Re: [xml] c14n - UNICODE normalization

[Daniel Veillard <veillard redhat com>]

On Tue, Jun 03, 2003 at 09:58:54AM -0400, Tres Seaver wrote:
> On Tue, 2003-06-03 at 08:38, Rich Salz wrote:
> > > If I'm not mistaken, the c14n.c doesn't handle UNICODE normalization.
> >
> > I'm not sure what Unicode normalization is, but you're probably right. :)
> > The code in c14n (and exc-c14n) implement canonicalization useful for
> > XML Digital signatures. See http://www.w3.org/Signature for details.
> >
> > > It should be possible to use the normalization code from IBM's ITU
> > > (http://www-124.ibm.com/icu/index.html), as it's under the
> > > X License, but the ITU codebase is rather huge and I don't know
> > > how easy it would be to extract a small normalization module.
> >
> > I seem to recall some postings on xml-dev that a couple of folks tried
> > to do this and failed.  It's apparently not really feasible to take
> > "just a piece" of ICU.
>
> Zope3 does this.  You are right, it is the classic "take a sip from a
> spitoon" problem.

  So is there some not to big, not too hard to maintain sample code
to check unicode normalization, (and somewhat another routine for
doing this normalization). The very concern about the code size and
efficiency is what blocked unicode normalization checking in XML-1.1
and kept it an optional feature. That's an area where so good tight 
focused code is really needed (C of course).

Daniel

-- 
Daniel Veillard      | Red Hat Network https://rhn.redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/