Fix forward reference in the presence of (?#( (open parens in comment).

git-svn-id: svn://vcs.exim.org/pcre/code/trunk@544 2f5784b3-3f2a-0410-8824-cb99058d5e15

4 files changed, 42 insertions, 13 deletions

diff --git a/ChangeLog b/ChangeLog
index 0fb949d..fb1b313 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -89,6 +89,11 @@ Version 8.10 03-Jun-2010
 20. Added the /T option to pcretest so as to be able to run tests with non-
     standard character tables, thus making it possible to include the tests
     used for 19 above in the standard set of tests.  
+    
+21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward 
+    reference to a subpattern the other side of a comment that contains an 
+    opening parenthesis caused either an internal compiling error, or a 
+    reference to the wrong subpattern.
 
 
 Version 8.02 19-Mar-2010
diff --git a/pcre_compile.c b/pcre_compile.c
index 3a23768..eab5991 100644
--- a/pcre_compile.c
+++ b/pcre_compile.c
@@ -1129,25 +1129,39 @@ dealing with. The very first call may not start with a parenthesis. */
 
 if (ptr[0] == CHAR_LEFT_PARENTHESIS)
   {
-  if (ptr[1] == CHAR_QUESTION_MARK &&
-      ptr[2] == CHAR_VERTICAL_LINE)
-    {
-    ptr += 3;
-    dup_parens = TRUE;
-    }
+  /* Handle specials such as (*SKIP) or (*UTF8) etc. */
+  
+  if (ptr[1] == CHAR_ASTERISK) ptr += 2;
+ 
+  /* Handle a normal, unnamed capturing parenthesis. */
 
-  /* Handle a normal, unnamed capturing parenthesis */
-
-  else if (ptr[1] != CHAR_QUESTION_MARK && ptr[1] != CHAR_ASTERISK)
+  else if (ptr[1] != CHAR_QUESTION_MARK)
     {
     *count += 1;
     if (name == NULL && *count == lorn) return *count;
     ptr++;
     }
 
+  /* All cases now have (? at the start. Remember when we are in a group
+  where the parenthesis numbers are duplicated. */
+
+  else if (ptr[2] == CHAR_VERTICAL_LINE)
+    {
+    ptr += 3;
+    dup_parens = TRUE;
+    }
+    
+  /* Handle comments; all characters are allowed until a ket is reached. */
+
+  else if (ptr[2] == CHAR_NUMBER_SIGN)
+    {
+    for (ptr += 3; *ptr != 0; ptr++) if (*ptr == CHAR_RIGHT_PARENTHESIS) break;
+    goto FAIL_EXIT;
+    }  
+
   /* Handle a condition. If it is an assertion, just carry on so that it
   is processed as normal. If not, skip to the closing parenthesis of the
-  condition (there can't be any nested parens. */
+  condition (there can't be any nested parens). */
 
   else if (ptr[2] == CHAR_LEFT_PARENTHESIS)
     {
@@ -1159,7 +1173,7 @@ if (ptr[0] == CHAR_LEFT_PARENTHESIS)
       }
     }
 
-  /* We have either (? or (* and not a condition */
+  /* Start with (? but not a condition. */
 
   else
     {
@@ -1281,8 +1295,7 @@ for (; *ptr != 0; ptr++)
   else if (*ptr == CHAR_RIGHT_PARENTHESIS)
     {
     if (dup_parens && *count < hwm_count) *count = hwm_count;
-    *ptrptr = ptr;
-    return -1;
+    goto FAIL_EXIT; 
     }
 
   else if (*ptr == CHAR_VERTICAL_LINE && dup_parens)
diff --git a/testdata/testinput11 b/testdata/testinput11
index c074176..795e1be 100644
--- a/testdata/testinput11
+++ b/testdata/testinput11
@@ -478,4 +478,9 @@ however, we need the complication for Perl. ---/
 /(\w+)b(*COMMIT)\w{2}/
     abbb
 
+/--- Check opening parens in comment when seeking forward reference. ---/ 
+
+/(?&t)(?#()(?(DEFINE)(?<t>a))/
+    bac
+
 /-- End of testinput11 --/
diff --git a/testdata/testoutput11 b/testdata/testoutput11
index 8b0f111..6270d5b 100644
--- a/testdata/testoutput11
+++ b/testdata/testoutput11
@@ -936,4 +936,10 @@ No match
     abbb
 No match
 
+/--- Check opening parens in comment when seeking forward reference. ---/ 
+
+/(?&t)(?#()(?(DEFINE)(?<t>a))/
+    bac
+ 0: a
+
 /-- End of testinput11 --/