Roll libxml to bdec2183f34b37ee89ae1d330c6ad2bb4d76605f

third_party/libxml/src/HTMLparser.c

 /*
  * HTMLparser.c : an HTML 4.0 non-verifying parser
  *
  * See Copyright for the status of this software.
  *
  * daniel@veillard.com
  */
 
 #define IN_LIBXML
 #include "libxml.h"
@@ skipping 87 matching lines @@
 /**
  * htmlParseErr:
  * @ctxt:  an HTML parser context
  * @error:  the error number
  * @msg:  the error message
  * @str1:  string infor
  * @str2:  string infor
  *
  * Handle a fatal parser error, i.e. violating Well-Formedness constraints
  */
-static void
+static void LIBXML_ATTR_FORMAT(3,0)
 htmlParseErr(xmlParserCtxtPtr ctxt, xmlParserErrors error,
              const char *msg, const xmlChar *str1, const xmlChar *str2)
 {
     if ((ctxt != NULL) && (ctxt->disableSAX != 0) &&
         (ctxt->instate == XML_PARSER_EOF))
         return;
     if (ctxt != NULL)
         ctxt->errNo = error;
     __xmlRaiseError(NULL, NULL, NULL, ctxt, NULL, XML_FROM_HTML, error,
                     XML_ERR_ERROR, NULL, 0,
                     (const char *) str1, (const char *) str2,
                     NULL, 0, 0,
                     msg, str1, str2);
     if (ctxt != NULL)
         ctxt->wellFormed = 0;
 }
 
 /**
  * htmlParseErrInt:
  * @ctxt:  an HTML parser context
  * @error:  the error number
  * @msg:  the error message
  * @val:  integer info
  *
  * Handle a fatal parser error, i.e. violating Well-Formedness constraints
  */
-static void
+static void LIBXML_ATTR_FORMAT(3,0)
 htmlParseErrInt(xmlParserCtxtPtr ctxt, xmlParserErrors error,
              const char *msg, int val)
 {
     if ((ctxt != NULL) && (ctxt->disableSAX != 0) &&
         (ctxt->instate == XML_PARSER_EOF))
         return;
     if (ctxt != NULL)
         ctxt->errNo = error;
     __xmlRaiseError(NULL, NULL, NULL, ctxt, NULL, XML_FROM_HTML, error,
                     XML_ERR_ERROR, NULL, 0, NULL, NULL,
@@ skipping 150 matching lines @@
 
 #define UPPER (toupper(*ctxt->input->cur))
 
 #define SKIP(val) ctxt->nbChars += (val),ctxt->input->cur += (val),ctxt->input->col+=(val)
 
 #define NXT(val) ctxt->input->cur[(val)]
 
 #define UPP(val) (toupper(ctxt->input->cur[(val)]))
 
 #define CUR_PTR ctxt->input->cur
+#define BASE_PTR ctxt->input->base
 
 #define SHRINK if ((ctxt->input->cur - ctxt->input->base > 2 * INPUT_CHUNK) && \
                    (ctxt->input->end - ctxt->input->cur < 2 * INPUT_CHUNK)) \
         xmlParserInputShrink(ctxt->input)
 
 #define GROW if ((ctxt->progressive == 0) &&                            \
                  (ctxt->input->end - ctxt->input->cur < INPUT_CHUNK))   \
         xmlParserInputGrow(ctxt->input, INPUT_CHUNK)
 
 #define CURRENT ((int) (*ctxt->input->cur))
@@ skipping 2148 matching lines @@
     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 == ctxt->input->end)
+            return(NULL);
+
         if ((*in > 0) && (*in < 0x80)) {
             count = in - ctxt->input->cur;
             ret = xmlDictLookup(ctxt->dict, ctxt->input->cur, count);
             ctxt->input->cur = in;
             ctxt->nbChars += count;
             ctxt->input->col += count;
             return(ret);
         }
     }
     return(htmlParseNameComplex(ctxt));
 }
 
 static const xmlChar *
 htmlParseNameComplex(xmlParserCtxtPtr ctxt) {
     int len = 0, l;
     int c;
     int count = 0;
+    const xmlChar *base = ctxt->input->base;
 
     /*
      * Handler for more complex cases
      */
     GROW;
     c = CUR_CHAR(l);
     if ((c == ' ') || (c == '>') || (c == '/') || /* accelerators */
         (!IS_LETTER(c) && (c != '_') &&
          (c != ':'))) {
         return(NULL);
     }
 
     while ((c != ' ') && (c != '>') && (c != '/') && /* test bigname.xml */
            ((IS_LETTER(c)) || (IS_DIGIT(c)) ||
             (c == '.') || (c == '-') ||
             (c == '_') || (c == ':') ||
             (IS_COMBINING(c)) ||
             (IS_EXTENDER(c)))) {
         if (count++ > 100) {
             count = 0;
             GROW;
         }
         len += l;
         NEXTL(l);
         c = CUR_CHAR(l);
+        if (ctxt->input->base != base) {
+            /*
+             * We changed encoding from an unknown encoding
+             * Input buffer changed location, so we better start again
+             */
+            return(htmlParseNameComplex(ctxt));
+        }
     }
+
+    if (ctxt->input->base > ctxt->input->cur - len)
+        return(NULL);
+
     return(xmlDictLookup(ctxt->dict, ctxt->input->cur - len, len));
 }
 
 
 /**
  * htmlParseHTMLAttribute:
  * @ctxt:  an HTML parser context
  * @stop:  a char stop value
  *
  * parse an HTML attribute value till the stop (quote), if
@@ skipping 231 matching lines @@
  *
  * parse an HTML Literal
  *
  * [11] SystemLiteral ::= ('"' [^"]* '"') | ("'" [^']* "'")
  *
  * Returns the SystemLiteral parsed or NULL
  */
 
 static xmlChar *
 htmlParseSystemLiteral(htmlParserCtxtPtr ctxt) {
-    const xmlChar *q;
+    size_t len = 0, startPosition = 0;
     xmlChar *ret = NULL;
 
     if (CUR == '"') {
         NEXT;
-        q = CUR_PTR;
-        while ((IS_CHAR_CH(CUR)) && (CUR != '"'))
+
+        if (CUR_PTR < BASE_PTR)
+            return(ret);
+        startPosition = CUR_PTR - BASE_PTR;
+
+        while ((IS_CHAR_CH(CUR)) && (CUR != '"')) {
             NEXT;
+            len++;
+        }
         if (!IS_CHAR_CH(CUR)) {
             htmlParseErr(ctxt, XML_ERR_LITERAL_NOT_FINISHED,
                          "Unfinished SystemLiteral\n", NULL, NULL);
         } else {
-            ret = xmlStrndup(q, CUR_PTR - q);
+            ret = xmlStrndup((BASE_PTR+startPosition), len);
             NEXT;
         }
     } else if (CUR == '\'') {
         NEXT;
-        q = CUR_PTR;
-        while ((IS_CHAR_CH(CUR)) && (CUR != '\''))
+
+        if (CUR_PTR < BASE_PTR)
+            return(ret);
+        startPosition = CUR_PTR - BASE_PTR;
+
+        while ((IS_CHAR_CH(CUR)) && (CUR != '\'')) {
             NEXT;
+            len++;
+        }
         if (!IS_CHAR_CH(CUR)) {
             htmlParseErr(ctxt, XML_ERR_LITERAL_NOT_FINISHED,
                          "Unfinished SystemLiteral\n", NULL, NULL);
         } else {
-            ret = xmlStrndup(q, CUR_PTR - q);
+            ret = xmlStrndup((BASE_PTR+startPosition), len);
             NEXT;
         }
     } else {
         htmlParseErr(ctxt, XML_ERR_LITERAL_NOT_STARTED,
                      " or ' expected\n", NULL, NULL);
     }
 
     return(ret);
 }
 
 /**
  * htmlParsePubidLiteral:
  * @ctxt:  an HTML parser context
  *
  * parse an HTML public literal
  *
  * [12] PubidLiteral ::= '"' PubidChar* '"' | "'" (PubidChar - "'")* "'"
  *
  * Returns the PubidLiteral parsed or NULL.
  */
 
 static xmlChar *
 htmlParsePubidLiteral(htmlParserCtxtPtr ctxt) {
-    const xmlChar *q;
+    size_t len = 0, startPosition = 0;
     xmlChar *ret = NULL;
     /*
      * Name ::= (Letter | '_') (NameChar)*
      */
     if (CUR == '"') {
         NEXT;
-        q = CUR_PTR;
-        while (IS_PUBIDCHAR_CH(CUR)) NEXT;
+
+        if (CUR_PTR < BASE_PTR)
+            return(ret);
+        startPosition = CUR_PTR - BASE_PTR;
+
+        while (IS_PUBIDCHAR_CH(CUR)) {
+            len++;
+            NEXT;
+        }
+
         if (CUR != '"') {
             htmlParseErr(ctxt, XML_ERR_LITERAL_NOT_FINISHED,
                          "Unfinished PubidLiteral\n", NULL, NULL);
         } else {
-            ret = xmlStrndup(q, CUR_PTR - q);
+            ret = xmlStrndup((BASE_PTR + startPosition), len);
             NEXT;
         }
     } else if (CUR == '\'') {
         NEXT;
-        q = CUR_PTR;
-        while ((IS_PUBIDCHAR_CH(CUR)) && (CUR != '\''))
-            NEXT;
+
+        if (CUR_PTR < BASE_PTR)
+            return(ret);
+        startPosition = CUR_PTR - BASE_PTR;
+
+        while ((IS_PUBIDCHAR_CH(CUR)) && (CUR != '\'')){
+            len++;
+            NEXT;
+        }
+
         if (CUR != '\'') {
             htmlParseErr(ctxt, XML_ERR_LITERAL_NOT_FINISHED,
                          "Unfinished PubidLiteral\n", NULL, NULL);
         } else {
-            ret = xmlStrndup(q, CUR_PTR - q);
+            ret = xmlStrndup((BASE_PTR + startPosition), len);
             NEXT;
         }
     } else {
         htmlParseErr(ctxt, XML_ERR_LITERAL_NOT_STARTED,
                      "PubidLiteral \" or ' expected\n", NULL, NULL);
     }
 
     return(ret);
 }
 
@@ skipping 4260 matching lines @@
         xmlFreeParserInputBuffer(input);
         return (NULL);
     }
     inputPush(ctxt, stream);
     return (htmlDoRead(ctxt, URL, encoding, options, 1));
 }
 
 #define bottom_HTMLparser
 #include "elfgcchack.h"
 #endif /* LIBXML_HTML_ENABLED */