📄 xmlparse.c
字号:
if (!tag->buf) return XML_ERROR_NO_MEMORY; tag->bufEnd = tag->buf + bufSize; } memcpy(tag->buf, tag->rawName, tag->rawNameLength); tag->rawName = tag->buf; } ++tagLevel; if (startElementHandler) { enum XML_Error result; XML_Char *toPtr; for (;;) { const char *rawNameEnd = tag->rawName + tag->rawNameLength; const char *fromPtr = tag->rawName; int bufSize; if (nextPtr) toPtr = (XML_Char *)(tag->buf + ROUND_UP(tag->rawNameLength, sizeof(XML_Char))); else toPtr = (XML_Char *)tag->buf; tag->name = toPtr; XmlConvert(enc, &fromPtr, rawNameEnd, (ICHAR **)&toPtr, (ICHAR *)tag->bufEnd - 1); if (fromPtr == rawNameEnd) break; bufSize = (tag->bufEnd - tag->buf) << 1; tag->buf = realloc(tag->buf, bufSize); if (!tag->buf) return XML_ERROR_NO_MEMORY; tag->bufEnd = tag->buf + bufSize; if (nextPtr) tag->rawName = tag->buf; } *toPtr = XML_T('\0'); result = storeAtts(parser, enc, tag->name, s); if (result) return result; startElementHandler(handlerArg, tag->name, (const XML_Char **)atts); poolClear(&tempPool); } else { tag->name = 0; if (defaultHandler) reportDefault(parser, enc, s, next); } break; } case XML_TOK_EMPTY_ELEMENT_WITH_ATTS: if (!startElementHandler) { enum XML_Error result = storeAtts(parser, enc, 0, s); if (result) return result; } /* fall through */ case XML_TOK_EMPTY_ELEMENT_NO_ATTS: if (startElementHandler || endElementHandler) { const char *rawName = s + enc->minBytesPerChar; const XML_Char *name = poolStoreString(&tempPool, enc, rawName, rawName + XmlNameLength(enc, rawName)); if (!name) return XML_ERROR_NO_MEMORY; poolFinish(&tempPool); if (startElementHandler) { enum XML_Error result = storeAtts(parser, enc, name, s); if (result) return result; startElementHandler(handlerArg, name, (const XML_Char **)atts); } if (endElementHandler) { if (startElementHandler) *eventPP = *eventEndPP; endElementHandler(handlerArg, name); } poolClear(&tempPool); } else if (defaultHandler) reportDefault(parser, enc, s, next); if (tagLevel == 0) return epilogProcessor(parser, next, end, nextPtr); break; case XML_TOK_END_TAG: if (tagLevel == startTagLevel) return XML_ERROR_ASYNC_ENTITY; else { int len; const char *rawName; TAG *tag = tagStack; tagStack = tag->parent; tag->parent = freeTagList; freeTagList = tag; rawName = s + enc->minBytesPerChar*2; len = XmlNameLength(enc, rawName); if (len != tag->rawNameLength || memcmp(tag->rawName, rawName, len) != 0) { *eventPP = rawName; return XML_ERROR_TAG_MISMATCH; } --tagLevel; if (endElementHandler) { if (tag->name) endElementHandler(handlerArg, tag->name); else { const XML_Char *name = poolStoreString(&tempPool, enc, rawName, rawName + len); if (!name) return XML_ERROR_NO_MEMORY; endElementHandler(handlerArg, name); poolClear(&tempPool); } } else if (defaultHandler) reportDefault(parser, enc, s, next); if (tagLevel == 0) return epilogProcessor(parser, next, end, nextPtr); } break; case XML_TOK_CHAR_REF: { int n = XmlCharRefNumber(enc, s); if (n < 0) return XML_ERROR_BAD_CHAR_REF; if (characterDataHandler) { XML_Char buf[XML_ENCODE_MAX]; characterDataHandler(handlerArg, buf, XmlEncode(n, (ICHAR *)buf)); } else if (defaultHandler) reportDefault(parser, enc, s, next); } break; case XML_TOK_XML_DECL: return XML_ERROR_MISPLACED_XML_PI; case XML_TOK_DATA_NEWLINE: if (characterDataHandler) { XML_Char c = XML_T('\n'); characterDataHandler(handlerArg, &c, 1); } else if (defaultHandler) reportDefault(parser, enc, s, next); break; case XML_TOK_CDATA_SECT_OPEN: { enum XML_Error result; if (characterDataHandler) characterDataHandler(handlerArg, dataBuf, 0); else if (defaultHandler) reportDefault(parser, enc, s, next); result = doCdataSection(parser, enc, &next, end, nextPtr); if (!next) { processor = cdataSectionProcessor; return result; } } break; case XML_TOK_TRAILING_RSQB: if (nextPtr) { *nextPtr = s; return XML_ERROR_NONE; } if (characterDataHandler) { if (MUST_CONVERT(enc, s)) { ICHAR *dataPtr = (ICHAR *)dataBuf; XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd); characterDataHandler(handlerArg, dataBuf, dataPtr - (ICHAR *)dataBuf); } else characterDataHandler(handlerArg, (XML_Char *)s, (XML_Char *)end - (XML_Char *)s); } else if (defaultHandler) reportDefault(parser, enc, s, end); if (startTagLevel == 0) { *eventPP = end; return XML_ERROR_NO_ELEMENTS; } if (tagLevel != startTagLevel) { *eventPP = end; return XML_ERROR_ASYNC_ENTITY; } return XML_ERROR_NONE; case XML_TOK_DATA_CHARS: if (characterDataHandler) { if (MUST_CONVERT(enc, s)) { for (;;) { ICHAR *dataPtr = (ICHAR *)dataBuf; XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd); *eventEndPP = s; characterDataHandler(handlerArg, dataBuf, dataPtr - (ICHAR *)dataBuf); if (s == next) break; *eventPP = s; } } else characterDataHandler(handlerArg, (XML_Char *)s, (XML_Char *)next - (XML_Char *)s); } else if (defaultHandler) reportDefault(parser, enc, s, next); break; case XML_TOK_PI: if (!reportProcessingInstruction(parser, enc, s, next)) return XML_ERROR_NO_MEMORY; break; default: if (defaultHandler) reportDefault(parser, enc, s, next); break; } *eventPP = s = next; } /* not reached */}/* If tagName is non-null, build a real list of attributes,otherwise just check the attributes for well-formedness. */static enum XML_Error storeAtts(XML_Parser parser, const ENCODING *enc, const XML_Char *tagName, const char *s){ ELEMENT_TYPE *elementType = 0; int nDefaultAtts = 0; const XML_Char **appAtts; int i; int n; if (tagName) { elementType = (ELEMENT_TYPE *)lookup(&dtd.elementTypes, tagName, 0); if (elementType) nDefaultAtts = elementType->nDefaultAtts; } n = XmlGetAttributes(enc, s, attsSize, atts); if (n + nDefaultAtts > attsSize) { int oldAttsSize = attsSize; attsSize = n + nDefaultAtts + INIT_ATTS_SIZE; atts = realloc((void *)atts, attsSize * sizeof(ATTRIBUTE)); if (!atts) return XML_ERROR_NO_MEMORY; if (n > oldAttsSize) XmlGetAttributes(enc, s, n, atts); } appAtts = (const XML_Char **)atts; for (i = 0; i < n; i++) { ATTRIBUTE_ID *attId = getAttributeId(parser, enc, atts[i].name, atts[i].name + XmlNameLength(enc, atts[i].name)); if (!attId) return XML_ERROR_NO_MEMORY; if ((attId->name)[-1]) { if (enc == encoding) eventPtr = atts[i].name; return XML_ERROR_DUPLICATE_ATTRIBUTE; } (attId->name)[-1] = 1; appAtts[i << 1] = attId->name; if (!atts[i].normalized) { enum XML_Error result; int isCdata = 1; if (attId->maybeTokenized) { int j; for (j = 0; j < nDefaultAtts; j++) { if (attId == elementType->defaultAtts[j].id) { isCdata = elementType->defaultAtts[j].isCdata; break; } } } result = storeAttributeValue(parser, enc, isCdata, atts[i].valuePtr, atts[i].valueEnd, &tempPool); if (result) return result; if (tagName) { appAtts[(i << 1) + 1] = poolStart(&tempPool); poolFinish(&tempPool); } else poolDiscard(&tempPool); } else if (tagName) { appAtts[(i << 1) + 1] = poolStoreString(&tempPool, enc, atts[i].valuePtr, atts[i].valueEnd); if (appAtts[(i << 1) + 1] == 0) return XML_ERROR_NO_MEMORY; poolFinish(&tempPool); } } if (tagName) { int j; for (j = 0; j < nDefaultAtts; j++) { const DEFAULT_ATTRIBUTE *da = elementType->defaultAtts + j; if (!(da->id->name)[-1] && da->value) { (da->id->name)[-1] = 1; appAtts[i << 1] = da->id->name; appAtts[(i << 1) + 1] = da->value; i++; } } appAtts[i << 1] = 0; } while (i-- > 0) ((XML_Char *)appAtts[i << 1])[-1] = 0; return XML_ERROR_NONE;}/* The idea here is to avoid using stack for each CDATA section whenthe whole file is parsed with one call. */staticenum XML_Error cdataSectionProcessor(XML_Parser parser, const char *start, const char *end, const char **endPtr){ enum XML_Error result = doCdataSection(parser, encoding, &start, end, endPtr); if (start) { processor = contentProcessor; return contentProcessor(parser, start, end, endPtr); } return result;}/* startPtr gets set to non-null is the section is closed, and to null ifthe section is not yet closed. */staticenum XML_Error doCdataSection(XML_Parser parser, const ENCODING *enc, const char **startPtr, const char *end, const char **nextPtr){ const char *s = *startPtr; const char *dummy; const char **eventPP; const char **eventEndPP; if (enc == encoding) { eventPP = &eventPtr; *eventPP = s; eventEndPP = &eventEndPtr; } else eventPP = eventEndPP = &dummy; *startPtr = 0; for (;;) { const char *next; int tok = XmlCdataSectionTok(enc, s, end, &next); *eventEndPP = next; switch (tok) { case XML_TOK_CDATA_SECT_CLOSE: if (characterDataHandler) characterDataHandler(handlerArg, dataBuf, 0); else if (defaultHandler) reportDefault(parser, enc, s, next); *startPtr = next; return XML_ERROR_NONE; case XML_TOK_DATA_NEWLINE: if (characterDataHandler) { XML_Char c = XML_T('\n'); characterDataHandler(handlerArg, &c, 1); } else if (defaultHandler) reportDefault(parser, enc, s, next); break; case XML_TOK_DATA_CHARS: if (characterDataHandler) { if (MUST_CONVERT(enc, s)) { for (;;) { ICHAR *dataPtr = (ICHAR *)dataBuf; XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd); *eventEndPP = next; characterDataHandler(handlerArg, dataBuf, dataPtr - (ICHAR *)dataBuf); if (s == next) break; *eventPP = s; } } else characterDataHandler(handlerArg, (XML_Char *)s, (XML_Char *)next - (XML_Char *)s); } else if (defaultHandler) reportDefault(parser, enc, s, next); break; case XML_TOK_INVALID: *eventPP = next; return XML_ERROR_INVALID_TOKEN; case XML_TOK_PARTIAL_CHAR: if (nextPtr) { *nextPtr = s; return XML_ERROR_NONE; } return XML_ERROR_PARTIAL_CHAR; case XML_TOK_PARTIAL: case XML_TOK_NONE: if (nextPtr) { *nextPtr = s; return XML_ERROR_NONE; } return XML_ERROR_UNCLOSED_CDATA_SECTION; default: abort(); } *eventPP = s = next; } /* not reached */}static enum XML_ErrorinitializeEncoding(XML_Parser parser){ const char *s;#ifdef XML_UNICODE char encodingBuf[128]; if (!protocolEncodingName) s = 0; else { int i; for (i = 0; protocolEncodingName[i]; i++) { if (i == sizeof(encodingBuf) - 1 || protocolEncodingName[i] >= 0x80 || protocolEncodingName[i] < 0) { encodingBuf[0] = '\0'; break; } encodingBuf[i] = (char)protocolEncodingName[i]; } encodingBuf[i] = '\0'; s = encodingBuf; }#else s = protocolEncodingName;#endif if (XmlInitEncoding(&initEncoding, &encoding, s)) return XML_ERROR_NONE; return handleUnknownEncoding(parser, protocolEncodingName);}static enum XML_ErrorprocessXmlDecl(XML_Parser parser, int isGeneralTextEntity, const char *s, const char *next){ const char *encodingName = 0; const ENCODING *newEncoding = 0; const char *version; int standalone = -1; if (!XmlParseXmlDecl(isGeneralTextEntity, encoding, s, next, &eventPtr, &version, &encodingName, &newEncoding, &standalone)) return XML_ERROR_SYNTAX; if (!isGeneralTextEntity && standalone == 1) dtd.standalone = 1; if (defaultHandler) reportDefault(parser, encoding, s, next); if (!protocolEncodingName) { if (newEncoding) { if (newEncoding->minBytesPerChar != encoding->minBytesPerChar) { eventPtr = encodingName; return XML_ERROR_INCORRECT_ENCODING; } encoding = newEncoding; } else if (encodingName) { enum XML_Error result; const XML_Char *s = poolStoreString(&tempPool, encoding, encodingName, encodingName + XmlNameLength(encoding, encodingName));
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -