00001 /* 00002 The contents of this file are subject to the Mozilla Public License 00003 Version 1.1 (the "License"); you may not use this file except in 00004 compliance with the License. You may obtain a copy of the License at 00005 http://www.mozilla.org/MPL/ 00006 00007 Software distributed under the License is distributed on an "AS IS" 00008 basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See the 00009 License for the specific language governing rights and limitations 00010 under the License. 00011 00012 The Original Code is expat. 00013 00014 The Initial Developer of the Original Code is James Clark. 00015 Portions created by James Clark are Copyright (C) 1998, 1999 00016 James Clark. All Rights Reserved. 00017 00018 Contributor(s): 00019 00020 Alternatively, the contents of this file may be used under the terms 00021 of the GNU General Public License (the "GPL"), in which case the 00022 provisions of the GPL are applicable instead of those above. If you 00023 wish to allow use of your version of this file only under the terms of 00024 the GPL and not to allow others to use your version of this file under 00025 the MPL, indicate your decision by deleting the provisions above and 00026 replace them with the notice and other provisions required by the 00027 GPL. If you do not delete the provisions above, a recipient may use 00028 your version of this file under either the MPL or the GPL. 00029 */ 00030 00031 #ifndef XmlParse_INCLUDED 00032 #define XmlParse_INCLUDED 1 00033 00034 #ifdef __cplusplus 00035 extern "C" { 00036 #endif 00037 00038 #ifndef XMLPARSEAPI 00039 #define XMLPARSEAPI /* as nothing */ 00040 #endif 00041 00042 typedef void *XML_Parser; 00043 00044 #ifdef XML_UNICODE_WCHAR_T 00045 00046 /* XML_UNICODE_WCHAR_T will work only if sizeof(wchar_t) == 2 and wchar_t 00047 uses Unicode. */ 00048 /* Information is UTF-16 encoded as wchar_ts */ 00049 00050 #ifndef XML_UNICODE 00051 #define XML_UNICODE 00052 #endif 00053 00054 #include <stddef.h> 00055 typedef wchar_t XML_Char; 00056 typedef wchar_t XML_LChar; 00057 00058 #else /* not XML_UNICODE_WCHAR_T */ 00059 00060 #ifdef XML_UNICODE 00061 00062 /* Information is UTF-16 encoded as unsigned shorts */ 00063 typedef unsigned short XML_Char; 00064 typedef char XML_LChar; 00065 00066 #else /* not XML_UNICODE */ 00067 00068 /* Information is UTF-8 encoded. */ 00069 typedef char XML_Char; 00070 typedef char XML_LChar; 00071 00072 #endif /* not XML_UNICODE */ 00073 00074 #endif /* not XML_UNICODE_WCHAR_T */ 00075 00076 00077 /* Constructs a new parser; encoding is the encoding specified by the external 00078 protocol or null if there is none specified. */ 00079 00080 XML_Parser XMLPARSEAPI 00081 XML_ParserCreate(const XML_Char *encoding); 00082 00083 /* Constructs a new parser and namespace processor. Element type names 00084 and attribute names that belong to a namespace will be expanded; 00085 unprefixed attribute names are never expanded; unprefixed element type 00086 names are expanded only if there is a default namespace. The expanded 00087 name is the concatenation of the namespace URI, the namespace separator character, 00088 and the local part of the name. If the namespace separator is '\0' then 00089 the namespace URI and the local part will be concatenated without any 00090 separator. When a namespace is not declared, the name and prefix will be 00091 passed through without expansion. */ 00092 00093 XML_Parser XMLPARSEAPI 00094 XML_ParserCreateNS(const XML_Char *encoding, XML_Char namespaceSeparator); 00095 00096 00097 /* atts is array of name/value pairs, terminated by 0; 00098 names and values are 0 terminated. */ 00099 00100 typedef void (*XML_StartElementHandler)(void *userData, 00101 const XML_Char *name, 00102 const XML_Char **atts); 00103 00104 typedef void (*XML_EndElementHandler)(void *userData, 00105 const XML_Char *name); 00106 00107 /* s is not 0 terminated. */ 00108 typedef void (*XML_CharacterDataHandler)(void *userData, 00109 const XML_Char *s, 00110 int len); 00111 00112 /* target and data are 0 terminated */ 00113 typedef void (*XML_ProcessingInstructionHandler)(void *userData, 00114 const XML_Char *target, 00115 const XML_Char *data); 00116 00117 /* data is 0 terminated */ 00118 typedef void (*XML_CommentHandler)(void *userData, const XML_Char *data); 00119 00120 typedef void (*XML_StartCdataSectionHandler)(void *userData); 00121 typedef void (*XML_EndCdataSectionHandler)(void *userData); 00122 00123 /* This is called for any characters in the XML document for 00124 which there is no applicable handler. This includes both 00125 characters that are part of markup which is of a kind that is 00126 not reported (comments, markup declarations), or characters 00127 that are part of a construct which could be reported but 00128 for which no handler has been supplied. The characters are passed 00129 exactly as they were in the XML document except that 00130 they will be encoded in UTF-8. Line boundaries are not normalized. 00131 Note that a byte order mark character is not passed to the default handler. 00132 There are no guarantees about how characters are divided between calls 00133 to the default handler: for example, a comment might be split between 00134 multiple calls. */ 00135 00136 typedef void (*XML_DefaultHandler)(void *userData, 00137 const XML_Char *s, 00138 int len); 00139 00140 /* This is called for a declaration of an unparsed (NDATA) 00141 entity. The base argument is whatever was set by XML_SetBase. 00142 The entityName, systemId and notationName arguments will never be null. 00143 The other arguments may be. */ 00144 00145 typedef void (*XML_UnparsedEntityDeclHandler)(void *userData, 00146 const XML_Char *entityName, 00147 const XML_Char *base, 00148 const XML_Char *systemId, 00149 const XML_Char *publicId, 00150 const XML_Char *notationName); 00151 00152 /* This is called for a declaration of notation. 00153 The base argument is whatever was set by XML_SetBase. 00154 The notationName will never be null. The other arguments can be. */ 00155 00156 typedef void (*XML_NotationDeclHandler)(void *userData, 00157 const XML_Char *notationName, 00158 const XML_Char *base, 00159 const XML_Char *systemId, 00160 const XML_Char *publicId); 00161 00162 /* When namespace processing is enabled, these are called once for 00163 each namespace declaration. The call to the start and end element 00164 handlers occur between the calls to the start and end namespace 00165 declaration handlers. For an xmlns attribute, prefix will be null. 00166 For an xmlns="" attribute, uri will be null. */ 00167 00168 typedef void (*XML_StartNamespaceDeclHandler)(void *userData, 00169 const XML_Char *prefix, 00170 const XML_Char *uri); 00171 00172 typedef void (*XML_EndNamespaceDeclHandler)(void *userData, 00173 const XML_Char *prefix); 00174 00175 /* This is called if the document is not standalone (it has an 00176 external subset or a reference to a parameter entity, but does not 00177 have standalone="yes"). If this handler returns 0, then processing 00178 will not continue, and the parser will return a 00179 XML_ERROR_NOT_STANDALONE error. */ 00180 00181 typedef int (*XML_NotStandaloneHandler)(void *userData); 00182 00183 /* This is called for a reference to an external parsed general entity. 00184 The referenced entity is not automatically parsed. 00185 The application can parse it immediately or later using 00186 XML_ExternalEntityParserCreate. 00187 The parser argument is the parser parsing the entity containing the reference; 00188 it can be passed as the parser argument to XML_ExternalEntityParserCreate. 00189 The systemId argument is the system identifier as specified in the entity declaration; 00190 it will not be null. 00191 The base argument is the system identifier that should be used as the base for 00192 resolving systemId if systemId was relative; this is set by XML_SetBase; 00193 it may be null. 00194 The publicId argument is the public identifier as specified in the entity declaration, 00195 or null if none was specified; the whitespace in the public identifier 00196 will have been normalized as required by the XML spec. 00197 The context argument specifies the parsing context in the format 00198 expected by the context argument to 00199 XML_ExternalEntityParserCreate; context is valid only until the handler 00200 returns, so if the referenced entity is to be parsed later, it must be copied. 00201 The handler should return 0 if processing should not continue because of 00202 a fatal error in the handling of the external entity. 00203 In this case the calling parser will return an XML_ERROR_EXTERNAL_ENTITY_HANDLING 00204 error. 00205 Note that unlike other handlers the first argument is the parser, not userData. */ 00206 00207 typedef int (*XML_ExternalEntityRefHandler)(XML_Parser parser, 00208 const XML_Char *context, 00209 const XML_Char *base, 00210 const XML_Char *systemId, 00211 const XML_Char *publicId); 00212 00213 /* This structure is filled in by the XML_UnknownEncodingHandler 00214 to provide information to the parser about encodings that are unknown 00215 to the parser. 00216 The map[b] member gives information about byte sequences 00217 whose first byte is b. 00218 If map[b] is c where c is >= 0, then b by itself encodes the Unicode scalar value c. 00219 If map[b] is -1, then the byte sequence is malformed. 00220 If map[b] is -n, where n >= 2, then b is the first byte of an n-byte 00221 sequence that encodes a single Unicode scalar value. 00222 The data member will be passed as the first argument to the convert function. 00223 The convert function is used to convert multibyte sequences; 00224 s will point to a n-byte sequence where map[(unsigned char)*s] == -n. 00225 The convert function must return the Unicode scalar value 00226 represented by this byte sequence or -1 if the byte sequence is malformed. 00227 The convert function may be null if the encoding is a single-byte encoding, 00228 that is if map[b] >= -1 for all bytes b. 00229 When the parser is finished with the encoding, then if release is not null, 00230 it will call release passing it the data member; 00231 once release has been called, the convert function will not be called again. 00232 00233 Expat places certain restrictions on the encodings that are supported 00234 using this mechanism. 00235 00236 1. Every ASCII character that can appear in a well-formed XML document, 00237 other than the characters 00238 00239 $@\^`{}~ 00240 00241 must be represented by a single byte, and that byte must be the 00242 same byte that represents that character in ASCII. 00243 00244 2. No character may require more than 4 bytes to encode. 00245 00246 3. All characters encoded must have Unicode scalar values <= 0xFFFF, 00247 (ie characters that would be encoded by surrogates in UTF-16 00248 are not allowed). Note that this restriction doesn't apply to 00249 the built-in support for UTF-8 and UTF-16. 00250 00251 4. No Unicode character may be encoded by more than one distinct sequence 00252 of bytes. */ 00253 00254 typedef struct { 00255 int map[256]; 00256 void *data; 00257 int (*convert)(void *data, const char *s); 00258 void (*release)(void *data); 00259 } XML_Encoding; 00260 00261 /* This is called for an encoding that is unknown to the parser. 00262 The encodingHandlerData argument is that which was passed as the 00263 second argument to XML_SetUnknownEncodingHandler. 00264 The name argument gives the name of the encoding as specified in 00265 the encoding declaration. 00266 If the callback can provide information about the encoding, 00267 it must fill in the XML_Encoding structure, and return 1. 00268 Otherwise it must return 0. 00269 If info does not describe a suitable encoding, 00270 then the parser will return an XML_UNKNOWN_ENCODING error. */ 00271 00272 typedef int (*XML_UnknownEncodingHandler)(void *encodingHandlerData, 00273 const XML_Char *name, 00274 XML_Encoding *info); 00275 00276 void XMLPARSEAPI 00277 XML_SetElementHandler(XML_Parser parser, 00278 XML_StartElementHandler start, 00279 XML_EndElementHandler end); 00280 00281 void XMLPARSEAPI 00282 XML_SetCharacterDataHandler(XML_Parser parser, 00283 XML_CharacterDataHandler handler); 00284 00285 void XMLPARSEAPI 00286 XML_SetProcessingInstructionHandler(XML_Parser parser, 00287 XML_ProcessingInstructionHandler handler); 00288 void XMLPARSEAPI 00289 XML_SetCommentHandler(XML_Parser parser, 00290 XML_CommentHandler handler); 00291 00292 void XMLPARSEAPI 00293 XML_SetCdataSectionHandler(XML_Parser parser, 00294 XML_StartCdataSectionHandler start, 00295 XML_EndCdataSectionHandler end); 00296 00297 /* This sets the default handler and also inhibits expansion of internal entities. 00298 The entity reference will be passed to the default handler. */ 00299 00300 void XMLPARSEAPI 00301 XML_SetDefaultHandler(XML_Parser parser, 00302 XML_DefaultHandler handler); 00303 00304 /* This sets the default handler but does not inhibit expansion of internal entities. 00305 The entity reference will not be passed to the default handler. */ 00306 00307 void XMLPARSEAPI 00308 XML_SetDefaultHandlerExpand(XML_Parser parser, 00309 XML_DefaultHandler handler); 00310 00311 void XMLPARSEAPI 00312 XML_SetUnparsedEntityDeclHandler(XML_Parser parser, 00313 XML_UnparsedEntityDeclHandler handler); 00314 00315 void XMLPARSEAPI 00316 XML_SetNotationDeclHandler(XML_Parser parser, 00317 XML_NotationDeclHandler handler); 00318 00319 void XMLPARSEAPI 00320 XML_SetNamespaceDeclHandler(XML_Parser parser, 00321 XML_StartNamespaceDeclHandler start, 00322 XML_EndNamespaceDeclHandler end); 00323 00324 void XMLPARSEAPI 00325 XML_SetNotStandaloneHandler(XML_Parser parser, 00326 XML_NotStandaloneHandler handler); 00327 00328 void XMLPARSEAPI 00329 XML_SetExternalEntityRefHandler(XML_Parser parser, 00330 XML_ExternalEntityRefHandler handler); 00331 00332 /* If a non-null value for arg is specified here, then it will be passed 00333 as the first argument to the external entity ref handler instead 00334 of the parser object. */ 00335 void XMLPARSEAPI 00336 XML_SetExternalEntityRefHandlerArg(XML_Parser, void *arg); 00337 00338 void XMLPARSEAPI 00339 XML_SetUnknownEncodingHandler(XML_Parser parser, 00340 XML_UnknownEncodingHandler handler, 00341 void *encodingHandlerData); 00342 00343 /* This can be called within a handler for a start element, end element, 00344 processing instruction or character data. It causes the corresponding 00345 markup to be passed to the default handler. */ 00346 void XMLPARSEAPI XML_DefaultCurrent(XML_Parser parser); 00347 00348 /* This value is passed as the userData argument to callbacks. */ 00349 void XMLPARSEAPI 00350 XML_SetUserData(XML_Parser parser, void *userData); 00351 00352 /* Returns the last value set by XML_SetUserData or null. */ 00353 #define XML_GetUserData(parser) (*(void **)(parser)) 00354 00355 /* This is equivalent to supplying an encoding argument 00356 to XML_CreateParser. It must not be called after XML_Parse 00357 or XML_ParseBuffer. */ 00358 00359 int XMLPARSEAPI 00360 XML_SetEncoding(XML_Parser parser, const XML_Char *encoding); 00361 00362 /* If this function is called, then the parser will be passed 00363 as the first argument to callbacks instead of userData. 00364 The userData will still be accessible using XML_GetUserData. */ 00365 00366 void XMLPARSEAPI 00367 XML_UseParserAsHandlerArg(XML_Parser parser); 00368 00369 /* Sets the base to be used for resolving relative URIs in system identifiers in 00370 declarations. Resolving relative identifiers is left to the application: 00371 this value will be passed through as the base argument to the 00372 XML_ExternalEntityRefHandler, XML_NotationDeclHandler 00373 and XML_UnparsedEntityDeclHandler. The base argument will be copied. 00374 Returns zero if out of memory, non-zero otherwise. */ 00375 00376 int XMLPARSEAPI 00377 XML_SetBase(XML_Parser parser, const XML_Char *base); 00378 00379 const XML_Char XMLPARSEAPI * 00380 XML_GetBase(XML_Parser parser); 00381 00382 /* Returns the number of the attributes passed in last call to the 00383 XML_StartElementHandler that were specified in the start-tag rather 00384 than defaulted. */ 00385 00386 int XMLPARSEAPI XML_GetSpecifiedAttributeCount(XML_Parser parser); 00387 00388 /* Parses some input. Returns 0 if a fatal error is detected. 00389 The last call to XML_Parse must have isFinal true; 00390 len may be zero for this call (or any other). */ 00391 int XMLPARSEAPI 00392 XML_Parse(XML_Parser parser, const char *s, int len, int isFinal); 00393 00394 void XMLPARSEAPI * 00395 XML_GetBuffer(XML_Parser parser, int len); 00396 00397 int XMLPARSEAPI 00398 XML_ParseBuffer(XML_Parser parser, int len, int isFinal); 00399 00400 /* Creates an XML_Parser object that can parse an external general entity; 00401 context is a '\0'-terminated string specifying the parse context; 00402 encoding is a '\0'-terminated string giving the name of the externally specified encoding, 00403 or null if there is no externally specified encoding. 00404 The context string consists of a sequence of tokens separated by formfeeds (\f); 00405 a token consisting of a name specifies that the general entity of the name 00406 is open; a token of the form prefix=uri specifies the namespace for a particular 00407 prefix; a token of the form =uri specifies the default namespace. 00408 This can be called at any point after the first call to an ExternalEntityRefHandler 00409 so longer as the parser has not yet been freed. 00410 The new parser is completely independent and may safely be used in a separate thread. 00411 The handlers and userData are initialized from the parser argument. 00412 Returns 0 if out of memory. Otherwise returns a new XML_Parser object. */ 00413 XML_Parser XMLPARSEAPI 00414 XML_ExternalEntityParserCreate(XML_Parser parser, 00415 const XML_Char *context, 00416 const XML_Char *encoding); 00417 00418 enum XML_Error { 00419 XML_ERROR_NONE, 00420 XML_ERROR_NO_MEMORY, 00421 XML_ERROR_SYNTAX, 00422 XML_ERROR_NO_ELEMENTS, 00423 XML_ERROR_INVALID_TOKEN, 00424 XML_ERROR_UNCLOSED_TOKEN, 00425 XML_ERROR_PARTIAL_CHAR, 00426 XML_ERROR_TAG_MISMATCH, 00427 XML_ERROR_DUPLICATE_ATTRIBUTE, 00428 XML_ERROR_JUNK_AFTER_DOC_ELEMENT, 00429 XML_ERROR_PARAM_ENTITY_REF, 00430 XML_ERROR_UNDEFINED_ENTITY, 00431 XML_ERROR_RECURSIVE_ENTITY_REF, 00432 XML_ERROR_ASYNC_ENTITY, 00433 XML_ERROR_BAD_CHAR_REF, 00434 XML_ERROR_BINARY_ENTITY_REF, 00435 XML_ERROR_ATTRIBUTE_EXTERNAL_ENTITY_REF, 00436 XML_ERROR_MISPLACED_XML_PI, 00437 XML_ERROR_UNKNOWN_ENCODING, 00438 XML_ERROR_INCORRECT_ENCODING, 00439 XML_ERROR_UNCLOSED_CDATA_SECTION, 00440 XML_ERROR_EXTERNAL_ENTITY_HANDLING, 00441 XML_ERROR_NOT_STANDALONE 00442 }; 00443 00444 /* If XML_Parse or XML_ParseBuffer have returned 0, then XML_GetErrorCode 00445 returns information about the error. */ 00446 00447 enum XML_Error XMLPARSEAPI XML_GetErrorCode(XML_Parser parser); 00448 00449 /* These functions return information about the current parse location. 00450 They may be called when XML_Parse or XML_ParseBuffer return 0; 00451 in this case the location is the location of the character at which 00452 the error was detected. 00453 They may also be called from any other callback called to report 00454 some parse event; in this the location is the location of the first 00455 of the sequence of characters that generated the event. */ 00456 00457 int XMLPARSEAPI XML_GetCurrentLineNumber(XML_Parser parser); 00458 int XMLPARSEAPI XML_GetCurrentColumnNumber(XML_Parser parser); 00459 long XMLPARSEAPI XML_GetCurrentByteIndex(XML_Parser parser); 00460 00461 /* Return the number of bytes in the current event. 00462 Returns 0 if the event is in an internal entity. */ 00463 00464 int XMLPARSEAPI XML_GetCurrentByteCount(XML_Parser parser); 00465 00466 /* For backwards compatibility with previous versions. */ 00467 #define XML_GetErrorLineNumber XML_GetCurrentLineNumber 00468 #define XML_GetErrorColumnNumber XML_GetCurrentColumnNumber 00469 #define XML_GetErrorByteIndex XML_GetCurrentByteIndex 00470 00471 /* Frees memory used by the parser. */ 00472 void XMLPARSEAPI 00473 XML_ParserFree(XML_Parser parser); 00474 00475 /* Returns a string describing the error. */ 00476 const XML_LChar XMLPARSEAPI *XML_ErrorString(int code); 00477 00478 #ifdef __cplusplus 00479 } 00480 #endif 00481 00482 #endif /* not XmlParse_INCLUDED */
1.6.1