encoding.html revision 8acca114eef1383702a6d6d97588e930cd650b9c
1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/1999/REC-html401-19991224/loose.dtd"> 2<html> 3<head> 4<meta content="text/html; charset=ISO-8859-1" http-equiv="Content-Type"> 5<style type="text/css"><!-- 6TD {font-size: 14pt; font-family: Verdana,Arial,Helvetica} 7BODY {font-size: 14pt; font-family: Verdana,Arial,Helvetica; margin-top: 2em; margin-left: 0em; margin-right: 0em} 8H1 {font-size: 20pt; font-family: Verdana,Arial,Helvetica} 9H2 {font-size: 18pt; font-family: Verdana,Arial,Helvetica} 10H3 {font-size: 16pt; font-family: Verdana,Arial,Helvetica} 11A:link, A:visited, A:active { text-decoration: underline } 12--></style> 13<title>Encodings support</title> 14</head> 15<body bgcolor="#8b7765" text="#000000" link="#000000" vlink="#000000"> 16<table border="0" width="100%" cellpadding="5" cellspacing="0" align="center"><tr> 17<td width="180"> 18<a href="http://www.gnome.org/"><img src="smallfootonly.gif" alt="Gnome Logo"></a><a href="http://www.w3.org/Status"><img src="w3c.png" alt="W3C Logo"></a><a href="http://www.redhat.com/"><img src="redhat.gif" alt="Red Hat Logo"></a> 19</td> 20<td><table border="0" width="90%" cellpadding="2" cellspacing="0" align="center" bgcolor="#000000"><tr><td><table width="100%" border="0" cellspacing="1" cellpadding="3" bgcolor="#fffacd"><tr><td align="center"> 21<h1>The XML C library for Gnome</h1> 22<h2>Encodings support</h2> 23</td></tr></table></td></tr></table></td> 24</tr></table> 25<table border="0" cellpadding="4" cellspacing="0" width="100%" align="center"><tr><td bgcolor="#8b7765"><table border="0" cellspacing="0" cellpadding="2" width="100%"><tr> 26<td valign="top" width="200" bgcolor="#8b7765"><table border="0" cellspacing="0" cellpadding="1" width="100%" bgcolor="#000000"><tr><td> 27<table width="100%" border="0" cellspacing="1" cellpadding="3"> 28<tr><td colspan="1" bgcolor="#eecfa1" align="center"><center><b>Main Menu</b></center></td></tr> 29<tr><td bgcolor="#fffacd"><ul> 30<li><a href="index.html">Home</a></li> 31<li><a href="intro.html">Introduction</a></li> 32<li><a href="FAQ.html">FAQ</a></li> 33<li><a href="docs.html">Documentation</a></li> 34<li><a href="bugs.html">Reporting bugs and getting help</a></li> 35<li><a href="help.html">How to help</a></li> 36<li><a href="downloads.html">Downloads</a></li> 37<li><a href="news.html">News</a></li> 38<li><a href="XMLinfo.html">XML</a></li> 39<li><a href="XSLT.html">XSLT</a></li> 40<li><a href="architecture.html">libxml architecture</a></li> 41<li><a href="tree.html">The tree output</a></li> 42<li><a href="interface.html">The SAX interface</a></li> 43<li><a href="xmldtd.html">Validation & DTDs</a></li> 44<li><a href="xmlmem.html">Memory Management</a></li> 45<li><a href="encoding.html">Encodings support</a></li> 46<li><a href="xmlio.html">I/O Interfaces</a></li> 47<li><a href="catalog.html">Catalog support</a></li> 48<li><a href="library.html">The parser interfaces</a></li> 49<li><a href="entities.html">Entities or no entities</a></li> 50<li><a href="namespaces.html">Namespaces</a></li> 51<li><a href="upgrade.html">Upgrading 1.x code</a></li> 52<li><a href="threads.html">Thread safety</a></li> 53<li><a href="DOM.html">DOM Principles</a></li> 54<li><a href="example.html">A real example</a></li> 55<li><a href="contribs.html">Contributions</a></li> 56<li> 57<a href="xml.html">flat page</a>, <a href="site.xsl">stylesheet</a> 58</li> 59</ul></td></tr> 60</table> 61<table width="100%" border="0" cellspacing="1" cellpadding="3"> 62<tr><td colspan="1" bgcolor="#eecfa1" align="center"><center><b>Related links</b></center></td></tr> 63<tr><td bgcolor="#fffacd"><ul> 64<li><a href="http://mail.gnome.org/archives/xml/">Mail archive</a></li> 65<li><a href="http://xmlsoft.org/XSLT/">XSLT libxslt</a></li> 66<li><a href="http://phd.cs.unibo.it/gdome2/">DOM gdome2</a></li> 67<li><a href="ftp://xmlsoft.org/">FTP</a></li> 68<li><a href="http://www.fh-frankfurt.de/~igor/projects/libxml/">Windows binaries</a></li> 69<li><a href="http://garypennington.net/libxml2/">Solaris binaries</a></li> 70<li><a href="http://bugzilla.gnome.org/buglist.cgi?product=libxml">Bug Tracker</a></li> 71</ul></td></tr> 72</table> 73</td></tr></table></td> 74<td valign="top" bgcolor="#8b7765"><table border="0" cellspacing="0" cellpadding="1" width="100%"><tr><td><table border="0" cellspacing="0" cellpadding="1" width="100%" bgcolor="#000000"><tr><td><table border="0" cellpadding="3" cellspacing="1" width="100%"><tr><td bgcolor="#fffacd"> 75<p>Table of Content:</p> 76<ol> 77<li><a href="encoding.html#What">What does internationalization support 78 mean ?</a></li> 79<li><a href="encoding.html#internal">The internal encoding, how and 80 why</a></li> 81<li><a href="encoding.html#implemente">How is it implemented ?</a></li> 82<li><a href="encoding.html#Default">Default supported encodings</a></li> 83<li><a href="encoding.html#extend">How to extend the existing 84 support</a></li> 85</ol> 86<h3><a name="What">What does internationalization support mean ?</a></h3> 87<p>XML was designed from the start to allow the support of any character set 88by using Unicode. Any conformant XML parser has to support the UTF-8 and 89UTF-16 default encodings which can both express the full unicode ranges. UTF8 90is a variable length encoding whose greatest point are to resuse the same 91emcoding for ASCII and to save space for Western encodings, but it is a bit 92more complex to handle in practice. UTF-16 use 2 bytes per characters (and 93sometimes combines two pairs), it makes implementation easier, but looks a 94bit overkill for Western languages encoding. Moreover the XML specification 95allows document to be encoded in other encodings at the condition that they 96are clearly labelled as such. For example the following is a wellformed XML 97document encoded in ISO-8859 1 and using accentuated letter that we French 98likes for both markup and content:</p> 99<pre><?xml version="1.0" encoding="ISO-8859-1"?> 100<tr�s>l�</tr�s></pre> 101<p>Having internationalization support in libxml means the foolowing:</p> 102<ul> 103<li>the document is properly parsed</li> 104<li>informations about it's encoding are saved</li> 105<li>it can be modified</li> 106<li>it can be saved in its original encoding</li> 107<li>it can also be saved in another encoding supported by libxml (for 108 example straight UTF8 or even an ASCII form)</li> 109</ul> 110<p>Another very important point is that the whole libxml API, with the 111exception of a few routines to read with a specific encoding or save to a 112specific encoding, is completely agnostic about the original encoding of the 113document.</p> 114<p>It should be noted too that the HTML parser embedded in libxml now obbey 115the same rules too, the following document will be (as of 2.2.2) handled in 116an internationalized fashion by libxml too:</p> 117<pre><!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" 118 "http://www.w3.org/TR/REC-html40/loose.dtd">; 119<html lang="fr"> 120<head> 121 <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=ISO-8859-1"> 122</head> 123<body> 124<p>W3C cr�e des standards pour le Web.</body> 125</html></pre> 126<h3><a name="internal">The internal encoding, how and why</a></h3> 127<p>One of the core decision was to force all documents to be converted to a 128default internal encoding, and that encoding to be UTF-8, here are the 129rationale for those choices:</p> 130<ul> 131<li>keeping the native encoding in the internal form would force the libxml 132 users (or the code associated) to be fully aware of the encoding of the 133 original document, for examples when adding a text node to a document, 134 the content would have to be provided in the document encoding, i.e. the 135 client code would have to check it before hand, make sure it's conformant 136 to the encoding, etc ... Very hard in practice, though in some specific 137 cases this may make sense.</li> 138<li>the second decision was which encoding. From the XML spec only UTF8 and 139 UTF16 really makes sense as being the two only encodings for which there 140 is amndatory support. UCS-4 (32 bits fixed size encoding) could be 141 considered an intelligent choice too since it's a direct Unicode mapping 142 support. I selected UTF-8 on the basis of efficiency and compatibility 143 with surrounding software: 144 <ul> 145<li>UTF-8 while a bit more complex to convert from/to (i.e. slightly 146 more costly to import and export CPU wise) is also far more compact 147 than UTF-16 (and UCS-4) for a majority of the documents I see it used 148 for right now (RPM RDF catalogs, advogato data, various configuration 149 file formats, etc.) and the key point for today's computer 150 architecture is efficient uses of caches. If one nearly double the 151 memory requirement to store the same amount of data, this will trash 152 caches (main memory/external caches/internal caches) and my take is 153 that this harms the system far more than the CPU requirements needed 154 for the conversion to UTF-8</li> 155<li>Most of libxml version 1 users were using it with straight ASCII 156 most of the time, doing the conversion with an internal encoding 157 requiring all their code to be rewritten was a serious show-stopper 158 for using UTF-16 or UCS-4.</li> 159<li>UTF-8 is being used as the de-facto internal encoding standard for 160 related code like the <a href="http://www.pango.org/">pango</a> 161 upcoming Gnome text widget, and a lot of Unix code (yep another place 162 where Unix programmer base takes a different approach from Microsoft 163 - they are using UTF-16)</li> 164</ul> 165</li> 166</ul> 167<p>What does this mean in practice for the libxml user:</p> 168<ul> 169<li>xmlChar, the libxml data type is a byte, those bytes must be assembled 170 as UTF-8 valid strings. The proper way to terminate an xmlChar * string 171 is simply to append 0 byte, as usual.</li> 172<li>One just need to make sure that when using chars outside the ASCII set, 173 the values has been properly converted to UTF-8</li> 174</ul> 175<h3><a name="implemente">How is it implemented ?</a></h3> 176<p>Let's describe how all this works within libxml, basically the I18N 177(internationalization) support get triggered only during I/O operation, i.e. 178when reading a document or saving one. Let's look first at the reading 179sequence:</p> 180<ol> 181<li>when a document is processed, we usually don't know the encoding, a 182 simple heuristic allows to detect UTF-18 and UCS-4 from whose where the 183 ASCII range (0-0x7F) maps with ASCII</li> 184<li>the xml declaration if available is parsed, including the encoding 185 declaration. At that point, if the autodetected encoding is different 186 from the one declared a call to xmlSwitchEncoding() is issued.</li> 187<li>If there is no encoding declaration, then the input has to be in either 188 UTF-8 or UTF-16, if it is not then at some point when processing the 189 input, the converter/checker of UTF-8 form will raise an encoding error. 190 You may end-up with a garbled document, or no document at all ! Example: 191 <pre>~/XML -> /xmllint err.xml 192err.xml:1: error: Input is not proper UTF-8, indicate encoding ! 193<tr�s>l�</tr�s> 194 ^ 195err.xml:1: error: Bytes: 0xE8 0x73 0x3E 0x6C 196<tr�s>l�</tr�s> 197 ^</pre> 198</li> 199<li>xmlSwitchEncoding() does an encoding name lookup, canonalize it, and 200 then search the default registered encoding converters for that encoding. 201 If it's not within the default set and iconv() support has been compiled 202 it, it will ask iconv for such an encoder. If this fails then the parser 203 will report an error and stops processing: 204 <pre>~/XML -> /xmllint err2.xml 205err2.xml:1: error: Unsupported encoding UnsupportedEnc 206<?xml version="1.0" encoding="UnsupportedEnc"?> 207 ^</pre> 208</li> 209<li>From that point the encoder process progressingly the input (it is 210 plugged as a front-end to the I/O module) for that entity. It captures 211 and convert on-the-fly the document to be parsed to UTF-8. The parser 212 itself just does UTF-8 checking of this input and process it 213 transparently. The only difference is that the encoding information has 214 been added to the parsing context (more precisely to the input 215 corresponding to this entity).</li> 216<li>The result (when using DOM) is an internal form completely in UTF-8 217 with just an encoding information on the document node.</li> 218</ol> 219<p>Ok then what's happen when saving the document (assuming you 220colllected/built an xmlDoc DOM like structure) ? It depends on the function 221called, xmlSaveFile() will just try to save in the original encoding, while 222xmlSaveFileTo() and xmlSaveFileEnc() can optionally save to a given 223encoding:</p> 224<ol> 225<li>if no encoding is given, libxml will look for an encoding value 226 associated to the document and if it exists will try to save to that 227 encoding, 228 <p>otherwise everything is written in the internal form, i.e. UTF-8</p> 229</li> 230<li>so if an encoding was specified, either at the API level or on the 231 document, libxml will again canonalize the encoding name, lookup for a 232 converter in the registered set or through iconv. If not found the 233 function will return an error code</li> 234<li>the converter is placed before the I/O buffer layer, as another kind of 235 buffer, then libxml will simply push the UTF-8 serialization to through 236 that buffer, which will then progressively be converted and pushed onto 237 the I/O layer.</li> 238<li>It is possible that the converter code fails on some input, for example 239 trying to push an UTF-8 encoded chinese character through the UTF-8 to 240 ISO-8859-1 converter won't work. Since the encoders are progressive they 241 will just report the error and the number of bytes converted, at that 242 point libxml will decode the offending character, remove it from the 243 buffer and replace it with the associated charRef encoding &#123; and 244 resume the convertion. This guarante that any document will be saved 245 without losses (except for markup names where this is not legal, this is 246 a problem in the current version, in pactice avoid using non-ascci 247 characters for tags or attributes names @@). A special "ascii" encoding 248 name is used to save documents to a pure ascii form can be used when 249 portability is really crucial</li> 250</ol> 251<p>Here is a few examples based on the same test document:</p> 252<pre>~/XML -> /xmllint isolat1 253<?xml version="1.0" encoding="ISO-8859-1"?> 254<tr�s>l�</tr�s> 255~/XML -> /xmllint --encode UTF-8 isolat1 256<?xml version="1.0" encoding="UTF-8"?> 257<très>l� �</très> 258~/XML -> </pre> 259<p>The same processing is applied (and reuse most of the code) for HTML I18N 260processing. Looking up and modifying the content encoding is a bit more 261difficult since it is located in a <meta> tag under the <head>, 262so a couple of functions htmlGetMetaEncoding() and htmlSetMetaEncoding() have 263been provided. The parser also attempts to switch encoding on the fly when 264detecting such a tag on input. Except for that the processing is the same 265(and again reuses the same code).</p> 266<h3><a name="Default">Default supported encodings</a></h3> 267<p>libxml has a set of default converters for the following encodings 268(located in encoding.c):</p> 269<ol> 270<li>UTF-8 is supported by default (null handlers)</li> 271<li>UTF-16, both little and big endian</li> 272<li>ISO-Latin-1 (ISO-8859-1) covering most western languages</li> 273<li>ASCII, useful mostly for saving</li> 274<li>HTML, a specific handler for the conversion of UTF-8 to ASCII with HTML 275 predefined entities like &copy; for the Copyright sign.</li> 276</ol> 277<p>More over when compiled on an Unix platfor with iconv support the full set 278of encodings supported by iconv can be instantly be used by libxml. On a 279linux machine with glibc-2.1 the list of supported encodings and aliases fill 2803 full pages, and include UCS-4, the full set of ISO-Latin encodings, and the 281various Japanese ones.</p> 282<h4>Encoding aliases</h4> 283<p>From 2.2.3, libxml has support to register encoding names aliases. The 284goal is to be able to parse document whose encoding is supported but where 285the name differs (for example from the default set of names accepted by 286iconv). The following functions allow to register and handle new aliases for 287existing encodings. Once registered libxml will automatically lookup the 288aliases when handling a document:</p> 289<ul> 290<li>int xmlAddEncodingAlias(const char *name, const char *alias);</li> 291<li>int xmlDelEncodingAlias(const char *alias);</li> 292<li>const char * xmlGetEncodingAlias(const char *alias);</li> 293<li>void xmlCleanupEncodingAliases(void);</li> 294</ul> 295<h3><a name="extend">How to extend the existing support</a></h3> 296<p>Well adding support for new encoding, or overriding one of the encoders 297(assuming it is buggy) should not be hard, just write an input and output 298conversion routines to/from UTF-8, and register them using 299xmlNewCharEncodingHandler(name, xxxToUTF8, UTF8Toxxx), and they will be 300called automatically if the parser(s) encounter such an encoding name 301(register it uppercase, this will help). The description of the encoders, 302their arguments and expected return values are described in the encoding.h 303header.</p> 304<p>A quick note on the topic of subverting the parser to use a different 305internal encoding than UTF-8, in some case people will absolutely want to 306keep the internal encoding different, I think it's still possible (but the 307encoding must be compliant with ASCII on the same subrange) though I didn't 308tried it. The key is to override the default conversion routines (by 309registering null encoders/decoders for your charsets), and bypass the UTF-8 310checking of the parser by setting the parser context charset 311(ctxt->charset) to something different than XML_CHAR_ENCODING_UTF8, but 312there is no guarantee taht this will work. You may also have some troubles 313saving back.</p> 314<p>Basically proper I18N support is important, this requires at least 315libxml-2.0.0, but a lot of features and corrections are really available only 316starting 2.2.</p> 317<p><a href="mailto:daniel@veillard.com">Daniel Veillard</a></p> 318</td></tr></table></td></tr></table></td></tr></table></td> 319</tr></table></td></tr></table> 320</body> 321</html> 322
