1 /*
2 * libjingle
3 * Copyright 2004--2005, Google Inc.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions are met:
7 *
8 * 1. Redistributions of source code must retain the above copyright notice,
9 * this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright notice,
11 * this list of conditions and the following disclaimer in the documentation
12 * and/or other materials provided with the distribution.
13 * 3. The name of the author may not be used to endorse or promote products
14 * derived from this software without specific prior written permission.
15 *
16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
17 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
18 * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
19 * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
20 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
21 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
22 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
23 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
24 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
25 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 */
27
28 #include "talk/xmllite/xmlparser.h"
29
30 #include <string>
31 #include <vector>
32 #include <iostream>
33 #include "talk/base/common.h"
34 #include "talk/xmllite/xmlelement.h"
35 #include "talk/xmllite/xmlnsstack.h"
36 #include "talk/xmllite/xmlconstants.h"
37 #include "talk/xmllite/xmlnsstack.h"
38 #ifdef EXPAT_RELATIVE_PATH
39 #include "lib/expat.h"
40 #else
41 #include "third_party/expat/v2_0_1/Source/lib/expat.h"
42 #endif // EXPAT_RELATIVE_PATH
43
44 namespace buzz {
45
46
47 static void
StartElementCallback(void * userData,const char * name,const char ** atts)48 StartElementCallback(void * userData, const char *name, const char **atts) {
49 (static_cast<XmlParser *>(userData))->ExpatStartElement(name, atts);
50 }
51
52 static void
EndElementCallback(void * userData,const char * name)53 EndElementCallback(void * userData, const char *name) {
54 (static_cast<XmlParser *>(userData))->ExpatEndElement(name);
55 }
56
57 static void
CharacterDataCallback(void * userData,const char * text,int len)58 CharacterDataCallback(void * userData, const char *text, int len) {
59 (static_cast<XmlParser *>(userData))->ExpatCharacterData(text, len);
60 }
61
62 static void
XmlDeclCallback(void * userData,const char * ver,const char * enc,int st)63 XmlDeclCallback(void * userData, const char * ver, const char * enc, int st) {
64 (static_cast<XmlParser *>(userData))->ExpatXmlDecl(ver, enc, st);
65 }
66
XmlParser(XmlParseHandler * pxph)67 XmlParser::XmlParser(XmlParseHandler *pxph) :
68 context_(this), pxph_(pxph), sentError_(false) {
69 expat_ = XML_ParserCreate(NULL);
70 XML_SetUserData(expat_, this);
71 XML_SetElementHandler(expat_, StartElementCallback, EndElementCallback);
72 XML_SetCharacterDataHandler(expat_, CharacterDataCallback);
73 XML_SetXmlDeclHandler(expat_, XmlDeclCallback);
74 }
75
76 void
Reset()77 XmlParser::Reset() {
78 if (!XML_ParserReset(expat_, NULL)) {
79 XML_ParserFree(expat_);
80 expat_ = XML_ParserCreate(NULL);
81 }
82 XML_SetUserData(expat_, this);
83 XML_SetElementHandler(expat_, StartElementCallback, EndElementCallback);
84 XML_SetCharacterDataHandler(expat_, CharacterDataCallback);
85 XML_SetXmlDeclHandler(expat_, XmlDeclCallback);
86 context_.Reset();
87 sentError_ = false;
88 }
89
90 static bool
XmlParser_StartsWithXmlns(const char * name)91 XmlParser_StartsWithXmlns(const char *name) {
92 return name[0] == 'x' &&
93 name[1] == 'm' &&
94 name[2] == 'l' &&
95 name[3] == 'n' &&
96 name[4] == 's';
97 }
98
99 void
ExpatStartElement(const char * name,const char ** atts)100 XmlParser::ExpatStartElement(const char *name, const char **atts) {
101 if (context_.RaisedError() != XML_ERROR_NONE)
102 return;
103 const char **att;
104 context_.StartElement();
105 for (att = atts; *att; att += 2) {
106 if (XmlParser_StartsWithXmlns(*att)) {
107 if ((*att)[5] == '\0') {
108 context_.StartNamespace("", *(att + 1));
109 }
110 else if ((*att)[5] == ':') {
111 if (**(att + 1) == '\0') {
112 // In XML 1.0 empty namespace illegal with prefix (not in 1.1)
113 context_.RaiseError(XML_ERROR_SYNTAX);
114 return;
115 }
116 context_.StartNamespace((*att) + 6, *(att + 1));
117 }
118 }
119 }
120 context_.SetPosition(XML_GetCurrentLineNumber(expat_),
121 XML_GetCurrentColumnNumber(expat_),
122 XML_GetCurrentByteIndex(expat_));
123 pxph_->StartElement(&context_, name, atts);
124 }
125
126 void
ExpatEndElement(const char * name)127 XmlParser::ExpatEndElement(const char *name) {
128 if (context_.RaisedError() != XML_ERROR_NONE)
129 return;
130 context_.EndElement();
131 context_.SetPosition(XML_GetCurrentLineNumber(expat_),
132 XML_GetCurrentColumnNumber(expat_),
133 XML_GetCurrentByteIndex(expat_));
134 pxph_->EndElement(&context_, name);
135 }
136
137 void
ExpatCharacterData(const char * text,int len)138 XmlParser::ExpatCharacterData(const char *text, int len) {
139 if (context_.RaisedError() != XML_ERROR_NONE)
140 return;
141 context_.SetPosition(XML_GetCurrentLineNumber(expat_),
142 XML_GetCurrentColumnNumber(expat_),
143 XML_GetCurrentByteIndex(expat_));
144 pxph_->CharacterData(&context_, text, len);
145 }
146
147 void
ExpatXmlDecl(const char * ver,const char * enc,int standalone)148 XmlParser::ExpatXmlDecl(const char * ver, const char * enc, int standalone) {
149 if (context_.RaisedError() != XML_ERROR_NONE)
150 return;
151
152 if (ver && std::string("1.0") != ver) {
153 context_.RaiseError(XML_ERROR_SYNTAX);
154 return;
155 }
156
157 if (standalone == 0) {
158 context_.RaiseError(XML_ERROR_SYNTAX);
159 return;
160 }
161
162 if (enc && !((enc[0] == 'U' || enc[0] == 'u') &&
163 (enc[1] == 'T' || enc[1] == 't') &&
164 (enc[2] == 'F' || enc[2] == 'f') &&
165 enc[3] == '-' && enc[4] =='8')) {
166 context_.RaiseError(XML_ERROR_INCORRECT_ENCODING);
167 return;
168 }
169
170 }
171
172 bool
Parse(const char * data,size_t len,bool isFinal)173 XmlParser::Parse(const char *data, size_t len, bool isFinal) {
174 if (sentError_)
175 return false;
176
177 if (XML_Parse(expat_, data, static_cast<int>(len), isFinal) !=
178 XML_STATUS_OK) {
179 context_.SetPosition(XML_GetCurrentLineNumber(expat_),
180 XML_GetCurrentColumnNumber(expat_),
181 XML_GetCurrentByteIndex(expat_));
182 context_.RaiseError(XML_GetErrorCode(expat_));
183 }
184
185 if (context_.RaisedError() != XML_ERROR_NONE) {
186 sentError_ = true;
187 pxph_->Error(&context_, context_.RaisedError());
188 return false;
189 }
190
191 return true;
192 }
193
~XmlParser()194 XmlParser::~XmlParser() {
195 XML_ParserFree(expat_);
196 }
197
198 void
ParseXml(XmlParseHandler * pxph,std::string text)199 XmlParser::ParseXml(XmlParseHandler *pxph, std::string text) {
200 XmlParser parser(pxph);
201 parser.Parse(text.c_str(), text.length(), true);
202 }
203
ParseContext(XmlParser * parser)204 XmlParser::ParseContext::ParseContext(XmlParser *parser) :
205 parser_(parser),
206 xmlnsstack_(),
207 raised_(XML_ERROR_NONE),
208 line_number_(0),
209 column_number_(0),
210 byte_index_(0) {
211 }
212
213 void
StartNamespace(const char * prefix,const char * ns)214 XmlParser::ParseContext::StartNamespace(const char *prefix, const char *ns) {
215 xmlnsstack_.AddXmlns(
216 *prefix ? std::string(prefix) : STR_EMPTY,
217 // ns == NS_CLIENT ? NS_CLIENT :
218 // ns == NS_ROSTER ? NS_ROSTER :
219 // ns == NS_GR ? NS_GR :
220 std::string(ns));
221 }
222
223 void
StartElement()224 XmlParser::ParseContext::StartElement() {
225 xmlnsstack_.PushFrame();
226 }
227
228 void
EndElement()229 XmlParser::ParseContext::EndElement() {
230 xmlnsstack_.PopFrame();
231 }
232
233 QName
ResolveQName(const char * qname,bool isAttr)234 XmlParser::ParseContext::ResolveQName(const char *qname, bool isAttr) {
235 const char *c;
236 for (c = qname; *c; ++c) {
237 if (*c == ':') {
238 const std::string * result;
239 result = xmlnsstack_.NsForPrefix(std::string(qname, c - qname));
240 if (result == NULL)
241 return QN_EMPTY;
242 const char * localname = c + 1;
243 return QName(*result, localname);
244 }
245 }
246 if (isAttr) {
247 return QName(STR_EMPTY, qname);
248 }
249
250 const std::string * result;
251 result = xmlnsstack_.NsForPrefix(STR_EMPTY);
252 if (result == NULL)
253 return QN_EMPTY;
254
255 return QName(*result, qname);
256 }
257
258 void
Reset()259 XmlParser::ParseContext::Reset() {
260 xmlnsstack_.Reset();
261 raised_ = XML_ERROR_NONE;
262 }
263
264 void
SetPosition(int line,int column,long byte_index)265 XmlParser::ParseContext::SetPosition(int line, int column,
266 long byte_index) {
267 line_number_ = line;
268 column_number_ = column;
269 byte_index_ = byte_index;
270 }
271
272 void
GetPosition(unsigned long * line,unsigned long * column,unsigned long * byte_index)273 XmlParser::ParseContext::GetPosition(unsigned long * line,
274 unsigned long * column,
275 unsigned long * byte_index) {
276 if (line != NULL) {
277 *line = static_cast<unsigned long>(line_number_);
278 }
279
280 if (column != NULL) {
281 *column = static_cast<unsigned long>(column_number_);
282 }
283
284 if (byte_index != NULL) {
285 *byte_index = static_cast<unsigned long>(byte_index_);
286 }
287 }
288
~ParseContext()289 XmlParser::ParseContext::~ParseContext() {
290 }
291
292 }
293