🚀 go-pugleaf

Can you post some code that exhibit this behavior?

"RK" <rkchats@yahoo.com> wrote in message
news:389ca595.0301060047.4034bfed@posting.google.com...
> Hi,
>
> I am using Xerces port for Mac from Apache. It's not an official port
> though.
>
> I am extracting DOMNodeLists for 4 different tags from a DOM document.
> I know these DOMNodeLists are "live" (pointers to the DOM tree). I am
> not changing anything, and am just getting some information out of
> those lists.
>
> But, the behaviour is not as expected. I get "unmapped memory
> exception" on the 2nd "doc->getElementByTagName(tag)" statement. I
> don't know why that (memory corruption) happens. Either there is a
> problem with my understanding of the usage of Xerces or there's a
> serious problem with Xerces on Mac.
>
> Could someone please help me here? I had decided to use Xerces for XML
> parsing in my project and it would be hard to back out.
>
> Thanks,
> RK

We switched from using Xerces to expat on the pc because expat ran many
times faster. It was pretty easy to switch, actually -- the structure of the
callbacks from both libraries are pretty similar. Btw, expat runs just fine
on the mac in case you're curious :)

RK wrote:
> Hi,
>
> I am using Xerces port for Mac from Apache. It's not an official port
> though.
>
> I am extracting DOMNodeLists for 4 different tags from a DOM document.
> I know these DOMNodeLists are "live" (pointers to the DOM tree). I am
> not changing anything, and am just getting some information out of
> those lists.
>
> But, the behaviour is not as expected. I get "unmapped memory
> exception" on the 2nd "doc->getElementByTagName(tag)" statement. I
> don't know why that (memory corruption) happens. Either there is a
> problem with my understanding of the usage of Xerces or there's a
> serious problem with Xerces on Mac.
>
> Could someone please help me here? I had decided to use Xerces for XML
> parsing in my project and it would be hard to back out.
>
> Thanks,
> RK

In article <%AhS9.1278$Af2.262617@news20.bellglobal.com>,
 "Jedidja Bourgeois" <jedidjab79@hotmail.com> wrote:

> We switched from using Xerces to expat on the pc because expat ran many
> times faster. It was pretty easy to switch, actually -- the structure of the
> callbacks from both libraries are pretty similar. Btw, expat runs just fine
> on the mac in case you're curious :)

Same here.  We are using it in a cross-platform product and it works 
just great.  Wrap it up in some C++ wrappers and it can pretty much drop 
in for your SAX implementation.

-- 

- rmgw

http://www.trustedmedianetworks.com/

----------------------------------------------------------------------------
Richard Wesley                                  Trusted Media Networks, Inc.

"Grownups have the most uninteresting explanations for things."
                 - C. S. Lewis, _The Magician's Nephew_

Here is some of the code I wrote. It's pretty simple. 

int ParserMain()
{
.......
 doc = parser->parseURI(xmlFile);
.......
    if(doc)
        {
          ParseHeader(doc);
          ParseFontTable(doc);
	  ParsePreferences(doc);
        }
.....
    parser->release();
    XMLPlatformUtils::Terminate();
}

void ParseHeader(DOMDocument *doc)
{
  XMLCh tag[] = L"header";
  unsigned int nH = 0, len = 0;
  
  DOMNodeList *HList = doc->getElementsByTagName(tag);

  if((nH = HList->getLength()) >= 1)
  {
	for(unsigned int i=0; i<nH; i++)
	{
	  DOMNode *firstChild = HList->item(i);
	  firstChild = (HList->item(i))->getFirstChild();

	  while(firstChild!=NULL)
	  {
	    bool bHasCNodes = firstChild->hasChildNodes();

	    if(bHasCNodes==true)
	    {
		DOMNode *tmp = firstChild->getFirstChild();
		if(tmp!=NULL)
		{
		  if(strcmp(tempNodeName, "id")==0)
		  {
		    XMLString::copyString(stHeader.id,
XMLString::transcode(tmp->getNodeValue()));
		  }
		  else if(strcmp(tempNodeName, "title")==0)
		  {
		    XMLString::copyString(stHeader.title,
XMLString::transcode(tmp->getNodeValue()));
		  }
		}
	    }
	    firstChild = firstChild->getNextSibling();
	  }//while ends
        }//for ends
     }//if ends
}

I have a few routines which are doing similar things. And execution of
2nd/3rd such routine gives me an exception.
__________________________________________________________


Also, I want to be very sure about moving to expat, if I have to. Any
comments on expat for Mac are also welcome.


-rk


Richard Wesley <hawkfish@trustedmedianetworks.com> wrote in message news:<hawkfish-D6AB2B.08053806012003@tribune.sj.sys.us.xo.net>...
> In article <%AhS9.1278$Af2.262617@news20.bellglobal.com>,
>  "Jedidja Bourgeois" <jedidjab79@hotmail.com> wrote:
> 
> > We switched from using Xerces to expat on the pc because expat ran many
> > times faster. It was pretty easy to switch, actually -- the structure of the
> > callbacks from both libraries are pretty similar. Btw, expat runs just fine
> > on the mac in case you're curious :)
> 
> Same here.  We are using it in a cross-platform product and it works 
> just great.  Wrap it up in some C++ wrappers and it can pretty much drop 
> in for your SAX implementation.
> 
> -- 
> 
> - rmgw
> 
> http://www.trustedmedianetworks.com/
> 
> ----------------------------------------------------------------------------
> Richard Wesley                                  Trusted Media Networks, Inc.
> 
> "Grownups have the most uninteresting explanations for things."
>                  - C. S. Lewis, _The Magician's Nephew_

We're using expat on pc/mac to parse our own XML files that store a variety
of data. Haven't had any issues and we're using the Unicode-enabled build on
both platforms. I can't see any reason not to use it on the mac :)

RK wrote:
> Here is some of the code I wrote. It's pretty simple.
>
> int ParserMain()
> {
> .......
>  doc = parser->parseURI(xmlFile);
> .......
>     if(doc)
>         {
>           ParseHeader(doc);
>           ParseFontTable(doc);
>   ParsePreferences(doc);
>         }
> .....
>     parser->release();
>     XMLPlatformUtils::Terminate();
> }
>
> void ParseHeader(DOMDocument *doc)
> {
>   XMLCh tag[] = L"header";
>   unsigned int nH = 0, len = 0;
>
>   DOMNodeList *HList = doc->getElementsByTagName(tag);
>
>   if((nH = HList->getLength()) >= 1)
>   {
> for(unsigned int i=0; i<nH; i++)
> {
>   DOMNode *firstChild = HList->item(i);
>   firstChild = (HList->item(i))->getFirstChild();
>
>   while(firstChild!=NULL)
>   {
>     bool bHasCNodes = firstChild->hasChildNodes();
>
>     if(bHasCNodes==true)
>     {
> DOMNode *tmp = firstChild->getFirstChild();
> if(tmp!=NULL)
> {
>   if(strcmp(tempNodeName, "id")==0)
>   {
>     XMLString::copyString(stHeader.id,
> XMLString::transcode(tmp->getNodeValue()));
>   }
>   else if(strcmp(tempNodeName, "title")==0)
>   {
>     XMLString::copyString(stHeader.title,
> XMLString::transcode(tmp->getNodeValue()));
>   }
> }
>     }
>     firstChild = firstChild->getNextSibling();
>   }//while ends
>         }//for ends
>      }//if ends
> }
>
> I have a few routines which are doing similar things. And execution of
> 2nd/3rd such routine gives me an exception.
> __________________________________________________________
>
>
> Also, I want to be very sure about moving to expat, if I have to. Any
> comments on expat for Mac are also welcome.
>
>
> -rk
>
>
> Richard Wesley <hawkfish@trustedmedianetworks.com> wrote in message
> news:<hawkfish-D6AB2B.08053806012003@tribune.sj.sys.us.xo.net>...
>> In article <%AhS9.1278$Af2.262617@news20.bellglobal.com>,
>>  "Jedidja Bourgeois" <jedidjab79@hotmail.com> wrote:
>>
>>> We switched from using Xerces to expat on the pc because expat ran
>>> many times faster. It was pretty easy to switch, actually -- the
>>> structure of the callbacks from both libraries are pretty similar.
>>> Btw, expat runs just fine on the mac in case you're curious :)
>>
>> Same here.  We are using it in a cross-platform product and it works
>> just great.  Wrap it up in some C++ wrappers and it can pretty much
>> drop in for your SAX implementation.
>>
>> --
>>
>> - rmgw
>>
>> http://www.trustedmedianetworks.com/
>>
>> -------------------------------------------------------------------------
---
>> Richard Wesley                                  Trusted Media
>> Networks, Inc.
>>
>> "Grownups have the most uninteresting explanations for things."
>>                  - C. S. Lewis, _The Magician's Nephew_

In article <389ca595.0301070412.70cf3c98@posting.google.com>,
 rkchats@yahoo.com (RK) wrote:

<snip> 

> Also, I want to be very sure about moving to expat, if I have to. Any
> comments on expat for Mac are also welcome.
> 
> 
> -rk
> 
<snip>

I used expat on a cross platform project a couple of years ago. I found 
it pretty easy to wrap it with the appropriate C++ code, since out app 
was in C++.

(Actually part of it also used JNI to call C++ which used expat).

At that time OS X was not stable or developed sufficiently to use it so 
we did everything under OS 8.*.

Before choosing expat I did examine and explore xerces but decided on 
expat for it's simplicity and sufficiency for our needs.

respect....


Peter