xml.html@ 67456

最後變更在這個檔案從67456是 65950,由 vboxsync 提交於 8 年前
libxml 2.9.4: fix export
屬性 svn:eol-style 設為 `native`
檔案大小: 291.6 KB

行
1	<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
2	"http://www.w3.org/TR/html4/loose.dtd">
3	<html>
4	<head>
5	<title>The XML C parser and toolkit of Gnome</title>
6	<meta http-equiv="Content-Type" content="text/html; charset=UTF-8"/>
7	</head>
8	<body bgcolor="#ffffff">
9	<h1 align="center">The XML C parser and toolkit of Gnome</h1>
10
11	<h1>Note: this is the flat content of the <a href="index.html">web
12	site</a></h1>
13
14	<h1 style="text-align: center">libxml, a.k.a. gnome-xml</h1>
15
16	<p></p>
17
18	<p
19	style="text-align: right; font-style: italic; font-size: 10pt">"Programming
20	with libxml2 is like the thrilling embrace of an exotic stranger." <a
21	href="http://diveintomark.org/archives/2004/02/18/libxml2">Mark
22	Pilgrim</a></p>
23
24	<p>Libxml2 is the XML C parser and toolkit developed for the Gnome project
25	(but usable outside of the Gnome platform), it is free software available
26	under the <a href="http://www.opensource.org/licenses/mit-license.html">MIT
27	License</a>. XML itself is a metalanguage to design markup languages, i.e.
28	text language where semantic and structure are added to the content using
29	extra "markup" information enclosed between angle brackets. HTML is the most
30	well-known markup language. Though the library is written in C <a
31	href="python.html">a variety of language bindings</a> make it available in
32	other environments.</p>
33
34	<p>Libxml2 is known to be very portable, the library should build and work
35	without serious troubles on a variety of systems (Linux, Unix, Windows,
36	CygWin, MacOS, MacOS X, RISC Os, OS/2, VMS, QNX, MVS, VxWorks, ...)</p>
37
38	<p>Libxml2 implements a number of existing standards related to markup
39	languages:</p>
40	<ul>
41	<li>the XML standard: <a
42	href="http://www.w3.org/TR/REC-xml">http://www.w3.org/TR/REC-xml</a></li>
43	<li>Namespaces in XML: <a
44	href="http://www.w3.org/TR/REC-xml-names/">http://www.w3.org/TR/REC-xml-names/</a></li>
45	<li>XML Base: <a
46	href="http://www.w3.org/TR/xmlbase/">http://www.w3.org/TR/xmlbase/</a></li>
47	<li><a href="http://www.cis.ohio-state.edu/rfc/rfc2396.txt">RFC 2396</a> :
48	Uniform Resource Identifiers <a
49	href="http://www.ietf.org/rfc/rfc2396.txt">http://www.ietf.org/rfc/rfc2396.txt</a></li>
50	<li>XML Path Language (XPath) 1.0: <a
51	href="http://www.w3.org/TR/xpath">http://www.w3.org/TR/xpath</a></li>
52	<li>HTML4 parser: <a
53	href="http://www.w3.org/TR/html401/">http://www.w3.org/TR/html401/</a></li>
54	<li>XML Pointer Language (XPointer) Version 1.0: <a
55	href="http://www.w3.org/TR/xptr">http://www.w3.org/TR/xptr</a></li>
56	<li>XML Inclusions (XInclude) Version 1.0: <a
57	href="http://www.w3.org/TR/xinclude/">http://www.w3.org/TR/xinclude/</a></li>
58	<li>ISO-8859-x encodings, as well as <a
59	href="http://www.cis.ohio-state.edu/rfc/rfc2044.txt">rfc2044</a> [UTF-8]
60	and <a href="http://www.cis.ohio-state.edu/rfc/rfc2781.txt">rfc2781</a>
61	[UTF-16] Unicode encodings, and more if using iconv support</li>
62	<li>part of SGML Open Technical Resolution TR9401:1997</li>
63	<li>XML Catalogs Working Draft 06 August 2001: <a
64	href="http://www.oasis-open.org/committees/entity/spec-2001-08-06.html">http://www.oasis-open.org/committees/entity/spec-2001-08-06.html</a></li>
65	<li>Canonical XML Version 1.0: <a
66	href="http://www.w3.org/TR/xml-c14n">http://www.w3.org/TR/xml-c14n</a>
67	and the Exclusive XML Canonicalization CR draft <a
68	href="http://www.w3.org/TR/xml-exc-c14n">http://www.w3.org/TR/xml-exc-c14n</a></li>
69	<li>Relax NG, ISO/IEC 19757-2:2003, <a
70	href="http://www.oasis-open.org/committees/relax-ng/spec-20011203.html">http://www.oasis-open.org/committees/relax-ng/spec-20011203.html</a></li>
71	<li>W3C XML Schemas Part 2: Datatypes <a
72	href="http://www.w3.org/TR/2001/REC-xmlschema-2-20010502/">REC 02 May
73	2001</a></li>
74	<li>W3C <a href="http://www.w3.org/TR/xml-id/">xml:id</a> Working Draft 7
75	April 2004</li>
76	</ul>
77
78	<p>In most cases libxml2 tries to implement the specifications in a
79	relatively strictly compliant way. As of release 2.4.16, libxml2 passed all
80	1800+ tests from the <a
81	href="http://www.oasis-open.org/committees/xml-conformance/">OASIS XML Tests
82	Suite</a>.</p>
83
84	<p>To some extent libxml2 provides support for the following additional
85	specifications but doesn't claim to implement them completely:</p>
86	<ul>
87	<li>Document Object Model (DOM) <a
88	href="http://www.w3.org/TR/DOM-Level-2-Core/">http://www.w3.org/TR/DOM-Level-2-Core/</a>
89	the document model, but it doesn't implement the API itself, gdome2 does
90	this on top of libxml2</li>
91	<li><a href="http://www.cis.ohio-state.edu/rfc/rfc959.txt">RFC 959</a> :
92	libxml2 implements a basic FTP client code</li>
93	<li><a href="http://www.cis.ohio-state.edu/rfc/rfc1945.txt">RFC 1945</a> :
94	HTTP/1.0, again a basic HTTP client code</li>
95	<li>SAX: a SAX2 like interface and a minimal SAX1 implementation compatible
96	with early expat versions</li>
97	</ul>
98
99	<p>A partial implementation of <a
100	href="http://www.w3.org/TR/2001/REC-xmlschema-1-20010502/">XML Schemas Part
101	1: Structure</a> is being worked on but it would be far too early to make any
102	conformance statement about it at the moment.</p>
103
104	<p>Separate documents:</p>
105	<ul>
106	<li><a href="http://xmlsoft.org/XSLT/">the libxslt page</a> providing an
107	implementation of XSLT 1.0 and common extensions like EXSLT for
108	libxml2</li>
109	<li><a href="http://gdome2.cs.unibo.it/">the gdome2 page</a>
110	: a standard DOM2 implementation for libxml2</li>
111	<li><a href="http://www.aleksey.com/xmlsec/">the XMLSec page</a>: an
112	implementation of <a href="http://www.w3.org/TR/xmldsig-core/">W3C XML
113	Digital Signature</a> for libxml2</li>
114	<li>also check the related links section for more related and active
115	projects.</li>
116	</ul>
117	<p> Hosting sponsored by <a href="http://www.aoemedia.de/opensource-cms.html"
118	>Open Source CMS services</a> from AOE media.</p>
119
120	<p>Logo designed by <a href="mailto:[email protected]">Marc Liyanage</a>.</p>
121
122	<h2><a name="Introducti">Introduction</a></h2>
123
124	<p>This document describes libxml, the <a
125	href="http://www.w3.org/XML/">XML</a> C parser and toolkit developed for the
126	<a href="http://www.gnome.org/">Gnome</a> project. <a
127	href="http://www.w3.org/XML/">XML is a standard</a> for building tag-based
128	structured documents/data.</p>
129
130	<p>Here are some key points about libxml:</p>
131	<ul>
132	<li>Libxml2 exports Push (progressive) and Pull (blocking) type parser
133	interfaces for both XML and HTML.</li>
134	<li>Libxml2 can do DTD validation at parse time, using a parsed document
135	instance, or with an arbitrary DTD.</li>
136	<li>Libxml2 includes complete <a
137	href="http://www.w3.org/TR/xpath">XPath</a>, <a
138	href="http://www.w3.org/TR/xptr">XPointer</a> and <a
139	href="http://www.w3.org/TR/xinclude">XInclude</a> implementations.</li>
140	<li>It is written in plain C, making as few assumptions as possible, and
141	sticking closely to ANSI C/POSIX for easy embedding. Works on
142	Linux/Unix/Windows, ported to a number of other platforms.</li>
143	<li>Basic support for HTTP and FTP client allowing applications to fetch
144	remote resources.</li>
145	<li>The design is modular, most of the extensions can be compiled out.</li>
146	<li>The internal document representation is as close as possible to the <a
147	href="http://www.w3.org/DOM/">DOM</a> interfaces.</li>
148	<li>Libxml2 also has a <a
149	href="http://www.megginson.com/SAX/index.html">SAX like interface</a>;
150	the interface is designed to be compatible with <a
151	href="http://www.jclark.com/xml/expat.html">Expat</a>.</li>
152	<li>This library is released under the <a
153	href="http://www.opensource.org/licenses/mit-license.html">MIT
154	License</a>. See the Copyright file in the distribution for the precise
155	wording.</li>
156	</ul>
157
158	<p>Warning: unless you are forced to because your application links with a
159	Gnome-1.X library requiring it, <strong><span
160	style="background-color: #FF0000">Do Not Use libxml1</span></strong>, use
161	libxml2</p>
162
163	<h2><a name="FAQ">FAQ</a></h2>
164
165	<p>Table of Contents:</p>
166	<ul>
167	<li><a href="FAQ.html#License">License(s)</a></li>
168	<li><a href="FAQ.html#Installati">Installation</a></li>
169	<li><a href="FAQ.html#Compilatio">Compilation</a></li>
170	<li><a href="FAQ.html#Developer">Developer corner</a></li>
171	</ul>
172
173	<h3><a name="License">License</a>(s)</h3>
174	<ol>
175	<li><em>Licensing Terms for libxml</em>
176	<p>libxml2 is released under the <a
177	href="http://www.opensource.org/licenses/mit-license.html">MIT
178	License</a>; see the file Copyright in the distribution for the precise
179	wording</p>
180	</li>
181	<li><em>Can I embed libxml2 in a proprietary application ?</em>
182	<p>Yes. The MIT License allows you to keep proprietary the changes you
183	made to libxml, but it would be graceful to send-back bug fixes and
184	improvements as patches for possible incorporation in the main
185	development tree.</p>
186	</li>
187	</ol>
188
189	<h3><a name="Installati">Installation</a></h3>
190	<ol>
191	<li><strong><span style="background-color: #FF0000">Do Not Use
192	libxml1</span></strong>, use libxml2</li>
193	<p></p>
194	<li><em>Where can I get libxml</em> ?
195	<p>The original distribution comes from <a
196	href="ftp://xmlsoft.org/libxml2/">xmlsoft.org</a> or <a
197	href="ftp://ftp.gnome.org/pub/GNOME/sources/libxml2/2.6/">gnome.org</a></p>
198	<p>Most Linux and BSD distributions include libxml, this is probably the
199	safer way for end-users to use libxml.</p>
200	<p>David Doolin provides precompiled Windows versions at <a
201	href="http://www.ce.berkeley.edu/~doolin/code/libxmlwin32/ ">http://www.ce.berkeley.edu/~doolin/code/libxmlwin32/</a></p>
202	</li>
203	<p></p>
204	<li><em>I see libxml and libxml2 releases, which one should I install ?</em>
205	<ul>
206	<li>If you are not constrained by backward compatibility issues with
207	existing applications, install libxml2 only</li>
208	<li>If you are not doing development, you can safely install both.
209	Usually the packages <a
210	href="http://rpmfind.net/linux/RPM/libxml.html">libxml</a> and <a
211	href="http://rpmfind.net/linux/RPM/libxml2.html">libxml2</a> are
212	compatible (this is not the case for development packages).</li>
213	<li>If you are a developer and your system provides separate packaging
214	for shared libraries and the development components, it is possible
215	to install libxml and libxml2, and also <a
216	href="http://rpmfind.net/linux/RPM/libxml-devel.html">libxml-devel</a>
217	and <a
218	href="http://rpmfind.net/linux/RPM/libxml2-devel.html">libxml2-devel</a>
219	too for libxml2 >= 2.3.0</li>
220	<li>If you are developing a new application, please develop against
221	libxml2(-devel)</li>
222	</ul>
223	</li>
224	<li><em>I can't install the libxml package, it conflicts with libxml0</em>
225	<p>You probably have an old libxml0 package used to provide the shared
226	library for libxml.so.0, you can probably safely remove it. The libxml
227	packages provided on <a
228	href="ftp://xmlsoft.org/libxml2/">xmlsoft.org</a> provide
229	libxml.so.0</p>
230	</li>
231	<li><em>I can't install the libxml(2) RPM package due to failed
232	dependencies</em>
233	<p>The most generic solution is to re-fetch the latest src.rpm , and
234	rebuild it locally with</p>
235	<p><code>rpm --rebuild libxml(2)-xxx.src.rpm</code>.</p>
236	<p>If everything goes well it will generate two binary rpm packages (one
237	providing the shared libs and xmllint, and the other one, the -devel
238	package, providing includes, static libraries and scripts needed to build
239	applications with libxml(2)) that you can install locally.</p>
240	</li>
241	</ol>
242
243	<h3><a name="Compilatio">Compilation</a></h3>
244	<ol>
245	<li><em>What is the process to compile libxml2 ?</em>
246	<p>As most UNIX libraries libxml2 follows the "standard":</p>
247	<p><code>gunzip -c xxx.tar.gz \| tar xvf -</code></p>
248	<p><code>cd libxml-xxxx</code></p>
249	<p><code>./configure --help</code></p>
250	<p>to see the options, then the compilation/installation proper</p>
251	<p><code>./configure [possible options]</code></p>
252	<p><code>make</code></p>
253	<p><code>make install</code></p>
254	<p>At that point you may have to rerun ldconfig or a similar utility to
255	update your list of installed shared libs.</p>
256	</li>
257	<li><em>What other libraries are needed to compile/install libxml2 ?</em>
258	<p>Libxml2 does not require any other library, the normal C ANSI API
259	should be sufficient (please report any violation to this rule you may
260	find).</p>
261	<p>However if found at configuration time libxml2 will detect and use the
262	following libs:</p>
263	<ul>
264	<li><a href="http://www.info-zip.org/pub/infozip/zlib/">libz</a> : a
265	highly portable and available widely compression library.</li>
266	<li>iconv: a powerful character encoding conversion library. It is
267	included by default in recent glibc libraries, so it doesn't need to
268	be installed specifically on Linux. It now seems a <a
269	href="http://www.opennc.org/onlinepubs/7908799/xsh/iconv.html">part
270	of the official UNIX</a> specification. Here is one <a
271	href="http://www.gnu.org/software/libiconv/">implementation of the
272	library</a> which source can be found <a
273	href="ftp://ftp.ilog.fr/pub/Users/haible/gnu/">here</a>.</li>
274	</ul>
275	</li>
276	<p></p>
277	<li><em>Make check fails on some platforms</em>
278	<p>Sometimes the regression tests' results don't completely match the
279	value produced by the parser, and the makefile uses diff to print the
280	delta. On some platforms the diff return breaks the compilation process;
281	if the diff is small this is probably not a serious problem.</p>
282	<p>Sometimes (especially on Solaris) make checks fail due to limitations
283	in make. Try using GNU-make instead.</p>
284	</li>
285	<li><em>I use the SVN version and there is no configure script</em>
286	<p>The configure script (and other Makefiles) are generated. Use the
287	autogen.sh script to regenerate the configure script and Makefiles,
288	like:</p>
289	<p><code>./autogen.sh --prefix=/usr --disable-shared</code></p>
290	</li>
291	<li><em>I have troubles when running make tests with gcc-3.0</em>
292	<p>It seems the initial release of gcc-3.0 has a problem with the
293	optimizer which miscompiles the URI module. Please use another
294	compiler.</p>
295	</li>
296	</ol>
297
298	<h3><a name="Developer">Developer</a> corner</h3>
299	<ol>
300	<li><em>Troubles compiling or linking programs using libxml2</em>
301	<p>Usually the problem comes from the fact that the compiler doesn't get
302	the right compilation or linking flags. There is a small shell script
303	<code>xml2-config</code> which is installed as part of libxml2 usual
304	install process which provides those flags. Use</p>
305	<p><code>xml2-config --cflags</code></p>
306	<p>to get the compilation flags and</p>
307	<p><code>xml2-config --libs</code></p>
308	<p>to get the linker flags. Usually this is done directly from the
309	Makefile as:</p>
310	<p><code>CFLAGS=`xml2-config --cflags`</code></p>
311	<p><code>LIBS=`xml2-config --libs`</code></p>
312	</li>
313	<li><em>I want to install my own copy of libxml2 in my home directory and
314	link my programs against it, but it doesn't work</em>
315	<p>There are many different ways to accomplish this. Here is one way to
316	do this under Linux. Suppose your home directory is <code>/home/user.
317	</code>Then:</p>
318	<ul>
319	<li>Create a subdirectory, let's call it <code>myxml</code></li>
320	<li>unpack the libxml2 distribution into that subdirectory</li>
321	<li>chdir into the unpacked distribution
322	(<code>/home/user/myxml/libxml2 </code>)</li>
323	<li>configure the library using the "<code>--prefix</code>" switch,
324	specifying an installation subdirectory in
325	<code>/home/user/myxml</code>, e.g.
326	<p><code>./configure --prefix /home/user/myxml/xmlinst</code> {other
327	configuration options}</p>
328	</li>
329	<li>now run <code>make</code> followed by <code>make install</code></li>
330	<li>At this point, the installation subdirectory contains the complete
331	"private" include files, library files and binary program files (e.g.
332	xmllint), located in
333	<p><code>/home/user/myxml/xmlinst/lib,
334	/home/user/myxml/xmlinst/include </code> and <code>
335	/home/user/myxml/xmlinst/bin</code></p>
336	respectively.</li>
337	<li>In order to use this "private" library, you should first add it to
338	the beginning of your default PATH (so that your own private program
339	files such as xmllint will be used instead of the normal system
340	ones). To do this, the Bash command would be
341	<p><code>export PATH=/home/user/myxml/xmlinst/bin:$PATH</code></p>
342	</li>
343	<li>Now suppose you have a program <code>test1.c</code> that you would
344	like to compile with your "private" library. Simply compile it using
345	the command
346	<p><code>gcc `xml2-config --cflags --libs` -o test test.c</code></p>
347	Note that, because your PATH has been set with <code>
348	/home/user/myxml/xmlinst/bin</code> at the beginning, the xml2-config
349	program which you just installed will be used instead of the system
350	default one, and this will <em>automatically</em> get the correct
351	libraries linked with your program.</li>
352	</ul>
353	</li>
354
355	<p></p>
356	<li><em>xmlDocDump() generates output on one line.</em>
357	<p>Libxml2 will not <strong>invent</strong> spaces in the content of a
358	document since <strong>all spaces in the content of a document are
359	significant</strong>. If you build a tree from the API and want
360	indentation:</p>
361	<ol>
362	<li>the correct way is to generate those yourself too.</li>
363	<li>the dangerous way is to ask libxml2 to add those blanks to your
364	content <strong>modifying the content of your document in the
365	process</strong>. The result may not be what you expect. There is
366	<strong>NO</strong> way to guarantee that such a modification won't
367	affect other parts of the content of your document. See <a
368	href="http://xmlsoft.org/html/libxml-parser.html#xmlKeepBlanksDefault">xmlKeepBlanksDefault
369	()</a> and <a
370	href="http://xmlsoft.org/html/libxml-tree.html#xmlSaveFormatFile">xmlSaveFormatFile
371	()</a></li>
372	</ol>
373	</li>
374	<p></p>
375	<li><em>Extra nodes in the document:</em>
376	<p><em>For an XML file as below:</em></p>
377	<pre><?xml version="1.0"?>
378	<PLAN xmlns="http://www.argus.ca/autotest/1.0/">
379	<NODE CommFlag="0"/>
380	<NODE CommFlag="1"/>
381	</PLAN></pre>
382	<p><em>after parsing it with the function
383	pxmlDoc=xmlParseFile(...);</em></p>
384	<p><em>I want to the get the content of the first node (node with the
385	CommFlag="0")</em></p>
386	<p><em>so I did it as following;</em></p>
387	<pre>xmlNodePtr pnode;
388	pnode=pxmlDoc->children->children;</pre>
389	<p><em>but it does not work. If I change it to</em></p>
390	<pre>pnode=pxmlDoc->children->children->next;</pre>
391	<p><em>then it works. Can someone explain it to me.</em></p>
392	<p></p>
393	<p>In XML all characters in the content of the document are significant
394	<strong>including blanks and formatting line breaks</strong>.</p>
395	<p>The extra nodes you are wondering about are just that, text nodes with
396	the formatting spaces which are part of the document but that people tend
397	to forget. There is a function <a
398	href="http://xmlsoft.org/html/libxml-parser.html">xmlKeepBlanksDefault
399	()</a> to remove those at parse time, but that's an heuristic, and its
400	use should be limited to cases where you are certain there is no
401	mixed-content in the document.</p>
402	</li>
403	<li><em>I get compilation errors of existing code like when accessing
404	<strong>root</strong> or <strong>child fields</strong> of nodes.</em>
405	<p>You are compiling code developed for libxml version 1 and using a
406	libxml2 development environment. Either switch back to libxml v1 devel or
407	even better fix the code to compile with libxml2 (or both) by <a
408	href="upgrade.html">following the instructions</a>.</p>
409	</li>
410	<li><em>I get compilation errors about non existing
411	<strong>xmlRootNode</strong> or <strong>xmlChildrenNode</strong>
412	fields.</em>
413	<p>The source code you are using has been <a
414	href="upgrade.html">upgraded</a> to be able to compile with both libxml
415	and libxml2, but you need to install a more recent version:
416	libxml(-devel) >= 1.8.8 or libxml2(-devel) >= 2.1.0</p>
417	</li>
418	<li><em>Random crashes in threaded applications</em>
419	<p>Read and follow all advices on the <a href="threads.html">thread
420	safety</a> page, and make 100% sure you never call xmlCleanupParser()
421	while the library or an XML document might still be in use by another
422	thread.</p>
423	</li>
424	<li><em>The example provided in the web page does not compile.</em>
425	<p>It's hard to maintain the documentation in sync with the code
426	<grin/> ...</p>
427	<p>Check the previous points 1/ and 2/ raised before, and please send
428	patches.</p>
429	</li>
430	<li><em>Where can I get more examples and information than provided on the
431	web page?</em>
432	<p>Ideally a libxml2 book would be nice. I have no such plan ... But you
433	can:</p>
434	<ul>
435	<li>check more deeply the <a href="html/libxml-lib.html">existing
436	generated doc</a></li>
437	<li>have a look at <a href="examples/index.html">the set of
438	examples</a>.</li>
439	<li>look for examples of use for libxml2 function using the Gnome code
440	or by asking on Google.</li>
441	<li><a
442	href="http://svn.gnome.org/viewvc/libxml2/trunk/">Browse
443	the libxml2 source</a> , I try to write code as clean and documented
444	as possible, so looking at it may be helpful. In particular the code
445	of <a href="http://svn.gnome.org/viewvc/libxml2/trunk/xmllint.c?view=markup">xmllint.c</a> and of the various testXXX.c test programs should
446	provide good examples of how to do things with the library.</li>
447	</ul>
448	</li>
449	<p></p>
450	<li><em>What about C++ ?</em>
451	<p>libxml2 is written in pure C in order to allow easy reuse on a number
452	of platforms, including embedded systems. I don't intend to convert to
453	C++.</p>
454	<p>There is however a C++ wrapper which may fulfill your needs:</p>
455	<ul>
456	<li>by Ari Johnson <[email protected]>:
457	<p>Website: <a
458	href="http://libxmlplusplus.sourceforge.net/">http://libxmlplusplus.sourceforge.net/</a></p>
459	<p>Download: <a
460	href="http://sourceforge.net/project/showfiles.php?group_id=12999">http://sourceforge.net/project/showfiles.php?group_id=12999</a></p>
461	</li>
462	</ul>
463	</li>
464	<li><em>How to validate a document a posteriori ?</em>
465	<p>It is possible to validate documents which had not been validated at
466	initial parsing time or documents which have been built from scratch
467	using the API. Use the <a
468	href="http://xmlsoft.org/html/libxml-valid.html#xmlValidateDtd">xmlValidateDtd()</a>
469	function. It is also possible to simply add a DTD to an existing
470	document:</p>
471	<pre>xmlDocPtr doc; /* your existing document */
472	xmlDtdPtr dtd = xmlParseDTD(NULL, filename_of_dtd); /* parse the DTD */
473
474	dtd->name = xmlStrDup((xmlChar)"root_name"); / use the given root */
475
476	doc->intSubset = dtd;
477	if (doc->children == NULL) xmlAddChild((xmlNodePtr)doc, (xmlNodePtr)dtd);
478	else xmlAddPrevSibling(doc->children, (xmlNodePtr)dtd);
479	</pre>
480	</li>
481	<li><em>So what is this funky "xmlChar" used all the time?</em>
482	<p>It is a null terminated sequence of utf-8 characters. And only utf-8!
483	You need to convert strings encoded in different ways to utf-8 before
484	passing them to the API. This can be accomplished with the iconv library
485	for instance.</p>
486	</li>
487	<li>etc ...</li>
488	</ol>
489
490	<p></p>
491
492	<h2><a name="Documentat">Developer Menu</a></h2>
493
494	<p>There are several on-line resources related to using libxml:</p>
495	<ol>
496	<li>Use the <a href="search.php">search engine</a> to look up
497	information.</li>
498	<li>Check the <a href="FAQ.html">FAQ.</a></li>
499	<li>Check the <a href="http://xmlsoft.org/html/libxml-lib.html">extensive
500	documentation</a> automatically extracted from code comments.</li>
501	<li>Look at the documentation about <a href="encoding.html">libxml
502	internationalization support</a>.</li>
503	<li>This page provides a global overview and <a href="example.html">some
504	examples</a> on how to use libxml.</li>
505	<li><a href="examples/index.html">Code examples</a></li>
506	<li>John Fleck's libxml2 tutorial: <a href="tutorial/index.html">html</a>
507	or <a href="tutorial/xmltutorial.pdf">pdf</a>.</li>
508	<li>If you need to parse large files, check the <a
509	href="xmlreader.html">xmlReader</a> API tutorial</li>
510	<li><a href="mailto:[email protected]">James Henstridge</a> wrote <a
511	href="http://www.jamesh.id.au/articles/libxml-sax/libxml-sax.html">some nice
512	documentation</a> explaining how to use the libxml SAX interface.</li>
513	<li>George Lebl wrote <a
514	href="http://www-106.ibm.com/developerworks/library/l-gnome3/">an article
515	for IBM developerWorks</a> about using libxml.</li>
516	<li>Check <a href="http://svn.gnome.org/viewvc/libxml2/trunk/TODO?view=markup">the TODO
517	file</a>.</li>
518	<li>Read the <a href="upgrade.html">1.x to 2.x upgrade path</a>
519	description. If you are starting a new project using libxml you should
520	really use the 2.x version.</li>
521	<li>And don't forget to look at the <a
522	href="http://mail.gnome.org/archives/xml/">mailing-list archive</a>.</li>
523	</ol>
524
525	<h2><a name="Reporting">Reporting bugs and getting help</a></h2>
526
527	<p>Well, bugs or missing features are always possible, and I will make a
528	point of fixing them in a timely fashion. The best way to report a bug is to
529	use the <a href="http://bugzilla.gnome.org/buglist.cgi?product=libxml2">Gnome
530	bug tracking database</a> (make sure to use the "libxml2" module name). I
531	look at reports there regularly and it's good to have a reminder when a bug
532	is still open. Be sure to specify that the bug is for the package libxml2.</p>
533
534	<p>For small problems you can try to get help on IRC, the #xml channel on
535	irc.gnome.org (port 6667) usually have a few person subscribed which may help
536	(but there is no guarantee and if a real issue is raised it should go on the
537	mailing-list for archival).</p>
538
539	<p>There is also a mailing-list <a
540	href="mailto:[email protected]">[email protected]</a> for libxml, with an <a
541	href="http://mail.gnome.org/archives/xml/">on-line archive</a> (<a
542	href="http://xmlsoft.org/messages">old</a>). To subscribe to this list,
543	please visit the <a
544	href="http://mail.gnome.org/mailman/listinfo/xml">associated Web</a> page and
545	follow the instructions. <strong>Do not send code, I won't debug it</strong>
546	(but patches are really appreciated!).</p>
547
548	<p>Please note that with the current amount of virus and SPAM, sending mail
549	to the list without being subscribed won't work. There is *far too many
550	bounces* (in the order of a thousand a day !) I cannot approve them manually
551	anymore. If your mail to the list bounced waiting for administrator approval,
552	it is LOST ! Repost it and fix the problem triggering the error. Also please
553	note that <span style="color: #FF0000; background-color: #FFFFFF">emails with
554	a legal warning asking to not copy or redistribute freely the information
555	they contain</span> are <strong>NOT</strong> acceptable for the mailing-list,
556	such mail will as much as possible be discarded automatically, and are less
557	likely to be answered if they made it to the list, <strong>DO NOT</strong>
558	post to the list from an email address where such legal requirements are
559	automatically added, get private paying support if you can't share
560	information.</p>
561
562	<p>Check the following <strong><span style="color: #FF0000">before
563	posting</span></strong>:</p>
564	<ul>
565	<li>Read the <a href="FAQ.html">FAQ</a> and <a href="search.php">use the
566	search engine</a> to get information related to your problem.</li>
567	<li>Make sure you are <a href="ftp://xmlsoft.org/libxml2/">using a recent
568	version</a>, and that the problem still shows up in a recent version.</li>
569	<li>Check the <a href="http://mail.gnome.org/archives/xml/">list
570	archives</a> to see if the problem was reported already. In this case
571	there is probably a fix available, similarly check the <a
572	href="http://bugzilla.gnome.org/buglist.cgi?product=libxml2">registered
573	open bugs</a>.</li>
574	<li>Make sure you can reproduce the bug with xmllint or one of the test
575	programs found in source in the distribution.</li>
576	<li>Please send the command showing the error as well as the input (as an
577	attachment)</li>
578	</ul>
579
580	<p>Then send the bug with associated information to reproduce it to the <a
581	href="mailto:[email protected]">[email protected]</a> list; if it's really libxml
582	related I will approve it. Please do not send mail to me directly, it makes
583	things really hard to track and in some cases I am not the best person to
584	answer a given question, ask on the list.</p>
585
586	<p>To <span style="color: #E50000">be really clear about support</span>:</p>
587	<ul>
588	<li>Support or help <span style="color: #E50000">requests MUST be sent to
589	the list or on bugzilla</span> in case of problems, so that the Question
590	and Answers can be shared publicly. Failing to do so carries the implicit
591	message "I want free support but I don't want to share the benefits with
592	others" and is not welcome. I will automatically Carbon-Copy the
593	[email protected] mailing list for any technical reply made about libxml2 or
594	libxslt.</li>
595	<li>There is <span style="color: #E50000">no guarantee of support</span>. If
596	your question remains unanswered after a week, repost it, making sure you
597	gave all the detail needed and the information requested.</li>
598	<li>Failing to provide information as requested or double checking first
599	for prior feedback also carries the implicit message "the time of the
600	library maintainers is less valuable than my time" and might not be
601	welcome.</li>
602	</ul>
603
604	<p>Of course, bugs reported with a suggested patch for fixing them will
605	probably be processed faster than those without.</p>
606
607	<p>If you're looking for help, a quick look at <a
608	href="http://mail.gnome.org/archives/xml/">the list archive</a> may actually
609	provide the answer. I usually send source samples when answering libxml2
610	usage questions. The <a
611	href="http://xmlsoft.org/html/book1.html">auto-generated documentation</a> is
612	not as polished as I would like (i need to learn more about DocBook), but
613	it's a good starting point.</p>
614
615	<h2><a name="help">How to help</a></h2>
616
617	<p>You can help the project in various ways, the best thing to do first is to
618	subscribe to the mailing-list as explained before, check the <a
619	href="http://mail.gnome.org/archives/xml/">archives </a>and the <a
620	href="http://bugzilla.gnome.org/buglist.cgi?product=libxml2">Gnome bug
621	database</a>:</p>
622	<ol>
623	<li>Provide patches when you find problems.</li>
624	<li>Provide the diffs when you port libxml2 to a new platform. They may not
625	be integrated in all cases but help pinpointing portability problems
626	and</li>
627	<li>Provide documentation fixes (either as patches to the code comments or
628	as HTML diffs).</li>
629	<li>Provide new documentations pieces (translations, examples, etc
630	...).</li>
631	<li>Check the TODO file and try to close one of the items.</li>
632	<li>Take one of the points raised in the archive or the bug database and
633	provide a fix. <a href="mailto:[email protected]">Get in touch with me
634	</a>before to avoid synchronization problems and check that the suggested
635	fix will fit in nicely :-)</li>
636	</ol>
637
638	<h2><a name="Downloads">Downloads</a></h2>
639
640	<p>The latest versions of libxml2 can be found on the <a
641	href="ftp://xmlsoft.org/libxml2/">xmlsoft.org</a> server ( <a
642	href="ftp://xmlsoft.org/libxml2/">FTP</a> and rsync are available), there are also
643	mirrors (<a href="ftp://fr.rpmfind.net/pub/libxml/">France</a> and
644	Antonin Sprinzl also provide <a href="ftp://gd.tuwien.ac.at/pub/libxml/">a
645	mirror in Austria</a>). (NOTE that you need both the <a
646	href="http://rpmfind.net/linux/RPM/libxml2.html">libxml(2)</a> and <a
647	href="http://rpmfind.net/linux/RPM/libxml2-devel.html">libxml(2)-devel</a>
648	packages installed to compile applications using libxml if using RPMs.)</p>
649
650	<p>You can find all the history of libxml(2) and libxslt releases in the <a
651	href="http://xmlsoft.org/sources/old/">old</a> directory. The precompiled
652	Windows binaries made by Igor Zlatovic are available in the <a
653	href="http://xmlsoft.org/sources/win32/">win32</a> directory.</p>
654
655	<p>Binary ports:</p>
656	<ul>
657	<li>RPMs for x86_64 are available directly on <a
658	href="ftp://xmlsoft.org/libxml2/">xmlsoft.org</a>, the source RPM will compile on
659	any architecture supported.</li>
660	<li><a href="mailto:[email protected]">Igor Zlatkovic</a> is now the
661	maintainer of the Windows port, <a
662	href="http://www.zlatkovic.com/projects/libxml/index.html">he provides
663	binaries</a>.</li>
664	<li>OpenCSW provides <a
665	href="http://opencsw.org/packages/libxml2">Solaris
666	binaries</a>.</li>
667	<li><a href="mailto:[email protected]">Steve Ball</a> provides <a
668	href="http://www.explain.com.au/oss/libxml2xslt.html">Mac Os X
669	binaries</a>.</li>
670	<li>The HP-UX porting center provides <a
671	href="http://hpux.connect.org.uk/hppd/hpux/Gnome/">HP-UX binaries</a></li>
672	<li>Bull provides precompiled <a
673	href="http://gnome.bullfreeware.com/new_index.html">RPMs for AIX</a> as
674	patr of their GNOME packages</li>
675	</ul>
676
677	<p>If you know other supported binary ports, please <a
678	href="http://veillard.com/">contact me</a>.</p>
679
680	<p><a name="Snapshot">Snapshot:</a></p>
681	<ul>
682	<li>Code from the GNOME GIT base libxml2 module, updated hourly <a
683	href="ftp://xmlsoft.org/libxml2/libxml2-git-snapshot.tar.gz">libxml2-git-snapshot.tar.gz</a>.</li>
684	<li>Docs, content of the web site, the list archive included <a
685	href="ftp://xmlsoft.org/libxml2/libxml-docs.tar.gz">libxml-docs.tar.gz</a>.</li>
686	</ul>
687
688	<p><a name="Contribs">Contributions:</a></p>
689
690	<p>I do accept external contributions, especially if compiling on another
691	platform, get in touch with the list to upload the package, wrappers for
692	various languages have been provided, and can be found in the <a
693	href="python.html">bindings section</a></p>
694
695	<p>Libxml2 is also available from GIT:</p>
696	<ul>
697	<li><p>See <a href="http://git.gnome.org/browse/libxml2/">libxml2 Git web</a>.
698	To checkout a local tree use:</p>
699	<pre>git clone git://git.gnome.org/libxml2</pre>
700	</li>
701	<li>The <strong>libxslt</strong> module is also present
702	<a href="http://git.gnome.org/browse/libxslt/">there</a>.</li>
703	</ul>
704
705	<h2><a name="News">Releases</a></h2>
706
707	<p>The <a href="ChangeLog.html">change log</a> describes the recents commits
708	to the <a href="http://git.gnome.org/browse/libxml2/">GIT</a> code base.</p>
709
710	<p>Here is the list of public releases:</p>
711
712	<h3>2.9.4: May 23 2016</h3>
713	<ul>
714	<li>Security:<br/>
715	More format string warnings with possible format string vulnerability (David Kilzer),<br/>
716	Avoid building recursive entities (Daniel Veillard),<br/>
717	Heap-based buffer overread in htmlCurrentChar (Pranjal Jumde),<br/>
718	Heap-based buffer-underreads due to xmlParseName (David Kilzer),<br/>
719	Heap use-after-free in xmlSAX2AttributeNs (Pranjal Jumde),<br/>
720	Heap use-after-free in htmlParsePubidLiteral and htmlParseSystemiteral (Pranjal Jumde),<br/>
721	Fix some format string warnings with possible format string vulnerability (David Kilzer),<br/>
722	Detect change of encoding when parsing HTML names (Hugh Davenport),<br/>
723	Fix inappropriate fetch of entities content (Daniel Veillard),<br/>
724	Bug 759398: Heap use-after-free in xmlDictComputeFastKey <https://bugzilla.gnome.org/show_bug.cgi?id=759398> (Pranjal Jumde),<br/>
725	Bug 758605: Heap-based buffer overread in xmlDictAddString <https://bugzilla.gnome.org/show_bug.cgi?id=758605> (Pranjal Jumde),<br/>
726	Bug 758588: Heap-based buffer overread in xmlParserPrintFileContextInternal <https://bugzilla.gnome.org/show_bug.cgi?id=758588> (David Kilzer),<br/>
727	Bug 757711: heap-buffer-overflow in xmlFAParsePosCharGroup <https://bugzilla.gnome.org/show_bug.cgi?id=757711> (Pranjal Jumde),<br/>
728	Add missing increments of recursion depth counter to XML parser. (Peter Simons)<br/>
729	</li>
730
731	<li>Documentation:<br/>
732	Fix typo: s{ ec -> cr }cipt (Jan Pokorný),<br/>
733	Fix typos: dictio{ nn -> n }ar{y,ies} (Jan Pokorný),<br/>
734	Fix typos: PATH_{ SEAPARATOR -> SEPARATOR } (Jan Pokorný),<br/>
735	Correct a typo. (Shlomi Fish)<br/>
736	</li>
737
738	<li>Portability:<br/>
739	Correct the usage of LDFLAGS (Mattias Hansson),<br/>
740	Revert the use of SAVE_LDFLAGS in configure.ac (Mattias Hansson),<br/>
741	libxml2 hardcodes -L/lib in zlib/lzma tests which breaks cross-compiles (Mike Frysinger),<br/>
742	Fix apibuild for a recently added construct (Daniel Veillard),<br/>
743	Use pkg-config to locate zlib when possible (Stewart Brodie),<br/>
744	Use pkg-config to locate ICU when possible (Stewart Brodie),<br/>
745	Portability to non C99 compliant compilers (Patrick Monnerat),<br/>
746	dict.h: Move xmlDictPtr definition before includes to allow direct inclusion. (Patrick Monnerat),<br/>
747	os400: tell about xmllint and xmlcatalog in README400. (Patrick Monnerat),<br/>
748	os400: properly process SGML add in XMLCATALOG command. (Patrick Monnerat),<br/>
749	os400: implement CL command XMLCATALOG. (Patrick Monnerat),<br/>
750	os400: compile and install program xmlcatalog (qshell-only). (Patrick Monnerat),<br/>
751	os400: expand tabs in sources, strip trailing blanks. (Patrick Monnerat),<br/>
752	os400: implement CL command XMLLINT. (Patrick Monnerat),<br/>
753	os400: compile and install program xmllint (qshell-only). (Patrick Monnerat),<br/>
754	os400: initscript make_module(): Use options instead of positional parameters. (Patrick Monnerat),<br/>
755	os400: c14n.rpgle: allow *omit for nullable reference parameters. (Patrick Monnerat),<br/>
756	os400: use like() for double type. (Patrick Monnerat),<br/>
757	os400: use like() for int type. (Patrick Monnerat),<br/>
758	os400: use like() for unsigned int type. (Patrick Monnerat),<br/>
759	os400: use like() for enum types. (Patrick Monnerat),<br/>
760	Add xz to xml2-config --libs output (Baruch Siach),<br/>
761	Bug 760190: configure.ac should be able to build --with-icu without icu-config tool <https://bugzilla.gnome.org/show_bug.cgi?id=760190> (David Kilzer),<br/>
762	win32\VC10\config.h and VS 2015 (Bruce Dawson),<br/>
763	Add configure maintainer mode (orzen)<br/>
764	</li>
765
766	<li>Bug Fixes:<br/>
767	Avoid an out of bound access when serializing malformed strings (Daniel Veillard),<br/>
768	Unsigned addition may overflow in xmlMallocAtomicLoc() (David Kilzer),<br/>
769	Integer signed/unsigned type mismatch in xmlParserInputGrow() (David Kilzer),<br/>
770	Bug 763071: heap-buffer-overflow in xmlStrncat <https://bugzilla.gnome.org/show_bug.cgi?id=763071> (Pranjal Jumde),<br/>
771	Integer overflow parsing port number in URI (Michael Paddon),<br/>
772	Fix an error with regexp on nullable counted char transition (Daniel Veillard),<br/>
773	Fix memory leak with XPath namespace nodes (Nick Wellnhofer),<br/>
774	Fix namespace axis traversal (Nick Wellnhofer),<br/>
775	Fix null pointer deref in docs with no root element (Hugh Davenport),<br/>
776	Fix XSD validation of URIs with ampersands (Alex Henrie),<br/>
777	xmlschemastypes.c: accept endOfDayFrag Times set to "24:00:00" mean "end of day" and should not cause an error. (Patrick Monnerat),<br/>
778	xmlcatalog: flush stdout before interactive shell input. (Patrick Monnerat),<br/>
779	xmllint: flush stdout before interactive shell input. (Patrick Monnerat),<br/>
780	Don't recurse into OP_VALUEs in xmlXPathOptimizeExpression (Nick Wellnhofer),<br/>
781	Fix namespace::node() XPath expression (Nick Wellnhofer),<br/>
782	Fix OOB write in xmlXPathEmptyNodeSet (Nick Wellnhofer),<br/>
783	Fix parsing of NCNames in XPath (Nick Wellnhofer),<br/>
784	Fix OOB read with invalid UTF-8 in xmlUTF8Strsize (Nick Wellnhofer),<br/>
785	Do normalize string-based datatype value in RelaxNG facet checking (Audric Schiltknecht),<br/>
786	Bug 760921: REGRESSION (8eb55d78): doc/examples/io1 test fails after fix for "xmlSaveUri() incorrectly recomposes URIs with rootless paths" <https://bugzilla.gnome.org/show_bug.cgi?id=760921> (David Kilzer),<br/>
787	Bug 760861: REGRESSION (bf9c1dad): Missing results for test/schemas/regexp-char-ref_[01].xsd <https://bugzilla.gnome.org/show_bug.cgi?id=760861> (David Kilzer),<br/>
788	error.c: *input->cur == 0 does not mean no error (Pavel Raiskup),<br/>
789	Add missing RNG test files (David Kilzer),<br/>
790	Bug 760183: REGRESSION (v2.9.3): XML push parser fails with bogus UTF-8 encoding error when multi-byte character in large CDATA section is split across buffer <https://bugzilla.gnome.org/show_bug.cgi?id=760183> (David Kilzer),<br/>
791	Bug 758572: ASAN crash in make check <https://bugzilla.gnome.org/show_bug.cgi?id=758572> (David Kilzer),<br/>
792	Bug 721158: Missing ICU string when doing --version on xmllint <https://bugzilla.gnome.org/show_bug.cgi?id=721158> (David Kilzer),<br/>
793	python 3: libxml2.c wrappers create Unicode str already (Michael Stahl),<br/>
794	Add autogen.sh to distrib (orzen),<br/>
795	Heap-based buffer overread in xmlNextChar (Daniel Veillard)<br/>
796	</li>
797
798	<li>Improvements:<br/>
799	Add more debugging info to runtest (Daniel Veillard),<br/>
800	Implement "runtest -u" mode (David Kilzer),<br/>
801	Add a make rule to rebuild for ASAN (Daniel Veillard)<br/>
802	</li>
803	</ul>
804	<h3>v2.9.3: Nov 20 2015</h3>
805	<ul>
806	<li>Security:<br/>
807	CVE-2015-8242 Buffer overead with HTML parser in push mode (Hugh Davenport),<br/>
808	CVE-2015-7500 Fix memory access error due to incorrect entities boundaries (Daniel Veillard),<br/>
809	CVE-2015-7499-2 Detect incoherency on GROW (Daniel Veillard),<br/>
810	CVE-2015-7499-1 Add xmlHaltParser() to stop the parser (Daniel Veillard),<br/>
811	CVE-2015-5312 Another entity expansion issue (David Drysdale),<br/>
812	CVE-2015-7497 Avoid an heap buffer overflow in xmlDictComputeFastQKey (David Drysdale),<br/>
813	CVE-2015-7498 Avoid processing entities after encoding conversion failures (Daniel Veillard),<br/>
814	CVE-2015-8035 Fix XZ compression support loop (Daniel Veillard),<br/>
815	CVE-2015-7942-2 Fix an error in previous Conditional section patch (Daniel Veillard),<br/>
816	CVE-2015-7942 Another variation of overflow in Conditional sections (Daniel Veillard),<br/>
817	CVE-2015-1819 Enforce the reader to run in constant memory (Daniel Veillard)<br/>
818	CVE-2015-7941_2 Cleanup conditional section error handling (Daniel Veillard),<br/>
819	CVE-2015-7941_1 Stop parsing on entities boundaries errors (Daniel Veillard),<br/>
820	</li>
821
822	<li>Documentation:<br/>
823	Correct spelling of "calling" (Alex Henrie),<br/>
824	Fix a small error in xmllint --format description (Fabien Degomme),<br/>
825	Avoid XSS on the search of xmlsoft.org (Daniel Veillard)<br/>
826	</li>
827
828	<li>Portability:<br/>
829	threads: use forward declarations only for glibc (Michael Heimpold),<br/>
830	Update Win32 configure.js to search for configure.ac (Daniel Veillard)<br/>
831	</li>
832
833	<li>Bug Fixes:<br/>
834	Bug on creating new stream from entity (Daniel Veillard),<br/>
835	Fix some loop issues embedding NEXT (Daniel Veillard),<br/>
836	Do not print error context when there is none (Daniel Veillard),<br/>
837	Avoid extra processing of MarkupDecl when EOF (Hugh Davenport),<br/>
838	Fix parsing short unclosed comment uninitialized access (Daniel Veillard),<br/>
839	Add missing Null check in xmlParseExternalEntityPrivate (Gaurav Gupta),<br/>
840	Fix a bug in CData error handling in the push parser (Daniel Veillard),<br/>
841	Fix a bug on name parsing at the end of current input buffer (Daniel Veillard),<br/>
842	Fix the spurious ID already defined error (Daniel Veillard),<br/>
843	Fix previous change to node sort order (Nick Wellnhofer),<br/>
844	Fix a self assignment issue raised by clang (Scott Graham),<br/>
845	Fail parsing early on if encoding conversion failed (Daniel Veillard),<br/>
846	Do not process encoding values if the declaration if broken (Daniel Veillard),<br/>
847	Silence clang's -Wunknown-attribute (Michael Catanzaro),<br/>
848	xmlMemUsed is not thread-safe (Martin von Gagern),<br/>
849	Fix support for except in nameclasses (Daniel Veillard),<br/>
850	Fix order of root nodes (Nick Wellnhofer),<br/>
851	Allow attributes on descendant-or-self axis (Nick Wellnhofer),<br/>
852	Fix the fix to Windows locking (Steve Nairn),<br/>
853	Fix timsort invariant loop re: Envisage article (Christopher Swenson),<br/>
854	Don't add IDs in xmlSetTreeDoc (Nick Wellnhofer),<br/>
855	Account for ID attributes in xmlSetTreeDoc (Nick Wellnhofer),<br/>
856	Remove various unused value assignments (Philip Withnall),<br/>
857	Fix missing entities after CVE-2014-3660 fix (Daniel Veillard),<br/>
858	Revert "Missing initialization for the catalog module" (Daniel Veillard)<br/>
859	</li>
860
861	<li>Improvements:<br/>
862	Reuse xmlHaltParser() where it makes sense (Daniel Veillard),<br/>
863	xmlStopParser reset errNo (Daniel Veillard),<br/>
864	Reenable xz support by default (Daniel Veillard),<br/>
865	Recover unescaped less-than character in HTML recovery parsing (Daniel Veillard),<br/>
866	Allow HTML serializer to output HTML5 DOCTYPE (Shaun McCance),<br/>
867	Regression test for bug #695699 (Nick Wellnhofer),<br/>
868	Add a couple of XPath tests (Nick Wellnhofer),<br/>
869	Add Python 3 rpm subpackage (Tomas Radej),<br/>
870	libxml2-config.cmake.in: update include directories (Samuel Martin),<br/>
871	Adding example from bugs 738805 to regression tests (Daniel Veillard)<br/>
872	</li>
873
874	<li>Cleanups:<br/>
875	</li>
876	</ul>
877	<h3>2.9.2: Oct 16 2014</h3>
878	<ul>
879	<li>Security:<br/>
880	Fix for CVE-2014-3660 billion laugh variant (Daniel Veillard),<br/>
881	CVE-2014-0191 Do not fetch external parameter entities (Daniel Veillard)<br/>
882	</li>
883
884	<li>Bug Fixes:<br/>
885	fix memory leak xml header encoding field with XML_PARSE_IGNORE_ENC (Bart De Schuymer),<br/>
886	xmlmemory: handle realloc properly (Yegor Yefremov),<br/>
887	Python generator bug raised by the const change (Daniel Veillard),<br/>
888	Windows Critical sections not released correctly (Daniel Veillard),<br/>
889	Parser error on repeated recursive entity expansion containing &lt; (Daniel Veillard),<br/>
890	xpointer : fixing Null Pointers (Gaurav Gupta),<br/>
891	Remove Unnecessary Null check in xpointer.c (Gaurav Gupta),<br/>
892	parser bug on misformed namespace attributes (Dennis Filder),<br/>
893	Pointer dereferenced before null check (Daniel Veillard),<br/>
894	Leak of struct addrinfo in xmlNanoFTPConnect() (Gaurav Gupta),<br/>
895	Possible overflow in HTMLParser.c (Daniel Veillard),<br/>
896	python/tests/sync.py assumes Python dictionaries are ordered (John Beck),<br/>
897	Fix Enum check and missing break (Gaurav Gupta),<br/>
898	xmlIO: Handle error returns from dup() (Philip Withnall),<br/>
899	Fix a problem properly saving URIs (Daniel Veillard),<br/>
900	wrong error column in structured error when parsing attribute values (Juergen Keil),<br/>
901	wrong error column in structured error when skipping whitespace in xml decl (Juergen Keil),<br/>
902	no error column in structured error handler for xml schema validation errors (Juergen Keil),<br/>
903	Couple of Missing Null checks (Gaurav Gupta),<br/>
904	Add couple of missing Null checks (Daniel Veillard),<br/>
905	xmlschemastypes: Fix potential array overflow (Philip Withnall),<br/>
906	runtest: Fix a memory leak on parse failure (Philip Withnall),<br/>
907	xmlIO: Fix an FD leak on gzdopen() failure (Philip Withnall),<br/>
908	xmlcatalog: Fix a memory leak on quit (Philip Withnall),<br/>
909	HTMLparser: Correctly initialise a stack allocated structure (Philip Withnall),<br/>
910	Check for tmon in _xmlSchemaDateAdd() is incorrect (David Kilzer),<br/>
911	Avoid Possible Null Pointer in trio.c (Gaurav Gupta),<br/>
912	Fix processing in SAX2 in case of an allocation failure (Daniel Veillard),<br/>
913	XML Shell command "cd" does not handle "/" at end of path (Daniel Veillard),<br/>
914	Fix various Missing Null checks (Gaurav Gupta),<br/>
915	Fix a potential NULL dereference (Daniel Veillard),<br/>
916	Add a couple of misisng check in xmlRelaxNGCleanupTree (Gaurav Gupta),<br/>
917	Add a missing argument check (Gaurav Gupta),<br/>
918	Adding a check in case of allocation error (Gaurav Gupta),<br/>
919	xmlSaveUri() incorrectly recomposes URIs with rootless paths (Dennis Filder),<br/>
920	Adding some missing NULL checks (Gaurav),<br/>
921	Fixes for xmlInitParserCtxt (Daniel Veillard),<br/>
922	Fix regressions introduced by CVE-2014-0191 patch (Daniel Veillard),<br/>
923	erroneously ignores a validation error if no error callback set (Daniel Veillard),<br/>
924	xmllint was not parsing the --c14n11 flag (Sérgio Batista),<br/>
925	Avoid Possible null pointer dereference in memory debug mode (Gaurav),<br/>
926	Avoid Double Null Check (Gaurav),<br/>
927	Restore context size and position after XPATH_OP_ARG (Nick Wellnhofer),<br/>
928	Fix xmlParseInNodeContext() if node is not element (Daniel Veillard),<br/>
929	Avoid a possible NULL pointer dereference (Gaurav),<br/>
930	Fix xmlTextWriterWriteElement when a null content is given (Daniel Veillard),<br/>
931	Fix an typo 'onrest' in htmlScriptAttributes (Daniel Veillard),<br/>
932	fixing a ptotential uninitialized access (Daniel Veillard),<br/>
933	Fix an fd leak in an error case (Daniel Veillard),<br/>
934	Missing initialization for the catalog module (Daniel Veillard),<br/>
935	Handling of XPath function arguments in error case (Nick Wellnhofer),<br/>
936	Fix a couple of missing NULL checks (Gaurav),<br/>
937	Avoid a possibility of dangling encoding handler (Gaurav),<br/>
938	Fix HTML push parser to accept HTML_PARSE_NODEFDTD (Arnold Hendriks),<br/>
939	Fix a bug loading some compressed files (Mike Alexander),<br/>
940	Fix XPath node comparison bug (Gaurav),<br/>
941	Type mismatch in xmlschemas.c (Gaurav),<br/>
942	Type mismatch in xmlschemastypes.c (Gaurav),<br/>
943	Avoid a deadcode in catalog.c (Daniel Veillard),<br/>
944	run close socket on Solaris, same as we do on other platforms (Denis Pauk),<br/>
945	Fix pointer dereferenced before null check (Gaurav),<br/>
946	Fix a potential NULL dereference in tree code (Daniel Veillard),<br/>
947	Fix potential NULL pointer dereferences in regexp code (Gaurav),<br/>
948	xmllint --pretty crashed without following numeric argument (Tim Galeckas),<br/>
949	Fix XPath expressions of the form '@ns:*' (Nick Wellnhofer),<br/>
950	Fix XPath '//' optimization with predicates (Nick Wellnhofer),<br/>
951	Clear up a potential NULL dereference (Daniel Veillard),<br/>
952	Fix a possible NULL dereference (Gaurav),<br/>
953	Avoid crash if allocation fails (Daniel Veillard),<br/>
954	Remove occasional leading space in XPath number formatting (Daniel Veillard),<br/>
955	Fix handling of mmap errors (Daniel Veillard),<br/>
956	Catch malloc error and exit accordingly (Daniel Veillard),<br/>
957	missing else in xlink.c (Ami Fischman),<br/>
958	Fix a parsing bug on non-ascii element and CR/LF usage (Daniel Veillard),<br/>
959	Fix a regression in xmlGetDocCompressMode() (Daniel Veillard),<br/>
960	properly quote the namespace uris written out during c14n (Aleksey Sanin),<br/>
961	Remove premature XInclude check on URI being relative (Alexey Neyman),<br/>
962	Fix missing break on last() function for attributes (dcb),<br/>
963	Do not URI escape in server side includes (Romain Bondue),<br/>
964	Fix an error in xmlCleanupParser (Alexander Pastukhov)<br/>
965	</li>
966
967	<li>Documentation:<br/>
968	typo in error messages "colon are forbidden from..." (Daniel Veillard),<br/>
969	Fix a link to James SAX documentation old page (Daniel Veillard),<br/>
970	Fix typos in relaxng.c (Jan Pokorný),<br/>
971	Fix a doc typo (Daniel Veillard),<br/>
972	Fix typos in {tree,xpath}.c (errror) (Jan Pokorný),<br/>
973	Add limitations about encoding conversion (Daniel Veillard),<br/>
974	Fix typos in xmlschemas{,types}.c (Jan Pokorný),<br/>
975	Fix incorrect spelling entites->entities (Jan Pokorný),<br/>
976	Forgot to document 2.9.1 release, regenerate docs (Daniel Veillard)<br/>
977	</li>
978
979	<li>Portability:<br/>
980	AC_CONFIG_FILES and executable bit (Roumen Petrov),<br/>
981	remove HAVE_CONFIG_H dependency in testlimits.c (Roumen Petrov),<br/>
982	fix some tabs mixing incompatible with python3 (Roumen Petrov),<br/>
983	Visual Studio 14 CTP defines snprintf() (Francis Dupont),<br/>
984	OS400: do not try to copy unexisting doc files (Patrick Monnerat),<br/>
985	OS400: use either configure.ac or configure.in. (Patrick Monnerat),<br/>
986	os400: make-src.sh: create physical file with target CCSID (Patrick Monnerat),<br/>
987	OS400: Add some more C macros equivalent procedures. (Patrick Monnerat),<br/>
988	OS400: use C macros to implement equivalent RPG support procedures. (Patrick Monnerat),<br/>
989	OS400: implement XPath macros as procedures for ILE/RPG support. (Patrick Monnerat),<br/>
990	OS400: include in distribution tarball. (Patrick Monnerat),<br/>
991	OS400: Add README: compilation directives and OS/400 specific stuff. (Patrick Monnerat),<br/>
992	OS400: Add compilation scripts. (Patrick Monnerat),<br/>
993	OS400: ILE RPG language header files. (Patrick Monnerat),<br/>
994	OS400: implement some macros as functions for ILE/RPG language support (that as no macros). (Patrick Monnerat),<br/>
995	OS400: UTF8<-->EBCDIC wrappers for system and external library calls (Patrick Monnerat),<br/>
996	OS400: Easy character transcoding support (Patrick Monnerat),<br/>
997	OS400: iconv functions compatibility wrappers and table builder. (Patrick Monnerat),<br/>
998	OS400: create architecture directory. Implement dlfcn emulation. (Patrick Monnerat),<br/>
999	Fix building when configuring without xpath and xptr (Daniel Veillard),<br/>
1000	configure: Add --with-python-install-dir (Jonas Eriksson),<br/>
1001	Fix compilation with minimum and xinclude. (Nicolas Le Cam),<br/>
1002	Compile out use of xmlValidateNCName() when not available. (Nicolas Le Cam),<br/>
1003	Fix compilation with minimum and schematron. (Nicolas Le Cam),<br/>
1004	Legacy needs xmlSAX2StartElement() and xmlSAX2EndElement(). (Nicolas Le Cam),<br/>
1005	Don't use xmlValidateName() when not available. (Nicolas Le Cam),<br/>
1006	Fix a portability issue on Windows (Longstreth Jon),<br/>
1007	Various portability patches for OpenVMS (Jacob (Jouk) Jansen),<br/>
1008	Use specific macros for portability to OS/400 (Patrick Monnerat),<br/>
1009	Add macros needed for OS/400 portability (Patrick Monnerat),<br/>
1010	Portability patch for fopen on OS/400 (Patrick Monnerat),<br/>
1011	Portability fixes for OS/400 (Patrick Monnerat),<br/>
1012	Improve va_list portability (Patrick Monnerat),<br/>
1013	Portability fix (Patrick Monnerat),<br/>
1014	Portability fix (Patrick Monnerat),<br/>
1015	Generic portability fix (Patrick Monnerat),<br/>
1016	Shortening lines in headers (Patrick Monnerat),<br/>
1017	build: Use pkg-config to find liblzma in preference to AC_CHECK_LIB (Philip Withnall),<br/>
1018	build: Add @LZMA_LIBS@ to libxml’s pkg-config files (Philip Withnall),<br/>
1019	fix some tabs mixing incompatible with python3 (Daniel Veillard),<br/>
1020	add additional defines checks for support "./configure --with-minimum" (Denis Pauk),<br/>
1021	Another round of fixes for older versions of Python (Arfrever Frehtes Taifersar Arahesis),<br/>
1022	python: fix drv_libxml2.py for python3 compatibility (Alexandre Rostovtsev),<br/>
1023	python: Fix compiler warnings when building python3 bindings (Armin K),<br/>
1024	Fix for compilation with python 2.6.8 (Petr Sumbera)<br/>
1025	</li>
1026
1027	<li>Improvements:<br/>
1028	win32/libxml2.def.src after rebuild in doc (Roumen Petrov),<br/>
1029	elfgcchack.h: more legacy needs xmlSAX2StartElement() and xmlSAX2EndElement() (Roumen Petrov),<br/>
1030	elfgcchack.h: add xmlXPathNodeEval and xmlXPathSetContextNode (Roumen Petrov),<br/>
1031	Provide cmake module (Samuel Martin),<br/>
1032	Fix a couple of issues raised by make dist (Daniel Veillard),<br/>
1033	Fix and add const qualifiers (Kurt Roeckx),<br/>
1034	Preparing for upcoming release of 2.9.2 (Daniel Veillard),<br/>
1035	Fix zlib and lzma libraries check via command line (Dmitriy),<br/>
1036	wrong error column in structured error when parsing end tag (Juergen Keil),<br/>
1037	doc/news.html: small update to avoid line join while generating NEWS. (Patrick Monnerat),<br/>
1038	Add methods for python3 iterator (Ron Angeles),<br/>
1039	Support element node traversal in document fragments. (Kyle VanderBeek),<br/>
1040	xmlNodeSetName: Allow setting the name to a substring of the currently set name (Tristan Van Berkom),<br/>
1041	Added macros for argument casts (Eric Zurcher),<br/>
1042	adding init calls to xml and html Read parsing entry points (Daniel Veillard),<br/>
1043	Get rid of 'REPLACEMENT CHARACTER' Unicode chars in xmlschemas.c (Jan Pokorný),<br/>
1044	Implement choice for name classes on attributes (Shaun McCance),<br/>
1045	Two small namespace tweaks (Daniel Veillard),<br/>
1046	xmllint --memory should fail on empty files (Daniel Veillard),<br/>
1047	Cast encoding name to char pointer to match arg type (Nikolay Sivov)<br/>
1048	</li>
1049
1050	<li>Cleanups:<br/>
1051	Removal of old configure.in (Daniel Veillard),<br/>
1052	Unreachable code in tree.c (Gaurav Gupta),<br/>
1053	Remove a couple of dead conditions (Gaurav Gupta),<br/>
1054	Avoid some dead code and cleanup in relaxng.c (Gaurav),<br/>
1055	Drop not needed checks (Denis Pauk),<br/>
1056	Fix a wrong test (Daniel Veillard)<br/>
1057	</li>
1058	</ul>
1059	<h3>2.9.1: Apr 19 2013</h3>
1060	<ul>
1061	<li> Features:<br/>
1062	Support for Python3 (Daniel Veillard),<br/>
1063	Add xmlXPathSetContextNode and xmlXPathNodeEval (Alex Bligh)<br/>
1064	</li>
1065
1066	<li> Documentation:<br/>
1067	Add documentation for xmllint --xpath (Daniel Veillard),<br/>
1068	Fix the URL of the SAX documentation from James (Daniel Veillard),<br/>
1069	Fix spelling of "length". (Michael Wood)<br/>
1070	</li>
1071
1072	<li> Portability:<br/>
1073	Fix python bindings with versions older than 2.7 (Daniel Veillard),<br/>
1074	rebuild docs:Makefile.am (Roumen Petrov),<br/>
1075	elfgcchack.h after rebuild in doc (Roumen Petrov),<br/>
1076	elfgcchack for buf module (Roumen Petrov),<br/>
1077	Fix a uneeded and wrong extra link parameter (Daniel Veillard),<br/>
1078	Few cleanup patches for Windows (Denis Pauk),<br/>
1079	Fix rpmbuild --nocheck (Mark Salter),<br/>
1080	Fix for win32/configure.js and WITH_THREAD_ALLOC (Daniel Richard),<br/>
1081	Fix Broken multi-arch support in xml2-config (Daniel Veillard),<br/>
1082	Fix a portability issue for GCC < 3.4.0 (Daniel Veillard),<br/>
1083	Windows build fixes (Daniel Richard),<br/>
1084	Fix a thread portability problem (Friedrich Haubensak),<br/>
1085	Downgrade autoconf requirement to 2.63 (Daniel Veillard)<br/>
1086	</li>
1087
1088	<li> Bug Fixes:<br/>
1089	Fix a linking error for python bindings (Daniel Veillard),<br/>
1090	Fix a couple of return without value (Jüri Aedla),<br/>
1091	Improve the hashing functions (Daniel Franke),<br/>
1092	Improve handling of xmlStopParser() (Daniel Veillard),<br/>
1093	Remove risk of lockup in dictionary initialization (Daniel Veillard),<br/>
1094	Activate detection of encoding in external subset (Daniel Veillard),<br/>
1095	Fix an output buffer flushing conversion bug (Mikhail Titov),<br/>
1096	Fix an old bug in xmlSchemaValidateOneElement (Csaba László),<br/>
1097	Fix configure cannot remove messages (Gilles Espinasse),<br/>
1098	fix schema validation in combination with xsi:nil (Daniel Veillard),<br/>
1099	xmlCtxtReadFile doesn't work with literal IPv6 URLs (Steve Wolf),<br/>
1100	Fix a few problems with setEntityLoader (Alexey Neyman),<br/>
1101	Detect excessive entities expansion upon replacement (Daniel Veillard),<br/>
1102	Fix the flushing out of raw buffers on encoding conversions (Daniel,<br/>
1103	Veillard),<br/>
1104	Fix some buffer conversion issues (Daniel Veillard),<br/>
1105	When calling xmlNodeDump make sure we grow the buffer quickly (Daniel,<br/>
1106	Veillard),<br/>
1107	Fix an error in the progressive DTD parsing code (Dan Winship),<br/>
1108	xmllint should not load DTD by default when using the reader (Daniel,<br/>
1109	Veillard),<br/>
1110	Try IBM-037 when looking for EBCDIC handlers (Petr Sumbera),<br/>
1111	Fix potential out of bound access (Daniel Veillard),<br/>
1112	Fix large parse of file from memory (Daniel Veillard),<br/>
1113	Fix a bug in the nsclean option of the parser (Daniel Veillard),<br/>
1114	Fix a regression in 2.9.0 breaking validation while streaming (Daniel,<br/>
1115	Veillard),<br/>
1116	Remove potential calls to exit() (Daniel Veillard)<br/>
1117	</li>
1118
1119	<li> Improvements:<br/>
1120	Regenerated API, and testapi, rebuild documentation (Daniel Veillard),<br/>
1121	Fix tree iterators broken by 2to3 script (Daniel Veillard),<br/>
1122	update all tests for Python3 and Python2 (Daniel Veillard),<br/>
1123	A few more fixes for python 3 affecting libxml2.py (Daniel Veillard),<br/>
1124	Fix compilation on Python3 (Daniel Veillard),<br/>
1125	Converting apibuild.py to python3 (Daniel Veillard),<br/>
1126	First pass at starting porting to python3 (Daniel Veillard),<br/>
1127	updated configure.in for python3 (Daniel Veillard),<br/>
1128	Add support for xpathRegisterVariable in Python (Shaun McCance),<br/>
1129	Added a regression tests from bug 694228 data (Daniel Veillard),<br/>
1130	Cache presence of '<' in entities content (Daniel Veillard),<br/>
1131	Avoid extra processing on entities (Daniel Veillard),<br/>
1132	Python binding for xmlRegisterInputCallback (Alexey Neyman),<br/>
1133	Python bindings: DOM casts everything to xmlNode (Alexey Neyman),<br/>
1134	Define LIBXML_THREAD_ALLOC_ENABLED via xmlversion.h (Tim Starling),<br/>
1135	Adding streaming validation to runtest checks (Daniel Veillard),<br/>
1136	Add a --pushsmall option to xmllint (Daniel Veillard)<br/>
1137	</li>
1138
1139	<li> Cleanups:<br/>
1140	Switched comment in file to UTF-8 encoding (Daniel Veillard),<br/>
1141	Extend gitignore (Daniel Veillard),<br/>
1142	Silent the new python test on input (Alexey Neyman),<br/>
1143	Cleanup of a duplicate test (Daniel Veillard),<br/>
1144	Cleanup on duplicate test expressions (Daniel Veillard),<br/>
1145	Fix compiler warning after 153cf15905cf4ec080612ada6703757d10caba1e (Patrick,<br/>
1146	Gansterer),<br/>
1147	Spec cleanups and a fix for multiarch support (Daniel Veillard),<br/>
1148	Silence a clang warning (Daniel Veillard),<br/>
1149	Cleanup the Copyright to be pure MIT Licence wording (Daniel Veillard),<br/>
1150	rand_seed should be static in dict.c (Wouter Van Rooy),<br/>
1151	Fix typos in parser comments (Jan Pokorný)<br/>
1152	</li>
1153	</ul>
1154	<h3>2.9.0: Sep 11 2012</h3>
1155	<ul>
1156	<li> Features:<br/>
1157	A few new API entry points,<br/>
1158	More resilient push parser mode,<br/>
1159	A lot of portability improvement,<br/>
1160	Faster XPath evaluation<br/>
1161	</li>
1162
1163	<li> Documentation:<br/>
1164	xml2-config.1 markup error (Christian Weisgerber),<br/>
1165	libxml(3) manpage typo fix (John Bradshaw),<br/>
1166	More cleanups to the documentation part of libxml2 (Daniel Richard G)<br/>
1167	</li>
1168
1169	<li> Portability:<br/>
1170	Bug 676544 - fails to build with --without-sax1 (Akira TAGOH),<br/>
1171	fix builds not having stdint.h (Rob Richards),<br/>
1172	GetProcAddressA is available only on WinCE (Daniel Veillard),<br/>
1173	More updates and cleanups on autotools and Makefiles (Daniel Richard G),<br/>
1174	More changes for Win32 compilation (Eric Zurcher),<br/>
1175	Basic changes for Win32 builds of release 2.9.0: compile buf.c (Eric Zurcher),<br/>
1176	Bundles all generated files for python into the distribution (Daniel Richard G),<br/>
1177	Fix compiler warnings of wincecompat.c (Patrick Gansterer),<br/>
1178	Fix non __GNUC__ build (Patrick Gansterer),<br/>
1179	Fix windows unicode build (Patrick Gansterer),<br/>
1180	clean redefinition of {v}snprintf in C-source (Roumen Petrov),<br/>
1181	use xmlBuf... if DEBUG_INPUT is defined (Roumen Petrov),<br/>
1182	fix runtests to use pthreads support for various Unix platforms (Daniel Richard G),<br/>
1183	Various "make distcheck" and portability fixups 2nd part (Daniel Richard G),<br/>
1184	Various "make distcheck" and portability fixups (Daniel Richard G),<br/>
1185	Fix compilation on older Visual Studio (Daniel Veillard)<br/>
1186	</li>
1187
1188	<li> Bug Fixes:<br/>
1189	Change the XPath code to percolate allocation errors (Daniel Veillard),<br/>
1190	Fix reuse of xmlInitParser (Daniel Veillard),<br/>
1191	Fix potential crash on entities errors (Daniel Veillard),<br/>
1192	initialize var (Rob Richards),<br/>
1193	Fix the XPath arity check to also check the XPath stack limits (Daniel Veillard),<br/>
1194	Fix problem with specific and generic error handlers (Pietro Cerutti),<br/>
1195	Avoid a potential infinite recursion (Daniel Veillard),<br/>
1196	Fix an XSD error when generating internal automata (Daniel Veillard),<br/>
1197	Patch for xinclude of text using multibyte characters (Vitaly Ostanin),<br/>
1198	Fix a segfault on XSD validation on pattern error (Daniel Veillard),<br/>
1199	Fix missing xmlsave.h module which was ignored in recent builds (Daniel Veillard),<br/>
1200	Add a missing element check (Daniel Veillard),<br/>
1201	Adding various checks on node type though the API (Daniel Veillard),<br/>
1202	Namespace nodes can't be unlinked with xmlUnlinkNode (Daniel Veillard),<br/>
1203	Fix make dist to include new private header files (Daniel Veillard),<br/>
1204	More fixups on the push parser behaviour (Daniel Veillard),<br/>
1205	Strengthen behaviour of the push parser in problematic situations (Daniel Veillard),<br/>
1206	Enforce XML_PARSER_EOF state handling through the parser (Daniel Veillard),<br/>
1207	Fixup limits parser (Daniel Veillard),<br/>
1208	Do not fetch external parsed entities (Daniel Veillard),<br/>
1209	Fix an error in previous commit (Aron Xu),<br/>
1210	Fix entities local buffers size problems (Daniel Veillard),<br/>
1211	Fix parser local buffers size problems (Daniel Veillard),<br/>
1212	Fix a failure to report xmlreader parsing failures (Daniel Veillard)<br/>
1213	</li>
1214
1215	<li> Improvements:<br/>
1216	Keep libxml2.syms when running "make distclean" (Daniel Veillard),<br/>
1217	Allow to set the quoting character of an xmlWriter (Csaba Raduly),<br/>
1218	Keep non-significant blanks node in HTML parser (Daniel Veillard),<br/>
1219	Add a forbidden variable error number and message to XPath (Daniel Veillard),<br/>
1220	Support long path names on WNT (Michael Stahl),<br/>
1221	Improve HTML escaping of attribute on output (Daniel Veillard),<br/>
1222	Handle ICU_LIBS as LIBADD, not LDFLAGS to prevent linking errors (Arfrever Frehtes Taifersar Arahesis),<br/>
1223	Switching XPath node sorting to Timsort (Vojtech Fried),<br/>
1224	Optimizing '//' in XPath expressions (Nick Wellnhofer),<br/>
1225	Expose xmlBufShrink in the public tree API (Daniel Veillard),<br/>
1226	Visible HTML elements close the head tag (Conrad Irwin),<br/>
1227	Fix file and line report for XSD SAX and reader streaming validation (Daniel Veillard),<br/>
1228	Fix const qualifyer to definition of xmlBufferDetach (Daniel Veillard),<br/>
1229	minimize use of HAVE_CONFIG_H (Roumen Petrov),<br/>
1230	fixup regression in Various "make distcheck" and portability fixups (Roumen Petrov),<br/>
1231	Add support for big line numbers in error reporting (Daniel Veillard),<br/>
1232	Avoid using xmlBuffer for serialization (Daniel Veillard),<br/>
1233	Improve compatibility between xmlBuf and xmlBuffer (Daniel Veillard),<br/>
1234	Provide new accessors for xmlOutputBuffer (Daniel Veillard),<br/>
1235	Improvements for old buffer compatibility (Daniel Veillard),<br/>
1236	Expand the limit test program (Daniel Veillard),<br/>
1237	Improve error reporting on parser errors (Daniel Veillard),<br/>
1238	Implement some default limits in the XPath module (Daniel Veillard),<br/>
1239	Introduce some default parser limits (Daniel Veillard),<br/>
1240	Cleanups and new limit APIs for dictionaries (Daniel Veillard),<br/>
1241	Fixup for buf.c (Daniel Veillard),<br/>
1242	Cleanup URI module memory allocation code (Daniel Veillard),<br/>
1243	Extend testlimits (Daniel Veillard),<br/>
1244	More avoid quadratic behaviour (Daniel Veillard),<br/>
1245	Impose a reasonable limit on PI size (Daniel Veillard),<br/>
1246	first version of testlimits new test (Daniel Veillard),<br/>
1247	Avoid quadratic behaviour in some push parsing cases (Daniel Veillard),<br/>
1248	Impose a reasonable limit on comment size (Daniel Veillard),<br/>
1249	Impose a reasonable limit on attribute size (Daniel Veillard),<br/>
1250	Harden the buffer code and make it more compatible (Daniel Veillard),<br/>
1251	More cleanups for input/buffers code (Daniel Veillard),<br/>
1252	Cleanup function xmlBufResetInput(),<br/> to set input from Buffer (Daniel Veillard)
1253	Swicth the test program for characters to new input buffers (Daniel Veillard),<br/>
1254	Convert the HTML tree module to the new buffers (Daniel Veillard),<br/>
1255	Convert of the HTML parser to new input buffers (Daniel Veillard),<br/>
1256	Convert the writer to new output buffer and save APIs (Daniel Veillard),<br/>
1257	Convert XMLReader to the new input buffers (Daniel Veillard),<br/>
1258	New saving functions using xmlBuf and conversion (Daniel Veillard),<br/>
1259	Provide new xmlBuf based saving functions (Daniel Veillard),<br/>
1260	Convert XInclude to the new input buffers (Daniel Veillard),<br/>
1261	Convert catalog code to the new input buffers (Daniel Veillard),<br/>
1262	Convert C14N to the new Input buffer (Daniel Veillard),<br/>
1263	Convert xmlIO.c to the new input and output buffers (Daniel Veillard),<br/>
1264	Convert XML parser to the new input buffers (Daniel Veillard),<br/>
1265	Incompatible change to the Input and Output buffers (Daniel Veillard),<br/>
1266	Adding new encoding function to deal with the new structures (Daniel Veillard),<br/>
1267	Convert XPath to xmlBuf (Daniel Veillard),<br/>
1268	Adding a new buf module for buffers (Daniel Veillard),<br/>
1269	Memory error within SAX2 reuse common framework (Daniel Veillard),<br/>
1270	Fix xmllint --xpath node initialization (Daniel Veillard)<br/>
1271	</li>
1272
1273	<li> Cleanups:<br/>
1274	Various cleanups to avoid compiler warnings (Daniel Veillard),<br/>
1275	Big space and tab cleanup (Daniel Veillard),<br/>
1276	Followup to LibXML2 docs/examples cleanup patch (Daniel Veillard),<br/>
1277	Second round of cleanups for LibXML2 docs/examples (Daniel Richard),<br/>
1278	Remove all .cvsignore as they are not used anymore (Daniel Veillard),<br/>
1279	Fix a Timsort function helper comment (Daniel Veillard),<br/>
1280	Small cleanup for valgrind target (Daniel Veillard),<br/>
1281	Patch for portability of latin characters in C files (Daniel Veillard),<br/>
1282	Cleanup some of the parser code (Daniel Veillard),<br/>
1283	Fix a variable name in comment (Daniel Veillard),<br/>
1284	Regenerated testapi.c (Daniel Veillard),<br/>
1285	Regenerating docs and API files (Daniel Veillard),<br/>
1286	Small cleanup of unused variables in test (Daniel Veillard),<br/>
1287	Expand .gitignore with more files (Daniel Veillard)<br/>
1288	</li>
1289	</ul>
1290	<h3>2.8.0: May 23 2012</h3>
1291	<ul>
1292	<li>Features:
1293	add lzma compression support (Anders F Bjorklund)
1294	</li>
1295
1296	<li>Documentation:
1297	xmlcatalog: Add uri and delegateURI to possible add types in man page. (Ville Skyttä),
1298	Update README.tests (Daniel Veillard),
1299	URI handling code is not OOM resilient (Daniel Veillard),
1300	Fix an error in comment (Daniel Veillard),
1301	Fixed bug #617016 (Daniel Mustieles),
1302	Fixed two typos in the README document (Daniel Neel),
1303	add generated html files (Anders F Bjorklund),
1304	Clarify the need to use xmlFreeNode after xmlUnlinkNode (Daniel Veillard),
1305	Improve documentation a bit (Daniel Veillard),
1306	Updated URL for lxml python bindings (Daniel Veillard)
1307	</li>
1308
1309	<li>Portability:
1310	Restore code for Windows compilation (Daniel Veillard),
1311	Remove git error message during configure (Christian Dywan),
1312	xmllint: Build fix for endTimer if !defined(HAVE_GETTIMEOFDAY) (Patrick R. Gansterer),
1313	remove a bashism in confgure.in (John Hein),
1314	undef ERROR if already defined (Patrick R. Gansterer),
1315	Fix library problems with mingw-w64 (Michael Cronenworth),
1316	fix windows build. ifdef addition from bug 666491 makes no sense (Rob Richards),
1317	prefer native threads on win32 (Sam Thursfield),
1318	Allow to compile with Visual Studio 2010 (Thomas Lemm),
1319	Fix mingw's snprintf configure check (Andoni Morales),
1320	fixed a 64bit big endian issue (Marcus Meissner),
1321	Fix portability failure if netdb.h lacks NO_ADDRESS (Daniel Veillard),
1322	Fix windows build from lzma addition (Rob Richards),
1323	autogen: Only check for libtoolize (Colin Walters),
1324	Fix the Windows build files (Patrick von Reth),
1325	634846 Remove a linking option breaking Windows VC10 (Daniel Veillard),
1326	599241 fix an initialization problem on Win64 (Andrew W. Nosenko),
1327	fix win build (Rob Richards)
1328	</li>
1329
1330	<li>Bug fixes:
1331	Part for rand_r checking missing (Daniel Veillard),
1332	Cleanup on randomization (Daniel Veillard),
1333	Fix undefined reference in python module (Pacho Ramos),
1334	Fix a race in xmlNewInputStream (Daniel Veillard),
1335	Fix weird streaming RelaxNG errors (Noam),
1336	Fix various bugs in new code raised by the API checking (Daniel Veillard),
1337	Fix various problems with "make dist" (Daniel Veillard),
1338	Fix a memory leak in the xzlib code (Daniel Veillard),
1339	HTML parser error with <noscript> in the <head> (Denis Pauk),
1340	XSD: optional element in complex type extension (Remi Gacogne),
1341	Fix html serialization error and htmlSetMetaEncoding() (Daniel Veillard),
1342	Fix a wrong return value in previous patch (Daniel Veillard),
1343	Fix an uninitialized variable use (Daniel Veillard),
1344	Fix a compilation problem with --minimum (Brandon Slack),
1345	Remove redundant and ungarded include of resolv.h (Daniel Veillard),
1346	xinclude with parse="text" does not use the entity loader (Shaun McCance),
1347	Allow to parse 1 byte HTML files (Denis Pauk),
1348	Patch that fixes the skipping of the HTML_PARSE_NOIMPLIED flag (Martin Schröder),
1349	Avoid memory leak if xmlParserInputBufferCreateIO fails (Lin Yi-Li),
1350	Prevent an infinite loop when dumping a node with encoding problems (Timothy Elliott),
1351	xmlParseNodeInContext problems with an empty document (Tim Elliott),
1352	HTML element position is not detected propperly (Pavel Andrejs),
1353	Fix an off by one pointer access (Jüri Aedla),
1354	Try to fix a problem with entities in SAX mode (Daniel Veillard),
1355	Fix a crash with xmllint --path on empty results (Daniel Veillard),
1356	Fixed bug #667946 (Daniel Mustieles),
1357	Fix a logic error in Schemas Component Constraints (Ryan Sleevi),
1358	Fix a wrong enum type use in Schemas Types (Nico Weber),
1359	Fix SAX2 builder in case of undefined attributes namespace (Daniel Veillard),
1360	Fix SAX2 builder in case of undefined element namespaces (Daniel Veillard),
1361	fix reference to STDOUT_FILENO on MSVC (Tay Ray Chuan),
1362	fix a pair of possible out of array char references (Daniel Veillard),
1363	Fix an allocation error when copying entities (Daniel Veillard),
1364	Make sure the parser returns when getting a Stop order (Chris Evans),
1365	Fix some potential problems on reallocation failures(parser.c) (Xia Xinfeng),
1366	Fix a schema type duration comparison overflow (Daniel Veillard),
1367	Fix an unimplemented part in RNG value validation (Daniel Veillard),
1368	Fix missing error status in XPath evaluation (Daniel Veillard),
1369	Hardening of XPath evaluation (Daniel Veillard),
1370	Fix an off by one error in encoding (Daniel Veillard),
1371	Fix RELAX NG include bug #655288 (Shaun McCance),
1372	Fix XSD validation bug #630130 (Toyoda Eizi),
1373	Fix some potential problems on reallocation failures (Chris Evans),
1374	__xmlRaiseError: fix use of the structured callback channel (Dmitry V. Levin),
1375	__xmlRaiseError: fix the structured callback channel's data initialization (Dmitry V. Levin),
1376	Fix memory corruption when xmlParseBalancedChunkMemoryInternal is called from xmlParseBalancedChunk (Rob Richards),
1377	Small fix for previous commit (Daniel Veillard),
1378	Fix a potential freeing error in XPath (Daniel Veillard),
1379	Fix a potential memory access error (Daniel Veillard),
1380	Reactivate the shared library versionning script (Daniel Veillard)
1381	</li>
1382
1383	<li>Improvements:
1384	use mingw C99 compatible functions {v}snprintf instead those from MSVC runtime (Roumen Petrov),
1385	New symbols added for the next release (Daniel Veillard),
1386	xmlTextReader bails too quickly on error (Andy Lutomirski),
1387	Use a hybrid allocation scheme in xmlNodeSetContent (Conrad Irwin),
1388	Use buffers when constructing string node lists. (Conrad Irwin),
1389	Add HTML parser support for HTML5 meta charset encoding declaration (Denis Pauk),
1390	wrong message for double hyphen in comment XML error (Bryan Henderson),
1391	Fix "make tst" to grab lzma lib too (Daniel Veillard),
1392	Add "whereis" command to xmllint shell (Ryan),
1393	Improve xmllint shell (Ryan),
1394	add function xmlTextReaderRelaxNGValidateCtxt() (Noam Postavsky),
1395	Add --system support to autogen.sh (Daniel Veillard),
1396	Add hash randomization to hash and dict structures (Daniel Veillard),
1397	included xzlib in dist (Anders F Bjorklund),
1398	move xz/lzma helpers to separate included files (Anders F Bjorklund),
1399	add generated devhelp files (Anders F Bjorklund),
1400	add XML_WITH_LZMA to api (Anders F Bjorklund),
1401	autogen.sh: Honor NOCONFIGURE environment variable (Colin Walters),
1402	Improve the error report on undefined REFs (Daniel Veillard),
1403	Add exception for new W3C PI xml-model (Daniel Veillard),
1404	Add options to ignore the internal encoding (Daniel Veillard),
1405	testapi: use the right type for the check (Stefan Kost),
1406	various: handle return values of write calls (Stefan Kost),
1407	testWriter: xmlTextWriterWriteFormatElement wants an int instead of a long int (Stefan Kost),
1408	runxmlconf: update to latest testsuite version (Stefan Kost),
1409	configure: add -Wno-long-long to CFLAGS (Stefan Kost),
1410	configure: support silent automake rules if possible (Stefan Kost),
1411	xmlmemory: add a cast as size_t has no portable printf modifier (Stefan Kost),
1412	__xmlRaiseError: remove redundant schannel initialization (Dmitry V. Levin),
1413	__xmlRaiseError: do cheap code check early (Dmitry V. Levin)
1414	</li>
1415
1416	<li>Cleanups:
1417	Cleanups before 2.8.0-rc2 (Daniel Veillard),
1418	Avoid an extra operation (Daniel Veillard),
1419	Remove vestigial de-ANSI-fication support. (Javier Jardón),
1420	autogen.sh: Fix typo (Javier Jardón),
1421	Do not use unsigned but unsigned int (Daniel Veillard),
1422	Remove two references to u_short (Daniel Veillard),
1423	Fix -Wempty-body warning from clang (Nico Weber),
1424	Cleanups of lzma support (Daniel Veillard),
1425	Augment the list of ignored files (Daniel Veillard),
1426	python: remove unused variable (Stefan Kost),
1427	python: flag two unused args (Stefan Kost),
1428	configure: acconfig.h is deprecated since autoconf-2.50 (Stefan Kost),
1429	xpath: remove unused variable (Stefan Kost)
1430	</li>
1431	</ul>
1432	<h3>2.7.8: Nov 4 2010</h3>
1433	<ul>
1434	<li> Features:
1435	480323 add code to plug in ICU converters by default (Giuseppe Iuculano),
1436	Add xmlSaveOption XML_SAVE_WSNONSIG (Adam Spragg)
1437	</li>
1438	<li> Documentation:
1439	Fix devhelp documentation installation (Mike Hommey),
1440	Fix web site encoding problems (Daniel Veillard),
1441	Fix a couple of typo in HTML parser error messages (Michael Day),
1442	Forgot to update the news page for 0.7.7 (Daniel Veillard)
1443	</li>
1444	<li> Portability:
1445	607273 Fix python detection on MSys/Windows (LRN),
1446	614087 Fix Socket API usage to allow Windows64 compilation (Ozkan Sezer),
1447	Fix compilation with Clang (Koop Mast),
1448	Fix Win32 build (Rob Richards)
1449	</li>
1450	<li> Bug Fixes:
1451	595789 fix a remaining potential Solaris problem (Daniel Veillard),
1452	617468 fix progressive HTML parsing with style using "'" (Denis Pauk),
1453	616478 Fix xmllint shell write command (Gwenn Kahz),
1454	614005 Possible erroneous HTML parsing on unterminated script (Pierre Belzile),
1455	627987 Fix XSD IDC errors in imported schemas (Jim Panetta),
1456	629325 XPath rounding errors first cleanup (Phil Shafer),
1457	630140 fix iso995x encoding error (Daniel Veillard),
1458	make sure htmlCtxtReset do reset the disableSAX field (Daniel Veillard),
1459	Fix a change of semantic on XPath preceding and following axis (Daniel Veillard),
1460	Fix a potential segfault due to weak symbols on pthreads (Mike Hommey),
1461	Fix a leak in XPath compilation (Daniel Veillard),
1462	Fix the semantic of XPath axis for namespace/attribute context nodes (Daniel Veillard),
1463	Avoid a descriptor leak in catalog loading code (Carlo Bramini),
1464	Fix a small bug in XPath evaluation code (Marius Wachtler),
1465	Fix handling of XML-1.0 XML namespace declaration (Daniel Veillard),
1466	Fix errors in XSD double validation check (Csaba Raduly),
1467	Fix handling of apos in URIs (Daniel Veillard),
1468	xmlTextReaderReadOuterXml should handle DTD (Rob Richards),
1469	Autogen.sh needs to create m4 directory (Rob Richards)
1470	</li>
1471	<li> Improvements:
1472	606592 update language ID parser to RFC 5646 (Daniel Veillard),
1473	Sort python generated stubs (Mike Hommey),
1474	Add an HTML parser option to avoid a default doctype (Daniel Veillard)
1475	</li>
1476	<li> Cleanups:
1477	618831 don't ship generated files in git (Adrian Bunk),
1478	Switch from the obsolete mkinstalldirs to AC_PROG_MKDIR_P (Adrian Bunk),
1479	Various cleanups on encoding handling (Daniel Veillard),
1480	Fix xmllint to use format=1 for default formatting (Adam Spragg),
1481	Force _xmlSaveCtxt.format to be 0 or 1 (Adam Spragg),
1482	Cleanup encoding pointer comparison (Nikolay Sivov),
1483	Small code cleanup on previous patch (Daniel Veillard)
1484	</li>
1485	</ul>
1486	<h3>2.7.7: Mar 15 2010</h3>
1487	<ul>
1488	<li> Improvements:
1489	Adding a --xpath option to xmllint (Daniel Veillard),
1490	Make HTML parser non-recursive (Eugene Pimenov)
1491	</li>
1492	<li> Portability:
1493	relaxng.c: cast to allow compilation with sun studio 11 (Ben Walton),
1494	Fix build failure on Sparc solaris (Roumen Petrov),
1495	use autoreconf in autogen.sh (Daniel Veillard),
1496	Fix build with mingw (Roumen Petrov),
1497	Upgrade some of the configure and autogen (Daniel Veillard),
1498	Fix relaxNG tests in runtest for Windows runtest.c: initialize ret (Rob Richards),
1499	Fix a const warning in xmlNodeSetBase (Martin Trappel),
1500	Fix python generator to not use deprecated xmllib (Daniel Veillard),
1501	Update some automake files (Daniel Veillard),
1502	598785 Fix nanohttp on Windows (spadix)
1503	</li>
1504	<li> Bug Fixes:
1505	libxml violates the zlib interface and crashes (Mark Adler),
1506	Fix broken escape behaviour in regexp ranges (Daniel Veillard),
1507	Fix missing win32 libraries in libxml-2.0.pc (Volker Grabsch),
1508	Fix detection of python linker flags (Daniel Macks),
1509	fix build error in libxml2/python (Paul Smith),
1510	ChunkParser: Incorrect decoding of small xml files (Raul Hudea),
1511	htmlCheckEncoding doesn't update input-end after shrink (Eugene Pimenov),
1512	Fix a missing #ifdef (Daniel Veillard),
1513	Fix encoding selection for xmlParseInNodeContext (Daniel Veillard),
1514	xmlPreviousElementSibling mistake (FranÃ§ois Delyon),
1515	608773 add a missing check in xmlGROW (Daniel Veillard),
1516	Fix xmlParseInNodeContext for HTML content (Daniel Veillard),
1517	Fix lost namespace when copying node * tree.c: reconcile namespace if not found (Rob Richards),
1518	Fix some missing commas in HTML element lists (Eugene Pimenov),
1519	Correct variable type to unsigned (Nikolay Sivov),
1520	Recognize ID attribute in HTML without DOCTYPE (Daniel Veillard),
1521	Fix memory leak in xmlXPathEvalExpression() (Martin),
1522	Fix an init bug in global.c (Kai Henning),
1523	Fix xmlNodeSetBase() comment (Daniel Veillard),
1524	Fix broken escape behaviour in regexp ranges (Daniel Veillard),
1525	Don't give default HTML boolean attribute values in parser (Daniel Veillard),
1526	xmlCtxtResetLastError should reset ctxt-errNo (Daniel Veillard)
1527	</li>
1528	<li> Cleanups:
1529	Cleanup a couple of weirdness in HTML parser (Eugene Pimenov)
1530	</li>
1531	</ul>
1532	<h3>2.7.6: Oct 6 2009</h3>
1533	<ul>
1534	<li> Bug Fixes:
1535	Restore thread support in default configuration (Andrew W. Nosenko),
1536	URI with no path parsing problem (Daniel Veillard),
1537	Minor patch for conditional defines in threads.c (Eric Zurcher)
1538	</li>
1539	</ul>
1540	<h3>2.7.5: Sep 24 2009</h3>
1541	<ul>
1542	<li> Bug Fixes:
1543	Restore behavior of --with-threads without argument (Andrew W. Nosenko),
1544	Fix memory leak when doc is NULL (Rob Richards),
1545	595792 fixing a RelaxNG bug introduced in 2.7.4 (Daniel Veillard),
1546	Fix a Relaxng bug raised by libvirt test suite (Daniel Veillard),
1547	Fix a parsing problem with little data at startup (Daniel Veillard),
1548	link python module with python library (Frederic Crozat),
1549	594874 Forgot an fclose in xmllint (Daniel Veillard)
1550	</li>
1551	<li> Cleanup:
1552	Adding symbols.xml to EXTRA_DIST (Daniel Veillard)
1553	</li>
1554	</ul>
1555	<h3>2.7.4: Sep 10 2009</h3>
1556	<ul>
1557	<li>Improvements:
1558	Switch to GIT (GNOME),
1559	Add symbol versioning to libxml2 shared libs (Daniel Veillard)
1560	</li>
1561	<li>Portability:
1562	593857 try to work around thread pbm MinGW 4.4 (Daniel Veillard),
1563	594250 rename ATTRIBUTE_ALLOC_SIZE to avoid clashes (Daniel Veillard),
1564	Fix Windows build * relaxng.c: fix windows build (Rob Richards),
1565	Fix the globals.h to use XMLPUBFUN (Paul Smith),
1566	Problem with extern extern in header (Daniel Veillard),
1567	Add -lnetwork for compiling on Haiku (Scott McCreary),
1568	Runtest portability patch for Solaris (Tim Rice),
1569	Small patch to accomodate the Haiku OS (Scott McCreary),
1570	584605 package VxWorks folder in the distribution (Daniel Veillard),
1571	574017 Realloc too expensive on most platform (Daniel Veillard),
1572	Fix windows build (Rob Richards),
1573	545579 doesn't compile without schema support (Daniel Veillard),
1574	xmllint use xmlGetNodePath when not compiled in (Daniel Veillard),
1575	Try to avoid __imp__xmlFree link trouble on msys (Daniel Veillard),
1576	Allow to select the threading system on Windows (LRN),
1577	Fix Solaris binary links, cleanups (Daniel Veillard),
1578	Bug 571059 â MSVC doesn't work with the bakefile (Intron),
1579	fix ATTRIBUTE_PRINTF header clash (Belgabor and Mike Hommey),
1580	fixes for Borland/CodeGear/Embarcadero compilers (Eric Zurcher)
1581	</li>
1582	<li>Documentation:
1583	544910 typo: "renciliateNs" (Leonid Evdokimov),
1584	Add VxWorks to list of OSes (Daniel Veillard),
1585	Regenerate the documentation and update for git (Daniel Veillard),
1586	560524 Â¿ xmlTextReaderLocalName description (Daniel Veillard),
1587	Added sponsoring by AOE media for the server (Daniel Veillard),
1588	updated URLs for GNOME (Vincent Lefevre),
1589	more warnings about xmlCleanupThreads and xmlCleanupParser (Daniel Veillard)
1590	</li>
1591	<li>Bug fixes:
1592	594514 memory leaks - duplicate initialization (MOD),
1593	Wrong block opening in htmlNodeDumpOutputInternal (Daniel Veillard),
1594	492317 Fix Relax-NG validation problems (Daniel Veillard),
1595	558452 fight with reg test and error report (Daniel Veillard),
1596	558452 RNG compilation of optional multiple child (Daniel Veillard),
1597	579746 XSD validation not correct / nilable groups (Daniel Veillard),
1598	502960 provide namespace stack when parsing entity (Daniel Veillard),
1599	566012 part 2 fix regresion tests and push mode (Daniel Veillard),
1600	566012 autodetected encoding and encoding conflict (Daniel Veillard),
1601	584220 xpointer(/) and xinclude problems (Daniel Veillard),
1602	587663 Incorrect Attribute-Value Normalization (Daniel Veillard),
1603	444994 HTML chunked failure for attribute with <> (Daniel Veillard),
1604	Fix end of buffer char being split in XML parser (Daniel Veillard),
1605	Non ASCII character may be split at buffer end (Adiel Mittmann),
1606	440226 Add xmlXIncludeProcessTreeFlagsData API (Stefan Behnel),
1607	572129 speed up parsing of large HTML text nodes (Markus Kull),
1608	Fix HTML parsing with 0 character in CDATA (Daniel Veillard),
1609	Fix SetGenericErrorFunc and SetStructured clash (Wang Lam),
1610	566012 Incomplete EBCDIC parsing support (Martin Kogler),
1611	541335 HTML avoid creating 2 head or 2 body element (Daniel Veillard),
1612	541237 error correcting missing end tags in HTML (Daniel Veillard),
1613	583439 missing line numbers in push mode (Daniel Veillard),
1614	587867 xmllint --html --xmlout serializing as HTML (Daniel Veillard),
1615	559501 avoid select and use poll for nanohttp (Raphael Prevost),
1616	559410 - Regexp bug on (...)? constructs (Daniel Veillard),
1617	Fix a small problem on previous HTML parser patch (Daniel Veillard),
1618	592430 - HTML parser runs into endless loop (Daniel Veillard),
1619	447899 potential double free in xmlFreeTextReader (Daniel Veillard),
1620	446613 small validation bug mixed content with NS (Daniel Veillard),
1621	Fix the problem of revalidating a doc with RNG (Daniel Veillard),
1622	Fix xmlKeepBlanksDefault to not break indent (Nick Wellnhofer),
1623	512131 refs from externalRef part need to be added (Daniel Veillard),
1624	512131 crash in xmlRelaxNGValidateFullElement (Daniel Veillard),
1625	588441 allow '.' in HTML Names even if invalid (Daniel Veillard),
1626	582913 Fix htmlSetMetaEncoding() to be nicer (Daniel Veillard),
1627	579317 Try to find the HTML encoding information (Daniel Veillard),
1628	575875 don't output charset=html (Daniel Veillard),
1629	571271 fix semantic of xsd:all with minOccurs=0 (Daniel Veillard),
1630	570702 fix a bug in regexp determinism checking (Daniel Veillard),
1631	567619 xmlValidateNotationUse missing param test (Daniel Veillard),
1632	574393 Â¿ utf-8 filename magic for compressed files (Hans Breuer),
1633	Fix a couple of problems in the parser (Daniel Veillard),
1634	585505 Â¿ Document ids and refs populated by XSD (Wayne Jensen),
1635	582906 XSD validating multiple imports of the same schema (Jason Childs),
1636	Bug 582887 Â¿ problems validating complex schemas (Jason Childs),
1637	Bug 579729 Â¿ fix XSD schemas parsing crash (Miroslav Bajtos),
1638	576368 Â¿ htmlChunkParser with special attributes (Jiri Netolicky),
1639	Bug 565747 Â¿ relax anyURI data character checking (Vincent Lefevre),
1640	Preserve attributes of include start on tree copy (Petr Pajas),
1641	Skip silently unrecognized XPointer schemes (Jakub Wilk),
1642	Fix leak on SAX1, xmllint --sax1 option and debug (Daniel Veillard),
1643	potential NULL dereference on non-glibc (Jim Meyering),
1644	Fix an XSD validation crash (Daniel Veillard),
1645	Fix a regression in streaming entities support (Daniel Veillard),
1646	Fix a couple of ABI issues with C14N 1.1 (Aleksey Sanin),
1647	Aleksey Sanin support for c14n 1.1 (Aleksey Sanin),
1648	reader bug fix with entities (Daniel Veillard),
1649	use options from current parser ctxt for external entities (Rob Richards),
1650	581612 use %s to printf strings (Christian Persch),
1651	584605 change the threading initialization sequence (Igor Novoseltsev),
1652	580705 keep line numbers in HTML parser (Aaron Patterson),
1653	581803 broken HTML table attributes init (Roland Steiner),
1654	do not set error code in xmlNsWarn (Rob Richards),
1655	564217 fix structured error handling problems,
1656	reuse options from current parser for entities (Rob Richards),
1657	xmlXPathRegisterNs should not allow enpty prefixes (Daniel Veillard),
1658	add a missing check in xmlAddSibling (Kris Breuker),
1659	avoid leaks on errors (Jinmei Tatuya)
1660	</li>
1661	<li>Cleanup:
1662	Chasing dead assignments reported by clang-scan (Daniel Veillard),
1663	A few more safety cleanup raised by scan (Daniel Veillard),
1664	Fixing assorted potential problems raised by scan (Daniel Veillard),
1665	Potential uninitialized arguments raised by scan (Daniel Veillard),
1666	Fix a bunch of scan 'dead increments' and cleanup (Daniel Veillard),
1667	Remove a pedantic warning (Daniel Veillard),
1668	555833 always use rm -f in uninstall-local (Daniel Veillard),
1669	542394 xmlRegisterOutputCallbacks MAX_INPUT_CALLBACK (Daniel Veillard),
1670	Autoregenerate libxml2.syms automated checkings (Daniel Veillard),
1671	Make xmlRecoverDoc const (Martin Trappel) (Daniel Veillard),
1672	Both args of xmlStrcasestr are const (Daniel Veillard),
1673	hide the nbParse* variables used for debugging (Mike Hommey),
1674	570806 changed include of config.h (William M. Brack),
1675	cleanups and error reports when xmlTextWriterVSprintf fails (Jinmei Tatuya)
1676	</li>
1677	</ul>
1678	<h3>2.7.3: Jan 18 2009</h3>
1679	<ul>
1680	<li>Build fix: fix build when HTML support is not included.</li>
1681	<li>Bug fixes: avoid memory overflow in gigantic text nodes,
1682	indentation problem on the writed (Rob Richards),
1683	xmlAddChildList pointer problem (Rob Richards and Kevin Milburn),
1684	xmlAddChild problem with attribute (Rob Richards and Kris Breuker),
1685	avoid a memory leak in an edge case (Daniel Zimmermann),
1686	deallocate some pthread data (Alex Ott).</li>
1687	<li>Improvements: configure option to avoid rebuilding docs (Adrian Bunk),
1688	limit text nodes to 10MB max by default, add element traversal
1689	APIs, add a parser option to enable pre 2.7 SAX behavior (Rob Richards),
1690	add gcc malloc checking (Marcus Meissner), add gcc printf like functions
1691	parameters checking (Marcus Meissner).</li>
1692	</ul>
1693	<h3>2.7.2: Oct 3 2008</h3>
1694	<ul>
1695	<li>Portability fix: fix solaris compilation problem, fix compilation
1696	if XPath is not configured in</li>
1697	<li>Bug fixes: nasty entity bug introduced in 2.7.0, restore old behaviour
1698	when saving an HTML doc with an xml dump function, HTML UTF-8 parsing
1699	bug, fix reader custom error handlers (Riccardo Scussat)
1700	<li>Improvement: xmlSave options for more flexibility to save as
1701	XML/HTML/XHTML, handle leading BOM in HTML documents</li>
1702	</ul>
1703
1704	<h3>2.7.1: Sep 1 2008</h3>
1705	<ul>
1706	<li>Portability fix: Borland C fix (Moritz Both)</li>
1707	<li>Bug fixes: python serialization wrappers, XPath QName corner
1708	case handking and leaks (Martin)</li>
1709	<li>Improvement: extend the xmlSave to handle HTML documents and trees</li>
1710	<li>Cleanup: python serialization wrappers</li>
1711	</ul>
1712
1713	<h3>2.7.0: Aug 30 2008</h3>
1714	<ul>
1715	<li>Documentation: switch ChangeLog to UTF-8, improve mutithreads and
1716	xmlParserCleanup docs</li>
1717	<li>Portability fixes: Older Win32 platforms (Rob Richards), MSVC
1718	porting fix (Rob Richards), Mac OS X regression tests (Sven Herzberg),
1719	non GNUCC builds (Rob Richards), compilation on Haiku (Andreas FÃ¤rber)
1720	</li>
1721	<li>Bug fixes: various realloc problems (Ashwin), potential double-free
1722	(Ashwin), regexp crash, icrash with invalid whitespace facets (Rob
1723	Richards), pattern fix when streaming (William Brack), various XML
1724	parsing and validation fixes based on the W3C regression tests, reader
1725	tree skipping function fix (Ashwin), Schemas regexps escaping fix
1726	(Volker Grabsch), handling of entity push errors (Ashwin), fix a slowdown
1727	when encoder cant serialize characters on output</li>
1728	<li>Code cleanup: compilation fix without the reader, without the output
1729	(Robert Schwebel), python whitespace (Martin), many space/tabs cleanups,
1730	serious cleanup of the entity handling code</li>
1731	<li>Improvement: switch parser to XML-1.0 5th edition, add parsing flags
1732	for old versions, switch URI parsing to RFC 3986,
1733	add xmlSchemaValidCtxtGetParserCtxt (Holger Kaelberer),
1734	new hashing functions for dictionnaries (based on Stefan Behnel work),
1735	improve handling of misplaced html/head/body in HTML parser, better
1736	regression test tools and code coverage display, better algorithms
1737	to detect various versions of the billion laughts attacks, make
1738	arbitrary parser limits avoidable as a parser option</li>
1739	</ul>
1740	<h3>2.6.32: Apr 8 2008</h3>
1741	<ul>
1742	<li>Documentation: returning heap memory to kernel (Wolfram Sang),
1743	trying to clarify xmlCleanupParser() use, xmlXPathContext improvement
1744	(Jack Jansen), improve the Recover functions documentation,
1745	XmlNodeType doc link fix (Martijn Arts)</li>
1746	<li>Bug fixes: internal subset memory leak (Ashwin), avoid problem with
1747	paths starting with // (Petr Sumbera), streaming XSD validation callback
1748	patches (Ashwin), fix redirection on port other than 80 (William Brack),
1749	SAX2 leak (Ashwin), XInclude fragment of own document (Chris Ryan),
1750	regexp bug with '.' (Andrew Tosh), flush the writer at the end of the
1751	document (Alfred Mickautsch), output I/O bug fix (William Brack),
1752	writer CDATA output after a text node (Alex Khesin), UTF-16 encoding
1753	detection (William Brack), fix handling of empty CDATA nodes for Safari
1754	team, python binding problem with namespace nodes, improve HTML parsing
1755	(Arnold Hendriks), regexp automata build bug, memory leak fix (Vasily
1756	Chekalkin), XSD test crash, weird system parameter entity parsing problem,
1757	allow save to file:///X:/ windows paths, various attribute normalisation
1758	problems, externalSubsetSplit fix (Ashwin), attribute redefinition in
1759	the DTD (Ashwin), fix in char ref parsing check (Alex Khesin), many
1760	out of memory handling fixes (Ashwin), XPath out of memory handling fixes
1761	(Alvaro Herrera), various realloc problems (Ashwin), UCS4 encoding
1762	conversion buffer size (Christian Fruth), problems with EatName
1763	functions on memory errors, BOM handling in external parsed entities
1764	(Mark Rowe)</li>
1765	<li>Code cleanup: fix build under VS 2008 (David Wimsey), remove useless
1766	mutex in xmlDict (Florent Guilian), Mingw32 compilation fix (Carlo
1767	Bramini), Win and MacOS EOL cleanups (Florent Guiliani), iconv need
1768	a const detection (Roumen Petrov), simplify xmlSetProp (Julien Charbon),
1769	cross compilation fixes for Mingw (Roumen Petrov), SCO Openserver build
1770	fix (Florent Guiliani), iconv uses const on Win32 (Rob Richards),
1771	duplicate code removal (Ashwin), missing malloc test and error reports
1772	(Ashwin), VMS makefile fix (Tycho Hilhorst)</li>
1773	<li>improvements: better plug of schematron in the normal error handling
1774	(Tobias Minich)</li>
1775	</ul>
1776
1777	<h3>2.6.31: Jan 11 2008</h3>
1778	<ul>
1779	<li>Security fix: missing of checks in UTF-8 parsing</li>
1780	<li>Bug fixes: regexp bug, dump attribute from XHTML document, fix
1781	xmlFree(NULL) to not crash in debug mode, Schematron parsing crash
1782	(Rob Richards), global lock free on Windows (Marc-Antoine Ruel),
1783	XSD crash due to double free (Rob Richards), indentation fix in
1784	xmlTextWriterFullEndElement (Felipe Pena), error in attribute type
1785	parsing if attribute redeclared, avoid crash in hash list scanner if
1786	deleting elements, column counter bug fix (Christian Schmidt),
1787	HTML embed element saving fix (Stefan Behnel), avoid -L/usr/lib
1788	output from xml2-config (Fred Crozat), avoid an xmllint crash
1789	(Stefan Kost), don't stop HTML parsing on out of range chars.
1790	</li>
1791	<li>Code cleanup: fix open() call third argument, regexp cut'n paste
1792	copy error, unused variable in __xmlGlobalInitMutexLock (Hannes Eder),
1793	some make distcheck realted fixes (John Carr)</li>
1794	<li>Improvements: HTTP Header: includes port number (William Brack),
1795	testURI --debug option, </li>
1796	</ul>
1797	<h3>2.6.30: Aug 23 2007</h3>
1798	<ul>
1799	<li>Portability: Solaris crash on error handling, windows path fixes
1800	(Roland Schwarz and Rob Richards), mingw build (Roland Schwarz)</li>
1801	<li>Bugfixes: xmlXPathNodeSetSort problem (William Brack), leak when
1802	reusing a writer for a new document (Dodji Seketeli), Schemas
1803	xsi:nil handling patch (Frank Gross), relative URI build problem
1804	(Patrik Fimml), crash in xmlDocFormatDump, invalid char in comment
1805	detection bug, fix disparity with xmlSAXUserParseMemory, automata
1806	generation for complex regexp counts problems, Schemas IDC import
1807	problems (Frank Gross), xpath predicate evailation error handling
1808	(William Brack)</li>
1809	</ul>
1810	<h3>2.6.29: Jun 12 2007</h3>
1811	<ul>
1812	<li>Portability: patches from Andreas Stricke for WinCEi,
1813	fix compilation warnings (William Brack), avoid warnings on Apple OS/X
1814	(Wendy Doyle and Mark Rowe), Windows compilation and threading
1815	improvements (Rob Richards), compilation against old Python versions,
1816	new GNU tar changes (Ryan Hill)</li>
1817	<li>Documentation: xmlURIUnescapeString comment, </li>
1818	<li>Bugfixes: xmlBufferAdd problem (Richard Jones), 'make valgrind'
1819	flag fix (Richard Jones), regexp interpretation of \,
1820	htmlCreateDocParserCtxt (Jean-Daniel Dupas), configure.in
1821	typo (Bjorn Reese), entity content failure, xmlListAppend() fix
1822	(Georges-AndrÃ© Silber), XPath number serialization (William Brack),
1823	nanohttp gzipped stream fix (William Brack and Alex Cornejo),
1824	xmlCharEncFirstLine typo (Mark Rowe), uri bug (FranÃ§ois Delyon),
1825	XPath string value of PI nodes (William Brack), XPath node set
1826	sorting bugs (William Brack), avoid outputting namespace decl
1827	dups in the writer (Rob Richards), xmlCtxtReset bug, UTF-8 encoding
1828	error handling, recustion on next in catalogs, fix a Relax-NG crash,
1829	workaround wrong file: URIs, htmlNodeDumpFormatOutput on attributes,
1830	invalid character in attribute detection bug, big comments before
1831	internal subset streaming bug, HTML parsing of attributes with : in
1832	the name, IDness of name in HTML (Dagfinn I. MannsÃ¥ker) </li>
1833	<li>Improvement: keep URI query parts in raw form (Richard Jones),
1834	embed tag support in HTML (Michael Day) </li>
1835	</ul>
1836
1837	<h3>2.6.28: Apr 17 2007</h3>
1838	<ul>
1839	<li>Documentation: comment fixes (Markus Keim), xpath comments fixes too
1840	(James Dennett)</li>
1841	<li>Bug fixes: XPath bug (William Brack), HTML parser autoclose stack usage
1842	(Usamah Malik), various regexp bug fixes (DV and William), path conversion
1843	on Windows (Igor Zlatkovic), htmlCtxtReset fix (Michael Day), XPath
1844	principal node of axis bug, HTML serialization of some codepoint
1845	(Steven Rainwater), user data propagation in XInclude (Michael Day),
1846	standalone and XML decl detection (Michael Day), Python id ouptut
1847	for some id, fix the big python string memory leak, URI parsing fixes
1848	(StÃ©phane Bidoul and William), long comments parsing bug (William),
1849	concurrent threads initialization (Ted Phelps), invalid char
1850	in text XInclude (William), XPath memory leak (William), tab in
1851	python problems (Andreas Hanke), XPath node comparison error
1852	(Oleg Paraschenko), cleanup patch for reader (Julien Reichel),
1853	XML Schemas attribute group (William), HTML parsing problem (William),
1854	fix char 0x2d in regexps (William), regexp quantifier range with
1855	min occurs of 0 (William), HTML script/style parsing (Mike Day)</li>
1856	<li>Improvement: make xmlTextReaderSetup() public</li>
1857	<li>Compilation and postability: fix a missing include problem (William),
1858	__ss_familly on AIX again (BjÃ¶rn Wiberg), compilation without zlib
1859	(Michael Day), catalog patch for Win32 (Christian Ehrlicher),
1860	Windows CE fixes (Andreas Stricke)</li>
1861	<li>Various CVS to SVN infrastructure changes</li>
1862	</ul>
1863	<h3>2.6.27: Oct 25 2006</h3>
1864	<ul>
1865	<li>Portability fixes: file names on windows (Roland Schwingel,
1866	Emelyanov Alexey), windows compile fixup (Rob Richards),
1867	AIX iconv() is apparently case sensitive</li>
1868	<li>improvements: Python XPath types mapping (Nic Ferrier), XPath optimization
1869	(Kasimier), add xmlXPathCompiledEvalToBoolean (Kasimier), Python node
1870	equality and comparison (Andreas Pakulat), xmlXPathCollectAndTest
1871	improvememt (Kasimier), expose if library was compiled with zlib
1872	support (Andrew Nosenko), cache for xmlSchemaIDCMatcher structs
1873	(Kasimier), xmlTextConcat should work with comments and PIs (Rob
1874	Richards), export htmlNewParserCtxt needed by Michael Day, refactoring
1875	of catalog entity loaders (Michael Day), add XPointer support to
1876	python bindings (Ross Reedstrom, Brian West and Stefan Anca),
1877	try to sort out most file path to URI conversions and xmlPathToUri,
1878	add --html --memory case to xmllint</li>
1879	<li>building fix: fix --with-minimum (Felipe Contreras), VMS fix,
1880	const'ification of HTML parser structures (Matthias Clasen),
1881	portability fix (Emelyanov Alexey), wget autodetection (Peter
1882	Breitenlohner), remove the build path recorded in the python
1883	shared module, separate library flags for shared and static builds
1884	(Mikhail Zabaluev), fix --with-minimum --with-sax1 builds, fix
1885	--with-minimum --with-schemas builds</li>
1886	<li>bug fix: xmlGetNodePath fix (Kasimier), xmlDOMWrapAdoptNode and
1887	attribute (Kasimier), crash when using the recover mode,
1888	xmlXPathEvalExpr problem (Kasimier), xmlXPathCompExprAdd bug (Kasimier),
1889	missing destry in xmlFreeRMutex (Andrew Nosenko), XML Schemas fixes
1890	(Kasimier), warning on entities processing, XHTML script and style
1891	serialization (Kasimier), python generator for long types, bug in
1892	xmlSchemaClearValidCtxt (Bertrand Fritsch), xmlSchemaXPathEvaluate
1893	allocation bug (Marton Illes), error message end of line (Rob Richards),
1894	fix attribute serialization in writer (Rob Richards), PHP4 DTD validation
1895	crasher, parser safety patch (Ben Darnell), _private context propagation
1896	when parsing entities (with Michael Day), fix entities behaviour when
1897	using SAX, URI to file path fix (Mikhail Zabaluev), disapearing validity
1898	context, arg error in SAX callback (Mike Hommey), fix mixed-content
1899	autodetect when using --noblanks, fix xmlIOParseDTD error handling,
1900	fix bug in xmlSplitQName on special Names, fix Relax-NG element content
1901	validation bug, fix xmlReconciliateNs bug, fix potential attribute
1902	XML parsing bug, fix line/column accounting in XML parser, chunking bug
1903	in the HTML parser on script, try to detect obviously buggy HTML
1904	meta encoding indications, bugs with encoding BOM and xmlSaveDoc,
1905	HTML entities in attributes parsing, HTML minimized attribute values,
1906	htmlReadDoc and htmlReadIO were broken, error handling bug in
1907	xmlXPathEvalExpression (Olaf Walkowiak), fix a problem in
1908	htmlCtxtUseOptions, xmlNewInputFromFile could leak (Marius Konitzer),
1909	bug on misformed SSD regexps (Christopher Boumenot)
1910	</li>
1911	<li>documentation: warning about XML_PARSE_COMPACT (Kasimier Buchcik),
1912	fix xmlXPathCastToString documentation, improve man pages for
1913	xmllitn and xmlcatalog (Daniel Leidert), fixed comments of a few
1914	functions</li>
1915	</ul>
1916	<h3>2.6.26: Jun 6 2006</h3>
1917	<ul>
1918	<li>portability fixes: Python detection (Joseph Sacco), compilation
1919	error(William Brack and Graham Bennett), LynxOS patch (Olli Savia)</li>
1920	<li>bug fixes: encoding buffer problem, mix of code and data in
1921	xmlIO.c(Kjartan Maraas), entities in XSD validation (Kasimier Buchcik),
1922	variousXSD validation fixes (Kasimier), memory leak in pattern (Rob
1923	Richards andKasimier), attribute with colon in name (Rob Richards), XPath
1924	leak inerror reporting (Aleksey Sanin), XInclude text include of
1925	selfdocument.</li>
1926	<li>improvements: Xpath optimizations (Kasimier), XPath object
1927	cache(Kasimier)</li>
1928	</ul>
1929
1930	<h3>2.6.25: Jun 6 2006:</h3>
1931
1932	<p>Do not use or package 2.6.25</p>
1933
1934	<h3>2.6.24: Apr 28 2006</h3>
1935	<ul>
1936	<li>Portability fixes: configure on Windows, testapi compile on windows
1937	(Kasimier Buchcik, venkat naidu), Borland C++ 6 compile (Eric Zurcher),
1938	HP-UX compiler workaround (Rick Jones), xml2-config bugfix, gcc-4.1
1939	cleanups, Python detection scheme (Joseph Sacco), UTF-8 file paths on
1940	Windows (Roland Schwingel).
1941	</li>
1942	<li>Improvements: xmlDOMWrapReconcileNamespaces xmlDOMWrapCloneNode (Kasimier
1943	Buchcik), XML catalog debugging (Rick Jones), update to Unicode 4.01.</li>
1944	<li>Bug fixes: xmlParseChunk() problem in 2.6.23, xmlParseInNodeContext()
1945	on HTML docs, URI behaviour on Windows (Rob Richards), comment streaming
1946	bug, xmlParseComment (with William Brack), regexp bug fixes (DV &
1947	Youri Golovanov), xmlGetNodePath on text/CDATA (Kasimier),
1948	one Relax-NG interleave bug, xmllint --path and --valid,
1949	XSD bugfixes (Kasimier), remove debug
1950	left in Python bindings (Nic Ferrier), xmlCatalogAdd bug (Martin Cole),
1951	xmlSetProp fixes (Rob Richards), HTML IDness (Rob Richards), a large
1952	number of cleanups and small fixes based on Coverity reports, bug
1953	in character ranges, Unicode tables const (Aivars Kalvans), schemas
1954	fix (Stefan Kost), xmlRelaxNGParse error deallocation,
1955	xmlSchemaAddSchemaDoc error deallocation, error handling on unallowed
1956	code point, ixmllint --nonet to never reach the net (Gary Coady),
1957	line break in writer after end PI (Jason Viers). </li>
1958	<li>Documentation: man pages updates and cleanups (Daniel Leidert).</li>
1959	<li>New features: Relax NG structure error handlers.</li>
1960	</ul>
1961
1962	<h3>2.6.23: Jan 5 2006</h3>
1963	<ul>
1964	<li>portability fixes: Windows (Rob Richards), getaddrinfo on Windows
1965	(Kolja Nowak, Rob Richards), icc warnings (Kjartan Maraas),
1966	--with-minimum compilation fixes (William Brack), error case handling fix
1967	on Solaris (Albert Chin), don't use 'list' as parameter name reported by
1968	Samuel Diaz Garcia, more old Unices portability fixes (Albert Chin),
1969	MinGW compilation (Mark Junker), HP-UX compiler warnings (Rick
1970	Jones),</li>
1971	<li>code cleanup: xmlReportError (Adrian Mouat), remove xmlBufferClose
1972	(Geert Jansen), unreachable code (Oleksandr Kononenko), refactoring
1973	parsing code (Bjorn Reese)</li>
1974	<li>bug fixes: xmlBuildRelativeURI and empty path (William Brack),
1975	combinatory explosion and performances in regexp code, leak in
1976	xmlTextReaderReadString(), xmlStringLenDecodeEntities problem (Massimo
1977	Morara), Identity Constraints bugs and a segfault (Kasimier Buchcik),
1978	XPath pattern based evaluation bugs (DV & Kasimier),
1979	xmlSchemaContentModelDump() memory leak (Kasimier), potential leak in
1980	xmlSchemaCheckCSelectorXPath(), xmlTextWriterVSprintf() misuse of
1981	vsnprintf (William Brack), XHTML serialization fix (Rob Richards), CRLF
1982	split problem (William), issues with non-namespaced attributes in
1983	xmlAddChild() xmlAddNextSibling() and xmlAddPrevSibling() (Rob Richards),
1984	HTML parsing of script, Python must not output to stdout (Nic Ferrier),
1985	exclusive C14N namespace visibility (Aleksey Sanin), XSD dataype
1986	totalDigits bug (Kasimier Buchcik), error handling when writing to an
1987	xmlBuffer (Rob Richards), runtest schemas error not reported (Hisashi
1988	Fujinaka), signed/unsigned problem in date/time code (Albert Chin), fix
1989	XSI driven XSD validation (Kasimier), parsing of xs:decimal (Kasimier),
1990	fix DTD writer output (Rob Richards), leak in xmlTextReaderReadInnerXml
1991	(Gary Coady), regexp bug affecting schemas (Kasimier), configuration of
1992	runtime debugging (Kasimier), xmlNodeBufGetContent bug on entity refs
1993	(Oleksandr Kononenko), xmlRegExecPushString2 bug (Sreeni Nair),
1994	compilation and build fixes (Michael Day), removed dependancies on
1995	xmlSchemaValidError (Kasimier), bug with <xml:foo/>, more XPath
1996	pattern based evaluation fixes (Kasimier)</li>
1997	<li>improvements: XSD Schemas redefinitions/restrictions (Kasimier
1998	Buchcik), node copy checks and fix for attribute (Rob Richards), counted
1999	transition bug in regexps, ctxt->standalone = -2 to indicate no
2000	standalone attribute was found, add xmlSchemaSetParserStructuredErrors()
2001	(Kasimier Buchcik), add xmlTextReaderSchemaValidateCtxt() to API
2002	(Kasimier), handle gzipped HTTP resources (Gary Coady), add
2003	htmlDocDumpMemoryFormat. (Rob Richards),</li>
2004	<li>documentation: typo (Michael Day), libxml man page (Albert Chin), save
2005	function to XML buffer (Geert Jansen), small doc fix (Aron Stansvik),</li>
2006	</ul>
2007
2008	<h3>2.6.22: Sep 12 2005</h3>
2009	<ul>
2010	<li>build fixes: compile without schematron (StÃ©phane Bidoul)</li>
2011	<li>bug fixes: xmlDebugDumpNode on namespace node (Oleg Paraschenko)i,
2012	CDATA push parser bug, xmlElemDump problem with XHTML1 doc,
2013	XML_FEATURE_xxx clash with expat headers renamed XML_WITH_xxx, fix some
2014	output formatting for meta element (Rob Richards), script and style
2015	XHTML1 serialization (David Madore), Attribute derivation fixups in XSD
2016	(Kasimier Buchcik), better IDC error reports (Kasimier Buchcik)</li>
2017	<li>improvements: add XML_SAVE_NO_EMPTY xmlSaveOption (Rob Richards), add
2018	XML_SAVE_NO_XHTML xmlSaveOption, XML Schemas improvements preparing for
2019	derive (Kasimier Buchcik).</li>
2020	<li>documentation: generation of gtk-doc like docs, integration with
2021	devhelp.</li>
2022	</ul>
2023
2024	<h3>2.6.21: Sep 4 2005</h3>
2025	<ul>
2026	<li>build fixes: Cygwin portability fixes (Gerrit P. Haase), calling
2027	convention problems on Windows (Marcus Boerger), cleanups based on Linus'
2028	sparse tool, update of win32/configure.js (Rob Richards), remove warnings
2029	on Windows(Marcus Boerger), compilation without SAX1, detection of the
2030	Python binary, use $GCC inestad of $CC = 'gcc' (Andrew W. Nosenko),
2031	compilation/link with threads and old gcc, compile problem by C370 on
2032	Z/OS,</li>
2033	<li>bug fixes: http_proxy environments (Peter Breitenlohner), HTML UTF-8
2034	bug (Jiri Netolicky), XPath NaN compare bug (William Brack),
2035	htmlParseScript potential bug, Schemas regexp handling of spaces, Base64
2036	Schemas comparisons NIST passes, automata build error xsd:all,
2037	xmlGetNodePath for namespaced attributes (Alexander Pohoyda), xmlSchemas
2038	foreign namespaces handling, XML Schemas facet comparison (Kupriyanov
2039	Anatolij), xmlSchemaPSimpleTypeErr error report (Kasimier Buchcik), xml:
2040	namespace ahndling in Schemas (Kasimier), empty model group in Schemas
2041	(Kasimier), wilcard in Schemas (Kasimier), URI composition (William),
2042	xs:anyType in Schemas (Kasimier), Python resolver emmitting error
2043	messages directly, Python xmlAttr.parent (Jakub Piotr Clapa), trying to
2044	fix the file path/URI conversion, xmlTextReaderGetAttribute fix (Rob
2045	Richards), xmlSchemaFreeAnnot memleak (Kasimier), HTML UTF-8
2046	serialization, streaming XPath, Schemas determinism detection problem,
2047	XInclude bug, Schemas context type (Dean Hill), validation fix (Derek
2048	Poon), xmlTextReaderGetAttribute[Ns] namespaces (Rob Richards), Schemas
2049	type fix (Kuba Nowakowski), UTF-8 parser bug, error in encoding handling,
2050	xmlGetLineNo fixes, bug on entities handling, entity name extraction in
2051	error handling with XInclude, text nodes in HTML body tags (Gary Coady),
2052	xml:id and IDness at the treee level fixes, XPath streaming patterns
2053	bugs.</li>
2054	<li>improvements: structured interfaces for schemas and RNG error reports
2055	(Marcus Boerger), optimization of the char data inner loop parsing
2056	(thanks to Behdad Esfahbod for the idea), schematron validation though
2057	not finished yet, xmlSaveOption to omit XML declaration, keyref match
2058	error reports (Kasimier), formal expression handling code not plugged
2059	yet, more lax mode for the HTML parser, parser XML_PARSE_COMPACT option
2060	for text nodes allocation.</li>
2061	<li>documentation: xmllint man page had --nonet duplicated</li>
2062	</ul>
2063
2064	<h3>2.6.20: Jul 10 2005</h3>
2065	<ul>
2066	<li>build fixes: Windows build (Rob Richards), Mingw compilation (Igor
2067	Zlatkovic), Windows Makefile (Igor), gcc warnings (Kasimier and
2068	[email protected]), use gcc weak references to pthread to avoid the
2069	pthread dependancy on Linux, compilation problem (Steve Nairn), compiling
2070	of subset (Morten Welinder), IPv6/ss_family compilation (William Brack),
2071	compilation when disabling parts of the library, standalone test
2072	distribution.</li>
2073	<li>bug fixes: bug in lang(), memory cleanup on errors (William Brack),
2074	HTTP query strings (Aron Stansvik), memory leak in DTD (William), integer
2075	overflow in XPath (William), nanoftp buffer size, pattern "." apth fixup
2076	(Kasimier), leak in tree reported by Malcolm Rowe, replaceNode patch
2077	(Brent Hendricks), CDATA with NULL content (Mark Vakoc), xml:base fixup
2078	on XInclude (William), pattern fixes (William), attribute bug in
2079	exclusive c14n (Aleksey Sanin), xml:space and xml:lang with SAX2 (Rob
2080	Richards), namespace trouble in complex parsing (Malcolm Rowe), XSD type
2081	QNames fixes (Kasimier), XPath streaming fixups (William), RelaxNG bug
2082	(Rob Richards), Schemas for Schemas fixes (Kasimier), removal of ID (Rob
2083	Richards), a small RelaxNG leak, HTML parsing in push mode bug (James
2084	Bursa), failure to detect UTF-8 parsing bugs in CDATA sections,
2085	areBlanks() heuristic failure, duplicate attributes in DTD bug
2086	(William).</li>
2087	<li>improvements: lot of work on Schemas by Kasimier Buchcik both on
2088	conformance and streaming, Schemas validation messages (Kasimier Buchcik,
2089	Matthew Burgess), namespace removal at the python level (Brent
2090	Hendricks), Update to new Schemas regression tests from W3C/Nist
2091	(Kasimier), xmlSchemaValidateFile() (Kasimier), implementation of
2092	xmlTextReaderReadInnerXml and xmlTextReaderReadOuterXml (James Wert),
2093	standalone test framework and programs, new DOM import APIs
2094	xmlDOMWrapReconcileNamespaces() xmlDOMWrapAdoptNode() and
2095	xmlDOMWrapRemoveNode(), extension of xmllint capabilities for SAX and
2096	Schemas regression tests, xmlStopParser() available in pull mode too,
2097	ienhancement to xmllint --shell namespaces support, Windows port of the
2098	standalone testing tools (Kasimier and William),
2099	xmlSchemaValidateStream() xmlSchemaSAXPlug() and xmlSchemaSAXUnplug() SAX
2100	Schemas APIs, Schemas xmlReader support.</li>
2101	</ul>
2102
2103	<h3>2.6.19: Apr 02 2005</h3>
2104	<ul>
2105	<li>build fixes: drop .la from RPMs, --with-minimum build fix (William
2106	Brack), use XML_SOCKLEN_T instead of SOCKLEN_T because it breaks with AIX
2107	5.3 compiler, fixed elfgcchack.h generation and PLT reduction code on
2108	Linux/ELF/gcc4</li>
2109	<li>bug fixes: schemas type decimal fixups (William Brack), xmmlint return
2110	code (Gerry Murphy), small schemas fixes (Matthew Burgess and GUY
2111	Fabrice), workaround "DAV:" namespace brokeness in c14n (Aleksey Sanin),
2112	segfault in Schemas (Kasimier Buchcik), Schemas attribute validation
2113	(Kasimier), Prop related functions and xmlNewNodeEatName (Rob Richards),
2114	HTML serialization of name attribute on a elements, Python error handlers
2115	leaks and improvement (Brent Hendricks), uninitialized variable in
2116	encoding code, Relax-NG validation bug, potential crash if
2117	gnorableWhitespace is NULL, xmlSAXParseDoc and xmlParseDoc signatures,
2118	switched back to assuming UTF-8 in case no encoding is given at
2119	serialization time</li>
2120	<li>improvements: lot of work on Schemas by Kasimier Buchcik on facets
2121	checking and also mixed handling.</li>
2122	<li></li>
2123	</ul>
2124
2125	<h3>2.6.18: Mar 13 2005</h3>
2126	<ul>
2127	<li>build fixes: warnings (Peter Breitenlohner), testapi.c generation,
2128	Bakefile support (Francesco Montorsi), Windows compilation (Joel Reed),
2129	some gcc4 fixes, HP-UX portability fixes (Rick Jones).</li>
2130	<li>bug fixes: xmlSchemaElementDump namespace (Kasimier Buchcik), push and
2131	xmlreader stopping on non-fatal errors, thread support for dictionnaries
2132	reference counting (Gary Coady), internal subset and push problem, URL
2133	saved in xmlCopyDoc, various schemas bug fixes (Kasimier), Python paths
2134	fixup (Stephane Bidoul), xmlGetNodePath and namespaces, xmlSetNsProp fix
2135	(Mike Hommey), warning should not count as error (William Brack),
2136	xmlCreatePushParser empty chunk, XInclude parser flags (William), cleanup
2137	FTP and HTTP code to reuse the uri parsing and IPv6 (William),
2138	xmlTextWriterStartAttributeNS fix (Rob Richards), XMLLINT_INDENT being
2139	empty (William), xmlWriter bugs (Rob Richards), multithreading on Windows
2140	(Rich Salz), xmlSearchNsByHref fix (Kasimier), Python binding leak (Brent
2141	Hendricks), aliasing bug exposed by gcc4 on s390, xmlTextReaderNext bug
2142	(Rob Richards), Schemas decimal type fixes (William Brack),
2143	xmlByteConsumed static buffer (Ben Maurer).</li>
2144	<li>improvement: speedup parsing comments and DTDs, dictionnary support for
2145	hash tables, Schemas Identity constraints (Kasimier), streaming XPath
2146	subset, xmlTextReaderReadString added (Bjorn Reese), Schemas canonical
2147	values handling (Kasimier), add xmlTextReaderByteConsumed (Aron
2148	Stansvik),</li>
2149	<li>Documentation: Wiki support (Joel Reed)</li>
2150	</ul>
2151
2152	<h3>2.6.17: Jan 16 2005</h3>
2153	<ul>
2154	<li>build fixes: Windows, warnings removal (William Brack),
2155	maintainer-clean dependency(William), build in a different directory
2156	(William), fixing --with-minimum configure build (William), BeOS build
2157	(Marcin Konicki), Python-2.4 detection (William), compilation on AIX (Dan
2158	McNichol)</li>
2159	<li>bug fixes: xmlTextReaderHasAttributes (Rob Richards), xmlCtxtReadFile()
2160	to use the catalog(s), loop on output (William Brack), XPath memory leak,
2161	ID deallocation problem (Steve Shepard), debugDumpNode crash (William),
2162	warning not using error callback (William), xmlStopParser bug (William),
2163	UTF-16 with BOM on DTDs (William), namespace bug on empty elements in
2164	push mode (Rob Richards), line and col computations fixups (Aleksey
2165	Sanin), xmlURIEscape fix (William), xmlXPathErr on bad range (William),
2166	patterns with too many steps, bug in RNG choice optimization, line number
2167	sometimes missing.</li>
2168	<li>improvements: XSD Schemas (Kasimier Buchcik), python generator
2169	(William), xmlUTF8Strpos speedup (William), unicode Python strings
2170	(William), XSD error reports (Kasimier Buchcik), Python __str__ call
2171	serialize().</li>
2172	<li>new APIs: added xmlDictExists(), GetLineNumber and GetColumnNumber for
2173	the xmlReader (Aleksey Sanin), Dynamic Shared Libraries APIs (mostly Joel
2174	Reed), error extraction API from regexps, new XMLSave option for format
2175	(Phil Shafer)</li>
2176	<li>documentation: site improvement (John Fleck), FAQ entries
2177	(William).</li>
2178	</ul>
2179
2180	<h3>2.6.16: Nov 10 2004</h3>
2181	<ul>
2182	<li>general hardening and bug fixing crossing all the API based on new
2183	automated regression testing</li>
2184	<li>build fix: IPv6 build and test on AIX (Dodji Seketeli)</li>
2185	<li>bug fixes: problem with XML::Libxml reported by Petr Pajas, encoding
2186	conversion functions return values, UTF-8 bug affecting XPath reported by
2187	Markus Bertheau, catalog problem with NULL entries (William Brack)</li>
2188	<li>documentation: fix to xmllint man page, some API function descritpion
2189	were updated.</li>
2190	<li>improvements: DTD validation APIs provided at the Python level (Brent
2191	Hendricks)</li>
2192	</ul>
2193
2194	<h3>2.6.15: Oct 27 2004</h3>
2195	<ul>
2196	<li>security fixes on the nanoftp and nanohttp modules</li>
2197	<li>build fixes: xmllint detection bug in configure, building outside the
2198	source tree (Thomas Fitzsimmons)</li>
2199	<li>bug fixes: HTML parser on broken ASCII chars in names (William), Python
2200	paths (Malcolm Tredinnick), xmlHasNsProp and default namespace (William),
2201	saving to python file objects (Malcolm Tredinnick), DTD lookup fix
2202	(Malcolm), save back <group> in catalogs (William), tree build
2203	fixes (DV and Rob Richards), Schemas memory bug, structured error handler
2204	on Python 64bits, thread local memory deallocation, memory leak reported
2205	by Volker Roth, xmlValidateDtd in the presence of an internal subset,
2206	entities and _private problem (William), xmlBuildRelativeURI error
2207	(William).</li>
2208	<li>improvements: better XInclude error reports (William), tree debugging
2209	module and tests, convenience functions at the Reader API (Graham
2210	Bennett), add support for PI in the HTML parser.</li>
2211	</ul>
2212
2213	<h3>2.6.14: Sep 29 2004</h3>
2214	<ul>
2215	<li>build fixes: configure paths for xmllint and xsltproc, compilation
2216	without HTML parser, compilation warning cleanups (William Brack &
2217	Malcolm Tredinnick), VMS makefile update (Craig Berry),</li>
2218	<li>bug fixes: xmlGetUTF8Char (William Brack), QName properties (Kasimier
2219	Buchcik), XInclude testing, Notation serialization, UTF8ToISO8859x
2220	transcoding (Mark Itzcovitz), lots of XML Schemas cleanup and fixes
2221	(Kasimier), ChangeLog cleanup (Stepan Kasal), memory fixes (Mark Vakoc),
2222	handling of failed realloc(), out of bound array adressing in Schemas
2223	date handling, Python space/tabs cleanups (Malcolm Tredinnick), NMTOKENS
2224	E20 validation fix (Malcolm),</li>
2225	<li>improvements: added W3C XML Schemas testsuite (Kasimier Buchcik), add
2226	xmlSchemaValidateOneElement (Kasimier), Python exception hierearchy
2227	(Malcolm Tredinnick), Python libxml2 driver improvement (Malcolm
2228	Tredinnick), Schemas support for xsi:schemaLocation,
2229	xsi:noNamespaceSchemaLocation, xsi:type (Kasimier Buchcik)</li>
2230	</ul>
2231
2232	<h3>2.6.13: Aug 31 2004</h3>
2233	<ul>
2234	<li>build fixes: Windows and zlib (Igor Zlatkovic), -O flag with gcc,
2235	Solaris compiler warning, fixing RPM BuildRequires,</li>
2236	<li>fixes: DTD loading on Windows (Igor), Schemas error reports APIs
2237	(Kasimier Buchcik), Schemas validation crash, xmlCheckUTF8 (William Brack
2238	and Julius Mittenzwei), Schemas facet check (Kasimier), default namespace
2239	problem (William), Schemas hexbinary empty values, encoding error could
2240	genrate a serialization loop.</li>
2241	<li>Improvements: Schemas validity improvements (Kasimier), added --path
2242	and --load-trace options to xmllint</li>
2243	<li>documentation: tutorial update (John Fleck)</li>
2244	</ul>
2245
2246	<h3>2.6.12: Aug 22 2004</h3>
2247	<ul>
2248	<li>build fixes: fix --with-minimum, elfgcchack.h fixes (Peter
2249	Breitenlohner), perl path lookup (William), diff on Solaris (Albert
2250	Chin), some 64bits cleanups.</li>
2251	<li>Python: avoid a warning with 2.3 (William Brack), tab and space mixes
2252	(William), wrapper generator fixes (William), Cygwin support (Gerrit P.
2253	Haase), node wrapper fix (Marc-Antoine Parent), XML Schemas support
2254	(Torkel Lyng)</li>
2255	<li>Schemas: a lot of bug fixes and improvements from Kasimier Buchcik</li>
2256	<li>fixes: RVT fixes (William), XPath context resets bug (William), memory
2257	debug (Steve Hay), catalog white space handling (Peter Breitenlohner),
2258	xmlReader state after attribute reading (William), structured error
2259	handler (William), XInclude generated xml:base fixup (William), Windows
2260	memory reallocation problem (Steve Hay), Out of Memory conditions
2261	handling (William and Olivier Andrieu), htmlNewDoc() charset bug,
2262	htmlReadMemory init (William), a posteriori validation DTD base
2263	(William), notations serialization missing, xmlGetNodePath (Dodji),
2264	xmlCheckUTF8 (Diego Tartara), missing line numbers on entity
2265	(William)</li>
2266	<li>improvements: DocBook catalog build scrip (William), xmlcatalog tool
2267	(Albert Chin), xmllint --c14n option, no_proxy environment (Mike Hommey),
2268	xmlParseInNodeContext() addition, extend xmllint --shell, allow XInclude
2269	to not generate start/end nodes, extend xmllint --version to include CVS
2270	tag (William)</li>
2271	<li>documentation: web pages fixes, validity API docs fixes (William)
2272	schemas API fix (Eric Haszlakiewicz), xmllint man page (John Fleck)</li>
2273	</ul>
2274
2275	<h3>2.6.11: July 5 2004</h3>
2276	<ul>
2277	<li>Schemas: a lot of changes and improvements by Kasimier Buchcik for
2278	attributes, namespaces and simple types.</li>
2279	<li>build fixes: --with-minimum (William Brack), some gcc cleanup
2280	(William), --with-thread-alloc (William)</li>
2281	<li>portability: Windows binary package change (Igor Zlatkovic), Catalog
2282	path on Windows</li>
2283	<li>documentation: update to the tutorial (John Fleck), xmllint return code
2284	(John Fleck), man pages (Ville Skytta),</li>
2285	<li>bug fixes: C14N bug serializing namespaces (Aleksey Sanin), testSAX
2286	properly initialize the library (William), empty node set in XPath
2287	(William), xmlSchemas errors (William), invalid charref problem pointed
2288	by Morus Walter, XInclude xml:base generation (William), Relax-NG bug
2289	with div processing (William), XPointer and xml:base problem(William),
2290	Reader and entities, xmllint return code for schemas (William), reader
2291	streaming problem (Steve Ball), DTD serialization problem (William),
2292	libxml.m4 fixes (Mike Hommey), do not provide destructors as methods on
2293	Python classes, xmlReader buffer bug, Python bindings memory interfaces
2294	improvement (with StÃ©phane Bidoul), Fixed the push parser to be back to
2295	synchronous behaviour.</li>
2296	<li>improvement: custom per-thread I/O enhancement (Rob Richards), register
2297	namespace in debug shell (Stefano Debenedetti), Python based regression
2298	test for non-Unix users (William), dynamically increase the number of
2299	XPath extension functions in Python and fix a memory leak (Marc-Antoine
2300	Parent and William)</li>
2301	<li>performance: hack done with Arjan van de Ven to reduce ELF footprint
2302	and generated code on Linux, plus use gcc runtime profiling to optimize
2303	the code generated in the RPM packages.</li>
2304	</ul>
2305
2306	<h3>2.6.10: May 17 2004</h3>
2307	<ul>
2308	<li>Web page generated for ChangeLog</li>
2309	<li>build fixes: --without-html problems, make check without make all</li>
2310	<li>portability: problem with xpath.c on Windows (MSC and Borland), memcmp
2311	vs. strncmp on Solaris, XPath tests on Windows (Mark Vakoc), C++ do not
2312	use "list" as parameter name, make tests work with Python 1.5 (Ed
2313	Davis),</li>
2314	<li>improvements: made xmlTextReaderMode public, small buffers resizing
2315	(Morten Welinder), add --maxmem option to xmllint, add
2316	xmlPopInputCallback() for Matt Sergeant, refactoring of serialization
2317	escaping, added escaping customization</li>
2318	<li>bugfixes: xsd:extension (Taihei Goi), assorted regexp bugs (William
2319	Brack), xmlReader end of stream problem, node deregistration with reader,
2320	URI escaping and filemanes, XHTML1 formatting (Nick Wellnhofer), regexp
2321	transition reduction (William), various XSD Schemas fixes (Kasimier
2322	Buchcik), XInclude fallback problem (William), weird problems with DTD
2323	(William), structured error handler callback context (William), reverse
2324	xmlEncodeSpecialChars() behaviour back to escaping '"'</li>
2325	</ul>
2326
2327	<h3>2.6.9: Apr 18 2004</h3>
2328	<ul>
2329	<li>implement xml:id Working Draft, relaxed XPath id() checking</li>
2330	<li>bugfixes: xmlCtxtReset (Brent Hendricks), line number and CDATA (Dave
2331	Beckett), Relax-NG compilation (William Brack), Regexp patches (with
2332	William), xmlUriEscape (Mark Vakoc), a Relax-NG notAllowed problem (with
2333	William), Relax-NG name classes compares (William), XInclude duplicate
2334	fallback (William), external DTD encoding detection (William), a DTD
2335	validation bug (William), xmlReader Close() fix, recusive extention
2336	schemas</li>
2337	<li>improvements: use xmlRead* APIs in test tools (Mark Vakoc), indenting
2338	save optimization, better handle IIS broken HTTP redirect behaviour (Ian
2339	Hummel), HTML parser frameset (James Bursa), libxml2-python RPM
2340	dependancy, XML Schemas union support (Kasimier Buchcik), warning removal
2341	clanup (William), keep ChangeLog compressed when installing from RPMs</li>
2342	<li>documentation: examples and xmlDocDumpMemory docs (John Fleck), new
2343	example (load, xpath, modify, save), xmlCatalogDump() comments,</li>
2344	<li>Windows: Borland C++ builder (Eric Zurcher), work around Microsoft
2345	compiler NaN handling bug (Mark Vakoc)</li>
2346	</ul>
2347
2348	<h3>2.6.8: Mar 23 2004</h3>
2349	<ul>
2350	<li>First step of the cleanup of the serialization code and APIs</li>
2351	<li>XML Schemas: mixed content (Adam Dickmeiss), QName handling fixes (Adam
2352	Dickmeiss), anyURI for "" (John Belmonte)</li>
2353	<li>Python: Canonicalization C14N support added (Anthony Carrico)</li>
2354	<li>xmlDocCopyNode() extension (William)</li>
2355	<li>Relax-NG: fix when processing XInclude results (William), external
2356	reference in interleave (William), missing error on <choice>
2357	failure (William), memory leak in schemas datatype facets.</li>
2358	<li>xmlWriter: patch for better DTD support (Alfred Mickautsch)</li>
2359	<li>bug fixes: xmlXPathLangFunction memory leak (Mike Hommey and William
2360	Brack), no ID errors if using HTML_PARSE_NOERROR, xmlcatalog fallbacks to
2361	URI on SYSTEM lookup failure, XInclude parse flags inheritance (William),
2362	XInclude and XPointer fixes for entities (William), XML parser bug
2363	reported by Holger Rauch, nanohttp fd leak (William), regexps char
2364	groups '-' handling (William), dictionnary reference counting problems,
2365	do not close stderr.</li>
2366	<li>performance patches from Petr Pajas</li>
2367	<li>Documentation fixes: XML_CATALOG_FILES in man pages (Mike Hommey)</li>
2368	<li>compilation and portability fixes: --without-valid, catalog cleanups
2369	(Peter Breitenlohner), MingW patch (Roland Schwingel), cross-compilation
2370	to Windows (Christophe de Vienne), --with-html-dir fixup (Julio Merino
2371	Vidal), Windows build (Eric Zurcher)</li>
2372	</ul>
2373
2374	<h3>2.6.7: Feb 23 2004</h3>
2375	<ul>
2376	<li>documentation: tutorial updates (John Fleck), benchmark results</li>
2377	<li>xmlWriter: updates and fixes (Alfred Mickautsch, Lucas Brasilino)</li>
2378	<li>XPath optimization (Petr Pajas)</li>
2379	<li>DTD ID handling optimization</li>
2380	<li>bugfixes: xpath number with > 19 fractional (William Brack), push
2381	mode with unescaped '>' characters, fix xmllint --stream --timing, fix
2382	xmllint --memory --stream memory usage, xmlAttrSerializeTxtContent
2383	handling NULL, trying to fix Relax-NG/Perl interface.</li>
2384	<li>python: 2.3 compatibility, whitespace fixes (Malcolm Tredinnick)</li>
2385	<li>Added relaxng option to xmllint --shell</li>
2386	</ul>
2387
2388	<h3>2.6.6: Feb 12 2004</h3>
2389	<ul>
2390	<li>nanohttp and nanoftp: buffer overflow error on URI parsing (Igor and
2391	William) reported by Yuuichi Teranishi</li>
2392	<li>bugfixes: make test and path issues, xmlWriter attribute serialization
2393	(William Brack), xmlWriter indentation (William), schemas validation
2394	(Eric Haszlakiewicz), XInclude dictionnaries issues (William and Oleg
2395	Paraschenko), XInclude empty fallback (William), HTML warnings (William),
2396	XPointer in XInclude (William), Python namespace serialization,
2397	isolat1ToUTF8 bound error (Alfred Mickautsch), output of parameter
2398	entities in internal subset (William), internal subset bug in push mode,
2399	<xs:all> fix (Alexey Sarytchev)</li>
2400	<li>Build: fix for automake-1.8 (Alexander Winston), warnings removal
2401	(Philip Ludlam), SOCKLEN_T detection fixes (Daniel Richard), fix
2402	--with-minimum configuration.</li>
2403	<li>XInclude: allow the 2001 namespace without warning.</li>
2404	<li>Documentation: missing example/index.html (John Fleck), version
2405	dependancies (John Fleck)</li>
2406	<li>reader API: structured error reporting (Steve Ball)</li>
2407	<li>Windows compilation: mingw, msys (Mikhail Grushinskiy), function
2408	prototype (Cameron Johnson), MSVC6 compiler warnings, _WINSOCKAPI_
2409	patch</li>
2410	<li>Parsers: added xmlByteConsumed(ctxt) API to get the byte offest in
2411	input.</li>
2412	</ul>
2413
2414	<h3>2.6.5: Jan 25 2004</h3>
2415	<ul>
2416	<li>Bugfixes: dictionnaries for schemas (William Brack), regexp segfault
2417	(William), xs:all problem (William), a number of XPointer bugfixes
2418	(William), xmllint error go to stderr, DTD validation problem with
2419	namespace, memory leak (William), SAX1 cleanup and minimal options fixes
2420	(Mark Vadoc), parser context reset on error (Shaun McCance), XPath union
2421	evaluation problem (William) , xmlReallocLoc with NULL (Aleksey Sanin),
2422	XML Schemas double free (Steve Ball), XInclude with no href, argument
2423	callbacks order for XPath callbacks (Frederic Peters)</li>
2424	<li>Documentation: python scripts (William Brack), xslt stylesheets (John
2425	Fleck), doc (Sven Zimmerman), I/O example.</li>
2426	<li>Python bindings: fixes (William), enum support (StÃ©phane Bidoul),
2427	structured error reporting (StÃ©phane Bidoul)</li>
2428	<li>XInclude: various fixes for conformance, problem related to dictionnary
2429	references (William & me), recursion (William)</li>
2430	<li>xmlWriter: indentation (Lucas Brasilino), memory leaks (Alfred
2431	Mickautsch),</li>
2432	<li>xmlSchemas: normalizedString datatype (John Belmonte)</li>
2433	<li>code cleanup for strings functions (William)</li>
2434	<li>Windows: compiler patches (Mark Vakoc)</li>
2435	<li>Parser optimizations, a few new XPath and dictionnary APIs for future
2436	XSLT optimizations.</li>
2437	</ul>
2438
2439	<h3>2.6.4: Dec 24 2003</h3>
2440	<ul>
2441	<li>Windows build fixes (Igor Zlatkovic)</li>
2442	<li>Some serious XInclude problems reported by Oleg Paraschenko and</li>
2443	<li>Unix and Makefile packaging fixes (me, William Brack,</li>
2444	<li>Documentation improvements (John Fleck, William Brack), example fix
2445	(Lucas Brasilino)</li>
2446	<li>bugfixes: xmlTextReaderExpand() with xmlReaderWalker, XPath handling of
2447	NULL strings (William Brack) , API building reader or parser from
2448	filedescriptor should not close it, changed XPath sorting to be stable
2449	again (William Brack), xmlGetNodePath() generating '(null)' (William
2450	Brack), DTD validation and namespace bug (William Brack), XML Schemas
2451	double inclusion behaviour</li>
2452	</ul>
2453
2454	<h3>2.6.3: Dec 10 2003</h3>
2455	<ul>
2456	<li>documentation updates and cleanup (DV, William Brack, John Fleck)</li>
2457	<li>added a repository of examples, examples from Aleksey Sanin, Dodji
2458	Seketeli, Alfred Mickautsch</li>
2459	<li>Windows updates: Mark Vakoc, Igor Zlatkovic, Eric Zurcher, Mingw
2460	(Kenneth Haley)</li>
2461	<li>Unicode range checking (William Brack)</li>
2462	<li>code cleanup (William Brack)</li>
2463	<li>Python bindings: doc (John Fleck), bug fixes</li>
2464	<li>UTF-16 cleanup and BOM issues (William Brack)</li>
2465	<li>bug fixes: ID and xmlReader validation, XPath (William Brack),
2466	xmlWriter (Alfred Mickautsch), hash.h inclusion problem, HTML parser
2467	(James Bursa), attribute defaulting and validation, some serialization
2468	cleanups, XML_GET_LINE macro, memory debug when using threads (William
2469	Brack), serialization of attributes and entities content, xmlWriter
2470	(Daniel Schulman)</li>
2471	<li>XInclude bugfix, new APIs and update to the last version including the
2472	namespace change.</li>
2473	<li>XML Schemas improvements: include (Robert Stepanek), import and
2474	namespace handling, fixed the regression tests troubles, added examples
2475	based on Eric van der Vlist book, regexp fixes</li>
2476	<li>preliminary pattern support for streaming (needed for schemas
2477	constraints), added xmlTextReaderPreservePattern() to collect subdocument
2478	when streaming.</li>
2479	<li>various fixes in the structured error handling</li>
2480	</ul>
2481
2482	<h3>2.6.2: Nov 4 2003</h3>
2483	<ul>
2484	<li>XPath context unregistration fixes</li>
2485	<li>text node coalescing fixes (Mark Lilback)</li>
2486	<li>API to screate a W3C Schemas from an existing document (Steve Ball)</li>
2487	<li>BeOS patches (Marcin 'Shard' Konicki)</li>
2488	<li>xmlStrVPrintf function added (Aleksey Sanin)</li>
2489	<li>compilation fixes (Mark Vakoc)</li>
2490	<li>stdin parsing fix (William Brack)</li>
2491	<li>a posteriori DTD validation fixes</li>
2492	<li>xmlReader bug fixes: Walker fixes, python bindings</li>
2493	<li>fixed xmlStopParser() to really stop the parser and errors</li>
2494	<li>always generate line numbers when using the new xmlReadxxx
2495	functions</li>
2496	<li>added XInclude support to the xmlReader interface</li>
2497	<li>implemented XML_PARSE_NONET parser option</li>
2498	<li>DocBook XSLT processing bug fixed</li>
2499	<li>HTML serialization for <p> elements (William Brack and me)</li>
2500	<li>XPointer failure in XInclude are now handled as resource errors</li>
2501	<li>fixed xmllint --html to use the HTML serializer on output (added
2502	--xmlout to implement the previous behaviour of saving it using the XML
2503	serializer)</li>
2504	</ul>
2505
2506	<h3>2.6.1: Oct 28 2003</h3>
2507	<ul>
2508	<li>Mostly bugfixes after the big 2.6.0 changes</li>
2509	<li>Unix compilation patches: libxml.m4 (Patrick Welche), warnings cleanup
2510	(William Brack)</li>
2511	<li>Windows compilation patches (Joachim Bauch, Stephane Bidoul, Igor
2512	Zlatkovic)</li>
2513	<li>xmlWriter bugfix (Alfred Mickautsch)</li>
2514	<li>chvalid.[ch]: couple of fixes from Stephane Bidoul</li>
2515	<li>context reset: error state reset, push parser reset (Graham
2516	Bennett)</li>
2517	<li>context reuse: generate errors if file is not readable</li>
2518	<li>defaulted attributes for element coming from internal entities
2519	(Stephane Bidoul)</li>
2520	<li>Python: tab and spaces mix (William Brack)</li>
2521	<li>Error handler could crash in DTD validation in 2.6.0</li>
2522	<li>xmlReader: do not use the document or element _private field</li>
2523	<li>testSAX.c: avoid a problem with some PIs (Massimo Morara)</li>
2524	<li>general bug fixes: mandatory encoding in text decl, serializing
2525	Document Fragment nodes, xmlSearchNs 2.6.0 problem (Kasimier Buchcik),
2526	XPath errors not reported, slow HTML parsing of large documents.</li>
2527	</ul>
2528
2529	<h3>2.6.0: Oct 20 2003</h3>
2530	<ul>
2531	<li>Major revision release: should be API and ABI compatible but got a lot
2532	of change</li>
2533	<li>Increased the library modularity, far more options can be stripped out,
2534	a --with-minimum configuration will weight around 160KBytes</li>
2535	<li>Use per parser and per document dictionnary, allocate names and small
2536	text nodes from the dictionnary</li>
2537	<li>Switch to a SAX2 like parser rewrote most of the XML parser core,
2538	provides namespace resolution and defaulted attributes, minimize memory
2539	allocations and copies, namespace checking and specific error handling,
2540	immutable buffers, make predefined entities static structures, etc...</li>
2541	<li>rewrote all the error handling in the library, all errors can be
2542	intercepted at a structured level, with precise information
2543	available.</li>
2544	<li>New simpler and more generic XML and HTML parser APIs, allowing to
2545	easilly modify the parsing options and reuse parser context for multiple
2546	consecutive documents.</li>
2547	<li>Similar new APIs for the xmlReader, for options and reuse, provided new
2548	functions to access content as const strings, use them for Python
2549	bindings</li>
2550	<li>a lot of other smaller API improvements: xmlStrPrintf (Aleksey Sanin),
2551	Walker i.e. reader on a document tree based on Alfred Mickautsch code,
2552	make room in nodes for line numbers, reference counting and future PSVI
2553	extensions, generation of character ranges to be checked with faster
2554	algorithm (William), xmlParserMaxDepth (Crutcher Dunnavant), buffer
2555	access</li>
2556	<li>New xmlWriter API provided by Alfred Mickautsch</li>
2557	<li>Schemas: base64 support by Anthony Carrico</li>
2558	<li>Parser<->HTTP integration fix, proper processing of the Mime-Type
2559	and charset information if available.</li>
2560	<li>Relax-NG: bug fixes including the one reported by Martijn Faassen and
2561	zeroOrMore, better error reporting.</li>
2562	<li>Python bindings (StÃ©phane Bidoul), never use stdout for errors
2563	output</li>
2564	<li>Portability: all the headers have macros for export and calling
2565	convention definitions (Igor Zlatkovic), VMS update (Craig A. Berry),
2566	Windows: threads (Jesse Pelton), Borland compiler (Eric Zurcher, Igor),
2567	Mingw (Igor), typos (Mark Vakoc), beta version (Stephane Bidoul),
2568	warning cleanups on AIX and MIPS compilers (William Brack), BeOS (Marcin
2569	'Shard' Konicki)</li>
2570	<li>Documentation fixes and README (William Brack), search fix (William),
2571	tutorial updates (John Fleck), namespace docs (Stefan Kost)</li>
2572	<li>Bug fixes: xmlCleanupParser (Dave Beckett), threading uninitialized
2573	mutexes, HTML doctype lowercase, SAX/IO (William), compression detection
2574	and restore (William), attribute declaration in DTDs (William), namespace
2575	on attribute in HTML output (William), input filename (Rob Richards),
2576	namespace DTD validation, xmlReplaceNode (Chris Ryland), I/O callbacks
2577	(Markus Keim), CDATA serialization (Shaun McCance), xmlReader (Peter
2578	Derr), high codepoint charref like &#x10FFFF;, buffer access in push
2579	mode (Justin Fletcher), TLS threads on Windows (Jesse Pelton), XPath bug
2580	(William), xmlCleanupParser (Marc Liyanage), CDATA output (William), HTTP
2581	error handling.</li>
2582	<li>xmllint options: --dtdvalidfpi for Tobias Reif, --sax1 for compat
2583	testing, --nodict for building without tree dictionnary, --nocdata to
2584	replace CDATA by text, --nsclean to remove surperfluous namespace
2585	declarations</li>
2586	<li>added xml2-config --libtool-libs option from Kevin P. Fleming</li>
2587	<li>a lot of profiling and tuning of the code, speedup patch for
2588	xmlSearchNs() by Luca Padovani. The xmlReader should do far less
2589	allocation and it speed should get closer to SAX. Chris Anderson worked
2590	on speeding and cleaning up repetitive checking code.</li>
2591	<li>cleanup of "make tests"</li>
2592	<li>libxml-2.0-uninstalled.pc from Malcolm Tredinnick</li>
2593	<li>deactivated the broken docBook SGML parser code and plugged the XML
2594	parser instead.</li>
2595	</ul>
2596
2597	<h3>2.5.11: Sep 9 2003</h3>
2598
2599	<p>A bugfix only release:</p>
2600	<ul>
2601	<li>risk of crash in Relax-NG</li>
2602	<li>risk of crash when using multithreaded programs</li>
2603	</ul>
2604
2605	<h3>2.5.10: Aug 15 2003</h3>
2606
2607	<p>A bugfixes only release</p>
2608	<ul>
2609	<li>Windows Makefiles (William Brack)</li>
2610	<li>UTF-16 support fixes (Mark Itzcovitz)</li>
2611	<li>Makefile and portability (William Brack) automake, Linux alpha, Mingw
2612	on Windows (Mikhail Grushinskiy)</li>
2613	<li>HTML parser (Oliver Stoeneberg)</li>
2614	<li>XInclude performance problem reported by Kevin Ruscoe</li>
2615	<li>XML parser performance problem reported by Grant Goodale</li>
2616	<li>xmlSAXParseDTD() bug fix from Malcolm Tredinnick</li>
2617	<li>and a couple other cleanup</li>
2618	</ul>
2619
2620	<h3>2.5.9: Aug 9 2003</h3>
2621	<ul>
2622	<li>bugfixes: IPv6 portability, xmlHasNsProp (Markus Keim), Windows build
2623	(Wiliam Brake, Jesse Pelton, Igor), Schemas (Peter Sobisch), threading
2624	(Rob Richards), hexBinary type (), UTF-16 BOM (Dodji Seketeli),
2625	xmlReader, Relax-NG schemas compilation, namespace handling, EXSLT (Sean
2626	Griffin), HTML parsing problem (William Brack), DTD validation for mixed
2627	content + namespaces, HTML serialization, library initialization,
2628	progressive HTML parser</li>
2629	<li>better interfaces for Relax-NG error handling (Joachim Bauch, )</li>
2630	<li>adding xmlXIncludeProcessTree() for XInclud'ing in a subtree</li>
2631	<li>doc fixes and improvements (John Fleck)</li>
2632	<li>configure flag for -with-fexceptions when embedding in C++</li>
2633	<li>couple of new UTF-8 helper functions (William Brack)</li>
2634	<li>general encoding cleanup + ISO-8859-x without iconv (Peter Jacobi)</li>
2635	<li>xmlTextReader cleanup + enum for node types (Bjorn Reese)</li>
2636	<li>general compilation/warning cleanup Solaris/HP-UX/... (William
2637	Brack)</li>
2638	</ul>
2639
2640	<h3>2.5.8: Jul 6 2003</h3>
2641	<ul>
2642	<li>bugfixes: XPath, XInclude, file/URI mapping, UTF-16 save (Mark
2643	Itzcovitz), UTF-8 checking, URI saving, error printing (William Brack),
2644	PI related memleak, compilation without schemas or without xpath (Joerg
2645	Schmitz-Linneweber/Garry Pennington), xmlUnlinkNode problem with DTDs,
2646	rpm problem on , i86_64, removed a few compilation problems from 2.5.7,
2647	xmlIOParseDTD, and xmlSAXParseDTD (Malcolm Tredinnick)</li>
2648	<li>portability: DJGPP (MsDos) , OpenVMS (Craig A. Berry)</li>
2649	<li>William Brack fixed multithreading lock problems</li>
2650	<li>IPv6 patch for FTP and HTTP accesses (Archana Shah/Wipro)</li>
2651	<li>Windows fixes (Igor Zlatkovic, Eric Zurcher), threading (StÃ©phane
2652	Bidoul)</li>
2653	<li>A few W3C Schemas Structure improvements</li>
2654	<li>W3C Schemas Datatype improvements (Charlie Bozeman)</li>
2655	<li>Python bindings for thread globals (StÃ©phane Bidoul), and method/class
2656	generator</li>
2657	<li>added --nonet option to xmllint</li>
2658	<li>documentation improvements (John Fleck)</li>
2659	</ul>
2660
2661	<h3>2.5.7: Apr 25 2003</h3>
2662	<ul>
2663	<li>Relax-NG: Compiling to regexp and streaming validation on top of the
2664	xmlReader interface, added to xmllint --stream</li>
2665	<li>xmlReader: Expand(), Next() and DOM access glue, bug fixes</li>
2666	<li>Support for large files: RGN validated a 4.5GB instance</li>
2667	<li>Thread support is now configured in by default</li>
2668	<li>Fixes: update of the Trio code (Bjorn), WXS Date and Duration fixes
2669	(Charles Bozeman), DTD and namespaces (Brent Hendricks), HTML push parser
2670	and zero bytes handling, some missing Windows file path conversions,
2671	behaviour of the parser and validator in the presence of "out of memory"
2672	error conditions</li>
2673	<li>extended the API to be able to plug a garbage collecting memory
2674	allocator, added xmlMallocAtomic() and modified the allocations
2675	accordingly.</li>
2676	<li>Performances: removed excessive malloc() calls, speedup of the push and
2677	xmlReader interfaces, removed excessive thread locking</li>
2678	<li>Documentation: man page (John Fleck), xmlReader documentation</li>
2679	<li>Python: adding binding for xmlCatalogAddLocal (Brent M Hendricks)</li>
2680	</ul>
2681
2682	<h3>2.5.6: Apr 1 2003</h3>
2683	<ul>
2684	<li>Fixed W3C XML Schemas datatype, should be compliant now except for
2685	binHex and base64 which are not supported yet.</li>
2686	<li>bug fixes: non-ASCII IDs, HTML output, XInclude on large docs and
2687	XInclude entities handling, encoding detection on external subsets, XML
2688	Schemas bugs and memory leaks, HTML parser (James Bursa)</li>
2689	<li>portability: python/trio (Albert Chin), Sun compiler warnings</li>
2690	<li>documentation: added --relaxng option to xmllint man page (John)</li>
2691	<li>improved error reporting: xml:space, start/end tag mismatches, Relax NG
2692	errors</li>
2693	</ul>
2694
2695	<h3>2.5.5: Mar 24 2003</h3>
2696	<ul>
2697	<li>Lot of fixes on the Relax NG implementation. More testing including
2698	DocBook and TEI examples.</li>
2699	<li>Increased the support for W3C XML Schemas datatype</li>
2700	<li>Several bug fixes in the URI handling layer</li>
2701	<li>Bug fixes: HTML parser, xmlReader, DTD validation, XPath, encoding
2702	conversion, line counting in the parser.</li>
2703	<li>Added support for $XMLLINT_INDENT environment variable, FTP delete</li>
2704	<li>Fixed the RPM spec file name</li>
2705	</ul>
2706
2707	<h3>2.5.4: Feb 20 2003</h3>
2708	<ul>
2709	<li>Conformance testing and lot of fixes on Relax NG and XInclude
2710	implementation</li>
2711	<li>Implementation of XPointer element() scheme</li>
2712	<li>Bug fixes: XML parser, XInclude entities merge, validity checking on
2713	namespaces,
2714	<p>2 serialization bugs, node info generation problems, a DTD regexp
2715	generation problem.</p>
2716	</li>
2717	<li>Portability: windows updates and path canonicalization (Igor)</li>
2718	<li>A few typo fixes (Kjartan Maraas)</li>
2719	<li>Python bindings generator fixes (Stephane Bidoul)</li>
2720	</ul>
2721
2722	<h3>2.5.3: Feb 10 2003</h3>
2723	<ul>
2724	<li>RelaxNG and XML Schemas datatypes improvements, and added a first
2725	version of RelaxNG Python bindings</li>
2726	<li>Fixes: XLink (Sean Chittenden), XInclude (Sean Chittenden), API fix for
2727	serializing namespace nodes, encoding conversion bug, XHTML1
2728	serialization</li>
2729	<li>Portability fixes: Windows (Igor), AMD 64bits RPM spec file</li>
2730	</ul>
2731
2732	<h3>2.5.2: Feb 5 2003</h3>
2733	<ul>
2734	<li>First implementation of RelaxNG, added --relaxng flag to xmllint</li>
2735	<li>Schemas support now compiled in by default.</li>
2736	<li>Bug fixes: DTD validation, namespace checking, XInclude and entities,
2737	delegateURI in XML Catalogs, HTML parser, XML reader (StÃ©phane Bidoul),
2738	XPath parser and evaluation, UTF8ToUTF8 serialization, XML reader memory
2739	consumption, HTML parser, HTML serialization in the presence of
2740	namespaces</li>
2741	<li>added an HTML API to check elements and attributes.</li>
2742	<li>Documentation improvement, PDF for the tutorial (John Fleck), doc
2743	patches (Stefan Kost)</li>
2744	<li>Portability fixes: NetBSD (Julio Merino), Windows (Igor Zlatkovic)</li>
2745	<li>Added python bindings for XPointer, contextual error reporting
2746	(StÃ©phane Bidoul)</li>
2747	<li>URI/file escaping problems (Stefano Zacchiroli)</li>
2748	</ul>
2749
2750	<h3>2.5.1: Jan 8 2003</h3>
2751	<ul>
2752	<li>Fixes a memory leak and configuration/compilation problems in 2.5.0</li>
2753	<li>documentation updates (John)</li>
2754	<li>a couple of XmlTextReader fixes</li>
2755	</ul>
2756
2757	<h3>2.5.0: Jan 6 2003</h3>
2758	<ul>
2759	<li>New <a href="xmlreader.html">XmltextReader interface</a> based on C#
2760	API (with help of StÃ©phane Bidoul)</li>
2761	<li>Windows: more exports, including the new API (Igor)</li>
2762	<li>XInclude fallback fix</li>
2763	<li>Python: bindings for the new API, packaging (StÃ©phane Bidoul),
2764	drv_libxml2.py Python xml.sax driver (StÃ©phane Bidoul), fixes, speedup
2765	and iterators for Python-2.2 (Hannu Krosing)</li>
2766	<li>Tutorial fixes (john Fleck and Niraj Tolia) xmllint man update
2767	(John)</li>
2768	<li>Fix an XML parser bug raised by Vyacheslav Pindyura</li>
2769	<li>Fix for VMS serialization (Nigel Hall) and config (Craig A. Berry)</li>
2770	<li>Entities handling fixes</li>
2771	<li>new API to optionally track node creation and deletion (Lukas
2772	Schroeder)</li>
2773	<li>Added documentation for the XmltextReader interface and some <a
2774	href="guidelines.html">XML guidelines</a></li>
2775	</ul>
2776
2777	<h3>2.4.30: Dec 12 2002</h3>
2778	<ul>
2779	<li>2.4.29 broke the python bindings, rereleasing</li>
2780	<li>Improvement/fixes of the XML API generator, and couple of minor code
2781	fixes.</li>
2782	</ul>
2783
2784	<h3>2.4.29: Dec 11 2002</h3>
2785	<ul>
2786	<li>Windows fixes (Igor): Windows CE port, pthread linking, python bindings
2787	(StÃ©phane Bidoul), Mingw (Magnus Henoch), and export list updates</li>
2788	<li>Fix for prev in python bindings (ERDI Gergo)</li>
2789	<li>Fix for entities handling (Marcus Clarke)</li>
2790	<li>Refactored the XML and HTML dumps to a single code path, fixed XHTML1
2791	dump</li>
2792	<li>Fix for URI parsing when handling URNs with fragment identifiers</li>
2793	<li>Fix for HTTP URL escaping problem</li>
2794	<li>added an TextXmlReader (C#) like API (work in progress)</li>
2795	<li>Rewrote the API in XML generation script, includes a C parser and saves
2796	more information needed for C# bindings</li>
2797	</ul>
2798
2799	<h3>2.4.28: Nov 22 2002</h3>
2800	<ul>
2801	<li>a couple of python binding fixes</li>
2802	<li>2 bug fixes in the XML push parser</li>
2803	<li>potential memory leak removed (Martin Stoilov)</li>
2804	<li>fix to the configure script for Unix (Dimitri Papadopoulos)</li>
2805	<li>added encoding support for XInclude parse="text"</li>
2806	<li>autodetection of XHTML1 and specific serialization rules added</li>
2807	<li>nasty threading bug fixed (William Brack)</li>
2808	</ul>
2809
2810	<h3>2.4.27: Nov 17 2002</h3>
2811	<ul>
2812	<li>fixes for the Python bindings</li>
2813	<li>a number of bug fixes: SGML catalogs, xmlParseBalancedChunkMemory(),
2814	HTML parser, Schemas (Charles Bozeman), document fragment support
2815	(Christian Glahn), xmlReconciliateNs (Brian Stafford), XPointer,
2816	xmlFreeNode(), xmlSAXParseMemory (Peter Jones), xmlGetNodePath (Petr
2817	Pajas), entities processing</li>
2818	<li>added grep to xmllint --shell</li>
2819	<li>VMS update patch from Craig A. Berry</li>
2820	<li>cleanup of the Windows build with support for more compilers (Igor),
2821	better thread support on Windows</li>
2822	<li>cleanup of Unix Makefiles and spec file</li>
2823	<li>Improvements to the documentation (John Fleck)</li>
2824	</ul>
2825
2826	<h3>2.4.26: Oct 18 2002</h3>
2827	<ul>
2828	<li>Patches for Windows CE port, improvements on Windows paths handling</li>
2829	<li>Fixes to the validation code (DTD and Schemas), xmlNodeGetPath() ,
2830	HTML serialization, Namespace compliance, and a number of small
2831	problems</li>
2832	</ul>
2833
2834	<h3>2.4.25: Sep 26 2002</h3>
2835	<ul>
2836	<li>A number of bug fixes: XPath, validation, Python bindings, DOM and
2837	tree, xmlI/O, Html</li>
2838	<li>Serious rewrite of XInclude</li>
2839	<li>Made XML Schemas regexp part of the default build and APIs, small fix
2840	and improvement of the regexp core</li>
2841	<li>Changed the validation code to reuse XML Schemas regexp APIs</li>
2842	<li>Better handling of Windows file paths, improvement of Makefiles (Igor,
2843	Daniel Gehriger, Mark Vakoc)</li>
2844	<li>Improved the python I/O bindings, the tests, added resolver and regexp
2845	APIs</li>
2846	<li>New logos from Marc Liyanage</li>
2847	<li>Tutorial improvements: John Fleck, Christopher Harris</li>
2848	<li>Makefile: Fixes for AMD x86_64 (Mandrake), DESTDIR (Christophe
2849	Merlet)</li>
2850	<li>removal of all stderr/perror use for error reporting</li>
2851	<li>Better error reporting: XPath and DTD validation</li>
2852	<li>update of the trio portability layer (Bjorn Reese)</li>
2853	</ul>
2854
2855	<p><strong>2.4.24: Aug 22 2002</strong></p>
2856	<ul>
2857	<li>XPath fixes (William), xf:escape-uri() (Wesley Terpstra)</li>
2858	<li>Python binding fixes: makefiles (William), generator, rpm build, x86-64
2859	(fcrozat)</li>
2860	<li>HTML <style> and boolean attributes serializer fixes</li>
2861	<li>C14N improvements by Aleksey</li>
2862	<li>doc cleanups: Rick Jones</li>
2863	<li>Windows compiler makefile updates: Igor and Elizabeth Barham</li>
2864	<li>XInclude: implementation of fallback and xml:base fixup added</li>
2865	</ul>
2866
2867	<h3>2.4.23: July 6 2002</h3>
2868	<ul>
2869	<li>performances patches: Peter Jacobi</li>
2870	<li>c14n fixes, testsuite and performances: Aleksey Sanin</li>
2871	<li>added xmlDocFormatDump: Chema Celorio</li>
2872	<li>new tutorial: John Fleck</li>
2873	<li>new hash functions and performances: Sander Vesik, portability fix from
2874	Peter Jacobi</li>
2875	<li>a number of bug fixes: XPath (William Brack, Richard Jinks), XML and
2876	HTML parsers, ID lookup function</li>
2877	<li>removal of all remaining sprintf: Aleksey Sanin</li>
2878	</ul>
2879
2880	<h3>2.4.22: May 27 2002</h3>
2881	<ul>
2882	<li>a number of bug fixes: configure scripts, base handling, parser, memory
2883	usage, HTML parser, XPath, documentation (Christian Cornelssen),
2884	indentation, URI parsing</li>
2885	<li>Optimizations for XMLSec, fixing and making public some of the network
2886	protocol handlers (Aleksey)</li>
2887	<li>performance patch from Gary Pennington</li>
2888	<li>Charles Bozeman provided date and time support for XML Schemas
2889	datatypes</li>
2890	</ul>
2891
2892	<h3>2.4.21: Apr 29 2002</h3>
2893
2894	<p>This release is both a bug fix release and also contains the early XML
2895	Schemas <a href="http://www.w3.org/TR/xmlschema-1/">structures</a> and <a
2896	href="http://www.w3.org/TR/xmlschema-2/">datatypes</a> code, beware, all
2897	interfaces are likely to change, there is huge holes, it is clearly a work in
2898	progress and don't even think of putting this code in a production system,
2899	it's actually not compiled in by default. The real fixes are:
2900	</p>
2901
2902	<ul>
2903	<li>a couple of bugs or limitations introduced in 2.4.20</li>
2904	<li>patches for Borland C++ and MSC by Igor</li>
2905	<li>some fixes on XPath strings and conformance patches by Richard
2906	Jinks</li>
2907	<li>patch from Aleksey for the ExcC14N specification</li>
2908	<li>OSF/1 bug fix by Bjorn</li>
2909	</ul>
2910
2911	<h3>2.4.20: Apr 15 2002</h3>
2912	<ul>
2913	<li>bug fixes: file descriptor leak, XPath, HTML output, DTD validation</li>
2914	<li>XPath conformance testing by Richard Jinks</li>
2915	<li>Portability fixes: Solaris, MPE/iX, Windows, OSF/1, python bindings,
2916	libxml.m4</li>
2917	</ul>
2918
2919	<h3>2.4.19: Mar 25 2002</h3>
2920	<ul>
2921	<li>bug fixes: half a dozen XPath bugs, Validation, ISO-Latin to UTF8
2922	encoder</li>
2923	<li>portability fixes in the HTTP code</li>
2924	<li>memory allocation checks using valgrind, and profiling tests</li>
2925	<li>revamp of the Windows build and Makefiles</li>
2926	</ul>
2927
2928	<h3>2.4.18: Mar 18 2002</h3>
2929	<ul>
2930	<li>bug fixes: tree, SAX, canonicalization, validation, portability,
2931	XPath</li>
2932	<li>removed the --with-buffer option it was becoming unmaintainable</li>
2933	<li>serious cleanup of the Python makefiles</li>
2934	<li>speedup patch to XPath very effective for DocBook stylesheets</li>
2935	<li>Fixes for Windows build, cleanup of the documentation</li>
2936	</ul>
2937
2938	<h3>2.4.17: Mar 8 2002</h3>
2939	<ul>
2940	<li>a lot of bug fixes, including "namespace nodes have no parents in
2941	XPath"</li>
2942	<li>fixed/improved the Python wrappers, added more examples and more
2943	regression tests, XPath extension functions can now return node-sets</li>
2944	<li>added the XML Canonicalization support from Aleksey Sanin</li>
2945	</ul>
2946
2947	<h3>2.4.16: Feb 20 2002</h3>
2948	<ul>
2949	<li>a lot of bug fixes, most of them were triggered by the XML Testsuite
2950	from OASIS and W3C. Compliance has been significantly improved.</li>
2951	<li>a couple of portability fixes too.</li>
2952	</ul>
2953
2954	<h3>2.4.15: Feb 11 2002</h3>
2955	<ul>
2956	<li>Fixed the Makefiles, especially the python module ones</li>
2957	<li>A few bug fixes and cleanup</li>
2958	<li>Includes cleanup</li>
2959	</ul>
2960
2961	<h3>2.4.14: Feb 8 2002</h3>
2962	<ul>
2963	<li>Change of License to the <a
2964	href="http://www.opensource.org/licenses/mit-license.html">MIT
2965	License</a> basically for integration in XFree86 codebase, and removing
2966	confusion around the previous dual-licensing</li>
2967	<li>added Python bindings, beta software but should already be quite
2968	complete</li>
2969	<li>a large number of fixes and cleanups, especially for all tree
2970	manipulations</li>
2971	<li>cleanup of the headers, generation of a reference API definition in
2972	XML</li>
2973	</ul>
2974
2975	<h3>2.4.13: Jan 14 2002</h3>
2976	<ul>
2977	<li>update of the documentation: John Fleck and Charlie Bozeman</li>
2978	<li>cleanup of timing code from Justin Fletcher</li>
2979	<li>fixes for Windows and initial thread support on Win32: Igor and Serguei
2980	Narojnyi</li>
2981	<li>Cygwin patch from Robert Collins</li>
2982	<li>added xmlSetEntityReferenceFunc() for Keith Isdale work on xsldbg</li>
2983	</ul>
2984
2985	<h3>2.4.12: Dec 7 2001</h3>
2986	<ul>
2987	<li>a few bug fixes: thread (Gary Pennington), xmllint (Geert Kloosterman),
2988	XML parser (Robin Berjon), XPointer (Danny Jamshy), I/O cleanups
2989	(robert)</li>
2990	<li>Eric Lavigne contributed project files for MacOS</li>
2991	<li>some makefiles cleanups</li>
2992	</ul>
2993
2994	<h3>2.4.11: Nov 26 2001</h3>
2995	<ul>
2996	<li>fixed a couple of errors in the includes, fixed a few bugs, some code
2997	cleanups</li>
2998	<li>xmllint man pages improvement by Heiko Rupp</li>
2999	<li>updated VMS build instructions from John A Fotheringham</li>
3000	<li>Windows Makefiles updates from Igor</li>
3001	</ul>
3002
3003	<h3>2.4.10: Nov 10 2001</h3>
3004	<ul>
3005	<li>URI escaping fix (Joel Young)</li>
3006	<li>added xmlGetNodePath() (for paths or XPointers generation)</li>
3007	<li>Fixes namespace handling problems when using DTD and validation</li>
3008	<li>improvements on xmllint: Morus Walter patches for --format and
3009	--encode, Stefan Kost and Heiko Rupp improvements on the --shell</li>
3010	<li>fixes for xmlcatalog linking pointed by Weiqi Gao</li>
3011	<li>fixes to the HTML parser</li>
3012	</ul>
3013
3014	<h3>2.4.9: Nov 6 2001</h3>
3015	<ul>
3016	<li>fixes more catalog bugs</li>
3017	<li>avoid a compilation problem, improve xmlGetLineNo()</li>
3018	</ul>
3019
3020	<h3>2.4.8: Nov 4 2001</h3>
3021	<ul>
3022	<li>fixed SGML catalogs broken in previous release, updated xmlcatalog
3023	tool</li>
3024	<li>fixed a compile errors and some includes troubles.</li>
3025	</ul>
3026
3027	<h3>2.4.7: Oct 30 2001</h3>
3028	<ul>
3029	<li>exported some debugging interfaces</li>
3030	<li>serious rewrite of the catalog code</li>
3031	<li>integrated Gary Pennington thread safety patch, added configure option
3032	and regression tests</li>
3033	<li>removed an HTML parser bug</li>
3034	<li>fixed a couple of potentially serious validation bugs</li>
3035	<li>integrated the SGML DocBook support in xmllint</li>
3036	<li>changed the nanoftp anonymous login passwd</li>
3037	<li>some I/O cleanup and a couple of interfaces for Perl wrapper</li>
3038	<li>general bug fixes</li>
3039	<li>updated xmllint man page by John Fleck</li>
3040	<li>some VMS and Windows updates</li>
3041	</ul>
3042
3043	<h3>2.4.6: Oct 10 2001</h3>
3044	<ul>
3045	<li>added an updated man pages by John Fleck</li>
3046	<li>portability and configure fixes</li>
3047	<li>an infinite loop on the HTML parser was removed (William)</li>
3048	<li>Windows makefile patches from Igor</li>
3049	<li>fixed half a dozen bugs reported for libxml or libxslt</li>
3050	<li>updated xmlcatalog to be able to modify SGML super catalogs</li>
3051	</ul>
3052
3053	<h3>2.4.5: Sep 14 2001</h3>
3054	<ul>
3055	<li>Remove a few annoying bugs in 2.4.4</li>
3056	<li>forces the HTML serializer to output decimal charrefs since some
3057	version of Netscape can't handle hexadecimal ones</li>
3058	</ul>
3059
3060	<h3>1.8.16: Sep 14 2001</h3>
3061	<ul>
3062	<li>maintenance release of the old libxml1 branch, couple of bug and
3063	portability fixes</li>
3064	</ul>
3065
3066	<h3>2.4.4: Sep 12 2001</h3>
3067	<ul>
3068	<li>added --convert to xmlcatalog, bug fixes and cleanups of XML
3069	Catalog</li>
3070	<li>a few bug fixes and some portability changes</li>
3071	<li>some documentation cleanups</li>
3072	</ul>
3073
3074	<h3>2.4.3: Aug 23 2001</h3>
3075	<ul>
3076	<li>XML Catalog support see the doc</li>
3077	<li>New NaN/Infinity floating point code</li>
3078	<li>A few bug fixes</li>
3079	</ul>
3080
3081	<h3>2.4.2: Aug 15 2001</h3>
3082	<ul>
3083	<li>adds xmlLineNumbersDefault() to control line number generation</li>
3084	<li>lot of bug fixes</li>
3085	<li>the Microsoft MSC projects files should now be up to date</li>
3086	<li>inheritance of namespaces from DTD defaulted attributes</li>
3087	<li>fixes a serious potential security bug</li>
3088	<li>added a --format option to xmllint</li>
3089	</ul>
3090
3091	<h3>2.4.1: July 24 2001</h3>
3092	<ul>
3093	<li>possibility to keep line numbers in the tree</li>
3094	<li>some computation NaN fixes</li>
3095	<li>extension of the XPath API</li>
3096	<li>cleanup for alpha and ia64 targets</li>
3097	<li>patch to allow saving through HTTP PUT or POST</li>
3098	</ul>
3099
3100	<h3>2.4.0: July 10 2001</h3>
3101	<ul>
3102	<li>Fixed a few bugs in XPath, validation, and tree handling.</li>
3103	<li>Fixed XML Base implementation, added a couple of examples to the
3104	regression tests</li>
3105	<li>A bit of cleanup</li>
3106	</ul>
3107
3108	<h3>2.3.14: July 5 2001</h3>
3109	<ul>
3110	<li>fixed some entities problems and reduce memory requirement when
3111	substituting them</li>
3112	<li>lots of improvements in the XPath queries interpreter can be
3113	substantially faster</li>
3114	<li>Makefiles and configure cleanups</li>
3115	<li>Fixes to XPath variable eval, and compare on empty node set</li>
3116	<li>HTML tag closing bug fixed</li>
3117	<li>Fixed an URI reference computation problem when validating</li>
3118	</ul>
3119
3120	<h3>2.3.13: June 28 2001</h3>
3121	<ul>
3122	<li>2.3.12 configure.in was broken as well as the push mode XML parser</li>
3123	<li>a few more fixes for compilation on Windows MSC by Yon Derek</li>
3124	</ul>
3125
3126	<h3>1.8.14: June 28 2001</h3>
3127	<ul>
3128	<li>Zbigniew Chyla gave a patch to use the old XML parser in push mode</li>
3129	<li>Small Makefile fix</li>
3130	</ul>
3131
3132	<h3>2.3.12: June 26 2001</h3>
3133	<ul>
3134	<li>lots of cleanup</li>
3135	<li>a couple of validation fix</li>
3136	<li>fixed line number counting</li>
3137	<li>fixed serious problems in the XInclude processing</li>
3138	<li>added support for UTF8 BOM at beginning of entities</li>
3139	<li>fixed a strange gcc optimizer bugs in xpath handling of float, gcc-3.0
3140	miscompile uri.c (William), Thomas Leitner provided a fix for the
3141	optimizer on Tru64</li>
3142	<li>incorporated Yon Derek and Igor Zlatkovic fixes and improvements for
3143	compilation on Windows MSC</li>
3144	<li>update of libxml-doc.el (Felix Natter)</li>
3145	<li>fixed 2 bugs in URI normalization code</li>
3146	</ul>
3147
3148	<h3>2.3.11: June 17 2001</h3>
3149	<ul>
3150	<li>updates to trio, Makefiles and configure should fix some portability
3151	problems (alpha)</li>
3152	<li>fixed some HTML serialization problems (pre, script, and block/inline
3153	handling), added encoding aware APIs, cleanup of this code</li>
3154	<li>added xmlHasNsProp()</li>
3155	<li>implemented a specific PI for encoding support in the DocBook SGML
3156	parser</li>
3157	<li>some XPath fixes (-Infinity, / as a function parameter and namespaces
3158	node selection)</li>
3159	<li>fixed a performance problem and an error in the validation code</li>
3160	<li>fixed XInclude routine to implement the recursive behaviour</li>
3161	<li>fixed xmlFreeNode problem when libxml is included statically twice</li>
3162	<li>added --version to xmllint for bug reports</li>
3163	</ul>
3164
3165	<h3>2.3.10: June 1 2001</h3>
3166	<ul>
3167	<li>fixed the SGML catalog support</li>
3168	<li>a number of reported bugs got fixed, in XPath, iconv detection,
3169	XInclude processing</li>
3170	<li>XPath string function should now handle unicode correctly</li>
3171	</ul>
3172
3173	<h3>2.3.9: May 19 2001</h3>
3174
3175	<p>Lots of bugfixes, and added a basic SGML catalog support:
3176	</p>
3177	<ul>
3178	<li>HTML push bugfix #54891 and another patch from Jonas Borgstrom</li>
3179	<li>some serious speed optimization again</li>
3180	<li>some documentation cleanups</li>
3181	<li>trying to get better linking on Solaris (-R)</li>
3182	<li>XPath API cleanup from Thomas Broyer</li>
3183	<li>Validation bug fixed #54631, added a patch from Gary Pennington, fixed
3184	xmlValidGetValidElements()</li>
3185	<li>Added an INSTALL file</li>
3186	<li>Attribute removal added to API: #54433</li>
3187	<li>added a basic support for SGML catalogs</li>
3188	<li>fixed xmlKeepBlanksDefault(0) API</li>
3189	<li>bugfix in xmlNodeGetLang()</li>
3190	<li>fixed a small configure portability problem</li>
3191	<li>fixed an inversion of SYSTEM and PUBLIC identifier in HTML document</li>
3192	</ul>
3193
3194	<h3>1.8.13: May 14 2001</h3>
3195	<ul>
3196	<li>bugfixes release of the old libxml1 branch used by Gnome</li>
3197	</ul>
3198
3199	<h3>2.3.8: May 3 2001</h3>
3200	<ul>
3201	<li>Integrated an SGML DocBook parser for the Gnome project</li>
3202	<li>Fixed a few things in the HTML parser</li>
3203	<li>Fixed some XPath bugs raised by XSLT use, tried to fix the floating
3204	point portability issue</li>
3205	<li>Speed improvement (8M/s for SAX, 3M/s for DOM, 1.5M/s for
3206	DOM+validation using the XML REC as input and a 700MHz celeron).</li>
3207	<li>incorporated more Windows cleanup</li>
3208	<li>added xmlSaveFormatFile()</li>
3209	<li>fixed problems in copying nodes with entities references (gdome)</li>
3210	<li>removed some troubles surrounding the new validation module</li>
3211	</ul>
3212
3213	<h3>2.3.7: April 22 2001</h3>
3214	<ul>
3215	<li>lots of small bug fixes, corrected XPointer</li>
3216	<li>Non deterministic content model validation support</li>
3217	<li>added xmlDocCopyNode for gdome2</li>
3218	<li>revamped the way the HTML parser handles end of tags</li>
3219	<li>XPath: corrections of namespaces support and number formatting</li>
3220	<li>Windows: Igor Zlatkovic patches for MSC compilation</li>
3221	<li>HTML output fixes from P C Chow and William M. Brack</li>
3222	<li>Improved validation speed sensible for DocBook</li>
3223	<li>fixed a big bug with ID declared in external parsed entities</li>
3224	<li>portability fixes, update of Trio from Bjorn Reese</li>
3225	</ul>
3226
3227	<h3>2.3.6: April 8 2001</h3>
3228	<ul>
3229	<li>Code cleanup using extreme gcc compiler warning options, found and
3230	cleared half a dozen potential problem</li>
3231	<li>the Eazel team found an XML parser bug</li>
3232	<li>cleaned up the user of some of the string formatting function. used the
3233	trio library code to provide the one needed when the platform is missing
3234	them</li>
3235	<li>xpath: removed a memory leak and fixed the predicate evaluation
3236	problem, extended the testsuite and cleaned up the result. XPointer seems
3237	broken ...</li>
3238	</ul>
3239
3240	<h3>2.3.5: Mar 23 2001</h3>
3241	<ul>
3242	<li>Biggest change is separate parsing and evaluation of XPath expressions,
3243	there is some new APIs for this too</li>
3244	<li>included a number of bug fixes(XML push parser, 51876, notations,
3245	52299)</li>
3246	<li>Fixed some portability issues</li>
3247	</ul>
3248
3249	<h3>2.3.4: Mar 10 2001</h3>
3250	<ul>
3251	<li>Fixed bugs #51860 and #51861</li>
3252	<li>Added a global variable xmlDefaultBufferSize to allow default buffer
3253	size to be application tunable.</li>
3254	<li>Some cleanup in the validation code, still a bug left and this part
3255	should probably be rewritten to support ambiguous content model :-\</li>
3256	<li>Fix a couple of serious bugs introduced or raised by changes in 2.3.3
3257	parser</li>
3258	<li>Fixed another bug in xmlNodeGetContent()</li>
3259	<li>Bjorn fixed XPath node collection and Number formatting</li>
3260	<li>Fixed a loop reported in the HTML parsing</li>
3261	<li>blank space are reported even if the Dtd content model proves that they
3262	are formatting spaces, this is for XML conformance</li>
3263	</ul>
3264
3265	<h3>2.3.3: Mar 1 2001</h3>
3266	<ul>
3267	<li>small change in XPath for XSLT</li>
3268	<li>documentation cleanups</li>
3269	<li>fix in validation by Gary Pennington</li>
3270	<li>serious parsing performances improvements</li>
3271	</ul>
3272
3273	<h3>2.3.2: Feb 24 2001</h3>
3274	<ul>
3275	<li>chasing XPath bugs, found a bunch, completed some TODO</li>
3276	<li>fixed a Dtd parsing bug</li>
3277	<li>fixed a bug in xmlNodeGetContent</li>
3278	<li>ID/IDREF support partly rewritten by Gary Pennington</li>
3279	</ul>
3280
3281	<h3>2.3.1: Feb 15 2001</h3>
3282	<ul>
3283	<li>some XPath and HTML bug fixes for XSLT</li>
3284	<li>small extension of the hash table interfaces for DOM gdome2
3285	implementation</li>
3286	<li>A few bug fixes</li>
3287	</ul>
3288
3289	<h3>2.3.0: Feb 8 2001 (2.2.12 was on 25 Jan but I didn't kept track)</h3>
3290	<ul>
3291	<li>Lots of XPath bug fixes</li>
3292	<li>Add a mode with Dtd lookup but without validation error reporting for
3293	XSLT</li>
3294	<li>Add support for text node without escaping (XSLT)</li>
3295	<li>bug fixes for xmlCheckFilename</li>
3296	<li>validation code bug fixes from Gary Pennington</li>
3297	<li>Patch from Paul D. Smith correcting URI path normalization</li>
3298	<li>Patch to allow simultaneous install of libxml-devel and
3299	libxml2-devel</li>
3300	<li>the example Makefile is now fixed</li>
3301	<li>added HTML to the RPM packages</li>
3302	<li>tree copying bugfixes</li>
3303	<li>updates to Windows makefiles</li>
3304	<li>optimization patch from Bjorn Reese</li>
3305	</ul>
3306
3307	<h3>2.2.11: Jan 4 2001</h3>
3308	<ul>
3309	<li>bunch of bug fixes (memory I/O, xpath, ftp/http, ...)</li>
3310	<li>added htmlHandleOmittedElem()</li>
3311	<li>Applied Bjorn Reese's IPV6 first patch</li>
3312	<li>Applied Paul D. Smith patches for validation of XInclude results</li>
3313	<li>added XPointer xmlns() new scheme support</li>
3314	</ul>
3315
3316	<h3>2.2.10: Nov 25 2000</h3>
3317	<ul>
3318	<li>Fix the Windows problems of 2.2.8</li>
3319	<li>integrate OpenVMS patches</li>
3320	<li>better handling of some nasty HTML input</li>
3321	<li>Improved the XPointer implementation</li>
3322	<li>integrate a number of provided patches</li>
3323	</ul>
3324
3325	<h3>2.2.9: Nov 25 2000</h3>
3326	<ul>
3327	<li>erroneous release :-(</li>
3328	</ul>
3329
3330	<h3>2.2.8: Nov 13 2000</h3>
3331	<ul>
3332	<li>First version of <a href="http://www.w3.org/TR/xinclude">XInclude</a>
3333	support</li>
3334	<li>Patch in conditional section handling</li>
3335	<li>updated MS compiler project</li>
3336	<li>fixed some XPath problems</li>
3337	<li>added an URI escaping function</li>
3338	<li>some other bug fixes</li>
3339	</ul>
3340
3341	<h3>2.2.7: Oct 31 2000</h3>
3342	<ul>
3343	<li>added message redirection</li>
3344	<li>XPath improvements (thanks TOM !)</li>
3345	<li>xmlIOParseDTD() added</li>
3346	<li>various small fixes in the HTML, URI, HTTP and XPointer support</li>
3347	<li>some cleanup of the Makefile, autoconf and the distribution content</li>
3348	</ul>
3349
3350	<h3>2.2.6: Oct 25 2000:</h3>
3351	<ul>
3352	<li>Added an hash table module, migrated a number of internal structure to
3353	those</li>
3354	<li>Fixed a posteriori validation problems</li>
3355	<li>HTTP module cleanups</li>
3356	<li>HTML parser improvements (tag errors, script/style handling, attribute
3357	normalization)</li>
3358	<li>coalescing of adjacent text nodes</li>
3359	<li>couple of XPath bug fixes, exported the internal API</li>
3360	</ul>
3361
3362	<h3>2.2.5: Oct 15 2000:</h3>
3363	<ul>
3364	<li>XPointer implementation and testsuite</li>
3365	<li>Lot of XPath fixes, added variable and functions registration, more
3366	tests</li>
3367	<li>Portability fixes, lots of enhancements toward an easy Windows build
3368	and release</li>
3369	<li>Late validation fixes</li>
3370	<li>Integrated a lot of contributed patches</li>
3371	<li>added memory management docs</li>
3372	<li>a performance problem when using large buffer seems fixed</li>
3373	</ul>
3374
3375	<h3>2.2.4: Oct 1 2000:</h3>
3376	<ul>
3377	<li>main XPath problem fixed</li>
3378	<li>Integrated portability patches for Windows</li>
3379	<li>Serious bug fixes on the URI and HTML code</li>
3380	</ul>
3381
3382	<h3>2.2.3: Sep 17 2000</h3>
3383	<ul>
3384	<li>bug fixes</li>
3385	<li>cleanup of entity handling code</li>
3386	<li>overall review of all loops in the parsers, all sprintf usage has been
3387	checked too</li>
3388	<li>Far better handling of larges Dtd. Validating against DocBook XML Dtd
3389	works smoothly now.</li>
3390	</ul>
3391
3392	<h3>1.8.10: Sep 6 2000</h3>
3393	<ul>
3394	<li>bug fix release for some Gnome projects</li>
3395	</ul>
3396
3397	<h3>2.2.2: August 12 2000</h3>
3398	<ul>
3399	<li>mostly bug fixes</li>
3400	<li>started adding routines to access xml parser context options</li>
3401	</ul>
3402
3403	<h3>2.2.1: July 21 2000</h3>
3404	<ul>
3405	<li>a purely bug fixes release</li>
3406	<li>fixed an encoding support problem when parsing from a memory block</li>
3407	<li>fixed a DOCTYPE parsing problem</li>
3408	<li>removed a bug in the function allowing to override the memory
3409	allocation routines</li>
3410	</ul>
3411
3412	<h3>2.2.0: July 14 2000</h3>
3413	<ul>
3414	<li>applied a lot of portability fixes</li>
3415	<li>better encoding support/cleanup and saving (content is now always
3416	encoded in UTF-8)</li>
3417	<li>the HTML parser now correctly handles encodings</li>
3418	<li>added xmlHasProp()</li>
3419	<li>fixed a serious problem with &#38;</li>
3420	<li>propagated the fix to FTP client</li>
3421	<li>cleanup, bugfixes, etc ...</li>
3422	<li>Added a page about <a href="encoding.html">libxml Internationalization
3423	support</a></li>
3424	</ul>
3425
3426	<h3>1.8.9: July 9 2000</h3>
3427	<ul>
3428	<li>fixed the spec the RPMs should be better</li>
3429	<li>fixed a serious bug in the FTP implementation, released 1.8.9 to solve
3430	rpmfind users problem</li>
3431	</ul>
3432
3433	<h3>2.1.1: July 1 2000</h3>
3434	<ul>
3435	<li>fixes a couple of bugs in the 2.1.0 packaging</li>
3436	<li>improvements on the HTML parser</li>
3437	</ul>
3438
3439	<h3>2.1.0 and 1.8.8: June 29 2000</h3>
3440	<ul>
3441	<li>1.8.8 is mostly a commodity package for upgrading to libxml2 according
3442	to <a href="upgrade.html">new instructions</a>. It fixes a nasty problem
3443	about &#38; charref parsing</li>
3444	<li>2.1.0 also ease the upgrade from libxml v1 to the recent version. it
3445	also contains numerous fixes and enhancements:
3446	<ul>
3447	<li>added xmlStopParser() to stop parsing</li>
3448	<li>improved a lot parsing speed when there is large CDATA blocs</li>
3449	<li>includes XPath patches provided by Picdar Technology</li>
3450	<li>tried to fix as much as possible DTD validation and namespace
3451	related problems</li>
3452	<li>output to a given encoding has been added/tested</li>
3453	<li>lot of various fixes</li>
3454	</ul>
3455	</li>
3456	</ul>
3457
3458	<h3>2.0.0: Apr 12 2000</h3>
3459	<ul>
3460	<li>First public release of libxml2. If you are using libxml, it's a good
3461	idea to check the 1.x to 2.x upgrade instructions. NOTE: while initially
3462	scheduled for Apr 3 the release occurred only on Apr 12 due to massive
3463	workload.</li>
3464	<li>The include are now located under $prefix/include/libxml (instead of
3465	$prefix/include/gnome-xml), they also are referenced by
3466	<pre>#include <libxml/xxx.h></pre>
3467	<p>instead of</p>
3468	<pre>#include "xxx.h"</pre>
3469	</li>
3470	<li>a new URI module for parsing URIs and following strictly RFC 2396</li>
3471	<li>the memory allocation routines used by libxml can now be overloaded
3472	dynamically by using xmlMemSetup()</li>
3473	<li>The previously CVS only tool tester has been renamed
3474	<strong>xmllint</strong> and is now installed as part of the libxml2
3475	package</li>
3476	<li>The I/O interface has been revamped. There is now ways to plug in
3477	specific I/O modules, either at the URI scheme detection level using
3478	xmlRegisterInputCallbacks() or by passing I/O functions when creating a
3479	parser context using xmlCreateIOParserCtxt()</li>
3480	<li>there is a C preprocessor macro LIBXML_VERSION providing the version
3481	number of the libxml module in use</li>
3482	<li>a number of optional features of libxml can now be excluded at
3483	configure time (FTP/HTTP/HTML/XPath/Debug)</li>
3484	</ul>
3485
3486	<h3>2.0.0beta: Mar 14 2000</h3>
3487	<ul>
3488	<li>This is a first Beta release of libxml version 2</li>
3489	<li>It's available only from<a href="ftp://xmlsoft.org/libxml2/">xmlsoft.org
3490	FTP</a>, it's packaged as libxml2-2.0.0beta and available as tar and
3491	RPMs</li>
3492	<li>This version is now the head in the Gnome CVS base, the old one is
3493	available under the tag LIB_XML_1_X</li>
3494	<li>This includes a very large set of changes. From a programmatic point
3495	of view applications should not have to be modified too much, check the
3496	<a href="upgrade.html">upgrade page</a></li>
3497	<li>Some interfaces may changes (especially a bit about encoding).</li>
3498	<li>the updates includes:
3499	<ul>
3500	<li>fix I18N support. ISO-Latin-x/UTF-8/UTF-16 (nearly) seems correctly
3501	handled now</li>
3502	<li>Better handling of entities, especially well-formedness checking
3503	and proper PEref extensions in external subsets</li>
3504	<li>DTD conditional sections</li>
3505	<li>Validation now correctly handle entities content</li>
3506	<li><a href="http://rpmfind.net/tools/gdome/messages/0039.html">change
3507	structures to accommodate DOM</a></li>
3508	</ul>
3509	</li>
3510	<li>Serious progress were made toward compliance, <a
3511	href="conf/result.html">here are the result of the test</a> against the
3512	OASIS testsuite (except the Japanese tests since I don't support that
3513	encoding yet). This URL is rebuilt every couple of hours using the CVS
3514	head version.</li>
3515	</ul>
3516
3517	<h3>1.8.7: Mar 6 2000</h3>
3518	<ul>
3519	<li>This is a bug fix release:</li>
3520	<li>It is possible to disable the ignorable blanks heuristic used by
3521	libxml-1.x, a new function xmlKeepBlanksDefault(0) will allow this. Note
3522	that for adherence to XML spec, this behaviour will be disabled by
3523	default in 2.x . The same function will allow to keep compatibility for
3524	old code.</li>
3525	<li>Blanks in <a> </a> constructs are not ignored anymore,
3526	avoiding heuristic is really the Right Way :-\</li>
3527	<li>The unchecked use of snprintf which was breaking libxml-1.8.6
3528	compilation on some platforms has been fixed</li>
3529	<li>nanoftp.c nanohttp.c: Fixed '#' and '?' stripping when processing
3530	URIs</li>
3531	</ul>
3532
3533	<h3>1.8.6: Jan 31 2000</h3>
3534	<ul>
3535	<li>added a nanoFTP transport module, debugged until the new version of <a
3536	href="http://rpmfind.net/linux/rpm2html/rpmfind.html">rpmfind</a> can use
3537	it without troubles</li>
3538	</ul>
3539
3540	<h3>1.8.5: Jan 21 2000</h3>
3541	<ul>
3542	<li>adding APIs to parse a well balanced chunk of XML (production <a
3543	href="http://www.w3.org/TR/REC-xml#NT-content">[43] content</a> of the
3544	XML spec)</li>
3545	<li>fixed a hideous bug in xmlGetProp pointed by [email protected]</li>
3546	<li>Jody Goldberg <[email protected]> provided another patch trying
3547	to solve the zlib checks problems</li>
3548	<li>The current state in gnome CVS base is expected to ship as 1.8.5 with
3549	gnumeric soon</li>
3550	</ul>
3551
3552	<h3>1.8.4: Jan 13 2000</h3>
3553	<ul>
3554	<li>bug fixes, reintroduced xmlNewGlobalNs(), fixed xmlNewNs()</li>
3555	<li>all exit() call should have been removed from libxml</li>
3556	<li>fixed a problem with INCLUDE_WINSOCK on WIN32 platform</li>
3557	<li>added newDocFragment()</li>
3558	</ul>
3559
3560	<h3>1.8.3: Jan 5 2000</h3>
3561	<ul>
3562	<li>a Push interface for the XML and HTML parsers</li>
3563	<li>a shell-like interface to the document tree (try tester --shell :-)</li>
3564	<li>lots of bug fixes and improvement added over XMas holidays</li>
3565	<li>fixed the DTD parsing code to work with the xhtml DTD</li>
3566	<li>added xmlRemoveProp(), xmlRemoveID() and xmlRemoveRef()</li>
3567	<li>Fixed bugs in xmlNewNs()</li>
3568	<li>External entity loading code has been revamped, now it uses
3569	xmlLoadExternalEntity(), some fix on entities processing were added</li>
3570	<li>cleaned up WIN32 includes of socket stuff</li>
3571	</ul>
3572
3573	<h3>1.8.2: Dec 21 1999</h3>
3574	<ul>
3575	<li>I got another problem with includes and C++, I hope this issue is fixed
3576	for good this time</li>
3577	<li>Added a few tree modification functions: xmlReplaceNode,
3578	xmlAddPrevSibling, xmlAddNextSibling, xmlNodeSetName and
3579	xmlDocSetRootElement</li>
3580	<li>Tried to improve the HTML output with help from <a
3581	href="mailto:[email protected]">Chris Lahey</a></li>
3582	</ul>
3583
3584	<h3>1.8.1: Dec 18 1999</h3>
3585	<ul>
3586	<li>various patches to avoid troubles when using libxml with C++ compilers
3587	the "namespace" keyword and C escaping in include files</li>
3588	<li>a problem in one of the core macros IS_CHAR was corrected</li>
3589	<li>fixed a bug introduced in 1.8.0 breaking default namespace processing,
3590	and more specifically the Dia application</li>
3591	<li>fixed a posteriori validation (validation after parsing, or by using a
3592	Dtd not specified in the original document)</li>
3593	<li>fixed a bug in</li>
3594	</ul>
3595
3596	<h3>1.8.0: Dec 12 1999</h3>
3597	<ul>
3598	<li>cleanup, especially memory wise</li>
3599	<li>the parser should be more reliable, especially the HTML one, it should
3600	not crash, whatever the input !</li>
3601	<li>Integrated various patches, especially a speedup improvement for large
3602	dataset from <a href="mailto:[email protected]">Carl Nygard</a>,
3603	configure with --with-buffers to enable them.</li>
3604	<li>attribute normalization, oops should have been added long ago !</li>
3605	<li>attributes defaulted from DTDs should be available, xmlSetProp() now
3606	does entities escaping by default.</li>
3607	</ul>
3608
3609	<h3>1.7.4: Oct 25 1999</h3>
3610	<ul>
3611	<li>Lots of HTML improvement</li>
3612	<li>Fixed some errors when saving both XML and HTML</li>
3613	<li>More examples, the regression tests should now look clean</li>
3614	<li>Fixed a bug with contiguous charref</li>
3615	</ul>
3616
3617	<h3>1.7.3: Sep 29 1999</h3>
3618	<ul>
3619	<li>portability problems fixed</li>
3620	<li>snprintf was used unconditionally, leading to link problems on system
3621	were it's not available, fixed</li>
3622	</ul>
3623
3624	<h3>1.7.1: Sep 24 1999</h3>
3625	<ul>
3626	<li>The basic type for strings manipulated by libxml has been renamed in
3627	1.7.1 from <strong>CHAR</strong> to <strong>xmlChar</strong>. The reason
3628	is that CHAR was conflicting with a predefined type on Windows. However
3629	on non WIN32 environment, compatibility is provided by the way of a
3630	<strong>#define </strong>.</li>
3631	<li>Changed another error : the use of a structure field called errno, and
3632	leading to troubles on platforms where it's a macro</li>
3633	</ul>
3634
3635	<h3>1.7.0: Sep 23 1999</h3>
3636	<ul>
3637	<li>Added the ability to fetch remote DTD or parsed entities, see the <a
3638	href="html/libxml-nanohttp.html">nanohttp</a> module.</li>
3639	<li>Added an errno to report errors by another mean than a simple printf
3640	like callback</li>
3641	<li>Finished ID/IDREF support and checking when validation</li>
3642	<li>Serious memory leaks fixed (there is now a <a
3643	href="html/libxml-xmlmemory.html">memory wrapper</a> module)</li>
3644	<li>Improvement of <a href="http://www.w3.org/TR/xpath">XPath</a>
3645	implementation</li>
3646	<li>Added an HTML parser front-end</li>
3647	</ul>
3648
3649	<h2><a name="XML">XML</a></h2>
3650
3651	<p><a href="http://www.w3.org/TR/REC-xml">XML is a standard</a> for
3652	markup-based structured documents. Here is <a name="example">an example XML
3653	document</a>:</p>
3654	<pre><?xml version="1.0"?>
3655	<EXAMPLE prop1="gnome is great" prop2="&amp; linux too">
3656	<head>
3657	<title>Welcome to Gnome</title>
3658	</head>
3659	<chapter>
3660	<title>The Linux adventure</title>
3661	<p>bla bla bla ...</p>
3662	<image href="linus.gif"/>
3663	<p>...</p>
3664	</chapter>
3665	</EXAMPLE></pre>
3666
3667	<p>The first line specifies that it is an XML document and gives useful
3668	information about its encoding. Then the rest of the document is a text
3669	format whose structure is specified by tags between brackets. <strong>Each
3670	tag opened has to be closed</strong>. XML is pedantic about this. However, if
3671	a tag is empty (no content), a single tag can serve as both the opening and
3672	closing tag if it ends with <code>/></code> rather than with
3673	<code>></code>. Note that, for example, the image tag has no content (just
3674	an attribute) and is closed by ending the tag with <code>/></code>.</p>
3675
3676	<p>XML can be applied successfully to a wide range of tasks, ranging from
3677	long term structured document maintenance (where it follows the steps of
3678	SGML) to simple data encoding mechanisms like configuration file formatting
3679	(glade), spreadsheets (gnumeric), or even shorter lived documents such as
3680	WebDAV where it is used to encode remote calls between a client and a
3681	server.</p>
3682
3683	<h2><a name="XSLT">XSLT</a></h2>
3684
3685	<p>Check <a href="http://xmlsoft.org/XSLT">the separate libxslt page</a></p>
3686
3687	<p><a href="http://www.w3.org/TR/xslt">XSL Transformations</a>, is a
3688	language for transforming XML documents into other XML documents (or
3689	HTML/textual output).</p>
3690
3691	<p>A separate library called libxslt is available implementing XSLT-1.0 for
3692	libxml2. This module "libxslt" too can be found in the Gnome SVN base.</p>
3693
3694	<p>You can check the progresses on the libxslt <a
3695	href="http://xmlsoft.org/XSLT/ChangeLog.html">Changelog</a>.</p>
3696
3697	<h2><a name="Python">Python and bindings</a></h2>
3698
3699	<p>There are a number of language bindings and wrappers available for
3700	libxml2, the list below is not exhaustive. Please contact the <a
3701	href="http://mail.gnome.org/mailman/listinfo/xml-bindings">[email protected]</a>
3702	(<a href="http://mail.gnome.org/archives/xml-bindings/">archives</a>) in
3703	order to get updates to this list or to discuss the specific topic of libxml2
3704	or libxslt wrappers or bindings:</p>
3705	<ul>
3706	<li><a href="http://libxmlplusplus.sourceforge.net/">Libxml++</a> seems the
3707	most up-to-date C++ bindings for libxml2, check the <a
3708	href="http://libxmlplusplus.sourceforge.net/reference/html/hierarchy.html">documentation</a>
3709	and the <a
3710	href="http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/libxmlplusplus/libxml%2b%2b/examples/">examples</a>.</li>
3711	<li>There is another <a href="http://libgdome-cpp.berlios.de/">C++ wrapper
3712	based on the gdome2 bindings</a> maintained by Tobias Peters.</li>
3713	<li>and a third C++ wrapper by Peter Jones <[email protected]>
3714	<p>Website: <a
3715	href="http://pmade.org/pjones/software/xmlwrapp/">http://pmade.org/pjones/software/xmlwrapp/</a></p>
3716	</li>
3717	<li>XML::LibXML <a href="http://cpan.uwinnipeg.ca/dist/XML-LibXML">Perl
3718	bindings</a> are available on CPAN, as well as XML::LibXSLT
3719	<a href="http://cpan.uwinnipeg.ca/dist/XML-LibXSLT">Perl libxslt
3720	bindings</a>.</li>
3721	<li>If you're interested into scripting XML processing, have a look at <a
3722	href="http://xsh.sourceforge.net/">XSH</a> an XML editing shell based on
3723	Libxml2 Perl bindings.</li>
3724	<li><a href="mailto:[email protected]">Dave Kuhlman</a> provides an
3725	earlier version of the libxml/libxslt <a
3726	href="http://www.rexx.com/~dkuhlman">wrappers for Python</a>.</li>
3727	<li>Gopal.V and Peter Minten develop <a
3728	href="http://savannah.gnu.org/projects/libxmlsharp">libxml#</a>, a set of
3729	C# libxml2 bindings.</li>
3730	<li>Petr Kozelka provides <a
3731	href="http://sourceforge.net/projects/libxml2-pas">Pascal units to glue
3732	libxml2</a> with Kylix, Delphi and other Pascal compilers.</li>
3733	<li>Uwe Fechner also provides <a
3734	href="http://sourceforge.net/projects/idom2-pas/">idom2</a>, a DOM2
3735	implementation for Kylix2/D5/D6 from Borland.</li>
3736	<li>There is <a href="http://libxml.rubyforge.org/">bindings for Ruby</a>
3737	and libxml2 bindings are also available in Ruby through the <a
3738	href="http://libgdome-ruby.berlios.de/">libgdome-ruby</a> module
3739	maintained by Tobias Peters.</li>
3740	<li>Steve Ball and contributors maintains <a
3741	href="http://tclxml.sourceforge.net/">libxml2 and libxslt bindings for
3742	Tcl</a>.</li>
3743	<li>libxml2 and libxslt are the default XML libraries for PHP5.</li>
3744	<li><a href="http://savannah.gnu.org/projects/classpathx/">LibxmlJ</a> is
3745	an effort to create a 100% JAXP-compatible Java wrapper for libxml2 and
3746	libxslt as part of GNU ClasspathX project.</li>
3747	<li>Patrick McPhee provides Rexx bindings fof libxml2 and libxslt, look for
3748	<a href="http://www.interlog.com/~ptjm/software.html">RexxXML</a>.</li>
3749	<li><a
3750	href="http://www.satimage.fr/software/en/xml_suite.html">Satimage</a>
3751	provides <a
3752	href="http://www.satimage.fr/software/en/downloads_osaxen.html">XMLLib
3753	osax</a>. This is an osax for Mac OS X with a set of commands to
3754	implement in AppleScript the XML DOM, XPATH and XSLT. Also includes
3755	commands for Property-lists (Apple's fast lookup table XML format.)</li>
3756	<li>Francesco Montorsi developped <a
3757	href="https://sourceforge.net/project/showfiles.php?group_id=51305&package_id=45182">wxXml2</a>
3758	wrappers that interface libxml2, allowing wxWidgets applications to
3759	load/save/edit XML instances.</li>
3760	</ul>
3761
3762	<p>The distribution includes a set of Python bindings, which are guaranteed
3763	to be maintained as part of the library in the future, though the Python
3764	interface have not yet reached the completeness of the C API.</p>
3765
3766	<p>Note that some of the Python purist dislike the default set of Python
3767	bindings, rather than complaining I suggest they have a look at <a
3768	href="http://lxml.de/">lxml the more pythonic bindings for libxml2
3769	and libxslt</a> and <a
3770	href="http://lxml.de/mailinglist/">check the mailing-list</a>.</p>
3771
3772	<p><a href="mailto:[email protected]">StÃ©phane Bidoul</a>
3773	maintains <a href="http://users.skynet.be/sbi/libxml-python/">a Windows port
3774	of the Python bindings</a>.</p>
3775
3776	<p>Note to people interested in building bindings, the API is formalized as
3777	<a href="libxml2-api.xml">an XML API description file</a> which allows to
3778	automate a large part of the Python bindings, this includes function
3779	descriptions, enums, structures, typedefs, etc... The Python script used to
3780	build the bindings is python/generator.py in the source distribution.</p>
3781
3782	<p>To install the Python bindings there are 2 options:</p>
3783	<ul>
3784	<li>If you use an RPM based distribution, simply install the <a
3785	href="http://rpmfind.net/linux/rpm2html/search.php?query=libxml2-python">libxml2-python
3786	RPM</a> (and if needed the <a
3787	href="http://rpmfind.net/linux/rpm2html/search.php?query=libxslt-python">libxslt-python
3788	RPM</a>).</li>
3789	<li>Otherwise use the <a href="ftp://xmlsoft.org/libxml2/python/">libxml2-python
3790	module distribution</a> corresponding to your installed version of
3791	libxml2 and libxslt. Note that to install it you will need both libxml2
3792	and libxslt installed and run "python setup.py build install" in the
3793	module tree.</li>
3794	</ul>
3795
3796	<p>The distribution includes a set of examples and regression tests for the
3797	python bindings in the <code>python/tests</code> directory. Here are some
3798	excerpts from those tests:</p>
3799
3800	<h3>tst.py:</h3>
3801
3802	<p>This is a basic test of the file interface and DOM navigation:</p>
3803	<pre>import libxml2, sys
3804
3805	doc = libxml2.parseFile("tst.xml")
3806	if doc.name != "tst.xml":
3807	print "doc.name failed"
3808	sys.exit(1)
3809	root = doc.children
3810	if root.name != "doc":
3811	print "root.name failed"
3812	sys.exit(1)
3813	child = root.children
3814	if child.name != "foo":
3815	print "child.name failed"
3816	sys.exit(1)
3817	doc.freeDoc()</pre>
3818
3819	<p>The Python module is called libxml2; parseFile is the equivalent of
3820	xmlParseFile (most of the bindings are automatically generated, and the xml
3821	prefix is removed and the casing convention are kept). All node seen at the
3822	binding level share the same subset of accessors:</p>
3823	<ul>
3824	<li><code>name</code> : returns the node name</li>
3825	<li><code>type</code> : returns a string indicating the node type</li>
3826	<li><code>content</code> : returns the content of the node, it is based on
3827	xmlNodeGetContent() and hence is recursive.</li>
3828	<li><code>parent</code> , <code>children</code>, <code>last</code>,
3829	<code>next</code>, <code>prev</code>, <code>doc</code>,
3830	<code>properties</code>: pointing to the associated element in the tree,
3831	those may return None in case no such link exists.</li>
3832	</ul>
3833
3834	<p>Also note the need to explicitly deallocate documents with freeDoc() .
3835	Reference counting for libxml2 trees would need quite a lot of work to
3836	function properly, and rather than risk memory leaks if not implemented
3837	correctly it sounds safer to have an explicit function to free a tree. The
3838	wrapper python objects like doc, root or child are them automatically garbage
3839	collected.</p>
3840
3841	<h3>validate.py:</h3>
3842
3843	<p>This test check the validation interfaces and redirection of error
3844	messages:</p>
3845	<pre>import libxml2
3846
3847	#deactivate error messages from the validation
3848	def noerr(ctx, str):
3849	pass
3850
3851	libxml2.registerErrorHandler(noerr, None)
3852
3853	ctxt = libxml2.createFileParserCtxt("invalid.xml")
3854	ctxt.validate(1)
3855	ctxt.parseDocument()
3856	doc = ctxt.doc()
3857	valid = ctxt.isValid()
3858	doc.freeDoc()
3859	if valid != 0:
3860	print "validity check failed"</pre>
3861
3862	<p>The first thing to notice is the call to registerErrorHandler(), it
3863	defines a new error handler global to the library. It is used to avoid seeing
3864	the error messages when trying to validate the invalid document.</p>
3865
3866	<p>The main interest of that test is the creation of a parser context with
3867	createFileParserCtxt() and how the behaviour can be changed before calling
3868	parseDocument() . Similarly the information resulting from the parsing phase
3869	is also available using context methods.</p>
3870
3871	<p>Contexts like nodes are defined as class and the libxml2 wrappers maps the
3872	C function interfaces in terms of objects method as much as possible. The
3873	best to get a complete view of what methods are supported is to look at the
3874	libxml2.py module containing all the wrappers.</p>
3875
3876	<h3>push.py:</h3>
3877
3878	<p>This test show how to activate the push parser interface:</p>
3879	<pre>import libxml2
3880
3881	ctxt = libxml2.createPushParser(None, "<foo", 4, "test.xml")
3882	ctxt.parseChunk("/>", 2, 1)
3883	doc = ctxt.doc()
3884
3885	doc.freeDoc()</pre>
3886
3887	<p>The context is created with a special call based on the
3888	xmlCreatePushParser() from the C library. The first argument is an optional
3889	SAX callback object, then the initial set of data, the length and the name of
3890	the resource in case URI-References need to be computed by the parser.</p>
3891
3892	<p>Then the data are pushed using the parseChunk() method, the last call
3893	setting the third argument terminate to 1.</p>
3894
3895	<h3>pushSAX.py:</h3>
3896
3897	<p>this test show the use of the event based parsing interfaces. In this case
3898	the parser does not build a document, but provides callback information as
3899	the parser makes progresses analyzing the data being provided:</p>
3900	<pre>import libxml2
3901	log = ""
3902
3903	class callback:
3904	def startDocument(self):
3905	global log
3906	log = log + "startDocument:"
3907
3908	def endDocument(self):
3909	global log
3910	log = log + "endDocument:"
3911
3912	def startElement(self, tag, attrs):
3913	global log
3914	log = log + "startElement %s %s:" % (tag, attrs)
3915
3916	def endElement(self, tag):
3917	global log
3918	log = log + "endElement %s:" % (tag)
3919
3920	def characters(self, data):
3921	global log
3922	log = log + "characters: %s:" % (data)
3923
3924	def warning(self, msg):
3925	global log
3926	log = log + "warning: %s:" % (msg)
3927
3928	def error(self, msg):
3929	global log
3930	log = log + "error: %s:" % (msg)
3931
3932	def fatalError(self, msg):
3933	global log
3934	log = log + "fatalError: %s:" % (msg)
3935
3936	handler = callback()
3937
3938	ctxt = libxml2.createPushParser(handler, "<foo", 4, "test.xml")
3939	chunk = " url='tst'>b"
3940	ctxt.parseChunk(chunk, len(chunk), 0)
3941	chunk = "ar</foo>"
3942	ctxt.parseChunk(chunk, len(chunk), 1)
3943
3944	reference = "startDocument:startElement foo {'url': 'tst'}:" + \
3945	"characters: bar:endElement foo:endDocument:"
3946	if log != reference:
3947	print "Error got: %s" % log
3948	print "Expected: %s" % reference</pre>
3949
3950	<p>The key object in that test is the handler, it provides a number of entry
3951	points which can be called by the parser as it makes progresses to indicate
3952	the information set obtained. The full set of callback is larger than what
3953	the callback class in that specific example implements (see the SAX
3954	definition for a complete list). The wrapper will only call those supplied by
3955	the object when activated. The startElement receives the names of the element
3956	and a dictionary containing the attributes carried by this element.</p>
3957
3958	<p>Also note that the reference string generated from the callback shows a
3959	single character call even though the string "bar" is passed to the parser
3960	from 2 different call to parseChunk()</p>
3961
3962	<h3>xpath.py:</h3>
3963
3964	<p>This is a basic test of XPath wrappers support</p>
3965	<pre>import libxml2
3966
3967	doc = libxml2.parseFile("tst.xml")
3968	ctxt = doc.xpathNewContext()
3969	res = ctxt.xpathEval("//*")
3970	if len(res) != 2:
3971	print "xpath query: wrong node set size"
3972	sys.exit(1)
3973	if res[0].name != "doc" or res[1].name != "foo":
3974	print "xpath query: wrong node set value"
3975	sys.exit(1)
3976	doc.freeDoc()
3977	ctxt.xpathFreeContext()</pre>
3978
3979	<p>This test parses a file, then create an XPath context to evaluate XPath
3980	expression on it. The xpathEval() method execute an XPath query and returns
3981	the result mapped in a Python way. String and numbers are natively converted,
3982	and node sets are returned as a tuple of libxml2 Python nodes wrappers. Like
3983	the document, the XPath context need to be freed explicitly, also not that
3984	the result of the XPath query may point back to the document tree and hence
3985	the document must be freed after the result of the query is used.</p>
3986
3987	<h3>xpathext.py:</h3>
3988
3989	<p>This test shows how to extend the XPath engine with functions written in
3990	python:</p>
3991	<pre>import libxml2
3992
3993	def foo(ctx, x):
3994	return x + 1
3995
3996	doc = libxml2.parseFile("tst.xml")
3997	ctxt = doc.xpathNewContext()
3998	libxml2.registerXPathFunction(ctxt._o, "foo", None, foo)
3999	res = ctxt.xpathEval("foo(1)")
4000	if res != 2:
4001	print "xpath extension failure"
4002	doc.freeDoc()
4003	ctxt.xpathFreeContext()</pre>
4004
4005	<p>Note how the extension function is registered with the context (but that
4006	part is not yet finalized, this may change slightly in the future).</p>
4007
4008	<h3>tstxpath.py:</h3>
4009
4010	<p>This test is similar to the previous one but shows how the extension
4011	function can access the XPath evaluation context:</p>
4012	<pre>def foo(ctx, x):
4013	global called
4014
4015	#
4016	# test that access to the XPath evaluation contexts
4017	#
4018	pctxt = libxml2.xpathParserContext(_obj=ctx)
4019	ctxt = pctxt.context()
4020	called = ctxt.function()
4021	return x + 1</pre>
4022
4023	<p>All the interfaces around the XPath parser(or rather evaluation) context
4024	are not finalized, but it should be sufficient to do contextual work at the
4025	evaluation point.</p>
4026
4027	<h3>Memory debugging:</h3>
4028
4029	<p>last but not least, all tests starts with the following prologue:</p>
4030	<pre>#memory debug specific
4031	libxml2.debugMemory(1)</pre>
4032
4033	<p>and ends with the following epilogue:</p>
4034	<pre>#memory debug specific
4035	libxml2.cleanupParser()
4036	if libxml2.debugMemory(1) == 0:
4037	print "OK"
4038	else:
4039	print "Memory leak %d bytes" % (libxml2.debugMemory(1))
4040	libxml2.dumpMemory()</pre>
4041
4042	<p>Those activate the memory debugging interface of libxml2 where all
4043	allocated block in the library are tracked. The prologue then cleans up the
4044	library state and checks that all allocated memory has been freed. If not it
4045	calls dumpMemory() which saves that list in a <code>.memdump</code> file.</p>
4046
4047	<h2><a name="architecture">libxml2 architecture</a></h2>
4048
4049	<p>Libxml2 is made of multiple components; some of them are optional, and
4050	most of the block interfaces are public. The main components are:</p>
4051	<ul>
4052	<li>an Input/Output layer</li>
4053	<li>FTP and HTTP client layers (optional)</li>
4054	<li>an Internationalization layer managing the encodings support</li>
4055	<li>a URI module</li>
4056	<li>the XML parser and its basic SAX interface</li>
4057	<li>an HTML parser using the same SAX interface (optional)</li>
4058	<li>a SAX tree module to build an in-memory DOM representation</li>
4059	<li>a tree module to manipulate the DOM representation</li>
4060	<li>a validation module using the DOM representation (optional)</li>
4061	<li>an XPath module for global lookup in a DOM representation
4062	(optional)</li>
4063	<li>a debug module (optional)</li>
4064	</ul>
4065
4066	<p>Graphically this gives the following:</p>
4067
4068	<p><img src="libxml.gif" alt="a graphical view of the various"></p>
4069
4070	<p></p>
4071
4072	<h2><a name="tree">The tree output</a></h2>
4073
4074	<p>The parser returns a tree built during the document analysis. The value
4075	returned is an <strong>xmlDocPtr</strong> (i.e., a pointer to an
4076	<strong>xmlDoc</strong> structure). This structure contains information such
4077	as the file name, the document type, and a <strong>children</strong> pointer
4078	which is the root of the document (or more exactly the first child under the
4079	root which is the document). The tree is made of <strong>xmlNode</strong>s,
4080	chained in double-linked lists of siblings and with a children<->parent
4081	relationship. An xmlNode can also carry properties (a chain of xmlAttr
4082	structures). An attribute may have a value which is a list of TEXT or
4083	ENTITY_REF nodes.</p>
4084
4085	<p>Here is an example (erroneous with respect to the XML spec since there
4086	should be only one ELEMENT under the root):</p>
4087
4088	<p><img src="structure.gif" alt=" structure.gif "></p>
4089
4090	<p>In the source package there is a small program (not installed by default)
4091	called <strong>xmllint</strong> which parses XML files given as argument and
4092	prints them back as parsed. This is useful for detecting errors both in XML
4093	code and in the XML parser itself. It has an option <strong>--debug</strong>
4094	which prints the actual in-memory structure of the document; here is the
4095	result with the <a href="#example">example</a> given before:</p>
4096	<pre>DOCUMENT
4097	version=1.0
4098	standalone=true
4099	ELEMENT EXAMPLE
4100	ATTRIBUTE prop1
4101	TEXT
4102	content=gnome is great
4103	ATTRIBUTE prop2
4104	ENTITY_REF
4105	TEXT
4106	content= linux too
4107	ELEMENT head
4108	ELEMENT title
4109	TEXT
4110	content=Welcome to Gnome
4111	ELEMENT chapter
4112	ELEMENT title
4113	TEXT
4114	content=The Linux adventure
4115	ELEMENT p
4116	TEXT
4117	content=bla bla bla ...
4118	ELEMENT image
4119	ATTRIBUTE href
4120	TEXT
4121	content=linus.gif
4122	ELEMENT p
4123	TEXT
4124	content=...</pre>
4125
4126	<p>This should be useful for learning the internal representation model.</p>
4127
4128	<h2><a name="interface">The SAX interface</a></h2>
4129
4130	<p>Sometimes the DOM tree output is just too large to fit reasonably into
4131	memory. In that case (and if you don't expect to save back the XML document
4132	loaded using libxml), it's better to use the SAX interface of libxml. SAX is
4133	a <strong>callback-based interface</strong> to the parser. Before parsing,
4134	the application layer registers a customized set of callbacks which are
4135	called by the library as it progresses through the XML input.</p>
4136
4137	<p>To get more detailed step-by-step guidance on using the SAX interface of
4138	libxml, see the <a
4139	href="http://www.jamesh.id.au/articles/libxml-sax/libxml-sax.html">nice
4140	documentation</a>.written by <a href="mailto:[email protected]">James
4141	Henstridge</a>.</p>
4142
4143	<p>You can debug the SAX behaviour by using the <strong>testSAX</strong>
4144	program located in the gnome-xml module (it's usually not shipped in the
4145	binary packages of libxml, but you can find it in the tar source
4146	distribution). Here is the sequence of callbacks that would be reported by
4147	testSAX when parsing the example XML document shown earlier:</p>
4148	<pre>SAX.setDocumentLocator()
4149	SAX.startDocument()
4150	SAX.getEntity(amp)
4151	SAX.startElement(EXAMPLE, prop1='gnome is great', prop2='&amp; linux too')
4152	SAX.characters( , 3)
4153	SAX.startElement(head)
4154	SAX.characters( , 4)
4155	SAX.startElement(title)
4156	SAX.characters(Welcome to Gnome, 16)
4157	SAX.endElement(title)
4158	SAX.characters( , 3)
4159	SAX.endElement(head)
4160	SAX.characters( , 3)
4161	SAX.startElement(chapter)
4162	SAX.characters( , 4)
4163	SAX.startElement(title)
4164	SAX.characters(The Linux adventure, 19)
4165	SAX.endElement(title)
4166	SAX.characters( , 4)
4167	SAX.startElement(p)
4168	SAX.characters(bla bla bla ..., 15)
4169	SAX.endElement(p)
4170	SAX.characters( , 4)
4171	SAX.startElement(image, href='linus.gif')
4172	SAX.endElement(image)
4173	SAX.characters( , 4)
4174	SAX.startElement(p)
4175	SAX.characters(..., 3)
4176	SAX.endElement(p)
4177	SAX.characters( , 3)
4178	SAX.endElement(chapter)
4179	SAX.characters( , 1)
4180	SAX.endElement(EXAMPLE)
4181	SAX.endDocument()</pre>
4182
4183	<p>Most of the other interfaces of libxml2 are based on the DOM tree-building
4184	facility, so nearly everything up to the end of this document presupposes the
4185	use of the standard DOM tree build. Note that the DOM tree itself is built by
4186	a set of registered default callbacks, without internal specific
4187	interface.</p>
4188
4189	<h2><a name="Validation">Validation & DTDs</a></h2>
4190
4191	<p>Table of Content:</p>
4192	<ol>
4193	<li><a href="#General5">General overview</a></li>
4194	<li><a href="#definition">The definition</a></li>
4195	<li><a href="#Simple">Simple rules</a>
4196	<ol>
4197	<li><a href="#reference">How to reference a DTD from a document</a></li>
4198	<li><a href="#Declaring">Declaring elements</a></li>
4199	<li><a href="#Declaring1">Declaring attributes</a></li>
4200	</ol>
4201	</li>
4202	<li><a href="#Some">Some examples</a></li>
4203	<li><a href="#validate">How to validate</a></li>
4204	<li><a href="#Other">Other resources</a></li>
4205	</ol>
4206
4207	<h3><a name="General5">General overview</a></h3>
4208
4209	<p>Well what is validation and what is a DTD ?</p>
4210
4211	<p>DTD is the acronym for Document Type Definition. This is a description of
4212	the content for a family of XML files. This is part of the XML 1.0
4213	specification, and allows one to describe and verify that a given document
4214	instance conforms to the set of rules detailing its structure and content.</p>
4215
4216	<p>Validation is the process of checking a document against a DTD (more
4217	generally against a set of construction rules).</p>
4218
4219	<p>The validation process and building DTDs are the two most difficult parts
4220	of the XML life cycle. Briefly a DTD defines all the possible elements to be
4221	found within your document, what is the formal shape of your document tree
4222	(by defining the allowed content of an element; either text, a regular
4223	expression for the allowed list of children, or mixed content i.e. both text
4224	and children). The DTD also defines the valid attributes for all elements and
4225	the types of those attributes.</p>
4226
4227	<h3><a name="definition1">The definition</a></h3>
4228
4229	<p>The <a href="http://www.w3.org/TR/REC-xml">W3C XML Recommendation</a> (<a
4230	href="http://www.xml.com/axml/axml.html">Tim Bray's annotated version of
4231	Rev1</a>):</p>
4232	<ul>
4233	<li><a href="http://www.w3.org/TR/REC-xml#elemdecls">Declaring
4234	elements</a></li>
4235	<li><a href="http://www.w3.org/TR/REC-xml#attdecls">Declaring
4236	attributes</a></li>
4237	</ul>
4238
4239	<p>(unfortunately) all this is inherited from the SGML world, the syntax is
4240	ancient...</p>
4241
4242	<h3><a name="Simple1">Simple rules</a></h3>
4243
4244	<p>Writing DTDs can be done in many ways. The rules to build them if you need
4245	something permanent or something which can evolve over time can be radically
4246	different. Really complex DTDs like DocBook ones are flexible but quite
4247	harder to design. I will just focus on DTDs for a formats with a fixed simple
4248	structure. It is just a set of basic rules, and definitely not exhaustive nor
4249	usable for complex DTD design.</p>
4250
4251	<h4><a name="reference1">How to reference a DTD from a document</a>:</h4>
4252
4253	<p>Assuming the top element of the document is <code>spec</code> and the dtd
4254	is placed in the file <code>mydtd</code> in the subdirectory
4255	<code>dtds</code> of the directory from where the document were loaded:</p>
4256
4257	<p><code><!DOCTYPE spec SYSTEM "dtds/mydtd"></code></p>
4258
4259	<p>Notes:</p>
4260	<ul>
4261	<li>The system string is actually an URI-Reference (as defined in <a
4262	href="http://www.ietf.org/rfc/rfc2396.txt">RFC 2396</a>) so you can use a
4263	full URL string indicating the location of your DTD on the Web. This is a
4264	really good thing to do if you want others to validate your document.</li>
4265	<li>It is also possible to associate a <code>PUBLIC</code> identifier (a
4266	magic string) so that the DTD is looked up in catalogs on the client side
4267	without having to locate it on the web.</li>
4268	<li>A DTD contains a set of element and attribute declarations, but they
4269	don't define what the root of the document should be. This is explicitly
4270	told to the parser/validator as the first element of the
4271	<code>DOCTYPE</code> declaration.</li>
4272	</ul>
4273
4274	<h4><a name="Declaring2">Declaring elements</a>:</h4>
4275
4276	<p>The following declares an element <code>spec</code>:</p>
4277
4278	<p><code><!ELEMENT spec (front, body, back?)></code></p>
4279
4280	<p>It also expresses that the spec element contains one <code>front</code>,
4281	one <code>body</code> and one optional <code>back</code> children elements in
4282	this order. The declaration of one element of the structure and its content
4283	are done in a single declaration. Similarly the following declares
4284	<code>div1</code> elements:</p>
4285
4286	<p><code><!ELEMENT div1 (head, (p \| list \| note)*, div2?)></code></p>
4287
4288	<p>which means div1 contains one <code>head</code> then a series of optional
4289	<code>p</code>, <code>list</code>s and <code>note</code>s and then an
4290	optional <code>div2</code>. And last but not least an element can contain
4291	text:</p>
4292
4293	<p><code><!ELEMENT b (#PCDATA)></code></p>
4294
4295	<p><code>b</code> contains text or being of mixed content (text and elements
4296	in no particular order):</p>
4297
4298	<p><code><!ELEMENT p (#PCDATA\|a\|ul\|b\|i\|em)*></code></p>
4299
4300	<p><code>p </code>can contain text or <code>a</code>, <code>ul</code>,
4301	<code>b</code>, <code>i </code>or <code>em</code> elements in no particular
4302	order.</p>
4303
4304	<h4><a name="Declaring1">Declaring attributes</a>:</h4>
4305
4306	<p>Again the attributes declaration includes their content definition:</p>
4307
4308	<p><code><!ATTLIST termdef name CDATA #IMPLIED></code></p>
4309
4310	<p>means that the element <code>termdef</code> can have a <code>name</code>
4311	attribute containing text (<code>CDATA</code>) and which is optional
4312	(<code>#IMPLIED</code>). The attribute value can also be defined within a
4313	set:</p>
4314
4315	<p><code><!ATTLIST list type (bullets\|ordered\|glossary)
4316	"ordered"></code></p>
4317
4318	<p>means <code>list</code> element have a <code>type</code> attribute with 3
4319	allowed values "bullets", "ordered" or "glossary" and which default to
4320	"ordered" if the attribute is not explicitly specified.</p>
4321
4322	<p>The content type of an attribute can be text (<code>CDATA</code>),
4323	anchor/reference/references
4324	(<code>ID</code>/<code>IDREF</code>/<code>IDREFS</code>), entity(ies)
4325	(<code>ENTITY</code>/<code>ENTITIES</code>) or name(s)
4326	(<code>NMTOKEN</code>/<code>NMTOKENS</code>). The following defines that a
4327	<code>chapter</code> element can have an optional <code>id</code> attribute
4328	of type <code>ID</code>, usable for reference from attribute of type
4329	IDREF:</p>
4330
4331	<p><code><!ATTLIST chapter id ID #IMPLIED></code></p>
4332
4333	<p>The last value of an attribute definition can be <code>#REQUIRED
4334	</code>meaning that the attribute has to be given, <code>#IMPLIED</code>
4335	meaning that it is optional, or the default value (possibly prefixed by
4336	<code>#FIXED</code> if it is the only allowed).</p>
4337
4338	<p>Notes:</p>
4339	<ul>
4340	<li>Usually the attributes pertaining to a given element are declared in a
4341	single expression, but it is just a convention adopted by a lot of DTD
4342	writers:
4343	<pre><!ATTLIST termdef
4344	id ID #REQUIRED
4345	name CDATA #IMPLIED></pre>
4346	<p>The previous construct defines both <code>id</code> and
4347	<code>name</code> attributes for the element <code>termdef</code>.</p>
4348	</li>
4349	</ul>
4350
4351	<h3><a name="Some1">Some examples</a></h3>
4352
4353	<p>The directory <code>test/valid/dtds/</code> in the libxml2 distribution
4354	contains some complex DTD examples. The example in the file
4355	<code>test/valid/dia.xml</code> shows an XML file where the simple DTD is
4356	directly included within the document.</p>
4357
4358	<h3><a name="validate1">How to validate</a></h3>
4359
4360	<p>The simplest way is to use the xmllint program included with libxml. The
4361	<code>--valid</code> option turns-on validation of the files given as input.
4362	For example the following validates a copy of the first revision of the XML
4363	1.0 specification:</p>
4364
4365	<p><code>xmllint --valid --noout test/valid/REC-xml-19980210.xml</code></p>
4366
4367	<p>the -- noout is used to disable output of the resulting tree.</p>
4368
4369	<p>The <code>--dtdvalid dtd</code> allows validation of the document(s)
4370	against a given DTD.</p>
4371
4372	<p>Libxml2 exports an API to handle DTDs and validation, check the <a
4373	href="http://xmlsoft.org/html/libxml-valid.html">associated
4374	description</a>.</p>
4375
4376	<h3><a name="Other1">Other resources</a></h3>
4377
4378	<p>DTDs are as old as SGML. So there may be a number of examples on-line, I
4379	will just list one for now, others pointers welcome:</p>
4380	<ul>
4381	<li><a href="http://www.xml101.com:8081/dtd/">XML-101 DTD</a></li>
4382	</ul>
4383
4384	<p>I suggest looking at the examples found under test/valid/dtd and any of
4385	the large number of books available on XML. The dia example in test/valid
4386	should be both simple and complete enough to allow you to build your own.</p>
4387
4388	<p></p>
4389
4390	<h2><a name="Memory">Memory Management</a></h2>
4391
4392	<p>Table of Content:</p>
4393	<ol>
4394	<li><a href="#General3">General overview</a></li>
4395	<li><a href="#setting">Setting libxml2 set of memory routines</a></li>
4396	<li><a href="#cleanup">Cleaning up after using the library</a></li>
4397	<li><a href="#Debugging">Debugging routines</a></li>
4398	<li><a href="#General4">General memory requirements</a></li>
4399	<li><a href="#Compacting">Returning memory to the kernel</a></li>
4400	</ol>
4401
4402	<h3><a name="General3">General overview</a></h3>
4403
4404	<p>The module <code><a
4405	href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlmemory.h</a></code>
4406	provides the interfaces to the libxml2 memory system:</p>
4407	<ul>
4408	<li>libxml2 does not use the libc memory allocator directly but xmlFree(),
4409	xmlMalloc() and xmlRealloc()</li>
4410	<li>those routines can be reallocated to a specific set of routine, by
4411	default the libc ones i.e. free(), malloc() and realloc()</li>
4412	<li>the xmlmemory.c module includes a set of debugging routine</li>
4413	</ul>
4414
4415	<h3><a name="setting">Setting libxml2 set of memory routines</a></h3>
4416
4417	<p>It is sometimes useful to not use the default memory allocator, either for
4418	debugging, analysis or to implement a specific behaviour on memory management
4419	(like on embedded systems). Two function calls are available to do so:</p>
4420	<ul>
4421	<li><a href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlMemGet
4422	()</a> which return the current set of functions in use by the parser</li>
4423	<li><a
4424	href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlMemSetup()</a>
4425	which allow to set up a new set of memory allocation functions</li>
4426	</ul>
4427
4428	<p>Of course a call to xmlMemSetup() should probably be done before calling
4429	any other libxml2 routines (unless you are sure your allocations routines are
4430	compatibles).</p>
4431
4432	<h3><a name="cleanup">Cleaning up after using the library</a></h3>
4433
4434	<p>Libxml2 is not stateless, there is a few set of memory structures needing
4435	allocation before the parser is fully functional (some encoding structures
4436	for example). This also mean that once parsing is finished there is a tiny
4437	amount of memory (a few hundred bytes) which can be recollected if you don't
4438	reuse the library or any document built with it:</p>
4439	<ul>
4440	<li><a href="http://xmlsoft.org/html/libxml-parser.html">xmlCleanupParser
4441	()</a> is a centralized routine to free the library state and data. Note
4442	that it won't deallocate any produced tree if any (use the xmlFreeDoc()
4443	and related routines for this). This should be called only when the library
4444	is not used anymore.</li>
4445	<li><a href="http://xmlsoft.org/html/libxml-parser.html">xmlInitParser
4446	()</a> is the dual routine allowing to preallocate the parsing state
4447	which can be useful for example to avoid initialization reentrancy
4448	problems when using libxml2 in multithreaded applications</li>
4449	</ul>
4450
4451	<p>Generally xmlCleanupParser() is safe assuming no parsing is ongoing and
4452	no document is still being used, if needed the state will be rebuild at the
4453	next invocation of parser routines (or by xmlInitParser()), but be careful
4454	of the consequences in multithreaded applications.</p>
4455
4456	<h3><a name="Debugging">Debugging routines</a></h3>
4457
4458	<p>When configured using --with-mem-debug flag (off by default), libxml2 uses
4459	a set of memory allocation debugging routines keeping track of all allocated
4460	blocks and the location in the code where the routine was called. A couple of
4461	other debugging routines allow to dump the memory allocated infos to a file
4462	or call a specific routine when a given block number is allocated:</p>
4463	<ul>
4464	<li><a
4465	href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlMallocLoc()</a>
4466	<a
4467	href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlReallocLoc()</a>
4468	and <a
4469	href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlMemStrdupLoc()</a>
4470	are the memory debugging replacement allocation routines</li>
4471	<li><a href="http://xmlsoft.org/html/libxml-xmlmemory.html">xmlMemoryDump
4472	()</a> dumps all the information about the allocated memory block lefts
4473	in the <code>.memdump</code> file</li>
4474	</ul>
4475
4476	<p>When developing libxml2 memory debug is enabled, the tests programs call
4477	xmlMemoryDump () and the "make test" regression tests will check for any
4478	memory leak during the full regression test sequence, this helps a lot
4479	ensuring that libxml2 does not leak memory and bullet proof memory
4480	allocations use (some libc implementations are known to be far too permissive
4481	resulting in major portability problems!).</p>
4482
4483	<p>If the .memdump reports a leak, it displays the allocation function and
4484	also tries to give some information about the content and structure of the
4485	allocated blocks left. This is sufficient in most cases to find the culprit,
4486	but not always. Assuming the allocation problem is reproducible, it is
4487	possible to find more easily:</p>
4488	<ol>
4489	<li>write down the block number xxxx not allocated</li>
4490	<li>export the environment variable XML_MEM_BREAKPOINT=xxxx , the easiest
4491	when using GDB is to simply give the command
4492	<p><code>set environment XML_MEM_BREAKPOINT xxxx</code></p>
4493	<p>before running the program.</p>
4494	</li>
4495	<li>run the program under a debugger and set a breakpoint on
4496	xmlMallocBreakpoint() a specific function called when this precise block
4497	is allocated</li>
4498	<li>when the breakpoint is reached you can then do a fine analysis of the
4499	allocation an step to see the condition resulting in the missing
4500	deallocation.</li>
4501	</ol>
4502
4503	<p>I used to use a commercial tool to debug libxml2 memory problems but after
4504	noticing that it was not detecting memory leaks that simple mechanism was
4505	used and proved extremely efficient until now. Lately I have also used <a
4506	href="http://developer.kde.org/~sewardj/">valgrind</a> with quite some
4507	success, it is tied to the i386 architecture since it works by emulating the
4508	processor and instruction set, it is slow but extremely efficient, i.e. it
4509	spot memory usage errors in a very precise way.</p>
4510
4511	<h3><a name="General4">General memory requirements</a></h3>
4512
4513	<p>How much libxml2 memory require ? It's hard to tell in average it depends
4514	of a number of things:</p>
4515	<ul>
4516	<li>the parser itself should work in a fixed amount of memory, except for
4517	information maintained about the stacks of names and entities locations.
4518	The I/O and encoding handlers will probably account for a few KBytes.
4519	This is true for both the XML and HTML parser (though the HTML parser
4520	need more state).</li>
4521	<li>If you are generating the DOM tree then memory requirements will grow
4522	nearly linear with the size of the data. In general for a balanced
4523	textual document the internal memory requirement is about 4 times the
4524	size of the UTF8 serialization of this document (example the XML-1.0
4525	recommendation is a bit more of 150KBytes and takes 650KBytes of main
4526	memory when parsed). Validation will add a amount of memory required for
4527	maintaining the external Dtd state which should be linear with the
4528	complexity of the content model defined by the Dtd</li>
4529	<li>If you need to work with fixed memory requirements or don't need the
4530	full DOM tree then using the <a href="xmlreader.html">xmlReader
4531	interface</a> is probably the best way to proceed, it still allows to
4532	validate or operate on subset of the tree if needed.</li>
4533	<li>If you don't care about the advanced features of libxml2 like
4534	validation, DOM, XPath or XPointer, don't use entities, need to work with
4535	fixed memory requirements, and try to get the fastest parsing possible
4536	then the SAX interface should be used, but it has known restrictions.</li>
4537	</ul>
4538
4539	<p></p>
4540	<h3><a name="Compacting">Returning memory to the kernel</a></h3>
4541
4542	<p>You may encounter that your process using libxml2 does not have a
4543	reduced memory usage although you freed the trees. This is because
4544	libxml2 allocates memory in a number of small chunks. When freeing one
4545	of those chunks, the OS may decide that giving this little memory back
4546	to the kernel will cause too much overhead and delay the operation. As
4547	all chunks are this small, they get actually freed but not returned to
4548	the kernel. On systems using glibc, there is a function call
4549	"malloc_trim" from malloc.h which does this missing operation (note that
4550	it is allowed to fail). Thus, after freeing your tree you may simply try
4551	"malloc_trim(0);" to really get the memory back. If your OS does not
4552	provide malloc_trim, try searching for a similar function.</p>
4553	<p></p>
4554
4555	<h2><a name="Encodings">Encodings support</a></h2>
4556
4557	<p>If you are not really familiar with Internationalization (usual shortcut
4558	is I18N) , Unicode, characters and glyphs, I suggest you read a <a
4559	href="http://www.tbray.org/ongoing/When/200x/2003/04/06/Unicode">presentation</a>
4560	by Tim Bray on Unicode and why you should care about it.</p>
4561
4562	<p>If you don't understand why <b>it does not make sense to have a string
4563	without knowing what encoding it uses</b>, then as Joel Spolsky said <a
4564	href="http://www.joelonsoftware.com/articles/Unicode.html">please do not
4565	write another line of code until you finish reading that article.</a>. It is
4566	a prerequisite to understand this page, and avoid a lot of problems with
4567	libxml2, XML or text processing in general.</p>
4568
4569	<p>Table of Content:</p>
4570	<ol>
4571	<li><a href="encoding.html#What">What does internationalization support
4572	mean ?</a></li>
4573	<li><a href="encoding.html#internal">The internal encoding, how and
4574	why</a></li>
4575	<li><a href="encoding.html#implemente">How is it implemented ?</a></li>
4576	<li><a href="encoding.html#Default">Default supported encodings</a></li>
4577	<li><a href="encoding.html#extend">How to extend the existing
4578	support</a></li>
4579	</ol>
4580
4581	<h3><a name="What">What does internationalization support mean ?</a></h3>
4582
4583	<p>XML was designed from the start to allow the support of any character set
4584	by using Unicode. Any conformant XML parser has to support the UTF-8 and
4585	UTF-16 default encodings which can both express the full unicode ranges. UTF8
4586	is a variable length encoding whose greatest points are to reuse the same
4587	encoding for ASCII and to save space for Western encodings, but it is a bit
4588	more complex to handle in practice. UTF-16 use 2 bytes per character (and
4589	sometimes combines two pairs), it makes implementation easier, but looks a
4590	bit overkill for Western languages encoding. Moreover the XML specification
4591	allows the document to be encoded in other encodings at the condition that
4592	they are clearly labeled as such. For example the following is a wellformed
4593	XML document encoded in ISO-8859-1 and using accentuated letters that we
4594	French like for both markup and content:</p>
4595	<pre><?xml version="1.0" encoding="ISO-8859-1"?>
4596	<très>là </très></pre>
4597
4598	<p>Having internationalization support in libxml2 means the following:</p>
4599	<ul>
4600	<li>the document is properly parsed</li>
4601	<li>information about it's encoding is saved</li>
4602	<li>it can be modified</li>
4603	<li>it can be saved in its original encoding</li>
4604	<li>it can also be saved in another encoding supported by libxml2 (for
4605	example straight UTF8 or even an ASCII form)</li>
4606	</ul>
4607
4608	<p>Another very important point is that the whole libxml2 API, with the
4609	exception of a few routines to read with a specific encoding or save to a
4610	specific encoding, is completely agnostic about the original encoding of the
4611	document.</p>
4612
4613	<p>It should be noted too that the HTML parser embedded in libxml2 now obey
4614	the same rules too, the following document will be (as of 2.2.2) handled in
4615	an internationalized fashion by libxml2 too:</p>
4616	<pre><!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"
4617	"http://www.w3.org/TR/REC-html40/loose.dtd">
4618	<html lang="fr">
4619	<head>
4620	<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=ISO-8859-1">
4621	</head>
4622	<body>
4623	<p>W3C crée des standards pour le Web.</body>
4624	</html></pre>
4625
4626	<h3><a name="internal">The internal encoding, how and why</a></h3>
4627
4628	<p>One of the core decisions was to force all documents to be converted to a
4629	default internal encoding, and that encoding to be UTF-8, here are the
4630	rationales for those choices:</p>
4631	<ul>
4632	<li>keeping the native encoding in the internal form would force the libxml
4633	users (or the code associated) to be fully aware of the encoding of the
4634	original document, for examples when adding a text node to a document,
4635	the content would have to be provided in the document encoding, i.e. the
4636	client code would have to check it before hand, make sure it's conformant
4637	to the encoding, etc ... Very hard in practice, though in some specific
4638	cases this may make sense.</li>
4639	<li>the second decision was which encoding. From the XML spec only UTF8 and
4640	UTF16 really makes sense as being the two only encodings for which there
4641	is mandatory support. UCS-4 (32 bits fixed size encoding) could be
4642	considered an intelligent choice too since it's a direct Unicode mapping
4643	support. I selected UTF-8 on the basis of efficiency and compatibility
4644	with surrounding software:
4645	<ul>
4646	<li>UTF-8 while a bit more complex to convert from/to (i.e. slightly
4647	more costly to import and export CPU wise) is also far more compact
4648	than UTF-16 (and UCS-4) for a majority of the documents I see it used
4649	for right now (RPM RDF catalogs, advogato data, various configuration
4650	file formats, etc.) and the key point for today's computer
4651	architecture is efficient uses of caches. If one nearly double the
4652	memory requirement to store the same amount of data, this will trash
4653	caches (main memory/external caches/internal caches) and my take is
4654	that this harms the system far more than the CPU requirements needed
4655	for the conversion to UTF-8</li>
4656	<li>Most of libxml2 version 1 users were using it with straight ASCII
4657	most of the time, doing the conversion with an internal encoding
4658	requiring all their code to be rewritten was a serious show-stopper
4659	for using UTF-16 or UCS-4.</li>
4660	<li>UTF-8 is being used as the de-facto internal encoding standard for
4661	related code like the <a href="http://www.pango.org/">pango</a>
4662	upcoming Gnome text widget, and a lot of Unix code (yet another place
4663	where Unix programmer base takes a different approach from Microsoft
4664	- they are using UTF-16)</li>
4665	</ul>
4666	</li>
4667	</ul>
4668
4669	<p>What does this mean in practice for the libxml2 user:</p>
4670	<ul>
4671	<li>xmlChar, the libxml2 data type is a byte, those bytes must be assembled
4672	as UTF-8 valid strings. The proper way to terminate an xmlChar * string
4673	is simply to append 0 byte, as usual.</li>
4674	<li>One just need to make sure that when using chars outside the ASCII set,
4675	the values has been properly converted to UTF-8</li>
4676	</ul>
4677
4678	<h3><a name="implemente">How is it implemented ?</a></h3>
4679
4680	<p>Let's describe how all this works within libxml, basically the I18N
4681	(internationalization) support get triggered only during I/O operation, i.e.
4682	when reading a document or saving one. Let's look first at the reading
4683	sequence:</p>
4684	<ol>
4685	<li>when a document is processed, we usually don't know the encoding, a
4686	simple heuristic allows to detect UTF-16 and UCS-4 from encodings where
4687	the ASCII range (0-0x7F) maps with ASCII</li>
4688	<li>the xml declaration if available is parsed, including the encoding
4689	declaration. At that point, if the autodetected encoding is different
4690	from the one declared a call to xmlSwitchEncoding() is issued.</li>
4691	<li>If there is no encoding declaration, then the input has to be in either
4692	UTF-8 or UTF-16, if it is not then at some point when processing the
4693	input, the converter/checker of UTF-8 form will raise an encoding error.
4694	You may end-up with a garbled document, or no document at all ! Example:
4695	<pre>~/XML -> ./xmllint err.xml
4696	err.xml:1: error: Input is not proper UTF-8, indicate encoding !
4697	<très>là </très>
4698	^
4699	err.xml:1: error: Bytes: 0xE8 0x73 0x3E 0x6C
4700	<très>là </très>
4701	^</pre>
4702	</li>
4703	<li>xmlSwitchEncoding() does an encoding name lookup, canonicalize it, and
4704	then search the default registered encoding converters for that encoding.
4705	If it's not within the default set and iconv() support has been compiled
4706	it, it will ask iconv for such an encoder. If this fails then the parser
4707	will report an error and stops processing:
4708	<pre>~/XML -> ./xmllint err2.xml
4709	err2.xml:1: error: Unsupported encoding UnsupportedEnc
4710	<?xml version="1.0" encoding="UnsupportedEnc"?>
4711	^</pre>
4712	</li>
4713	<li>From that point the encoder processes progressively the input (it is
4714	plugged as a front-end to the I/O module) for that entity. It captures
4715	and converts on-the-fly the document to be parsed to UTF-8. The parser
4716	itself just does UTF-8 checking of this input and process it
4717	transparently. The only difference is that the encoding information has
4718	been added to the parsing context (more precisely to the input
4719	corresponding to this entity).</li>
4720	<li>The result (when using DOM) is an internal form completely in UTF-8
4721	with just an encoding information on the document node.</li>
4722	</ol>
4723
4724	<p>Ok then what happens when saving the document (assuming you
4725	collected/built an xmlDoc DOM like structure) ? It depends on the function
4726	called, xmlSaveFile() will just try to save in the original encoding, while
4727	xmlSaveFileTo() and xmlSaveFileEnc() can optionally save to a given
4728	encoding:</p>
4729	<ol>
4730	<li>if no encoding is given, libxml2 will look for an encoding value
4731	associated to the document and if it exists will try to save to that
4732	encoding,
4733	<p>otherwise everything is written in the internal form, i.e. UTF-8</p>
4734	</li>
4735	<li>so if an encoding was specified, either at the API level or on the
4736	document, libxml2 will again canonicalize the encoding name, lookup for a
4737	converter in the registered set or through iconv. If not found the
4738	function will return an error code</li>
4739	<li>the converter is placed before the I/O buffer layer, as another kind of
4740	buffer, then libxml2 will simply push the UTF-8 serialization to through
4741	that buffer, which will then progressively be converted and pushed onto
4742	the I/O layer.</li>
4743	<li>It is possible that the converter code fails on some input, for example
4744	trying to push an UTF-8 encoded Chinese character through the UTF-8 to
4745	ISO-8859-1 converter won't work. Since the encoders are progressive they
4746	will just report the error and the number of bytes converted, at that
4747	point libxml2 will decode the offending character, remove it from the
4748	buffer and replace it with the associated charRef encoding &#123; and
4749	resume the conversion. This guarantees that any document will be saved
4750	without losses (except for markup names where this is not legal, this is
4751	a problem in the current version, in practice avoid using non-ascii
4752	characters for tag or attribute names). A special "ascii" encoding name
4753	is used to save documents to a pure ascii form can be used when
4754	portability is really crucial</li>
4755	</ol>
4756
4757	<p>Here are a few examples based on the same test document and assumin a
4758	terminal using ISO-8859-1 as the text encoding:</p>
4759	<pre>~/XML -> ./xmllint isolat1
4760	<?xml version="1.0" encoding="ISO-8859-1"?>
4761	<très>lÃ </très>
4762	~/XML -> ./xmllint --encode UTF-8 isolat1
4763	<?xml version="1.0" encoding="UTF-8"?>
4764	<trÃ¨s>lÃ  </trÃ¨s>
4765	~/XML -> </pre>
4766
4767	<p>The same processing is applied (and reuse most of the code) for HTML I18N
4768	processing. Looking up and modifying the content encoding is a bit more
4769	difficult since it is located in a <meta> tag under the <head>,
4770	so a couple of functions htmlGetMetaEncoding() and htmlSetMetaEncoding() have
4771	been provided. The parser also attempts to switch encoding on the fly when
4772	detecting such a tag on input. Except for that the processing is the same
4773	(and again reuses the same code).</p>
4774
4775	<h3><a name="Default">Default supported encodings</a></h3>
4776
4777	<p>libxml2 has a set of default converters for the following encodings
4778	(located in encoding.c):</p>
4779	<ol>
4780	<li>UTF-8 is supported by default (null handlers)</li>
4781	<li>UTF-16, both little and big endian</li>
4782	<li>ISO-Latin-1 (ISO-8859-1) covering most western languages</li>
4783	<li>ASCII, useful mostly for saving</li>
4784	<li>HTML, a specific handler for the conversion of UTF-8 to ASCII with HTML
4785	predefined entities like &copy; for the Copyright sign.</li>
4786	</ol>
4787
4788	<p>More over when compiled on an Unix platform with iconv support the full
4789	set of encodings supported by iconv can be instantly be used by libxml. On a
4790	linux machine with glibc-2.1 the list of supported encodings and aliases fill
4791	3 full pages, and include UCS-4, the full set of ISO-Latin encodings, and the
4792	various Japanese ones.</p>
4793
4794	<p>To convert from the UTF-8 values returned from the API to another encoding
4795	then it is possible to use the function provided from <a
4796	href="html/libxml-encoding.html">the encoding module</a> like <a
4797	href="html/libxml-encoding.html#UTF8Toisolat1">UTF8Toisolat1</a>, or use the
4798	POSIX <a
4799	href="http://www.opengroup.org/onlinepubs/009695399/functions/iconv.html">iconv()</a>
4800	API directly.</p>
4801
4802	<h4>Encoding aliases</h4>
4803
4804	<p>From 2.2.3, libxml2 has support to register encoding names aliases. The
4805	goal is to be able to parse document whose encoding is supported but where
4806	the name differs (for example from the default set of names accepted by
4807	iconv). The following functions allow to register and handle new aliases for
4808	existing encodings. Once registered libxml2 will automatically lookup the
4809	aliases when handling a document:</p>
4810	<ul>
4811	<li>int xmlAddEncodingAlias(const char name, const char alias);</li>
4812	<li>int xmlDelEncodingAlias(const char *alias);</li>
4813	<li>const char * xmlGetEncodingAlias(const char *alias);</li>
4814	<li>void xmlCleanupEncodingAliases(void);</li>
4815	</ul>
4816
4817	<h3><a name="extend">How to extend the existing support</a></h3>
4818
4819	<p>Well adding support for new encoding, or overriding one of the encoders
4820	(assuming it is buggy) should not be hard, just write input and output
4821	conversion routines to/from UTF-8, and register them using
4822	xmlNewCharEncodingHandler(name, xxxToUTF8, UTF8Toxxx), and they will be
4823	called automatically if the parser(s) encounter such an encoding name
4824	(register it uppercase, this will help). The description of the encoders,
4825	their arguments and expected return values are described in the encoding.h
4826	header.</p>
4827
4828	<h2><a name="IO">I/O Interfaces</a></h2>
4829
4830	<p>Table of Content:</p>
4831	<ol>
4832	<li><a href="#General1">General overview</a></li>
4833	<li><a href="#basic">The basic buffer type</a></li>
4834	<li><a href="#Input">Input I/O handlers</a></li>
4835	<li><a href="#Output">Output I/O handlers</a></li>
4836	<li><a href="#entities">The entities loader</a></li>
4837	<li><a href="#Example2">Example of customized I/O</a></li>
4838	</ol>
4839
4840	<h3><a name="General1">General overview</a></h3>
4841
4842	<p>The module <code><a
4843	href="http://xmlsoft.org/html/libxml-xmlio.html">xmlIO.h</a></code> provides
4844	the interfaces to the libxml2 I/O system. This consists of 4 main parts:</p>
4845	<ul>
4846	<li>Entities loader, this is a routine which tries to fetch the entities
4847	(files) based on their PUBLIC and SYSTEM identifiers. The default loader
4848	don't look at the public identifier since libxml2 do not maintain a
4849	catalog. You can redefine you own entity loader by using
4850	<code>xmlGetExternalEntityLoader()</code> and
4851	<code>xmlSetExternalEntityLoader()</code>. <a href="#entities">Check the
4852	example</a>.</li>
4853	<li>Input I/O buffers which are a commodity structure used by the parser(s)
4854	input layer to handle fetching the information to feed the parser. This
4855	provides buffering and is also a placeholder where the encoding
4856	converters to UTF8 are piggy-backed.</li>
4857	<li>Output I/O buffers are similar to the Input ones and fulfill similar
4858	task but when generating a serialization from a tree.</li>
4859	<li>A mechanism to register sets of I/O callbacks and associate them with
4860	specific naming schemes like the protocol part of the URIs.
4861	<p>This affect the default I/O operations and allows to use specific I/O
4862	handlers for certain names.</p>
4863	</li>
4864	</ul>
4865
4866	<p>The general mechanism used when loading http://rpmfind.net/xml.html for
4867	example in the HTML parser is the following:</p>
4868	<ol>
4869	<li>The default entity loader calls <code>xmlNewInputFromFile()</code> with
4870	the parsing context and the URI string.</li>
4871	<li>the URI string is checked against the existing registered handlers
4872	using their match() callback function, if the HTTP module was compiled
4873	in, it is registered and its match() function will succeeds</li>
4874	<li>the open() function of the handler is called and if successful will
4875	return an I/O Input buffer</li>
4876	<li>the parser will the start reading from this buffer and progressively
4877	fetch information from the resource, calling the read() function of the
4878	handler until the resource is exhausted</li>
4879	<li>if an encoding change is detected it will be installed on the input
4880	buffer, providing buffering and efficient use of the conversion
4881	routines</li>
4882	<li>once the parser has finished, the close() function of the handler is
4883	called once and the Input buffer and associated resources are
4884	deallocated.</li>
4885	</ol>
4886
4887	<p>The user defined callbacks are checked first to allow overriding of the
4888	default libxml2 I/O routines.</p>
4889
4890	<h3><a name="basic">The basic buffer type</a></h3>
4891
4892	<p>All the buffer manipulation handling is done using the
4893	<code>xmlBuffer</code> type define in <code><a
4894	href="http://xmlsoft.org/html/libxml-tree.html">tree.h</a> </code>which is a
4895	resizable memory buffer. The buffer allocation strategy can be selected to be
4896	either best-fit or use an exponential doubling one (CPU vs. memory use
4897	trade-off). The values are <code>XML_BUFFER_ALLOC_EXACT</code> and
4898	<code>XML_BUFFER_ALLOC_DOUBLEIT</code>, and can be set individually or on a
4899	system wide basis using <code>xmlBufferSetAllocationScheme()</code>. A number
4900	of functions allows to manipulate buffers with names starting with the
4901	<code>xmlBuffer...</code> prefix.</p>
4902
4903	<h3><a name="Input">Input I/O handlers</a></h3>
4904
4905	<p>An Input I/O handler is a simple structure
4906	<code>xmlParserInputBuffer</code> containing a context associated to the
4907	resource (file descriptor, or pointer to a protocol handler), the read() and
4908	close() callbacks to use and an xmlBuffer. And extra xmlBuffer and a charset
4909	encoding handler are also present to support charset conversion when
4910	needed.</p>
4911
4912	<h3><a name="Output">Output I/O handlers</a></h3>
4913
4914	<p>An Output handler <code>xmlOutputBuffer</code> is completely similar to an
4915	Input one except the callbacks are write() and close().</p>
4916
4917	<h3><a name="entities">The entities loader</a></h3>
4918
4919	<p>The entity loader resolves requests for new entities and create inputs for
4920	the parser. Creating an input from a filename or an URI string is done
4921	through the xmlNewInputFromFile() routine. The default entity loader do not
4922	handle the PUBLIC identifier associated with an entity (if any). So it just
4923	calls xmlNewInputFromFile() with the SYSTEM identifier (which is mandatory in
4924	XML).</p>
4925
4926	<p>If you want to hook up a catalog mechanism then you simply need to
4927	override the default entity loader, here is an example:</p>
4928	<pre>#include <libxml/xmlIO.h>
4929
4930	xmlExternalEntityLoader defaultLoader = NULL;
4931
4932	xmlParserInputPtr
4933	xmlMyExternalEntityLoader(const char URL, const char ID,
4934	xmlParserCtxtPtr ctxt) {
4935	xmlParserInputPtr ret;
4936	const char *fileID = NULL;
4937	/* lookup for the fileID depending on ID */
4938
4939	ret = xmlNewInputFromFile(ctxt, fileID);
4940	if (ret != NULL)
4941	return(ret);
4942	if (defaultLoader != NULL)
4943	ret = defaultLoader(URL, ID, ctxt);
4944	return(ret);
4945	}
4946
4947	int main(..) {
4948	...
4949
4950	/*
4951	* Install our own entity loader
4952	*/
4953	defaultLoader = xmlGetExternalEntityLoader();
4954	xmlSetExternalEntityLoader(xmlMyExternalEntityLoader);
4955
4956	...
4957	}</pre>
4958
4959	<h3><a name="Example2">Example of customized I/O</a></h3>
4960
4961	<p>This example come from <a href="http://xmlsoft.org/messages/0708.html">a
4962	real use case</a>, xmlDocDump() closes the FILE * passed by the application
4963	and this was a problem. The <a
4964	href="http://xmlsoft.org/messages/0711.html">solution</a> was to redefine a
4965	new output handler with the closing call deactivated:</p>
4966	<ol>
4967	<li>First define a new I/O output allocator where the output don't close
4968	the file:
4969	<pre>xmlOutputBufferPtr
4970	xmlOutputBufferCreateOwn(FILE *file, xmlCharEncodingHandlerPtr encoder) {
4971	xmlOutputBufferPtr ret;
4972
4973	if (xmlOutputCallbackInitialized == 0)
4974	xmlRegisterDefaultOutputCallbacks();
4975
4976	if (file == NULL) return(NULL);
4977	ret = xmlAllocOutputBuffer(encoder);
4978	if (ret != NULL) {
4979	ret->context = file;
4980	ret->writecallback = xmlFileWrite;
4981	ret->closecallback = NULL; /* No close callback */
4982	}
4983	return(ret);
4984	} </pre>
4985	</li>
4986	<li>And then use it to save the document:
4987	<pre>FILE *f;
4988	xmlOutputBufferPtr output;
4989	xmlDocPtr doc;
4990	int res;
4991
4992	f = ...
4993	doc = ....
4994
4995	output = xmlOutputBufferCreateOwn(f, NULL);
4996	res = xmlSaveFileTo(output, doc, NULL);
4997	</pre>
4998	</li>
4999	</ol>
5000
5001	<h2><a name="Catalog">Catalog support</a></h2>
5002
5003	<p>Table of Content:</p>
5004	<ol>
5005	<li><a href="General2">General overview</a></li>
5006	<li><a href="#definition">The definition</a></li>
5007	<li><a href="#Simple">Using catalogs</a></li>
5008	<li><a href="#Some">Some examples</a></li>
5009	<li><a href="#reference">How to tune catalog usage</a></li>
5010	<li><a href="#validate">How to debug catalog processing</a></li>
5011	<li><a href="#Declaring">How to create and maintain catalogs</a></li>
5012	<li><a href="#implemento">The implementor corner quick review of the
5013	API</a></li>
5014	<li><a href="#Other">Other resources</a></li>
5015	</ol>
5016
5017	<h3><a name="General2">General overview</a></h3>
5018
5019	<p>What is a catalog? Basically it's a lookup mechanism used when an entity
5020	(a file or a remote resource) references another entity. The catalog lookup
5021	is inserted between the moment the reference is recognized by the software
5022	(XML parser, stylesheet processing, or even images referenced for inclusion
5023	in a rendering) and the time where loading that resource is actually
5024	started.</p>
5025
5026	<p>It is basically used for 3 things:</p>
5027	<ul>
5028	<li>mapping from "logical" names, the public identifiers and a more
5029	concrete name usable for download (and URI). For example it can associate
5030	the logical name
5031	<p>"-//OASIS//DTD DocBook XML V4.1.2//EN"</p>
5032	<p>of the DocBook 4.1.2 XML DTD with the actual URL where it can be
5033	downloaded</p>
5034	<p>http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd</p>
5035	</li>
5036	<li>remapping from a given URL to another one, like an HTTP indirection
5037	saying that
5038	<p>"http://www.oasis-open.org/committes/tr.xsl"</p>
5039	<p>should really be looked at</p>
5040	<p>"http://www.oasis-open.org/committes/entity/stylesheets/base/tr.xsl"</p>
5041	</li>
5042	<li>providing a local cache mechanism allowing to load the entities
5043	associated to public identifiers or remote resources, this is a really
5044	important feature for any significant deployment of XML or SGML since it
5045	allows to avoid the aleas and delays associated to fetching remote
5046	resources.</li>
5047	</ul>
5048
5049	<h3><a name="definition">The definitions</a></h3>
5050
5051	<p>Libxml, as of 2.4.3 implements 2 kind of catalogs:</p>
5052	<ul>
5053	<li>the older SGML catalogs, the official spec is SGML Open Technical
5054	Resolution TR9401:1997, but is better understood by reading <a
5055	href="http://www.jclark.com/sp/catalog.htm">the SP Catalog page</a> from
5056	James Clark. This is relatively old and not the preferred mode of
5057	operation of libxml.</li>
5058	<li><a href="http://www.oasis-open.org/committees/entity/spec.html">XML
5059	Catalogs</a> is far more flexible, more recent, uses an XML syntax and
5060	should scale quite better. This is the default option of libxml.</li>
5061	</ul>
5062
5063	<p></p>
5064
5065	<h3><a name="Simple">Using catalog</a></h3>
5066
5067	<p>In a normal environment libxml2 will by default check the presence of a
5068	catalog in /etc/xml/catalog, and assuming it has been correctly populated,
5069	the processing is completely transparent to the document user. To take a
5070	concrete example, suppose you are authoring a DocBook document, this one
5071	starts with the following DOCTYPE definition:</p>
5072	<pre><?xml version='1.0'?>
5073	<!DOCTYPE book PUBLIC "-//Norman Walsh//DTD DocBk XML V3.1.4//EN"
5074	"http://nwalsh.com/docbook/xml/3.1.4/db3xml.dtd"></pre>
5075
5076	<p>When validating the document with libxml, the catalog will be
5077	automatically consulted to lookup the public identifier "-//Norman Walsh//DTD
5078	DocBk XML V3.1.4//EN" and the system identifier
5079	"http://nwalsh.com/docbook/xml/3.1.4/db3xml.dtd", and if these entities have
5080	been installed on your system and the catalogs actually point to them, libxml
5081	will fetch them from the local disk.</p>
5082
5083	<p style="font-size: 10pt"><strong>Note</strong>: Really don't use this
5084	DOCTYPE example it's a really old version, but is fine as an example.</p>
5085
5086	<p>Libxml2 will check the catalog each time that it is requested to load an
5087	entity, this includes DTD, external parsed entities, stylesheets, etc ... If
5088	your system is correctly configured all the authoring phase and processing
5089	should use only local files, even if your document stays portable because it
5090	uses the canonical public and system ID, referencing the remote document.</p>
5091
5092	<h3><a name="Some">Some examples:</a></h3>
5093
5094	<p>Here is a couple of fragments from XML Catalogs used in libxml2 early
5095	regression tests in <code>test/catalogs</code> :</p>
5096	<pre><?xml version="1.0"?>
5097	<!DOCTYPE catalog PUBLIC
5098	"-//OASIS//DTD Entity Resolution XML Catalog V1.0//EN"
5099	"http://www.oasis-open.org/committees/entity/release/1.0/catalog.dtd">
5100	<catalog xmlns="urn:oasis:names:tc:entity:xmlns:xml:catalog">
5101	<public publicId="-//OASIS//DTD DocBook XML V4.1.2//EN"
5102	uri="http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd"/>
5103	...</pre>
5104
5105	<p>This is the beginning of a catalog for DocBook 4.1.2, XML Catalogs are
5106	written in XML, there is a specific namespace for catalog elements
5107	"urn:oasis:names:tc:entity:xmlns:xml:catalog". The first entry in this
5108	catalog is a <code>public</code> mapping it allows to associate a Public
5109	Identifier with an URI.</p>
5110	<pre>...
5111	<rewriteSystem systemIdStartString="http://www.oasis-open.org/docbook/"
5112	rewritePrefix="file:///usr/share/xml/docbook/"/>
5113	...</pre>
5114
5115	<p>A <code>rewriteSystem</code> is a very powerful instruction, it says that
5116	any URI starting with a given prefix should be looked at another URI
5117	constructed by replacing the prefix with an new one. In effect this acts like
5118	a cache system for a full area of the Web. In practice it is extremely useful
5119	with a file prefix if you have installed a copy of those resources on your
5120	local system.</p>
5121	<pre>...
5122	<delegatePublic publicIdStartString="-//OASIS//DTD XML Catalog //"
5123	catalog="file:///usr/share/xml/docbook.xml"/>
5124	<delegatePublic publicIdStartString="-//OASIS//ENTITIES DocBook XML"
5125	catalog="file:///usr/share/xml/docbook.xml"/>
5126	<delegatePublic publicIdStartString="-//OASIS//DTD DocBook XML"
5127	catalog="file:///usr/share/xml/docbook.xml"/>
5128	<delegateSystem systemIdStartString="http://www.oasis-open.org/docbook/"
5129	catalog="file:///usr/share/xml/docbook.xml"/>
5130	<delegateURI uriStartString="http://www.oasis-open.org/docbook/"
5131	catalog="file:///usr/share/xml/docbook.xml"/>
5132	...</pre>
5133
5134	<p>Delegation is the core features which allows to build a tree of catalogs,
5135	easier to maintain than a single catalog, based on Public Identifier, System
5136	Identifier or URI prefixes it instructs the catalog software to look up
5137	entries in another resource. This feature allow to build hierarchies of
5138	catalogs, the set of entries presented should be sufficient to redirect the
5139	resolution of all DocBook references to the specific catalog in
5140	<code>/usr/share/xml/docbook.xml</code> this one in turn could delegate all
5141	references for DocBook 4.2.1 to a specific catalog installed at the same time
5142	as the DocBook resources on the local machine.</p>
5143
5144	<h3><a name="reference">How to tune catalog usage:</a></h3>
5145
5146	<p>The user can change the default catalog behaviour by redirecting queries
5147	to its own set of catalogs, this can be done by setting the
5148	<code>XML_CATALOG_FILES</code> environment variable to a list of catalogs, an
5149	empty one should deactivate loading the default <code>/etc/xml/catalog</code>
5150	default catalog</p>
5151
5152	<h3><a name="validate">How to debug catalog processing:</a></h3>
5153
5154	<p>Setting up the <code>XML_DEBUG_CATALOG</code> environment variable will
5155	make libxml2 output debugging information for each catalog operations, for
5156	example:</p>
5157	<pre>orchis:~/XML -> xmllint --memory --noout test/ent2
5158	warning: failed to load external entity "title.xml"
5159	orchis:~/XML -> export XML_DEBUG_CATALOG=
5160	orchis:~/XML -> xmllint --memory --noout test/ent2
5161	Failed to parse catalog /etc/xml/catalog
5162	Failed to parse catalog /etc/xml/catalog
5163	warning: failed to load external entity "title.xml"
5164	Catalogs cleanup
5165	orchis:~/XML -> </pre>
5166
5167	<p>The test/ent2 references an entity, running the parser from memory makes
5168	the base URI unavailable and the the "title.xml" entity cannot be loaded.
5169	Setting up the debug environment variable allows to detect that an attempt is
5170	made to load the <code>/etc/xml/catalog</code> but since it's not present the
5171	resolution fails.</p>
5172
5173	<p>But the most advanced way to debug XML catalog processing is to use the
5174	<strong>xmlcatalog</strong> command shipped with libxml2, it allows to load
5175	catalogs and make resolution queries to see what is going on. This is also
5176	used for the regression tests:</p>
5177	<pre>orchis:~/XML -> ./xmlcatalog test/catalogs/docbook.xml \
5178	"-//OASIS//DTD DocBook XML V4.1.2//EN"
5179	http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd
5180	orchis:~/XML -> </pre>
5181
5182	<p>For debugging what is going on, adding one -v flags increase the verbosity
5183	level to indicate the processing done (adding a second flag also indicate
5184	what elements are recognized at parsing):</p>
5185	<pre>orchis:~/XML -> ./xmlcatalog -v test/catalogs/docbook.xml \
5186	"-//OASIS//DTD DocBook XML V4.1.2//EN"
5187	Parsing catalog test/catalogs/docbook.xml's content
5188	Found public match -//OASIS//DTD DocBook XML V4.1.2//EN
5189	http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd
5190	Catalogs cleanup
5191	orchis:~/XML -> </pre>
5192
5193	<p>A shell interface is also available to debug and process multiple queries
5194	(and for regression tests):</p>
5195	<pre>orchis:~/XML -> ./xmlcatalog -shell test/catalogs/docbook.xml \
5196	"-//OASIS//DTD DocBook XML V4.1.2//EN"
5197	> help
5198	Commands available:
5199	public PublicID: make a PUBLIC identifier lookup
5200	system SystemID: make a SYSTEM identifier lookup
5201	resolve PublicID SystemID: do a full resolver lookup
5202	add 'type' 'orig' 'replace' : add an entry
5203	del 'values' : remove values
5204	dump: print the current catalog state
5205	debug: increase the verbosity level
5206	quiet: decrease the verbosity level
5207	exit: quit the shell
5208	> public "-//OASIS//DTD DocBook XML V4.1.2//EN"
5209	http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd
5210	> quit
5211	orchis:~/XML -> </pre>
5212
5213	<p>This should be sufficient for most debugging purpose, this was actually
5214	used heavily to debug the XML Catalog implementation itself.</p>
5215
5216	<h3><a name="Declaring">How to create and maintain</a> catalogs:</h3>
5217
5218	<p>Basically XML Catalogs are XML files, you can either use XML tools to
5219	manage them or use <strong>xmlcatalog</strong> for this. The basic step is
5220	to create a catalog the -create option provide this facility:</p>
5221	<pre>orchis:~/XML -> ./xmlcatalog --create tst.xml
5222	<?xml version="1.0"?>
5223	<!DOCTYPE catalog PUBLIC "-//OASIS//DTD Entity Resolution XML Catalog V1.0//EN"
5224	"http://www.oasis-open.org/committees/entity/release/1.0/catalog.dtd">
5225	<catalog xmlns="urn:oasis:names:tc:entity:xmlns:xml:catalog"/>
5226	orchis:~/XML -> </pre>
5227
5228	<p>By default xmlcatalog does not overwrite the original catalog and save the
5229	result on the standard output, this can be overridden using the -noout
5230	option. The <code>-add</code> command allows to add entries in the
5231	catalog:</p>
5232	<pre>orchis:~/XML -> ./xmlcatalog --noout --create --add "public" \
5233	"-//OASIS//DTD DocBook XML V4.1.2//EN" \
5234	http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd tst.xml
5235	orchis:~/XML -> cat tst.xml
5236	<?xml version="1.0"?>
5237	<!DOCTYPE catalog PUBLIC "-//OASIS//DTD Entity Resolution XML Catalog V1.0//EN" \
5238	"http://www.oasis-open.org/committees/entity/release/1.0/catalog.dtd">
5239	<catalog xmlns="urn:oasis:names:tc:entity:xmlns:xml:catalog">
5240	<public publicId="-//OASIS//DTD DocBook XML V4.1.2//EN"
5241	uri="http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd"/>
5242	</catalog>
5243	orchis:~/XML -> </pre>
5244
5245	<p>The <code>-add</code> option will always take 3 parameters even if some of
5246	the XML Catalog constructs (like nextCatalog) will have only a single
5247	argument, just pass a third empty string, it will be ignored.</p>
5248
5249	<p>Similarly the <code>-del</code> option remove matching entries from the
5250	catalog:</p>
5251	<pre>orchis:~/XML -> ./xmlcatalog --del \
5252	"http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" tst.xml
5253	<?xml version="1.0"?>
5254	<!DOCTYPE catalog PUBLIC "-//OASIS//DTD Entity Resolution XML Catalog V1.0//EN"
5255	"http://www.oasis-open.org/committees/entity/release/1.0/catalog.dtd">
5256	<catalog xmlns="urn:oasis:names:tc:entity:xmlns:xml:catalog"/>
5257	orchis:~/XML -> </pre>
5258
5259	<p>The catalog is now empty. Note that the matching of <code>-del</code> is
5260	exact and would have worked in a similar fashion with the Public ID
5261	string.</p>
5262
5263	<p>This is rudimentary but should be sufficient to manage a not too complex
5264	catalog tree of resources.</p>
5265
5266	<h3><a name="implemento">The implementor corner quick review of the
5267	API:</a></h3>
5268
5269	<p>First, and like for every other module of libxml, there is an
5270	automatically generated <a href="html/libxml-catalog.html">API page for
5271	catalog support</a>.</p>
5272
5273	<p>The header for the catalog interfaces should be included as:</p>
5274	<pre>#include <libxml/catalog.h></pre>
5275
5276	<p>The API is voluntarily kept very simple. First it is not obvious that
5277	applications really need access to it since it is the default behaviour of
5278	libxml2 (Note: it is possible to completely override libxml2 default catalog
5279	by using <a href="html/libxml-parser.html">xmlSetExternalEntityLoader</a> to
5280	plug an application specific resolver).</p>
5281
5282	<p>Basically libxml2 support 2 catalog lists:</p>
5283	<ul>
5284	<li>the default one, global shared by all the application</li>
5285	<li>a per-document catalog, this one is built if the document uses the
5286	<code>oasis-xml-catalog</code> PIs to specify its own catalog list, it is
5287	associated to the parser context and destroyed when the parsing context
5288	is destroyed.</li>
5289	</ul>
5290
5291	<p>the document one will be used first if it exists.</p>
5292
5293	<h4>Initialization routines:</h4>
5294
5295	<p>xmlInitializeCatalog(), xmlLoadCatalog() and xmlLoadCatalogs() should be
5296	used at startup to initialize the catalog, if the catalog should be
5297	initialized with specific values xmlLoadCatalog() or xmlLoadCatalogs()
5298	should be called before xmlInitializeCatalog() which would otherwise do a
5299	default initialization first.</p>
5300
5301	<p>The xmlCatalogAddLocal() call is used by the parser to grow the document
5302	own catalog list if needed.</p>
5303
5304	<h4>Preferences setup:</h4>
5305
5306	<p>The XML Catalog spec requires the possibility to select default
5307	preferences between public and system delegation,
5308	xmlCatalogSetDefaultPrefer() allows this, xmlCatalogSetDefaults() and
5309	xmlCatalogGetDefaults() allow to control if XML Catalogs resolution should
5310	be forbidden, allowed for global catalog, for document catalog or both, the
5311	default is to allow both.</p>
5312
5313	<p>And of course xmlCatalogSetDebug() allows to generate debug messages
5314	(through the xmlGenericError() mechanism).</p>
5315
5316	<h4>Querying routines:</h4>
5317
5318	<p>xmlCatalogResolve(), xmlCatalogResolveSystem(), xmlCatalogResolvePublic()
5319	and xmlCatalogResolveURI() are relatively explicit if you read the XML
5320	Catalog specification they correspond to section 7 algorithms, they should
5321	also work if you have loaded an SGML catalog with a simplified semantic.</p>
5322
5323	<p>xmlCatalogLocalResolve() and xmlCatalogLocalResolveURI() are the same but
5324	operate on the document catalog list</p>
5325
5326	<h4>Cleanup and Miscellaneous:</h4>
5327
5328	<p>xmlCatalogCleanup() free-up the global catalog, xmlCatalogFreeLocal() is
5329	the per-document equivalent.</p>
5330
5331	<p>xmlCatalogAdd() and xmlCatalogRemove() are used to dynamically modify the
5332	first catalog in the global list, and xmlCatalogDump() allows to dump a
5333	catalog state, those routines are primarily designed for xmlcatalog, I'm not
5334	sure that exposing more complex interfaces (like navigation ones) would be
5335	really useful.</p>
5336
5337	<p>The xmlParseCatalogFile() is a function used to load XML Catalog files,
5338	it's similar as xmlParseFile() except it bypass all catalog lookups, it's
5339	provided because this functionality may be useful for client tools.</p>
5340
5341	<h4>threaded environments:</h4>
5342
5343	<p>Since the catalog tree is built progressively, some care has been taken to
5344	try to avoid troubles in multithreaded environments. The code is now thread
5345	safe assuming that the libxml2 library has been compiled with threads
5346	support.</p>
5347
5348	<p></p>
5349
5350	<h3><a name="Other">Other resources</a></h3>
5351
5352	<p>The XML Catalog specification is relatively recent so there isn't much
5353	literature to point at:</p>
5354	<ul>
5355	<li>You can find a good rant from Norm Walsh about <a
5356	href="http://www.arbortext.com/Think_Tank/XML_Resources/Issue_Three/issue_three.html">the
5357	need for catalogs</a>, it provides a lot of context information even if
5358	I don't agree with everything presented. Norm also wrote a more recent
5359	article <a
5360	href="http://wwws.sun.com/software/xml/developers/resolver/article/">XML
5361	entities and URI resolvers</a> describing them.</li>
5362	<li>An <a href="http://home.ccil.org/~cowan/XML/XCatalog.html">old XML
5363	catalog proposal</a> from John Cowan</li>
5364	<li>The <a href="http://www.rddl.org/">Resource Directory Description
5365	Language</a> (RDDL) another catalog system but more oriented toward
5366	providing metadata for XML namespaces.</li>
5367	<li>the page from the OASIS Technical <a
5368	href="http://www.oasis-open.org/committees/entity/">Committee on Entity
5369	Resolution</a> who maintains XML Catalog, you will find pointers to the
5370	specification update, some background and pointers to others tools
5371	providing XML Catalog support</li>
5372	<li>There is a <a href="buildDocBookCatalog">shell script</a> to generate
5373	XML Catalogs for DocBook 4.1.2 . If it can write to the /etc/xml/
5374	directory, it will set-up /etc/xml/catalog and /etc/xml/docbook based on
5375	the resources found on the system. Otherwise it will just create
5376	~/xmlcatalog and ~/dbkxmlcatalog and doing:
5377	<p><code>export XML_CATALOG_FILES=$HOME/xmlcatalog</code></p>
5378	<p>should allow to process DocBook documentations without requiring
5379	network accesses for the DTD or stylesheets</p>
5380	</li>
5381	<li>I have uploaded <a href="ftp://xmlsoft.org/libxml2/test/dbk412catalog.tar.gz">a
5382	small tarball</a> containing XML Catalogs for DocBook 4.1.2 which seems
5383	to work fine for me too</li>
5384	<li>The <a href="http://www.xmlsoft.org/xmlcatalog_man.html">xmlcatalog
5385	manual page</a></li>
5386	</ul>
5387
5388	<p>If you have suggestions for corrections or additions, simply contact
5389	me:</p>
5390
5391	<h2><a name="library">The parser interfaces</a></h2>
5392
5393	<p>This section is directly intended to help programmers getting bootstrapped
5394	using the XML tollkit from the C language. It is not intended to be
5395	extensive. I hope the automatically generated documents will provide the
5396	completeness required, but as a separate set of documents. The interfaces of
5397	the XML parser are by principle low level, Those interested in a higher level
5398	API should <a href="#DOM">look at DOM</a>.</p>
5399
5400	<p>The <a href="html/libxml-parser.html">parser interfaces for XML</a> are
5401	separated from the <a href="html/libxml-htmlparser.html">HTML parser
5402	interfaces</a>. Let's have a look at how the XML parser can be called:</p>
5403
5404	<h3><a name="Invoking">Invoking the parser : the pull method</a></h3>
5405
5406	<p>Usually, the first thing to do is to read an XML input. The parser accepts
5407	documents either from in-memory strings or from files. The functions are
5408	defined in "parser.h":</p>
5409	<dl>
5410	<dt><code>xmlDocPtr xmlParseMemory(char *buffer, int size);</code></dt>
5411	<dd><p>Parse a null-terminated string containing the document.</p>
5412	</dd>
5413	</dl>
5414	<dl>
5415	<dt><code>xmlDocPtr xmlParseFile(const char *filename);</code></dt>
5416	<dd><p>Parse an XML document contained in a (possibly compressed)
5417	file.</p>
5418	</dd>
5419	</dl>
5420
5421	<p>The parser returns a pointer to the document structure (or NULL in case of
5422	failure).</p>
5423
5424	<h3 id="Invoking1">Invoking the parser: the push method</h3>
5425
5426	<p>In order for the application to keep the control when the document is
5427	being fetched (which is common for GUI based programs) libxml2 provides a
5428	push interface, too, as of version 1.8.3. Here are the interface
5429	functions:</p>
5430	<pre>xmlParserCtxtPtr xmlCreatePushParserCtxt(xmlSAXHandlerPtr sax,
5431	void *user_data,
5432	const char *chunk,
5433	int size,
5434	const char *filename);
5435	int xmlParseChunk (xmlParserCtxtPtr ctxt,
5436	const char *chunk,
5437	int size,
5438	int terminate);</pre>
5439
5440	<p>and here is a simple example showing how to use the interface:</p>
5441	<pre> FILE *f;
5442
5443	f = fopen(filename, "r");
5444	if (f != NULL) {
5445	int res, size = 1024;
5446	char chars[1024];
5447	xmlParserCtxtPtr ctxt;
5448
5449	res = fread(chars, 1, 4, f);
5450	if (res > 0) {
5451	ctxt = xmlCreatePushParserCtxt(NULL, NULL,
5452	chars, res, filename);
5453	while ((res = fread(chars, 1, size, f)) > 0) {
5454	xmlParseChunk(ctxt, chars, res, 0);
5455	}
5456	xmlParseChunk(ctxt, chars, 0, 1);
5457	doc = ctxt->myDoc;
5458	xmlFreeParserCtxt(ctxt);
5459	}
5460	}</pre>
5461
5462	<p>The HTML parser embedded into libxml2 also has a push interface; the
5463	functions are just prefixed by "html" rather than "xml".</p>
5464
5465	<h3 id="Invoking2">Invoking the parser: the SAX interface</h3>
5466
5467	<p>The tree-building interface makes the parser memory-hungry, first loading
5468	the document in memory and then building the tree itself. Reading a document
5469	without building the tree is possible using the SAX interfaces (see SAX.h and
5470	<a href="http://www.daa.com.au/~james/gnome/xml-sax/xml-sax.html">James
5471	Henstridge's documentation</a>). Note also that the push interface can be
5472	limited to SAX: just use the two first arguments of
5473	<code>xmlCreatePushParserCtxt()</code>.</p>
5474
5475	<h3><a name="Building">Building a tree from scratch</a></h3>
5476
5477	<p>The other way to get an XML tree in memory is by building it. Basically
5478	there is a set of functions dedicated to building new elements. (These are
5479	also described in <libxml/tree.h>.) For example, here is a piece of
5480	code that produces the XML document used in the previous examples:</p>
5481	<pre> #include <libxml/tree.h>
5482	xmlDocPtr doc;
5483	xmlNodePtr tree, subtree;
5484
5485	doc = xmlNewDoc("1.0");
5486	doc->children = xmlNewDocNode(doc, NULL, "EXAMPLE", NULL);
5487	xmlSetProp(doc->children, "prop1", "gnome is great");
5488	xmlSetProp(doc->children, "prop2", "& linux too");
5489	tree = xmlNewChild(doc->children, NULL, "head", NULL);
5490	subtree = xmlNewChild(tree, NULL, "title", "Welcome to Gnome");
5491	tree = xmlNewChild(doc->children, NULL, "chapter", NULL);
5492	subtree = xmlNewChild(tree, NULL, "title", "The Linux adventure");
5493	subtree = xmlNewChild(tree, NULL, "p", "bla bla bla ...");
5494	subtree = xmlNewChild(tree, NULL, "image", NULL);
5495	xmlSetProp(subtree, "href", "linus.gif");</pre>
5496
5497	<p>Not really rocket science ...</p>
5498
5499	<h3><a name="Traversing">Traversing the tree</a></h3>
5500
5501	<p>Basically by <a href="html/libxml-tree.html">including "tree.h"</a> your
5502	code has access to the internal structure of all the elements of the tree.
5503	The names should be somewhat simple like <strong>parent</strong>,
5504	<strong>children</strong>, <strong>next</strong>, <strong>prev</strong>,
5505	<strong>properties</strong>, etc... For example, still with the previous
5506	example:</p>
5507	<pre><code>doc->children->children->children</code></pre>
5508
5509	<p>points to the title element,</p>
5510	<pre>doc->children->children->next->children->children</pre>
5511
5512	<p>points to the text node containing the chapter title "The Linux
5513	adventure".</p>
5514
5515	<p><strong>NOTE</strong>: XML allows <em>PI</em>s and <em>comments</em> to be
5516	present before the document root, so <code>doc->children</code> may point
5517	to an element which is not the document Root Element; a function
5518	<code>xmlDocGetRootElement()</code> was added for this purpose.</p>
5519
5520	<h3><a name="Modifying">Modifying the tree</a></h3>
5521
5522	<p>Functions are provided for reading and writing the document content. Here
5523	is an excerpt from the <a href="html/libxml-tree.html">tree API</a>:</p>
5524	<dl>
5525	<dt><code>xmlAttrPtr xmlSetProp(xmlNodePtr node, const xmlChar *name, const
5526	xmlChar *value);</code></dt>
5527	<dd><p>This sets (or changes) an attribute carried by an ELEMENT node.
5528	The value can be NULL.</p>
5529	</dd>
5530	</dl>
5531	<dl>
5532	<dt><code>const xmlChar *xmlGetProp(xmlNodePtr node, const xmlChar
5533	*name);</code></dt>
5534	<dd><p>This function returns a pointer to new copy of the property
5535	content. Note that the user must deallocate the result.</p>
5536	</dd>
5537	</dl>
5538
5539	<p>Two functions are provided for reading and writing the text associated
5540	with elements:</p>
5541	<dl>
5542	<dt><code>xmlNodePtr xmlStringGetNodeList(xmlDocPtr doc, const xmlChar
5543	*value);</code></dt>
5544	<dd><p>This function takes an "external" string and converts it to one
5545	text node or possibly to a list of entity and text nodes. All
5546	non-predefined entity references like &Gnome; will be stored
5547	internally as entity nodes, hence the result of the function may not be
5548	a single node.</p>
5549	</dd>
5550	</dl>
5551	<dl>
5552	<dt><code>xmlChar *xmlNodeListGetString(xmlDocPtr doc, xmlNodePtr list, int
5553	inLine);</code></dt>
5554	<dd><p>This function is the inverse of
5555	<code>xmlStringGetNodeList()</code>. It generates a new string
5556	containing the content of the text and entity nodes. Note the extra
5557	argument inLine. If this argument is set to 1, the function will expand
5558	entity references. For example, instead of returning the &Gnome;
5559	XML encoding in the string, it will substitute it with its value (say,
5560	"GNU Network Object Model Environment").</p>
5561	</dd>
5562	</dl>
5563
5564	<h3><a name="Saving">Saving a tree</a></h3>
5565
5566	<p>Basically 3 options are possible:</p>
5567	<dl>
5568	<dt><code>void xmlDocDumpMemory(xmlDocPtr cur, xmlChar**mem, int
5569	*size);</code></dt>
5570	<dd><p>Returns a buffer into which the document has been saved.</p>
5571	</dd>
5572	</dl>
5573	<dl>
5574	<dt><code>extern void xmlDocDump(FILE *f, xmlDocPtr doc);</code></dt>
5575	<dd><p>Dumps a document to an open file descriptor.</p>
5576	</dd>
5577	</dl>
5578	<dl>
5579	<dt><code>int xmlSaveFile(const char *filename, xmlDocPtr cur);</code></dt>
5580	<dd><p>Saves the document to a file. In this case, the compression
5581	interface is triggered if it has been turned on.</p>
5582	</dd>
5583	</dl>
5584
5585	<h3><a name="Compressio">Compression</a></h3>
5586
5587	<p>The library transparently handles compression when doing file-based
5588	accesses. The level of compression on saves can be turned on either globally
5589	or individually for one file:</p>
5590	<dl>
5591	<dt><code>int xmlGetDocCompressMode (xmlDocPtr doc);</code></dt>
5592	<dd><p>Gets the document compression ratio (0-9).</p>
5593	</dd>
5594	</dl>
5595	<dl>
5596	<dt><code>void xmlSetDocCompressMode (xmlDocPtr doc, int mode);</code></dt>
5597	<dd><p>Sets the document compression ratio.</p>
5598	</dd>
5599	</dl>
5600	<dl>
5601	<dt><code>int xmlGetCompressMode(void);</code></dt>
5602	<dd><p>Gets the default compression ratio.</p>
5603	</dd>
5604	</dl>
5605	<dl>
5606	<dt><code>void xmlSetCompressMode(int mode);</code></dt>
5607	<dd><p>Sets the default compression ratio.</p>
5608	</dd>
5609	</dl>
5610
5611	<h2><a name="Entities">Entities or no entities</a></h2>
5612
5613	<p>Entities in principle are similar to simple C macros. An entity defines an
5614	abbreviation for a given string that you can reuse many times throughout the
5615	content of your document. Entities are especially useful when a given string
5616	may occur frequently within a document, or to confine the change needed to a
5617	document to a restricted area in the internal subset of the document (at the
5618	beginning). Example:</p>
5619	<pre>1 <?xml version="1.0"?>
5620	2 <!DOCTYPE EXAMPLE SYSTEM "example.dtd" [
5621	3 <!ENTITY xml "Extensible Markup Language">
5622	4 ]>
5623	5 <EXAMPLE>
5624	6 &xml;
5625	7 </EXAMPLE></pre>
5626
5627	<p>Line 3 declares the xml entity. Line 6 uses the xml entity, by prefixing
5628	its name with '&' and following it by ';' without any spaces added. There
5629	are 5 predefined entities in libxml2 allowing you to escape characters with
5630	predefined meaning in some parts of the xml document content:
5631	<strong>&lt;</strong> for the character '<', <strong>&gt;</strong>
5632	for the character '>', <strong>&apos;</strong> for the character ''',
5633	<strong>&quot;</strong> for the character '"', and
5634	<strong>&amp;</strong> for the character '&'.</p>
5635
5636	<p>One of the problems related to entities is that you may want the parser to
5637	substitute an entity's content so that you can see the replacement text in
5638	your application. Or you may prefer to keep entity references as such in the
5639	content to be able to save the document back without losing this usually
5640	precious information (if the user went through the pain of explicitly
5641	defining entities, he may have a a rather negative attitude if you blindly
5642	substitute them as saving time). The <a
5643	href="html/libxml-parser.html#xmlSubstituteEntitiesDefault">xmlSubstituteEntitiesDefault()</a>
5644	function allows you to check and change the behaviour, which is to not
5645	substitute entities by default.</p>
5646
5647	<p>Here is the DOM tree built by libxml2 for the previous document in the
5648	default case:</p>
5649	<pre>/gnome/src/gnome-xml -> ./xmllint --debug test/ent1
5650	DOCUMENT
5651	version=1.0
5652	ELEMENT EXAMPLE
5653	TEXT
5654	content=
5655	ENTITY_REF
5656	INTERNAL_GENERAL_ENTITY xml
5657	content=Extensible Markup Language
5658	TEXT
5659	content=</pre>
5660
5661	<p>And here is the result when substituting entities:</p>
5662	<pre>/gnome/src/gnome-xml -> ./tester --debug --noent test/ent1
5663	DOCUMENT
5664	version=1.0
5665	ELEMENT EXAMPLE
5666	TEXT
5667	content= Extensible Markup Language</pre>
5668
5669	<p>So, entities or no entities? Basically, it depends on your use case. I
5670	suggest that you keep the non-substituting default behaviour and avoid using
5671	entities in your XML document or data if you are not willing to handle the
5672	entity references elements in the DOM tree.</p>
5673
5674	<p>Note that at save time libxml2 enforces the conversion of the predefined
5675	entities where necessary to prevent well-formedness problems, and will also
5676	transparently replace those with chars (i.e. it will not generate entity
5677	reference elements in the DOM tree or call the reference() SAX callback when
5678	finding them in the input).</p>
5679
5680	<p><span style="background-color: #FF0000">WARNING</span>: handling entities
5681	on top of the libxml2 SAX interface is difficult!!! If you plan to use
5682	non-predefined entities in your documents, then the learning curve to handle
5683	then using the SAX API may be long. If you plan to use complex documents, I
5684	strongly suggest you consider using the DOM interface instead and let libxml
5685	deal with the complexity rather than trying to do it yourself.</p>
5686
5687	<h2><a name="Namespaces">Namespaces</a></h2>
5688
5689	<p>The libxml2 library implements <a
5690	href="http://www.w3.org/TR/REC-xml-names/">XML namespaces</a> support by
5691	recognizing namespace constructs in the input, and does namespace lookup
5692	automatically when building the DOM tree. A namespace declaration is
5693	associated with an in-memory structure and all elements or attributes within
5694	that namespace point to it. Hence testing the namespace is a simple and fast
5695	equality operation at the user level.</p>
5696
5697	<p>I suggest that people using libxml2 use a namespace, and declare it in the
5698	root element of their document as the default namespace. Then they don't need
5699	to use the prefix in the content but we will have a basis for future semantic
5700	refinement and merging of data from different sources. This doesn't increase
5701	the size of the XML output significantly, but significantly increases its
5702	value in the long-term. Example:</p>
5703	<pre><mydoc xmlns="http://mydoc.example.org/schemas/">
5704	<elem1>...</elem1>
5705	<elem2>...</elem2>
5706	</mydoc></pre>
5707
5708	<p>The namespace value has to be an absolute URL, but the URL doesn't have to
5709	point to any existing resource on the Web. It will bind all the element and
5710	attributes with that URL. I suggest to use an URL within a domain you
5711	control, and that the URL should contain some kind of version information if
5712	possible. For example, <code>"http://www.gnome.org/gnumeric/1.0/"</code> is a
5713	good namespace scheme.</p>
5714
5715	<p>Then when you load a file, make sure that a namespace carrying the
5716	version-independent prefix is installed on the root element of your document,
5717	and if the version information don't match something you know, warn the user
5718	and be liberal in what you accept as the input. Also do not try to base
5719	namespace checking on the prefix value. <foo:text> may be exactly the
5720	same as <bar:text> in another document. What really matters is the URI
5721	associated with the element or the attribute, not the prefix string (which is
5722	just a shortcut for the full URI). In libxml, element and attributes have an
5723	<code>ns</code> field pointing to an xmlNs structure detailing the namespace
5724	prefix and its URI.</p>
5725
5726	<p>@@Interfaces@@</p>
5727	<pre>xmlNodePtr node;
5728	if(!strncmp(node->name,"mytag",5)
5729	&& node->ns
5730	&& !strcmp(node->ns->href,"http://www.mysite.com/myns/1.0")) {
5731	...
5732	}</pre>
5733
5734	<p>Usually people object to using namespaces together with validity checking.
5735	I will try to make sure that using namespaces won't break validity checking,
5736	so even if you plan to use or currently are using validation I strongly
5737	suggest adding namespaces to your document. A default namespace scheme
5738	<code>xmlns="http://...."</code> should not break validity even on less
5739	flexible parsers. Using namespaces to mix and differentiate content coming
5740	from multiple DTDs will certainly break current validation schemes. To check
5741	such documents one needs to use schema-validation, which is supported in
5742	libxml2 as well. See <a href="http://www.relaxng.org/">relagx-ng</a> and <a
5743	href="http://www.w3c.org/XML/Schema">w3c-schema</a>.</p>
5744
5745	<h2><a name="Upgrading">Upgrading 1.x code</a></h2>
5746
5747	<p>Incompatible changes:</p>
5748
5749	<p>Version 2 of libxml2 is the first version introducing serious backward
5750	incompatible changes. The main goals were:</p>
5751	<ul>
5752	<li>a general cleanup. A number of mistakes inherited from the very early
5753	versions couldn't be changed due to compatibility constraints. Example
5754	the "childs" element in the nodes.</li>
5755	<li>Uniformization of the various nodes, at least for their header and link
5756	parts (doc, parent, children, prev, next), the goal is a simpler
5757	programming model and simplifying the task of the DOM implementors.</li>
5758	<li>better conformances to the XML specification, for example version 1.x
5759	had an heuristic to try to detect ignorable white spaces. As a result the
5760	SAX event generated were ignorableWhitespace() while the spec requires
5761	character() in that case. This also mean that a number of DOM node
5762	containing blank text may populate the DOM tree which were not present
5763	before.</li>
5764	</ul>
5765
5766	<h3>How to fix libxml-1.x code:</h3>
5767
5768	<p>So client code of libxml designed to run with version 1.x may have to be
5769	changed to compile against version 2.x of libxml. Here is a list of changes
5770	that I have collected, they may not be sufficient, so in case you find other
5771	change which are required, <a href="mailto:[email protected]">drop me a
5772	mail</a>:</p>
5773	<ol>
5774	<li>The package name have changed from libxml to libxml2, the library name
5775	is now -lxml2 . There is a new xml2-config script which should be used to
5776	select the right parameters libxml2</li>
5777	<li>Node <strong>childs</strong> field has been renamed
5778	<strong>children</strong> so s/childs/children/g should be applied
5779	(probability of having "childs" anywhere else is close to 0+</li>
5780	<li>The document don't have anymore a <strong>root</strong> element it has
5781	been replaced by <strong>children</strong> and usually you will get a
5782	list of element here. For example a Dtd element for the internal subset
5783	and it's declaration may be found in that list, as well as processing
5784	instructions or comments found before or after the document root element.
5785	Use <strong>xmlDocGetRootElement(doc)</strong> to get the root element of
5786	a document. Alternatively if you are sure to not reference DTDs nor have
5787	PIs or comments before or after the root element
5788	s/->root/->children/g will probably do it.</li>
5789	<li>The white space issue, this one is more complex, unless special case of
5790	validating parsing, the line breaks and spaces usually used for indenting
5791	and formatting the document content becomes significant. So they are
5792	reported by SAX and if your using the DOM tree, corresponding nodes are
5793	generated. Too approach can be taken:
5794	<ol>
5795	<li>lazy one, use the compatibility call
5796	<strong>xmlKeepBlanksDefault(0)</strong> but be aware that you are
5797	relying on a special (and possibly broken) set of heuristics of
5798	libxml to detect ignorable blanks. Don't complain if it breaks or
5799	make your application not 100% clean w.r.t. to it's input.</li>
5800	<li>the Right Way: change you code to accept possibly insignificant
5801	blanks characters, or have your tree populated with weird blank text
5802	nodes. You can spot them using the commodity function
5803	<strong>xmlIsBlankNode(node)</strong> returning 1 for such blank
5804	nodes.</li>
5805	</ol>
5806	<p>Note also that with the new default the output functions don't add any
5807	extra indentation when saving a tree in order to be able to round trip
5808	(read and save) without inflating the document with extra formatting
5809	chars.</p>
5810	</li>
5811	<li>The include path has changed to $prefix/libxml/ and the includes
5812	themselves uses this new prefix in includes instructions... If you are
5813	using (as expected) the
5814	<pre>xml2-config --cflags</pre>
5815	<p>output to generate you compile commands this will probably work out of
5816	the box</p>
5817	</li>
5818	<li>xmlDetectCharEncoding takes an extra argument indicating the length in
5819	byte of the head of the document available for character detection.</li>
5820	</ol>
5821
5822	<h3>Ensuring both libxml-1.x and libxml-2.x compatibility</h3>
5823
5824	<p>Two new version of libxml (1.8.11) and libxml2 (2.3.4) have been released
5825	to allow smooth upgrade of existing libxml v1code while retaining
5826	compatibility. They offers the following:</p>
5827	<ol>
5828	<li>similar include naming, one should use
5829	<strong>#include<libxml/...></strong> in both cases.</li>
5830	<li>similar identifiers defined via macros for the child and root fields:
5831	respectively <strong>xmlChildrenNode</strong> and
5832	<strong>xmlRootNode</strong></li>
5833	<li>a new macro <strong>LIBXML_TEST_VERSION</strong> which should be
5834	inserted once in the client code</li>
5835	</ol>
5836
5837	<p>So the roadmap to upgrade your existing libxml applications is the
5838	following:</p>
5839	<ol>
5840	<li>install the libxml-1.8.8 (and libxml-devel-1.8.8) packages</li>
5841	<li>find all occurrences where the xmlDoc <strong>root</strong> field is
5842	used and change it to <strong>xmlRootNode</strong></li>
5843	<li>similarly find all occurrences where the xmlNode
5844	<strong>childs</strong> field is used and change it to
5845	<strong>xmlChildrenNode</strong></li>
5846	<li>add a <strong>LIBXML_TEST_VERSION</strong> macro somewhere in your
5847	<strong>main()</strong> or in the library init entry point</li>
5848	<li>Recompile, check compatibility, it should still work</li>
5849	<li>Change your configure script to look first for xml2-config and fall
5850	back using xml-config . Use the --cflags and --libs output of the command
5851	as the Include and Linking parameters needed to use libxml.</li>
5852	<li>install libxml2-2.3.x and libxml2-devel-2.3.x (libxml-1.8.y and
5853	libxml-devel-1.8.y can be kept simultaneously)</li>
5854	<li>remove your config.cache, relaunch your configuration mechanism, and
5855	recompile, if steps 2 and 3 were done right it should compile as-is</li>
5856	<li>Test that your application is still running correctly, if not this may
5857	be due to extra empty nodes due to formating spaces being kept in libxml2
5858	contrary to libxml1, in that case insert xmlKeepBlanksDefault(1) in your
5859	code before calling the parser (next to
5860	<strong>LIBXML_TEST_VERSION</strong> is a fine place).</li>
5861	</ol>
5862
5863	<p>Following those steps should work. It worked for some of my own code.</p>
5864
5865	<p>Let me put some emphasis on the fact that there is far more changes from
5866	libxml 1.x to 2.x than the ones you may have to patch for. The overall code
5867	has been considerably cleaned up and the conformance to the XML specification
5868	has been drastically improved too. Don't take those changes as an excuse to
5869	not upgrade, it may cost a lot on the long term ...</p>
5870
5871	<h2><a name="Thread">Thread safety</a></h2>
5872
5873	<p>Starting with 2.4.7, libxml2 makes provisions to ensure that concurrent
5874	threads can safely work in parallel parsing different documents. There is
5875	however a couple of things to do to ensure it:</p>
5876	<ul>
5877	<li>configure the library accordingly using the --with-threads options</li>
5878	<li>call xmlInitParser() in the "main" thread before using any of the
5879	libxml2 API (except possibly selecting a different memory allocator)</li>
5880	</ul>
5881
5882	<p>Note that the thread safety cannot be ensured for multiple threads sharing
5883	the same document, the locking must be done at the application level, libxml
5884	exports a basic mutex and reentrant mutexes API in <libxml/threads.h>.
5885	The parts of the library checked for thread safety are:</p>
5886	<ul>
5887	<li>concurrent loading</li>
5888	<li>file access resolution</li>
5889	<li>catalog access</li>
5890	<li>catalog building</li>
5891	<li>entities lookup/accesses</li>
5892	<li>validation</li>
5893	<li>global variables per-thread override</li>
5894	<li>memory handling</li>
5895	</ul>
5896
5897	<p>XPath has been tested for threaded usage on non-modified document
5898	for example when using libxslt, but make 100% sure the documents
5899	are accessed read-only !</p>
5900
5901	<h2><a name="DOM"></a><a name="Principles">DOM Principles</a></h2>
5902
5903	<p><a href="http://www.w3.org/DOM/">DOM</a> stands for the <em>Document
5904	Object Model</em>; this is an API for accessing XML or HTML structured
5905	documents. Native support for DOM in Gnome is on the way (module gnome-dom),
5906	and will be based on gnome-xml. This will be a far cleaner interface to
5907	manipulate XML files within Gnome since it won't expose the internal
5908	structure.</p>
5909
5910	<p>The current DOM implementation on top of libxml2 is the <a
5911	href="http://svn.gnome.org/viewvc/gdome2/trunk/">gdome2 Gnome module</a>, this
5912	is a full DOM interface, thanks to Paolo Casarini, check the <a
5913	href="http://gdome2.cs.unibo.it/">Gdome2 homepage</a> for more
5914	information.</p>
5915
5916	<h2><a name="Example"></a><a name="real">A real example</a></h2>
5917
5918	<p>Here is a real size example, where the actual content of the application
5919	data is not kept in the DOM tree but uses internal structures. It is based on
5920	a proposal to keep a database of jobs related to Gnome, with an XML based
5921	storage structure. Here is an <a href="gjobs.xml">XML encoded jobs
5922	base</a>:</p>
5923	<pre><?xml version="1.0"?>
5924	<gjob:Helping xmlns:gjob="http://www.gnome.org/some-location">
5925	<gjob:Jobs>
5926
5927	<gjob:Job>
5928	<gjob:Project ID="3"/>
5929	<gjob:Application>GBackup</gjob:Application>
5930	<gjob:Category>Development</gjob:Category>
5931
5932	<gjob:Update>
5933	<gjob:Status>Open</gjob:Status>
5934	<gjob:Modified>Mon, 07 Jun 1999 20:27:45 -0400 MET DST</gjob:Modified>
5935	<gjob:Salary>USD 0.00</gjob:Salary>
5936	</gjob:Update>
5937
5938	<gjob:Developers>
5939	<gjob:Developer>
5940	</gjob:Developer>
5941	</gjob:Developers>
5942
5943	<gjob:Contact>
5944	<gjob:Person>Nathan Clemons</gjob:Person>
5945	<gjob:Email>[email protected]</gjob:Email>
5946	<gjob:Company>
5947	</gjob:Company>
5948	<gjob:Organisation>
5949	</gjob:Organisation>
5950	<gjob:Webpage>
5951	</gjob:Webpage>
5952	<gjob:Snailmail>
5953	</gjob:Snailmail>
5954	<gjob:Phone>
5955	</gjob:Phone>
5956	</gjob:Contact>
5957
5958	<gjob:Requirements>
5959	The program should be released as free software, under the GPL.
5960	</gjob:Requirements>
5961
5962	<gjob:Skills>
5963	</gjob:Skills>
5964
5965	<gjob:Details>
5966	A GNOME based system that will allow a superuser to configure
5967	compressed and uncompressed files and/or file systems to be backed
5968	up with a supported media in the system. This should be able to
5969	perform via find commands generating a list of files that are passed
5970	to tar, dd, cpio, cp, gzip, etc., to be directed to the tape machine
5971	or via operations performed on the filesystem itself. Email
5972	notification and GUI status display very important.
5973	</gjob:Details>
5974
5975	</gjob:Job>
5976
5977	</gjob:Jobs>
5978	</gjob:Helping></pre>
5979
5980	<p>While loading the XML file into an internal DOM tree is a matter of
5981	calling only a couple of functions, browsing the tree to gather the data and
5982	generate the internal structures is harder, and more error prone.</p>
5983
5984	<p>The suggested principle is to be tolerant with respect to the input
5985	structure. For example, the ordering of the attributes is not significant,
5986	the XML specification is clear about it. It's also usually a good idea not to
5987	depend on the order of the children of a given node, unless it really makes
5988	things harder. Here is some code to parse the information for a person:</p>
5989	<pre>/*
5990	* A person record
5991	*/
5992	typedef struct person {
5993	char *name;
5994	char *email;
5995	char *company;
5996	char *organisation;
5997	char *smail;
5998	char *webPage;
5999	char *phone;
6000	} person, *personPtr;
6001
6002	/*
6003	* And the code needed to parse it
6004	*/
6005	personPtr parsePerson(xmlDocPtr doc, xmlNsPtr ns, xmlNodePtr cur) {
6006	personPtr ret = NULL;
6007
6008	DEBUG("parsePerson\n");
6009	/*
6010	* allocate the struct
6011	*/
6012	ret = (personPtr) malloc(sizeof(person));
6013	if (ret == NULL) {
6014	fprintf(stderr,"out of memory\n");
6015	return(NULL);
6016	}
6017	memset(ret, 0, sizeof(person));
6018
6019	/* We don't care what the top level element name is */
6020	cur = cur->xmlChildrenNode;
6021	while (cur != NULL) {
6022	if ((!strcmp(cur->name, "Person")) && (cur->ns == ns))
6023	ret->name = xmlNodeListGetString(doc, cur->xmlChildrenNode, 1);
6024	if ((!strcmp(cur->name, "Email")) && (cur->ns == ns))
6025	ret->email = xmlNodeListGetString(doc, cur->xmlChildrenNode, 1);
6026	cur = cur->next;
6027	}
6028
6029	return(ret);
6030	}</pre>
6031
6032	<p>Here are a couple of things to notice:</p>
6033	<ul>
6034	<li>Usually a recursive parsing style is the more convenient one: XML data
6035	is by nature subject to repetitive constructs and usually exhibits highly
6036	structured patterns.</li>
6037	<li>The two arguments of type <em>xmlDocPtr</em> and <em>xmlNsPtr</em>,
6038	i.e. the pointer to the global XML document and the namespace reserved to
6039	the application. Document wide information are needed for example to
6040	decode entities and it's a good coding practice to define a namespace for
6041	your application set of data and test that the element and attributes
6042	you're analyzing actually pertains to your application space. This is
6043	done by a simple equality test (cur->ns == ns).</li>
6044	<li>To retrieve text and attributes value, you can use the function
6045	<em>xmlNodeListGetString</em> to gather all the text and entity reference
6046	nodes generated by the DOM output and produce an single text string.</li>
6047	</ul>
6048
6049	<p>Here is another piece of code used to parse another level of the
6050	structure:</p>
6051	<pre>#include <libxml/tree.h>
6052	/*
6053	* a Description for a Job
6054	*/
6055	typedef struct job {
6056	char *projectID;
6057	char *application;
6058	char *category;
6059	personPtr contact;
6060	int nbDevelopers;
6061	personPtr developers[100]; /* using dynamic alloc is left as an exercise */
6062	} job, *jobPtr;
6063
6064	/*
6065	* And the code needed to parse it
6066	*/
6067	jobPtr parseJob(xmlDocPtr doc, xmlNsPtr ns, xmlNodePtr cur) {
6068	jobPtr ret = NULL;
6069
6070	DEBUG("parseJob\n");
6071	/*
6072	* allocate the struct
6073	*/
6074	ret = (jobPtr) malloc(sizeof(job));
6075	if (ret == NULL) {
6076	fprintf(stderr,"out of memory\n");
6077	return(NULL);
6078	}
6079	memset(ret, 0, sizeof(job));
6080
6081	/* We don't care what the top level element name is */
6082	cur = cur->xmlChildrenNode;
6083	while (cur != NULL) {
6084
6085	if ((!strcmp(cur->name, "Project")) && (cur->ns == ns)) {
6086	ret->projectID = xmlGetProp(cur, "ID");
6087	if (ret->projectID == NULL) {
6088	fprintf(stderr, "Project has no ID\n");
6089	}
6090	}
6091	if ((!strcmp(cur->name, "Application")) && (cur->ns == ns))
6092	ret->application = xmlNodeListGetString(doc, cur->xmlChildrenNode, 1);
6093	if ((!strcmp(cur->name, "Category")) && (cur->ns == ns))
6094	ret->category = xmlNodeListGetString(doc, cur->xmlChildrenNode, 1);
6095	if ((!strcmp(cur->name, "Contact")) && (cur->ns == ns))
6096	ret->contact = parsePerson(doc, ns, cur);
6097	cur = cur->next;
6098	}
6099
6100	return(ret);
6101	}</pre>
6102
6103	<p>Once you are used to it, writing this kind of code is quite simple, but
6104	boring. Ultimately, it could be possible to write stubbers taking either C
6105	data structure definitions, a set of XML examples or an XML DTD and produce
6106	the code needed to import and export the content between C data and XML
6107	storage. This is left as an exercise to the reader :-)</p>
6108
6109	<p>Feel free to use <a href="example/gjobread.c">the code for the full C
6110	parsing example</a> as a template, it is also available with Makefile in the
6111	Gnome SVN base under libxml2/example</p>
6112
6113	<h2><a name="Contributi">Contributions</a></h2>
6114	<ul>
6115	<li>Bjorn Reese, William Brack and Thomas Broyer have provided a number of
6116	patches, Gary Pennington worked on the validation API, threading support
6117	and Solaris port.</li>
6118	<li>John Fleck helps maintaining the documentation and man pages.</li>
6119	<li><a href="mailto:[email protected]">Igor Zlatkovic</a> is now the
6120	maintainer of the Windows port, <a
6121	href="http://www.zlatkovic.com/projects/libxml/index.html">he provides
6122	binaries</a></li>
6123	<li><a href="mailto:[email protected]">Gary Pennington</a> provides
6124	<a href="http://garypennington.net/libxml2/">Solaris binaries</a></li>
6125	<li><a
6126	href="http://mail.gnome.org/archives/xml/2001-March/msg00014.html">Matt
6127	Sergeant</a> developed <a
6128	href="http://axkit.org/download/">XML::LibXSLT</a>, a Perl wrapper for
6129	libxml2/libxslt as part of the <a href="http://axkit.com/">AxKit XML
6130	application server</a></li>
6131	<li><a href="mailto:[email protected]">Felix Natter</a> and <a
6132	href="mailto:[email protected]">Geert Kloosterman</a> provide <a
6133	href="libxml-doc.el">an emacs module</a> to lookup libxml(2) functions
6134	documentation</li>
6135	<li><a href="mailto:[email protected]">Ziying Sherwin</a> provided <a
6136	href="http://xmlsoft.org/messages/0488.html">man pages</a></li>
6137	<li>there is a module for <a
6138	href="http://acs-misc.sourceforge.net/nsxml.html">libxml/libxslt support
6139	in OpenNSD/AOLServer</a></li>
6140	<li><a href="mailto:[email protected]">Dave Kuhlman</a> provided the
6141	first version of libxml/libxslt <a
6142	href="http://www.rexx.com/~dkuhlman">wrappers for Python</a></li>
6143	<li>Petr Kozelka provides <a
6144	href="http://sourceforge.net/projects/libxml2-pas">Pascal units to glue
6145	libxml2</a> with Kylix and Delphi and other Pascal compilers</li>
6146	<li><a href="mailto:[email protected]">Aleksey Sanin</a> implemented the
6147	<a href="http://www.w3.org/Signature/">XML Canonicalization and XML
6148	Digital Signature</a> <a
6149	href="http://www.aleksey.com/xmlsec/">implementations for libxml2</a></li>
6150	<li><a href="mailto:[email protected]">Steve Ball</a> and
6151	contributors maintain <a href="http://tclxml.sourceforge.net/">tcl
6152	bindings for libxml2 and libxslt</a>, as well as <a
6153	href="http://tclxml.sf.net/tkxmllint.html">tkxmllint</a> a GUI for
6154	xmllint and <a href="http://tclxml.sf.net/tkxsltproc.html">tkxsltproc</a>
6155	a GUI for xsltproc.</li>
6156	</ul>
6157
6158	<p></p>
6159	</body>
6160	</html>

注意: 瀏覽 TracBrowser 來幫助您使用儲存庫瀏覽器

source: vbox/trunk/src/libs/libxml2-2.9.4/doc/xml.html@ 67456

以其他格式下載: