libxmlplusplus-general — Open discusions about libXML++

Re: [libxml++] Portability patches for 1.0.1

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Albert Chin a =E9crit :

| The patches below were required to get libxml++ build using the
| Sun, HP, SGI, and AIX C++ compilers.


Thanks very much. The patches are now in the CVS repositery (in both
branches).

|
| Why do you add @LIBXML_LIBS@ to Libs: in libxml++-1.0.pc.in? Why
| not add libxml-2.0 to the Requires: line? pkg-config will do the
| right thing and search for libxml-2.0.pc.


I've got no idea, since I did not created the file. I changed it as
you suggested.


Thanks for your contribution,

Christophe

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAMJFPB+sU3TyOQjARAoRhAJ4qmXqiz/8GxZ0J1OEkXtN2Igc2KgCffIb8
sgP6+n2RPHnmOKEcVOPXKgo=3D
=3DKso3
-----END PGP SIGNATURE-----

[libxml++] Portability patches for 1.0.1

[Albert C. <lib...@ml...>]

The patches below were required to get libxml++ build using the Sun,
HP, SGI, and AIX C++ compilers.

Why do you add @LIBXML_LIBS@ to Libs: in libxml++-1.0.pc.in? Why not
add libxml-2.0 to the Requires: line? pkg-config will do the right
thing and search for libxml-2.0.pc.

Also, libtool 1.5.x won't correctly build libxml++ using the Sun C++
compiler. I've submitted a patch to lib...@gn... so
hopefully it'll make it into 1.5.3. The problem is that convenience
libraries are not passed correctly to the linker so all the symbols
from the convenience libraries don't make their way into libxml++.so.

-- 
albert chin (ch...@th...)

-- snip snip
--- libxml++/nodes/node.h.orig	2004-02-15 21:46:51.285069000 -0600
+++ libxml++/nodes/node.h	2004-02-15 21:46:55.536383000 -0600
@@ -62,7 +62,7 @@
   /** Discover at what line number this node occurs in the XML file.
    * @returns The line number.
    */
-  int get_line() const;;
+  int get_line() const;
 
   /** Obtain the list of child nodes. You may optionally obtain a list of only the child nodes which have a certain name.
    * @param name The names of the child nodes to get. If you do not specigy a name, then the list will contain all nodes, regardless of their names.
--- libxml++/nodes/node.cc.orig	2004-02-16 01:50:36.618977000 -0600
+++ libxml++/nodes/node.cc	2004-02-16 01:50:42.811534000 -0600
@@ -4,6 +4,7 @@
  * included with libxml++ as the file COPYING.
  */
 
+#include <libxml++/nodes/element.h>
 #include <libxml++/nodes/node.h>
 #include <libxml++/exceptions/internal_error.h>
 #include <iostream>
--- libxml++/parsers/parser.h.orig	2004-02-15 23:56:42.619652000 -0600
+++ libxml++/parsers/parser.h	2004-02-16 00:59:01.690901000 -0600
@@ -15,6 +15,8 @@
 #include <libxml++/exceptions/validity_error.h>
 #include <libxml++/exceptions/internal_error.h>
 
+#include <istream>
+
 #ifdef WIN32 //TODO: Why do we do this? murrayc.
 #define vsnprintf _vsnprintf
 #endif

Re: [libxml++] A few minor CVS issues

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jim Garrison a =E9crit :

| Hello everyone,
|
| There's a few things that I would like to do to the CVS tree, and I
| will do these if there are no complaints.
|

No objection for both issues.

| 1. Add .cvsignore files
|
| 2. Remove unnecessary semicolons.  g++ 3.4 (snapshots available)
| has made "-pedantic" error out on unnecessary semicolons.  All but
| one of these in the libxml++ code is in the following form:
|
| extern "C" { #include "something.h" };
|
| or...
|
| namespace xmlpp { // stuff };
|
| The semicolon at the end is unnecessary in each one.  If nobody
| objects, I'll commit a patch that removes these semicolons.


Thanks,

Christophe

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFALKfiB+sU3TyOQjARAgxcAJoCrBXe0iqlFvUhDW1rHek03gdysQCbB/v2
ugThvA+OiFtnPFB1dNdKmVc=3D
=3DXLyB
-----END PGP SIGNATURE-----

[libxml++] A few minor CVS issues

[Jim G. <gar...@ca...>]

Hello everyone,

There's a few things that I would like to do to the CVS tree, and I will
do these if there are no complaints.

1. Add .cvsignore files

2. Remove unnecessary semicolons.  g++ 3.4 (snapshots available) has
made "-pedantic" error out on unnecessary semicolons.  All but one of
these in the libxml++ code is in the following form:

extern "C" {
#include "something.h"
};

or...

namespace xmlpp
{
// stuff
};

The semicolon at the end is unnecessary in each one.  If nobody objects,
I'll commit a patch that removes these semicolons.

Jim

[libxml++] Validation grammar languages

[Gunther L. <gu...@sb...>]

Hi!

First: Thank you for libxml++! Its really great to have an easy portable
XML C++ API.=20

But now my question: I try to validate my XML File. The documentation
tells me about Parser::set_validate, but does not explain which grammar
languages are possible. Is this limited to DTDs or are XML Schema or RNG
useable?

Thanks,

Gunther

Re: [libxml++] Running under Windows XP?

[Philipp K. <lib...@py...>]

Hi Ephraim,

> take a look at:  http://www.cygwin.com/ml/cygwin/2002-12/msg00686.html
> it seems this message has to do with file access permissions.
> 
> do the dlls have the correct access permissions?
> are they accessible at all in the path?

Thanks a lot, that helped. One of the DLLs (a downloaded binary one) 
hadn't set the executable flag. I searched on Google for the error 
message and probably got that page as a result too but looked over it 
because the error occured only in MSVC and explicitly NOT in cygwin.

Cool operating system giving a developer hints like that if something 
went wrong... ;-)

Thanks again.

Philipp

Re: [libxml++] Running under Windows XP?

[Ephraim V. <ef...@ep...>]

Hi,

take a look at:  http://www.cygwin.com/ml/cygwin/2002-12/msg00686.html
it seems this message has to do with file access permissions.

do the dlls have the correct access permissions?
are they accessible at all in the path?

-eff

----- Original Message ----- 
From: "Philipp Klaus" <lib...@py...>
To: <lib...@li...>
Sent: Friday, February 06, 2004 4:42 PM
Subject: Re: [libxml++] Running under Windows XP?


> Hi Ephraim,
>
> thank you for your help.
>
> > I used libxml 2.6.5 and iconv 1.9.1 prebuilt binaries -  both from
> > http://www.zlatkovic.com/libxml.en.html
> > Note that with the latest builds of libxml2 you also need zlib.dll to
run
> > the programs. (headers and library are not used)
> >
> > First try to use a more updated iconv, then make sure you compiled
libxml2
> > with thread support, I'm not sure if this is the default, but this is
the
> > setting in the prebuilt binaries.
>
> I used now the precompiled binaries from zlatkovic.com but still get the
> same annoying error message. I compiled the first five examples projects
> in the workspace but each of them produces the this error message,
> always with the same "address" (?) of 0xc0000022.
>
> To repeat, this is the message I get:
> "The application failed to initialize properly (0xc0000022). Click OK to
> terminate the application."
>
> Does anybody know what the error message means? What's the sense of that
> address? Is this from the loader, from the C++ initialization stage or
> is already normal code running?
>
> I get that same error message even when I start the demo applications
> with the debugger (the one built into MSVC-6), but perhaps that's
> because I don't know how to handle the debugger, it feels strange if
> you're used to gdb :-)
>
>
> Thanx
>
> Philipp
>
>

Re: [libxml++] Running under Windows XP?

[Philipp K. <lib...@py...>]

Hi Ephraim,

thank you for your help.

> I used libxml 2.6.5 and iconv 1.9.1 prebuilt binaries -  both from
> http://www.zlatkovic.com/libxml.en.html
> Note that with the latest builds of libxml2 you also need zlib.dll to run
> the programs. (headers and library are not used)
> 
> First try to use a more updated iconv, then make sure you compiled libxml2
> with thread support, I'm not sure if this is the default, but this is the
> setting in the prebuilt binaries.

I used now the precompiled binaries from zlatkovic.com but still get the 
same annoying error message. I compiled the first five examples projects 
in the workspace but each of them produces the this error message, 
always with the same "address" (?) of 0xc0000022.

To repeat, this is the message I get:
"The application failed to initialize properly (0xc0000022). Click OK to 
terminate the application."

Does anybody know what the error message means? What's the sense of that 
address? Is this from the loader, from the C++ initialization stage or 
is already normal code running?

I get that same error message even when I start the demo applications 
with the debugger (the one built into MSVC-6), but perhaps that's 
because I don't know how to handle the debugger, it feels strange if 
you're used to gdb :-)


Thanx

Philipp

Re: [libxml++] My proposed patches

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jim Garrison wrote:

| On Thu, 2004-02-05 at 12:46, Christophe de VIENNE wrote:
|
|> | The first patch adds cobj() to Document.
|>
|> Is it a real need to have it ? I'd prefer not to introduce it.
|
|
| There is no need for it as long as the API is and always will be
| all-encompassing.  In the same way, there is no need for
| Node::cobj() as far as I know.  If Document::cobj() existed in 1.0,
|  it would be easier to work around the lack of the
| create_root_node_by_import function.
|
| Really, I just consider it standard that the application programmer
|  can access the underlying C object if necessary, and I question
| why it is allowed in Node but not Document.  At the same time, I
| cannot speculate on how useful the Document::cobj() function will
| be in the future.


We had long threads about this subject some time ago. But you're right
the lib is not consistent by providing a cobj() on Node and not on
Document. Let's add it.

Cheers,

Christophe




-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAI2X7B+sU3TyOQjARAnvRAJ40mcMwwVv8lCH3csP+G2Lu32seMwCaA62L
TkbzQMmcg0MEJ4MRGvSFeFk=
=yX+S
-----END PGP SIGNATURE-----

Re: [libxml++] My proposed patches

[Jim G. <gar...@ca...>]

On Thu, 2004-02-05 at 12:46, Christophe de VIENNE wrote:
> | The first patch adds cobj() to Document.
>
> Is it a real need to have it ? I'd prefer not to introduce it.

There is no need for it as long as the API is and always will be
all-encompassing.  In the same way, there is no need for Node::cobj() as
far as I know.  If Document::cobj() existed in 1.0, it would be easier
to work around the lack of the create_root_node_by_import function.

Really, I just consider it standard that the application programmer can
access the underlying C object if necessary, and I question why it is
allowed in Node but not Document.  At the same time, I cannot speculate
on how useful the Document::cobj() function will be in the future.

Jim

Re: [libxml++] Running under Windows XP?

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ephraim Vider wrote:

|
| First try to use a more updated iconv, then make sure you compiled
| libxml2 with thread support, I'm not sure if this is the default,
| but this is the setting in the prebuilt binaries.


I think it's the default since version 2.5.7


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAIoZSB+sU3TyOQjARAkC5AJ46oXau56Zakaw4CTg4iYLOUWlJbQCg5/Jk
fWH85qDvgi4FSeIf4ScoHv0=
=Ek4x
-----END PGP SIGNATURE-----

Re: [libxml++] Running under Windows XP?

[Ephraim V. <ef...@ep...>]

I tested 1.0.1 on win2k and the example runs fine.

I used libxml 2.6.5 and iconv 1.9.1 prebuilt binaries -  both from
http://www.zlatkovic.com/libxml.en.html
Note that with the latest builds of libxml2 you also need zlib.dll to run
the programs. (headers and library are not used)

First try to use a more updated iconv, then make sure you compiled libxml2
with thread support, I'm not sure if this is the default, but this is the
setting in the prebuilt binaries.

-eff

----- Original Message ----- 
From: "Philipp Klaus" <lib...@py...>
To: <lib...@li...>
Sent: Thursday, February 05, 2004 6:51 PM
Subject: [libxml++] Running under Windows XP?


> Hi all,
>
> I tried to compile and run libxml++ in version 1.0.1 under Windows XP
> (MS Visual C++ 6). I developed an application using this library. This
> compiles fine but when I run the application I get
>
> "The application failed to initialize properly (0xc0000022). Click OK to
> terminate the application."
>
> After checking my application I tried the dom_xpath example program
> (using the supplied MSVC6 project) and I got exactly the same error
message.
>
> I'm using libxml++-1.0.1 (compiled using the supplied MSVC project),
> libxml2-2.6.5 (compiled using nmake and the makefile.vc in the
> distribution) and libiconv-1.7 (binary version included in the binary
> Sablotron distribution). The OS is WinXP with SP1. VC++ is version 6.0.
>
> Has anyone used libxml++ on Windows XP before (I suppose yes because
> there is a workspace with all example projects in the win32_msvc6
> folder)? Is the error message mentioned before known to somebody? What
> does it mean?
>
> Please note that I'm not very familiar with Windows development, my
> applications get developed under Linux but this one must be used under
> Win32 too, so I'm obliged to care about this.
>
> Thanks to everyone in advance.
>
> Philipp
>
>
>
> -------------------------------------------------------
> The SF.Net email is sponsored by EclipseCon 2004
> Premiere Conference on Open Tools Development and Integration
> See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
> http://www.eclipsecon.org/osdn
> _______________________________________________
> Libxmlplusplus-general mailing list
> Lib...@li...
> https://lists.sourceforge.net/lists/listinfo/libxmlplusplus-general
>

Re: [libxml++] My proposed patches

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jim Garrison wrote:

| The first patch adds cobj() to Document.


Is it a real need to have it ? I'd prefer not to introduce it.

| The second one adds the function
| Document::create_root_node_by_import.  I decided to leave the long
| name and not overload create_root_node for the sake of clarity; in
| the same manner, Node::import_node and Node::add_child are two
| differently named functions with analogous functionality.


Looks good to me.

|
| If you give me the word, I'll go ahead and commit it to the module,
|  or feel free to do it on your own.


If you have an access to gnome-cvs, please commit your second patch,
but don't forget to update and commit the ChangeLog at the same time.

Thanks,

Christophe

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAIoFzB+sU3TyOQjARAohcAKDVe4C/X8cZifC63aynO3G8y4gHGgCeODvE
pl6PQl3ova4lY7mbF4mPDKg=
=pH/2
-----END PGP SIGNATURE-----

[libxml++] Running under Windows XP?

[Philipp K. <lib...@py...>]

Hi all,

I tried to compile and run libxml++ in version 1.0.1 under Windows XP 
(MS Visual C++ 6). I developed an application using this library. This 
compiles fine but when I run the application I get

"The application failed to initialize properly (0xc0000022). Click OK to 
terminate the application."

After checking my application I tried the dom_xpath example program 
(using the supplied MSVC6 project) and I got exactly the same error message.

I'm using libxml++-1.0.1 (compiled using the supplied MSVC project), 
libxml2-2.6.5 (compiled using nmake and the makefile.vc in the 
distribution) and libiconv-1.7 (binary version included in the binary 
Sablotron distribution). The OS is WinXP with SP1. VC++ is version 6.0.

Has anyone used libxml++ on Windows XP before (I suppose yes because 
there is a workspace with all example projects in the win32_msvc6 
folder)? Is the error message mentioned before known to somebody? What 
does it mean?

Please note that I'm not very familiar with Windows development, my 
applications get developed under Linux but this one must be used under 
Win32 too, so I'm obliged to care about this.

Thanks to everyone in advance.

Philipp

[libxml++] My proposed patches

[Jim G. <gar...@ca...>]

The first patch adds cobj() to Document.  The second one adds the
function Document::create_root_node_by_import.  I decided to leave the
long name and not overload create_root_node for the sake of clarity; in
the same manner, Node::import_node and Node::add_child are two
differently named functions with analogous functionality.

If you give me the word, I'll go ahead and commit it to the module, or
feel free to do it on your own.

Thanks,
Jim

Re: [libxml++] Propose Document::create_root_node_by_import ()

[Jim G. <gar...@ca...>]

In trying to work around the problem, I also realized that there is no
xmlpp::Document::cobj() function.  While it is not entirely essential to
what I am doing, it would make things easier, and I think it is still
useful.  If nobody objects, I'll also make a patch for this function for
the 2.5 branch.

Jim

On Wed, 2004-02-04 at 05:10, Christophe de VIENNE wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> 
> 
> Jim Garrison wrote:
> 
> |
> | -----Forwarded Message----- From: Jim Garrison <gar...@ca...>
> | To: lib...@li... Subject: Propose
> | Document::create_root_node_by_import () Date: Wed, 28 Jan 2004
> | 22:21:16 -0500
> |
> | Hello,
> |
> | I would like to propose the addition of a function to
> | xmlpp::Document. Right now, one can only create a root node with
> | the given function create_root_node(), and then manipulate that
> | Node.  My proposed function (something like
> | Document::create_root_node_by_import) would take a Node from an
> | existing Document and import it in the spirit of
> | Node::import_node().  So basicly, this adds the import
> | functionality so that an entire document can be created from a
> | arbitrary Node in a different document.
> |
> | The biggest reason I need to do this is so that I can create a
> | string that holds all the XML within a certain node.  If there's a
> | better way to do this, please let me know.  If not, would this type
> | of patch be accepted?  If so, would it be able to go in the 1.0
> | branch?  I don't think it would break anything in terms of the ABI.
> |
> |
> 
> Hi,
> 
> The way you propose to implement the functionnality, you will have the
> xml header in the resulting string. If it's what you want then it's
> probably the best way to do it.
> A patch is welcome, and will probably go in 1.0 branch is ABI is not
> broken.
> 
> Cheers,
> 
> Christophe
> 
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.2.3 (GNU/Linux)
> Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
> 
> iD8DBQFAIMUpB+sU3TyOQjARAgSJAJ9SZOkKdCU2O5K1MjsepDJkv4f8lwCggllB
> 19JqHRtGO/OMvo2R/GBv4PM=
> =f3qP
> -----END PGP SIGNATURE-----
> 
> 
> 
> -------------------------------------------------------
> The SF.Net email is sponsored by EclipseCon 2004
> Premiere Conference on Open Tools Development and Integration
> See the breadth of Eclipse activity. February 3-5 in Anaheim, CA.
> http://www.eclipsecon.org/osdn
> _______________________________________________
> Libxmlplusplus-general mailing list
> Lib...@li...
> https://lists.sourceforge.net/lists/listinfo/libxmlplusplus-general

Re: [libxml++] Propose Document::create_root_node_by_import ()

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Christophe de VIENNE wrote:

| and will probably go in 1.0 branch is ABI is not broken.


I've spoke with Murray about this, and in fact it will probably not go
in the stable API.
Anyway the next stable API will come quite soon, so it shouldn't be a
problem.

Cheers,

Christophe


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAIOp9B+sU3TyOQjARAvX/AJoCSyXSCTq64/l1EqRhR2QwqccewACgllU3
9e5gi3wZr0g+mmyeRLktDoU=
=sFu1
-----END PGP SIGNATURE-----

Re: [libxml++] Propose Document::create_root_node_by_import ()

[Christophe de V. <cde...@al...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1



Jim Garrison wrote:

|
| -----Forwarded Message----- From: Jim Garrison <gar...@ca...>
| To: lib...@li... Subject: Propose
| Document::create_root_node_by_import () Date: Wed, 28 Jan 2004
| 22:21:16 -0500
|
| Hello,
|
| I would like to propose the addition of a function to
| xmlpp::Document. Right now, one can only create a root node with
| the given function create_root_node(), and then manipulate that
| Node.  My proposed function (something like
| Document::create_root_node_by_import) would take a Node from an
| existing Document and import it in the spirit of
| Node::import_node().  So basicly, this adds the import
| functionality so that an entire document can be created from a
| arbitrary Node in a different document.
|
| The biggest reason I need to do this is so that I can create a
| string that holds all the XML within a certain node.  If there's a
| better way to do this, please let me know.  If not, would this type
| of patch be accepted?  If so, would it be able to go in the 1.0
| branch?  I don't think it would break anything in terms of the ABI.
|
|

Hi,

The way you propose to implement the functionnality, you will have the
xml header in the resulting string. If it's what you want then it's
probably the best way to do it.
A patch is welcome, and will probably go in 1.0 branch is ABI is not
broken.

Cheers,

Christophe

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAIMUpB+sU3TyOQjARAgSJAJ9SZOkKdCU2O5K1MjsepDJkv4f8lwCggllB
19JqHRtGO/OMvo2R/GBv4PM=
=f3qP
-----END PGP SIGNATURE-----

[libxml++] Propose Document::create_root_node_by_import ()

[Jim G. <gar...@ca...>]

I sent this a few days ago, but it has not appeared in the archive yet,
so I am sending it again.  My apologies if you have already received
this.

-----Forwarded Message-----
From: Jim Garrison <gar...@ca...>
To: lib...@li...
Subject: Propose Document::create_root_node_by_import ()
Date: Wed, 28 Jan 2004 22:21:16 -0500

Hello,

I would like to propose the addition of a function to xmlpp::Document. 
Right now, one can only create a root node with the given function
create_root_node(), and then manipulate that Node.  My proposed function
(something like Document::create_root_node_by_import) would take a Node
from an existing Document and import it in the spirit of
Node::import_node().  So basicly, this adds the import functionality so
that an entire document can be created from a arbitrary Node in a
different document.

The biggest reason I need to do this is so that I can create a string
that holds all the XML within a certain node.  If there's a better way
to do this, please let me know.  If not, would this type of patch be
accepted?  If so, would it be able to go in the 1.0 branch?  I don't
think it would break anything in terms of the ABI.

Please let me know.

Jim Garrison
http://gnomoradio.org/

Re: [libxml++] solaris 9 sparc problems

[Christophe de V. <cde...@al...>]

Hi,

I think those problems occurs when Document::Init is not instanciated. 
It should be automatic but depending on the platform/linker/compiler it 
may not work properly.
Try to declare a global var of type Document::Init.

BTW, I was wondering if making a singleton from Document::Init could 
eventually solve this problem. We could make sure it is instanciated in 
Parser and Document constructor.

Cheers,

Christophe

kike fernandez a écrit :

>Hello,
>
>I'm trying to use libxml++ in one project but I'm having some problems.
>Our environment is sparc/Solaris 9. Is libxml++ known to work in this
>environment?
>
>We use gcc2.95.3 and the library compiles and installs OK.
>but some of the example programs segfault and others work (using the
>same versions in Linux seems to work ok and libxml2 and libxslt work ok
>in our solaris box)
>
>We are very interested in using libxml++ (we would like to help to find
>and resolve any bugs in libxml++) but I'm relatively new to c++
>programming so I'm sorry if I'm doing something wrong. 
>
>
>
>---> dom_build
> 
>40          xmlpp::Element* nodeRoot =
>document.create_root_node("exampleroot", "http://foo", "foo");
>//Declares the namespace and uses its prefix for this node
>(gdb) n
>  
>Program received signal SIGSEGV, Segmentation fault.
>0xff378f18 in xmlpp::Node::cobj (this=0x0) at node.cc:131
> 
> 
>---> dom_parse_entities
> 
>54          xmlpp::Node::NodeList list = node->get_children();
>(gdb) n
>  
>Program received signal SIGSEGV, Segmentation fault.
>xmlpp::Node::get_children (this=0x0, name=@0xffbff5d0) at node.cc:28
>28         xmlNode* child = impl_->children;
> 
>---> dom_parser
> 
>115       return impl_->name ? (const char*)impl_->name : "";
>(gdb) n
>  
>Program terminated with signal SIGSEGV, Segmentation fault.
>The program no longer exists.
> 
>
>
>
>  
>

[libxml++] solaris 9 sparc problems

[kike f. <sis...@si...>]

Hello,

I'm trying to use libxml++ in one project but I'm having some problems.
Our environment is sparc/Solaris 9. Is libxml++ known to work in this
environment?

We use gcc2.95.3 and the library compiles and installs OK.
but some of the example programs segfault and others work (using the
same versions in Linux seems to work ok and libxml2 and libxslt work ok
in our solaris box)

We are very interested in using libxml++ (we would like to help to find
and resolve any bugs in libxml++) but I'm relatively new to c++
programming so I'm sorry if I'm doing something wrong. 



---> dom_build
 
40          xmlpp::Element* nodeRoot =
document.create_root_node("exampleroot", "http://foo", "foo");
//Declares the namespace and uses its prefix for this node
(gdb) n
  
Program received signal SIGSEGV, Segmentation fault.
0xff378f18 in xmlpp::Node::cobj (this=0x0) at node.cc:131
 
 
---> dom_parse_entities
 
54          xmlpp::Node::NodeList list = node->get_children();
(gdb) n
  
Program received signal SIGSEGV, Segmentation fault.
xmlpp::Node::get_children (this=0x0, name=@0xffbff5d0) at node.cc:28
28         xmlNode* child = impl_->children;
 
---> dom_parser
 
115       return impl_->name ? (const char*)impl_->name : "";
(gdb) n
  
Program terminated with signal SIGSEGV, Segmentation fault.
The program no longer exists.
 



-- 
kike fernandez <sis...@si...>
IXA

[libxml++] ANNOUNCE: libxml++ 1.0.1 (stable)

[Christophe de V. <cde...@al...>]

Hi,


This is a bugfix release.


*** libxml++

libxml++ is a C++ wrapper for the libxml XML parser library. It has SAX
and DOM-like APIs, but does not attempt to conform exactly to the DOM
specifications because they are not aimed at C++. Its API is much
simpler than the underlying libxml C API.


*** Homepage

http://libxmlplusplus.sourceforge.net/


*** Notes

This is a bugfix release.


*** Changes

* Fixed Bug #131329 [http://bugzilla.gnome.org/show_bug.cgi?id=131329] 
on Document::write_to_stream() raised by Guillaume Arreckx.
* Fixed Bug #131018 [http://bugzilla.gnome.org/show_bug.cgi?id=131018] 
on Document::set_internal_subset raised by pa...@la....

*** Download

You can download libxml++ 1.0.1 from here :

http://ftp.gnome.org/pub/GNOME/sources/libxml++/1.0/



Best Regards,

Christophe de Vienne

Re: [libxml++] Re: Formatting problem

[Daniel V. <vei...@re...>]

On Tue, Jan 13, 2004 at 01:02:11PM +0100, Christophe de VIENNE wrote:
> Removing formatting linebreaks and spaces at parsing time imply a 
> decision on wether or not a whitespace is significant (See 
> http://www.xmlsoft.org/html/libxml-parser.html#xmlKeepBlanksDefault for 
> more information about it).

  See function areBlanks() in parser.c in libxml2 sources. It's not
public but you will see the criteria used by libxml2 to select whether
a whitespace node is significant.

Daniel

-- 
Daniel Veillard      | Red Hat Network https://rhn.redhat.com/
vei...@re...  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/

[libxml++] Re: Formatting problem

[Christophe de V. <cde...@ne...>]

Bastian_Pfennigschmidt/CODESCO/DE...@co... a =E9crit :

>
> Hello Christophe,
>

Hi Bastian,

<snip>

> Why the parser doesn't remove the formatting linebreaks and spaces, if=20
> I parse it again?
> If I call write_to_xxx I don't want to get a formatted output, even if=20
> the input was formatted.
> Is there any way to get a unformatted output if the input was formatted=
?
>

Removing formatting linebreaks and spaces at parsing time imply a=20
decision on wether or not a whitespace is significant (See=20
http://www.xmlsoft.org/html/libxml-parser.html#xmlKeepBlanksDefault for=20
more information about it).

However if you are able to decide if a whitespace is significant or not,=20
then you can walk your dom tree and remove them. This shouldn't be hard=20
to implement : you just have to remove all content nodes that are white=20
space (see bool ContentNode::is_white_space=20
<http://libxmlplusplus.sourceforge.net/reference/html/classxmlpp_1_1Conte=
ntNode.html#a4>=20
() const).

There is another possibility, which is to change the behavior of=20
libxml++ regarding whitespaces by changing KeepBlanks::Default to false.=20
But this solution will be heavy to maintain for you.
Eventually we can think about a little modification in the next API to=20
allow a dynamic change of this setting, but we'll have to convince=20
Murray :-).

>
> Thanks for helping


You're welcome,

Christophe

Re: [libxml++] IMPORTANT: libxml++ CVS Repositery is now on GNOME

[Christophe de V. <cde...@al...>]

Ephraim Vider wrote:

>  
>
>>Christophe de VIENNE wrote:
>>
>>    
>>
>>>Hi all,
>>>
>>>I've just imported libxml++ to the gnome CVS. From now it is the
>>>official repositery of libxml++. I've also updated the homepage to
>>>reflect this.
>>>
>>>      
>>>
>>Ephraim, do you have a gnome cvs access ?
>>
>>Christophe
>>
>>
>>    
>>
>
>Nope. How do I get one?
>
>  
>

http://developer.gnome.org/doc/policies/accounts/index.html

You need a cvs account. 


>BTW: the ViewCVS web interface is _so_ much easier than the gnome bonsai
>interface...
>
>  
>

yep :-\


Christophe