From SheelR@aol.com  Wed Aug  1 08:06:12 2001
From: SheelR@aol.com (SheelR@aol.com)
Date: Wed, 1 Aug 2001 03:06:12 EDT
Subject: [Expat-discuss] is expat a reentrant parser
Message-ID: <66.1243c99b.28990464@aol.com>

Folks,
     I am planning to use expat parser in favor of xerces since 
     expat has a very small footprint when compared to xerces. 
     The questions that I have are 
        Is expat reentrant?
        Is expat thread safe?
      Also, with xerces , to process 40000 elements using SAX 
      ( size of xml buffer is roughly 2MB)
      takes 8 secs. I want to get it down to a couple of secs. Please advice.

~RR


From fdrake@acm.org  Thu Aug  2 18:36:11 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Thu, 2 Aug 2001 13:36:11 -0400 (EDT)
Subject: [Expat-discuss] is expat a reentrant parser
In-Reply-To: <66.1243c99b.28990464@aol.com>
References: <66.1243c99b.28990464@aol.com>
Message-ID: <15209.36747.450396.320026@cj42289-a.reston1.va.home.com>

SheelR@aol.com writes:
 >         Is expat reentrant?

  I'm not entirely sure what you mean here.  If you create several
parsers, you should be able to use them at the same time (from
different threads or whatever).  You probably don't want to muck
around with a parser while it is actively parsing, aside from calling
the defined APIs to add/change callbacks.

 >         Is expat thread safe?

  My answer above should be sufficient; if not, please explain what
you mean.  Realize that while Expat is parsing, it controls the flow
in its thread.

 >       Also, with xerces , to process 40000 elements using SAX 
 >       ( size of xml buffer is roughly 2MB)
 >       takes 8 secs. I want to get it down to a couple of secs. Please advice.

  Just measure it -- you didn't say anything about your hardware, so
no one has any way to judge, even if they've already done relevant
measurements (I haven't).
  If Expat is fast enough, that's great.  If not, there are other
parsers available.  I don't really have the resources to spend
performance-tuning Expat unless you want to fund the work through my
employer.  ;-)


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From josh.martin@abq.sc.philips.com  Fri Aug  3 17:49:28 2001
From: josh.martin@abq.sc.philips.com (Josh Martin)
Date: Fri, 3 Aug 2001 09:49:28 -0700
Subject: [Expat-discuss] is expat a reentrant parser
Message-ID: <200108031649.JAA17751@geocrawler.com>

This message was sent from Geocrawler.com by "Josh Martin" <josh.martin@abq.sc.philips.com>


>>SheelR@aol.com writes:
>>         Is expat reentrant?
>
>  I'm not entirely sure what you mean here.  If
you create several
>parsers, you should be able to use them at the
same time (from
>different threads or whatever).  You probably
don't want to muck
>around with a parser while it is actively
parsing, aside from calling
>the defined APIs to add/change callbacks.
>

<snip>

>  -Fred
>
>-- 
>Fred L. Drake, Jr.  <fdrake at acm.org>
>PythonLabs at Zope Corporation

Reentrant Functions:

  Reentrant functions (and functions which are not
interruptable
by signals) are defined as functions that may be
invoked, without
restriction, from signal-catching functions. A
function is reentrant
only if, when invoked inside a signal-catching
function, it does
not adversly affect the normal flow of operations
of the function
or code that the signal-catching function
interrupted. In other
words, reentrant functions aren't going to
unexpectedly change
any critical values, and thus the result of the
operations, if they
are invoked in the middle of a function.

  I'm not positive that this is the meaning that
SheelR was going
for, but this is the "standard" meaning of a
reentrant function.
Personally, I would like to know the answer to
this question,
given this definition of reentrant.

 - Josh Martin

_______________________________________________
Expat-discuss mailing list
Expat-discuss@lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/expat-discuss
		

Geocrawler.com - The Knowledge Archive



From mballen@erols.com  Fri Aug  3 19:16:38 2001
From: mballen@erols.com (Michael B. Allen)
Date: Fri, 3 Aug 2001 14:16:38 -0400
Subject: [Expat-discuss] is expat a reentrant parser
In-Reply-To: <200108031649.JAA17751@geocrawler.com>; from josh.martin@abq.sc.philips.com on Fri, Aug 03, 2001 at 09:49:28AM -0700
References: <200108031649.JAA17751@geocrawler.com>
Message-ID: <20010803141638.A842@nano.foo.net>

On Fri, Aug 03, 2001 at 09:49:28AM -0700, Josh Martin wrote:
> This message was sent from Geocrawler.com by "Josh Martin" <josh.martin@abq.sc.philips.com>
> >>         Is expat reentrant?
> >
> >  I'm not entirely sure what you mean here.  If

I recall seeing a message by James Clark that it is indeed reentrant
but unforunately I don't remember where. This could have been changed
of course. The sourceforge code is considerably different from Clark's
original expat-1.2 source.

Incedentally, how do these distributions differ? Is there a list of
features that have been added?

Mike


From fdrake@acm.org  Sat Aug  4 01:26:09 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Fri, 3 Aug 2001 20:26:09 -0400 (EDT)
Subject: [Expat-discuss] is expat a reentrant parser
In-Reply-To: <20010803141638.A842@nano.foo.net>
References: <200108031649.JAA17751@geocrawler.com>
	<20010803141638.A842@nano.foo.net>
Message-ID: <15211.16673.369393.251165@cj42289-a.reston1.va.home.com>

Michael B. Allen writes:
 > I recall seeing a message by James Clark that it is indeed reentrant
 > but unforunately I don't remember where. This could have been changed
 > of course. The sourceforge code is considerably different from Clark's
 > original expat-1.2 source.

  I would not expect that quality of the code to have changed
terribly, so I'll say that it *should* still hold.  However, I've not
analyzed it for that, and don't expect that I'll have time to do so.
  I suspect that it's also re-entrant using the thread-safe definition
of re-entrancy, but again, have not done a serious analysis of it.
  It would be great for someone to do the analysis, but I don't think
I'll have the time.  If anyone can say that these forms of re-entrance
are still allowable in the current codebase, we can be especially
careful to preserve that.

 > Incedentally, how do these distributions differ? Is there a list of
 > features that have been added?

  This has been asked many times, but I don't recall 1.2 well enough
off-hand.  Please submit this as a documentation feature request in
the Expat bug manager on SourceForge:

	http://sourceforge.net/projects/expat/

and I'll try to get to it in a few weeks, when my current swampage has
a chance at subsiding.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From mballen@erols.com  Sat Aug  4 04:11:36 2001
From: mballen@erols.com (Michael B. Allen)
Date: Fri, 3 Aug 2001 23:11:36 -0400
Subject: [Expat-discuss] is expat a reentrant parser
In-Reply-To: <15211.16673.369393.251165@cj42289-a.reston1.va.home.com>; from fdrake@acm.org on Fri, Aug 03, 2001 at 08:26:09PM -0400
References: <200108031649.JAA17751@geocrawler.com> <20010803141638.A842@nano.foo.net> <15211.16673.369393.251165@cj42289-a.reston1.va.home.com>
Message-ID: <20010803231136.A5568@nano.foo.net>

On Fri, Aug 03, 2001 at 08:26:09PM -0400, Fred L. Drake, Jr. wrote:
> 
> Michael B. Allen writes:
>  > I recall seeing a message by James Clark that it is indeed reentrant
>  > but unforunately I don't remember where. This could have been changed
>  > of course. The sourceforge code is considerably different from Clark's
>  > original expat-1.2 source.
> 
>   I would not expect that quality of the code to have changed
> terribly, so I'll say that it *should* still hold.

Maintaining reentrancey is not a matter of quality but rather actively
being conscientious about not modify data outside of the current stack
context. From glancing at the code it does look like everything is
carried with the XML_Parser object. At first glance something like:

void XML_UseParserAsHandlerArg(XML_Parser parser) 
{
  handlerArg = parser; 
}

Looks like setting a global but this, as well as many others like it,
is actually a macro that sets the associated member of the XML_Parser
object so it's ok of course.

#define handlerArg (((Parser *)parser)->m_handlerArg)

Mike


From garys@ihug.com.au  Tue Aug  7 06:29:39 2001
From: garys@ihug.com.au (Gary Stephenson)
Date: Tue, 7 Aug 2001 15:29:39 +1000
Subject: [Expat-discuss] NSTriplets work?
Message-ID: <000401c11f01$fb780620$fad7fea9@gateway>

Hi all,

I've just started work on interfacing expat to my Xbase++ XML processor.  I am
making excellent progress, and have learned a heap along the way - both about
XML and C coding!  Many thanks to all concerned, particularly JC <g>.  But....
(you just knew there was going to be a "but" somewhere didn't you?)

I can't get the function XML_setReturnNSTriplet( ..) to work for element
names - it only works for attributes.  Browsing the source code confirms the
observation - the only place where the parser's ns_triplets field is referred
to is within the storeAtts(..) function.

I am using expat (1.95.2) on Win2K.

I am not terribly experienced as a C programmer - but I would be happy to have
a go at fixing the problem if I could get confirmation that it really is a
bug - and not a feature <g>, and that nobody else is likely to be fixing it
anytime soon.

Many thanks in advance,

gary stephenson



From fdrake@acm.org  Tue Aug  7 21:36:12 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Tue, 7 Aug 2001 16:36:12 -0400 (EDT)
Subject: [Expat-discuss] NSTriplets work?
In-Reply-To: <000401c11f01$fb780620$fad7fea9@gateway>
References: <000401c11f01$fb780620$fad7fea9@gateway>
Message-ID: <15216.20796.802962.436108@cj42289-a.reston1.va.home.com>

Gary Stephenson writes:
 > I can't get the function XML_setReturnNSTriplet( ..) to work for element
 > names - it only works for attributes.  Browsing the source code confirms the
 > observation - the only place where the parser's ns_triplets field is referred
 > to is within the storeAtts(..) function.

  This is a known bug:

http://sourceforge.net/tracker/?func=detail&aid=231864&group_id=10127&atid=110127

but I have no idea when I'm going to have time to go code-diving after
this one.

 > I am using expat (1.95.2) on Win2K.

  The bug also exists in at least 1.95.1, if not earlier.

 > I am not terribly experienced as a C programmer - but I would be happy to have
 > a go at fixing the problem if I could get confirmation that it really is a
 > bug - and not a feature <g>, and that nobody else is likely to be fixing it
 > anytime soon.

  Any pointers to a solution would be welcome!  If you develop a
patch, please attach it to the existing bug report and I'll see about
getting it checked in.
  Thanks for your interest!


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From sam@uchicago.edu  Wed Aug  8 00:01:15 2001
From: sam@uchicago.edu (Sam TH)
Date: Tue, 7 Aug 2001 18:01:15 -0500
Subject: [Expat-discuss] Anyone using BeOS?
In-Reply-To: <15205.45464.632584.320171@cj42289-a.reston1.va.home.com>;
	from fdrake@acm.org on Mon, Jul 30, 2001 at 03:12:24PM -0400
References: <15205.45464.632584.320171@cj42289-a.reston1.va.home.com>
Message-ID: <20010807180115.B15677@uchicago.edu>

---------------------- multipart/signed attachment
On Mon, Jul 30, 2001 at 03:12:24PM -0400, Fred L. Drake, Jr. wrote:
>=20
>   Does anyone here have a BeOS machine?  I'd like to learn a little
> more about the appearant __declspec support on that platform, since
> past patches indicate that it differs from the MSVC support.
>   If anyone has a pointer to relevant documentation, I'd really
> appreciate it.

Sorry about the late response. =20

I'm the author of those patches.  I no longer have acess to the BeOS
machines they were tested on.  As I remember it, the fundamental
problem was that gcc, which we (AbiWord) use on x86 BeOS doesn't
support __declspec.  The Metroworks compiler does, I think. =20
          =20
sam th --- sam@uchicago.edu --- http://www.abisource.com/~sam/
OpenPGP Key: CABD33FC --- http://samth.dyndns.org/key
DeCSS: http://samth.dynds.org/decss


---------------------- multipart/signed attachment
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
Url : http://mail.libexpat.org/pipermail-21/expat-discuss/attachments/20010807/05400885/attachment.bin

---------------------- multipart/signed attachment--


From gstein@lyra.org  Wed Aug  8 00:21:48 2001
From: gstein@lyra.org (Greg Stein)
Date: Tue, 7 Aug 2001 16:21:48 -0700
Subject: [Expat-discuss] Anyone using BeOS?
In-Reply-To: <20010807180115.B15677@uchicago.edu>; from sam@uchicago.edu on Tue, Aug 07, 2001 at 06:01:15PM -0500
References: <15205.45464.632584.320171@cj42289-a.reston1.va.home.com> <20010807180115.B15677@uchicago.edu>
Message-ID: <20010807162148.P1414@lyra.org>

On Tue, Aug 07, 2001 at 06:01:15PM -0500, Sam TH wrote:
> On Mon, Jul 30, 2001 at 03:12:24PM -0400, Fred L. Drake, Jr. wrote:
> > 
> >   Does anyone here have a BeOS machine?  I'd like to learn a little
> > more about the appearant __declspec support on that platform, since
> > past patches indicate that it differs from the MSVC support.
> >   If anyone has a pointer to relevant documentation, I'd really
> > appreciate it.
> 
> Sorry about the late response.  
> 
> I'm the author of those patches.  I no longer have acess to the BeOS
> machines they were tested on.  As I remember it, the fundamental
> problem was that gcc, which we (AbiWord) use on x86 BeOS doesn't
> support __declspec.  The Metroworks compiler does, I think.  


Shouldn't be a problem. I had seen some -checkins related to that stuff. I
plan to just snarf what Apache is doing in that area. Apache's got lots of
portability stuff :-)

We should be able to clear it right up.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/


From asehgal@research.att.com  Wed Aug  8 03:06:05 2001
From: asehgal@research.att.com (Amit Sehgal)
Date: Tue, 7 Aug 2001 22:06:05 -0400
Subject: [Expat-discuss] Parsing Attributes.
Message-ID: <Pine.SGI.4.33.0108072200531.325515-100000@chips.research.att.com>

What function in expat handles parsing of attributes ?

Ex.

<Store  Manager="128.12.19.134"
	Port='23456"
/>

The start and end element handlers return the store tag.
But the Manager and Port attributes are skipped by the parser.

Comments, ideas ?

Thanks
Amit




From daniel@mailgratis.com.ar  Wed Aug  8 03:07:32 2001
From: daniel@mailgratis.com.ar (Daniel Crown)
Date: Tue, 7 Aug 2001 23:07:32 -0300
Subject: [Expat-discuss] Implementing other encodings on Expat
Message-ID: <HIELIBIHFFDAGGNCIENAGEGLEIAA.daniel@mailgratis.com.ar>

Hi there,

Do you guys plan to implement BIG5, JIS, ISO-8859-* and other encodings on
Expat?
If not, how can I insert my own encoding tabs for other charsets in the
source code?

Thank you!


Regards,
Daniel Crown
FRBA UTN





From Josh.Martin@abq.sc.philips.com  Wed Aug  8 17:29:56 2001
From: Josh.Martin@abq.sc.philips.com (Josh Martin)
Date: Wed, 8 Aug 2001 10:29:56 -0600 (MDT)
Subject: [Expat-discuss] is expat a reentrant parser
Message-ID: <200108081629.KAA25277@abqn42.abq.sc.philips.com>

Mike,

  The first sentace is paraphrased from the HP-UX Release 11.00 man page for 
sigaction(2). The rest is distilled from my learning about and use of reentrant 
functions. Anyone is of course free to quote or paraphrase "my" definition if it 
is useful to them. I also apologize for the odd formatting of the definition, 
web form email entry was never my forte.
  
  - Josh Martin


> From: MIke Wilson <mike.wilson@tradeweb.com>
> To: Josh Martin <josh.martin@abq.sc.philips.com>
> Subject: RE: [Expat-discuss] is expat a reentrant parser
> Date: Fri, 3 Aug 2001 13:27:09 -0400 
> MIME-Version: 1.0
> 
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> 
> Josh,
> 
> I just saw this post.  Where did you get this definition?  It's
> wonderful!  I'd love to be able to cite it from various points in my
> own project documentation.
> 
> TIA,
> 
> - - Mike
> 
> - -----Original Message-----
> From:	Josh Martin [SMTP:josh.martin@abq.sc.philips.com]
> Sent:	Friday, August 03, 2001 12:49 PM
> To:	SheelR@aol.com; expat-discuss@lists.sourceforge.net
> Subject:	RE: [Expat-discuss] is expat a reentrant parser
> 
> This message was sent from Geocrawler.com by "Josh Martin"
> <josh.martin@abq.sc.philips.com>
> 
> 
> >>SheelR@aol.com writes:
> >>         Is expat reentrant?
> >
> >  I'm not entirely sure what you mean here.  If
> you create several
> >parsers, you should be able to use them at the
> same time (from
> >different threads or whatever).  You probably
> don't want to muck
> >around with a parser while it is actively
> parsing, aside from calling
> >the defined APIs to add/change callbacks.
> >
> 
> <snip>
> 
> >  -Fred
> >
> >-- 
> >Fred L. Drake, Jr.  <fdrake at acm.org>
> >PythonLabs at Zope Corporation
> 
> Reentrant Functions:
> 
>   Reentrant functions (and functions which are not
> interruptable
> by signals) are defined as functions that may be
> invoked, without
> restriction, from signal-catching functions. A
> function is reentrant
> only if, when invoked inside a signal-catching
> function, it does
> not adversly affect the normal flow of operations
> of the function
> or code that the signal-catching function
> interrupted. In other
> words, reentrant functions aren't going to
> unexpectedly change
> any critical values, and thus the result of the
> operations, if they
> are invoked in the middle of a function.
> 
>   I'm not positive that this is the meaning that
> SheelR was going
> for, but this is the "standard" meaning of a
> reentrant function.
> Personally, I would like to know the answer to
> this question,
> given this definition of reentrant.
> 
>  - Josh Martin
> 
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss
> 		
> 
> Geocrawler.com - The Knowledge Archive
> 
> 
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss
> 
> -----BEGIN PGP SIGNATURE-----
> Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>
> 
> iQA/AwUBO2rf9owiya8i+ZOTEQLdlwCdF4RwCoa/phPg0Amuz34H3x2+G6YAnAsK
> qesUwwwrwKd1FNVuKsafDp/w
> =Nwae
> -----END PGP SIGNATURE-----



From gstein@lyra.org  Wed Aug  8 20:31:50 2001
From: gstein@lyra.org (Greg Stein)
Date: Wed, 8 Aug 2001 12:31:50 -0700
Subject: [Expat-discuss] is expat a reentrant parser
In-Reply-To: <200108081629.KAA25277@abqn42.abq.sc.philips.com>; from Josh.Martin@abq.sc.philips.com on Wed, Aug 08, 2001 at 10:29:56AM -0600
References: <200108081629.KAA25277@abqn42.abq.sc.philips.com>
Message-ID: <20010808123150.Z1414@lyra.org>

By that definition, Expat is not "reentrant". It allocates memory.

However, as Fred points out: it *is* thread-safe if you use different
parsers in each thread. (one parser across multiple threads will hork)

Cheers,
-g

On Wed, Aug 08, 2001 at 10:29:56AM -0600, Josh Martin wrote:
> Mike,
> 
>   The first sentace is paraphrased from the HP-UX Release 11.00 man page for 
> sigaction(2). The rest is distilled from my learning about and use of reentrant 
> functions. Anyone is of course free to quote or paraphrase "my" definition if it 
> is useful to them. I also apologize for the odd formatting of the definition, 
> web form email entry was never my forte.
>   
>   - Josh Martin
> 
> 
> > From: MIke Wilson <mike.wilson@tradeweb.com>
> > To: Josh Martin <josh.martin@abq.sc.philips.com>
> > Subject: RE: [Expat-discuss] is expat a reentrant parser
> > Date: Fri, 3 Aug 2001 13:27:09 -0400 
> > MIME-Version: 1.0
> > 
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> > 
> > 
> > Josh,
> > 
> > I just saw this post.  Where did you get this definition?  It's
> > wonderful!  I'd love to be able to cite it from various points in my
> > own project documentation.
> > 
> > TIA,
> > 
> > - - Mike
> > 
> > - -----Original Message-----
> > From:	Josh Martin [SMTP:josh.martin@abq.sc.philips.com]
> > Sent:	Friday, August 03, 2001 12:49 PM
> > To:	SheelR@aol.com; expat-discuss@lists.sourceforge.net
> > Subject:	RE: [Expat-discuss] is expat a reentrant parser
> > 
> > This message was sent from Geocrawler.com by "Josh Martin"
> > <josh.martin@abq.sc.philips.com>
> > 
> > 
> > >>SheelR@aol.com writes:
> > >>         Is expat reentrant?
> > >
> > >  I'm not entirely sure what you mean here.  If
> > you create several
> > >parsers, you should be able to use them at the
> > same time (from
> > >different threads or whatever).  You probably
> > don't want to muck
> > >around with a parser while it is actively
> > parsing, aside from calling
> > >the defined APIs to add/change callbacks.
> > >
> > 
> > <snip>
> > 
> > >  -Fred
> > >
> > >-- 
> > >Fred L. Drake, Jr.  <fdrake at acm.org>
> > >PythonLabs at Zope Corporation
> > 
> > Reentrant Functions:
> > 
> >   Reentrant functions (and functions which are not
> > interruptable
> > by signals) are defined as functions that may be
> > invoked, without
> > restriction, from signal-catching functions. A
> > function is reentrant
> > only if, when invoked inside a signal-catching
> > function, it does
> > not adversly affect the normal flow of operations
> > of the function
> > or code that the signal-catching function
> > interrupted. In other
> > words, reentrant functions aren't going to
> > unexpectedly change
> > any critical values, and thus the result of the
> > operations, if they
> > are invoked in the middle of a function.
> > 
> >   I'm not positive that this is the meaning that
> > SheelR was going
> > for, but this is the "standard" meaning of a
> > reentrant function.
> > Personally, I would like to know the answer to
> > this question,
> > given this definition of reentrant.
> > 
> >  - Josh Martin
> > 
> > _______________________________________________
> > Expat-discuss mailing list
> > Expat-discuss@lists.sourceforge.net
> > http://lists.sourceforge.net/lists/listinfo/expat-discuss
> > 		
> > 
> > Geocrawler.com - The Knowledge Archive
> > 
> > 
> > _______________________________________________
> > Expat-discuss mailing list
> > Expat-discuss@lists.sourceforge.net
> > http://lists.sourceforge.net/lists/listinfo/expat-discuss
> > 
> > -----BEGIN PGP SIGNATURE-----
> > Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>
> > 
> > iQA/AwUBO2rf9owiya8i+ZOTEQLdlwCdF4RwCoa/phPg0Amuz34H3x2+G6YAnAsK
> > qesUwwwrwKd1FNVuKsafDp/w
> > =Nwae
> > -----END PGP SIGNATURE-----
> 
> 
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss

-- 
Greg Stein, http://www.lyra.org/


From fdrake@acm.org  Wed Aug  8 20:59:55 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Wed, 8 Aug 2001 15:59:55 -0400 (EDT)
Subject: [Expat-discuss] Parsing Attributes.
In-Reply-To: <Pine.SGI.4.33.0108072200531.325515-100000@chips.research.att.com>
References: <Pine.SGI.4.33.0108072200531.325515-100000@chips.research.att.com>
Message-ID: <15217.39483.21856.169425@cj42289-a.reston1.va.home.com>

Amit Sehgal writes:
 > <Store  Manager="128.12.19.134"
 > 	Port='23456"
 > />
 > 
 > The start and end element handlers return the store tag.
 > But the Manager and Port attributes are skipped by the parser.

  You should be getting the attributes along with the start tag.  See
the examples in the article at:

	http://www.xml.com/pub/a/1999/09/expat/


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From garys@ihug.com.au  Fri Aug 10 07:33:50 2001
From: garys@ihug.com.au (Gary Stephenson)
Date: Fri, 10 Aug 2001 16:33:50 +1000
Subject: [Expat-discuss] problems with dtddestroy()
Message-ID: <001f01c12166$6e5e9e40$fad7fea9@gateway>

Hi all ( "Expatriates",  ?).

I am trying to get Expat to parse the OASIS conformance test Suite
document(s). It barfs with a debug assertion failure (dbghead.c : line 1017):

_BLOCK_TYPE_IS_VALID(pHead->nBlockUse)

on either of the following lines from the dtddestroy() function:

  if (p->scaffIndex)
    FREE(p->scaffIndex);

  if (p->scaffold)
    FREE(p->scaffold);
  traceLog( "dtd 2" ) ;

Note that the failure only occurs after successfully destroying the DTD itself
(testcases.dtd), and also successfully destroying the first external entity
("xmltest\xmltest.xml").  Only then does it fail when trying to destroy the
second external entity ("japanese\japanese.xml").

I assume the error is that the "scaffold"ing has already been free()'ed, and I
notice in the dtdCopy function the following lines appear:

  /* Don't want deep copying for scaffolding */
    ...
    newDtd->scaffLevel = oldDtd->scaffLevel;
    newDtd->scaffIndex = oldDtd->scaffIndex;

which may or may not be significant.

Anybody else experienced similar problems?  Can anybody tell me the right way
to fix it?

All processed files are in UTF-8 encoding. Using 1.95.2 on Win2K (SP2).

many tias,

gary




From syprat@yahoo.fr  Fri Aug 10 13:37:08 2001
From: syprat@yahoo.fr (=?iso-8859-1?q?Sylvain=20PRAT?=)
Date: Fri, 10 Aug 2001 14:37:08 +0200 (CEST)
Subject: [Expat-discuss] empty tags
Message-ID: <20010810123708.97732.qmail@web14809.mail.yahoo.com>

Hi,

I would like to know if there's a mean to know if the
read tag is empty : in fact, i would like to re-fit
the tags i don't want to recognize (both start and end
elements), so i expected to use the input context
associated to the byte count to do that, but i think
empty tags will cause problems because they will be
duplicated... Another way to do that ?

Thanks...

Sylvain



___________________________________________________________
Do You Yahoo!? -- Vos albums photos en ligne, 
Yahoo! Photos : http://fr.photos.yahoo.com


From fdrake@acm.org  Fri Aug 10 14:10:51 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Fri, 10 Aug 2001 09:10:51 -0400 (EDT)
Subject: [Expat-discuss] empty tags
In-Reply-To: <20010810123708.97732.qmail@web14809.mail.yahoo.com>
References: <20010810123708.97732.qmail@web14809.mail.yahoo.com>
Message-ID: <15219.56667.408959.764629@cj42289-a.reston1.va.home.com>

Sylvain PRAT writes:
 > I would like to know if there's a mean to know if the
 > read tag is empty : in fact, i would like to re-fit
 > the tags i don't want to recognize (both start and end
 > elements), so i expected to use the input context
 > associated to the byte count to do that, but i think
 > empty tags will cause problems because they will be
 > duplicated... Another way to do that ?

  I think the only way to tell the difference betwee <e/> and <e></e>
is to use the input context.  The simplest way would be to check it
for the end tags; if it starts with "</", the element did not use
empty element notation.  Checking this should be easy & fast (the call
to get the input context will be more expensive than the check!), but
you should be careful; if you have an empty element tag with a *lot*
of attributes, the beginning of the tag may not be there any more.
See the caveats in the documentation for more information.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From vnemchin@hotmail.com  Sun Aug 12 03:40:12 2001
From: vnemchin@hotmail.com (Vassilii Nemtchinov)
Date: Sun, 12 Aug 2001 02:40:12 +0000
Subject: [Expat-discuss] Character Data
Message-ID: <F263BnccYTjr8cU6xfP00005068@hotmail.com>

I know that the subject of handling character data has been already 
discussed here. Still I would like somebody to provide some suggestions on 
the subject. Since we can not assume that the character data arrive in one 
chunk for various reasons, in my solution I am allocating a buffer in the 
start tag handler and I am also setting a flag indicating that I've 
encountered the beginning of the element. I keep adding character data to a 
buffer in the character handler until I reset back the flag in the end 
handler. I can see several problems in using this method. First, it seems 
that the whole purpose of event-driven parser has been defied since I have 
to set sentinels  myself and not rely entirely on the parser. Secondly, in 
the worst case I have to allocate as many sentinels as I have elements in 
the document (same goes for separate buffers for character data). I am sure 
that somebody found a better solution for getting character data.


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp



From syprat@yahoo.fr  Mon Aug 13 08:15:10 2001
From: syprat@yahoo.fr (=?iso-8859-1?q?Sylvain=20PRAT?=)
Date: Mon, 13 Aug 2001 09:15:10 +0200 (CEST)
Subject: [Expat-discuss] empty tags
In-Reply-To: <15219.56667.408959.764629@cj42289-a.reston1.va.home.com>
Message-ID: <20010813071510.24523.qmail@web14802.mail.yahoo.com>

 --- "Fred L. Drake, Jr." <fdrake@acm.org> a écrit : >

> Sylvain PRAT writes:
>  > I would like to know if there's a mean to know if
> the
>  > read tag is empty : in fact, i would like to
> re-fit
>  > the tags i don't want to recognize (both start
> and end
>  > elements), so i expected to use the input context
>  > associated to the byte count to do that, but i
> think
>  > empty tags will cause problems because they will
> be
>  > duplicated... Another way to do that ?
> 
>   I think the only way to tell the difference betwee
> <e/> and <e></e>
> is to use the input context.  

It's what i had guessed reading the docs

> The simplest way would
> be to check it
> for the end tags; if it starts with "</", the
> element did not use
> empty element notation.  Checking this should be
> easy & fast (the call
> to get the input context will be more expensive than
> the check!), 

yes, but we should be aware of the charset...

>but
> you should be careful; if you have an empty element
> tag with a *lot*
> of attributes, the beginning of the tag may not be
> there any more.
> See the caveats in the documentation for more
> information.
> 

I think the parser is aware of the empty tag, so why
couldn't this be a feature (as reading the input
context is one too), especially because start tags are
possibly erased before reading the end tags...  

___________________________________________________________
Do You Yahoo!? -- Vos albums photos en ligne, 
Yahoo! Photos : http://fr.photos.yahoo.com


From brc@fourlittlemice.com  Mon Aug 13 08:20:07 2001
From: brc@fourlittlemice.com (Dirk Dierckx)
Date: Mon, 13 Aug 2001 09:20:07 +0200
Subject: [Expat-discuss] RE: Character Data
In-Reply-To: <E15W0WT-0007ol-00@usw-sf-list1.sourceforge.net>
Message-ID: <FBEPJFEBIOECCAJIGEKMKEONCDAA.brc@fourlittlemice.com>

"Vassilii Nemtchinov" <vnemchin@hotmail.com> wrote:
>I know that the subject of handling character data has been already
>discussed here. Still I would like somebody to provide some suggestions on
>the subject. Since we can not assume that the character data arrive in one
>chunk for various reasons, in my solution I am allocating a buffer in the
>start tag handler and I am also setting a flag indicating that I've
>encountered the beginning of the element. I keep adding character data to a
>buffer in the character handler until I reset back the flag in the end
>handler. I can see several problems in using this method. First, it seems
>that the whole purpose of event-driven parser has been defied since I have
>to set sentinels  myself and not rely entirely on the parser. Secondly, in
>the worst case I have to allocate as many sentinels as I have elements in
>the document (same goes for separate buffers for character data). I am sure
>that somebody found a better solution for getting character data.

I'll explain (in short) how I do it.  I use something like the
following:

struct SParserContext
{	
	size_t m_szElementValueSize; /* init: m_szElementValueSize = (size_t)0U; */
	size_t m_szElementValueLen; /* init: m_szElementValueLen = (size_t)0U; */
	char *m_pchElementValue; /* init: m_pchElementValue = NULL; */
	char **m_ppchCurrentElementValue; /* init: m_ppchCurrentElementValue = NULL; */
	...
};

and use this structure in the following way:

void
assignElementValueToCurrentElement(struct SParserContext *psCtx)
{
	if(psCtx->m_ppchCurrentElementValue
		&& NULL == *(psCtx->m_ppchCurrentElementValue)
		&& psCtx->m_szElementValueLen > (size_t)0U)
	{
		*(psCtx->m_ppchCurrentElementValue) = (char*)malloc(
			psCtx->m_szElementValueLen + (size_t)1U);
		strcpy(*(psCtx->m_ppchCurrentElementValue),
			 psCtx->m_pchElementValue;
	}

	/* Make sure the element value is empty. */
	psCtx->m_szElementValueLen = (size_t)0U;
}

void
processOpenTag(void *pvUserData,
		   const char *pcchElement,
		   const char **ppcchAttributes)
{
	struct SParserContext *psCtx = (struct SParserContext*)pvUserData;

	/* 
		If we have a value stored in m_pchElementValue at this point
		=> it will be the full value of the *previously encountered* element.
	*/
	assignElementValueToCurrentElement(psCtx);

	...
	/* With your new element (pcchElement) you need a NULL char *
	   to it's value (~pchThisElementValueBuffer) , so ... */
	pchThisElementValueBuffer = NULL;
	psCtx->m_ppchCurrentElementValue = &pchThisElementValueBuffer;
	...
}

void
processCloseTag(void *pvUserData,
		    const char *pcchElement)
{
	struct SParserContext *psCtx = (struct SParserContext*)pvUserData;

	/*
		If we have a value stored in m_pchElementValue at this point
		=> it will be the full value of the *current* (~ pcchElement) element.
	*/
	assignElementValueToCurrentElement(psCtx);
}

/*
	Beware: Decoding of pcchData has been left out of this sample code !!!
*/
void
processTagData(void *pvUserData,
		   const char *pcchData,
		   int iDataLen)
{
	const size_t cszIncrement = (size_t)512U;
	struct SParserContext *psCtx = (struct SParserContext*)pvUserData;
	size_t szDataLen = (size_t)iDataLen, szBufferSize;

	szBufferSize = psCtx->m_szElementValueLen + szDataLen + (size_t)1U;
	if(szBufferSize > psCtx->m_szElementValueSize)
	{
		char *pchNew = NULL;

		/*
			Not enough memory available in m_pchElementValue to append
			pcchData to it, create/enlarge m_pchElementValue.
		*/
		szBufferSize = ((szBufferSize / cszIncrement) + (size_t)1U)
			* cszIncrement;
		if(psCtx->m_szElementValueSize)
			pchNew = (char*)realloc(psCtx->m_pchElementValue, szBufferSize);
		else
			pchNew = (char*)malloc(szBufferSize);
		if(pchNew)
		{
			psCtx->m_szElementValueSize = szBufferSize;
			psCtx->m_pchElementValue = pchNew;
		}
	}

	if(szBufferSize <= psCtx->m_szElementValueSize)
	{
		char *pchString = &(psCtx->m_pchElementValue[psCtx->m_szElementValueLen]);

		memcpy((void*)pchString,
 (void*)pcchData, szDataLen);
		pchString[szDataLen] = '\0'; /* m_pchElementValue is always terminated. */
		psCtx->m_szElementValueLen += szDataLen;
	}
}

PS.: The use of psCtx->m_ppchCurrentElementValue is purely to make this
sample
complete, in your particular impl. you should use a method that is
appropriate
for your problem off course.

Hope this code has done more good then bad ;-).

Regards,
Dirk.

-----Original Message-----
From: expat-discuss-admin@lists.sourceforge.net
[mailto:expat-discuss-admin@lists.sourceforge.net]On Behalf Of
expat-discuss-request@lists.sourceforge.net
Sent: Sunday, August 12, 2001 9:04 PM
To: expat-discuss@lists.sourceforge.net
Subject: Expat-discuss digest, Vol 1 #94 - 1 msg


Send Expat-discuss mailing list submissions to
	expat-discuss@lists.sourceforge.net

To subscribe or unsubscribe via the World Wide Web, visit
	http://lists.sourceforge.net/lists/listinfo/expat-discuss
or, via email, send a message with subject or body 'help' to
	expat-discuss-request@lists.sourceforge.net

You can reach the person managing the list at
	expat-discuss-admin@lists.sourceforge.net

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Expat-discuss digest..."


Today's Topics:

   1. Character Data (Vassilii Nemtchinov)

--__--__--

Message: 1
From: "Vassilii Nemtchinov" <vnemchin@hotmail.com>
To: expat-discuss@lists.sourceforge.net
Date: Sun, 12 Aug 2001 02:40:12 +0000
Subject: [Expat-discuss] Character Data

I know that the subject of handling character data has been already
discussed here. Still I would like somebody to provide some suggestions on
the subject. Since we can not assume that the character data arrive in one
chunk for various reasons, in my solution I am allocating a buffer in the
start tag handler and I am also setting a flag indicating that I've
encountered the beginning of the element. I keep adding character data to a
buffer in the character handler until I reset back the flag in the end
handler. I can see several problems in using this method. First, it seems
that the whole purpose of event-driven parser has been defied since I have
to set sentinels  myself and not rely entirely on the parser. Secondly, in
the worst case I have to allocate as many sentinels as I have elements in
the document (same goes for separate buffers for character data). I am sure
that somebody found a better solution for getting character data.


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp




--__--__--

_______________________________________________
Expat-discuss mailing list
Expat-discuss@lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/expat-discuss


End of Expat-discuss Digest



From fdrake@acm.org  Mon Aug 13 15:51:58 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Mon, 13 Aug 2001 10:51:58 -0400 (EDT)
Subject: [Expat-discuss] empty tags
In-Reply-To: <20010813071510.24523.qmail@web14802.mail.yahoo.com>
References: <15219.56667.408959.764629@cj42289-a.reston1.va.home.com>
	<20010813071510.24523.qmail@web14802.mail.yahoo.com>
Message-ID: <15223.59790.73199.257638@cj42289-a.reston1.va.home.com>

=?iso-8859-1?q?Sylvain=20PRAT?= writes:
 > yes, but we should be aware of the charset...

  Yes, pretty flaky stuff.  Here's another approach: for every start
tag, get the current source index, then for end tags, you know
it was an empty tag if the position didn't change.  That can be
optimized a little bit by maintaining a flag:

static int  maybe_empty_element_tag;
static long byte_index;

void
start(void *data, const char *el, const char **attr)
{
    maybe_empty_element_tag = 1;
    byte_index = XML_GetCurrentByteIndex(parser);

    ...
}

static void
end(void *data, const char *el)
    if (maybe_empty_element_tag) {
        maybe_empty_element_tag = 0;
        if (byte_index == XML_GetCurrentByteIndex(parser)) {
           /* empty-element tag */
           return
        }
    }
    ...
}

  All other handlers could optionally clear maybe_empty_element_tag to
avoid the call back into expat for elements like <e>characters
only</e>.  Whether that would be a win depends how isolated you want
this aspect of the processing, or if the performance improvement (very
small) is worth the maintenance cost.

 > I think the parser is aware of the empty tag, so why
 > couldn't this be a feature (as reading the input
 > context is one too), especially because start tags are
 > possibly erased before reading the end tags...  

  The parser has to be aware of it, but I'm not sure what would be the
best way to expose it.  Perhaps something similar to the
XML_GetSpecifiedAttributeCount() function, which is only valid during
the start-element callback?  That's certainly possible, but would
incur higher overhead that the approach outlined here (because it
would always result in a function call).  Providing that would not
invalidate this approach, so they could coexist.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From rsalz@zolera.com  Mon Aug 13 16:14:05 2001
From: rsalz@zolera.com (Rich Salz)
Date: Mon, 13 Aug 2001 11:14:05 -0400
Subject: [Expat-discuss] Re-entrant
Message-ID: <3B77EEBD.AFDB440A@zolera.com>

Sorry, but that definition from HP is completely bogus.  First,
according to ISO C, the only safe thing to do from a signal handler is
set a variable of type sig_atomic_t.  Second, it's not what reentrant
means. :)

Informally, re-entrant means that it is safe to call the function while
an existing call is pending:  can I *re-enter* the function?  An
excellent example is yacc.  Yacc exports a single function yyparse(). 
Suppose a write a parser for some language, and that language has an
import statement.  Can my import handler call yyparse to parse the new
file, while the containing file is being parsed.  Is yyparse
re-entrant?  The answer is no, it's not.  Yacc stores its state in
internal global variables.

For another example, look at the C function "strtok."

I'm not sure what the answer is to expat, except that it usually doesn't
matter.  Rather than re-enter, the thing to do is create another parser
and have that parser do the "recursive" or "internal" work.

Hope this helps.
	/r$

-- 
Zolera Systems, Your Key to Online Integrity
Securing Web services: XML, SOAP, Dig-sig, Encryption
http://www.zolera.com


From fdrake@acm.org  Mon Aug 13 16:25:05 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Mon, 13 Aug 2001 11:25:05 -0400 (EDT)
Subject: [Expat-discuss] Re-entrant
In-Reply-To: <3B77EEBD.AFDB440A@zolera.com>
References: <3B77EEBD.AFDB440A@zolera.com>
Message-ID: <15223.61777.423858.420431@cj42289-a.reston1.va.home.com>

Rich Salz writes:
 > Informally, re-entrant means that it is safe to call the function while
 > an existing call is pending:  can I *re-enter* the function?  An

  This is certainly what I'm more familiar with...


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From michael.isard@compaq.com  Mon Aug 13 19:16:33 2001
From: michael.isard@compaq.com (Michael Isard)
Date: Mon, 13 Aug 2001 11:16:33 -0700 (PDT)
Subject: [Expat-discuss] empty tags
Message-ID: <15224.6529.299234.691317@diamond.pa.dec.com>

>On Mon, 13 Aug 2001 10:51:58 -0400 (EDT), "Fred L. Drake, Jr." <fdrake@acm.org> said:

> =?iso-8859-1?q?Sylvain=20PRAT?= writes:
>> yes, but we should be aware of the charset...

>   Yes, pretty flaky stuff.  Here's another approach: for every start
> tag, get the current source index, then for end tags, you know it
> was an empty tag if the position didn't change.  That can be
> optimized a little bit by maintaining a flag:

My code relies on the fact that

  XML_GetCurrentByteCount(parser) == 0

in the end element callback iff the element is empty. I think this is
equivalent to what you propose but avoids storing flags.

Michael.


From fdrake@acm.org  Mon Aug 13 20:04:56 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Mon, 13 Aug 2001 15:04:56 -0400 (EDT)
Subject: [Expat-discuss] empty tags
In-Reply-To: <15224.6529.299234.691317@diamond.pa.dec.com>
References: <15224.6529.299234.691317@diamond.pa.dec.com>
Message-ID: <15224.9432.61787.444160@cj42289-a.reston1.va.home.com>

Michael Isard writes:
 > My code relies on the fact that
 > 
 >   XML_GetCurrentByteCount(parser) == 0
 > 
 > in the end element callback iff the element is empty. I think this is
 > equivalent to what you propose but avoids storing flags.

  Aha!  This is good.
  Maintaining a flag may yield better performance, but this is
probably better.  I'll add a note to the documentation for
XML_GetCurrentByteCount() to indicate this.
  I should probably write a cookbook-style document containing tips
for using Expat to make these detailed distinctions and show good
usage examples, but I'd really rather someone else write it and
contribute it to the documentation.
  Besides, I should be writing a test suite for Expat.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From fdrake@acm.org  Mon Aug 13 19:50:02 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Mon, 13 Aug 2001 14:50:02 -0400 (EDT)
Subject: [Expat-discuss] Character Data
In-Reply-To: <F263BnccYTjr8cU6xfP00005068@hotmail.com>
References: <F263BnccYTjr8cU6xfP00005068@hotmail.com>
Message-ID: <15224.8538.631600.976693@cj42289-a.reston1.va.home.com>

Vassilii Nemtchinov writes:
 > that the whole purpose of event-driven parser has been defied since I have 
 > to set sentinels  myself and not rely entirely on the parser. Secondly, in 

  No; you still avoid an enormous portion of the work involved with
something like building a DOM has been avoided.  Unless you're having
to keep everything indefinately, you get a win there.  Internal state
tracking and short-term data accumulation are common with event-based
parsing, XML or otherwise.

 > the worst case I have to allocate as many sentinels as I have elements in 
 > the document (same goes for separate buffers for character data). I am sure 
 > that somebody found a better solution for getting character data.

  How many buffers you need to allocate and how many state variables
you need depends entirely on your application.  If you're searching
the text for a match, for example, there are certainly incremental
algorithms that can be applied.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From JWieman@daktronics.com  Mon Aug 13 21:04:51 2001
From: JWieman@daktronics.com (Joe Wiemann)
Date: Mon, 13 Aug 2001 15:04:51 -0500
Subject: [Expat-discuss] Definition Of Re-entrant
Message-ID: <sb77ecbc.021@gw.daktronics.com>

Reentrancy -=20

One of the real beauties of multi threading is that the same C function =
can be called from multiple threads.... This provides great power and =
reduces code space... however, it does require that C functions called =
from multiple threads are reentrant.

What does reentrant mean? Basically a reentrant function stores the =
caller's return address on the stack and does not rely on global or static =
C variables that it previously set up.  Most compilers place the return =
address on the stack. hence application developers must only worry about =
the use of globals and statics.

An example of a non-reentrant function is strtok found in the standard C =
library.  This function remembers the previous string pointer on subsequent=
 calls. it does this with a static string pointer.  if this function is =
called from multiple threads, it would most likely return an invalid =
pointer.

 - Taken from threadx user manual.


>>> "Fred L. Drake, Jr." <fdrake@acm.org> 08/13/01 10:25AM >>>

Rich Salz writes:
 > Informally, re-entrant means that it is safe to call the function while
 > an existing call is pending:  can I *re-enter* the function?  An

  This is certainly what I'm more familiar with...


  -Fred

--=20
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation


_______________________________________________
Expat-discuss mailing list
Expat-discuss@lists.sourceforge.net=20
http://lists.sourceforge.net/lists/listinfo/expat-discuss



From dcrowley@scitegic.com  Mon Aug 13 22:01:54 2001
From: dcrowley@scitegic.com (David Crowley)
Date: Mon, 13 Aug 2001 14:01:54 -0700
Subject: [Expat-discuss] Proposal for XML_ParserReset() function
Message-ID: <5.1.0.14.0.20010813135650.02d77fd8@pop.business.earthlink.net>

---------------------- multipart/mixed attachment

I just posted this into the Patches on sourceforge, but thought I would 
present it here for further discussion and so others may test it/refine 
it.  This functionality is quite important for my application (SOAP server) 
which can parse many thousands of XML documents.  When doing some 
profiling, about 12-15% of the time was being spent in memory allocation 
routines when creating and destroying the Parser object.  I know others 
have asked for this functionality and so I spent some time the past weekend 
looking into this.

David



This is my first cut at adding a XML_ParserReset function.  My idea was to 
reset the parser to a state that was almost identical to what it is after 
XML_ParserCreate() except that any allocated memory is preserved.  As this 
patch is currently, I think it misght still has some potential problems 
with dtdInit() and possibly internalEncoding and setContext().  But for my 
documents/application it seems to work great.  It passes Purify without any 
memory leaks and when parsing 5000 documents, I only get ~40 memory 
allocations instead of ~200,000 :)

The function declartion needed for expat.h:

/* Resets an existing parser to a state comparable to that after
    XML_ParserCreate but preserves any allocated memory. */

XMLPARSEAPI(void)
XML_ParserReset(XML_Parser parser, const XML_Char *encoding);


---------------------- multipart/mixed attachment
A non-text attachment was scrubbed...
Name: reset.diff
Type: application/octet-stream
Size: 5937 bytes
Desc: not available
Url : http://mail.libexpat.org/pipermail-21/expat-discuss/attachments/20010813/83be080c/reset.exe

---------------------- multipart/mixed attachment

---------------------- multipart/mixed attachment--



From mballen@erols.com  Tue Aug 14 04:22:31 2001
From: mballen@erols.com (Michael B. Allen)
Date: Mon, 13 Aug 2001 23:22:31 -0400
Subject: [Expat-discuss] Re-entrant
In-Reply-To: <3B77EEBD.AFDB440A@zolera.com>; from rsalz@zolera.com on Mon, Aug 13, 2001 at 11:14:05AM -0400
References: <3B77EEBD.AFDB440A@zolera.com>
Message-ID: <20010813232231.C1095@nano.foo.net>

On Mon, Aug 13, 2001 at 11:14:05AM -0400, Rich Salz wrote:
> file, while the containing file is being parsed.  Is yyparse
> re-entrant?  The answer is no, it's not.  Yacc stores its state in
> internal global variables.

Actually, bison does have a %pure_parser option that supposedly will
un-static-ify these state variables and therefore make the parser
reentrant but I think I tried it once and couldn't get it to work.

About reentrance and threading: It looks like all of the functions of
the Expat API are clearly reentrant (and I distinctly recall seeing a
message to this effect by James Clark on a mailing list somewhere). The
design is such that all state is stored in the XML_Parser object which is
passed to all (almost all) functions of the Expat API. AFAIK there is no
static storage allocated within functions or global writable data. All
this means is that it is ok to call the same function before a previous
call to that function has completed. This might occur in a multi-threaded
environment or if perhaps some function was called recursively (which
I'm sure happends quite natrually internally to the parser).

However, this does not mean you can use the same parser object recursively
or in a multi-threaded environment. This is a completely different
issue. That XML_Parser object exists so that _separate_ parser objects
may be used in these ways.

I could be wrong though. I don't know the Expat source very well at all
and it would be easy to introduce a static variable at the beginnging
of a function or somesuch that inadvertantly uses static storage.

I think this has been said in less words already, so sorry if I'm being
redundant.

Mike

-- 
Wow a memory-mapped fork bomb! Now what on earth did you expect? - lkml


From syprat@yahoo.fr  Tue Aug 14 08:29:01 2001
From: syprat@yahoo.fr (=?iso-8859-1?q?Sylvain=20PRAT?=)
Date: Tue, 14 Aug 2001 09:29:01 +0200 (CEST)
Subject: [Expat-discuss] empty tags
In-Reply-To: <15223.59790.73199.257638@cj42289-a.reston1.va.home.com>
Message-ID: <20010814072901.33304.qmail@web14803.mail.yahoo.com>

 --- "Fred L. Drake, Jr." <fdrake@acm.org> a écrit : >

> =?iso-8859-1?q?Sylvain=20PRAT?= writes:
>  > yes, but we should be aware of the charset...
> 
>   Yes, pretty flaky stuff.  Here's another approach:
> for every start
> tag, get the current source index, then for end
> tags, you know
> it was an empty tag if the position didn't change. 
> That can be
> optimized a little bit by maintaining a flag:
> 
> static int  maybe_empty_element_tag;
> static long byte_index;
> 
> void
> start(void *data, const char *el, const char **attr)
> {
>     maybe_empty_element_tag = 1;
>     byte_index = XML_GetCurrentByteIndex(parser);
> 
>     ...
> }
> 
> static void
> end(void *data, const char *el)
>     if (maybe_empty_element_tag) {
>         maybe_empty_element_tag = 0;
>         if (byte_index ==
> XML_GetCurrentByteIndex(parser)) {
>            /* empty-element tag */
>            return
>         }
>     }
>     ...
> }
> 
>   All other handlers could optionally clear
> maybe_empty_element_tag to
> avoid the call back into expat for elements like
> <e>characters
> only</e>.  Whether that would be a win depends how
> isolated you want
> this aspect of the processing, or if the performance
> improvement (very
> small) is worth the maintenance cost.
> 
>  > I think the parser is aware of the empty tag, so
> why
>  > couldn't this be a feature (as reading the input
>  > context is one too), especially because start
> tags are
>  > possibly erased before reading the end tags...  
> 
>   The parser has to be aware of it, but I'm not sure
> what would be the
> best way to expose it.  Perhaps something similar to
> the
> XML_GetSpecifiedAttributeCount() function, which is
> only valid during
> the start-element callback?  That's certainly
> possible, but would
> incur higher overhead that the approach outlined
> here (because it
> would always result in a function call).  Providing
> that would not
> invalidate this approach, so they could coexist.
> 

Finally, there's a best solution to this, using the
XML_GetCurrentByteCount which returns 0 for end
elements when using an empty tag... but it results in
a function call (i don't mind)...

Sylvain  

___________________________________________________________
Do You Yahoo!? -- Vos albums photos en ligne, 
Yahoo! Photos : http://fr.photos.yahoo.com


From k.buckley@lancaster.ac.uk  Wed Aug 15 15:55:18 2001
From: k.buckley@lancaster.ac.uk (Kev Buckley)
Date: Wed, 15 Aug 2001 15:55:18 +0100 (BST)
Subject: [Expat-discuss] Parser query
Message-ID: <E15X24l-0006Nt-00@unixa.lancs.ac.uk>

Hello,

I'm trying to understand what characters the expat parser is throwing
when it reads the numerical entity code for the &#szlig; and &#oslah;
characters.

Bit of background.

I'm dumping data off a Palm which produces 

octal 337 (dec 223)  for the &#szlig;
      370 (dec 248)  for the &#oslash;

But if I have &#223; and &#248; in my XML doc, when expat parses the
doc I seem to get back two bytes as follows:

octal 303 + octal 237 for the &#223;
octal 303 + octal 248 for the &#248;


Other "extended shift" characters from the Palm seem to translate OK,
in that their &#nnn; codes get parsed as 

octal 302 + octal NNN  

where NNN equates to the decimal nnn used in the numeric entity code.

eg, the trademark symbol 

octal 231 (dec 153)  ->   &#153  which is parsed as 

octal 302 + octal 231


If the above makes any sense to anyone, then do you have any clues as
to what I have missed in trying to get these characters out of
(through) the expat parser ?


Kevin

-- 
Regards,

----------------------------------------------------------------------
*  Kevin M. Buckley              e-mail: K.Buckley@lancaster.ac.uk   *
*                                                                    *
*  Systems Administrator                                             *
*  Computer Centre                                                   *
*  Lancaster University          Voice:  +44 (0) 1524 5 93718        *
*  LANCASTER. LA1 4YW            Fax  :  +44 (0) 1524 5 25113        *
*  England.                                                          *
*                                                                    *
*  My PC runs Linux/GNU, you still computing the Bill Gate$' way ?   *
----------------------------------------------------------------------


From fdrake@acm.org  Wed Aug 15 15:59:23 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Wed, 15 Aug 2001 10:59:23 -0400 (EDT)
Subject: [Expat-discuss] Parser query
In-Reply-To: <E15X24l-0006Nt-00@unixa.lancs.ac.uk>
References: <E15X24l-0006Nt-00@unixa.lancs.ac.uk>
Message-ID: <15226.36427.253508.408143@cj42289-a.reston1.va.home.com>

Kev Buckley writes:
 > But if I have &#223; and &#248; in my XML doc, when expat parses the
 > doc I seem to get back two bytes as follows:
 > 
 > octal 303 + octal 237 for the &#223;
 > octal 303 + octal 248 for the &#248;
...
 > octal 231 (dec 153)  ->   &#153  which is parsed as 
 > 
 > octal 302 + octal 231
...
 > If the above makes any sense to anyone, then do you have any clues as
 > to what I have missed in trying to get these characters out of
 > (through) the expat parser ?

  The characters are coming through just fine if you're telling Expat
that your input is Latin-1.  Expat's output is encoded in UTF-8,
though it can be compiled to return 16-bit characters directly.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From brc@fourlittlemice.com  Thu Aug 16 07:40:10 2001
From: brc@fourlittlemice.com (Dirk Dierckx)
Date: Thu, 16 Aug 2001 08:40:10 +0200
Subject: [Expat-discuss] RE: Parser query
In-Reply-To: <E15X5xl-0002qM-00@usw-sf-list1.sourceforge.net>
Message-ID: <FBEPJFEBIOECCAJIGEKMAEOPCDAA.brc@fourlittlemice.com>

As Fred L. Drake already pointed out, you get it UTF-8 encoded by default
and using Latin-1 requires Expat to be recompiled to work with UTF-16.  If
you're like me and just want to retrieve your &szlig; in a char with value
(dec 223), etc. without having to recompile expat nor using 16bit wide chars
(wchar_t) you can use the following function to convert from UTF-8 to ANSI
(0-255).

int
ishUtilUTF8toANSI(char *pchString, int iStringLen)
{
  const size_t cszStringLen = iStringLen >= 0
    ? (size_t)iStringLen
    : (pchString
       ? strlen(pchString)
       : (size_t)0U);
  int bConverted = 1;
  size_t szInputIdx, szOutputIdx = (size_t)0U;

  for(szInputIdx = (size_t)0U;
      bConverted
	&& szInputIdx < cszStringLen;
      ++szInputIdx)
    {
      /* If input_bin(0xxxxxxx) ~ ASCII_bin(0xxxxxxx)
	 If input_bin(110000yy 10xxxxxx) ~ ANSI_bin(yyxxxxxx)
	 All other UTF-8 encodings don't map to ANSI so we don't
	 convert them and fail if we encounter them.
	 See: http://www.unicode.org for more information about
	 UTF-8 encoding.
      */
      if(0x00 == (pchString[szInputIdx] & 0x80)) /* Plain ascii char */
		pchString[szOutputIdx++] = pchString[szInputIdx];
      else if(szInputIdx + (size_t)1U < cszStringLen
	      && 0xC0 == (pchString[szInputIdx] & 0xFC)
	      && 0x80 == (pchString[szInputIdx + (size_t)1U] & 0xC0))
	{
	  /* UTF-8 encoded char that maps to ANSI. */
	  pchString[szOutputIdx++] = ((pchString[szInputIdx] & 0x03) << 6)
	    + (pchString[szInputIdx + (size_t)1U] & 0x3F);
	  ++szInputIdx; /* We must skip the second input char. */
	}
      else /* UTF-8 encoding that doesn't map to ANSI or illegal input. */
		bConverted = 0;
    }
  return bConverted ? (int)szOutputIdx : -1;
}

Note: As you can see from the code, this function converts the string
inplace (modifying the data pointed to by pchString directly).  This is
possible because the resulting string length will be <= iStringLen.

---
Regards,
Dirk



From willievu@hotmail.com  Sat Aug 18 03:35:49 2001
From: willievu@hotmail.com (Willie Vu)
Date: Fri, 17 Aug 2001 22:35:49 -0400
Subject: [Expat-discuss] undefined reference when compile examples
Message-ID: <F229Aj5p1Fjxe0ymQ1W0000b170@hotmail.com>

I'm a newbie here.  I run into a dump problem.  Please help.  I failed to 
compile the examples due the following errors.  I'm running cygwin gcc 
version 2.95.3-5 (Reading specs from 
/usr/lib/gcc-lib/i686-pc-cygwin/2.95.3-5/specs, gcc version 2.95.3-5 (cygwin 
special)) on Windows 2000 SP2.

----

d:/temp/expat/expat-1.95.2/examples> make
gcc -o elements elements.o -static -L../lib/.libs -lexpat
elements.o: In function `main':
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:31: undefined 
reference to `_imp__XML_ParserCreate'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:34: undefined 
reference to `_imp__XML_SetUserData'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:35: undefined 
reference to `_imp__XML_SetElementHandler'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:39: undefined 
reference to `_imp__XML_Parse'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:47: undefined 
reference to `_imp__XML_ParserFree'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:40: undefined 
reference to `_imp__XML_GetCurrentLineNumber'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:40: undefined 
reference to `_imp__XML_GetErrorCode'
/cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:40: undefined 
reference to `_imp__XML_ErrorString'
collect2: ld returned 1 exit status
make: *** [elements] Error 1

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp



From mballen@erols.com  Sat Aug 18 05:04:42 2001
From: mballen@erols.com (Michael B. Allen)
Date: Sat, 18 Aug 2001 00:04:42 -0400
Subject: [Expat-discuss] undefined reference when compile examples
In-Reply-To: <F229Aj5p1Fjxe0ymQ1W0000b170@hotmail.com>; from willievu@hotmail.com on Fri, Aug 17, 2001 at 10:35:49PM -0400
References: <F229Aj5p1Fjxe0ymQ1W0000b170@hotmail.com>
Message-ID: <20010818000442.A4548@nano.foo.net>

On Fri, Aug 17, 2001 at 10:35:49PM -0400, Willie Vu wrote:
> 
> d:/temp/expat/expat-1.95.2/examples> make
> gcc -o elements elements.o -static -L../lib/.libs -lexpat
> elements.o: In function `main':
> /cygdrive/d/temp/expat/expat-1.95.2/examples/elements.c:31: undefined 
> reference to `_imp__XML_ParserCreate'

I believe you need to include the .h path. Like -I../include/? or
similar. Also, is that third dot just before libs in the library path
really supposed to be there? This is a simple gcc options problem. You
need to get your include and library paths right.

Mike

-- 
Wow a memory-mapped fork bomb! Now what on earth did you expect? - lkml


From JWieman@daktronics.com  Mon Aug 20 21:16:23 2001
From: JWieman@daktronics.com (Joe Wiemann)
Date: Mon, 20 Aug 2001 15:16:23 -0500
Subject: [Expat-discuss] RE: Parser query
Message-ID: <sb8129e8.030@gw.daktronics.com>

Is there anyway I can pull expat out of the library implementation and =
compile it statically into my C code...

I am sure there is a way to do this =AF but I can't figure it out... I am =
working in an embedded environment so a dll is not a viable option using =
the os that we are using... I need to be able to compile this statically =
into my binary file.





From fdrake@acm.org  Mon Aug 20 21:22:56 2001
From: fdrake@acm.org (Fred L. Drake, Jr.)
Date: Mon, 20 Aug 2001 16:22:56 -0400 (EDT)
Subject: [Expat-discuss] RE: Parser query
In-Reply-To: <sb8129e8.030@gw.daktronics.com>
References: <sb8129e8.030@gw.daktronics.com>
Message-ID: <15233.29088.634733.835654@cj42289-a.reston1.va.home.com>

Joe Wiemann wrote:
 > Is there anyway I can pull expat out of the library implementation and
 > compile it statically into my C code...
 >
 > I am sure there is a way to do this but I can't figure it out... I am
 > working in an embedded environment so a dll is not a viable option using
 > the os that we are using... I need to be able to compile this statically
 > into my binary file.

  You should be able to use the C files in the lib/ directory; there
really aren't any special compilation issues.  You should be able to
determine what pre-processor flags are appropriate for your platform.


  -Fred

-- 
Fred L. Drake, Jr.  <fdrake at acm.org>
PythonLabs at Zope Corporation



From gstein@lyra.org  Mon Aug 20 21:40:55 2001
From: gstein@lyra.org (Greg Stein)
Date: Mon, 20 Aug 2001 13:40:55 -0700
Subject: [Expat-discuss] RE: Parser query
In-Reply-To: <15233.29088.634733.835654@cj42289-a.reston1.va.home.com>; from fdrake@acm.org on Mon, Aug 20, 2001 at 04:22:56PM -0400
References: <sb8129e8.030@gw.daktronics.com> <15233.29088.634733.835654@cj42289-a.reston1.va.home.com>
Message-ID: <20010820134055.F11073@lyra.org>

On Mon, Aug 20, 2001 at 04:22:56PM -0400, Fred L. Drake, Jr. wrote:
> Joe Wiemann wrote:
>  > Is there anyway I can pull expat out of the library implementation and
>  > compile it statically into my C code...
>  >
>  > I am sure there is a way to do this but I can't figure it out... I am
>  > working in an embedded environment so a dll is not a viable option using
>  > the os that we are using... I need to be able to compile this statically
>  > into my binary file.
> 
>   You should be able to use the C files in the lib/ directory; there
> really aren't any special compilation issues.  You should be able to
> determine what pre-processor flags are appropriate for your platform.

You could also link the Expat library into your application using static
linking.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/


From syprat@yahoo.fr  Tue Aug 21 14:47:49 2001
From: syprat@yahoo.fr (=?iso-8859-1?q?Sylvain=20PRAT?=)
Date: Tue, 21 Aug 2001 15:47:49 +0200 (CEST)
Subject: [Expat-discuss] Stopping the parser
Message-ID: <20010821134749.28831.qmail@web14810.mail.yahoo.com>

Hi,

I'm experiencing a new problem using expat : 
I have to do an incremental parsing of a file , i have
a file describing a set of events and i need to read
it event by event (with first, next functions), so on
the event end tag, i would like to stop the parser.
But i see no convenient way to do that. 

I first used a boolean in my loop (reading from the
file, parsing the resulting buffer) to stop it (which
is set in the event end element handler), but the
parser still parsed the rest of the buffer, and i lost
the next data. We can't go back parsing (i think), the
only way is to create a new parser to parse the next
event (starting after the event end tag file
position), but obviously, the parser will state an
error because at the last event, it will encounter the
end tag of the set of events (dealing with errors,
possible but very bad solution i think)... I have also
thought of using a stack/set/list to remember the
events, but it's not easy to make it (it's c++, i
couldn't use the stl, and moreover the events are
quite complicated to be pushed)
So is there's a solution to this, is it difficult to
add a stop parser feature, something else ???

Thanks



___________________________________________________________
Do You Yahoo!? -- Vos albums photos en ligne, 
Yahoo! Photos : http://fr.photos.yahoo.com


From lshen@cisco.com  Wed Aug 22 02:11:10 2001
From: lshen@cisco.com (Lin Shen)
Date: Tue, 21 Aug 2001 18:11:10 -0700
Subject: [Expat-discuss] return tag contents as it is
Message-ID: <006201c12aa7$548ac490$738b6b80@cisco.com>

Hi,

Is it possible to have the parser return the contents of certain tags as it
is? For instance, for the following xml segment

<grammar>
    <rule id="root">
        text <ruleref tag="&lt tag &gt"/>
    </rule>
</grammar>

is there is a way to let the parser to return whatever contained in
<grammar> tag w/out any parsing, including the escape char and the <rule>
tag?

Please send reply to my personal account lshen@cisco.com

thanks
lin



From Michael_B_Allen@ml.com  Wed Aug 22 02:16:18 2001
From: Michael_B_Allen@ml.com (Allen, Michael B (RSCH))
Date: Tue, 21 Aug 2001 21:16:18 -0400
Subject: [Expat-discuss] return tag contents as it is
Message-ID: <B27EB33BAB29D2119ABF0001FA7EF289053BF176@ewfd04.exchange.ml.com>

I don't think so. That would be an extra feature of an XML parser. Perhaps you
can use the special CDATA tag into which you may put anything including xml.

> -----Original Message-----
> From:	Lin Shen [SMTP:lshen@cisco.com]
> Sent:	Tuesday, August 21, 2001 9:11 PM
> To:	expat-discuss@lists.sourceforge.net
> Subject:	[Expat-discuss] return tag contents as it is
> 
> Hi,
> 
> Is it possible to have the parser return the contents of certain tags as it
> is? For instance, for the following xml segment
> 
> <grammar>
>     <rule id="root">
>         text <ruleref tag="&lt tag &gt"/>
>     </rule>
> </grammar>
> 
> is there is a way to let the parser to return whatever contained in
> <grammar> tag w/out any parsing, including the escape char and the <rule>
> tag?
> 
> Please send reply to my personal account lshen@cisco.com
> 
> thanks
> lin
> 
> 
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss



From lshen@cisco.com  Wed Aug 22 02:57:52 2001
From: lshen@cisco.com (Lin Shen)
Date: Tue, 21 Aug 2001 18:57:52 -0700
Subject: [Expat-discuss] return tag contents as it is
References: <B27EB33BAB29D2119ABF0001FA7EF289053BF176@ewfd04.exchange.ml.com>
Message-ID: <007001c12aad$daecdef0$738b6b80@cisco.com>

What about getting the begin and end positions of the <grammar> tag in the
buffer (whatever that is)? So that the contents can be copied out.

lin

----- Original Message -----
From: "Allen, Michael B (RSCH)" <Michael_B_Allen@ml.com>
To: "'Lin Shen'" <lshen@cisco.com>; <expat-discuss@lists.sourceforge.net>
Sent: Tuesday, August 21, 2001 6:16 PM
Subject: RE: [Expat-discuss] return tag contents as it is


> I don't think so. That would be an extra feature of an XML parser. Perhaps
you
> can use the special CDATA tag into which you may put anything including
xml.
>
> > -----Original Message-----
> > From: Lin Shen [SMTP:lshen@cisco.com]
> > Sent: Tuesday, August 21, 2001 9:11 PM
> > To: expat-discuss@lists.sourceforge.net
> > Subject: [Expat-discuss] return tag contents as it is
> >
> > Hi,
> >
> > Is it possible to have the parser return the contents of certain tags as
it
> > is? For instance, for the following xml segment
> >
> > <grammar>
> >     <rule id="root">
> >         text <ruleref tag="&lt tag &gt"/>
> >     </rule>
> > </grammar>
> >
> > is there is a way to let the parser to return whatever contained in
> > <grammar> tag w/out any parsing, including the escape char and the
<rule>
> > tag?
> >
> > Please send reply to my personal account lshen@cisco.com
> >
> > thanks
> > lin
> >
> >
> > _______________________________________________
> > Expat-discuss mailing list
> > Expat-discuss@lists.sourceforge.net
> > http://lists.sourceforge.net/lists/listinfo/expat-discuss
>



From Michael_B_Allen@ml.com  Wed Aug 22 03:34:15 2001
From: Michael_B_Allen@ml.com (Allen, Michael B (RSCH))
Date: Tue, 21 Aug 2001 22:34:15 -0400
Subject: [Expat-discuss] return tag contents as it is
Message-ID: <B27EB33BAB29D2119ABF0001FA7EF289053BF178@ewfd04.exchange.ml.com>

Well I don't think there is an existing facility for this however I suppose you could add a function to extract that information. If you can identify where exactly begin and end tags are discovered
you might set an unsigned int in the XML_Parser structure. Of course you would need to call your function in the begin and end tag handlers before the value is lost and it might need some positional
translation.

IMHO I don't think you should do what you're doing in the first place though. Storing tagged stuff in XML is error prone to say the least. You could write a much less obfuscated grammer to be parsed
with a yacc parser.

Mike

> -----Original Message-----
> From:	Lin Shen [SMTP:lshen@cisco.com]
> Sent:	Tuesday, August 21, 2001 9:58 PM
> To:	Allen, Michael B (RSCH); expat-discuss@lists.sourceforge.net
> Subject:	Re: [Expat-discuss] return tag contents as it is
> 
> What about getting the begin and end positions of the <grammar> tag in the
> buffer (whatever that is)? So that the contents can be copied out.
> 
> lin
> 
> ----- Original Message -----
> From: "Allen, Michael B (RSCH)" <Michael_B_Allen@ml.com>
> To: "'Lin Shen'" <lshen@cisco.com>; <expat-discuss@lists.sourceforge.net>
> Sent: Tuesday, August 21, 2001 6:16 PM
> Subject: RE: [Expat-discuss] return tag contents as it is
> 
> 
> > I don't think so. That would be an extra feature of an XML parser. Perhaps
> you
> > can use the special CDATA tag into which you may put anything including
> xml.
> >
> > > -----Original Message-----
> > > From: Lin Shen [SMTP:lshen@cisco.com]
> > > Sent: Tuesday, August 21, 2001 9:11 PM
> > > To: expat-discuss@lists.sourceforge.net
> > > Subject: [Expat-discuss] return tag contents as it is
> > >
> > > Hi,
> > >
> > > Is it possible to have the parser return the contents of certain tags as
> it
> > > is? For instance, for the following xml segment
> > >
> > > <grammar>
> > >     <rule id="root">
> > >         text <ruleref tag="&lt tag &gt"/>
> > >     </rule>
> > > </grammar>
> > >
> > > is there is a way to let the parser to return whatever contained in
> > > <grammar> tag w/out any parsing, including the escape char and the
> <rule>
> > > tag?
> > >
> > > Please send reply to my personal account lshen@cisco.com
> > >
> > > thanks
> > > lin
> > >
> > >
> > > _______________________________________________
> > > Expat-discuss mailing list
> > > Expat-discuss@lists.sourceforge.net
> > > http://lists.sourceforge.net/lists/listinfo/expat-discuss
> >
> 



From syprat@yahoo.fr  Wed Aug 22 08:10:13 2001
From: syprat@yahoo.fr (=?iso-8859-1?q?Sylvain=20PRAT?=)
Date: Wed, 22 Aug 2001 09:10:13 +0200 (CEST)
Subject: [Expat-discuss] return tag contents as it is
Message-ID: <20010822071013.90450.qmail@web14808.mail.yahoo.com>

---------------------- multipart/mixed attachment
Hi,

Yes it is possible (i do it in my application).
You must use the XML_GetInputContext function
associated with the XML_GetCurrentByteCount to rebuild
a tag parsed, but you can't get what is inside your
grammar tag easily, so you must rebuild the inside of
the tag with appropriate callbacks (i set a cdata,
start element and end element, which concat all the
strings together - i had to ignore xhtml tags in some
tags)

Bye,

___________________________________________________________
Do You Yahoo!? -- Vos albums photos en ligne, 
Yahoo! Photos : http://fr.photos.yahoo.com
---------------------- multipart/mixed attachment--


From lshen@cisco.com  Fri Aug 24 03:33:56 2001
From: lshen@cisco.com (Lin Shen)
Date: Thu, 23 Aug 2001 19:33:56 -0700
Subject: [Expat-discuss] XML_GetInputContext
Message-ID: <000f01c12c45$394763b0$738b6b80@cisco.com>

Hi,

I'm trying to add XML_GetInputContext api to Expat 1.1.
By just copying the 1.2 code  doesn't work and that's all I experimented so
far.
Any pointers?

thanks
lin



From gstein@lyra.org  Fri Aug 24 04:24:45 2001
From: gstein@lyra.org (Greg Stein)
Date: Thu, 23 Aug 2001 20:24:45 -0700
Subject: [Expat-discuss] XML_GetInputContext
In-Reply-To: <000f01c12c45$394763b0$738b6b80@cisco.com>; from lshen@cisco.com on Thu, Aug 23, 2001 at 07:33:56PM -0700
References: <000f01c12c45$394763b0$738b6b80@cisco.com>
Message-ID: <20010823202445.O26054@lyra.org>

On Thu, Aug 23, 2001 at 07:33:56PM -0700, Lin Shen wrote:
> Hi,
> 
> I'm trying to add XML_GetInputContext api to Expat 1.1.
> By just copying the 1.2 code  doesn't work and that's all I experimented so
> far.
> Any pointers?

Upgrade your Expat. The APIs are compatible, so why not upgrade?

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/


From lshen@cisco.com  Fri Aug 24 06:04:00 2001
From: lshen@cisco.com (Lin Shen)
Date: Thu, 23 Aug 2001 22:04:00 -0700
Subject: [Expat-discuss] XML_GetInputContext
References: <000f01c12c45$394763b0$738b6b80@cisco.com> <20010823202445.O26054@lyra.org>
Message-ID: <000e01c12c5a$300b72e0$7ccf150a@cisco.com>

We've ported the parser onto another platform, and it's a hell lot of work
to upgrade the parser in terms of technical and legal issues.

lin

----- Original Message -----
From: "Greg Stein" <gstein@lyra.org>
To: "Lin Shen" <lshen@cisco.com>
Cc: <expat-discuss@lists.sourceforge.net>
Sent: Thursday, August 23, 2001 8:24 PM
Subject: Re: [Expat-discuss] XML_GetInputContext


> On Thu, Aug 23, 2001 at 07:33:56PM -0700, Lin Shen wrote:
> > Hi,
> >
> > I'm trying to add XML_GetInputContext api to Expat 1.1.
> > By just copying the 1.2 code  doesn't work and that's all I experimented
so
> > far.
> > Any pointers?
>
> Upgrade your Expat. The APIs are compatible, so why not upgrade?
>
> Cheers,
> -g
>
> --
> Greg Stein, http://www.lyra.org/



From gstein@lyra.org  Fri Aug 24 09:31:35 2001
From: gstein@lyra.org (Greg Stein)
Date: Fri, 24 Aug 2001 01:31:35 -0700
Subject: [Expat-discuss] XML_GetInputContext
In-Reply-To: <000e01c12c5a$300b72e0$7ccf150a@cisco.com>; from lshen@cisco.com on Thu, Aug 23, 2001 at 10:04:00PM -0700
References: <000f01c12c45$394763b0$738b6b80@cisco.com> <20010823202445.O26054@lyra.org> <000e01c12c5a$300b72e0$7ccf150a@cisco.com>
Message-ID: <20010824013135.D27852@lyra.org>

Ah... yah, I can see how that could be a problem :-)

I think it would be helpful if you explain what "doesn't work" means.
Compile error? The code doesn't run right? What kind of error...

Cheers,
-g

On Thu, Aug 23, 2001 at 10:04:00PM -0700, Lin Shen wrote:
> We've ported the parser onto another platform, and it's a hell lot of work
> to upgrade the parser in terms of technical and legal issues.
> 
> lin
> 
> ----- Original Message -----
> From: "Greg Stein" <gstein@lyra.org>
> To: "Lin Shen" <lshen@cisco.com>
> Cc: <expat-discuss@lists.sourceforge.net>
> Sent: Thursday, August 23, 2001 8:24 PM
> Subject: Re: [Expat-discuss] XML_GetInputContext
> 
> 
> > On Thu, Aug 23, 2001 at 07:33:56PM -0700, Lin Shen wrote:
> > > Hi,
> > >
> > > I'm trying to add XML_GetInputContext api to Expat 1.1.
> > > By just copying the 1.2 code  doesn't work and that's all I experimented
> so
> > > far.
> > > Any pointers?
> >
> > Upgrade your Expat. The APIs are compatible, so why not upgrade?
> >
> > Cheers,
> > -g
> >
> > --
> > Greg Stein, http://www.lyra.org/
> 
> 
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss

-- 
Greg Stein, http://www.lyra.org/


From kfox@vulpes.com  Sun Aug 26 00:58:16 2001
From: kfox@vulpes.com (Ken Fox)
Date: Sat, 25 Aug 2001 19:58:16 -0400
Subject: [Expat-discuss] Documentation on Expat internals?
Message-ID: <3B883B98.CC3BA797@vulpes.com>

I've been using Expat for quite a while now under the
guise of the Perl XML::Parser module. Just recently, i.e.
today, I started getting into Expat internals.

Are there any docs available?

I've started some bits on the input buffer and I'm wondering
if anybody else is doing this already. Google didn't turn
anything up, which seems remarkable considering how popular
Expat is.

My docs are at <http://www.msen.com/~fox/expat-internals/>.

The thing that started me looking deeper into Expat is
a performance problem with XML::Parser. I've always thought
it was "fast enough", but recently (yesterday!) I had to
replace it with a nest of regular expressions. Something's
not right with that.

The first part of the plan is to reduce buffer copying and
improve I/O performance because that looks pretty easy. The
second is to speed up the callbacks by eliminating some
of the Perl sub call overhead. (It will be faster, but less
general.)

- Ken


From shashank@arkionsystems.com  Mon Aug 27 02:24:39 2001
From: shashank@arkionsystems.com (Shashank Banerjea)
Date: Sun, 26 Aug 2001 21:24:39 -0400
Subject: [Expat-discuss] Basic Build Question on expat on VxWorks
Message-ID: <000101c12e97$0afe6c70$04d95b18@arkserver>

This is a multi-part message in MIME format.

---------------------- multipart/alternative attachment
Hi,
 
I downloaded the latest source code for expat XML parser from the Source
Forge website.
I plan to use this act to an XML parser on VxWorks to parse incoming XML
messages.
 
Anyone has pointers on building expat on VxWorks 5.4 (Tornado 2.0) with
gnu C complier?
 
Thanks in advance
 
With best regards
Shashank Banerjea
 
 
 

---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: http://mail.libexpat.org/pipermail-21/expat-discuss/attachments/20010826/c424a483/attachment.html

---------------------- multipart/alternative attachment--



From mrityu@aplion.stpn.soft.net  Mon Aug 27 05:00:28 2001
From: mrityu@aplion.stpn.soft.net (Mrityunjay Kumar)
Date: Mon, 27 Aug 2001 09:30:28 +0530
Subject: [Expat-discuss] Basic Build Question on expat on VxWorks
In-Reply-To: <000101c12e97$0afe6c70$04d95b18@arkserver>
Message-ID: <007401c12eac$cfb7b3e0$08080808@mrityunjay>

This is a multi-part message in MIME format.

---------------------- multipart/alternative attachment
Well, I built it recently, without any hitch; the core part of the parser
(expat/lib). A good idea is to first build it on a linux machine (using the
Configure script was straightforward for me), try it to verify it works, and
then take the generated config.h and the makefile to VxWorks.

Hope it helps.
-Mrityunjay

  -----Original Message-----
  From: expat-discuss-admin@lists.sourceforge.net
[mailto:expat-discuss-admin@lists.sourceforge.net]On Behalf Of Shashank
Banerjea
  Sent: Monday, August 27, 2001 6:55 AM
  To: expat-discuss@lists.sourceforge.net
  Subject: [Expat-discuss] Basic Build Question on expat on VxWorks


  Hi,



  I downloaded the latest source code for expat XML parser from the Source
Forge website.

  I plan to use this act to an XML parser on VxWorks to parse incoming XML
messages.



  Anyone has pointers on building expat on VxWorks 5.4 (Tornado 2.0) with
gnu C complier?



  Thanks in advance



  With best regards

  Shashank Banerjea








---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: http://mail.libexpat.org/pipermail-21/expat-discuss/attachments/20010827/f3e76755/attachment.html

---------------------- multipart/alternative attachment--



From ghein@thisisa.com  Tue Aug 28 00:03:29 2001
From: ghein@thisisa.com (Glen Hein)
Date: Mon, 27 Aug 2001 16:03:29 -0700 (MST)
Subject: [Expat-discuss] whitespace trouble
Message-ID: <998953409.3b8ad1c161583@thisisa.com>

Hello Everybody!

I'm new to expat and I recently adopted a project that was
using expat.  Unfortunately, I do not have contact with the
orginal programmer.  

I've having a problem with whitespace in the xml structure that
is to be parsed.  It appears that any whitespace bewtween the 
xml tags casuse the parser to fail.  

I have not yet determined if the problem is in expat or in the
overall application.  

Is there a known problem with whitespace?  Is there a particular
area of expat that I should be looking at?  Is whitespace an
option in expat?

Thanks,
Glen Hein
ghein@thisisa.com




From michael@vivtek.com  Tue Aug 28 00:11:20 2001
From: michael@vivtek.com (Michael Roberts)
Date: Mon, 27 Aug 2001 18:11:20 -0500
Subject: [Expat-discuss] whitespace trouble
References: <998953409.3b8ad1c161583@thisisa.com>
Message-ID: <3B8AD398.F89A652E@vivtek.com>

I've certainly never had any trouble with whitespace.  Would it be
possible for you to include a sample of a *small* XML document which
causes a failure, and the error code you get?  That would help narrow
things down.

Glen Hein wrote:

> Hello Everybody!
>
> I'm new to expat and I recently adopted a project that was
> using expat.  Unfortunately, I do not have contact with the
> orginal programmer.
>
> I've having a problem with whitespace in the xml structure that
> is to be parsed.  It appears that any whitespace bewtween the
> xml tags casuse the parser to fail.
>
> I have not yet determined if the problem is in expat or in the
> overall application.
>
> Is there a known problem with whitespace?  Is there a particular
> area of expat that I should be looking at?  Is whitespace an
> option in expat?
>
> Thanks,
> Glen Hein
> ghein@thisisa.com
>
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss



From Michael_B_Allen@ml.com  Tue Aug 28 00:41:58 2001
From: Michael_B_Allen@ml.com (Allen, Michael B (RSCH))
Date: Mon, 27 Aug 2001 19:41:58 -0400
Subject: [Expat-discuss] whitespace trouble
Message-ID: <B27EB33BAB29D2119ABF0001FA7EF289053BF19E@ewfd04.exchange.ml.com>

Be very carefull about the code in your CharacterDataHandler
function. The strings passed to it are NOT null terminated and
may not be all of the text (may be broken into adjacent
peices). This is a great place for an error. See how DOMC
handles it in the chardata_fn when building the DOM tree here:

http://auditorymodels.org/domc/src/lib/expatls.c

Mike
> -----Original Message-----
> From:	Glen Hein [SMTP:ghein@thisisa.com]
> Sent:	Monday, August 27, 2001 7:03 PM
> To:	expat-discuss@lists.sourceforge.net
> Subject:	[Expat-discuss] whitespace trouble
> 
> 
> Hello Everybody!
> 
> I'm new to expat and I recently adopted a project that was
> using expat.  Unfortunately, I do not have contact with the
> orginal programmer.  
> 
> I've having a problem with whitespace in the xml structure that
> is to be parsed.  It appears that any whitespace bewtween the 
> xml tags casuse the parser to fail.  
> 
> I have not yet determined if the problem is in expat or in the
> overall application.  
> 
> Is there a known problem with whitespace?  Is there a particular
> area of expat that I should be looking at?  Is whitespace an
> option in expat?
> 
> Thanks,
> Glen Hein
> ghein@thisisa.com
> 
> 
> 
> _______________________________________________
> Expat-discuss mailing list
> Expat-discuss@lists.sourceforge.net
> http://lists.sourceforge.net/lists/listinfo/expat-discuss



From ghein@thisisa.com  Tue Aug 28 02:51:13 2001
From: ghein@thisisa.com (Glen Hein)
Date: Mon, 27 Aug 2001 18:51:13 -0700 (MST)
Subject: [Expat-discuss] whitespace trouble
In-Reply-To: <B27EB33BAB29D2119ABF0001FA7EF289053BF19E@ewfd04.exchange.ml.com>
References: <B27EB33BAB29D2119ABF0001FA7EF289053BF19E@ewfd04.exchange.ml.com>
Message-ID: <998963473.3b8af911f17c1@thisisa.com>

I'd like to thank you from the bottom of my keyboard :-)

I examined the fundamental differences between my 
CharacterDataHandler and the version you cited.  All I had
to do was to have my handler skip the whitespace:

for (i = 0; i < len && !isgraph(s[i]); i++) {
  ;
}
if (i == len) {
  return;
}

-Glen Hein

Quoting "Allen, Michael B (RSCH)" <Michael_B_Allen@ml.com>:

> Be very carefull about the code in your CharacterDataHandler
> function. The strings passed to it are NOT null terminated and
> may not be all of the text (may be broken into adjacent
> peices). This is a great place for an error. See how DOMC
> handles it in the chardata_fn when building the DOM tree here:
> 
> http://auditorymodels.org/domc/src/lib/expatls.c
> 
> Mike
> > -----Original Message-----
> > From:	Glen Hein [SMTP:ghein@thisisa.com]
> > Sent:	Monday, August 27, 2001 7:03 PM
> > To:	expat-discuss@lists.sourceforge.net
> > Subject:	[Expat-discuss] whitespace trouble
> > 
> > 
> > Hello Everybody!
> > 
> > I'm new to expat and I recently adopted a project that was
> > using expat.  Unfortunately, I do not have contact with the
> > orginal programmer.  
> > 
> > I've having a problem with whitespace in the xml structure that
> > is to be parsed.  It appears that any whitespace bewtween the 
> > xml tags casuse the parser to fail.  
> > 
> > I have not yet determined if the problem is in expat or in the
> > overall application.  
> > 
> > Is there a known problem with whitespace?  Is there a particular
> > area of expat that I should be looking at?  Is whitespace an
> > option in expat?
> > 
> > Thanks,
> > Glen Hein
> > ghein@thisisa.com
> > 
> > 
> > 
> > _______________________________________________
> > Expat-discuss mailing list
> > Expat-discuss@lists.sourceforge.net
> > http://lists.sourceforge.net/lists/listinfo/expat-discuss
> 


From Michael_B_Allen@ml.com  Tue Aug 28 02:53:16 2001
From: Michael_B_Allen@ml.com (Allen, Michael B (RSCH))
Date: Mon, 27 Aug 2001 21:53:16 -0400
Subject: [Expat-discuss] whitespace trouble
Message-ID: <B27EB33BAB29D2119ABF0001FA7EF289053BF1A3@ewfd04.exchange.ml.com>

You're welcome. Ironically this is a peice of code that should be taken out of DOMC because it assumes the user does not need it. But that's my problem.

Later,
Mike

> -----Original Message-----
> From:	Glen Hein [SMTP:ghein@thisisa.com]
> Sent:	Monday, August 27, 2001 9:51 PM
> To:	Allen, Michael B (RSCH)
> Cc:	'Glen Hein'; expat-discuss@lists.sourceforge.net
> Subject:	RE: [Expat-discuss] whitespace trouble
> 
> 
> I'd like to thank you from the bottom of my keyboard :-)
> 
> I examined the fundamental differences between my 
> CharacterDataHandler and the version you cited.  All I had
> to do was to have my handler skip the whitespace:
> 
> for (i = 0; i < len && !isgraph(s[i]); i++) {
>   ;
> }
> if (i == len) {
>   return;
> }
> 
> -Glen Hein
> 
> Quoting "Allen, Michael B (RSCH)" <Michael_B_Allen@ml.com>:
> 
> > Be very carefull about the code in your CharacterDataHandler
> > function. The strings passed to it are NOT null terminated and
> > may not be all of the text (may be broken into adjacent
> > peices). This is a great place for an error. See how DOMC
> > handles it in the chardata_fn when building the DOM tree here:
> > 
> > http://auditorymodels.org/domc/src/lib/expatls.c
> > 
> > Mike
> > > -----Original Message-----
> > > From:	Glen Hein [SMTP:ghein@thisisa.com]
> > > Sent:	Monday, August 27, 2001 7:03 PM
> > > To:	expat-discuss@lists.sourceforge.net
> > > Subject:	[Expat-discuss] whitespace trouble
> > > 
> > > 
> > > Hello Everybody!
> > > 
> > > I'm new to expat and I recently adopted a project that was
> > > using expat.  Unfortunately, I do not have contact with the
> > > orginal programmer.  
> > > 
> > > I've having a problem with whitespace in the xml structure that
> > > is to be parsed.  It appears that any whitespace bewtween the 
> > > xml tags casuse the parser to fail.  
> > > 
> > > I have not yet determined if the problem is in expat or in the
> > > overall application.  
> > > 
> > > Is there a known problem with whitespace?  Is there a particular
> > > area of expat that I should be looking at?  Is whitespace an
> > > option in expat?
> > > 
> > > Thanks,
> > > Glen Hein
> > > ghein@thisisa.com
> > > 
> > > 
> > > 
> > > _______________________________________________
> > > Expat-discuss mailing list
> > > Expat-discuss@lists.sourceforge.net
> > > http://lists.sourceforge.net/lists/listinfo/expat-discuss
> > 



From shashank@arkionsystems.com  Tue Aug 28 05:14:54 2001
From: shashank@arkionsystems.com (Shashank Banerjea)
Date: Tue, 28 Aug 2001 00:14:54 -0400
Subject: [Expat-discuss] Basic Build Question on expat on VxWorks
References: <007401c12eac$cfb7b3e0$08080808@mrityunjay>
Message-ID: <006c01c12f77$fe155870$0adb5b18@SBVAIO>

Thanks Mr. Mrityunjay Kumar . Before I take this step, I clarify one doubt,
I am on a x86 architecture on Linux while my target is VxWorks on ARM. Will
it require tweaking the makefile?

Shashank
----- Original Message -----
From: Mrityunjay Kumar
To: 'Shashank Banerjea' ; expat-discuss@lists.sourceforge.net
Sent: Monday, August 27, 2001 12:00 AM
Subject: RE: [Expat-discuss] Basic Build Question on expat on VxWorks


Well, I built it recently, without any hitch; the core part of the parser
(expat/lib). A good idea is to first build it on a linux machine (using the
Configure script was straightforward for me), try it to verify it works, and
then take the generated config.h and the makefile to VxWorks.

Hope it helps.
-Mrityunjay

-----Original Message-----
From: expat-discuss-admin@lists.sourceforge.net
[mailto:expat-discuss-admin@lists.sourceforge.net]On Behalf Of Shashank
Banerjea
Sent: Monday, August 27, 2001 6:55 AM
To: expat-discuss@lists.sourceforge.net
Subject: [Expat-discuss] Basic Build Question on expat on VxWorks


Hi,

I downloaded the latest source code for expat XML parser from the Source
Forge website.
I plan to use this act to an XML parser on VxWorks to parse incoming XML
messages.

Anyone has pointers on building expat on VxWorks 5.4 (Tornado 2.0) with gnu
C complier?

Thanks in advance

With best regards
Shashank Banerjea






From JWieman@daktronics.com  Wed Aug 29 15:52:35 2001
From: JWieman@daktronics.com (Joe Wiemann)
Date: Wed, 29 Aug 2001 09:52:35 -0500
Subject: [Expat-discuss] How can I add a function to skip a set number of bytes
Message-ID: <sb8cbb7c.064@gw.daktronics.com>

How can I add a function to skip a set number of bytes in the expat parser =
buffer...


Such as if I get a certain begin tag ? it gives me a length of text that =
it has in it.

I want to skip the text that is in the tag.




From mballen@erols.com  Wed Aug 29 18:43:57 2001
From: mballen@erols.com (Michael B. Allen)
Date: Wed, 29 Aug 2001 13:43:57 -0400
Subject: [Expat-discuss] How can I add a function to skip a set number of bytes
In-Reply-To: <sb8cbb7c.064@gw.daktronics.com>; from JWieman@daktronics.com on Wed, Aug 29, 2001 at 09:52:35AM -0500
References: <sb8cbb7c.064@gw.daktronics.com>
Message-ID: <20010829134357.C836@nano.foo.net>

On Wed, Aug 29, 2001 at 09:52:35AM -0500, Joe Wiemann wrote:
> How can I add a function to skip a set number of bytes in the expat parser buffer...
> 
> Such as if I get a certain begin tag ? it gives me a length of text that it has in it.
> 
> I want to skip the text that is in the tag.

You don't want to "skip some bytes", you want to skip that tag and all
it's children? This is very simple to do with Expat. Look at the outline.c
or elements.c example programs that come with Expat. Modify the example
to pass a flag with or as the user ponter and use it to indicate that
the input should(n't) be ignored. IOW, start with the flag off. Test
it at the beginning of you start tag handler. If it's off, process as
normal. When you see that flag on, don't do anything in the begin tag
handler, just return. When you see the end tag for that element, flip
the flag off. Get it?

I beleive the Expat tutorial at XML.com describes exactly this problem.

Mike

-- 
Wow a memory-mapped fork bomb! Now what on earth did you expect? - lkml


From mballen@erols.com  Thu Aug 30 19:24:12 2001
From: mballen@erols.com (Michael B. Allen)
Date: Thu, 30 Aug 2001 14:24:12 -0400
Subject: [Expat-discuss] How can I add a function to skip a set number of bytes
In-Reply-To: <sb8cf0ed.015@gw.daktronics.com>; from JWieman@daktronics.com on Wed, Aug 29, 2001 at 01:40:27PM -0500
References: <sb8cf0ed.015@gw.daktronics.com>
Message-ID: <20010830142412.A932@nano.foo.net>

Sorry if I lost track of the thread but did you participate in the
recent XML_GetInputContext discussion? If not, maybe that's something
you can use?

Mike

On Wed, Aug 29, 2001 at 01:40:27PM -0500, Joe Wiemann wrote:
> But I cannot 100 percent guarantee that the closing tag won't appear
within the text. so it would be better if I could just skip a set number
of bytes in the parser buffer.
> 
> You don't want to "skip some bytes", you want to skip that tag and all
> it's children? This is very simple to do with Expat. Look at the outline.c

-- 
Wow a memory-mapped fork bomb! Now what on earth did you expect? - lkml