Bug#20795: extractvalue() won't accept names containing a dot (.)

Dot character was not considered as a valid identifier body character.

mysql-test/r/xml.result

@@ -704,3 +704,9 @@ a
 select extractValue('<ns:element xmlns:ns="myns">a</ns:element>','/ns:element/@xmlns:ns');
 extractValue('<ns:element xmlns:ns="myns">a</ns:element>','/ns:element/@xmlns:ns')
 myns
+select extractValue('<foo><foo.bar>Data</foo.bar><something>Otherdata</something></foo>','/foo/foo.bar');
+extractValue('<foo><foo.bar>Data</foo.bar><something>Otherdata</something></foo>','/foo/foo.bar')
+Data
+select extractValue('<foo><foo.bar>Data</foo.bar><something>Otherdata</something></foo>','/foo/something');
+extractValue('<foo><foo.bar>Data</foo.bar><something>Otherdata</something></foo>','/foo/something')
+Otherdata

mysql-test/t/xml.test

@@ -354,3 +354,9 @@ DROP PROCEDURE p2;
 select extractValue('<ns:element xmlns:ns="myns"/>','count(ns:element)');
 select extractValue('<ns:element xmlns:ns="myns">a</ns:element>','/ns:element');
 select extractValue('<ns:element xmlns:ns="myns">a</ns:element>','/ns:element/@xmlns:ns');
+
+#
+# Bug#20795 extractvalue() won't accept names containing a dot (.)
+#
+select extractValue('<foo><foo.bar>Data</foo.bar><something>Otherdata</something></foo>','/foo/foo.bar');
+select extractValue('<foo><foo.bar>Data</foo.bar><something>Otherdata</something></foo>','/foo/something');

sql/item_xmlfunc.cc

@@ -1356,7 +1356,8 @@ my_xpath_lex_scan(MY_XPATH *xpath,
          (length= xpath->cs->cset->ctype(xpath->cs, &ctype,
                                          (const uchar*) beg,
                                          (const uchar*) end)) > 0 &&
-         ((ctype & (_MY_L | _MY_U | _MY_NMR)) || *beg == '_' || *beg == '-') ;
+         ((ctype & (_MY_L | _MY_U | _MY_NMR)) ||
+          *beg == '_' || *beg == '-' || *beg == '.') ;
          beg+= length) /* no op */;
     lex->end= beg;
 

strings/ctype.c

@@ -80,35 +80,35 @@ struct my_cs_file_section_st
 static struct my_cs_file_section_st sec[] =
 {
   {_CS_MISC,		"xml"},
-  {_CS_MISC,		"xml.version"},
-  {_CS_MISC,		"xml.encoding"},
+  {_CS_MISC,		"xml/version"},
+  {_CS_MISC,		"xml/encoding"},
   {_CS_MISC,		"charsets"},
-  {_CS_MISC,		"charsets.max-id"},
-  {_CS_CHARSET,		"charsets.charset"},
-  {_CS_PRIMARY_ID,	"charsets.charset.primary-id"},
-  {_CS_BINARY_ID,	"charsets.charset.binary-id"},
-  {_CS_CSNAME,		"charsets.charset.name"},
-  {_CS_FAMILY,		"charsets.charset.family"},
-  {_CS_CSDESCRIPT,	"charsets.charset.description"},
-  {_CS_MISC,		"charsets.charset.alias"},
-  {_CS_MISC,		"charsets.charset.ctype"},
-  {_CS_CTYPEMAP,	"charsets.charset.ctype.map"},
-  {_CS_MISC,		"charsets.charset.upper"},
-  {_CS_UPPERMAP,	"charsets.charset.upper.map"},
-  {_CS_MISC,		"charsets.charset.lower"},
-  {_CS_LOWERMAP,	"charsets.charset.lower.map"},
-  {_CS_MISC,		"charsets.charset.unicode"},
-  {_CS_UNIMAP,		"charsets.charset.unicode.map"},
-  {_CS_COLLATION,	"charsets.charset.collation"},
-  {_CS_COLNAME,		"charsets.charset.collation.name"},
-  {_CS_ID,		"charsets.charset.collation.id"},
-  {_CS_ORDER,		"charsets.charset.collation.order"},
-  {_CS_FLAG,		"charsets.charset.collation.flag"},
-  {_CS_COLLMAP,		"charsets.charset.collation.map"},
-  {_CS_RESET,		"charsets.charset.collation.rules.reset"},
-  {_CS_DIFF1,		"charsets.charset.collation.rules.p"},
-  {_CS_DIFF2,		"charsets.charset.collation.rules.s"},
-  {_CS_DIFF3,		"charsets.charset.collation.rules.t"},
+  {_CS_MISC,		"charsets/max-id"},
+  {_CS_CHARSET,		"charsets/charset"},
+  {_CS_PRIMARY_ID,	"charsets/charset/primary-id"},
+  {_CS_BINARY_ID,	"charsets/charset/binary-id"},
+  {_CS_CSNAME,		"charsets/charset/name"},
+  {_CS_FAMILY,		"charsets/charset/family"},
+  {_CS_CSDESCRIPT,	"charsets/charset/description"},
+  {_CS_MISC,		"charsets/charset/alias"},
+  {_CS_MISC,		"charsets/charset/ctype"},
+  {_CS_CTYPEMAP,	"charsets/charset/ctype/map"},
+  {_CS_MISC,		"charsets/charset/upper"},
+  {_CS_UPPERMAP,	"charsets/charset/upper/map"},
+  {_CS_MISC,		"charsets/charset/lower"},
+  {_CS_LOWERMAP,	"charsets/charset/lower/map"},
+  {_CS_MISC,		"charsets/charset/unicode"},
+  {_CS_UNIMAP,		"charsets/charset/unicode/map"},
+  {_CS_COLLATION,	"charsets/charset/collation"},
+  {_CS_COLNAME,		"charsets/charset/collation/name"},
+  {_CS_ID,		"charsets/charset/collation/id"},
+  {_CS_ORDER,		"charsets/charset/collation/order"},
+  {_CS_FLAG,		"charsets/charset/collation/flag"},
+  {_CS_COLLMAP,		"charsets/charset/collation/map"},
+  {_CS_RESET,		"charsets/charset/collation/rules/reset"},
+  {_CS_DIFF1,		"charsets/charset/collation/rules/p"},
+  {_CS_DIFF2,		"charsets/charset/collation/rules/s"},
+  {_CS_DIFF3,		"charsets/charset/collation/rules/t"},
   {0,	NULL}
 };
 

strings/xml.c

@@ -159,7 +159,7 @@ static int my_xml_enter(MY_XML_PARSER *st, const char *str, uint len)
   }
   if (st->attrend > st->attr)
   {
-    st->attrend[0]='.';
+    st->attrend[0]= '/';
     st->attrend++;
   }
   memcpy(st->attrend,str,len);
@@ -188,9 +188,9 @@ static int my_xml_leave(MY_XML_PARSER *p, const char *str, uint slen)
   char g[32];
   int  rc;
 
-  /* Find previous '.' or beginning */
-  for( e=p->attrend; (e>p->attr) && (e[0] != '.') ; e--);
-  glen = (uint) ((e[0] == '.') ? (p->attrend-e-1) : p->attrend-e);
+  /* Find previous '/' or beginning */
+  for( e=p->attrend; (e>p->attr) && (e[0] != '/') ; e--);
+  glen = (uint) ((e[0] == '/') ? (p->attrend-e-1) : p->attrend-e);
   
   if (str && (slen != glen))
   {