Jean-Marc Lasgouttes wrote:
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
The solution IMO is to detect and transform the quotes if a french
language is applied to a given snippets. Unicode makes a distinction
between quotes and french quotes (aka guillemets) so should we: this a
change in contents.
The distinction I was making in my message was between «quote» and
« quote ». The later form is used in french, but other countries use
the former.
I realized that after posting indeed.
Exactly and my opinion is that metrics and drawing should not
depending on drawing.
and not on metrics either ;)
I meant 'language' of course :-)
Yes, for french document, the quote character should trigger a lookup
for an associated opening quotes and act accordingly: insert opening
or closing guillemets. If we implement that, I don't really understand
why we would need an InsetQuote. I may be missing something...
I think you are not considering the same problem as the one I have in
mind.
Right.
The algorithm we use for opening/closing quote (are we after a
space?) is simpler than the one you propose, but I think it works
well. The question of using an inset or not is orthogonal to that. I
think the best reason is that we want to use double ascii quotes in
the most common case (english). Also, each babel language setting has
gorwn its own way of making quotes and I think we need to follow that
somehow. In this sense a quote is more on the semantic side, and plain
ascii my not cut it.
So, shouldn't we transform that to a full InsetText that will also
contains the text between the quotes?
Abdel.
