commit 7a7620af4b806d9a67ac883a0f84479e3a9fda6b
Author: Olli Pottonen <olli pottonen iki fi>
Date:   Sun Jun 28 11:38:26 2015 +1000

    Improve encoding detection.
    
    Call xmlDetectCharEncoding() even if there is less than 4 bytes of
    input; 2 or 3 bytes may be enough. Avoid unnecessary copying of data
    to local array.

diff --git a/HTMLparser.c b/HTMLparser.c
index 8717d0b..9c4ec04 100644
--- a/HTMLparser.c
+++ b/HTMLparser.c
@@ -4619,7 +4619,6 @@ __htmlParseContent(void *ctxt) {
 
 int
 htmlParseDocument(htmlParserCtxtPtr ctxt) {
-    xmlChar start[4];
     xmlCharEncoding enc;
     xmlDtdPtr dtd;
 
@@ -4641,18 +4640,12 @@ htmlParseDocument(htmlParserCtxtPtr ctxt) {
     if ((ctxt->sax) && (ctxt->sax->setDocumentLocator))
         ctxt->sax->setDocumentLocator(ctxt->userData, &xmlDefaultSAXLocator);
 
-    if ((ctxt->encoding == (const xmlChar *)XML_CHAR_ENCODING_NONE) &&
-        ((ctxt->input->end - ctxt->input->cur) >= 4)) {
+    if (ctxt->encoding == NULL) {
 	/*
-	 * Get the 4 first bytes and decode the charset
-	 * if enc != XML_CHAR_ENCODING_NONE
 	 * plug some encoding conversion routines.
 	 */
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(&start[0], 4);
+	int length = ctxt->input->end - ctxt->input->cur;
+	enc = xmlDetectCharEncoding(ctxt->input->cur, length);
 	if (enc != XML_CHAR_ENCODING_NONE) {
 	    xmlSwitchEncoding(ctxt, enc);
 	}
diff --git a/parser.c b/parser.c
index 0e73d47..af637ca 100644
--- a/parser.c
+++ b/parser.c
@@ -2623,9 +2623,7 @@ xmlParserHandlePEReference(xmlParserCtxtPtr ctxt) {
 	    } else {
 	        if ((entity->etype == XML_INTERNAL_PARAMETER_ENTITY) ||
 		    (entity->etype == XML_EXTERNAL_PARAMETER_ENTITY)) {
-		    xmlChar start[4];
 		    xmlCharEncoding enc;
-
 		    /*
 		     * Note: external parameter entities will not be loaded, it
 		     * is not required for a non-validating parser, unless the
@@ -2664,16 +2662,11 @@ xmlParserHandlePEReference(xmlParserCtxtPtr ctxt) {
 		    GROW
                     if (ctxt->instate == XML_PARSER_EOF)
                         return;
-		    if ((ctxt->input->end - ctxt->input->cur)>=4) {
-			start[0] = RAW;
-			start[1] = NXT(1);
-			start[2] = NXT(2);
-			start[3] = NXT(3);
-			enc = xmlDetectCharEncoding(start, 4);
-			if (enc != XML_CHAR_ENCODING_NONE) {
-			    xmlSwitchEncoding(ctxt, enc);
-			}
-		    }
+		    int len = ctxt->input->end - ctxt->input->cur;
+		    enc = xmlDetectCharEncoding(ctxt->input->cur, len);
+		    if (enc != XML_CHAR_ENCODING_NONE) {
+		        xmlSwitchEncoding(ctxt, enc);
+                    }
 
 		    if ((entity->etype == XML_EXTERNAL_PARAMETER_ENTITY) &&
 			(CMP5(CUR_PTR, '<', '?', 'x', 'm', 'l' )) &&
@@ -7080,18 +7073,12 @@ xmlParseExternalSubset(xmlParserCtxtPtr ctxt, const xmlChar *ExternalID,
     xmlDetectSAX2(ctxt);
     GROW;
 
-    if ((ctxt->encoding == NULL) &&
-        (ctxt->input->end - ctxt->input->cur >= 4)) {
-        xmlChar start[4];
-	xmlCharEncoding enc;
-
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(start, 4);
-	if (enc != XML_CHAR_ENCODING_NONE)
+    if (ctxt->encoding == NULL) {
+        int length = ctxt->input->end - ctxt->input->cur;
+	xmlCharEncoding enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+	if (enc != XML_CHAR_ENCODING_NONE) {
 	    xmlSwitchEncoding(ctxt, enc);
+	}
     }
 
     if (CMP5(CUR_PTR, '<', '?', 'x', 'm', 'l')) {
@@ -10803,7 +10790,6 @@ xmlParseMisc(xmlParserCtxtPtr ctxt) {
 
 int
 xmlParseDocument(xmlParserCtxtPtr ctxt) {
-    xmlChar start[4];
     xmlCharEncoding enc;
 
     xmlInitParser();
@@ -10826,18 +10812,9 @@ xmlParseDocument(xmlParserCtxtPtr ctxt) {
     if (ctxt->instate == XML_PARSER_EOF)
 	return(-1);
 
-    if ((ctxt->encoding == NULL) &&
-        ((ctxt->input->end - ctxt->input->cur) >= 4)) {
-	/*
-	 * Get the 4 first bytes and decode the charset
-	 * if enc != XML_CHAR_ENCODING_NONE
-	 * plug some encoding conversion routines.
-	 */
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(&start[0], 4);
+    if (ctxt->encoding == NULL) {
+        int length = ctxt->input->end - ctxt->input->cur;
+        enc = xmlDetectCharEncoding(ctxt->input->cur, length);
 	if (enc != XML_CHAR_ENCODING_NONE) {
 	    xmlSwitchEncoding(ctxt, enc);
 	}
@@ -10997,9 +10974,7 @@ xmlParseDocument(xmlParserCtxtPtr ctxt) {
 
 int
 xmlParseExtParsedEnt(xmlParserCtxtPtr ctxt) {
-    xmlChar start[4];
     xmlCharEncoding enc;
-
     if ((ctxt == NULL) || (ctxt->input == NULL))
         return(-1);
 
@@ -11016,19 +10991,12 @@ xmlParseExtParsedEnt(xmlParserCtxtPtr ctxt) {
         ctxt->sax->setDocumentLocator(ctxt->userData, &xmlDefaultSAXLocator);
 
     /*
-     * Get the 4 first bytes and decode the charset
-     * if enc != XML_CHAR_ENCODING_NONE
      * plug some encoding conversion routines.
      */
-    if ((ctxt->input->end - ctxt->input->cur) >= 4) {
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(start, 4);
-	if (enc != XML_CHAR_ENCODING_NONE) {
-	    xmlSwitchEncoding(ctxt, enc);
-	}
+    int length = ctxt->input->end - ctxt->input->cur;
+    enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+    if (enc != XML_CHAR_ENCODING_NONE) {
+        xmlSwitchEncoding(ctxt, enc);
     }
 
 
@@ -11428,28 +11396,19 @@ xmlParseTryOrFinish(xmlParserCtxtPtr ctxt, int terminate) {
 	        goto done;
             case XML_PARSER_START:
 		if (ctxt->charset == XML_CHAR_ENCODING_NONE) {
-		    xmlChar start[4];
 		    xmlCharEncoding enc;
-
 		    /*
 		     * Very first chars read from the document flow.
 		     */
 		    if (avail < 4)
 			goto done;
 
-		    /*
-		     * Get the 4 first bytes and decode the charset
-		     * if enc != XML_CHAR_ENCODING_NONE
-		     * plug some encoding conversion routines,
-		     * else xmlSwitchEncoding will set to (default)
-		     * UTF8.
-		     */
-		    start[0] = RAW;
-		    start[1] = NXT(1);
-		    start[2] = NXT(2);
-		    start[3] = NXT(3);
-		    enc = xmlDetectCharEncoding(start, 4);
-		    xmlSwitchEncoding(ctxt, enc);
+		    int length = ctxt->input->end - ctxt->input->cur;
+		    enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+		    if (enc != XML_CHAR_ENCODING_NONE) {
+		        xmlSwitchEncoding(ctxt, enc);
+		    }
+
 		    break;
 		}
 
@@ -12583,13 +12542,10 @@ xmlCreatePushParserCtxt(xmlSAXHandlerPtr sax, void *user_data,
     xmlParserInputBufferPtr buf;
     xmlCharEncoding enc = XML_CHAR_ENCODING_NONE;
 
-    /*
-     * plug some encoding conversion routines
-     */
-    if ((chunk != NULL) && (size >= 4))
-	enc = xmlDetectCharEncoding((const xmlChar *) chunk, size);
+    if (chunk != NULL)
+        enc = xmlDetectCharEncoding((const xmlChar *) chunk, size);
 
-    buf = xmlAllocParserInputBuffer(enc);
+    buf = xmlAllocParserInputBuffer(XML_CHAR_ENCODING_NONE);
     if (buf == NULL) return(NULL);
 
     ctxt = xmlNewParserCtxt();
@@ -12791,7 +12747,6 @@ xmlIOParseDTD(xmlSAXHandlerPtr sax, xmlParserInputBufferPtr input,
     xmlDtdPtr ret = NULL;
     xmlParserCtxtPtr ctxt;
     xmlParserInputPtr pinput = NULL;
-    xmlChar start[4];
 
     if (input == NULL)
 	return(NULL);
@@ -12838,6 +12793,12 @@ xmlIOParseDTD(xmlSAXHandlerPtr sax, xmlParserInputBufferPtr input,
     }
     if (enc != XML_CHAR_ENCODING_NONE) {
         xmlSwitchEncoding(ctxt, enc);
+    } else {
+        int length = ctxt->input->end - ctxt->input->cur;
+	enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+	if (enc != XML_CHAR_ENCODING_NONE) {
+	    xmlSwitchEncoding(ctxt, enc);
+	}
     }
 
     pinput->filename = NULL;
@@ -12860,23 +12821,6 @@ xmlIOParseDTD(xmlSAXHandlerPtr sax, xmlParserInputBufferPtr input,
     ctxt->myDoc->extSubset = xmlNewDtd(ctxt->myDoc, BAD_CAST "none",
 	                               BAD_CAST "none", BAD_CAST "none");
 
-    if ((enc == XML_CHAR_ENCODING_NONE) &&
-        ((ctxt->input->end - ctxt->input->cur) >= 4)) {
-	/*
-	 * Get the 4 first bytes and decode the charset
-	 * if enc != XML_CHAR_ENCODING_NONE
-	 * plug some encoding conversion routines.
-	 */
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(start, 4);
-	if (enc != XML_CHAR_ENCODING_NONE) {
-	    xmlSwitchEncoding(ctxt, enc);
-	}
-    }
-
     xmlParseExternalSubset(ctxt, BAD_CAST "none", BAD_CAST "none");
 
     if (ctxt->myDoc != NULL) {
@@ -12979,9 +12923,10 @@ xmlSAXParseDTD(xmlSAXHandlerPtr sax, const xmlChar *ExternalID,
 	    xmlFree(systemIdCanonic);
 	return(NULL);
     }
-    if ((ctxt->input->end - ctxt->input->cur) >= 4) {
-	enc = xmlDetectCharEncoding(ctxt->input->cur, 4);
-	xmlSwitchEncoding(ctxt, enc);
+    int length = ctxt->input->end - ctxt->input->cur;
+    enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+    if (enc != XML_CHAR_ENCODING_NONE) {
+        xmlSwitchEncoding(ctxt, enc);
     }
 
     if (input->filename == NULL)
@@ -13084,7 +13029,6 @@ xmlParseCtxtExternalEntity(xmlParserCtxtPtr ctx, const xmlChar *URL,
     xmlNodePtr newRoot;
     xmlSAXHandlerPtr oldsax = NULL;
     int ret = 0;
-    xmlChar start[4];
     xmlCharEncoding enc;
 
     if (ctx == NULL) return(-1);
@@ -13150,15 +13094,11 @@ xmlParseCtxtExternalEntity(xmlParserCtxtPtr ctx, const xmlChar *URL,
      * plug some encoding conversion routines.
      */
     GROW
-    if ((ctxt->input->end - ctxt->input->cur) >= 4) {
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(start, 4);
-	if (enc != XML_CHAR_ENCODING_NONE) {
-	    xmlSwitchEncoding(ctxt, enc);
-	}
+
+    int length = ctxt->input->end - ctxt->input->cur;
+    enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+    if (enc != XML_CHAR_ENCODING_NONE) {
+        xmlSwitchEncoding(ctxt, enc);
     }
 
     /*
@@ -13293,7 +13233,6 @@ xmlParseExternalEntityPrivate(xmlDocPtr doc, xmlParserCtxtPtr oldctxt,
     xmlNodePtr newRoot;
     xmlSAXHandlerPtr oldsax = NULL;
     xmlParserErrors ret = XML_ERR_OK;
-    xmlChar start[4];
     xmlCharEncoding enc;
 
     if (((depth > 40) &&
@@ -13375,20 +13314,13 @@ xmlParseExternalEntityPrivate(xmlDocPtr doc, xmlParserCtxtPtr oldctxt,
     newRoot->doc = doc;
 
     /*
-     * Get the 4 first bytes and decode the charset
-     * if enc != XML_CHAR_ENCODING_NONE
      * plug some encoding conversion routines.
      */
     GROW;
-    if ((ctxt->input->end - ctxt->input->cur) >= 4) {
-	start[0] = RAW;
-	start[1] = NXT(1);
-	start[2] = NXT(2);
-	start[3] = NXT(3);
-	enc = xmlDetectCharEncoding(start, 4);
-	if (enc != XML_CHAR_ENCODING_NONE) {
-	    xmlSwitchEncoding(ctxt, enc);
-	}
+    int length = ctxt->input->end - ctxt->input->cur;
+    enc = xmlDetectCharEncoding(ctxt->input->cur, length);
+    if (enc != XML_CHAR_ENCODING_NONE) {
+        xmlSwitchEncoding(ctxt, enc);
     }
 
     /*
@@ -15132,8 +15064,9 @@ xmlCtxtResetPush(xmlParserCtxtPtr ctxt, const char *chunk,
     if (ctxt == NULL)
         return(1);
 
-    if ((encoding == NULL) && (chunk != NULL) && (size >= 4))
-        enc = xmlDetectCharEncoding((const xmlChar *) chunk, size);
+    if ((encoding == NULL) && (chunk != NULL)) {
+	enc = xmlDetectCharEncoding(BAD_CAST chunk, size);
+    }
 
     buf = xmlAllocParserInputBuffer(XML_CHAR_ENCODING_NONE);
     if (buf == NULL)