Update to libxml-2.11.4

NEWS

@@ -1,5 +1,12 @@
 NEWS file for libxml2
 
+v2.11.4: May 18 2023
+
+Fixes a serious regression.
+
+- parser: Fix regression when push parsing UTF-8 sequences
+
+
 v2.11.3: May 11 2023
 
 Fixes more regressions.

README.md

@@ -2,7 +2,7 @@
 
 libxml2 Windows build with Visual Studio.
 
-This version is libxml2-2.11.3.
+This version is libxml2-2.11.4.
 
 To build, simply open the required solution file, and
 you know how to use Visual Studio, right?

distfiles/download.url

@@ -1 +1 @@
-https://download.gnome.org/sources/libxml2/2.11/libxml2-2.11.3.tar.xz
+https://download.gnome.org/sources/libxml2/2.11/libxml2-2.11.4.tar.xz

distfiles/libxml2-2.11.3-import.md5 → distfiles/libxml2-2.11.4-import.md5

@@ -2,7 +2,7 @@
 5aa30e4781d7c802b78ad4346632485d  HTMLparser.c
 7bcee4dcda730f1f34029529f3cc24a4  HTMLtree.c
 dd63184811cb2ff705c3e466364d3773  INSTALL
-39aea02633e5243ad3798ee268a34a9d  NEWS
+201a96e1d38a965133ab1e20247dc32a  NEWS
 c67f840acaa24a4752ec5a09c5b7c9ab  README.libxml2.md
 2774afd7cdd927d278dcc64f73112e0a  SAX.c
 5d0ea4ec1675df4c37ed947df682ff39  SAX2.c
@@ -86,8 +86,8 @@ e857d2ed47c0848d6f469ebc16ef31f1  include/wsockcompat.h
 678e168c1dacffd4da8d528e4e733d05  list.c
 a591291901040a0c5a8a4c187154d1de  nanoftp.c
 e0bf81a96bd2674a5c51710b76b3b20a  nanohttp.c
-e73e07a7e57282942ac51b34094e45c9  parser.c
-f3a661d72d9c0335fcd5a5ee948ddbee  parserInternals.c
+524e02fec3580c1095a6ed1650ec1b6d  parser.c
+f22c5bb3a2fa2ed3567e83a25dbcc774  parserInternals.c
 247aeee5286dc32f9212a71764c8916b  pattern.c
 be5da77f293be48c4611cc4dd456954d  relaxng.c
 b171015076b703294c403dd53c4ef090  runsuite.c

include/libxml/xmlversion.h

@@ -29,21 +29,21 @@ XMLPUBFUN void xmlCheckVersion(int version);
  *
  * the version string like "1.2.3"
  */
-#define LIBXML_DOTTED_VERSION "2.11.3"
+#define LIBXML_DOTTED_VERSION "2.11.4"
 
 /**
  * LIBXML_VERSION:
  *
  * the version number: 1.2.3 value is 10203
  */
-#define LIBXML_VERSION 21103
+#define LIBXML_VERSION 21104
 
 /**
  * LIBXML_VERSION_STRING:
  *
  * the version number string, 1.2.3 value is "10203"
  */
-#define LIBXML_VERSION_STRING "21103"
+#define LIBXML_VERSION_STRING "21104"
 
 /**
  * LIBXML_VERSION_EXTRA:
@@ -58,7 +58,7 @@ XMLPUBFUN void xmlCheckVersion(int version);
  * Macro to check that the libxml version in use is compatible with
  * the version the software has been compiled against
  */
-#define LIBXML_TEST_VERSION xmlCheckVersion(21103);
+#define LIBXML_TEST_VERSION xmlCheckVersion(21104);
 
 #ifndef VMS
 #if 0

parser.c

@@ -4332,7 +4332,7 @@ xmlParsePubidLiteral(xmlParserCtxtPtr ctxt) {
     return(buf);
 }
 
-static void xmlParseCharDataComplex(xmlParserCtxtPtr ctxt);
+static void xmlParseCharDataComplex(xmlParserCtxtPtr ctxt, int partial);
 
 /*
  * used for the test in the inner loop of the char data testing
@@ -4373,27 +4373,22 @@ static const unsigned char test_char_data[256] = {
 };
 
 /**
- * xmlParseCharData:
+ * xmlParseCharDataInternal:
  * @ctxt:  an XML parser context
- * @cdata:  unused
- *
- * DEPRECATED: Internal function, don't use.
+ * @partial:  buffer may contain partial UTF-8 sequences
  *
  * Parse character data. Always makes progress if the first char isn't
  * '<' or '&'.
  *
- * if we are within a CDATA section ']]>' marks an end of section.
- *
  * The right angle bracket (>) may be represented using the string "&gt;",
  * and must, for compatibility, be escaped using "&gt;" 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 ::= [^<&]* - ([^<&]* ']]>' [^<&]*)
  */
-
-void
-xmlParseCharData(xmlParserCtxtPtr ctxt, ATTRIBUTE_UNUSED int cdata) {
+static void
+xmlParseCharDataInternal(xmlParserCtxtPtr ctxt, int partial) {
     const xmlChar *in;
     int nbchar = 0;
     int line = ctxt->input->line;
@@ -4526,7 +4521,7 @@ xmlParseCharData(xmlParserCtxtPtr ctxt, ATTRIBUTE_UNUSED int cdata) {
              (*in == 0x09) || (*in == 0x0a));
     ctxt->input->line = line;
     ctxt->input->col = col;
-    xmlParseCharDataComplex(ctxt);
+    xmlParseCharDataComplex(ctxt, partial);
 }
 
 /**
@@ -4541,7 +4536,7 @@ xmlParseCharData(xmlParserCtxtPtr ctxt, ATTRIBUTE_UNUSED int cdata) {
  * of non-ASCII characters.
  */
 static void
-xmlParseCharDataComplex(xmlParserCtxtPtr ctxt) {
+xmlParseCharDataComplex(xmlParserCtxtPtr ctxt, int partial) {
     xmlChar buf[XML_PARSER_BIG_BUFFER_SIZE + 5];
     int nbchar = 0;
     int cur, l;
@@ -4604,15 +4599,42 @@ xmlParseCharDataComplex(xmlParserCtxtPtr ctxt) {
 	    }
 	}
     }
-    if ((ctxt->input->cur < ctxt->input->end) && (!IS_CHAR(cur))) {
-	/* Generate the error and skip the offending character */
-        xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR,
-                          "PCDATA invalid Char value %d\n",
-	                  cur ? cur : CUR);
-	NEXT;
+    /*
+     * cur == 0 can mean
+     *
+     * - XML_PARSER_EOF or memory error. This is checked above.
+     * - An actual 0 character.
+     * - End of buffer.
+     * - 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 != '&')) {
+            /* Generate the error and skip the offending character */
+            xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR,
+                              "PCDATA invalid Char value %d\n", cur);
+            NEXTL(l);
+        }
     }
 }
 
+/**
+ * xmlParseCharData:
+ * @ctxt:  an XML parser context
+ * @cdata:  unused
+ *
+ * DEPRECATED: Internal function, don't use.
+ */
+void
+xmlParseCharData(xmlParserCtxtPtr ctxt, ATTRIBUTE_UNUSED int cdata) {
+    xmlParseCharDataInternal(ctxt, 0);
+}
+
 /**
  * xmlParseExternalID:
  * @ctxt:  an XML parser context
@@ -9875,7 +9897,7 @@ xmlParseContentInternal(xmlParserCtxtPtr ctxt) {
 	 * Last case, text. Note that References are handled directly.
 	 */
 	else {
-	    xmlParseCharData(ctxt, 0);
+	    xmlParseCharDataInternal(ctxt, 0);
 	}
 
 	SHRINK;
@@ -11662,7 +11684,7 @@ xmlParseTryOrFinish(xmlParserCtxtPtr ctxt, int terminate) {
 			    goto done;
                     }
                     ctxt->checkIndex = 0;
-		    xmlParseCharData(ctxt, 0);
+		    xmlParseCharDataInternal(ctxt, !terminate);
 		}
 		break;
 	    }

parserInternals.c

@@ -715,14 +715,20 @@ xmlCurrentChar(xmlParserCtxtPtr ctxt, int *len) {
 
             avail = ctxt->input->end - ctxt->input->cur;
 
-	    if ((avail < 2) || (cur[1] & 0xc0) != 0x80)
+            if (avail < 2)
+                goto incomplete_sequence;
+	    if ((cur[1] & 0xc0) != 0x80)
 		goto encoding_error;
 	    if ((c & 0xe0) == 0xe0) {
-		if ((avail < 3) || (cur[2] & 0xc0) != 0x80)
+                if (avail < 3)
+                    goto incomplete_sequence;
+		if ((cur[2] & 0xc0) != 0x80)
 		    goto encoding_error;
 		if ((c & 0xf0) == 0xf0) {
+                    if (avail < 4)
+                        goto incomplete_sequence;
 		    if (((c & 0xf8) != 0xf0) ||
-			(avail < 4) || ((cur[3] & 0xc0) != 0x80))
+			((cur[3] & 0xc0) != 0x80))
 			goto encoding_error;
 		    /* 4-byte code */
 		    *len = 4;
@@ -784,25 +790,20 @@ xmlCurrentChar(xmlParserCtxtPtr ctxt, int *len) {
 	return(0xA);
     }
     return(*ctxt->input->cur);
-encoding_error:
-    /*
-     * An encoding problem may arise from a truncated input buffer
-     * splitting a character in the middle. In that case do not raise
-     * an error but return 0 to indicate an end of stream problem
-     */
-    if (ctxt->input->end - ctxt->input->cur < 4) {
-	*len = 0;
-	return(0);
-    }
 
+encoding_error:
     /*
      * If we detect an UTF8 error that probably mean that the
      * input encoding didn't get properly advertised in the
      * declaration header. Report the error and switch the encoding
      * to ISO-Latin-1 (if you don't like this policy, just declare the
      * encoding !)
      */
-    {
+    if (ctxt->input->end - ctxt->input->cur < 4) {
+	__xmlErrEncoding(ctxt, XML_ERR_INVALID_CHAR,
+		     "Input is not proper UTF-8, indicate encoding !\n",
+		     NULL, NULL);
+    } else {
         char buffer[150];
 
 	snprintf(&buffer[0], 149, "Bytes: 0x%02X 0x%02X 0x%02X 0x%02X\n",
@@ -815,6 +816,16 @@ xmlCurrentChar(xmlParserCtxtPtr ctxt, int *len) {
     ctxt->charset = XML_CHAR_ENCODING_8859_1;
     *len = 1;
     return(*ctxt->input->cur);
+
+incomplete_sequence:
+    /*
+     * An encoding problem may arise from a truncated input buffer
+     * splitting a character in the middle. In that case do not raise
+     * an error but return 0. This should only happen when push parsing
+     * char data.
+     */
+    *len = 0;
+    return(0);
 }
 
 /**

win32/rcVersion.h

@@ -1,4 +1,4 @@
 #define LIBXML_MAJOR_VERSION 2
 #define LIBXML_MINOR_VERSION 11
-#define LIBXML_MICRO_VERSION 3
-#define LIBXML_DOTTED_VERSION "2.11.3"
+#define LIBXML_MICRO_VERSION 4
+#define LIBXML_DOTTED_VERSION "2.11.4"