Hi, I realized that the problem was actually in the instance document of your example:
<note xmlns="http://www.w3schools.com" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://localhost/test.xsd"> If xsi:schemaLocation is used then the value must consist of (namespace name, document URI) tuples. Thus is should be: xsi:schemaLocation="http://www.w3schools.com http://localhost/test.xsd" However, we'll let the processor better report the following: 1) A warning if the value is malformed (Xerces reports a warning here; Saxon does not report it) 2) A warning if the document could not be located (Note that this is still no error, since the location is just a "hint" for the processor; i.e., the information is just ignored if no schema document is found) Regards, Kasimier > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Buchcik, Kasimier > Hi (what's your first name? Hei or Chan?), > > > -----Original Message----- > > From: Hei Chan [mailto:[EMAIL PROTECTED] > > > Hi Kasimier, > > > > Thanks for your quick reply. > > I didn't mention clearly that I used the same piece of > > C code you provided in > > http://mail.gnome.org/archives/xml/2005-March/msg00048.html > > to validate my XML. > > > > I was not using xmllint. > > > > I take a look at the bug (you listed): > > http://bugzilla.gnome.org/show_bug.cgi?id=157205 > > > > It seems to me that xmlSchemaValidateDoc won't work > > with xmlSchemaNewValidCtxtPtr(which has a NULL URL). > > > > Is it true? > > > > Thanks for your time. > > Ah. Well, it used to work, but since we don't have regression tests > for this, we might have broken it in the meantime without noticing. > I'll try to look into this today and will keep you informed. _______________________________________________ xml mailing list, project page http://xmlsoft.org/ [email protected] http://mail.gnome.org/mailman/listinfo/xml
