, Guenter Milde:
>> I cannot say when. Have a look inside t1enc.def, there you see definitions
>> like
>> \DeclareTextComposite{\`}{T1}{\i}{236}
>> \DeclareTextComposite{\'}{T1}{i}{237}
>> ...
> It looks like this happened in May 2001, and nobody told me... I'm
> shocked. One less LaTeX idiosyncrasy.
> So we should treat \'i and \'\i on equal footing in tex2lyx. I am not
> sure how to do that except via hardcoding.
Please be aware, that the Unicode <-> LICR mapping is not one-to-one.
While there is always one unique code point per Unicode character, there are
many characters with multiple LICR LaTeX macros.
This means that for tex2lyx, we should rather use an additional ressource
for the LICR alias -> Unicode mappings. We could either extend the
"unicodesymbols" syntax by adding an "aliases" field (remember, there can
be more than one alias for one character) or use a different "aliases"
file.
Günter
