xpath: Fix parsing of non-ASCII names

Fix a long-standing issue where QNames starting with a non-ASCII
character would be rejected. This became more visible after "streaming"
XPath evaluation was disabled since the latter handled non-ASCII names
correctly.

Fixes #818.

Author: nwellnhof

result/XPath/tests/unicodesimple

@@ -0,0 +1,6 @@
+
+========================
+Expression: /文書
+Object is a Node Set :
+Set contains 1 nodes:
+1  ELEMENT #E6#96#87#E6#9B#B8

test/XPath/docs/unicode

@@ -0,0 +1 @@
+<文書>text1</文書>

test/XPath/tests/unicodesimple

@@ -0,0 +1 @@
+/文書

xpath.c

@@ -10434,8 +10434,9 @@ xmlXPathCompLocationPath(xmlXPathParserContextPtr ctxt) {
 	    } else if (CUR == '/') {
 		NEXT;
 		SKIP_BLANKS;
-		if ((CUR != 0 ) &&
-		    ((IS_ASCII_LETTER(CUR)) || (CUR == '_') || (CUR == '.') ||
+		if ((CUR != 0) &&
+		    ((IS_ASCII_LETTER(CUR)) || (CUR >= 0x80) ||
+                     (CUR == '_') || (CUR == '.') ||
 		     (CUR == '@') || (CUR == '*')))
 		    xmlXPathCompRelativeLocationPath(ctxt);
 	    }