Re: Parsing XML file retrieved through HTTP

John Gentilin 1 Jul 2004 16:54:51 -0000

I have been doing this for a while very successfully, the only difference I see is that I actually parse the incoming stream into a DOM document then run the transformation from there. As a brute force test, just spool your input stream to a file, then run the transformation off the file. If it doesn't work, you can poke through the file with an editor to see where the problem was.

I had this issue before where the XML document I was retrieving was dynamically generated from a MS SQL server DB connection where the tables were populated through Access. The original content was generated in Notepad then cut & pasted into the Access table. The cut & paste action added the quote characters as special windows font value and the resulting parse threw an error "Single byte Unicode violation". Since the columns were declared as varchar and not nvarchar, I assumed it was Xalan on the server side... After about 4 hours of second guessing myself, I spooled the content to a file and use a Hex Editor to find the char....

HTH
John G

Kransen, J. wrote:

Yes, I saw the file through a browser, and I'm sure that it's valid. Also I
tried a completely empty XML file with only open and close tag. Either way
it parses all right when found as a local file, but the file as retrieved
over HTTP does not work. So I am quite sure that the problem lies there. I
think the HTTP headers are not stripped. If this is the case, can anybody
tell me a straightforward way of stripping the headers in Java? I'm sure I
don't have to write something common and low-level like that ;-) Maybe in
your case the header stripping was done by PHP? I sort of would expect that
from Java as well when I request an XML file over HTTP...

So far what I did was write a custom URIResolver like this, this gives
exactly the same result (same error as in original email):

transformer.setURIResolver (new URIResolver() {
        public Source resolve(String href, String base) {
                StreamSource source = null;
                URL context = null, url = null;
                try {
                        context = new URL(base);
                        url = new URL(context, href);
                        InputStream in = url.openStream();
                        source = new StreamSource(in, url.toString());
                }
                catch (MalformedURLException urle) {
                        //
                }
                catch (IOException ioe) {
                        //
                }
                return source;
        }
});

Since my code obviously doesn't do header stripping and gives the same
result, I strongly feel like headers are still there when doing a
StreamSource("http://localhost/risc/readonly.jsp";).

Any additional thoughts/help would be appreciated.

Jeroen

-----Oorspronkelijk bericht-----
Van: Russell Simpkins [mailto:[EMAIL PROTECTED]
Verzonden: donderdag 1 juli 2004 13:55
Aan: 'Kransen, J.'; [EMAIL PROTECTED]
Onderwerp: RE: Parsing XML file retrieved through HTTP

Jeroen,

Did you look at the xml file through a brower?  Maybe there is something
there you aren't seeing.  I have done the exact same thing, opening an
http page as a streamsource, using php as the source.  Every error I saw
came in malformed xml.  Which is what it sounds like you have.

Russ

-----Original Message-----
From: Kransen, J. [mailto:[EMAIL PROTECTED]
Sent: Thursday, July 01, 2004 5:16 AM
To: '[EMAIL PROTECTED]'
Subject: Parsing XML file retrieved through HTTP

Hello,

I have a .jsp that outputs a XML file. I have another .jsp in which I
want
to parse the XML file using an XSL file. So in the latter .jsp I want to
do
something like this:

<textarea><%

String xslFile = getServletContext().getRealPath("/xml/risc2cvs.xsl");

TransformerFactory tFactory = TransformerFactory.newInstance();
Transformer
transformer = tFactory.newTransformer(new StreamSource(xslFile));

// write the content of the parsed XML file
transformer.transform(new
StreamSource("http://localhost/risc/readonly.jsp";), new
StreamResult(out));

%></textarea>

However, when I do this, I get the following error message:

The element type "base" must be terminated by the matching end-tag "".

With this stack trace:
javax.xml.transform.TransformerException: The element type "base" must
be
terminated by the matching end-tag "".
        at
org.apache.xalan.transformer.TransformerImpl.fatalError(TransformerImpl.
java
:744)
        at
org.apache.xalan.transformer.TransformerImpl.transform(TransformerImpl.j
ava:
720)
        at
org.apache.xalan.transformer.TransformerImpl.transform(TransformerImpl.j
ava:
1192)
        at
org.apache.xalan.transformer.TransformerImpl.transform(TransformerImpl.j
ava:
1170)
        at
org.apache.jsp.cvs_005fuittreksel_jsp._jspService(cvs_005fuittreksel_jsp
jav
a:109)
..


When instead I try to parse a local XML file, there are no problems:
String xmlFile = getServletContext().getRealPath("/xml/temp.xml");
transformer.transform(new StreamSource(xmlFile), new StreamResult(out));

But then, when I try to parse the very same file as accessed through
HTTP, I
get the very same error:
transformer.transform(new
StreamSource("http://localhost/risc/xml/temp.xml";), new
StreamResult(out));

I was thinking that maybe the HTTP headers aren't stripped before the
parsing. Does anybody know what to do here?

Thanks in advance!

Jeroen

Re: Parsing XML file retrieved through HTTP

Reply via email to