/* * parser.c : an XML 1.0 parser, namespaces and validity support are mostly * implemented on top of the SAX interfaces * * References: * The XML specification: * http://www.w3.org/TR/REC-xml * Original 1.0 version: * http://www.w3.org/TR/1998/REC-xml-19980210 * XML second edition working draft * http://www.w3.org/TR/2000/WD-xml-2e-20000814 * * Okay this is a big file, the parser core is around 7000 lines, then it * is followed by the progressive parser top routines, then the various * high level APIs to call the parser and a few miscellaneous functions. * A number of helper functions and deprecated ones have been moved to * parserInternals.c to reduce this file size. * As much as possible the functions are associated with their relative * production in the XML specification. A few productions defining the * different ranges of character are actually implanted either in * parserInternals.h or parserInternals.c * The DOM tree build is realized from the default SAX callbacks in * the module SAX2.c. * The routines doing the validation checks are in valid.c and called either * from the SAX callbacks or as standalone functions using a preparsed * document. * * See Copyright for the status of this software. * * Author: Daniel Veillard */ /* To avoid EBCDIC trouble when parsing on zOS */ #if defined(__MVS__) #pragma convert("ISO8859-1") #endif #define IN_LIBXML #include "libxml.h" #if defined(_WIN32) #define XML_DIR_SEP '\\' #else #define XML_DIR_SEP '/' #endif #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef LIBXML_CATALOG_ENABLED #include #endif #include "private/buf.h" #include "private/dict.h" #include "private/entities.h" #include "private/error.h" #include "private/html.h" #include "private/io.h" #include "private/memory.h" #include "private/parser.h" #include "private/tree.h" #define NS_INDEX_EMPTY INT_MAX #define NS_INDEX_XML (INT_MAX - 1) #define URI_HASH_EMPTY 0xD943A04E #define URI_HASH_XML 0xF0451F02 #ifndef STDIN_FILENO #define STDIN_FILENO 0 #endif #ifndef SIZE_MAX #define SIZE_MAX ((size_t) -1) #endif #define XML_MAX_ATTRS 100000000 /* 100 million */ #define XML_SPECIAL_EXTERNAL (1 << 20) #define XML_SPECIAL_TYPE_MASK (XML_SPECIAL_EXTERNAL - 1) #define XML_ATTVAL_ALLOC (1 << 0) #define XML_ATTVAL_NORM_CHANGE (1 << 1) struct _xmlStartTag { const xmlChar *prefix; const xmlChar *URI; int line; int nsNr; }; typedef struct { void *saxData; unsigned prefixHashValue; unsigned uriHashValue; unsigned elementId; int oldIndex; } xmlParserNsExtra; typedef struct { unsigned hashValue; int index; } xmlParserNsBucket; struct _xmlParserNsData { xmlParserNsExtra *extra; unsigned hashSize; unsigned hashElems; xmlParserNsBucket *hash; unsigned elementId; int defaultNsIndex; int minNsIndex; }; static int xmlParseElementStart(xmlParserCtxtPtr ctxt); static void xmlParseElementEnd(xmlParserCtxtPtr ctxt); static xmlEntityPtr xmlLookupGeneralEntity(xmlParserCtxtPtr ctxt, const xmlChar *name, int inAttr); static const xmlChar * xmlParseEntityRefInternal(xmlParserCtxtPtr ctxt); /************************************************************************ * * * Arbitrary limits set in the parser. See XML_PARSE_HUGE * * * ************************************************************************/ #define XML_PARSER_BIG_ENTITY 1000 #define XML_PARSER_LOT_ENTITY 5000 /* * Constants for protection against abusive entity expansion * ("billion laughs"). */ /* * A certain amount of entity expansion which is always allowed. */ #define XML_PARSER_ALLOWED_EXPANSION 1000000 /* * Fixed cost for each entity reference. This crudely models processing time * as well to protect, for example, against exponential expansion of empty * or very short entities. */ #define XML_ENT_FIXED_COST 20 #define XML_PARSER_BIG_BUFFER_SIZE 300 #define XML_PARSER_BUFFER_SIZE 100 #define SAX_COMPAT_MODE BAD_CAST "SAX compatibility mode document" /** * XML_PARSER_CHUNK_SIZE * * When calling GROW that's the minimal amount of data * the parser expected to have received. It is not a hard * limit but an optimization when reading strings like Names * It is not strictly needed as long as inputs available characters * are followed by 0, which should be provided by the I/O level */ #define XML_PARSER_CHUNK_SIZE 100 /** * Constant string describing the version of the library used at * run-time. */ const char *const xmlParserVersion = LIBXML_VERSION_STRING LIBXML_VERSION_EXTRA; /* * List of XML prefixed PI allowed by W3C specs */ static const char* const xmlW3CPIs[] = { "xml-stylesheet", "xml-model", NULL }; /* DEPR void xmlParserHandleReference(xmlParserCtxtPtr ctxt); */ static xmlEntityPtr xmlParseStringPEReference(xmlParserCtxtPtr ctxt, const xmlChar **str); static void xmlCtxtParseEntity(xmlParserCtxtPtr ctxt, xmlEntityPtr ent); static int xmlLoadEntityContent(xmlParserCtxtPtr ctxt, xmlEntityPtr entity); static void xmlParsePERefInternal(xmlParserCtxt *ctxt, int markupDecl); /************************************************************************ * * * Some factorized error routines * * * ************************************************************************/ static void xmlErrMemory(xmlParserCtxtPtr ctxt) { xmlCtxtErrMemory(ctxt); } /** * Handle a redefinition of attribute error * * @param ctxt an XML parser context * @param prefix the attribute prefix * @param localname the attribute localname */ static void xmlErrAttributeDup(xmlParserCtxtPtr ctxt, const xmlChar * prefix, const xmlChar * localname) { if (prefix == NULL) xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, XML_ERR_ATTRIBUTE_REDEFINED, XML_ERR_FATAL, localname, NULL, NULL, 0, "Attribute %s redefined\n", localname); else xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, XML_ERR_ATTRIBUTE_REDEFINED, XML_ERR_FATAL, prefix, localname, NULL, 0, "Attribute %s:%s redefined\n", prefix, localname); } /** * Handle a fatal parser error, i.e. violating Well-Formedness constraints * * @param ctxt an XML parser context * @param error the error number * @param msg the error message */ static void LIBXML_ATTR_FORMAT(3,0) xmlFatalErrMsg(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg) { xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, error, XML_ERR_FATAL, NULL, NULL, NULL, 0, "%s", msg); } /** * Handle a warning. * * @param ctxt an XML parser context * @param error the error number * @param msg the error message * @param str1 extra data * @param str2 extra data */ void LIBXML_ATTR_FORMAT(3,0) xmlWarningMsg(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar *str1, const xmlChar *str2) { xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, error, XML_ERR_WARNING, str1, str2, NULL, 0, msg, str1, str2); } #ifdef LIBXML_VALID_ENABLED /** * Handle a validity error. * * @param ctxt an XML parser context * @param error the error number * @param msg the error message * @param str1 extra data * @param str2 extra data */ static void LIBXML_ATTR_FORMAT(3,0) xmlValidityError(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar *str1, const xmlChar *str2) { ctxt->valid = 0; xmlCtxtErr(ctxt, NULL, XML_FROM_DTD, error, XML_ERR_ERROR, str1, str2, NULL, 0, msg, str1, str2); } #endif /** * Handle a fatal parser error, i.e. violating Well-Formedness constraints * * @param ctxt an XML parser context * @param error the error number * @param msg the error message * @param val an integer value */ static void LIBXML_ATTR_FORMAT(3,0) xmlFatalErrMsgInt(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, int val) { xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, error, XML_ERR_FATAL, NULL, NULL, NULL, val, msg, val); } /** * Handle a fatal parser error, i.e. violating Well-Formedness constraints * * @param ctxt an XML parser context * @param error the error number * @param msg the error message * @param str1 an string info * @param val an integer value * @param str2 an string info */ static void LIBXML_ATTR_FORMAT(3,0) xmlFatalErrMsgStrIntStr(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar *str1, int val, const xmlChar *str2) { xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, error, XML_ERR_FATAL, str1, str2, NULL, val, msg, str1, val, str2); } /** * Handle a fatal parser error, i.e. violating Well-Formedness constraints * * @param ctxt an XML parser context * @param error the error number * @param msg the error message * @param val a string value */ static void LIBXML_ATTR_FORMAT(3,0) xmlFatalErrMsgStr(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar * val) { xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, error, XML_ERR_FATAL, val, NULL, NULL, 0, msg, val); } /** * Handle a non fatal parser error * * @param ctxt an XML parser context * @param error the error number * @param msg the error message * @param val a string value */ static void LIBXML_ATTR_FORMAT(3,0) xmlErrMsgStr(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar * val) { xmlCtxtErr(ctxt, NULL, XML_FROM_PARSER, error, XML_ERR_ERROR, val, NULL, NULL, 0, msg, val); } /** * Handle a fatal parser error, i.e. violating Well-Formedness constraints * * @param ctxt an XML parser context * @param error the error number * @param msg the message * @param info1 extra information string * @param info2 extra information string * @param info3 extra information string */ static void LIBXML_ATTR_FORMAT(3,0) xmlNsErr(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar * info1, const xmlChar * info2, const xmlChar * info3) { ctxt->nsWellFormed = 0; xmlCtxtErr(ctxt, NULL, XML_FROM_NAMESPACE, error, XML_ERR_ERROR, info1, info2, info3, 0, msg, info1, info2, info3); } /** * Handle a namespace warning error * * @param ctxt an XML parser context * @param error the error number * @param msg the message * @param info1 extra information string * @param info2 extra information string * @param info3 extra information string */ static void LIBXML_ATTR_FORMAT(3,0) xmlNsWarn(xmlParserCtxtPtr ctxt, xmlParserErrors error, const char *msg, const xmlChar * info1, const xmlChar * info2, const xmlChar * info3) { xmlCtxtErr(ctxt, NULL, XML_FROM_NAMESPACE, error, XML_ERR_WARNING, info1, info2, info3, 0, msg, info1, info2, info3); } /** * Check for non-linear entity expansion behaviour. * * In some cases like xmlExpandEntityInAttValue, this function is called * for each, possibly nested entity and its unexpanded content length. * * In other cases like #xmlParseReference, it's only called for each * top-level entity with its unexpanded content length plus the sum of * the unexpanded content lengths (plus fixed cost) of all nested * entities. * * Summing the unexpanded lengths also adds the length of the reference. * This is by design. Taking the length of the entity name into account * discourages attacks that try to waste CPU time with abusively long * entity names. See test/recurse/lol6.xml for example. Each call also * adds some fixed cost XML_ENT_FIXED_COST to discourage attacks with * short entities. * * @param ctxt parser context * @param extra sum of unexpanded entity sizes * @returns 1 on error, 0 on success. */ static int xmlParserEntityCheck(xmlParserCtxtPtr ctxt, unsigned long extra) { unsigned long consumed; unsigned long *expandedSize; xmlParserInputPtr input = ctxt->input; xmlEntityPtr entity = input->entity; if ((entity) && (entity->flags & XML_ENT_CHECKED)) return(0); /* * Compute total consumed bytes so far, including input streams of * external entities. */ consumed = input->consumed; xmlSaturatedAddSizeT(&consumed, input->cur - input->base); xmlSaturatedAdd(&consumed, ctxt->sizeentities); if (entity) expandedSize = &entity->expandedSize; else expandedSize = &ctxt->sizeentcopy; /* * Add extra cost and some fixed cost. */ xmlSaturatedAdd(expandedSize, extra); xmlSaturatedAdd(expandedSize, XML_ENT_FIXED_COST); /* * It's important to always use saturation arithmetic when tracking * entity sizes to make the size checks reliable. If "sizeentcopy" * overflows, we have to abort. */ if ((*expandedSize > XML_PARSER_ALLOWED_EXPANSION) && ((*expandedSize >= ULONG_MAX) || (*expandedSize / ctxt->maxAmpl > consumed))) { xmlFatalErrMsg(ctxt, XML_ERR_RESOURCE_LIMIT, "Maximum entity amplification factor exceeded, see " "xmlCtxtSetMaxAmplification.\n"); return(1); } return(0); } /************************************************************************ * * * Library wide options * * * ************************************************************************/ /** * Examines if the library has been compiled with a given feature. * * @param feature the feature to be examined * @returns zero (0) if the feature does not exist or an unknown * feature is requested, non-zero otherwise. */ int xmlHasFeature(xmlFeature feature) { switch (feature) { case XML_WITH_THREAD: #ifdef LIBXML_THREAD_ENABLED return(1); #else return(0); #endif case XML_WITH_TREE: return(1); case XML_WITH_OUTPUT: #ifdef LIBXML_OUTPUT_ENABLED return(1); #else return(0); #endif case XML_WITH_PUSH: #ifdef LIBXML_PUSH_ENABLED return(1); #else return(0); #endif case XML_WITH_READER: #ifdef LIBXML_READER_ENABLED return(1); #else return(0); #endif case XML_WITH_PATTERN: #ifdef LIBXML_PATTERN_ENABLED return(1); #else return(0); #endif case XML_WITH_WRITER: #ifdef LIBXML_WRITER_ENABLED return(1); #else return(0); #endif case XML_WITH_SAX1: #ifdef LIBXML_SAX1_ENABLED return(1); #else return(0); #endif case XML_WITH_HTTP: return(0); case XML_WITH_VALID: #ifdef LIBXML_VALID_ENABLED return(1); #else return(0); #endif case XML_WITH_HTML: #ifdef LIBXML_HTML_ENABLED return(1); #else return(0); #endif case XML_WITH_LEGACY: return(0); case XML_WITH_C14N: #ifdef LIBXML_C14N_ENABLED return(1); #else return(0); #endif case XML_WITH_CATALOG: #ifdef LIBXML_CATALOG_ENABLED return(1); #else return(0); #endif case XML_WITH_XPATH: #ifdef LIBXML_XPATH_ENABLED return(1); #else return(0); #endif case XML_WITH_XPTR: #ifdef LIBXML_XPTR_ENABLED return(1); #else return(0); #endif case XML_WITH_XINCLUDE: #ifdef LIBXML_XINCLUDE_ENABLED return(1); #else return(0); #endif case XML_WITH_ICONV: #ifdef LIBXML_ICONV_ENABLED return(1); #else return(0); #endif case XML_WITH_ISO8859X: #ifdef LIBXML_ISO8859X_ENABLED return(1); #else return(0); #endif case XML_WITH_UNICODE: return(0); case XML_WITH_REGEXP: #ifdef LIBXML_REGEXP_ENABLED return(1); #else return(0); #endif case XML_WITH_AUTOMATA: #ifdef LIBXML_REGEXP_ENABLED return(1); #else return(0); #endif case XML_WITH_EXPR: return(0); case XML_WITH_RELAXNG: #ifdef LIBXML_RELAXNG_ENABLED return(1); #else return(0); #endif case XML_WITH_SCHEMAS: #ifdef LIBXML_SCHEMAS_ENABLED return(1); #else return(0); #endif case XML_WITH_SCHEMATRON: #ifdef LIBXML_SCHEMATRON_ENABLED return(1); #else return(0); #endif case XML_WITH_MODULES: #ifdef LIBXML_MODULES_ENABLED return(1); #else return(0); #endif case XML_WITH_DEBUG: #ifdef LIBXML_DEBUG_ENABLED return(1); #else return(0); #endif case XML_WITH_DEBUG_MEM: return(0); case XML_WITH_ZLIB: #ifdef LIBXML_ZLIB_ENABLED return(1); #else return(0); #endif case XML_WITH_LZMA: return(0); case XML_WITH_ICU: #ifdef LIBXML_ICU_ENABLED return(1); #else return(0); #endif default: break; } return(0); } /************************************************************************ * * * Simple string buffer * * * ************************************************************************/ typedef struct { xmlChar *mem; unsigned size; unsigned cap; /* size < cap */ unsigned max; /* size <= max */ xmlParserErrors code; } xmlSBuf; static void xmlSBufInit(xmlSBuf *buf, unsigned max) { buf->mem = NULL; buf->size = 0; buf->cap = 0; buf->max = max; buf->code = XML_ERR_OK; } static int xmlSBufGrow(xmlSBuf *buf, unsigned len) { xmlChar *mem; unsigned cap; if (len >= UINT_MAX / 2 - buf->size) { if (buf->code == XML_ERR_OK) buf->code = XML_ERR_RESOURCE_LIMIT; return(-1); } cap = (buf->size + len) * 2; if (cap < 240) cap = 240; mem = xmlRealloc(buf->mem, cap); if (mem == NULL) { buf->code = XML_ERR_NO_MEMORY; return(-1); } buf->mem = mem; buf->cap = cap; return(0); } static void xmlSBufAddString(xmlSBuf *buf, const xmlChar *str, unsigned len) { if (buf->max - buf->size < len) { if (buf->code == XML_ERR_OK) buf->code = XML_ERR_RESOURCE_LIMIT; return; } if (buf->cap - buf->size <= len) { if (xmlSBufGrow(buf, len) < 0) return; } if (len > 0) memcpy(buf->mem + buf->size, str, len); buf->size += len; } static void xmlSBufAddCString(xmlSBuf *buf, const char *str, unsigned len) { xmlSBufAddString(buf, (const xmlChar *) str, len); } static void xmlSBufAddChar(xmlSBuf *buf, int c) { xmlChar *end; if (buf->max - buf->size < 4) { if (buf->code == XML_ERR_OK) buf->code = XML_ERR_RESOURCE_LIMIT; return; } if (buf->cap - buf->size <= 4) { if (xmlSBufGrow(buf, 4) < 0) return; } end = buf->mem + buf->size; if (c < 0x80) { *end = (xmlChar) c; buf->size += 1; } else { buf->size += xmlCopyCharMultiByte(end, c); } } static void xmlSBufAddReplChar(xmlSBuf *buf) { xmlSBufAddCString(buf, "\xEF\xBF\xBD", 3); } static void xmlSBufReportError(xmlSBuf *buf, xmlParserCtxtPtr ctxt, const char *errMsg) { if (buf->code == XML_ERR_NO_MEMORY) xmlCtxtErrMemory(ctxt); else xmlFatalErr(ctxt, buf->code, errMsg); } static xmlChar * xmlSBufFinish(xmlSBuf *buf, int *sizeOut, xmlParserCtxtPtr ctxt, const char *errMsg) { if (buf->mem == NULL) { buf->mem = xmlMalloc(1); if (buf->mem == NULL) { buf->code = XML_ERR_NO_MEMORY; } else { buf->mem[0] = 0; } } else { buf->mem[buf->size] = 0; } if (buf->code == XML_ERR_OK) { if (sizeOut != NULL) *sizeOut = buf->size; return(buf->mem); } xmlSBufReportError(buf, ctxt, errMsg); xmlFree(buf->mem); if (sizeOut != NULL) *sizeOut = 0; return(NULL); } static void xmlSBufCleanup(xmlSBuf *buf, xmlParserCtxtPtr ctxt, const char *errMsg) { if (buf->code != XML_ERR_OK) xmlSBufReportError(buf, ctxt, errMsg); xmlFree(buf->mem); } static int xmlUTF8MultibyteLen(xmlParserCtxtPtr ctxt, const xmlChar *str, const char *errMsg) { int c = str[0]; int c1 = str[1]; if ((c1 & 0xC0) != 0x80) goto encoding_error; if (c < 0xE0) { /* 2-byte sequence */ if (c < 0xC2) goto encoding_error; return(2); } else { int c2 = str[2]; if ((c2 & 0xC0) != 0x80) goto encoding_error; if (c < 0xF0) { /* 3-byte sequence */ if (c == 0xE0) { /* overlong */ if (c1 < 0xA0) goto encoding_error; } else if (c == 0xED) { /* surrogate */ if (c1 >= 0xA0) goto encoding_error; } else if (c == 0xEF) { /* U+FFFE and U+FFFF are invalid Chars */ if ((c1 == 0xBF) && (c2 >= 0xBE)) xmlFatalErrMsg(ctxt, XML_ERR_INVALID_CHAR, errMsg); } return(3); } else { /* 4-byte sequence */ if ((str[3] & 0xC0) != 0x80) goto encoding_error; if (c == 0xF0) { /* overlong */ if (c1 < 0x90) goto encoding_error; } else if (c >= 0xF4) { /* greater than 0x10FFFF */ if ((c > 0xF4) || (c1 >= 0x90)) goto encoding_error; } return(4); } } encoding_error: /* Only report the first error */ if ((ctxt->input->flags & XML_INPUT_ENCODING_ERROR) == 0) { xmlCtxtErrIO(ctxt, XML_ERR_INVALID_ENCODING, NULL); ctxt->input->flags |= XML_INPUT_ENCODING_ERROR; } return(0); } /************************************************************************ * * * SAX2 defaulted attributes handling * * * ************************************************************************/ /** * Final initialization of the parser context before starting to parse. * * This accounts for users modifying struct members of parser context * directly. * * @param ctxt an XML parser context */ static void xmlCtxtInitializeLate(xmlParserCtxtPtr ctxt) { xmlSAXHandlerPtr sax; /* Avoid unused variable warning if features are disabled. */ (void) sax; /* * Changing the SAX struct directly is still widespread practice * in internal and external code. */ if (ctxt == NULL) return; sax = ctxt->sax; #ifdef LIBXML_SAX1_ENABLED /* * Only enable SAX2 if there SAX2 element handlers, except when there * are no element handlers at all. */ if (((ctxt->options & XML_PARSE_SAX1) == 0) && (sax) && (sax->initialized == XML_SAX2_MAGIC) && ((sax->startElementNs != NULL) || (sax->endElementNs != NULL) || ((sax->startElement == NULL) && (sax->endElement == NULL)))) ctxt->sax2 = 1; #else ctxt->sax2 = 1; #endif /* LIBXML_SAX1_ENABLED */ /* * Some users replace the dictionary directly in the context struct. * We really need an API function to do that cleanly. */ ctxt->str_xml = xmlDictLookup(ctxt->dict, BAD_CAST "xml", 3); ctxt->str_xmlns = xmlDictLookup(ctxt->dict, BAD_CAST "xmlns", 5); ctxt->str_xml_ns = xmlDictLookup(ctxt->dict, XML_XML_NAMESPACE, 36); if ((ctxt->str_xml==NULL) || (ctxt->str_xmlns==NULL) || (ctxt->str_xml_ns == NULL)) { xmlErrMemory(ctxt); } xmlDictSetLimit(ctxt->dict, (ctxt->options & XML_PARSE_HUGE) ? 0 : XML_MAX_DICTIONARY_LIMIT); #ifdef LIBXML_VALID_ENABLED if (ctxt->validate) ctxt->vctxt.flags |= XML_VCTXT_VALIDATE; else ctxt->vctxt.flags &= ~XML_VCTXT_VALIDATE; #endif /* LIBXML_VALID_ENABLED */ } typedef struct { xmlHashedString prefix; xmlHashedString name; xmlHashedString value; const xmlChar *valueEnd; int external; int expandedSize; } xmlDefAttr; typedef struct _xmlDefAttrs xmlDefAttrs; typedef xmlDefAttrs *xmlDefAttrsPtr; struct _xmlDefAttrs { int nbAttrs; /* number of defaulted attributes on that element */ int maxAttrs; /* the size of the array */ #if __STDC_VERSION__ >= 199901L /* Using a C99 flexible array member avoids UBSan errors. */ xmlDefAttr attrs[] ATTRIBUTE_COUNTED_BY(maxAttrs); #else xmlDefAttr attrs[1]; #endif }; /** * Normalize the space in non CDATA attribute values: * If the attribute type is not CDATA, then the XML processor MUST further * process the normalized attribute value by discarding any leading and * trailing space (\#x20) characters, and by replacing sequences of space * (\#x20) characters by a single space (\#x20) character. * Note that the size of dst need to be at least src, and if one doesn't need * to preserve dst (and it doesn't come from a dictionary or read-only) then * passing src as dst is just fine. * * @param src the source string * @param dst the target string * @returns a pointer to the normalized value (dst) or NULL if no conversion * is needed. */ static xmlChar * xmlAttrNormalizeSpace(const xmlChar *src, xmlChar *dst) { if ((src == NULL) || (dst == NULL)) return(NULL); while (*src == 0x20) src++; while (*src != 0) { if (*src == 0x20) { while (*src == 0x20) src++; if (*src != 0) *dst++ = 0x20; } else { *dst++ = *src++; } } *dst = 0; if (dst == src) return(NULL); return(dst); } /** * Add a defaulted attribute for an element * * @param ctxt an XML parser context * @param fullname the element fullname * @param fullattr the attribute fullname * @param value the attribute value */ static void xmlAddDefAttrs(xmlParserCtxtPtr ctxt, const xmlChar *fullname, const xmlChar *fullattr, const xmlChar *value) { xmlDefAttrsPtr defaults; xmlDefAttr *attr; int len, expandedSize; xmlHashedString name; xmlHashedString prefix; xmlHashedString hvalue; const xmlChar *localname; /* * Allows to detect attribute redefinitions */ if (ctxt->attsSpecial != NULL) { if (xmlHashLookup2(ctxt->attsSpecial, fullname, fullattr) != NULL) return; } if (ctxt->attsDefault == NULL) { ctxt->attsDefault = xmlHashCreateDict(10, ctxt->dict); if (ctxt->attsDefault == NULL) goto mem_error; } /* * split the element name into prefix:localname , the string found * are within the DTD and then not associated to namespace names. */ localname = xmlSplitQName3(fullname, &len); if (localname == NULL) { name = xmlDictLookupHashed(ctxt->dict, fullname, -1); prefix.name = NULL; } else { name = xmlDictLookupHashed(ctxt->dict, localname, -1); prefix = xmlDictLookupHashed(ctxt->dict, fullname, len); if (prefix.name == NULL) goto mem_error; } if (name.name == NULL) goto mem_error; /* * make sure there is some storage */ defaults = xmlHashLookup2(ctxt->attsDefault, name.name, prefix.name); if ((defaults == NULL) || (defaults->nbAttrs >= defaults->maxAttrs)) { xmlDefAttrsPtr temp; int newSize; if (defaults == NULL) { newSize = 4; } else { if ((defaults->maxAttrs >= XML_MAX_ATTRS) || ((size_t) defaults->maxAttrs > SIZE_MAX / 2 / sizeof(temp[0]) - sizeof(*defaults))) goto mem_error; if (defaults->maxAttrs > XML_MAX_ATTRS / 2) newSize = XML_MAX_ATTRS; else newSize = defaults->maxAttrs * 2; } temp = xmlRealloc(defaults, sizeof(*defaults) + newSize * sizeof(xmlDefAttr)); if (temp == NULL) goto mem_error; if (defaults == NULL) temp->nbAttrs = 0; temp->maxAttrs = newSize; defaults = temp; if (xmlHashUpdateEntry2(ctxt->attsDefault, name.name, prefix.name, defaults, NULL) < 0) { xmlFree(defaults); goto mem_error; } } /* * Split the attribute name into prefix:localname , the string found * are within the DTD and hen not associated to namespace names. */ localname = xmlSplitQName3(fullattr, &len); if (localname == NULL) { name = xmlDictLookupHashed(ctxt->dict, fullattr, -1); prefix.name = NULL; } else { name = xmlDictLookupHashed(ctxt->dict, localname, -1); prefix = xmlDictLookupHashed(ctxt->dict, fullattr, len); if (prefix.name == NULL) goto mem_error; } if (name.name == NULL) goto mem_error; /* intern the string and precompute the end */ len = strlen((const char *) value); hvalue = xmlDictLookupHashed(ctxt->dict, value, len); if (hvalue.name == NULL) goto mem_error; expandedSize = strlen((const char *) name.name); if (prefix.name != NULL) expandedSize += strlen((const char *) prefix.name); expandedSize += len; attr = &defaults->attrs[defaults->nbAttrs++]; attr->name = name; attr->prefix = prefix; attr->value = hvalue; attr->valueEnd = hvalue.name + len; attr->external = PARSER_EXTERNAL(ctxt); attr->expandedSize = expandedSize; return; mem_error: xmlErrMemory(ctxt); } /** * Register this attribute type * * @param ctxt an XML parser context * @param fullname the element fullname * @param fullattr the attribute fullname * @param type the attribute type */ static void xmlAddSpecialAttr(xmlParserCtxtPtr ctxt, const xmlChar *fullname, const xmlChar *fullattr, int type) { if (ctxt->attsSpecial == NULL) { ctxt->attsSpecial = xmlHashCreateDict(10, ctxt->dict); if (ctxt->attsSpecial == NULL) goto mem_error; } if (PARSER_EXTERNAL(ctxt)) type |= XML_SPECIAL_EXTERNAL; if (xmlHashAdd2(ctxt->attsSpecial, fullname, fullattr, XML_INT_TO_PTR(type)) < 0) goto mem_error; return; mem_error: xmlErrMemory(ctxt); } /** * Removes CDATA attributes from the special attribute table */ static void xmlCleanSpecialAttrCallback(void *payload, void *data, const xmlChar *fullname, const xmlChar *fullattr, const xmlChar *unused ATTRIBUTE_UNUSED) { xmlParserCtxtPtr ctxt = (xmlParserCtxtPtr) data; if (XML_PTR_TO_INT(payload) == XML_ATTRIBUTE_CDATA) { xmlHashRemoveEntry2(ctxt->attsSpecial, fullname, fullattr, NULL); } } /** * Trim the list of attributes defined to remove all those of type * CDATA as they are not special. This call should be done when finishing * to parse the DTD and before starting to parse the document root. * * @param ctxt an XML parser context */ static void xmlCleanSpecialAttr(xmlParserCtxtPtr ctxt) { if (ctxt->attsSpecial == NULL) return; xmlHashScanFull(ctxt->attsSpecial, xmlCleanSpecialAttrCallback, ctxt); if (xmlHashSize(ctxt->attsSpecial) == 0) { xmlHashFree(ctxt->attsSpecial, NULL); ctxt->attsSpecial = NULL; } } /** * Checks that the value conforms to the LanguageID production: * * @deprecated Internal function, do not use. * * NOTE: this is somewhat deprecated, those productions were removed from * the XML Second edition. * * [33] LanguageID ::= Langcode ('-' Subcode)* * [34] Langcode ::= ISO639Code | IanaCode | UserCode * [35] ISO639Code ::= ([a-z] | [A-Z]) ([a-z] | [A-Z]) * [36] IanaCode ::= ('i' | 'I') '-' ([a-z] | [A-Z])+ * [37] UserCode ::= ('x' | 'X') '-' ([a-z] | [A-Z])+ * [38] Subcode ::= ([a-z] | [A-Z])+ * * The current REC reference the successors of RFC 1766, currently 5646 * * http://www.rfc-editor.org/rfc/rfc5646.txt * * langtag = language * ["-" script] * ["-" region] * *("-" variant) * *("-" extension) * ["-" privateuse] * language = 2*3ALPHA ; shortest ISO 639 code * ["-" extlang] ; sometimes followed by * ; extended language subtags * / 4ALPHA ; or reserved for future use * / 5*8ALPHA ; or registered language subtag * * extlang = 3ALPHA ; selected ISO 639 codes * *2("-" 3ALPHA) ; permanently reserved * * script = 4ALPHA ; ISO 15924 code * * region = 2ALPHA ; ISO 3166-1 code * / 3DIGIT ; UN M.49 code * * variant = 5*8alphanum ; registered variants * / (DIGIT 3alphanum) * * extension = singleton 1*("-" (2*8alphanum)) * * ; Single alphanumerics * ; "x" reserved for private use * singleton = DIGIT ; 0 - 9 * / %x41-57 ; A - W * / %x59-5A ; Y - Z * / %x61-77 ; a - w * / %x79-7A ; y - z * * it sounds right to still allow Irregular i-xxx IANA and user codes too * The parser below doesn't try to cope with extension or privateuse * that could be added but that's not interoperable anyway * * @param lang pointer to the string value * @returns 1 if correct 0 otherwise **/ int xmlCheckLanguageID(const xmlChar * lang) { const xmlChar *cur = lang, *nxt; if (cur == NULL) return (0); if (((cur[0] == 'i') && (cur[1] == '-')) || ((cur[0] == 'I') && (cur[1] == '-')) || ((cur[0] == 'x') && (cur[1] == '-')) || ((cur[0] == 'X') && (cur[1] == '-'))) { /* * Still allow IANA code and user code which were coming * from the previous version of the XML-1.0 specification * it's deprecated but we should not fail */ cur += 2; while (((cur[0] >= 'A') && (cur[0] <= 'Z')) || ((cur[0] >= 'a') && (cur[0] <= 'z'))) cur++; return(cur[0] == 0); } nxt = cur; while (((nxt[0] >= 'A') && (nxt[0] <= 'Z')) || ((nxt[0] >= 'a') && (nxt[0] <= 'z'))) nxt++; if (nxt - cur >= 4) { /* * Reserved */ if ((nxt - cur > 8) || (nxt[0] != 0)) return(0); return(1); } if (nxt - cur < 2) return(0); /* we got an ISO 639 code */ if (nxt[0] == 0) return(1); if (nxt[0] != '-') return(0); nxt++; cur = nxt; /* now we can have extlang or script or region or variant */ if ((nxt[0] >= '0') && (nxt[0] <= '9')) goto region_m49; while (((nxt[0] >= 'A') && (nxt[0] <= 'Z')) || ((nxt[0] >= 'a') && (nxt[0] <= 'z'))) nxt++; if (nxt - cur == 4) goto script; if (nxt - cur == 2) goto region; if ((nxt - cur >= 5) && (nxt - cur <= 8)) goto variant; if (nxt - cur != 3) return(0); /* we parsed an extlang */ if (nxt[0] == 0) return(1); if (nxt[0] != '-') return(0); nxt++; cur = nxt; /* now we can have script or region or variant */ if ((nxt[0] >= '0') && (nxt[0] <= '9')) goto region_m49; while (((nxt[0] >= 'A') && (nxt[0] <= 'Z')) || ((nxt[0] >= 'a') && (nxt[0] <= 'z'))) nxt++; if (nxt - cur == 2) goto region; if ((nxt - cur >= 5) && (nxt - cur <= 8)) goto variant; if (nxt - cur != 4) return(0); /* we parsed a script */ script: if (nxt[0] == 0) return(1); if (nxt[0] != '-') return(0); nxt++; cur = nxt; /* now we can have region or variant */ if ((nxt[0] >= '0') && (nxt[0] <= '9')) goto region_m49; while (((nxt[0] >= 'A') && (nxt[0] <= 'Z')) || ((nxt[0] >= 'a') && (nxt[0] <= 'z'))) nxt++; if ((nxt - cur >= 5) && (nxt - cur <= 8)) goto variant; if (nxt - cur != 2) return(0); /* we parsed a region */ region: if (nxt[0] == 0) return(1); if (nxt[0] != '-') return(0); nxt++; cur = nxt; /* now we can just have a variant */ while (((nxt[0] >= 'A') && (nxt[0] <= 'Z')) || ((nxt[0] >= 'a') && (nxt[0] <= 'z'))) nxt++; if ((nxt - cur < 5) || (nxt - cur > 8)) return(0); /* we parsed a variant */ variant: if (nxt[0] == 0) return(1); if (nxt[0] != '-') return(0); /* extensions and private use subtags not checked */ return (1); region_m49: if (((nxt[1] >= '0') && (nxt[1] <= '9')) && ((nxt[2] >= '0') && (nxt[2] <= '9'))) { nxt += 3; goto region; } return(0); } /************************************************************************ * * * Parser stacks related functions and macros * * * ************************************************************************/ static xmlChar * xmlParseStringEntityRef(xmlParserCtxtPtr ctxt, const xmlChar **str); /** * Create a new namespace database. * * @returns the new obejct. */ xmlParserNsData * xmlParserNsCreate(void) { xmlParserNsData *nsdb = xmlMalloc(sizeof(*nsdb)); if (nsdb == NULL) return(NULL); memset(nsdb, 0, sizeof(*nsdb)); nsdb->defaultNsIndex = INT_MAX; return(nsdb); } /** * Free a namespace database. * * @param nsdb namespace database */ void xmlParserNsFree(xmlParserNsData *nsdb) { if (nsdb == NULL) return; xmlFree(nsdb->extra); xmlFree(nsdb->hash); xmlFree(nsdb); } /** * Reset a namespace database. * * @param nsdb namespace database */ static void xmlParserNsReset(xmlParserNsData *nsdb) { if (nsdb == NULL) return; nsdb->hashElems = 0; nsdb->elementId = 0; nsdb->defaultNsIndex = INT_MAX; if (nsdb->hash) memset(nsdb->hash, 0, nsdb->hashSize * sizeof(nsdb->hash[0])); } /** * Signal that a new element has started. * * @param nsdb namespace database * @returns 0 on success, -1 if the element counter overflowed. */ static int xmlParserNsStartElement(xmlParserNsData *nsdb) { if (nsdb->elementId == UINT_MAX) return(-1); nsdb->elementId++; return(0); } /** * Lookup namespace with given prefix. If `bucketPtr` is non-NULL, it will * be set to the matching bucket, or the first empty bucket if no match * was found. * * @param ctxt parser context * @param prefix namespace prefix * @param bucketPtr optional bucket (return value) * @returns the namespace index on success, INT_MAX if no namespace was * found. */ static int xmlParserNsLookup(xmlParserCtxtPtr ctxt, const xmlHashedString *prefix, xmlParserNsBucket **bucketPtr) { xmlParserNsBucket *bucket, *tombstone; unsigned index, hashValue; if (prefix->name == NULL) return(ctxt->nsdb->defaultNsIndex); if (ctxt->nsdb->hashSize == 0) return(INT_MAX); hashValue = prefix->hashValue; index = hashValue & (ctxt->nsdb->hashSize - 1); bucket = &ctxt->nsdb->hash[index]; tombstone = NULL; while (bucket->hashValue) { if (bucket->index == INT_MAX) { if (tombstone == NULL) tombstone = bucket; } else if (bucket->hashValue == hashValue) { if (ctxt->nsTab[bucket->index * 2] == prefix->name) { if (bucketPtr != NULL) *bucketPtr = bucket; return(bucket->index); } } index++; bucket++; if (index == ctxt->nsdb->hashSize) { index = 0; bucket = ctxt->nsdb->hash; } } if (bucketPtr != NULL) *bucketPtr = tombstone ? tombstone : bucket; return(INT_MAX); } /** * Lookup namespace URI with given prefix. * * @param ctxt parser context * @param prefix namespace prefix * @returns the namespace URI on success, NULL if no namespace was found. */ static const xmlChar * xmlParserNsLookupUri(xmlParserCtxtPtr ctxt, const xmlHashedString *prefix) { const xmlChar *ret; int nsIndex; if (prefix->name == ctxt->str_xml) return(ctxt->str_xml_ns); /* * minNsIndex is used when building an entity tree. We must * ignore namespaces declared outside the entity. */ nsIndex = xmlParserNsLookup(ctxt, prefix, NULL); if ((nsIndex == INT_MAX) || (nsIndex < ctxt->nsdb->minNsIndex)) return(NULL); ret = ctxt->nsTab[nsIndex * 2 + 1]; if (ret[0] == 0) ret = NULL; return(ret); } /** * Lookup extra data for the given prefix. This returns data stored * with xmlParserNsUdpateSax. * * @param ctxt parser context * @param prefix namespace prefix * @returns the data on success, NULL if no namespace was found. */ void * xmlParserNsLookupSax(xmlParserCtxt *ctxt, const xmlChar *prefix) { xmlHashedString hprefix; int nsIndex; if (prefix == ctxt->str_xml) return(NULL); hprefix.name = prefix; if (prefix != NULL) hprefix.hashValue = xmlDictComputeHash(ctxt->dict, prefix); else hprefix.hashValue = 0; nsIndex = xmlParserNsLookup(ctxt, &hprefix, NULL); if ((nsIndex == INT_MAX) || (nsIndex < ctxt->nsdb->minNsIndex)) return(NULL); return(ctxt->nsdb->extra[nsIndex].saxData); } /** * Sets or updates extra data for the given prefix. This value will be * returned by xmlParserNsLookupSax as long as the namespace with the * given prefix is in scope. * * @param ctxt parser context * @param prefix namespace prefix * @param saxData extra data for SAX handler * @returns the data on success, NULL if no namespace was found. */ int xmlParserNsUpdateSax(xmlParserCtxt *ctxt, const xmlChar *prefix, void *saxData) { xmlHashedString hprefix; int nsIndex; if (prefix == ctxt->str_xml) return(-1); hprefix.name = prefix; if (prefix != NULL) hprefix.hashValue = xmlDictComputeHash(ctxt->dict, prefix); else hprefix.hashValue = 0; nsIndex = xmlParserNsLookup(ctxt, &hprefix, NULL); if ((nsIndex == INT_MAX) || (nsIndex < ctxt->nsdb->minNsIndex)) return(-1); ctxt->nsdb->extra[nsIndex].saxData = saxData; return(0); } /** * Grows the namespace tables. * * @param ctxt parser context * @returns 0 on success, -1 if a memory allocation failed. */ static int xmlParserNsGrow(xmlParserCtxtPtr ctxt) { const xmlChar **table; xmlParserNsExtra *extra; int newSize; newSize = xmlGrowCapacity(ctxt->nsMax, sizeof(table[0]) + sizeof(extra[0]), 16, XML_MAX_ITEMS); if (newSize < 0) goto error; table = xmlRealloc(ctxt->nsTab, 2 * newSize * sizeof(table[0])); if (table == NULL) goto error; ctxt->nsTab = table; extra = xmlRealloc(ctxt->nsdb->extra, newSize * sizeof(extra[0])); if (extra == NULL) goto error; ctxt->nsdb->extra = extra; ctxt->nsMax = newSize; return(0); error: xmlErrMemory(ctxt); return(-1); } /** * Push a new namespace on the table. * * @param ctxt parser context * @param prefix prefix with hash value * @param uri uri with hash value * @param saxData extra data for SAX handler * @param defAttr whether the namespace comes from a default attribute * @returns 1 if the namespace was pushed, 0 if the namespace was ignored, * -1 if a memory allocation failed. */ static int xmlParserNsPush(xmlParserCtxtPtr ctxt, const xmlHashedString *prefix, const xmlHashedString *uri, void *saxData, int defAttr) { xmlParserNsBucket *bucket = NULL; xmlParserNsExtra *extra; const xmlChar **ns; unsigned hashValue, nsIndex, oldIndex; if ((prefix != NULL) && (prefix->name == ctxt->str_xml)) return(0); if ((ctxt->nsNr >= ctxt->nsMax) && (xmlParserNsGrow(ctxt) < 0)) { xmlErrMemory(ctxt); return(-1); } /* * Default namespace and 'xml' namespace */ if ((prefix == NULL) || (prefix->name == NULL)) { oldIndex = ctxt->nsdb->defaultNsIndex; if (oldIndex != INT_MAX) { extra = &ctxt->nsdb->extra[oldIndex]; if (extra->elementId == ctxt->nsdb->elementId) { if (defAttr == 0) xmlErrAttributeDup(ctxt, NULL, BAD_CAST "xmlns"); return(0); } if ((ctxt->options & XML_PARSE_NSCLEAN) && (uri->name == ctxt->nsTab[oldIndex * 2 + 1])) return(0); } ctxt->nsdb->defaultNsIndex = ctxt->nsNr; goto populate_entry; } /* * Hash table lookup */ oldIndex = xmlParserNsLookup(ctxt, prefix, &bucket); if (oldIndex != INT_MAX) { extra = &ctxt->nsdb->extra[oldIndex]; /* * Check for duplicate definitions on the same element. */ if (extra->elementId == ctxt->nsdb->elementId) { if (defAttr == 0) xmlErrAttributeDup(ctxt, BAD_CAST "xmlns", prefix->name); return(0); } if ((ctxt->options & XML_PARSE_NSCLEAN) && (uri->name == ctxt->nsTab[bucket->index * 2 + 1])) return(0); bucket->index = ctxt->nsNr; goto populate_entry; } /* * Insert new bucket */ hashValue = prefix->hashValue; /* * Grow hash table, 50% fill factor */ if (ctxt->nsdb->hashElems + 1 > ctxt->nsdb->hashSize / 2) { xmlParserNsBucket *newHash; unsigned newSize, i, index; if (ctxt->nsdb->hashSize > UINT_MAX / 2) { xmlErrMemory(ctxt); return(-1); } newSize = ctxt->nsdb->hashSize ? ctxt->nsdb->hashSize * 2 : 16; newHash = xmlMalloc(newSize * sizeof(newHash[0])); if (newHash == NULL) { xmlErrMemory(ctxt); return(-1); } memset(newHash, 0, newSize * sizeof(newHash[0])); for (i = 0; i < ctxt->nsdb->hashSize; i++) { unsigned hv = ctxt->nsdb->hash[i].hashValue; unsigned newIndex; if ((hv == 0) || (ctxt->nsdb->hash[i].index == INT_MAX)) continue; newIndex = hv & (newSize - 1); while (newHash[newIndex].hashValue != 0) { newIndex++; if (newIndex == newSize) newIndex = 0; } newHash[newIndex] = ctxt->nsdb->hash[i]; } xmlFree(ctxt->nsdb->hash); ctxt->nsdb->hash = newHash; ctxt->nsdb->hashSize = newSize; /* * Relookup */ index = hashValue & (newSize - 1); while (newHash[index].hashValue != 0) { index++; if (index == newSize) index = 0; } bucket = &newHash[index]; } bucket->hashValue = hashValue; bucket->index = ctxt->nsNr; ctxt->nsdb->hashElems++; oldIndex = INT_MAX; populate_entry: nsIndex = ctxt->nsNr; ns = &ctxt->nsTab[nsIndex * 2]; ns[0] = prefix ? prefix->name : NULL; ns[1] = uri->name; extra = &ctxt->nsdb->extra[nsIndex]; extra->saxData = saxData; extra->prefixHashValue = prefix ? prefix->hashValue : 0; extra->uriHashValue = uri->hashValue; extra->elementId = ctxt->nsdb->elementId; extra->oldIndex = oldIndex; ctxt->nsNr++; return(1); } /** * Pops the top `nr` namespaces and restores the hash table. * * @param ctxt an XML parser context * @param nr the number to pop * @returns the number of namespaces popped. */ static int xmlParserNsPop(xmlParserCtxtPtr ctxt, int nr) { int i; /* assert(nr <= ctxt->nsNr); */ for (i = ctxt->nsNr - 1; i >= ctxt->nsNr - nr; i--) { const xmlChar *prefix = ctxt->nsTab[i * 2]; xmlParserNsExtra *extra = &ctxt->nsdb->extra[i]; if (prefix == NULL) { ctxt->nsdb->defaultNsIndex = extra->oldIndex; } else { xmlHashedString hprefix; xmlParserNsBucket *bucket = NULL; hprefix.name = prefix; hprefix.hashValue = extra->prefixHashValue; xmlParserNsLookup(ctxt, &hprefix, &bucket); /* assert(bucket && bucket->hashValue); */ bucket->index = extra->oldIndex; } } ctxt->nsNr -= nr; return(nr); } static int xmlCtxtGrowAttrs(xmlParserCtxtPtr ctxt) { const xmlChar **atts; unsigned *attallocs; int newSize; newSize = xmlGrowCapacity(ctxt->maxatts / 5, sizeof(atts[0]) * 5 + sizeof(attallocs[0]), 10, XML_MAX_ATTRS); if (newSize < 0) { xmlFatalErr(ctxt, XML_ERR_RESOURCE_LIMIT, "Maximum number of attributes exceeded"); return(-1); } atts = xmlRealloc(ctxt->atts, newSize * sizeof(atts[0]) * 5); if (atts == NULL) goto mem_error; ctxt->atts = atts; attallocs = xmlRealloc(ctxt->attallocs, newSize * sizeof(attallocs[0])); if (attallocs == NULL) goto mem_error; ctxt->attallocs = attallocs; ctxt->maxatts = newSize * 5; return(0); mem_error: xmlErrMemory(ctxt); return(-1); } /** * Pushes a new parser input on top of the input stack * * @param ctxt an XML parser context * @param value the parser input * @returns -1 in case of error, the index in the stack otherwise */ int xmlCtxtPushInput(xmlParserCtxt *ctxt, xmlParserInput *value) { char *directory = NULL; int maxDepth; if ((ctxt == NULL) || (value == NULL)) return(-1); maxDepth = (ctxt->options & XML_PARSE_HUGE) ? 40 : 20; if (ctxt->inputNr >= ctxt->inputMax) { xmlParserInputPtr *tmp; int newSize; newSize = xmlGrowCapacity(ctxt->inputMax, sizeof(tmp[0]), 5, maxDepth); if (newSize < 0) { xmlFatalErrMsg(ctxt, XML_ERR_RESOURCE_LIMIT, "Maximum entity nesting depth exceeded"); return(-1); } tmp = xmlRealloc(ctxt->inputTab, newSize * sizeof(tmp[0])); if (tmp == NULL) { xmlErrMemory(ctxt); return(-1); } ctxt->inputTab = tmp; ctxt->inputMax = newSize; } if ((ctxt->inputNr == 0) && (value->filename != NULL)) { directory = xmlParserGetDirectory(value->filename); if (directory == NULL) { xmlErrMemory(ctxt); return(-1); } } if (ctxt->input_id >= INT_MAX) { xmlFatalErrMsg(ctxt, XML_ERR_RESOURCE_LIMIT, "Input ID overflow\n"); return(-1); } ctxt->inputTab[ctxt->inputNr] = value; ctxt->input = value; if (ctxt->inputNr == 0) { xmlFree(ctxt->directory); ctxt->directory = directory; } /* * The input ID is unused internally, but there are entity * loaders in downstream code that detect the main document * by checking for "input_id == 1". */ value->id = ctxt->input_id++; return(ctxt->inputNr++); } /** * Pops the top parser input from the input stack * * @param ctxt an XML parser context * @returns the input just removed */ xmlParserInput * xmlCtxtPopInput(xmlParserCtxt *ctxt) { xmlParserInputPtr ret; if (ctxt == NULL) return(NULL); if (ctxt->inputNr <= 0) return (NULL); ctxt->inputNr--; if (ctxt->inputNr > 0) ctxt->input = ctxt->inputTab[ctxt->inputNr - 1]; else ctxt->input = NULL; ret = ctxt->inputTab[ctxt->inputNr]; ctxt->inputTab[ctxt->inputNr] = NULL; return (ret); } /** * Pushes a new element node on top of the node stack * * @deprecated Internal function, do not use. * * @param ctxt an XML parser context * @param value the element node * @returns -1 in case of error, the index in the stack otherwise */ int nodePush(xmlParserCtxt *ctxt, xmlNode *value) { if (ctxt == NULL) return(0); if (ctxt->nodeNr >= ctxt->nodeMax) { int maxDepth = (ctxt->options & XML_PARSE_HUGE) ? 2048 : 256; xmlNodePtr *tmp; int newSize; newSize = xmlGrowCapacity(ctxt->nodeMax, sizeof(tmp[0]), 10, maxDepth); if (newSize < 0) { xmlFatalErrMsgInt(ctxt, XML_ERR_RESOURCE_LIMIT, "Excessive depth in document: %d," " use XML_PARSE_HUGE option\n", ctxt->nodeNr); return(-1); } tmp = xmlRealloc(ctxt->nodeTab, newSize * sizeof(tmp[0])); if (tmp == NULL) { xmlErrMemory(ctxt); return (-1); } ctxt->nodeTab = tmp; ctxt->nodeMax = newSize; } ctxt->nodeTab[ctxt->nodeNr] = value; ctxt->node = value; return (ctxt->nodeNr++); } /** * Pops the top element node from the node stack * * @deprecated Internal function, do not use. * * @param ctxt an XML parser context * @returns the node just removed */ xmlNode * nodePop(xmlParserCtxt *ctxt) { xmlNodePtr ret; if (ctxt == NULL) return(NULL); if (ctxt->nodeNr <= 0) return (NULL); ctxt->nodeNr--; if (ctxt->nodeNr > 0) ctxt->node = ctxt->nodeTab[ctxt->nodeNr - 1]; else ctxt->node = NULL; ret = ctxt->nodeTab[ctxt->nodeNr]; ctxt->nodeTab[ctxt->nodeNr] = NULL; return (ret); } /** * Pushes a new element name/prefix/URL on top of the name stack * * @param ctxt an XML parser context * @param value the element name * @param prefix the element prefix * @param URI the element namespace name * @param line the current line number for error messages * @param nsNr the number of namespaces pushed on the namespace table * @returns -1 in case of error, the index in the stack otherwise */ static int nameNsPush(xmlParserCtxtPtr ctxt, const xmlChar * value, const xmlChar *prefix, const xmlChar *URI, int line, int nsNr) { xmlStartTag *tag; if (ctxt->nameNr >= ctxt->nameMax) { const xmlChar **tmp; xmlStartTag *tmp2; int newSize; newSize = xmlGrowCapacity(ctxt->nameMax, sizeof(tmp[0]) + sizeof(tmp2[0]), 10, XML_MAX_ITEMS); if (newSize < 0) goto mem_error; tmp = xmlRealloc(ctxt->nameTab, newSize * sizeof(tmp[0])); if (tmp == NULL) goto mem_error; ctxt->nameTab = tmp; tmp2 = xmlRealloc(ctxt->pushTab, newSize * sizeof(tmp2[0])); if (tmp2 == NULL) goto mem_error; ctxt->pushTab = tmp2; ctxt->nameMax = newSize; } else if (ctxt->pushTab == NULL) { ctxt->pushTab = xmlMalloc(ctxt->nameMax * sizeof(ctxt->pushTab[0])); if (ctxt->pushTab == NULL) goto mem_error; } ctxt->nameTab[ctxt->nameNr] = value; ctxt->name = value; tag = &ctxt->pushTab[ctxt->nameNr]; tag->prefix = prefix; tag->URI = URI; tag->line = line; tag->nsNr = nsNr; return (ctxt->nameNr++); mem_error: xmlErrMemory(ctxt); return (-1); } #ifdef LIBXML_PUSH_ENABLED /** * Pops the top element/prefix/URI name from the name stack * * @param ctxt an XML parser context * @returns the name just removed */ static const xmlChar * nameNsPop(xmlParserCtxtPtr ctxt) { const xmlChar *ret; if (ctxt->nameNr <= 0) return (NULL); ctxt->nameNr--; if (ctxt->nameNr > 0) ctxt->name = ctxt->nameTab[ctxt->nameNr - 1]; else ctxt->name = NULL; ret = ctxt->nameTab[ctxt->nameNr]; ctxt->nameTab[ctxt->nameNr] = NULL; return (ret); } #endif /* LIBXML_PUSH_ENABLED */ /** * Pops the top element name from the name stack * * @deprecated Internal function, do not use. * * @param ctxt an XML parser context * @returns the name just removed */ static const xmlChar * namePop(xmlParserCtxtPtr ctxt) { const xmlChar *ret; if ((ctxt == NULL) || (ctxt->nameNr <= 0)) return (NULL); ctxt->nameNr--; if (ctxt->nameNr > 0) ctxt->name = ctxt->nameTab[ctxt->nameNr - 1]; else ctxt->name = NULL; ret = ctxt->nameTab[ctxt->nameNr]; ctxt->nameTab[ctxt->nameNr] = NULL; return (ret); } static int spacePush(xmlParserCtxtPtr ctxt, int val) { if (ctxt->spaceNr >= ctxt->spaceMax) { int *tmp; int newSize; newSize = xmlGrowCapacity(ctxt->spaceMax, sizeof(tmp[0]), 10, XML_MAX_ITEMS); if (newSize < 0) { xmlErrMemory(ctxt); return(-1); } tmp = xmlRealloc(ctxt->spaceTab, newSize * sizeof(tmp[0])); if (tmp == NULL) { xmlErrMemory(ctxt); return(-1); } ctxt->spaceTab = tmp; ctxt->spaceMax = newSize; } ctxt->spaceTab[ctxt->spaceNr] = val; ctxt->space = &ctxt->spaceTab[ctxt->spaceNr]; return(ctxt->spaceNr++); } static int spacePop(xmlParserCtxtPtr ctxt) { int ret; if (ctxt->spaceNr <= 0) return(0); ctxt->spaceNr--; if (ctxt->spaceNr > 0) ctxt->space = &ctxt->spaceTab[ctxt->spaceNr - 1]; else ctxt->space = &ctxt->spaceTab[0]; ret = ctxt->spaceTab[ctxt->spaceNr]; ctxt->spaceTab[ctxt->spaceNr] = -1; return(ret); } /* * Macros for accessing the content. Those should be used only by the parser, * and not exported. * * Dirty macros, i.e. one often need to make assumption on the context to * use them * * CUR_PTR return the current pointer to the xmlChar to be parsed. * To be used with extreme caution since operations consuming * characters may move the input buffer to a different location ! * CUR returns the current xmlChar value, i.e. a 8 bit value if compiled * This should be used internally by the parser * only to compare to ASCII values otherwise it would break when * running with UTF-8 encoding. * RAW same as CUR but in the input buffer, bypass any token * extraction that may have been done * NXT(n) returns the n'th next xmlChar. Same as CUR is should be used only * to compare on ASCII based substring. * SKIP(n) Skip n xmlChar, and must also be used only to skip ASCII defined * strings without newlines within the parser. * NEXT1(l) Skip 1 xmlChar, and must also be used only to skip 1 non-newline ASCII * defined char within the parser. * Clean macros, not dependent of an ASCII context, expect UTF-8 encoding * * NEXT Skip to the next character, this does the proper decoding * in UTF-8 mode. It also pop-up unfinished entities on the fly. * NEXTL(l) Skip the current unicode character of l xmlChars long. * COPY_BUF copy the current unicode char to the target buffer, increment * the index * GROW, SHRINK handling of input buffers */ #define RAW (*ctxt->input->cur) #define CUR (*ctxt->input->cur) #define NXT(val) ctxt->input->cur[(val)] #define CUR_PTR ctxt->input->cur #define BASE_PTR ctxt->input->base #define CMP4( s, c1, c2, c3, c4 ) \ ( ((unsigned char *) s)[ 0 ] == c1 && ((unsigned char *) s)[ 1 ] == c2 && \ ((unsigned char *) s)[ 2 ] == c3 && ((unsigned char *) s)[ 3 ] == c4 ) #define CMP5( s, c1, c2, c3, c4, c5 ) \ ( CMP4( s, c1, c2, c3, c4 ) && ((unsigned char *) s)[ 4 ] == c5 ) #define CMP6( s, c1, c2, c3, c4, c5, c6 ) \ ( CMP5( s, c1, c2, c3, c4, c5 ) && ((unsigned char *) s)[ 5 ] == c6 ) #define CMP7( s, c1, c2, c3, c4, c5, c6, c7 ) \ ( CMP6( s, c1, c2, c3, c4, c5, c6 ) && ((unsigned char *) s)[ 6 ] == c7 ) #define CMP8( s, c1, c2, c3, c4, c5, c6, c7, c8 ) \ ( CMP7( s, c1, c2, c3, c4, c5, c6, c7 ) && ((unsigned char *) s)[ 7 ] == c8 ) #define CMP9( s, c1, c2, c3, c4, c5, c6, c7, c8, c9 ) \ ( CMP8( s, c1, c2, c3, c4, c5, c6, c7, c8 ) && \ ((unsigned char *) s)[ 8 ] == c9 ) #define CMP10( s, c1, c2, c3, c4, c5, c6, c7, c8, c9, c10 ) \ ( CMP9( s, c1, c2, c3, c4, c5, c6, c7, c8, c9 ) && \ ((unsigned char *) s)[ 9 ] == c10 ) #define SKIP(val) do { \ ctxt->input->cur += (val),ctxt->input->col+=(val); \ if (*ctxt->input->cur == 0) \ xmlParserGrow(ctxt); \ } while (0) #define SKIPL(val) do { \ int skipl; \ for(skipl=0; skiplinput->cur) == '\n') { \ ctxt->input->line++; ctxt->input->col = 1; \ } else ctxt->input->col++; \ ctxt->input->cur++; \ } \ if (*ctxt->input->cur == 0) \ xmlParserGrow(ctxt); \ } while (0) #define SHRINK \ if (!PARSER_PROGRESSIVE(ctxt)) \ xmlParserShrink(ctxt); #define GROW \ if ((!PARSER_PROGRESSIVE(ctxt)) && \ (ctxt->input->end - ctxt->input->cur < INPUT_CHUNK)) \ xmlParserGrow(ctxt); #define SKIP_BLANKS xmlSkipBlankChars(ctxt) #define SKIP_BLANKS_PE xmlSkipBlankCharsPE(ctxt) #define NEXT xmlNextChar(ctxt) #define NEXT1 { \ ctxt->input->col++; \ ctxt->input->cur++; \ if (*ctxt->input->cur == 0) \ xmlParserGrow(ctxt); \ } #define NEXTL(l) do { \ if (*(ctxt->input->cur) == '\n') { \ ctxt->input->line++; ctxt->input->col = 1; \ } else ctxt->input->col++; \ ctxt->input->cur += l; \ } while (0) #define COPY_BUF(b, i, v) \ if (v < 0x80) b[i++] = v; \ else i += xmlCopyCharMultiByte(&b[i],v) static int xmlCurrentCharRecover(xmlParserCtxtPtr ctxt, int *len) { int c = xmlCurrentChar(ctxt, len); if (c == XML_INVALID_CHAR) c = 0xFFFD; /* replacement character */ return(c); } /** * Skip whitespace in the input stream. * * @deprecated Internal function, do not use. * * @param ctxt the XML parser context * @returns the number of space chars skipped */ int xmlSkipBlankChars(xmlParserCtxt *ctxt) { const xmlChar *cur; int res = 0; cur = ctxt->input->cur; while (IS_BLANK_CH(*cur)) { if (*cur == '\n') { ctxt->input->line++; ctxt->input->col = 1; } else { ctxt->input->col++; } cur++; if (res < INT_MAX) res++; if (*cur == 0) { ctxt->input->cur = cur; xmlParserGrow(ctxt); cur = ctxt->input->cur; } } ctxt->input->cur = cur; if (res > 4) GROW; return(res); } static void xmlPopPE(xmlParserCtxtPtr ctxt) { unsigned long consumed; xmlEntityPtr ent; ent = ctxt->input->entity; ent->flags &= ~XML_ENT_EXPANDING; if ((ent->flags & XML_ENT_CHECKED) == 0) { int result; /* * Read the rest of the stream in case of errors. We want * to account for the whole entity size. */ do { ctxt->input->cur = ctxt->input->end; xmlParserShrink(ctxt); result = xmlParserGrow(ctxt); } while (result > 0); consumed = ctxt->input->consumed; xmlSaturatedAddSizeT(&consumed, ctxt->input->end - ctxt->input->base); xmlSaturatedAdd(&ent->expandedSize, consumed); /* * Add to sizeentities when parsing an external entity * for the first time. */ if (ent->etype == XML_EXTERNAL_PARAMETER_ENTITY) { xmlSaturatedAdd(&ctxt->sizeentities, consumed); } ent->flags |= XML_ENT_CHECKED; } xmlFreeInputStream(xmlCtxtPopInput(ctxt)); xmlParserEntityCheck(ctxt, ent->expandedSize); GROW; } /** * Skip whitespace in the input stream, also handling parameter * entities. * * @param ctxt the XML parser context * @returns the number of space chars skipped */ static int xmlSkipBlankCharsPE(xmlParserCtxtPtr ctxt) { int res = 0; int inParam; int expandParam; inParam = PARSER_IN_PE(ctxt); expandParam = PARSER_EXTERNAL(ctxt); if (!inParam && !expandParam) return(xmlSkipBlankChars(ctxt)); /* * It's Okay to use CUR/NEXT here since all the blanks are on * the ASCII range. */ while (PARSER_STOPPED(ctxt) == 0) { if (IS_BLANK_CH(CUR)) { /* CHECKED tstblanks.xml */ NEXT; } else if (CUR == '%') { if ((expandParam == 0) || (IS_BLANK_CH(NXT(1))) || (NXT(1) == 0)) break; /* * Expand parameter entity. We continue to consume * whitespace at the start of the entity and possible * even consume the whole entity and pop it. We might * even pop multiple PEs in this loop. */ xmlParsePERefInternal(ctxt, 0); inParam = PARSER_IN_PE(ctxt); expandParam = PARSER_EXTERNAL(ctxt); } else if (CUR == 0) { if (inParam == 0) break; /* * Don't pop parameter entities that start a markup * declaration to detect Well-formedness constraint: * PE Between Declarations. */ if (ctxt->input->flags & XML_INPUT_MARKUP_DECL) break; xmlPopPE(ctxt); inParam = PARSER_IN_PE(ctxt); expandParam = PARSER_EXTERNAL(ctxt); } else { break; } /* * Also increase the counter when entering or exiting a PERef. * The spec says: "When a parameter-entity reference is recognized * in the DTD and included, its replacement text MUST be enlarged * by the attachment of one leading and one following space (#x20) * character." */ if (res < INT_MAX) res++; } return(res); } /************************************************************************ * * * Commodity functions to handle entities * * * ************************************************************************/ /** * @deprecated Internal function, don't use. * * @param ctxt an XML parser context * @returns the current xmlChar in the parser context */ xmlChar xmlPopInput(xmlParserCtxt *ctxt) { xmlParserInputPtr input; if ((ctxt == NULL) || (ctxt->inputNr <= 1)) return(0); input = xmlCtxtPopInput(ctxt); xmlFreeInputStream(input); if (*ctxt->input->cur == 0) xmlParserGrow(ctxt); return(CUR); } /** * Push an input stream onto the stack. * * @deprecated Internal function, don't use. * * @param ctxt an XML parser context * @param input an XML parser input fragment (entity, XML fragment ...). * @returns -1 in case of error or the index in the input stack */ int xmlPushInput(xmlParserCtxt *ctxt, xmlParserInput *input) { int ret; if ((ctxt == NULL) || (input == NULL)) return(-1); ret = xmlCtxtPushInput(ctxt, input); if (ret >= 0) GROW; return(ret); } /** * Parse a numeric character reference. Always consumes '&'. * * @deprecated Internal function, don't use. * * [66] CharRef ::= '&#' [0-9]+ ';' | * '&#x' [0-9a-fA-F]+ ';' * * [ WFC: Legal Character ] * Characters referred to using character references must match the * production for Char. * * @param ctxt an XML parser context * @returns the value parsed (as an int), 0 in case of error */ int xmlParseCharRef(xmlParserCtxt *ctxt) { int val = 0; int count = 0; /* * Using RAW/CUR/NEXT is okay since we are working on ASCII range here */ if ((RAW == '&') && (NXT(1) == '#') && (NXT(2) == 'x')) { SKIP(3); GROW; while ((RAW != ';') && (PARSER_STOPPED(ctxt) == 0)) { if (count++ > 20) { count = 0; GROW; } if ((RAW >= '0') && (RAW <= '9')) val = val * 16 + (CUR - '0'); else if ((RAW >= 'a') && (RAW <= 'f') && (count < 20)) val = val * 16 + (CUR - 'a') + 10; else if ((RAW >= 'A') && (RAW <= 'F') && (count < 20)) val = val * 16 + (CUR - 'A') + 10; else { xmlFatalErr(ctxt, XML_ERR_INVALID_HEX_CHARREF, NULL); val = 0; break; } if (val > 0x110000) val = 0x110000; NEXT; count++; } if (RAW == ';') { /* on purpose to avoid reentrancy problems with NEXT and SKIP */ ctxt->input->col++; ctxt->input->cur++; } } else if ((RAW == '&') && (NXT(1) == '#')) { SKIP(2); GROW; while (RAW != ';') { /* loop blocked by count */ if (count++ > 20) { count = 0; GROW; } if ((RAW >= '0') && (RAW <= '9')) val = val * 10 + (CUR - '0'); else { xmlFatalErr(ctxt, XML_ERR_INVALID_DEC_CHARREF, NULL); val = 0; break; } if (val > 0x110000) val = 0x110000; NEXT; count++; } if (RAW == ';') { /* on purpose to avoid reentrancy problems with NEXT and SKIP */ ctxt->input->col++; ctxt->input->cur++; } } else { if (RAW == '&') SKIP(1); xmlFatalErr(ctxt, XML_ERR_INVALID_CHARREF, NULL); } /* * [ WFC: Legal Character ] * Characters referred to using character references must match the * production for Char. */ if (val >= 0x110000) { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "xmlParseCharRef: character reference out of bounds\n", val); val = 0xFFFD; } else if (!IS_CHAR(val)) { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "xmlParseCharRef: invalid xmlChar value %d\n", val); } return(val); } /** * Parse Reference declarations, variant parsing from a string rather * than an an input flow. * * [66] CharRef ::= '&#' [0-9]+ ';' | * '&#x' [0-9a-fA-F]+ ';' * * [ WFC: Legal Character ] * Characters referred to using character references must match the * production for Char. * * @param ctxt an XML parser context * @param str a pointer to an index in the string * @returns the value parsed (as an int), 0 in case of error, str will be * updated to the current value of the index */ static int xmlParseStringCharRef(xmlParserCtxtPtr ctxt, const xmlChar **str) { const xmlChar *ptr; xmlChar cur; int val = 0; if ((str == NULL) || (*str == NULL)) return(0); ptr = *str; cur = *ptr; if ((cur == '&') && (ptr[1] == '#') && (ptr[2] == 'x')) { ptr += 3; cur = *ptr; while (cur != ';') { /* Non input consuming loop */ if ((cur >= '0') && (cur <= '9')) val = val * 16 + (cur - '0'); else if ((cur >= 'a') && (cur <= 'f')) val = val * 16 + (cur - 'a') + 10; else if ((cur >= 'A') && (cur <= 'F')) val = val * 16 + (cur - 'A') + 10; else { xmlFatalErr(ctxt, XML_ERR_INVALID_HEX_CHARREF, NULL); val = 0; break; } if (val > 0x110000) val = 0x110000; ptr++; cur = *ptr; } if (cur == ';') ptr++; } else if ((cur == '&') && (ptr[1] == '#')){ ptr += 2; cur = *ptr; while (cur != ';') { /* Non input consuming loops */ if ((cur >= '0') && (cur <= '9')) val = val * 10 + (cur - '0'); else { xmlFatalErr(ctxt, XML_ERR_INVALID_DEC_CHARREF, NULL); val = 0; break; } if (val > 0x110000) val = 0x110000; ptr++; cur = *ptr; } if (cur == ';') ptr++; } else { xmlFatalErr(ctxt, XML_ERR_INVALID_CHARREF, NULL); return(0); } *str = ptr; /* * [ WFC: Legal Character ] * Characters referred to using character references must match the * production for Char. */ if (val >= 0x110000) { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "xmlParseStringCharRef: character reference out of bounds\n", val); } else if (IS_CHAR(val)) { return(val); } else { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "xmlParseStringCharRef: invalid xmlChar value %d\n", val); } return(0); } /** * [69] PEReference ::= '%' Name ';' * * @deprecated Internal function, do not use. * * [ WFC: No Recursion ] * A parsed entity must not contain a recursive * reference to itself, either directly or indirectly. * * [ WFC: Entity Declared ] * In a document without any DTD, a document with only an internal DTD * subset which contains no parameter entity references, or a document * with "standalone='yes'", ... ... The declaration of a parameter * entity must precede any reference to it... * * [ VC: Entity Declared ] * In a document with an external subset or external parameter entities * with "standalone='no'", ... ... The declaration of a parameter entity * must precede any reference to it... * * [ WFC: In DTD ] * Parameter-entity references may only appear in the DTD. * NOTE: misleading but this is handled. * * A PEReference may have been detected in the current input stream * the handling is done accordingly to * http://www.w3.org/TR/REC-xml#entproc * i.e. * - Included in literal in entity values * - Included as Parameter Entity reference within DTDs * @param ctxt the parser context */ void xmlParserHandlePEReference(xmlParserCtxt *ctxt) { xmlParsePERefInternal(ctxt, 0); } /** * @deprecated Internal function, don't use. * * @param ctxt the parser context * @param str the input string * @param len the string length * @param what combination of XML_SUBSTITUTE_REF and XML_SUBSTITUTE_PEREF * @param end an end marker xmlChar, 0 if none * @param end2 an end marker xmlChar, 0 if none * @param end3 an end marker xmlChar, 0 if none * @returns A newly allocated string with the substitution done. The caller * must deallocate it ! */ xmlChar * xmlStringLenDecodeEntities(xmlParserCtxt *ctxt, const xmlChar *str, int len, int what ATTRIBUTE_UNUSED, xmlChar end, xmlChar end2, xmlChar end3) { if ((ctxt == NULL) || (str == NULL) || (len < 0)) return(NULL); if ((str[len] != 0) || (end != 0) || (end2 != 0) || (end3 != 0)) return(NULL); return(xmlExpandEntitiesInAttValue(ctxt, str, 0)); } /** * @deprecated Internal function, don't use. * * @param ctxt the parser context * @param str the input string * @param what combination of XML_SUBSTITUTE_REF and XML_SUBSTITUTE_PEREF * @param end an end marker xmlChar, 0 if none * @param end2 an end marker xmlChar, 0 if none * @param end3 an end marker xmlChar, 0 if none * @returns A newly allocated string with the substitution done. The caller * must deallocate it ! */ xmlChar * xmlStringDecodeEntities(xmlParserCtxt *ctxt, const xmlChar *str, int what ATTRIBUTE_UNUSED, xmlChar end, xmlChar end2, xmlChar end3) { if ((ctxt == NULL) || (str == NULL)) return(NULL); if ((end != 0) || (end2 != 0) || (end3 != 0)) return(NULL); return(xmlExpandEntitiesInAttValue(ctxt, str, 0)); } /************************************************************************ * * * Commodity functions, cleanup needed ? * * * ************************************************************************/ /** * Is this a sequence of blank chars that one can ignore ? * * @param ctxt an XML parser context * @param str a xmlChar * * @param len the size of `str` * @param blank_chars we know the chars are blanks * @returns 1 if ignorable 0 otherwise. */ static int areBlanks(xmlParserCtxtPtr ctxt, const xmlChar *str, int len, int blank_chars) { int i; xmlNodePtr lastChild; /* * Check for xml:space value. */ if ((ctxt->space == NULL) || (*(ctxt->space) == 1) || (*(ctxt->space) == -2)) return(0); /* * Check that the string is made of blanks */ if (blank_chars == 0) { for (i = 0;i < len;i++) if (!(IS_BLANK_CH(str[i]))) return(0); } /* * Look if the element is mixed content in the DTD if available */ if (ctxt->node == NULL) return(0); if (ctxt->myDoc != NULL) { xmlElementPtr elemDecl = NULL; xmlDocPtr doc = ctxt->myDoc; const xmlChar *prefix = NULL; if (ctxt->node->ns) prefix = ctxt->node->ns->prefix; if (doc->intSubset != NULL) elemDecl = xmlHashLookup2(doc->intSubset->elements, ctxt->node->name, prefix); if ((elemDecl == NULL) && (doc->extSubset != NULL)) elemDecl = xmlHashLookup2(doc->extSubset->elements, ctxt->node->name, prefix); if (elemDecl != NULL) { if (elemDecl->etype == XML_ELEMENT_TYPE_ELEMENT) return(1); if ((elemDecl->etype == XML_ELEMENT_TYPE_ANY) || (elemDecl->etype == XML_ELEMENT_TYPE_MIXED)) return(0); } } /* * Otherwise, heuristic :-\ * * When push parsing, we could be at the end of a chunk. * This makes the look-ahead and consequently the NOBLANKS * option unreliable. */ if ((RAW != '<') && (RAW != 0xD)) return(0); if ((ctxt->node->children == NULL) && (RAW == '<') && (NXT(1) == '/')) return(0); lastChild = xmlGetLastChild(ctxt->node); if (lastChild == NULL) { if ((ctxt->node->type != XML_ELEMENT_NODE) && (ctxt->node->content != NULL)) return(0); } else if (xmlNodeIsText(lastChild)) return(0); else if ((ctxt->node->children != NULL) && (xmlNodeIsText(ctxt->node->children))) return(0); return(1); } /************************************************************************ * * * Extra stuff for namespace support * * Relates to http://www.w3.org/TR/WD-xml-names * * * ************************************************************************/ /** * Parse an UTF8 encoded XML qualified name string * * @deprecated Don't use. * * @param ctxt an XML parser context * @param name an XML parser context * @param prefixOut a xmlChar ** * @returns the local part, and prefix is updated * to get the Prefix if any. */ xmlChar * xmlSplitQName(xmlParserCtxt *ctxt, const xmlChar *name, xmlChar **prefixOut) { xmlChar *ret; const xmlChar *localname; localname = xmlSplitQName4(name, prefixOut); if (localname == NULL) { xmlCtxtErrMemory(ctxt); return(NULL); } ret = xmlStrdup(localname); if (ret == NULL) { xmlCtxtErrMemory(ctxt); xmlFree(*prefixOut); } return(ret); } /************************************************************************ * * * The parser itself * * Relates to http://www.w3.org/TR/REC-xml * * * ************************************************************************/ /************************************************************************ * * * Routines to parse Name, NCName and NmToken * * * ************************************************************************/ /* * The two following functions are related to the change of accepted * characters for Name and NmToken in the Revision 5 of XML-1.0 * They correspond to the modified production [4] and the new production [4a] * changes in that revision. Also note that the macros used for the * productions Letter, Digit, CombiningChar and Extender are not needed * anymore. * We still keep compatibility to pre-revision5 parsing semantic if the * new XML_PARSE_OLD10 option is given to the parser. */ static int xmlIsNameStartCharNew(int c) { /* * Use the new checks of production [4] [4a] amd [5] of the * Update 5 of XML-1.0 */ if ((c != ' ') && (c != '>') && (c != '/') && /* accelerators */ (((c >= 'a') && (c <= 'z')) || ((c >= 'A') && (c <= 'Z')) || (c == '_') || (c == ':') || ((c >= 0xC0) && (c <= 0xD6)) || ((c >= 0xD8) && (c <= 0xF6)) || ((c >= 0xF8) && (c <= 0x2FF)) || ((c >= 0x370) && (c <= 0x37D)) || ((c >= 0x37F) && (c <= 0x1FFF)) || ((c >= 0x200C) && (c <= 0x200D)) || ((c >= 0x2070) && (c <= 0x218F)) || ((c >= 0x2C00) && (c <= 0x2FEF)) || ((c >= 0x3001) && (c <= 0xD7FF)) || ((c >= 0xF900) && (c <= 0xFDCF)) || ((c >= 0xFDF0) && (c <= 0xFFFD)) || ((c >= 0x10000) && (c <= 0xEFFFF)))) return(1); return(0); } static int xmlIsNameCharNew(int c) { /* * Use the new checks of production [4] [4a] amd [5] of the * Update 5 of XML-1.0 */ if ((c != ' ') && (c != '>') && (c != '/') && /* accelerators */ (((c >= 'a') && (c <= 'z')) || ((c >= 'A') && (c <= 'Z')) || ((c >= '0') && (c <= '9')) || /* !start */ (c == '_') || (c == ':') || (c == '-') || (c == '.') || (c == 0xB7) || /* !start */ ((c >= 0xC0) && (c <= 0xD6)) || ((c >= 0xD8) && (c <= 0xF6)) || ((c >= 0xF8) && (c <= 0x2FF)) || ((c >= 0x300) && (c <= 0x36F)) || /* !start */ ((c >= 0x370) && (c <= 0x37D)) || ((c >= 0x37F) && (c <= 0x1FFF)) || ((c >= 0x200C) && (c <= 0x200D)) || ((c >= 0x203F) && (c <= 0x2040)) || /* !start */ ((c >= 0x2070) && (c <= 0x218F)) || ((c >= 0x2C00) && (c <= 0x2FEF)) || ((c >= 0x3001) && (c <= 0xD7FF)) || ((c >= 0xF900) && (c <= 0xFDCF)) || ((c >= 0xFDF0) && (c <= 0xFFFD)) || ((c >= 0x10000) && (c <= 0xEFFFF)))) return(1); return(0); } static int xmlIsNameStartCharOld(int c) { if ((c != ' ') && (c != '>') && (c != '/') && /* accelerators */ ((IS_LETTER(c) || (c == '_') || (c == ':')))) return(1); return(0); } static int xmlIsNameCharOld(int c) { if ((c != ' ') && (c != '>') && (c != '/') && /* accelerators */ ((IS_LETTER(c)) || (IS_DIGIT(c)) || (c == '.') || (c == '-') || (c == '_') || (c == ':') || (IS_COMBINING(c)) || (IS_EXTENDER(c)))) return(1); return(0); } static int xmlIsNameStartChar(int c, int old10) { if (!old10) return(xmlIsNameStartCharNew(c)); else return(xmlIsNameStartCharOld(c)); } static int xmlIsNameChar(int c, int old10) { if (!old10) return(xmlIsNameCharNew(c)); else return(xmlIsNameCharOld(c)); } /* * Scan an XML Name, NCName or Nmtoken. * * Returns a pointer to the end of the name on success. If the * name is invalid, returns `ptr`. If the name is longer than * `maxSize` bytes, returns NULL. * * @param ptr pointer to the start of the name * @param maxSize maximum size in bytes * @param flags XML_SCAN_* flags * @returns a pointer to the end of the name or NULL */ const xmlChar * xmlScanName(const xmlChar *ptr, size_t maxSize, int flags) { int stop = flags & XML_SCAN_NC ? ':' : 0; int old10 = flags & XML_SCAN_OLD10 ? 1 : 0; while (1) { int c, len; c = *ptr; if (c < 0x80) { if (c == stop) break; len = 1; } else { len = 4; c = xmlGetUTF8Char(ptr, &len); if (c < 0) break; } if (flags & XML_SCAN_NMTOKEN ? !xmlIsNameChar(c, old10) : !xmlIsNameStartChar(c, old10)) break; if ((size_t) len > maxSize) return(NULL); ptr += len; maxSize -= len; flags |= XML_SCAN_NMTOKEN; } return(ptr); } static const xmlChar * xmlParseNameComplex(xmlParserCtxtPtr ctxt) { const xmlChar *ret; int len = 0, l; int c; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; int old10 = (ctxt->options & XML_PARSE_OLD10) ? 1 : 0; /* * Handler for more complex cases */ c = xmlCurrentChar(ctxt, &l); if (!xmlIsNameStartChar(c, old10)) return(NULL); len += l; NEXTL(l); c = xmlCurrentChar(ctxt, &l); while (xmlIsNameChar(c, old10)) { if (len <= INT_MAX - l) len += l; NEXTL(l); c = xmlCurrentChar(ctxt, &l); } if (len > maxLength) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "Name"); return(NULL); } if (ctxt->input->cur - ctxt->input->base < len) { /* * There were a couple of bugs where PERefs lead to to a change * of the buffer. Check the buffer size to avoid passing an invalid * pointer to xmlDictLookup. */ xmlFatalErr(ctxt, XML_ERR_INTERNAL_ERROR, "unexpected change of input buffer"); return (NULL); } if ((*ctxt->input->cur == '\n') && (ctxt->input->cur[-1] == '\r')) ret = xmlDictLookup(ctxt->dict, ctxt->input->cur - (len + 1), len); else ret = xmlDictLookup(ctxt->dict, ctxt->input->cur - len, len); if (ret == NULL) xmlErrMemory(ctxt); return(ret); } /** * Parse an XML name. * * @deprecated Internal function, don't use. * * [4] NameChar ::= Letter | Digit | '.' | '-' | '_' | ':' | * CombiningChar | Extender * * [5] Name ::= (Letter | '_' | ':') (NameChar)* * * [6] Names ::= Name (#x20 Name)* * * @param ctxt an XML parser context * @returns the Name parsed or NULL */ const xmlChar * xmlParseName(xmlParserCtxt *ctxt) { const xmlChar *in; const xmlChar *ret; size_t count = 0; size_t maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; GROW; /* * Accelerator for simple ASCII names */ in = ctxt->input->cur; if (((*in >= 0x61) && (*in <= 0x7A)) || ((*in >= 0x41) && (*in <= 0x5A)) || (*in == '_') || (*in == ':')) { in++; while (((*in >= 0x61) && (*in <= 0x7A)) || ((*in >= 0x41) && (*in <= 0x5A)) || ((*in >= 0x30) && (*in <= 0x39)) || (*in == '_') || (*in == '-') || (*in == ':') || (*in == '.')) in++; if ((*in > 0) && (*in < 0x80)) { count = in - ctxt->input->cur; if (count > maxLength) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "Name"); return(NULL); } ret = xmlDictLookup(ctxt->dict, ctxt->input->cur, count); ctxt->input->cur = in; ctxt->input->col += count; if (ret == NULL) xmlErrMemory(ctxt); return(ret); } } /* accelerator for special cases */ return(xmlParseNameComplex(ctxt)); } static xmlHashedString xmlParseNCNameComplex(xmlParserCtxtPtr ctxt) { xmlHashedString ret; int len = 0, l; int c; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; int old10 = (ctxt->options & XML_PARSE_OLD10) ? 1 : 0; size_t startPosition = 0; ret.name = NULL; ret.hashValue = 0; /* * Handler for more complex cases */ startPosition = CUR_PTR - BASE_PTR; c = xmlCurrentChar(ctxt, &l); if ((c == ' ') || (c == '>') || (c == '/') || /* accelerators */ (!xmlIsNameStartChar(c, old10) || (c == ':'))) { return(ret); } while ((c != ' ') && (c != '>') && (c != '/') && /* test bigname.xml */ (xmlIsNameChar(c, old10) && (c != ':'))) { if (len <= INT_MAX - l) len += l; NEXTL(l); c = xmlCurrentChar(ctxt, &l); } if (len > maxLength) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "NCName"); return(ret); } ret = xmlDictLookupHashed(ctxt->dict, (BASE_PTR + startPosition), len); if (ret.name == NULL) xmlErrMemory(ctxt); return(ret); } /** * Parse an XML name. * * [4NS] NCNameChar ::= Letter | Digit | '.' | '-' | '_' | * CombiningChar | Extender * * [5NS] NCName ::= (Letter | '_') (NCNameChar)* * * @param ctxt an XML parser context * @returns the Name parsed or NULL */ static xmlHashedString xmlParseNCName(xmlParserCtxtPtr ctxt) { const xmlChar *in, *e; xmlHashedString ret; size_t count = 0; size_t maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; ret.name = NULL; /* * Accelerator for simple ASCII names */ in = ctxt->input->cur; e = ctxt->input->end; if ((((*in >= 0x61) && (*in <= 0x7A)) || ((*in >= 0x41) && (*in <= 0x5A)) || (*in == '_')) && (in < e)) { in++; while ((((*in >= 0x61) && (*in <= 0x7A)) || ((*in >= 0x41) && (*in <= 0x5A)) || ((*in >= 0x30) && (*in <= 0x39)) || (*in == '_') || (*in == '-') || (*in == '.')) && (in < e)) in++; if (in >= e) goto complex; if ((*in > 0) && (*in < 0x80)) { count = in - ctxt->input->cur; if (count > maxLength) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "NCName"); return(ret); } ret = xmlDictLookupHashed(ctxt->dict, ctxt->input->cur, count); ctxt->input->cur = in; ctxt->input->col += count; if (ret.name == NULL) { xmlErrMemory(ctxt); } return(ret); } } complex: return(xmlParseNCNameComplex(ctxt)); } /** * Parse an XML name and compares for match * (specialized for endtag parsing) * * @param ctxt an XML parser context * @param other the name to compare with * @returns NULL for an illegal name, (xmlChar*) 1 for success * and the name for mismatch */ static const xmlChar * xmlParseNameAndCompare(xmlParserCtxtPtr ctxt, xmlChar const *other) { register const xmlChar *cmp = other; register const xmlChar *in; const xmlChar *ret; GROW; in = ctxt->input->cur; while (*in != 0 && *in == *cmp) { ++in; ++cmp; } if (*cmp == 0 && (*in == '>' || IS_BLANK_CH (*in))) { /* success */ ctxt->input->col += in - ctxt->input->cur; ctxt->input->cur = in; return (const xmlChar*) 1; } /* failure (or end of input buffer), check with full function */ ret = xmlParseName (ctxt); /* strings coming from the dictionary direct compare possible */ if (ret == other) { return (const xmlChar*) 1; } return ret; } /** * Parse an XML name. * * @param ctxt an XML parser context * @param str a pointer to the string pointer (IN/OUT) * @returns the Name parsed or NULL. The `str` pointer * is updated to the current location in the string. */ static xmlChar * xmlParseStringName(xmlParserCtxtPtr ctxt, const xmlChar** str) { xmlChar *ret; const xmlChar *cur = *str; int flags = 0; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; if (ctxt->options & XML_PARSE_OLD10) flags |= XML_SCAN_OLD10; cur = xmlScanName(*str, maxLength, flags); if (cur == NULL) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "NCName"); return(NULL); } if (cur == *str) return(NULL); ret = xmlStrndup(*str, cur - *str); if (ret == NULL) xmlErrMemory(ctxt); *str = cur; return(ret); } /** * Parse an XML Nmtoken. * * @deprecated Internal function, don't use. * * [7] Nmtoken ::= (NameChar)+ * * [8] Nmtokens ::= Nmtoken (#x20 Nmtoken)* * * @param ctxt an XML parser context * @returns the Nmtoken parsed or NULL */ xmlChar * xmlParseNmtoken(xmlParserCtxt *ctxt) { xmlChar buf[XML_MAX_NAMELEN + 5]; xmlChar *ret; int len = 0, l; int c; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; int old10 = (ctxt->options & XML_PARSE_OLD10) ? 1 : 0; c = xmlCurrentChar(ctxt, &l); while (xmlIsNameChar(c, old10)) { COPY_BUF(buf, len, c); NEXTL(l); c = xmlCurrentChar(ctxt, &l); if (len >= XML_MAX_NAMELEN) { /* * Okay someone managed to make a huge token, so he's ready to pay * for the processing speed. */ xmlChar *buffer; int max = len * 2; buffer = xmlMalloc(max); if (buffer == NULL) { xmlErrMemory(ctxt); return(NULL); } memcpy(buffer, buf, len); while (xmlIsNameChar(c, old10)) { if (len + 10 > max) { xmlChar *tmp; int newSize; newSize = xmlGrowCapacity(max, 1, 1, maxLength); if (newSize < 0) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "NmToken"); xmlFree(buffer); return(NULL); } tmp = xmlRealloc(buffer, newSize); if (tmp == NULL) { xmlErrMemory(ctxt); xmlFree(buffer); return(NULL); } buffer = tmp; max = newSize; } COPY_BUF(buffer, len, c); NEXTL(l); c = xmlCurrentChar(ctxt, &l); } buffer[len] = 0; return(buffer); } } if (len == 0) return(NULL); if (len > maxLength) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "NmToken"); return(NULL); } ret = xmlStrndup(buf, len); if (ret == NULL) xmlErrMemory(ctxt); return(ret); } /** * Validate an entity value and expand parameter entities. * * @param ctxt parser context * @param buf string buffer * @param str entity value * @param length size of entity value * @param depth nesting depth */ static void xmlExpandPEsInEntityValue(xmlParserCtxtPtr ctxt, xmlSBuf *buf, const xmlChar *str, int length, int depth) { int maxDepth = (ctxt->options & XML_PARSE_HUGE) ? 40 : 20; const xmlChar *end, *chunk; int c, l; if (str == NULL) return; depth += 1; if (depth > maxDepth) { xmlFatalErrMsg(ctxt, XML_ERR_RESOURCE_LIMIT, "Maximum entity nesting depth exceeded"); return; } end = str + length; chunk = str; while ((str < end) && (!PARSER_STOPPED(ctxt))) { c = *str; if (c >= 0x80) { l = xmlUTF8MultibyteLen(ctxt, str, "invalid character in entity value\n"); if (l == 0) { if (chunk < str) xmlSBufAddString(buf, chunk, str - chunk); xmlSBufAddReplChar(buf); str += 1; chunk = str; } else { str += l; } } else if (c == '&') { if (str[1] == '#') { if (chunk < str) xmlSBufAddString(buf, chunk, str - chunk); c = xmlParseStringCharRef(ctxt, &str); if (c == 0) return; xmlSBufAddChar(buf, c); chunk = str; } else { xmlChar *name; /* * General entity references are checked for * syntactic validity. */ str++; name = xmlParseStringName(ctxt, &str); if ((name == NULL) || (*str++ != ';')) { xmlFatalErrMsg(ctxt, XML_ERR_ENTITY_CHAR_ERROR, "EntityValue: '&' forbidden except for entities " "references\n"); xmlFree(name); return; } xmlFree(name); } } else if (c == '%') { xmlEntityPtr ent; if (chunk < str) xmlSBufAddString(buf, chunk, str - chunk); ent = xmlParseStringPEReference(ctxt, &str); if (ent == NULL) return; if (!PARSER_EXTERNAL(ctxt)) { xmlFatalErr(ctxt, XML_ERR_ENTITY_PE_INTERNAL, NULL); return; } if (ent->content == NULL) { /* * Note: external parsed entities will not be loaded, * it is not required for a non-validating parser to * complete external PEReferences coming from the * internal subset */ if (((ctxt->options & XML_PARSE_NO_XXE) == 0) && ((ctxt->replaceEntities) || (ctxt->validate))) { xmlLoadEntityContent(ctxt, ent); } else { xmlWarningMsg(ctxt, XML_ERR_ENTITY_PROCESSING, "not validating will not read content for " "PE entity %s\n", ent->name, NULL); } } /* * TODO: Skip if ent->content is still NULL. */ if (xmlParserEntityCheck(ctxt, ent->length)) return; if (ent->flags & XML_ENT_EXPANDING) { xmlFatalErr(ctxt, XML_ERR_ENTITY_LOOP, NULL); return; } ent->flags |= XML_ENT_EXPANDING; xmlExpandPEsInEntityValue(ctxt, buf, ent->content, ent->length, depth); ent->flags &= ~XML_ENT_EXPANDING; chunk = str; } else { /* Normal ASCII char */ if (!IS_BYTE_CHAR(c)) { xmlFatalErrMsg(ctxt, XML_ERR_INVALID_CHAR, "invalid character in entity value\n"); if (chunk < str) xmlSBufAddString(buf, chunk, str - chunk); xmlSBufAddReplChar(buf); str += 1; chunk = str; } else { str += 1; } } } if (chunk < str) xmlSBufAddString(buf, chunk, str - chunk); } /** * Parse a value for ENTITY declarations * * @deprecated Internal function, don't use. * * [9] EntityValue ::= '"' ([^%&"] | PEReference | Reference)* '"' | * "'" ([^%&'] | PEReference | Reference)* "'" * * @param ctxt an XML parser context * @param orig if non-NULL store a copy of the original entity value * @returns the EntityValue parsed with reference substituted or NULL */ xmlChar * xmlParseEntityValue(xmlParserCtxt *ctxt, xmlChar **orig) { unsigned maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_HUGE_LENGTH : XML_MAX_TEXT_LENGTH; xmlSBuf buf; const xmlChar *start; int quote, length; xmlSBufInit(&buf, maxLength); GROW; quote = CUR; if ((quote != '"') && (quote != '\'')) { xmlFatalErr(ctxt, XML_ERR_ATTRIBUTE_NOT_STARTED, NULL); return(NULL); } CUR_PTR++; length = 0; /* * Copy raw content of the entity into a buffer */ while (1) { int c; if (PARSER_STOPPED(ctxt)) goto error; if (CUR_PTR >= ctxt->input->end) { xmlFatalErrMsg(ctxt, XML_ERR_ENTITY_NOT_FINISHED, NULL); goto error; } c = CUR; if (c == 0) { xmlFatalErrMsg(ctxt, XML_ERR_INVALID_CHAR, "invalid character in entity value\n"); goto error; } if (c == quote) break; NEXTL(1); length += 1; /* * TODO: Check growth threshold */ if (ctxt->input->end - CUR_PTR < 10) GROW; } start = CUR_PTR - length; if (orig != NULL) { *orig = xmlStrndup(start, length); if (*orig == NULL) xmlErrMemory(ctxt); } xmlExpandPEsInEntityValue(ctxt, &buf, start, length, ctxt->inputNr); NEXTL(1); return(xmlSBufFinish(&buf, NULL, ctxt, "entity length too long")); error: xmlSBufCleanup(&buf, ctxt, "entity length too long"); return(NULL); } /** * Check an entity reference in an attribute value for validity * without expanding it. * * @param ctxt parser context * @param pent entity * @param depth nesting depth */ static void xmlCheckEntityInAttValue(xmlParserCtxtPtr ctxt, xmlEntityPtr pent, int depth) { int maxDepth = (ctxt->options & XML_PARSE_HUGE) ? 40 : 20; const xmlChar *str; unsigned long expandedSize = pent->length; int c, flags; depth += 1; if (depth > maxDepth) { xmlFatalErrMsg(ctxt, XML_ERR_RESOURCE_LIMIT, "Maximum entity nesting depth exceeded"); return; } if (pent->flags & XML_ENT_EXPANDING) { xmlFatalErr(ctxt, XML_ERR_ENTITY_LOOP, NULL); return; } /* * If we're parsing a default attribute value in DTD content, * the entity might reference other entities which weren't * defined yet, so the check isn't reliable. */ if (ctxt->inSubset == 0) flags = XML_ENT_CHECKED | XML_ENT_VALIDATED; else flags = XML_ENT_VALIDATED; str = pent->content; if (str == NULL) goto done; /* * Note that entity values are already validated. We only check * for illegal less-than signs and compute the expanded size * of the entity. No special handling for multi-byte characters * is needed. */ while (!PARSER_STOPPED(ctxt)) { c = *str; if (c != '&') { if (c == 0) break; if (c == '<') xmlFatalErrMsgStr(ctxt, XML_ERR_LT_IN_ATTRIBUTE, "'<' in entity '%s' is not allowed in attributes " "values\n", pent->name); str += 1; } else if (str[1] == '#') { int val; val = xmlParseStringCharRef(ctxt, &str); if (val == 0) { pent->content[0] = 0; break; } } else { xmlChar *name; xmlEntityPtr ent; name = xmlParseStringEntityRef(ctxt, &str); if (name == NULL) { pent->content[0] = 0; break; } ent = xmlLookupGeneralEntity(ctxt, name, /* inAttr */ 1); xmlFree(name); if ((ent != NULL) && (ent->etype != XML_INTERNAL_PREDEFINED_ENTITY)) { if ((ent->flags & flags) != flags) { pent->flags |= XML_ENT_EXPANDING; xmlCheckEntityInAttValue(ctxt, ent, depth); pent->flags &= ~XML_ENT_EXPANDING; } xmlSaturatedAdd(&expandedSize, ent->expandedSize); xmlSaturatedAdd(&expandedSize, XML_ENT_FIXED_COST); } } } done: if (ctxt->inSubset == 0) pent->expandedSize = expandedSize; pent->flags |= flags; } /** * Expand general entity references in an entity or attribute value. * Perform attribute value normalization. * * @param ctxt parser context * @param buf string buffer * @param str entity or attribute value * @param pent entity for entity value, NULL for attribute values * @param normalize whether to collapse whitespace * @param inSpace whitespace state * @param depth nesting depth * @param check whether to check for amplification * @returns whether there was a normalization change */ static int xmlExpandEntityInAttValue(xmlParserCtxtPtr ctxt, xmlSBuf *buf, const xmlChar *str, xmlEntityPtr pent, int normalize, int *inSpace, int depth, int check) { int maxDepth = (ctxt->options & XML_PARSE_HUGE) ? 40 : 20; int c, chunkSize; int normChange = 0; if (str == NULL) return(0); depth += 1; if (depth > maxDepth) { xmlFatalErrMsg(ctxt, XML_ERR_RESOURCE_LIMIT, "Maximum entity nesting depth exceeded"); return(0); } if (pent != NULL) { if (pent->flags & XML_ENT_EXPANDING) { xmlFatalErr(ctxt, XML_ERR_ENTITY_LOOP, NULL); return(0); } if (check) { if (xmlParserEntityCheck(ctxt, pent->length)) return(0); } } chunkSize = 0; /* * Note that entity values are already validated. No special * handling for multi-byte characters is needed. */ while (!PARSER_STOPPED(ctxt)) { c = *str; if (c != '&') { if (c == 0) break; /* * If this function is called without an entity, it is used to * expand entities in an attribute content where less-than was * already unscaped and is allowed. */ if ((pent != NULL) && (c == '<')) { xmlFatalErrMsgStr(ctxt, XML_ERR_LT_IN_ATTRIBUTE, "'<' in entity '%s' is not allowed in attributes " "values\n", pent->name); break; } if (c <= 0x20) { if ((normalize) && (*inSpace)) { /* Skip char */ if (chunkSize > 0) { xmlSBufAddString(buf, str - chunkSize, chunkSize); chunkSize = 0; } normChange = 1; } else if (c < 0x20) { if (chunkSize > 0) { xmlSBufAddString(buf, str - chunkSize, chunkSize); chunkSize = 0; } xmlSBufAddCString(buf, " ", 1); } else { chunkSize += 1; } *inSpace = 1; } else { chunkSize += 1; *inSpace = 0; } str += 1; } else if (str[1] == '#') { int val; if (chunkSize > 0) { xmlSBufAddString(buf, str - chunkSize, chunkSize); chunkSize = 0; } val = xmlParseStringCharRef(ctxt, &str); if (val == 0) { if (pent != NULL) pent->content[0] = 0; break; } if (val == ' ') { if ((normalize) && (*inSpace)) normChange = 1; else xmlSBufAddCString(buf, " ", 1); *inSpace = 1; } else { xmlSBufAddChar(buf, val); *inSpace = 0; } } else { xmlChar *name; xmlEntityPtr ent; if (chunkSize > 0) { xmlSBufAddString(buf, str - chunkSize, chunkSize); chunkSize = 0; } name = xmlParseStringEntityRef(ctxt, &str); if (name == NULL) { if (pent != NULL) pent->content[0] = 0; break; } ent = xmlLookupGeneralEntity(ctxt, name, /* inAttr */ 1); xmlFree(name); if ((ent != NULL) && (ent->etype == XML_INTERNAL_PREDEFINED_ENTITY)) { if (ent->content == NULL) { xmlFatalErrMsg(ctxt, XML_ERR_INTERNAL_ERROR, "predefined entity has no content\n"); break; } xmlSBufAddString(buf, ent->content, ent->length); *inSpace = 0; } else if ((ent != NULL) && (ent->content != NULL)) { if (pent != NULL) pent->flags |= XML_ENT_EXPANDING; normChange |= xmlExpandEntityInAttValue(ctxt, buf, ent->content, ent, normalize, inSpace, depth, check); if (pent != NULL) pent->flags &= ~XML_ENT_EXPANDING; } } } if (chunkSize > 0) xmlSBufAddString(buf, str - chunkSize, chunkSize); return(normChange); } /** * Expand general entity references in an entity or attribute value. * Perform attribute value normalization. * * @param ctxt parser context * @param str entity or attribute value * @param normalize whether to collapse whitespace * @returns the expanded attribtue value. */ xmlChar * xmlExpandEntitiesInAttValue(xmlParserCtxt *ctxt, const xmlChar *str, int normalize) { unsigned maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_HUGE_LENGTH : XML_MAX_TEXT_LENGTH; xmlSBuf buf; int inSpace = 1; xmlSBufInit(&buf, maxLength); xmlExpandEntityInAttValue(ctxt, &buf, str, NULL, normalize, &inSpace, ctxt->inputNr, /* check */ 0); if ((normalize) && (inSpace) && (buf.size > 0)) buf.size--; return(xmlSBufFinish(&buf, NULL, ctxt, "AttValue length too long")); } /** * Parse a value for an attribute. * * NOTE: if no normalization is needed, the routine will return pointers * directly from the data buffer. * * 3.3.3 Attribute-Value Normalization: * * Before the value of an attribute is passed to the application or * checked for validity, the XML processor must normalize it as follows: * * - a character reference is processed by appending the referenced * character to the attribute value * - an entity reference is processed by recursively processing the * replacement text of the entity * - a whitespace character (\#x20, \#xD, \#xA, \#x9) is processed by * appending \#x20 to the normalized value, except that only a single * \#x20 is appended for a "#xD#xA" sequence that is part of an external * parsed entity or the literal entity value of an internal parsed entity * - other characters are processed by appending them to the normalized value * * If the declared value is not CDATA, then the XML processor must further * process the normalized attribute value by discarding any leading and * trailing space (\#x20) characters, and by replacing sequences of space * (\#x20) characters by a single space (\#x20) character. * All attributes for which no declaration has been read should be treated * by a non-validating parser as if declared CDATA. * * @param ctxt an XML parser context * @param attlen attribute len result * @param outFlags resulting XML_ATTVAL_* flags * @param special value from attsSpecial * @param isNamespace whether this is a namespace declaration * @returns the AttValue parsed or NULL. The value has to be freed by the * caller if it was copied, this can be detected by val[*len] == 0. */ static xmlChar * xmlParseAttValueInternal(xmlParserCtxtPtr ctxt, int *attlen, int *outFlags, int special, int isNamespace) { unsigned maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_HUGE_LENGTH : XML_MAX_TEXT_LENGTH; xmlSBuf buf; xmlChar *ret; int c, l, quote, entFlags, chunkSize; int inSpace = 1; int replaceEntities; int normalize = (special & XML_SPECIAL_TYPE_MASK) > XML_ATTRIBUTE_CDATA; int attvalFlags = 0; /* Always expand namespace URIs */ replaceEntities = (ctxt->replaceEntities) || (isNamespace); xmlSBufInit(&buf, maxLength); GROW; quote = CUR; if ((quote != '"') && (quote != '\'')) { xmlFatalErr(ctxt, XML_ERR_ATTRIBUTE_NOT_STARTED, NULL); return(NULL); } NEXTL(1); if (ctxt->inSubset == 0) entFlags = XML_ENT_CHECKED | XML_ENT_VALIDATED; else entFlags = XML_ENT_VALIDATED; inSpace = 1; chunkSize = 0; while (1) { if (PARSER_STOPPED(ctxt)) goto error; if (CUR_PTR >= ctxt->input->end) { xmlFatalErrMsg(ctxt, XML_ERR_ATTRIBUTE_NOT_FINISHED, "AttValue: ' expected\n"); goto error; } /* * TODO: Check growth threshold */ if (ctxt->input->end - CUR_PTR < 10) GROW; c = CUR; if (c >= 0x80) { l = xmlUTF8MultibyteLen(ctxt, CUR_PTR, "invalid character in attribute value\n"); if (l == 0) { if (chunkSize > 0) { xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); chunkSize = 0; } xmlSBufAddReplChar(&buf); NEXTL(1); } else { chunkSize += l; NEXTL(l); } inSpace = 0; } else if (c != '&') { if (c > 0x20) { if (c == quote) break; if (c == '<') xmlFatalErr(ctxt, XML_ERR_LT_IN_ATTRIBUTE, NULL); chunkSize += 1; inSpace = 0; } else if (!IS_BYTE_CHAR(c)) { xmlFatalErrMsg(ctxt, XML_ERR_INVALID_CHAR, "invalid character in attribute value\n"); if (chunkSize > 0) { xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); chunkSize = 0; } xmlSBufAddReplChar(&buf); inSpace = 0; } else { /* Whitespace */ if ((normalize) && (inSpace)) { /* Skip char */ if (chunkSize > 0) { xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); chunkSize = 0; } attvalFlags |= XML_ATTVAL_NORM_CHANGE; } else if (c < 0x20) { /* Convert to space */ if (chunkSize > 0) { xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); chunkSize = 0; } xmlSBufAddCString(&buf, " ", 1); } else { chunkSize += 1; } inSpace = 1; if ((c == 0xD) && (NXT(1) == 0xA)) CUR_PTR++; } NEXTL(1); } else if (NXT(1) == '#') { int val; if (chunkSize > 0) { xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); chunkSize = 0; } val = xmlParseCharRef(ctxt); if (val == 0) goto error; if ((val == '&') && (!replaceEntities)) { /* * The reparsing will be done in xmlNodeParseContent() * called from SAX2.c */ xmlSBufAddCString(&buf, "&", 5); inSpace = 0; } else if (val == ' ') { if ((normalize) && (inSpace)) attvalFlags |= XML_ATTVAL_NORM_CHANGE; else xmlSBufAddCString(&buf, " ", 1); inSpace = 1; } else { xmlSBufAddChar(&buf, val); inSpace = 0; } } else { const xmlChar *name; xmlEntityPtr ent; if (chunkSize > 0) { xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); chunkSize = 0; } name = xmlParseEntityRefInternal(ctxt); if (name == NULL) { /* * Probably a literal '&' which wasn't escaped. * TODO: Handle gracefully in recovery mode. */ continue; } ent = xmlLookupGeneralEntity(ctxt, name, /* isAttr */ 1); if (ent == NULL) continue; if (ent->etype == XML_INTERNAL_PREDEFINED_ENTITY) { if ((ent->content[0] == '&') && (!replaceEntities)) xmlSBufAddCString(&buf, "&", 5); else xmlSBufAddString(&buf, ent->content, ent->length); inSpace = 0; } else if (replaceEntities) { if (xmlExpandEntityInAttValue(ctxt, &buf, ent->content, ent, normalize, &inSpace, ctxt->inputNr, /* check */ 1) > 0) attvalFlags |= XML_ATTVAL_NORM_CHANGE; } else { if ((ent->flags & entFlags) != entFlags) xmlCheckEntityInAttValue(ctxt, ent, ctxt->inputNr); if (xmlParserEntityCheck(ctxt, ent->expandedSize)) { ent->content[0] = 0; goto error; } /* * Just output the reference */ xmlSBufAddCString(&buf, "&", 1); xmlSBufAddString(&buf, ent->name, xmlStrlen(ent->name)); xmlSBufAddCString(&buf, ";", 1); inSpace = 0; } } } if ((buf.mem == NULL) && (outFlags != NULL)) { ret = (xmlChar *) CUR_PTR - chunkSize; if (attlen != NULL) *attlen = chunkSize; if ((normalize) && (inSpace) && (chunkSize > 0)) { attvalFlags |= XML_ATTVAL_NORM_CHANGE; *attlen -= 1; } /* Report potential error */ xmlSBufCleanup(&buf, ctxt, "AttValue length too long"); } else { if (chunkSize > 0) xmlSBufAddString(&buf, CUR_PTR - chunkSize, chunkSize); if ((normalize) && (inSpace) && (buf.size > 0)) { attvalFlags |= XML_ATTVAL_NORM_CHANGE; buf.size--; } ret = xmlSBufFinish(&buf, attlen, ctxt, "AttValue length too long"); attvalFlags |= XML_ATTVAL_ALLOC; if (ret != NULL) { if (attlen != NULL) *attlen = buf.size; } } if (outFlags != NULL) *outFlags = attvalFlags; NEXTL(1); return(ret); error: xmlSBufCleanup(&buf, ctxt, "AttValue length too long"); return(NULL); } /** * Parse a value for an attribute * Note: the parser won't do substitution of entities here, this * will be handled later in #xmlStringGetNodeList * * @deprecated Internal function, don't use. * * [10] AttValue ::= '"' ([^<&"] | Reference)* '"' | * "'" ([^<&'] | Reference)* "'" * * 3.3.3 Attribute-Value Normalization: * * Before the value of an attribute is passed to the application or * checked for validity, the XML processor must normalize it as follows: * * - a character reference is processed by appending the referenced * character to the attribute value * - an entity reference is processed by recursively processing the * replacement text of the entity * - a whitespace character (\#x20, \#xD, \#xA, \#x9) is processed by * appending \#x20 to the normalized value, except that only a single * \#x20 is appended for a "#xD#xA" sequence that is part of an external * parsed entity or the literal entity value of an internal parsed entity * - other characters are processed by appending them to the normalized value * * If the declared value is not CDATA, then the XML processor must further * process the normalized attribute value by discarding any leading and * trailing space (\#x20) characters, and by replacing sequences of space * (\#x20) characters by a single space (\#x20) character. * All attributes for which no declaration has been read should be treated * by a non-validating parser as if declared CDATA. * * @param ctxt an XML parser context * @returns the AttValue parsed or NULL. The value has to be freed by the * caller. */ xmlChar * xmlParseAttValue(xmlParserCtxt *ctxt) { if ((ctxt == NULL) || (ctxt->input == NULL)) return(NULL); return(xmlParseAttValueInternal(ctxt, NULL, NULL, 0, 0)); } /** * Parse an XML Literal * * @deprecated Internal function, don't use. * * [11] SystemLiteral ::= ('"' [^"]* '"') | ("'" [^']* "'") * * @param ctxt an XML parser context * @returns the SystemLiteral parsed or NULL */ xmlChar * xmlParseSystemLiteral(xmlParserCtxt *ctxt) { xmlChar *buf = NULL; int len = 0; int size = XML_PARSER_BUFFER_SIZE; int cur, l; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; xmlChar stop; if (RAW == '"') { NEXT; stop = '"'; } else if (RAW == '\'') { NEXT; stop = '\''; } else { xmlFatalErr(ctxt, XML_ERR_LITERAL_NOT_STARTED, NULL); return(NULL); } buf = xmlMalloc(size); if (buf == NULL) { xmlErrMemory(ctxt); return(NULL); } cur = xmlCurrentCharRecover(ctxt, &l); while ((IS_CHAR(cur)) && (cur != stop)) { /* checked */ if (len + 5 >= size) { xmlChar *tmp; int newSize; newSize = xmlGrowCapacity(size, 1, 1, maxLength); if (newSize < 0) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "SystemLiteral"); xmlFree(buf); return(NULL); } tmp = xmlRealloc(buf, newSize); if (tmp == NULL) { xmlFree(buf); xmlErrMemory(ctxt); return(NULL); } buf = tmp; size = newSize; } COPY_BUF(buf, len, cur); NEXTL(l); cur = xmlCurrentCharRecover(ctxt, &l); } buf[len] = 0; if (!IS_CHAR(cur)) { xmlFatalErr(ctxt, XML_ERR_LITERAL_NOT_FINISHED, NULL); } else { NEXT; } return(buf); } /** * Parse an XML public literal * * @deprecated Internal function, don't use. * * [12] PubidLiteral ::= '"' PubidChar* '"' | "'" (PubidChar - "'")* "'" * * @param ctxt an XML parser context * @returns the PubidLiteral parsed or NULL. */ xmlChar * xmlParsePubidLiteral(xmlParserCtxt *ctxt) { xmlChar *buf = NULL; int len = 0; int size = XML_PARSER_BUFFER_SIZE; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_TEXT_LENGTH : XML_MAX_NAME_LENGTH; xmlChar cur; xmlChar stop; if (RAW == '"') { NEXT; stop = '"'; } else if (RAW == '\'') { NEXT; stop = '\''; } else { xmlFatalErr(ctxt, XML_ERR_LITERAL_NOT_STARTED, NULL); return(NULL); } buf = xmlMalloc(size); if (buf == NULL) { xmlErrMemory(ctxt); return(NULL); } cur = CUR; while ((IS_PUBIDCHAR_CH(cur)) && (cur != stop) && (PARSER_STOPPED(ctxt) == 0)) { /* checked */ if (len + 1 >= size) { xmlChar *tmp; int newSize; newSize = xmlGrowCapacity(size, 1, 1, maxLength); if (newSize < 0) { xmlFatalErr(ctxt, XML_ERR_NAME_TOO_LONG, "Public ID"); xmlFree(buf); return(NULL); } tmp = xmlRealloc(buf, newSize); if (tmp == NULL) { xmlErrMemory(ctxt); xmlFree(buf); return(NULL); } buf = tmp; size = newSize; } buf[len++] = cur; NEXT; cur = CUR; } buf[len] = 0; if (cur != stop) { xmlFatalErr(ctxt, XML_ERR_LITERAL_NOT_FINISHED, NULL); } else { NEXTL(1); } return(buf); } static void xmlParseCharDataComplex(xmlParserCtxtPtr ctxt, int partial); /* * used for the test in the inner loop of the char data testing */ static const unsigned char test_char_data[256] = { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x9, CR/LF separated */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x00, 0x27, /* & */ 0x28, 0x29, 0x2A, 0x2B, 0x2C, 0x2D, 0x2E, 0x2F, 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3A, 0x3B, 0x00, 0x3D, 0x3E, 0x3F, /* < */ 0x40, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, 0x48, 0x49, 0x4A, 0x4B, 0x4C, 0x4D, 0x4E, 0x4F, 0x50, 0x51, 0x52, 0x53, 0x54, 0x55, 0x56, 0x57, 0x58, 0x59, 0x5A, 0x5B, 0x5C, 0x00, 0x5E, 0x5F, /* ] */ 0x60, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6A, 0x6B, 0x6C, 0x6D, 0x6E, 0x6F, 0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7A, 0x7B, 0x7C, 0x7D, 0x7E, 0x7F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* non-ascii */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 }; static void xmlCharacters(xmlParserCtxtPtr ctxt, const xmlChar *buf, int size, int isBlank) { int checkBlanks; if ((ctxt->sax == NULL) || (ctxt->disableSAX)) return; checkBlanks = (!ctxt->keepBlanks) || (ctxt->sax->ignorableWhitespace != ctxt->sax->characters); /* * Calling areBlanks with only parts of a text node * is fundamentally broken, making the NOBLANKS option * essentially unusable. */ if ((checkBlanks) && (areBlanks(ctxt, buf, size, isBlank))) { if ((ctxt->sax->ignorableWhitespace != NULL) && (ctxt->keepBlanks)) ctxt->sax->ignorableWhitespace(ctxt->userData, buf, size); } else { if (ctxt->sax->characters != NULL) ctxt->sax->characters(ctxt->userData, buf, size); /* * The old code used to update this value for "complex" data * even if checkBlanks was false. This was probably a bug. */ if ((checkBlanks) && (*ctxt->space == -1)) *ctxt->space = -2; } } /** * Parse character data. Always makes progress if the first char isn't * '<' or '&'. * * The right angle bracket (>) may be represented using the string ">", * and must, for compatibility, be escaped using ">" or a character * reference when it appears in the string "]]>" in content, when that * string is not marking the end of a CDATA section. * * [14] CharData ::= [^<&]* - ([^<&]* ']]>' [^<&]*) * @param ctxt an XML parser context * @param partial buffer may contain partial UTF-8 sequences */ static void xmlParseCharDataInternal(xmlParserCtxtPtr ctxt, int partial) { const xmlChar *in; int line = ctxt->input->line; int col = ctxt->input->col; int ccol; int terminate = 0; GROW; /* * Accelerated common case where input don't need to be * modified before passing it to the handler. */ in = ctxt->input->cur; do { get_more_space: while (*in == 0x20) { in++; ctxt->input->col++; } if (*in == 0xA) { do { ctxt->input->line++; ctxt->input->col = 1; in++; } while (*in == 0xA); goto get_more_space; } if (*in == '<') { while (in > ctxt->input->cur) { const xmlChar *tmp = ctxt->input->cur; size_t nbchar = in - tmp; if (nbchar > XML_MAX_ITEMS) nbchar = XML_MAX_ITEMS; ctxt->input->cur += nbchar; xmlCharacters(ctxt, tmp, nbchar, 1); } return; } get_more: ccol = ctxt->input->col; while (test_char_data[*in]) { in++; ccol++; } ctxt->input->col = ccol; if (*in == 0xA) { do { ctxt->input->line++; ctxt->input->col = 1; in++; } while (*in == 0xA); goto get_more; } if (*in == ']') { size_t avail = ctxt->input->end - in; if (partial && avail < 2) { terminate = 1; goto invoke_callback; } if (in[1] == ']') { if (partial && avail < 3) { terminate = 1; goto invoke_callback; } if (in[2] == '>') xmlFatalErr(ctxt, XML_ERR_MISPLACED_CDATA_END, NULL); } in++; ctxt->input->col++; goto get_more; } invoke_callback: while (in > ctxt->input->cur) { const xmlChar *tmp = ctxt->input->cur; size_t nbchar = in - tmp; if (nbchar > XML_MAX_ITEMS) nbchar = XML_MAX_ITEMS; ctxt->input->cur += nbchar; xmlCharacters(ctxt, tmp, nbchar, 0); line = ctxt->input->line; col = ctxt->input->col; } ctxt->input->cur = in; if (*in == 0xD) { in++; if (*in == 0xA) { ctxt->input->cur = in; in++; ctxt->input->line++; ctxt->input->col = 1; continue; /* while */ } in--; } if (*in == '<') { return; } if (*in == '&') { return; } if (terminate) { return; } SHRINK; GROW; in = ctxt->input->cur; } while (((*in >= 0x20) && (*in <= 0x7F)) || (*in == 0x09) || (*in == 0x0a)); ctxt->input->line = line; ctxt->input->col = col; xmlParseCharDataComplex(ctxt, partial); } /** * Always makes progress if the first char isn't '<' or '&'. * * parse a CharData section.this is the fallback function * of #xmlParseCharData when the parsing requires handling * of non-ASCII characters. * * @param ctxt an XML parser context * @param partial whether the input can end with truncated UTF-8 */ static void xmlParseCharDataComplex(xmlParserCtxtPtr ctxt, int partial) { xmlChar buf[XML_PARSER_BIG_BUFFER_SIZE + 5]; int nbchar = 0; int cur, l; cur = xmlCurrentCharRecover(ctxt, &l); while ((cur != '<') && /* checked */ (cur != '&') && (IS_CHAR(cur))) { if (cur == ']') { size_t avail = ctxt->input->end - ctxt->input->cur; if (partial && avail < 2) break; if (NXT(1) == ']') { if (partial && avail < 3) break; if (NXT(2) == '>') xmlFatalErr(ctxt, XML_ERR_MISPLACED_CDATA_END, NULL); } } COPY_BUF(buf, nbchar, cur); /* move current position before possible calling of ctxt->sax->characters */ NEXTL(l); if (nbchar >= XML_PARSER_BIG_BUFFER_SIZE) { buf[nbchar] = 0; xmlCharacters(ctxt, buf, nbchar, 0); nbchar = 0; SHRINK; } cur = xmlCurrentCharRecover(ctxt, &l); } if (nbchar != 0) { buf[nbchar] = 0; xmlCharacters(ctxt, buf, nbchar, 0); } /* * cur == 0 can mean * * - End of buffer. * - An actual 0 character. * - An incomplete UTF-8 sequence. This is allowed if partial is set. */ if (ctxt->input->cur < ctxt->input->end) { if ((cur == 0) && (CUR != 0)) { if (partial == 0) { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "Incomplete UTF-8 sequence starting with %02X\n", CUR); NEXTL(1); } } else if ((cur != '<') && (cur != '&') && (cur != ']')) { /* Generate the error and skip the offending character */ xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "PCDATA invalid Char value %d\n", cur); NEXTL(l); } } } /** * @deprecated Internal function, don't use. * @param ctxt an XML parser context * @param cdata unused */ void xmlParseCharData(xmlParserCtxt *ctxt, ATTRIBUTE_UNUSED int cdata) { xmlParseCharDataInternal(ctxt, 0); } /** * Parse an External ID or a Public ID * * @deprecated Internal function, don't use. * * NOTE: Productions [75] and [83] interact badly since [75] can generate * `'PUBLIC' S PubidLiteral S SystemLiteral` * * [75] ExternalID ::= 'SYSTEM' S SystemLiteral * | 'PUBLIC' S PubidLiteral S SystemLiteral * * [83] PublicID ::= 'PUBLIC' S PubidLiteral * * @param ctxt an XML parser context * @param publicId a xmlChar** receiving PubidLiteral * @param strict indicate whether we should restrict parsing to only * production [75], see NOTE below * @returns the function returns SystemLiteral and in the second * case publicID receives PubidLiteral, is strict is off * it is possible to return NULL and have publicID set. */ xmlChar * xmlParseExternalID(xmlParserCtxt *ctxt, xmlChar **publicId, int strict) { xmlChar *URI = NULL; *publicId = NULL; if (CMP6(CUR_PTR, 'S', 'Y', 'S', 'T', 'E', 'M')) { SKIP(6); if (SKIP_BLANKS == 0) { xmlFatalErrMsg(ctxt, XML_ERR_SPACE_REQUIRED, "Space required after 'SYSTEM'\n"); } URI = xmlParseSystemLiteral(ctxt); if (URI == NULL) { xmlFatalErr(ctxt, XML_ERR_URI_REQUIRED, NULL); } } else if (CMP6(CUR_PTR, 'P', 'U', 'B', 'L', 'I', 'C')) { SKIP(6); if (SKIP_BLANKS == 0) { xmlFatalErrMsg(ctxt, XML_ERR_SPACE_REQUIRED, "Space required after 'PUBLIC'\n"); } *publicId = xmlParsePubidLiteral(ctxt); if (*publicId == NULL) { xmlFatalErr(ctxt, XML_ERR_PUBID_REQUIRED, NULL); } if (strict) { /* * We don't handle [83] so "S SystemLiteral" is required. */ if (SKIP_BLANKS == 0) { xmlFatalErrMsg(ctxt, XML_ERR_SPACE_REQUIRED, "Space required after the Public Identifier\n"); } } else { /* * We handle [83] so we return immediately, if * "S SystemLiteral" is not detected. We skip blanks if no * system literal was found, but this is harmless since we must * be at the end of a NotationDecl. */ if (SKIP_BLANKS == 0) return(NULL); if ((CUR != '\'') && (CUR != '"')) return(NULL); } URI = xmlParseSystemLiteral(ctxt); if (URI == NULL) { xmlFatalErr(ctxt, XML_ERR_URI_REQUIRED, NULL); } } return(URI); } /** * Skip an XML (SGML) comment * The spec says that "For compatibility, the string "--" (double-hyphen) * must not occur within comments. " * This is the slow routine in case the accelerator for ascii didn't work * * [15] Comment ::= '' * @param ctxt an XML parser context * @param buf the already parsed part of the buffer * @param len number of bytes in the buffer * @param size allocated size of the buffer */ static void xmlParseCommentComplex(xmlParserCtxtPtr ctxt, xmlChar *buf, size_t len, size_t size) { int q, ql; int r, rl; int cur, l; int maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_HUGE_LENGTH : XML_MAX_TEXT_LENGTH; if (buf == NULL) { len = 0; size = XML_PARSER_BUFFER_SIZE; buf = xmlMalloc(size); if (buf == NULL) { xmlErrMemory(ctxt); return; } } q = xmlCurrentCharRecover(ctxt, &ql); if (q == 0) goto not_terminated; if (!IS_CHAR(q)) { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "xmlParseComment: invalid xmlChar value %d\n", q); xmlFree (buf); return; } NEXTL(ql); r = xmlCurrentCharRecover(ctxt, &rl); if (r == 0) goto not_terminated; if (!IS_CHAR(r)) { xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR, "xmlParseComment: invalid xmlChar value %d\n", r); xmlFree (buf); return; } NEXTL(rl); cur = xmlCurrentCharRecover(ctxt, &l); if (cur == 0) goto not_terminated; while (IS_CHAR(cur) && /* checked */ ((cur != '>') || (r != '-') || (q != '-'))) { if ((r == '-') && (q == '-')) { xmlFatalErr(ctxt, XML_ERR_HYPHEN_IN_COMMENT, NULL); } if (len + 5 >= size) { xmlChar *tmp; int newSize; newSize = xmlGrowCapacity(size, 1, 1, maxLength); if (newSize < 0) { xmlFatalErrMsgStr(ctxt, XML_ERR_COMMENT_NOT_FINISHED, "Comment too big found", NULL); xmlFree (buf); return; } tmp = xmlRealloc(buf, newSize); if (tmp == NULL) { xmlErrMemory(ctxt); xmlFree(buf); return; } buf = tmp; size = newSize; } COPY_BUF(buf, len, q); q = r; ql = rl; r = cur; rl = l; NEXTL(l); cur = xmlCurrentCharRecover(ctxt, &l); } buf[len] = 0; if (cur == 0) { xmlFatalErrMsgStr(ctxt, XML_ERR_COMMENT_NOT_FINISHED, "Comment not terminated \n' * @param ctxt an XML parser context */ void xmlParseComment(xmlParserCtxt *ctxt) { xmlChar *buf = NULL; size_t size = XML_PARSER_BUFFER_SIZE; size_t len = 0; size_t maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_HUGE_LENGTH : XML_MAX_TEXT_LENGTH; const xmlChar *in; size_t nbchar = 0; int ccol; /* * Check that there is a comment right here. */ if ((RAW != '<') || (NXT(1) != '!')) return; SKIP(2); if ((RAW != '-') || (NXT(1) != '-')) return; SKIP(2); GROW; /* * Accelerated common case where input don't need to be * modified before passing it to the handler. */ in = ctxt->input->cur; do { if (*in == 0xA) { do { ctxt->input->line++; ctxt->input->col = 1; in++; } while (*in == 0xA); } get_more: ccol = ctxt->input->col; while (((*in > '-') && (*in <= 0x7F)) || ((*in >= 0x20) && (*in < '-')) || (*in == 0x09)) { in++; ccol++; } ctxt->input->col = ccol; if (*in == 0xA) { do { ctxt->input->line++; ctxt->input->col = 1; in++; } while (*in == 0xA); goto get_more; } nbchar = in - ctxt->input->cur; /* * save current set of data */ if (nbchar > 0) { if (nbchar > maxLength - len) { xmlFatalErrMsgStr(ctxt, XML_ERR_COMMENT_NOT_FINISHED, "Comment too big found", NULL); xmlFree(buf); return; } if (buf == NULL) { if ((*in == '-') && (in[1] == '-')) size = nbchar + 1; else size = XML_PARSER_BUFFER_SIZE + nbchar; buf = xmlMalloc(size); if (buf == NULL) { xmlErrMemory(ctxt); return; } len = 0; } else if (len + nbchar + 1 >= size) { xmlChar *new_buf; size += len + nbchar + XML_PARSER_BUFFER_SIZE; new_buf = xmlRealloc(buf, size); if (new_buf == NULL) { xmlErrMemory(ctxt); xmlFree(buf); return; } buf = new_buf; } memcpy(&buf[len], ctxt->input->cur, nbchar); len += nbchar; buf[len] = 0; } ctxt->input->cur = in; if (*in == 0xA) { in++; ctxt->input->line++; ctxt->input->col = 1; } if (*in == 0xD) { in++; if (*in == 0xA) { ctxt->input->cur = in; in++; ctxt->input->line++; ctxt->input->col = 1; goto get_more; } in--; } SHRINK; GROW; in = ctxt->input->cur; if (*in == '-') { if (in[1] == '-') { if (in[2] == '>') { SKIP(3); if ((ctxt->sax != NULL) && (ctxt->sax->comment != NULL) && (!ctxt->disableSAX)) { if (buf != NULL) ctxt->sax->comment(ctxt->userData, buf); else ctxt->sax->comment(ctxt->userData, BAD_CAST ""); } if (buf != NULL) xmlFree(buf); return; } if (buf != NULL) { xmlFatalErrMsgStr(ctxt, XML_ERR_HYPHEN_IN_COMMENT, "Double hyphen within comment: " " as comment */ start = cur; continue; } } else if ((*cur == '"') || (*cur == '\'') || (*cur == ']')) { state = *cur; } cur++; } /* * Rescan the three last characters to detect "" * split across chunks. */ if ((state == 0) || (state == '-')) { if (cur - start < 3) cur = start; else cur -= 3; } index = cur - ctxt->input->cur; if (index > LONG_MAX) { ctxt->checkIndex = 0; ctxt->endCheckState = 0; return(1); } ctxt->checkIndex = index; ctxt->endCheckState = state; return(0); } /** * Try to progress on parsing * * @param ctxt an XML parser context * @param terminate last chunk indicator * @returns zero if no parsing was possible */ static int xmlParseTryOrFinish(xmlParserCtxtPtr ctxt, int terminate) { int ret = 0; size_t avail; xmlChar cur, next; if (ctxt->input == NULL) return(0); if ((ctxt->input != NULL) && (ctxt->input->cur - ctxt->input->base > 4096)) { xmlParserShrink(ctxt); } while (ctxt->disableSAX == 0) { avail = ctxt->input->end - ctxt->input->cur; if (avail < 1) goto done; switch (ctxt->instate) { case XML_PARSER_EOF: /* * Document parsing is done ! */ goto done; case XML_PARSER_START: /* * Very first chars read from the document flow. */ if ((!terminate) && (avail < 4)) goto done; /* * We need more bytes to detect EBCDIC code pages. * See xmlDetectEBCDIC. */ if ((CMP4(CUR_PTR, 0x4C, 0x6F, 0xA7, 0x94)) && (!terminate) && (avail < 200)) goto done; xmlDetectEncoding(ctxt); ctxt->instate = XML_PARSER_XML_DECL; break; case XML_PARSER_XML_DECL: if ((!terminate) && (avail < 2)) goto done; cur = ctxt->input->cur[0]; next = ctxt->input->cur[1]; if ((cur == '<') && (next == '?')) { /* PI or XML decl */ if ((!terminate) && (!xmlParseLookupString(ctxt, 2, "?>", 2))) goto done; if ((ctxt->input->cur[2] == 'x') && (ctxt->input->cur[3] == 'm') && (ctxt->input->cur[4] == 'l') && (IS_BLANK_CH(ctxt->input->cur[5]))) { ret += 5; xmlParseXMLDecl(ctxt); } else { ctxt->version = xmlCharStrdup(XML_DEFAULT_VERSION); if (ctxt->version == NULL) { xmlErrMemory(ctxt); break; } } } else { ctxt->version = xmlCharStrdup(XML_DEFAULT_VERSION); if (ctxt->version == NULL) { xmlErrMemory(ctxt); break; } } if ((ctxt->sax) && (ctxt->sax->setDocumentLocator)) { ctxt->sax->setDocumentLocator(ctxt->userData, (xmlSAXLocator *) &xmlDefaultSAXLocator); } if ((ctxt->sax) && (ctxt->sax->startDocument) && (!ctxt->disableSAX)) ctxt->sax->startDocument(ctxt->userData); ctxt->instate = XML_PARSER_MISC; break; case XML_PARSER_START_TAG: { const xmlChar *name; const xmlChar *prefix = NULL; const xmlChar *URI = NULL; int line = ctxt->input->line; int nbNs = 0; if ((!terminate) && (avail < 2)) goto done; cur = ctxt->input->cur[0]; if (cur != '<') { xmlFatalErrMsg(ctxt, XML_ERR_DOCUMENT_EMPTY, "Start tag expected, '<' not found"); ctxt->instate = XML_PARSER_EOF; xmlFinishDocument(ctxt); goto done; } if ((!terminate) && (!xmlParseLookupGt(ctxt))) goto done; if (ctxt->spaceNr == 0) spacePush(ctxt, -1); else if (*ctxt->space == -2) spacePush(ctxt, -1); else spacePush(ctxt, *ctxt->space); #ifdef LIBXML_SAX1_ENABLED if (ctxt->sax2) #endif /* LIBXML_SAX1_ENABLED */ name = xmlParseStartTag2(ctxt, &prefix, &URI, &nbNs); #ifdef LIBXML_SAX1_ENABLED else name = xmlParseStartTag(ctxt); #endif /* LIBXML_SAX1_ENABLED */ if (name == NULL) { spacePop(ctxt); ctxt->instate = XML_PARSER_EOF; xmlFinishDocument(ctxt); goto done; } #ifdef LIBXML_VALID_ENABLED /* * [ VC: Root Element Type ] * The Name in the document type declaration must match * the element type of the root element. */ if (ctxt->validate && ctxt->wellFormed && ctxt->myDoc && ctxt->node && (ctxt->node == ctxt->myDoc->children)) ctxt->valid &= xmlValidateRoot(&ctxt->vctxt, ctxt->myDoc); #endif /* LIBXML_VALID_ENABLED */ /* * Check for an Empty Element. */ if ((RAW == '/') && (NXT(1) == '>')) { SKIP(2); if (ctxt->sax2) { if ((ctxt->sax != NULL) && (ctxt->sax->endElementNs != NULL) && (!ctxt->disableSAX)) ctxt->sax->endElementNs(ctxt->userData, name, prefix, URI); if (nbNs > 0) xmlParserNsPop(ctxt, nbNs); #ifdef LIBXML_SAX1_ENABLED } else { if ((ctxt->sax != NULL) && (ctxt->sax->endElement != NULL) && (!ctxt->disableSAX)) ctxt->sax->endElement(ctxt->userData, name); #endif /* LIBXML_SAX1_ENABLED */ } spacePop(ctxt); } else if (RAW == '>') { NEXT; nameNsPush(ctxt, name, prefix, URI, line, nbNs); } else { xmlFatalErrMsgStr(ctxt, XML_ERR_GT_REQUIRED, "Couldn't find end of Start Tag %s\n", name); nodePop(ctxt); spacePop(ctxt); if (nbNs > 0) xmlParserNsPop(ctxt, nbNs); } if (ctxt->nameNr == 0) ctxt->instate = XML_PARSER_EPILOG; else ctxt->instate = XML_PARSER_CONTENT; break; } case XML_PARSER_CONTENT: { cur = ctxt->input->cur[0]; if (cur == '<') { if ((!terminate) && (avail < 2)) goto done; next = ctxt->input->cur[1]; if (next == '/') { ctxt->instate = XML_PARSER_END_TAG; break; } else if (next == '?') { if ((!terminate) && (!xmlParseLookupString(ctxt, 2, "?>", 2))) goto done; xmlParsePI(ctxt); ctxt->instate = XML_PARSER_CONTENT; break; } else if (next == '!') { if ((!terminate) && (avail < 3)) goto done; next = ctxt->input->cur[2]; if (next == '-') { if ((!terminate) && (avail < 4)) goto done; if (ctxt->input->cur[3] == '-') { if ((!terminate) && (!xmlParseLookupString(ctxt, 4, "-->", 3))) goto done; xmlParseComment(ctxt); ctxt->instate = XML_PARSER_CONTENT; break; } } else if (next == '[') { if ((!terminate) && (avail < 9)) goto done; if ((ctxt->input->cur[2] == '[') && (ctxt->input->cur[3] == 'C') && (ctxt->input->cur[4] == 'D') && (ctxt->input->cur[5] == 'A') && (ctxt->input->cur[6] == 'T') && (ctxt->input->cur[7] == 'A') && (ctxt->input->cur[8] == '[')) { if ((!terminate) && (!xmlParseLookupString(ctxt, 9, "]]>", 3))) goto done; ctxt->instate = XML_PARSER_CDATA_SECTION; xmlParseCDSect(ctxt); ctxt->instate = XML_PARSER_CONTENT; break; } } } } else if (cur == '&') { if ((!terminate) && (!xmlParseLookupChar(ctxt, ';'))) goto done; xmlParseReference(ctxt); break; } else { /* TODO Avoid the extra copy, handle directly !!! */ /* * Goal of the following test is: * - minimize calls to the SAX 'character' callback * when they are mergeable * - handle an problem for isBlank when we only parse * a sequence of blank chars and the next one is * not available to check against '<' presence. * - tries to homogenize the differences in SAX * callbacks between the push and pull versions * of the parser. */ if (avail < XML_PARSER_BIG_BUFFER_SIZE) { if ((!terminate) && (!xmlParseLookupCharData(ctxt))) goto done; } ctxt->checkIndex = 0; xmlParseCharDataInternal(ctxt, !terminate); break; } ctxt->instate = XML_PARSER_START_TAG; break; } case XML_PARSER_END_TAG: if ((!terminate) && (!xmlParseLookupChar(ctxt, '>'))) goto done; if (ctxt->sax2) { xmlParseEndTag2(ctxt, &ctxt->pushTab[ctxt->nameNr - 1]); nameNsPop(ctxt); } #ifdef LIBXML_SAX1_ENABLED else xmlParseEndTag1(ctxt, 0); #endif /* LIBXML_SAX1_ENABLED */ if (ctxt->nameNr == 0) { ctxt->instate = XML_PARSER_EPILOG; } else { ctxt->instate = XML_PARSER_CONTENT; } break; case XML_PARSER_MISC: case XML_PARSER_PROLOG: case XML_PARSER_EPILOG: SKIP_BLANKS; avail = ctxt->input->end - ctxt->input->cur; if (avail < 1) goto done; if (ctxt->input->cur[0] == '<') { if ((!terminate) && (avail < 2)) goto done; next = ctxt->input->cur[1]; if (next == '?') { if ((!terminate) && (!xmlParseLookupString(ctxt, 2, "?>", 2))) goto done; xmlParsePI(ctxt); break; } else if (next == '!') { if ((!terminate) && (avail < 3)) goto done; if (ctxt->input->cur[2] == '-') { if ((!terminate) && (avail < 4)) goto done; if (ctxt->input->cur[3] == '-') { if ((!terminate) && (!xmlParseLookupString(ctxt, 4, "-->", 3))) goto done; xmlParseComment(ctxt); break; } } else if (ctxt->instate == XML_PARSER_MISC) { if ((!terminate) && (avail < 9)) goto done; if ((ctxt->input->cur[2] == 'D') && (ctxt->input->cur[3] == 'O') && (ctxt->input->cur[4] == 'C') && (ctxt->input->cur[5] == 'T') && (ctxt->input->cur[6] == 'Y') && (ctxt->input->cur[7] == 'P') && (ctxt->input->cur[8] == 'E')) { if ((!terminate) && (!xmlParseLookupGt(ctxt))) goto done; ctxt->inSubset = 1; xmlParseDocTypeDecl(ctxt); if (RAW == '[') { ctxt->instate = XML_PARSER_DTD; } else { if (RAW == '>') NEXT; /* * Create and update the external subset. */ ctxt->inSubset = 2; if ((ctxt->sax != NULL) && (!ctxt->disableSAX) && (ctxt->sax->externalSubset != NULL)) ctxt->sax->externalSubset( ctxt->userData, ctxt->intSubName, ctxt->extSubSystem, ctxt->extSubURI); ctxt->inSubset = 0; xmlCleanSpecialAttr(ctxt); ctxt->instate = XML_PARSER_PROLOG; } break; } } } } if (ctxt->instate == XML_PARSER_EPILOG) { if (ctxt->errNo == XML_ERR_OK) xmlFatalErr(ctxt, XML_ERR_DOCUMENT_END, NULL); ctxt->instate = XML_PARSER_EOF; xmlFinishDocument(ctxt); } else { ctxt->instate = XML_PARSER_START_TAG; } break; case XML_PARSER_DTD: { if ((!terminate) && (!xmlParseLookupInternalSubset(ctxt))) goto done; xmlParseInternalSubset(ctxt); ctxt->inSubset = 2; if ((ctxt->sax != NULL) && (!ctxt->disableSAX) && (ctxt->sax->externalSubset != NULL)) ctxt->sax->externalSubset(ctxt->userData, ctxt->intSubName, ctxt->extSubSystem, ctxt->extSubURI); ctxt->inSubset = 0; xmlCleanSpecialAttr(ctxt); ctxt->instate = XML_PARSER_PROLOG; break; } default: xmlFatalErrMsg(ctxt, XML_ERR_INTERNAL_ERROR, "PP: internal error\n"); ctxt->instate = XML_PARSER_EOF; break; } } done: return(ret); } /** * Parse a chunk of memory in push parser mode. * * Assumes that the parser context was initialized with * #xmlCreatePushParserCtxt. * * The last chunk, which will often be empty, must be marked with * the `terminate` flag. With the default SAX callbacks, the resulting * document will be available in ctxt->myDoc. This pointer will not * be freed when calling #xmlFreeParserCtxt and must be freed by the * caller. If the document isn't well-formed, it will still be returned * in ctxt->myDoc. * * As an exception, #xmlCtxtResetPush will free the document in * ctxt->myDoc. So ctxt->myDoc should be set to NULL after extracting * the document. * * Since 2.14.0, #xmlCtxtGetDocument can be used to retrieve the * result document. * * @param ctxt an XML parser context * @param chunk chunk of memory * @param size size of chunk in bytes * @param terminate last chunk indicator * @returns an xmlParserErrors code (0 on success). */ int xmlParseChunk(xmlParserCtxt *ctxt, const char *chunk, int size, int terminate) { size_t curBase; size_t maxLength; size_t pos; int end_in_lf = 0; int res; if ((ctxt == NULL) || (size < 0)) return(XML_ERR_ARGUMENT); if ((chunk == NULL) && (size > 0)) return(XML_ERR_ARGUMENT); if ((ctxt->input == NULL) || (ctxt->input->buf == NULL)) return(XML_ERR_ARGUMENT); if (ctxt->disableSAX != 0) return(ctxt->errNo); ctxt->input->flags |= XML_INPUT_PROGRESSIVE; if (ctxt->instate == XML_PARSER_START) xmlCtxtInitializeLate(ctxt); if ((size > 0) && (chunk != NULL) && (!terminate) && (chunk[size - 1] == '\r')) { end_in_lf = 1; size--; } /* * Also push an empty chunk to make sure that the raw buffer * will be flushed if there is an encoder. */ pos = ctxt->input->cur - ctxt->input->base; res = xmlParserInputBufferPush(ctxt->input->buf, size, chunk); xmlBufUpdateInput(ctxt->input->buf->buffer, ctxt->input, pos); if (res < 0) { xmlCtxtErrIO(ctxt, ctxt->input->buf->error, NULL); return(ctxt->errNo); } xmlParseTryOrFinish(ctxt, terminate); curBase = ctxt->input->cur - ctxt->input->base; maxLength = (ctxt->options & XML_PARSE_HUGE) ? XML_MAX_HUGE_LENGTH : XML_MAX_LOOKUP_LIMIT; if (curBase > maxLength) { xmlFatalErr(ctxt, XML_ERR_RESOURCE_LIMIT, "Buffer size limit exceeded, try XML_PARSE_HUGE\n"); } if ((ctxt->errNo != XML_ERR_OK) && (ctxt->disableSAX != 0)) return(ctxt->errNo); if (end_in_lf == 1) { pos = ctxt->input->cur - ctxt->input->base; res = xmlParserInputBufferPush(ctxt->input->buf, 1, "\r"); xmlBufUpdateInput(ctxt->input->buf->buffer, ctxt->input, pos); if (res < 0) { xmlCtxtErrIO(ctxt, ctxt->input->buf->error, NULL); return(ctxt->errNo); } } if (terminate) { /* * Check for termination */ if ((ctxt->instate != XML_PARSER_EOF) && (ctxt->instate != XML_PARSER_EPILOG)) { if (ctxt->nameNr > 0) { const xmlChar *name = ctxt->nameTab[ctxt->nameNr - 1]; int line = ctxt->pushTab[ctxt->nameNr - 1].line; xmlFatalErrMsgStrIntStr(ctxt, XML_ERR_TAG_NOT_FINISHED, "Premature end of data in tag %s line %d\n", name, line, NULL); } else if (ctxt->instate == XML_PARSER_START) { xmlFatalErr(ctxt, XML_ERR_DOCUMENT_EMPTY, NULL); } else { xmlFatalErrMsg(ctxt, XML_ERR_DOCUMENT_EMPTY, "Start tag expected, '<' not found\n"); } } else { xmlParserCheckEOF(ctxt, XML_ERR_DOCUMENT_END); } if (ctxt->instate != XML_PARSER_EOF) { ctxt->instate = XML_PARSER_EOF; xmlFinishDocument(ctxt); } } if (ctxt->wellFormed == 0) return((xmlParserErrors) ctxt->errNo); else return(0); } /************************************************************************ * * * I/O front end functions to the parser * * * ************************************************************************/ /** * Create a parser context for using the XML parser in push mode. * See #xmlParseChunk. * * Passing an initial chunk is useless and deprecated. * * The push parser doesn't support recovery mode or the * XML_PARSE_NOBLANKS option. * * `filename` is used as base URI to fetch external entities and for * error reports. * * @param sax a SAX handler (optional) * @param user_data user data for SAX callbacks (optional) * @param chunk initial chunk (optional, deprecated) * @param size size of initial chunk in bytes * @param filename file name or URI (optional) * @returns the new parser context or NULL if a memory allocation * failed. */ xmlParserCtxt * xmlCreatePushParserCtxt(xmlSAXHandler *sax, void *user_data, const char *chunk, int size, const char *filename) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; ctxt = xmlNewSAXParserCtxt(sax, user_data); if (ctxt == NULL) return(NULL); ctxt->options &= ~XML_PARSE_NODICT; ctxt->dictNames = 1; input = xmlNewPushInput(filename, chunk, size); if (input == NULL) { xmlFreeParserCtxt(ctxt); return(NULL); } if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); xmlFreeParserCtxt(ctxt); return(NULL); } return(ctxt); } #endif /* LIBXML_PUSH_ENABLED */ /** * Blocks further parser processing * * @param ctxt an XML parser context */ void xmlStopParser(xmlParserCtxt *ctxt) { if (ctxt == NULL) return; /* This stops the parser */ ctxt->disableSAX = 2; /* * xmlStopParser is often called from error handlers, * so we can't raise an error here to avoid infinite * loops. Just make sure that an error condition is * reported. */ if (ctxt->errNo == XML_ERR_OK) { ctxt->errNo = XML_ERR_USER_STOP; ctxt->lastError.code = XML_ERR_USER_STOP; ctxt->wellFormed = 0; } } /** * Create a parser context for using the XML parser with an existing * I/O stream * * @param sax a SAX handler (optional) * @param user_data user data for SAX callbacks (optional) * @param ioread an I/O read function * @param ioclose an I/O close function (optional) * @param ioctx an I/O handler * @param enc the charset encoding if known (deprecated) * @returns the new parser context or NULL */ xmlParserCtxt * xmlCreateIOParserCtxt(xmlSAXHandler *sax, void *user_data, xmlInputReadCallback ioread, xmlInputCloseCallback ioclose, void *ioctx, xmlCharEncoding enc) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; const char *encoding; ctxt = xmlNewSAXParserCtxt(sax, user_data); if (ctxt == NULL) return(NULL); encoding = xmlGetCharEncodingName(enc); input = xmlCtxtNewInputFromIO(ctxt, NULL, ioread, ioclose, ioctx, encoding, 0); if (input == NULL) { xmlFreeParserCtxt(ctxt); return (NULL); } if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); xmlFreeParserCtxt(ctxt); return(NULL); } return(ctxt); } #ifdef LIBXML_VALID_ENABLED /************************************************************************ * * * Front ends when parsing a DTD * * * ************************************************************************/ /** * Parse a DTD. * * Option XML_PARSE_DTDLOAD should be enabled in the parser context * to make external entities work. * * @since 2.14.0 * * @param ctxt a parser context * @param input a parser input * @param publicId public ID of the DTD (optional) * @param systemId system ID of the DTD (optional) * @returns the resulting xmlDtd or NULL in case of error. * `input` will be freed by the function in any case. */ xmlDtd * xmlCtxtParseDtd(xmlParserCtxt *ctxt, xmlParserInput *input, const xmlChar *publicId, const xmlChar *systemId) { xmlDtdPtr ret = NULL; if ((ctxt == NULL) || (input == NULL)) { xmlFatalErr(ctxt, XML_ERR_ARGUMENT, NULL); xmlFreeInputStream(input); return(NULL); } if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); return(NULL); } if (publicId == NULL) publicId = BAD_CAST "none"; if (systemId == NULL) systemId = BAD_CAST "none"; ctxt->myDoc = xmlNewDoc(BAD_CAST "1.0"); if (ctxt->myDoc == NULL) { xmlErrMemory(ctxt); goto error; } ctxt->myDoc->properties = XML_DOC_INTERNAL; ctxt->myDoc->extSubset = xmlNewDtd(ctxt->myDoc, BAD_CAST "none", publicId, systemId); if (ctxt->myDoc->extSubset == NULL) { xmlErrMemory(ctxt); xmlFreeDoc(ctxt->myDoc); goto error; } xmlParseExternalSubset(ctxt, publicId, systemId); if (ctxt->wellFormed) { ret = ctxt->myDoc->extSubset; ctxt->myDoc->extSubset = NULL; if (ret != NULL) { xmlNodePtr tmp; ret->doc = NULL; tmp = ret->children; while (tmp != NULL) { tmp->doc = NULL; tmp = tmp->next; } } } else { ret = NULL; } xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL; error: xmlFreeInputStream(xmlCtxtPopInput(ctxt)); return(ret); } /** * Load and parse a DTD * * @deprecated Use #xmlCtxtParseDtd. * * @param sax the SAX handler block or NULL * @param input an Input Buffer * @param enc the charset encoding if known * @returns the resulting xmlDtd or NULL in case of error. * `input` will be freed by the function in any case. */ xmlDtd * xmlIOParseDTD(xmlSAXHandler *sax, xmlParserInputBuffer *input, xmlCharEncoding enc) { xmlDtdPtr ret = NULL; xmlParserCtxtPtr ctxt; xmlParserInputPtr pinput = NULL; if (input == NULL) return(NULL); ctxt = xmlNewSAXParserCtxt(sax, NULL); if (ctxt == NULL) { xmlFreeParserInputBuffer(input); return(NULL); } xmlCtxtSetOptions(ctxt, XML_PARSE_DTDLOAD); /* * generate a parser input from the I/O handler */ pinput = xmlNewIOInputStream(ctxt, input, XML_CHAR_ENCODING_NONE); if (pinput == NULL) { xmlFreeParserInputBuffer(input); xmlFreeParserCtxt(ctxt); return(NULL); } if (enc != XML_CHAR_ENCODING_NONE) { xmlSwitchEncoding(ctxt, enc); } ret = xmlCtxtParseDtd(ctxt, pinput, NULL, NULL); xmlFreeParserCtxt(ctxt); return(ret); } /** * Load and parse an external subset. * * @deprecated Use #xmlCtxtParseDtd. * * @param sax the SAX handler block * @param publicId public identifier of the DTD (optional) * @param systemId system identifier (URL) of the DTD * @returns the resulting xmlDtd or NULL in case of error. */ xmlDtd * xmlSAXParseDTD(xmlSAXHandler *sax, const xmlChar *publicId, const xmlChar *systemId) { xmlDtdPtr ret = NULL; xmlParserCtxtPtr ctxt; xmlParserInputPtr input = NULL; xmlChar* systemIdCanonic; if ((publicId == NULL) && (systemId == NULL)) return(NULL); ctxt = xmlNewSAXParserCtxt(sax, NULL); if (ctxt == NULL) { return(NULL); } xmlCtxtSetOptions(ctxt, XML_PARSE_DTDLOAD); /* * Canonicalise the system ID */ systemIdCanonic = xmlCanonicPath(systemId); if ((systemId != NULL) && (systemIdCanonic == NULL)) { xmlFreeParserCtxt(ctxt); return(NULL); } /* * Ask the Entity resolver to load the damn thing */ if ((ctxt->sax != NULL) && (ctxt->sax->resolveEntity != NULL)) input = ctxt->sax->resolveEntity(ctxt->userData, publicId, systemIdCanonic); if (input == NULL) { xmlFreeParserCtxt(ctxt); if (systemIdCanonic != NULL) xmlFree(systemIdCanonic); return(NULL); } if (input->filename == NULL) input->filename = (char *) systemIdCanonic; else xmlFree(systemIdCanonic); ret = xmlCtxtParseDtd(ctxt, input, publicId, systemId); xmlFreeParserCtxt(ctxt); return(ret); } /** * Load and parse an external subset. * * @param publicId public identifier of the DTD (optional) * @param systemId system identifier (URL) of the DTD * @returns the resulting xmlDtd or NULL in case of error. */ xmlDtd * xmlParseDTD(const xmlChar *publicId, const xmlChar *systemId) { return(xmlSAXParseDTD(NULL, publicId, systemId)); } #endif /* LIBXML_VALID_ENABLED */ /************************************************************************ * * * Front ends when parsing an Entity * * * ************************************************************************/ static xmlNodePtr xmlCtxtParseContentInternal(xmlParserCtxtPtr ctxt, xmlParserInputPtr input, int hasTextDecl, int buildTree) { xmlNodePtr root = NULL; xmlNodePtr list = NULL; xmlChar *rootName = BAD_CAST "#root"; int result; if (buildTree) { root = xmlNewDocNode(ctxt->myDoc, NULL, rootName, NULL); if (root == NULL) { xmlErrMemory(ctxt); goto error; } } if (xmlCtxtPushInput(ctxt, input) < 0) goto error; nameNsPush(ctxt, rootName, NULL, NULL, 0, 0); spacePush(ctxt, -1); if (buildTree) nodePush(ctxt, root); if (hasTextDecl) { xmlDetectEncoding(ctxt); /* * Parse a possible text declaration first */ if ((CMP5(CUR_PTR, '<', '?', 'x', 'm', 'l')) && (IS_BLANK_CH(NXT(5)))) { xmlParseTextDecl(ctxt); /* * An XML-1.0 document can't reference an entity not XML-1.0 */ if ((xmlStrEqual(ctxt->version, BAD_CAST "1.0")) && (!xmlStrEqual(ctxt->input->version, BAD_CAST "1.0"))) { xmlFatalErrMsg(ctxt, XML_ERR_VERSION_MISMATCH, "Version mismatch between document and " "entity\n"); } } } xmlParseContentInternal(ctxt); if (ctxt->input->cur < ctxt->input->end) xmlFatalErr(ctxt, XML_ERR_NOT_WELL_BALANCED, NULL); if ((ctxt->wellFormed) || ((ctxt->recovery) && (!xmlCtxtIsCatastrophicError(ctxt)))) { if (root != NULL) { xmlNodePtr cur; /* * Unlink newly created node list. */ list = root->children; root->children = NULL; root->last = NULL; for (cur = list; cur != NULL; cur = cur->next) cur->parent = NULL; } } /* * Read the rest of the stream in case of errors. We want * to account for the whole entity size. */ do { ctxt->input->cur = ctxt->input->end; xmlParserShrink(ctxt); result = xmlParserGrow(ctxt); } while (result > 0); if (buildTree) nodePop(ctxt); namePop(ctxt); spacePop(ctxt); xmlCtxtPopInput(ctxt); error: xmlFreeNode(root); return(list); } static void xmlCtxtParseEntity(xmlParserCtxtPtr ctxt, xmlEntityPtr ent) { xmlParserInputPtr input; xmlNodePtr list; unsigned long consumed; int isExternal; int buildTree; int oldMinNsIndex; int oldNodelen, oldNodemem; isExternal = (ent->etype == XML_EXTERNAL_GENERAL_PARSED_ENTITY); buildTree = (ctxt->node != NULL); /* * Recursion check */ if (ent->flags & XML_ENT_EXPANDING) { xmlFatalErr(ctxt, XML_ERR_ENTITY_LOOP, NULL); goto error; } /* * Load entity */ input = xmlNewEntityInputStream(ctxt, ent); if (input == NULL) goto error; /* * When building a tree, we need to limit the scope of namespace * declarations, so that entities don't reference xmlNs structs * from the parent of a reference. */ oldMinNsIndex = ctxt->nsdb->minNsIndex; if (buildTree) ctxt->nsdb->minNsIndex = ctxt->nsNr; oldNodelen = ctxt->nodelen; oldNodemem = ctxt->nodemem; ctxt->nodelen = 0; ctxt->nodemem = 0; /* * Parse content * * This initiates a recursive call chain: * * - xmlCtxtParseContentInternal * - xmlParseContentInternal * - xmlParseReference * - xmlCtxtParseEntity * * The nesting depth is limited by the maximum number of inputs, * see xmlCtxtPushInput. * * It's possible to make this non-recursive (minNsIndex must be * stored in the input struct) at the expense of code readability. */ ent->flags |= XML_ENT_EXPANDING; list = xmlCtxtParseContentInternal(ctxt, input, isExternal, buildTree); ent->flags &= ~XML_ENT_EXPANDING; ctxt->nsdb->minNsIndex = oldMinNsIndex; ctxt->nodelen = oldNodelen; ctxt->nodemem = oldNodemem; /* * Entity size accounting */ consumed = input->consumed; xmlSaturatedAddSizeT(&consumed, input->end - input->base); if ((ent->flags & XML_ENT_CHECKED) == 0) xmlSaturatedAdd(&ent->expandedSize, consumed); if ((ent->flags & XML_ENT_PARSED) == 0) { if (isExternal) xmlSaturatedAdd(&ctxt->sizeentities, consumed); ent->children = list; while (list != NULL) { list->parent = (xmlNodePtr) ent; /* * Downstream code like the nginx xslt module can set * ctxt->myDoc->extSubset to a separate DTD, so the entity * might have a different or a NULL document. */ if (list->doc != ent->doc) xmlSetTreeDoc(list, ent->doc); if (list->next == NULL) ent->last = list; list = list->next; } } else { xmlFreeNodeList(list); } xmlFreeInputStream(input); error: ent->flags |= XML_ENT_PARSED | XML_ENT_CHECKED; } /** * Parse an external general entity within an existing parsing context * An external general parsed entity is well-formed if it matches the * production labeled extParsedEnt. * * [78] extParsedEnt ::= TextDecl? content * * @param ctxt the existing parsing context * @param URL the URL for the entity to load * @param ID the System ID for the entity to load * @param listOut the return value for the set of parsed nodes * @returns 0 if the entity is well formed, -1 in case of args problem and * the parser error code otherwise */ int xmlParseCtxtExternalEntity(xmlParserCtxt *ctxt, const xmlChar *URL, const xmlChar *ID, xmlNode **listOut) { xmlParserInputPtr input; xmlNodePtr list; if (listOut != NULL) *listOut = NULL; if (ctxt == NULL) return(XML_ERR_ARGUMENT); input = xmlLoadResource(ctxt, (char *) URL, (char *) ID, XML_RESOURCE_GENERAL_ENTITY); if (input == NULL) return(ctxt->errNo); xmlCtxtInitializeLate(ctxt); list = xmlCtxtParseContentInternal(ctxt, input, /* hasTextDecl */ 1, 1); if (listOut != NULL) *listOut = list; else xmlFreeNodeList(list); xmlFreeInputStream(input); return(ctxt->errNo); } #ifdef LIBXML_SAX1_ENABLED /** * Parse an external general entity * An external general parsed entity is well-formed if it matches the * production labeled extParsedEnt. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlParseCtxtExternalEntity. * * [78] extParsedEnt ::= TextDecl? content * * @param doc the document the chunk pertains to * @param sax the SAX handler block (possibly NULL) * @param user_data The user data returned on SAX callbacks (possibly NULL) * @param depth Used for loop detection, use 0 * @param URL the URL for the entity to load * @param ID the System ID for the entity to load * @param list the return value for the set of parsed nodes * @returns 0 if the entity is well formed, -1 in case of args problem and * the parser error code otherwise */ int xmlParseExternalEntity(xmlDoc *doc, xmlSAXHandler *sax, void *user_data, int depth, const xmlChar *URL, const xmlChar *ID, xmlNode **list) { xmlParserCtxtPtr ctxt; int ret; if (list != NULL) *list = NULL; if (doc == NULL) return(XML_ERR_ARGUMENT); ctxt = xmlNewSAXParserCtxt(sax, user_data); if (ctxt == NULL) return(XML_ERR_NO_MEMORY); ctxt->depth = depth; ctxt->myDoc = doc; ret = xmlParseCtxtExternalEntity(ctxt, URL, ID, list); xmlFreeParserCtxt(ctxt); return(ret); } /** * Parse a well-balanced chunk of an XML document * called by the parser * The allowed sequence for the Well Balanced Chunk is the one defined by * the content production in the XML grammar: * * [43] content ::= (element | CharData | Reference | CDSect | PI | * Comment)* * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @param doc the document the chunk pertains to (must not be NULL) * @param sax the SAX handler block (possibly NULL) * @param user_data The user data returned on SAX callbacks (possibly NULL) * @param depth Used for loop detection, use 0 * @param string the input string in UTF8 or ISO-Latin (zero terminated) * @param lst the return value for the set of parsed nodes * @returns 0 if the chunk is well balanced, -1 in case of args problem and * the parser error code otherwise */ int xmlParseBalancedChunkMemory(xmlDoc *doc, xmlSAXHandler *sax, void *user_data, int depth, const xmlChar *string, xmlNode **lst) { return xmlParseBalancedChunkMemoryRecover( doc, sax, user_data, depth, string, lst, 0 ); } #endif /* LIBXML_SAX1_ENABLED */ /** * Parse a well-balanced chunk of XML matching the 'content' production. * * Namespaces in scope of `node` and entities of `node`'s document are * recognized. When validating, the DTD of `node`'s document is used. * * Always consumes `input` even in error case. * * @since 2.14.0 * * @param ctxt parser context * @param input parser input * @param node target node or document * @param hasTextDecl whether to parse text declaration * @returns a node list or NULL in case of error. */ xmlNode * xmlCtxtParseContent(xmlParserCtxt *ctxt, xmlParserInput *input, xmlNode *node, int hasTextDecl) { xmlDocPtr doc; xmlNodePtr cur, list = NULL; int nsnr = 0; xmlDictPtr oldDict; int oldOptions, oldDictNames, oldLoadSubset; if ((ctxt == NULL) || (input == NULL) || (node == NULL)) { xmlFatalErr(ctxt, XML_ERR_ARGUMENT, NULL); goto exit; } doc = node->doc; if (doc == NULL) { xmlFatalErr(ctxt, XML_ERR_ARGUMENT, NULL); goto exit; } switch (node->type) { case XML_ELEMENT_NODE: case XML_DOCUMENT_NODE: case XML_HTML_DOCUMENT_NODE: break; case XML_ATTRIBUTE_NODE: case XML_TEXT_NODE: case XML_CDATA_SECTION_NODE: case XML_ENTITY_REF_NODE: case XML_PI_NODE: case XML_COMMENT_NODE: for (cur = node->parent; cur != NULL; cur = cur->parent) { if ((cur->type == XML_ELEMENT_NODE) || (cur->type == XML_DOCUMENT_NODE) || (cur->type == XML_HTML_DOCUMENT_NODE)) { node = cur; break; } } break; default: xmlFatalErr(ctxt, XML_ERR_ARGUMENT, NULL); goto exit; } xmlCtxtReset(ctxt); oldDict = ctxt->dict; oldOptions = ctxt->options; oldDictNames = ctxt->dictNames; oldLoadSubset = ctxt->loadsubset; /* * Use input doc's dict if present, else assure XML_PARSE_NODICT is set. */ if (doc->dict != NULL) { ctxt->dict = doc->dict; } else { ctxt->options |= XML_PARSE_NODICT; ctxt->dictNames = 0; } /* * Disable IDs */ ctxt->loadsubset |= XML_SKIP_IDS; ctxt->options |= XML_PARSE_SKIP_IDS; ctxt->myDoc = doc; #ifdef LIBXML_HTML_ENABLED if (ctxt->html) { /* * When parsing in context, it makes no sense to add implied * elements like html/body/etc... */ ctxt->options |= HTML_PARSE_NOIMPLIED; list = htmlCtxtParseContentInternal(ctxt, input); } else #endif { xmlCtxtInitializeLate(ctxt); /* * initialize the SAX2 namespaces stack */ cur = node; while ((cur != NULL) && (cur->type == XML_ELEMENT_NODE)) { xmlNsPtr ns = cur->nsDef; xmlHashedString hprefix, huri; while (ns != NULL) { hprefix = xmlDictLookupHashed(ctxt->dict, ns->prefix, -1); huri = xmlDictLookupHashed(ctxt->dict, ns->href, -1); if (xmlParserNsPush(ctxt, &hprefix, &huri, ns, 1) > 0) nsnr++; ns = ns->next; } cur = cur->parent; } list = xmlCtxtParseContentInternal(ctxt, input, hasTextDecl, 1); if (nsnr > 0) xmlParserNsPop(ctxt, nsnr); } ctxt->dict = oldDict; ctxt->options = oldOptions; ctxt->dictNames = oldDictNames; ctxt->loadsubset = oldLoadSubset; ctxt->myDoc = NULL; ctxt->node = NULL; exit: xmlFreeInputStream(input); return(list); } /** * Parse a well-balanced chunk of an XML document * within the context (DTD, namespaces, etc ...) of the given node. * * The allowed sequence for the data is a Well Balanced Chunk defined by * the content production in the XML grammar: * * [43] content ::= (element | CharData | Reference | CDSect | PI | * Comment)* * * This function assumes the encoding of `node`'s document which is * typically not what you want. A better alternative is * #xmlCtxtParseContent. * * @param node the context node * @param data the input string * @param datalen the input string length in bytes * @param options a combination of xmlParserOption * @param listOut the return value for the set of parsed nodes * @returns XML_ERR_OK if the chunk is well balanced, and the parser * error code otherwise */ xmlParserErrors xmlParseInNodeContext(xmlNode *node, const char *data, int datalen, int options, xmlNode **listOut) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlDocPtr doc; xmlNodePtr list; xmlParserErrors ret; if (listOut == NULL) return(XML_ERR_INTERNAL_ERROR); *listOut = NULL; if ((node == NULL) || (data == NULL) || (datalen < 0)) return(XML_ERR_INTERNAL_ERROR); doc = node->doc; if (doc == NULL) return(XML_ERR_INTERNAL_ERROR); #ifdef LIBXML_HTML_ENABLED if (doc->type == XML_HTML_DOCUMENT_NODE) { ctxt = htmlNewParserCtxt(); } else #endif ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(XML_ERR_NO_MEMORY); input = xmlCtxtNewInputFromMemory(ctxt, NULL, data, datalen, (const char *) doc->encoding, XML_INPUT_BUF_STATIC); if (input == NULL) { xmlFreeParserCtxt(ctxt); return(XML_ERR_NO_MEMORY); } xmlCtxtUseOptions(ctxt, options); list = xmlCtxtParseContent(ctxt, input, node, /* hasTextDecl */ 0); if (list == NULL) { ret = ctxt->errNo; if (ret == XML_ERR_ARGUMENT) ret = XML_ERR_INTERNAL_ERROR; } else { ret = XML_ERR_OK; *listOut = list; } xmlFreeParserCtxt(ctxt); return(ret); } #ifdef LIBXML_SAX1_ENABLED /** * Parse a well-balanced chunk of an XML document * * The allowed sequence for the Well Balanced Chunk is the one defined by * the content production in the XML grammar: * * [43] content ::= (element | CharData | Reference | CDSect | PI | * Comment)* * * In case recover is set to 1, the nodelist will not be empty even if * the parsed chunk is not well balanced, assuming the parsing succeeded to * some extent. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @param doc the document the chunk pertains to (must not be NULL) * @param sax the SAX handler block (possibly NULL) * @param user_data The user data returned on SAX callbacks (possibly NULL) * @param depth Used for loop detection, use 0 * @param string the input string in UTF8 or ISO-Latin (zero terminated) * @param listOut the return value for the set of parsed nodes * @param recover return nodes even if the data is broken (use 0) * @returns 0 if the chunk is well balanced, or thehe parser error code * otherwise. */ int xmlParseBalancedChunkMemoryRecover(xmlDoc *doc, xmlSAXHandler *sax, void *user_data, int depth, const xmlChar *string, xmlNode **listOut, int recover) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlNodePtr list; int ret; if (listOut != NULL) *listOut = NULL; if (string == NULL) return(XML_ERR_ARGUMENT); ctxt = xmlNewSAXParserCtxt(sax, user_data); if (ctxt == NULL) return(XML_ERR_NO_MEMORY); xmlCtxtInitializeLate(ctxt); ctxt->depth = depth; ctxt->myDoc = doc; if (recover) { ctxt->options |= XML_PARSE_RECOVER; ctxt->recovery = 1; } input = xmlNewStringInputStream(ctxt, string); if (input == NULL) { ret = ctxt->errNo; goto error; } list = xmlCtxtParseContentInternal(ctxt, input, /* hasTextDecl */ 0, 1); if (listOut != NULL) *listOut = list; else xmlFreeNodeList(list); if (!ctxt->wellFormed) ret = ctxt->errNo; else ret = XML_ERR_OK; error: xmlFreeInputStream(input); xmlFreeParserCtxt(ctxt); return(ret); } /** * Parse an XML external entity out of context and build a tree. * It use the given SAX function block to handle the parsing callback. * If sax is NULL, fallback to the default DOM tree building routines. * * @deprecated Don't use. * * [78] extParsedEnt ::= TextDecl? content * * This correspond to a "Well Balanced" chunk * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @param sax the SAX handler block * @param filename the filename * @returns the resulting document tree */ xmlDoc * xmlSAXParseEntity(xmlSAXHandler *sax, const char *filename) { xmlDocPtr ret; xmlParserCtxtPtr ctxt; ctxt = xmlCreateFileParserCtxt(filename); if (ctxt == NULL) { return(NULL); } if (sax != NULL) { if (sax->initialized == XML_SAX2_MAGIC) { *ctxt->sax = *sax; } else { memset(ctxt->sax, 0, sizeof(*ctxt->sax)); memcpy(ctxt->sax, sax, sizeof(xmlSAXHandlerV1)); } ctxt->userData = NULL; } xmlParseExtParsedEnt(ctxt); if (ctxt->wellFormed) { ret = ctxt->myDoc; } else { ret = NULL; xmlFreeDoc(ctxt->myDoc); } xmlFreeParserCtxt(ctxt); return(ret); } /** * Parse an XML external entity out of context and build a tree. * * [78] extParsedEnt ::= TextDecl? content * * This correspond to a "Well Balanced" chunk * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Don't use. * * @param filename the filename * @returns the resulting document tree */ xmlDoc * xmlParseEntity(const char *filename) { return(xmlSAXParseEntity(NULL, filename)); } #endif /* LIBXML_SAX1_ENABLED */ /** * Create a parser context for an external entity * Automatic support for ZLIB/Compress compressed document is provided * by default if found at compile-time. * * @deprecated Don't use. * * @param URL the entity URL * @param ID the entity PUBLIC ID * @param base a possible base for the target URI * @returns the new parser context or NULL */ xmlParserCtxt * xmlCreateEntityParserCtxt(const xmlChar *URL, const xmlChar *ID, const xmlChar *base) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlChar *uri = NULL; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); if (base != NULL) { if (xmlBuildURISafe(URL, base, &uri) < 0) goto error; if (uri != NULL) URL = uri; } input = xmlLoadResource(ctxt, (char *) URL, (char *) ID, XML_RESOURCE_UNKNOWN); if (input == NULL) goto error; if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); goto error; } xmlFree(uri); return(ctxt); error: xmlFree(uri); xmlFreeParserCtxt(ctxt); return(NULL); } /************************************************************************ * * * Front ends when parsing from a file * * * ************************************************************************/ /** * Create a parser context for a file or URL content. * Automatic support for ZLIB/Compress compressed document is provided * by default if found at compile-time and for file accesses * * @deprecated Use #xmlNewParserCtxt and #xmlCtxtReadFile. * * @param filename the filename or URL * @param options a combination of xmlParserOption * @returns the new parser context or NULL */ xmlParserCtxt * xmlCreateURLParserCtxt(const char *filename, int options) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); xmlCtxtUseOptions(ctxt, options); input = xmlLoadResource(ctxt, filename, NULL, XML_RESOURCE_MAIN_DOCUMENT); if (input == NULL) { xmlFreeParserCtxt(ctxt); return(NULL); } if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); xmlFreeParserCtxt(ctxt); return(NULL); } return(ctxt); } /** * Create a parser context for a file content. * Automatic support for ZLIB/Compress compressed document is provided * by default if found at compile-time. * * @deprecated Use #xmlNewParserCtxt and #xmlCtxtReadFile. * * @param filename the filename * @returns the new parser context or NULL */ xmlParserCtxt * xmlCreateFileParserCtxt(const char *filename) { return(xmlCreateURLParserCtxt(filename, 0)); } #ifdef LIBXML_SAX1_ENABLED /** * Parse an XML file and build a tree. Automatic support for ZLIB/Compress * compressed document is provided by default if found at compile-time. * It use the given SAX function block to handle the parsing callback. * If sax is NULL, fallback to the default DOM tree building routines. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadFile. * * User data (void *) is stored within the parser context in the * context's _private member, so it is available nearly everywhere in libxml * * @param sax the SAX handler block * @param filename the filename * @param recovery work in recovery mode, i.e. tries to read no Well Formed * documents * @param data the userdata * @returns the resulting document tree */ xmlDoc * xmlSAXParseFileWithData(xmlSAXHandler *sax, const char *filename, int recovery, void *data) { xmlDocPtr ret = NULL; xmlParserCtxtPtr ctxt; xmlParserInputPtr input; ctxt = xmlNewSAXParserCtxt(sax, NULL); if (ctxt == NULL) return(NULL); if (data != NULL) ctxt->_private = data; if (recovery) { ctxt->options |= XML_PARSE_RECOVER; ctxt->recovery = 1; } if ((filename != NULL) && (filename[0] == '-') && (filename[1] == 0)) input = xmlCtxtNewInputFromFd(ctxt, filename, STDIN_FILENO, NULL, 0); else input = xmlCtxtNewInputFromUrl(ctxt, filename, NULL, NULL, 0); if (input != NULL) ret = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(ret); } /** * Parse an XML file and build a tree. Automatic support for ZLIB/Compress * compressed document is provided by default if found at compile-time. * It use the given SAX function block to handle the parsing callback. * If sax is NULL, fallback to the default DOM tree building routines. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadFile. * * @param sax the SAX handler block * @param filename the filename * @param recovery work in recovery mode, i.e. tries to read no Well Formed * documents * @returns the resulting document tree */ xmlDoc * xmlSAXParseFile(xmlSAXHandler *sax, const char *filename, int recovery) { return(xmlSAXParseFileWithData(sax,filename,recovery,NULL)); } /** * Parse an XML in-memory document and build a tree. * In the case the document is not Well Formed, a attempt to build a * tree is tried anyway * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT | XML_PARSE_RECOVER. * * @deprecated Use #xmlReadDoc with XML_PARSE_RECOVER. * * @param cur a pointer to an array of xmlChar * @returns the resulting document tree or NULL in case of failure */ xmlDoc * xmlRecoverDoc(const xmlChar *cur) { return(xmlSAXParseDoc(NULL, cur, 1)); } /** * Parse an XML file and build a tree. Automatic support for ZLIB/Compress * compressed document is provided by default if found at compile-time. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlReadFile. * * @param filename the filename * @returns the resulting document tree if the file was wellformed, * NULL otherwise. */ xmlDoc * xmlParseFile(const char *filename) { return(xmlSAXParseFile(NULL, filename, 0)); } /** * Parse an XML file and build a tree. Automatic support for ZLIB/Compress * compressed document is provided by default if found at compile-time. * In the case the document is not Well Formed, it attempts to build * a tree anyway * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT | XML_PARSE_RECOVER. * * @deprecated Use #xmlReadFile with XML_PARSE_RECOVER. * * @param filename the filename * @returns the resulting document tree or NULL in case of failure */ xmlDoc * xmlRecoverFile(const char *filename) { return(xmlSAXParseFile(NULL, filename, 1)); } /** * Setup the parser context to parse a new buffer; Clears any prior * contents from the parser context. The buffer parameter must not be * NULL, but the filename parameter can be * * @deprecated Don't use. * * @param ctxt an XML parser context * @param buffer a xmlChar * buffer * @param filename a file name */ void xmlSetupParserForBuffer(xmlParserCtxt *ctxt, const xmlChar* buffer, const char* filename) { xmlParserInputPtr input; if ((ctxt == NULL) || (buffer == NULL)) return; xmlCtxtReset(ctxt); input = xmlCtxtNewInputFromString(ctxt, filename, (const char *) buffer, NULL, 0); if (input == NULL) return; if (xmlCtxtPushInput(ctxt, input) < 0) xmlFreeInputStream(input); } /** * Parse an XML file and call the given SAX handler routines. * Automatic support for ZLIB/Compress compressed document is provided * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadFile. * * @param sax a SAX handler * @param user_data The user data returned on SAX callbacks * @param filename a file name * @returns 0 in case of success or a error number otherwise */ int xmlSAXUserParseFile(xmlSAXHandler *sax, void *user_data, const char *filename) { int ret = 0; xmlParserCtxtPtr ctxt; ctxt = xmlCreateFileParserCtxt(filename); if (ctxt == NULL) return -1; if (sax != NULL) { if (sax->initialized == XML_SAX2_MAGIC) { *ctxt->sax = *sax; } else { memset(ctxt->sax, 0, sizeof(*ctxt->sax)); memcpy(ctxt->sax, sax, sizeof(xmlSAXHandlerV1)); } ctxt->userData = user_data; } xmlParseDocument(ctxt); if (ctxt->wellFormed) ret = 0; else { if (ctxt->errNo != 0) ret = ctxt->errNo; else ret = -1; } if (ctxt->myDoc != NULL) { xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL; } xmlFreeParserCtxt(ctxt); return ret; } #endif /* LIBXML_SAX1_ENABLED */ /************************************************************************ * * * Front ends when parsing from memory * * * ************************************************************************/ /** * Create a parser context for an XML in-memory document. The input buffer * must not contain a terminating null byte. * * @param buffer a pointer to a char array * @param size the size of the array * @returns the new parser context or NULL */ xmlParserCtxt * xmlCreateMemoryParserCtxt(const char *buffer, int size) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; if (size < 0) return(NULL); ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); input = xmlCtxtNewInputFromMemory(ctxt, NULL, buffer, size, NULL, 0); if (input == NULL) { xmlFreeParserCtxt(ctxt); return(NULL); } if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); xmlFreeParserCtxt(ctxt); return(NULL); } return(ctxt); } #ifdef LIBXML_SAX1_ENABLED /** * Parse an XML in-memory block and use the given SAX function block * to handle the parsing callback. If sax is NULL, fallback to the default * DOM tree building routines. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadMemory. * * User data (void *) is stored within the parser context in the * context's _private member, so it is available nearly everywhere in libxml * * @param sax the SAX handler block * @param buffer an pointer to a char array * @param size the size of the array * @param recovery work in recovery mode, i.e. tries to read no Well Formed * documents * @param data the userdata * @returns the resulting document tree */ xmlDoc * xmlSAXParseMemoryWithData(xmlSAXHandler *sax, const char *buffer, int size, int recovery, void *data) { xmlDocPtr ret = NULL; xmlParserCtxtPtr ctxt; xmlParserInputPtr input; if (size < 0) return(NULL); ctxt = xmlNewSAXParserCtxt(sax, NULL); if (ctxt == NULL) return(NULL); if (data != NULL) ctxt->_private=data; if (recovery) { ctxt->options |= XML_PARSE_RECOVER; ctxt->recovery = 1; } input = xmlCtxtNewInputFromMemory(ctxt, NULL, buffer, size, NULL, XML_INPUT_BUF_STATIC); if (input != NULL) ret = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(ret); } /** * Parse an XML in-memory block and use the given SAX function block * to handle the parsing callback. If sax is NULL, fallback to the default * DOM tree building routines. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadMemory. * * @param sax the SAX handler block * @param buffer an pointer to a char array * @param size the size of the array * @param recovery work in recovery mode, i.e. tries to read not Well Formed * documents * @returns the resulting document tree */ xmlDoc * xmlSAXParseMemory(xmlSAXHandler *sax, const char *buffer, int size, int recovery) { return xmlSAXParseMemoryWithData(sax, buffer, size, recovery, NULL); } /** * Parse an XML in-memory block and build a tree. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlReadMemory. * * @param buffer an pointer to a char array * @param size the size of the array * @returns the resulting document tree */ xmlDoc *xmlParseMemory(const char *buffer, int size) { return(xmlSAXParseMemory(NULL, buffer, size, 0)); } /** * Parse an XML in-memory block and build a tree. * In the case the document is not Well Formed, an attempt to * build a tree is tried anyway * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT | XML_PARSE_RECOVER. * * @deprecated Use #xmlReadMemory with XML_PARSE_RECOVER. * * @param buffer an pointer to a char array * @param size the size of the array * @returns the resulting document tree or NULL in case of error */ xmlDoc *xmlRecoverMemory(const char *buffer, int size) { return(xmlSAXParseMemory(NULL, buffer, size, 1)); } /** * Parse an XML in-memory buffer and call the given SAX handler routines. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadMemory. * * @param sax a SAX handler * @param user_data The user data returned on SAX callbacks * @param buffer an in-memory XML document input * @param size the length of the XML document in bytes * @returns 0 in case of success or a error number otherwise */ int xmlSAXUserParseMemory(xmlSAXHandler *sax, void *user_data, const char *buffer, int size) { int ret = 0; xmlParserCtxtPtr ctxt; ctxt = xmlCreateMemoryParserCtxt(buffer, size); if (ctxt == NULL) return -1; if (sax != NULL) { if (sax->initialized == XML_SAX2_MAGIC) { *ctxt->sax = *sax; } else { memset(ctxt->sax, 0, sizeof(*ctxt->sax)); memcpy(ctxt->sax, sax, sizeof(xmlSAXHandlerV1)); } ctxt->userData = user_data; } xmlParseDocument(ctxt); if (ctxt->wellFormed) ret = 0; else { if (ctxt->errNo != 0) ret = ctxt->errNo; else ret = -1; } if (ctxt->myDoc != NULL) { xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL; } xmlFreeParserCtxt(ctxt); return ret; } #endif /* LIBXML_SAX1_ENABLED */ /** * Creates a parser context for an XML in-memory document. * * @param str a pointer to an array of xmlChar * @returns the new parser context or NULL */ xmlParserCtxt * xmlCreateDocParserCtxt(const xmlChar *str) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); input = xmlCtxtNewInputFromString(ctxt, NULL, (const char *) str, NULL, 0); if (input == NULL) { xmlFreeParserCtxt(ctxt); return(NULL); } if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); xmlFreeParserCtxt(ctxt); return(NULL); } return(ctxt); } #ifdef LIBXML_SAX1_ENABLED /** * Parse an XML in-memory document and build a tree. * It use the given SAX function block to handle the parsing callback. * If sax is NULL, fallback to the default DOM tree building routines. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlNewSAXParserCtxt and #xmlCtxtReadDoc. * * @param sax the SAX handler block * @param cur a pointer to an array of xmlChar * @param recovery work in recovery mode, i.e. tries to read no Well Formed * documents * @returns the resulting document tree */ xmlDoc * xmlSAXParseDoc(xmlSAXHandler *sax, const xmlChar *cur, int recovery) { xmlDocPtr ret; xmlParserCtxtPtr ctxt; xmlSAXHandlerPtr oldsax = NULL; if (cur == NULL) return(NULL); ctxt = xmlCreateDocParserCtxt(cur); if (ctxt == NULL) return(NULL); if (sax != NULL) { oldsax = ctxt->sax; ctxt->sax = sax; ctxt->userData = NULL; } xmlParseDocument(ctxt); if ((ctxt->wellFormed) || recovery) ret = ctxt->myDoc; else { ret = NULL; xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL; } if (sax != NULL) ctxt->sax = oldsax; xmlFreeParserCtxt(ctxt); return(ret); } /** * Parse an XML in-memory document and build a tree. * * This function uses deprecated global variables to set parser options * which default to XML_PARSE_NODICT. * * @deprecated Use #xmlReadDoc. * * @param cur a pointer to an array of xmlChar * @returns the resulting document tree */ xmlDoc * xmlParseDoc(const xmlChar *cur) { return(xmlSAXParseDoc(NULL, cur, 0)); } #endif /* LIBXML_SAX1_ENABLED */ /************************************************************************ * * * New set (2.6.0) of simpler and more flexible APIs * * * ************************************************************************/ /** * Reset a parser context * * @param ctxt an XML parser context */ void xmlCtxtReset(xmlParserCtxt *ctxt) { xmlParserInputPtr input; if (ctxt == NULL) return; while ((input = xmlCtxtPopInput(ctxt)) != NULL) { /* Non consuming */ xmlFreeInputStream(input); } ctxt->inputNr = 0; ctxt->input = NULL; ctxt->spaceNr = 0; if (ctxt->spaceTab != NULL) { ctxt->spaceTab[0] = -1; ctxt->space = &ctxt->spaceTab[0]; } else { ctxt->space = NULL; } ctxt->nodeNr = 0; ctxt->node = NULL; ctxt->nameNr = 0; ctxt->name = NULL; ctxt->nsNr = 0; xmlParserNsReset(ctxt->nsdb); if (ctxt->version != NULL) { xmlFree(ctxt->version); ctxt->version = NULL; } if (ctxt->encoding != NULL) { xmlFree(ctxt->encoding); ctxt->encoding = NULL; } if (ctxt->extSubURI != NULL) { xmlFree(ctxt->extSubURI); ctxt->extSubURI = NULL; } if (ctxt->extSubSystem != NULL) { xmlFree(ctxt->extSubSystem); ctxt->extSubSystem = NULL; } if (ctxt->directory != NULL) { xmlFree(ctxt->directory); ctxt->directory = NULL; } if (ctxt->myDoc != NULL) xmlFreeDoc(ctxt->myDoc); ctxt->myDoc = NULL; ctxt->standalone = -1; ctxt->hasExternalSubset = 0; ctxt->hasPErefs = 0; ctxt->html = ctxt->html ? 1 : 0; ctxt->instate = XML_PARSER_START; ctxt->wellFormed = 1; ctxt->nsWellFormed = 1; ctxt->disableSAX = 0; ctxt->valid = 1; ctxt->record_info = 0; ctxt->checkIndex = 0; ctxt->endCheckState = 0; ctxt->inSubset = 0; ctxt->errNo = XML_ERR_OK; ctxt->depth = 0; ctxt->catalogs = NULL; ctxt->sizeentities = 0; ctxt->sizeentcopy = 0; xmlInitNodeInfoSeq(&ctxt->node_seq); if (ctxt->attsDefault != NULL) { xmlHashFree(ctxt->attsDefault, xmlHashDefaultDeallocator); ctxt->attsDefault = NULL; } if (ctxt->attsSpecial != NULL) { xmlHashFree(ctxt->attsSpecial, NULL); ctxt->attsSpecial = NULL; } #ifdef LIBXML_CATALOG_ENABLED if (ctxt->catalogs != NULL) xmlCatalogFreeLocal(ctxt->catalogs); #endif ctxt->nbErrors = 0; ctxt->nbWarnings = 0; if (ctxt->lastError.code != XML_ERR_OK) xmlResetError(&ctxt->lastError); } /** * Reset a push parser context * * @param ctxt an XML parser context * @param chunk a pointer to an array of chars * @param size number of chars in the array * @param filename an optional file name or URI * @param encoding the document encoding, or NULL * @returns 0 in case of success and 1 in case of error */ int xmlCtxtResetPush(xmlParserCtxt *ctxt, const char *chunk, int size, const char *filename, const char *encoding) { xmlParserInputPtr input; if (ctxt == NULL) return(1); xmlCtxtReset(ctxt); input = xmlNewPushInput(filename, chunk, size); if (input == NULL) return(1); if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); return(1); } if (encoding != NULL) xmlSwitchEncodingName(ctxt, encoding); return(0); } static int xmlCtxtSetOptionsInternal(xmlParserCtxtPtr ctxt, int options, int keepMask) { int allMask; if (ctxt == NULL) return(-1); /* * XInclude options aren't handled by the parser. * * XML_PARSE_XINCLUDE * XML_PARSE_NOXINCNODE * XML_PARSE_NOBASEFIX */ allMask = XML_PARSE_RECOVER | XML_PARSE_NOENT | XML_PARSE_DTDLOAD | XML_PARSE_DTDATTR | XML_PARSE_DTDVALID | XML_PARSE_NOERROR | XML_PARSE_NOWARNING | XML_PARSE_PEDANTIC | XML_PARSE_NOBLANKS | #ifdef LIBXML_SAX1_ENABLED XML_PARSE_SAX1 | #endif XML_PARSE_NONET | XML_PARSE_NODICT | XML_PARSE_NSCLEAN | XML_PARSE_NOCDATA | XML_PARSE_COMPACT | XML_PARSE_OLD10 | XML_PARSE_HUGE | XML_PARSE_OLDSAX | XML_PARSE_IGNORE_ENC | XML_PARSE_BIG_LINES | XML_PARSE_NO_XXE | XML_PARSE_UNZIP | XML_PARSE_NO_SYS_CATALOG | XML_PARSE_CATALOG_PI; ctxt->options = (ctxt->options & keepMask) | (options & allMask); /* * For some options, struct members are historically the source * of truth. The values are initalized from global variables and * old code could also modify them directly. Several older API * functions that don't take an options argument rely on these * deprecated mechanisms. * * Once public access to struct members and the globals are * disabled, we can use the options bitmask as source of * truth, making all these struct members obsolete. * * The XML_DETECT_IDS flags is misnamed. It simply enables * loading of the external subset. */ ctxt->recovery = (options & XML_PARSE_RECOVER) ? 1 : 0; ctxt->replaceEntities = (options & XML_PARSE_NOENT) ? 1 : 0; ctxt->loadsubset = (options & XML_PARSE_DTDLOAD) ? XML_DETECT_IDS : 0; ctxt->loadsubset |= (options & XML_PARSE_DTDATTR) ? XML_COMPLETE_ATTRS : 0; ctxt->loadsubset |= (options & XML_PARSE_SKIP_IDS) ? XML_SKIP_IDS : 0; ctxt->validate = (options & XML_PARSE_DTDVALID) ? 1 : 0; ctxt->pedantic = (options & XML_PARSE_PEDANTIC) ? 1 : 0; ctxt->keepBlanks = (options & XML_PARSE_NOBLANKS) ? 0 : 1; ctxt->dictNames = (options & XML_PARSE_NODICT) ? 0 : 1; return(options & ~allMask); } /** * Applies the options to the parser context. Unset options are * cleared. * * @since 2.13.0 * * With older versions, you can use #xmlCtxtUseOptions. * * @param ctxt an XML parser context * @param options a bitmask of xmlParserOption values * @returns 0 in case of success, the set of unknown or unimplemented options * in case of error. */ int xmlCtxtSetOptions(xmlParserCtxt *ctxt, int options) { #ifdef LIBXML_HTML_ENABLED if ((ctxt != NULL) && (ctxt->html)) return(htmlCtxtSetOptions(ctxt, options)); #endif return(xmlCtxtSetOptionsInternal(ctxt, options, 0)); } /** * Get the current options of the parser context. * * @since 2.14.0 * * @param ctxt an XML parser context * @returns the current options set in the parser context, or -1 if ctxt is NULL. */ int xmlCtxtGetOptions(xmlParserCtxt *ctxt) { if (ctxt == NULL) return(-1); return(ctxt->options); } /** * Applies the options to the parser context. The following options * are never cleared and can only be enabled: * * - XML_PARSE_NOERROR * - XML_PARSE_NOWARNING * - XML_PARSE_NONET * - XML_PARSE_NSCLEAN * - XML_PARSE_NOCDATA * - XML_PARSE_COMPACT * - XML_PARSE_OLD10 * - XML_PARSE_HUGE * - XML_PARSE_OLDSAX * - XML_PARSE_IGNORE_ENC * - XML_PARSE_BIG_LINES * * @deprecated Use #xmlCtxtSetOptions. * * @param ctxt an XML parser context * @param options a combination of xmlParserOption * @returns 0 in case of success, the set of unknown or unimplemented options * in case of error. */ int xmlCtxtUseOptions(xmlParserCtxt *ctxt, int options) { int keepMask; #ifdef LIBXML_HTML_ENABLED if ((ctxt != NULL) && (ctxt->html)) return(htmlCtxtUseOptions(ctxt, options)); #endif /* * For historic reasons, some options can only be enabled. */ keepMask = XML_PARSE_NOERROR | XML_PARSE_NOWARNING | XML_PARSE_NONET | XML_PARSE_NSCLEAN | XML_PARSE_NOCDATA | XML_PARSE_COMPACT | XML_PARSE_OLD10 | XML_PARSE_HUGE | XML_PARSE_OLDSAX | XML_PARSE_IGNORE_ENC | XML_PARSE_BIG_LINES; return(xmlCtxtSetOptionsInternal(ctxt, options, keepMask)); } /** * To protect against exponential entity expansion ("billion laughs"), the * size of serialized output is (roughly) limited to the input size * multiplied by this factor. The default value is 5. * * When working with documents making heavy use of entity expansion, it can * be necessary to increase the value. For security reasons, this should only * be considered when processing trusted input. * * @param ctxt an XML parser context * @param maxAmpl maximum amplification factor */ void xmlCtxtSetMaxAmplification(xmlParserCtxt *ctxt, unsigned maxAmpl) { if (ctxt == NULL) return; ctxt->maxAmpl = maxAmpl; } /** * Parse an XML document and return the resulting document tree. * Takes ownership of the input object. * * @since 2.13.0 * * @param ctxt an XML parser context * @param input parser input * @returns the resulting document tree or NULL */ xmlDoc * xmlCtxtParseDocument(xmlParserCtxt *ctxt, xmlParserInput *input) { xmlDocPtr ret = NULL; if ((ctxt == NULL) || (input == NULL)) { xmlFatalErr(ctxt, XML_ERR_ARGUMENT, NULL); xmlFreeInputStream(input); return(NULL); } /* assert(ctxt->inputNr == 0); */ while (ctxt->inputNr > 0) xmlFreeInputStream(xmlCtxtPopInput(ctxt)); if (xmlCtxtPushInput(ctxt, input) < 0) { xmlFreeInputStream(input); return(NULL); } xmlParseDocument(ctxt); ret = xmlCtxtGetDocument(ctxt); /* assert(ctxt->inputNr == 1); */ while (ctxt->inputNr > 0) xmlFreeInputStream(xmlCtxtPopInput(ctxt)); return(ret); } /** * Convenience function to parse an XML document from a * zero-terminated string. * * See #xmlCtxtReadDoc for details. * * @param cur a pointer to a zero terminated string * @param URL base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlReadDoc(const xmlChar *cur, const char *URL, const char *encoding, int options) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlDocPtr doc = NULL; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromString(ctxt, URL, (const char *) cur, encoding, XML_INPUT_BUF_STATIC); if (input != NULL) doc = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(doc); } /** * Convenience function to parse an XML file from the filesystem * or a global, user-defined resource loader. * * If a "-" filename is passed, the function will read from stdin. * This feature is potentially insecure and might be removed from * later versions. * * See #xmlCtxtReadFile for details. * * @param filename a file or URL * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlReadFile(const char *filename, const char *encoding, int options) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlDocPtr doc = NULL; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); xmlCtxtUseOptions(ctxt, options); /* * Backward compatibility for users of command line utilities like * xmlstarlet expecting "-" to mean stdin. This is dangerous and * should be removed at some point. */ if ((filename != NULL) && (filename[0] == '-') && (filename[1] == 0)) input = xmlCtxtNewInputFromFd(ctxt, filename, STDIN_FILENO, encoding, 0); else input = xmlCtxtNewInputFromUrl(ctxt, filename, NULL, encoding, 0); if (input != NULL) doc = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(doc); } /** * Parse an XML in-memory document and build a tree. The input buffer must * not contain a terminating null byte. * * See #xmlCtxtReadMemory for details. * * @param buffer a pointer to a char array * @param size the size of the array * @param url base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlReadMemory(const char *buffer, int size, const char *url, const char *encoding, int options) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlDocPtr doc = NULL; if (size < 0) return(NULL); ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromMemory(ctxt, url, buffer, size, encoding, XML_INPUT_BUF_STATIC); if (input != NULL) doc = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(doc); } /** * Parse an XML from a file descriptor and build a tree. * * See #xmlCtxtReadFd for details. * * NOTE that the file descriptor will not be closed when the * context is freed or reset. * * @param fd an open file descriptor * @param URL base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlReadFd(int fd, const char *URL, const char *encoding, int options) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlDocPtr doc = NULL; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromFd(ctxt, URL, fd, encoding, 0); if (input != NULL) doc = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(doc); } /** * Parse an XML document from I/O functions and context and build a tree. * * See #xmlCtxtReadIO for details. * * @param ioread an I/O read function * @param ioclose an I/O close function (optional) * @param ioctx an I/O handler * @param URL base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlReadIO(xmlInputReadCallback ioread, xmlInputCloseCallback ioclose, void *ioctx, const char *URL, const char *encoding, int options) { xmlParserCtxtPtr ctxt; xmlParserInputPtr input; xmlDocPtr doc = NULL; ctxt = xmlNewParserCtxt(); if (ctxt == NULL) return(NULL); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromIO(ctxt, URL, ioread, ioclose, ioctx, encoding, 0); if (input != NULL) doc = xmlCtxtParseDocument(ctxt, input); xmlFreeParserCtxt(ctxt); return(doc); } /** * Parse an XML in-memory document and build a tree. * * `URL` is used as base to resolve external entities and for error * reporting. * * @param ctxt an XML parser context * @param str a pointer to a zero terminated string * @param URL base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlCtxtReadDoc(xmlParserCtxt *ctxt, const xmlChar *str, const char *URL, const char *encoding, int options) { xmlParserInputPtr input; if (ctxt == NULL) return(NULL); xmlCtxtReset(ctxt); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromString(ctxt, URL, (const char *) str, encoding, XML_INPUT_BUF_STATIC); if (input == NULL) return(NULL); return(xmlCtxtParseDocument(ctxt, input)); } /** * Parse an XML file from the filesystem or a global, user-defined * resource loader. * * @param ctxt an XML parser context * @param filename a file or URL * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlCtxtReadFile(xmlParserCtxt *ctxt, const char *filename, const char *encoding, int options) { xmlParserInputPtr input; if (ctxt == NULL) return(NULL); xmlCtxtReset(ctxt); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromUrl(ctxt, filename, NULL, encoding, 0); if (input == NULL) return(NULL); return(xmlCtxtParseDocument(ctxt, input)); } /** * Parse an XML in-memory document and build a tree. The input buffer must * not contain a terminating null byte. * * `URL` is used as base to resolve external entities and for error * reporting. * * @param ctxt an XML parser context * @param buffer a pointer to a char array * @param size the size of the array * @param URL base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlCtxtReadMemory(xmlParserCtxt *ctxt, const char *buffer, int size, const char *URL, const char *encoding, int options) { xmlParserInputPtr input; if ((ctxt == NULL) || (size < 0)) return(NULL); xmlCtxtReset(ctxt); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromMemory(ctxt, URL, buffer, size, encoding, XML_INPUT_BUF_STATIC); if (input == NULL) return(NULL); return(xmlCtxtParseDocument(ctxt, input)); } /** * Parse an XML document from a file descriptor and build a tree. * * NOTE that the file descriptor will not be closed when the * context is freed or reset. * * `URL` is used as base to resolve external entities and for error * reporting. * * @param ctxt an XML parser context * @param fd an open file descriptor * @param URL base URL (optional) * @param encoding the document encoding (optional) * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlCtxtReadFd(xmlParserCtxt *ctxt, int fd, const char *URL, const char *encoding, int options) { xmlParserInputPtr input; if (ctxt == NULL) return(NULL); xmlCtxtReset(ctxt); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromFd(ctxt, URL, fd, encoding, 0); if (input == NULL) return(NULL); return(xmlCtxtParseDocument(ctxt, input)); } /** * Parse an XML document from I/O functions and source and build a tree. * This reuses the existing `ctxt` parser context * * `URL` is used as base to resolve external entities and for error * reporting. * * @param ctxt an XML parser context * @param ioread an I/O read function * @param ioclose an I/O close function * @param ioctx an I/O handler * @param URL the base URL to use for the document * @param encoding the document encoding, or NULL * @param options a combination of xmlParserOption * @returns the resulting document tree */ xmlDoc * xmlCtxtReadIO(xmlParserCtxt *ctxt, xmlInputReadCallback ioread, xmlInputCloseCallback ioclose, void *ioctx, const char *URL, const char *encoding, int options) { xmlParserInputPtr input; if (ctxt == NULL) return(NULL); xmlCtxtReset(ctxt); xmlCtxtUseOptions(ctxt, options); input = xmlCtxtNewInputFromIO(ctxt, URL, ioread, ioclose, ioctx, encoding, 0); if (input == NULL) return(NULL); return(xmlCtxtParseDocument(ctxt, input)); }