- From: Pavithran <pavithran.a@ap.sony.com>
- Date: Sun, 7 Oct 2001 10:30:38 +0530
- To: "Christian Parpart" <cparpart@surakware.net>, "Philippe Le Hegaret" <plh@w3.org>, "Rowland Shaw" <Rowland.Shaw@crystaldecisions.com>, "Boris Zbarsky" <bzbarsky@MIT.EDU>
- Cc: <www-dom@w3.org>
Hi Parpart, I am also not understanding this namepsaceNormalization alg, I feel only auther can understand it,hope editor too, one thing i am very clear ,this algorithm is not covering all scenarios, even if u implement also it will not normalize...i am sure... i wasted almost 1 week, DONT give up, bye pavithran ----- Original Message ----- From: "Christian Parpart" <cparpart@surakware.net> To: "Philippe Le Hegaret" <plh@w3.org>; "Rowland Shaw" <Rowland.Shaw@crystaldecisions.com>; "Boris Zbarsky" <bzbarsky@MIT.EDU> Cc: <www-dom@w3.org> Sent: Monday, October 07, 2002 9:09 AM Subject: how to understand: void Element.normalizeNamespaces() > > WARNING: Unsanitized content follows. > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi all, > > as recommented by the current DOM Level 3 Core spec, I should > take a look into its Appendix B, Namespace Algorithms, wich > also describes the function void Element.normalizeNamespaces() > in pseudo code. > > Unfortunately the pseudo code is as pseudo as I wasn't able to > understand it fully. Furthermore, does there exist _any_ > implementation of that function so that I can look there to > understand? I didn't find it yet in Xerces C and J nor in current > Mozilla sources. libxml2 doesn't have it of course, too. > > The only starting point I had, was one of them having the interface > declared for void Document.normalizeDocument(). nothing more. I'd > be happy if someone can point me to any project who implemented > these namespace normalization. > > Am I the only one, trying to implement DOM Level 3 modules? I > implemented all algorithms as suggested by that namespace algorithms > appendix, but not that Element.normalizeNamespaces() one :( > > 1.) "if (Element's prefix/namespace pair (or default NS if no > prefix) are within the scope of binding)" > > "within scope of binding" is unclear. > does it mean both can be found using lookupNamespaceXXX? > > In that case I'd code: > if (lookupNamespacePrefix(prefix, false) && lookupNamespaceURI(namespaceURI)) > right? > > 2.) "there's a conflicting local default namespace declaration" > > What exactly is meant with conflicting? I have really no clue how to > implement this :( > > 3.) "if one ore more prefix bindings are available" > > A prefix binding is the corresponding namespaceURI, isn't it? so > there should be used the first one found starting from current > element up to the root. Am I wrong? Is that what is meant with > "binding"? > > Unfortunately namespace normalization seems not well documented > in DOM Level3 Core, or should it be part of Load&Save, or XML > (Namespaces?)? > > It's now the third day I'm trying to fight against my misunderstandings > caused by this pseudo code. Please help. > > Best Regards, > Christian Parpart. > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.0.7 (GNU/Linux) > > iD8DBQE9oQIJPpa2GmDVhK0RAkiTAJ9UxGnK1c0HaVDqbIE3jrFHPOK7hACfUFDf > osf6aPYJf4pkgIOo989EGyU= > =lKLE > -----END PGP SIGNATURE-----
Received on Monday, 7 October 2002 00:57:26 UTC