added two new macros IS_ASCII_LETTER and IS_ASCII_DIGIT used with (html)

* include/libxml/parserInternals.h: added two new macros
  IS_ASCII_LETTER and IS_ASCII_DIGIT used with (html)
  parsing and xpath for testing data not necessarily
  unicode.
* HTMLparser.c, xpath.c: changed use of IS_LETTER_CH and
  IS_DIGIT_CH macros to ascii versions (bug 153936).

ChangeLog

@@ -1,3 +1,12 @@
+Sat Oct  2 15:03:14 PDT 2004 William Brack <wbrack@mmm.com.hk>
+
+	* include/libxml/parserInternals.h: added two new macros
+	  IS_ASCII_LETTER and IS_ASCII_DIGIT used with (html)
+	  parsing and xpath for testing data not necessarily
+	  unicode.
+	* HTMLparser.c, xpath.c: changed use of IS_LETTER_CH and
+	  IS_DIGIT_CH macros to ascii versions (bug 153936).
+	  
 Fri Oct  1 20:37:25 PDT 2004 William Brack <wbrack@mmm.com.hk>
 
 	* error.c: added some coding to attempt to display which file

HTMLparser.c

@@ -2168,11 +2168,11 @@ htmlParseHTMLName(htmlParserCtxtPtr ctxt) {
     int i = 0;
     xmlChar loc[HTML_PARSER_BUFFER_SIZE];
 
-    if (!IS_LETTER_CH(CUR) && (CUR != '_') &&
+    if (!IS_ASCII_LETTER(CUR) && (CUR != '_') &&
         (CUR != ':')) return(NULL);
 
     while ((i < HTML_PARSER_BUFFER_SIZE) &&
-           ((IS_LETTER_CH(CUR)) || (IS_DIGIT_CH(CUR)) ||
+           ((IS_ASCII_LETTER(CUR)) || (IS_ASCII_DIGIT(CUR)) ||
 	   (CUR == ':') || (CUR == '-') || (CUR == '_'))) {
 	if ((CUR >= 'A') && (CUR <= 'Z')) loc[i] = CUR + 0x20;
         else loc[i] = CUR;

include/libxml/parserInternals.h

@@ -197,6 +197,26 @@ XMLPUBVAR unsigned int xmlParserMaxDepth;
  *
  */
 #define IS_LETTER_CH(c) xmlIsBaseChar_ch(c)
+
+/**
+ * IS_ASCII_LETTER(c)
+ * @c: an xmlChar value
+ *
+ * Macro to check [a-zA-Z]
+ *
+ */
+#define IS_ASCII_LETTER(c)	(((0x41 <= (c)) && ((c) <= 0x5a)) || \
+				 ((0x61 <= (c)) && ((c) <= 0x7a)))
+
+/**
+ * IS_ASCII_DIGIT(c)
+ * @c: an xmlChar value
+ *
+ * Macro to check [0-9]
+ *
+ */
+#define IS_ASCII_DIGIT(c)	((0x30 <= (c)) && ((c) <= 0x39))
+
 /**
  * IS_PUBIDCHAR:
  * @c:  an UNICODE value (int)

xpath.c

@@ -57,6 +57,13 @@
 	    "Unimplemented block at %s:%d\n",				\
             __FILE__, __LINE__);
 
+/*
+ * TODO:
+ * There are a few spots where some tests are done which depend upon ascii
+ * data.  These should be enhanced for full UTF8 support (see particularly
+ * any use of the macros IS_ASCII_CHARACTER and IS_ASCII_DIGIT)
+ */
+ 
 #if defined(LIBXML_SCHEMAS_ENABLED) || defined(LIBXML_XPATH_ENABLED)
 /************************************************************************
  * 									*
@@ -7947,7 +7954,7 @@ xmlXPathCompPrimaryExpr(xmlXPathParserContextPtr ctxt) {
 	}
 	NEXT;
 	SKIP_BLANKS;
-    } else if (IS_DIGIT_CH(CUR) || (CUR == '.' && IS_DIGIT_CH(NXT(1)))) {
+    } else if (IS_ASCII_DIGIT(CUR) || (CUR == '.' && IS_ASCII_DIGIT(NXT(1)))) {
 	xmlXPathCompNumber(ctxt);
     } else if ((CUR == '\'') || (CUR == '"')) {
 	xmlXPathCompLiteral(ctxt);
@@ -8009,12 +8016,12 @@ xmlXPathScanName(xmlXPathParserContextPtr ctxt) {
     int len = 0;
 
     SKIP_BLANKS;
-    if (!IS_LETTER_CH(CUR) && (CUR != '_') &&
+    if (!IS_ASCII_LETTER(CUR) && (CUR != '_') &&
         (CUR != ':')) {
 	return(NULL);
     }
 
-    while ((IS_LETTER_CH(NXT(len))) || (IS_DIGIT_CH(NXT(len))) ||
+    while ((IS_ASCII_LETTER(NXT(len))) || (IS_ASCII_DIGIT(NXT(len))) ||
            (NXT(len) == '.') || (NXT(len) == '-') ||
 	   (NXT(len) == '_') || (NXT(len) == ':') || 
 	   (IS_COMBINING_CH(NXT(len))) ||
@@ -8024,7 +8031,7 @@ xmlXPathScanName(xmlXPathParserContextPtr ctxt) {
 	if (len >= XML_MAX_NAMELEN) {
 	    xmlGenericError(xmlGenericErrorContext, 
 	       "xmlScanName: reached XML_MAX_NAMELEN limit\n");
-	    while ((IS_LETTER_CH(NXT(len))) || (IS_DIGIT_CH(NXT(len))) ||
+	    while ((IS_ASCII_LETTER(NXT(len))) || (IS_ASCII_DIGIT(NXT(len))) ||
 		   (NXT(len) == '.') || (NXT(len) == '-') ||
 		   (NXT(len) == '_') || (NXT(len) == ':') || 
 		   (IS_COMBINING_CH(NXT(len))) ||
@@ -8060,8 +8067,10 @@ xmlXPathCompPathExpr(xmlXPathParserContextPtr ctxt) {
     xmlChar *name = NULL; /* we may have to preparse a name to find out */
 
     SKIP_BLANKS;
-    if ((CUR == '$') || (CUR == '(') || (IS_DIGIT_CH(CUR)) ||
+    if ((CUR == '$') || (CUR == '(') || 
-        (CUR == '\'') || (CUR == '"') || (CUR == '.' && IS_DIGIT_CH(NXT(1)))) {
+    	(IS_ASCII_DIGIT(CUR)) ||
+        (CUR == '\'') || (CUR == '"') ||
+	(CUR == '.' && IS_ASCII_DIGIT(NXT(1)))) {
 	lc = 0;
     } else if (CUR == '*') {
 	/* relative or absolute location path */
@@ -8957,7 +8966,7 @@ xmlXPathCompLocationPath(xmlXPathParserContextPtr ctxt) {
 		NEXT;
 		SKIP_BLANKS;
 		if ((CUR != 0 ) &&
-		    ((IS_LETTER_CH(CUR)) || (CUR == '_') || (CUR == '.') ||
+		    ((IS_ASCII_LETTER(CUR)) || (CUR == '_') || (CUR == '.') ||
 		     (CUR == '@') || (CUR == '*')))
 		    xmlXPathCompRelativeLocationPath(ctxt);
 	    }