263450 - nsIRDFRemoteDataSource.Refresh doesn't load files with other files extensions as rdf

Status: RESOLVED INCOMPLETE

(Core Graveyard :: RDF, defect)

Description

[Matthias Baumann]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.3) Gecko/20040910
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.3) Gecko/20040910

I have tried load an File as nsIRDFRemoteDataSource.
var remote = ds.QueryInterface(Components.interfaces.nsIRDFRemoteDataSource);

    try {
      remote.Init(src); // throws an exception if URL already in use
    }
    catch(err) {
      // loading already
      load = false;
    }

    if (load) {
      try {
        remote.Refresh((flags & JSLIB_RDF_FLAG_SYNC) ? true: false);
      }
      catch(err) {
        dump(err);
        return;
      }
    }

if src an uri like : file:///d:/test.rdf I get a remote Objcet with the URI
:file:///d:/test.rdf and lodded =true

if src an URI like file:///d:/test.sed (the file contains the same as test.rdf)
the methode remote.Refresh throw an err ( but no err.Message !!)
and the remote.loaded is still false !!

is it ab Bug or is there a other reason ?


Reproducible: Always
Steps to Reproduce:
1. see Details
2.
3.



Expected Results:  
load all files with an RDF inside regardless there file extension.

Comment 1

[Axel Hecht]

Sounds like a MIME type issue. I wonder if we trigger the sniffer for file://
URLs.
Could you try to load the .sed file in the browser itself? For me, this gives an
xml pretty printer on linux.

Comment 2

[Matthias Baumann]

here is the content of *.sed file:
<?xml version="1.0" ?><RDF:RDF xmlns:EBMTVC="http://www.ebmt.de/ebmt/mapping#"
xmlns:RDF="http://www.w3.org/1999/02/22-rdf-syntax-ns#"><RDF:Seq
about="urn:ebmt:data"></RDF:Seq></RDF:RDF>

The Browser can open this file.
Note: the encondig is not defined ! Is there the Problem ?

Matthias

Comment 3

[Boris Zbarsky [:bzbarsky]]

This is definitely a MIME type issue.  The parser doesn't get an XML type, and
then tries to use CNavDTD, which expects an HTML content sink, which we don't
give it (we give it an XML content sink).  So the DTD throws an error, which
propagates back.

I suppose we could force stuff to be parsed as XML, but I think we _do_ want to
throw errors on things like 404 pages and the like...  Note that an XHTML 404
page with the XHTML MIME type will get parsed as RDF with the current setup,
though.  ;)

Comment 4

[Axel Hecht]

This should get better once we load stuff right. I.e., we should probably throw
a rich error message like "wrong mime type", hinting an xml mime type 
nevertheless.