Fix non-diagnosis of (?=a)(?R) (positive lookaheads not skipped when checking

for an empty match).


git-svn-id: svn://vcs.exim.org/pcre/code/trunk@282 2f5784b3-3f2a-0410-8824-cb99058d5e15

4 files changed, 38 insertions, 2 deletions

diff --git a/ChangeLog b/ChangeLog
index f8f3ac8..7cfc09d 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -75,6 +75,11 @@ Version 7.5 12-Nov-07
     line for every non-matching line. GNU grep prints nothing, and pcregrep now 
     does the same. The return code can be used to tell if there were any 
     non-matching lines. 
+    
+14. The pattern (?=something)(?R) was not being diagnosed as a potentially
+    infinitely looping recursion. The bug was that positive lookaheads were not
+    being skipped when checking for a possible empty match (negative lookaheads 
+    and both kinds of lookbehind were skipped). 
 
 
 Version 7.4 21-Sep-07
diff --git a/pcre_compile.c b/pcre_compile.c
index 885e239..8c5c4a1 100644
--- a/pcre_compile.c
+++ b/pcre_compile.c
@@ -1508,8 +1508,9 @@ for (;;)
 can match the empty string or not. It is called from could_be_empty()
 below and from compile_branch() when checking for an unlimited repeat of a
 group that can match nothing. Note that first_significant_code() skips over
-assertions. If we hit an unclosed bracket, we return "empty" - this means we've
-struck an inner bracket whose current branch will already have been scanned.
+backward and negative forward assertions when its final argument is TRUE. If we
+hit an unclosed bracket, we return "empty" - this means we've struck an inner
+bracket whose current branch will already have been scanned.
 
 Arguments:
   code        points to start of search
@@ -1530,6 +1531,16 @@ for (code = first_significant_code(code + _pcre_OP_lengths[*code], NULL, 0, TRUE
   const uschar *ccode;
 
   c = *code;
+  
+  /* Skip over forward assertions; the other assertions are skipped by 
+  first_significant_code() with a TRUE final argument. */
+  
+  if (c == OP_ASSERT)
+    {  
+    do code += GET(code, 1); while (*code == OP_ALT);
+    c = *code;
+    continue;
+    }  
 
   /* Groups with zero repeats can of course be empty; skip them. */
 
diff --git a/testdata/testinput2 b/testdata/testinput2
index 9b153dc..32c5ef2 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -2563,4 +2563,12 @@ a random value. /Ix
 
 /(?P>)/
 
+/(?!\w)(?R)/
+
+/(?=\w)(?R)/
+
+/(?<!\w)(?R)/
+
+/(?<=\w)(?R)/
+
 / End of testinput2 /
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index c7ae26f..72ac0af 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -9346,4 +9346,16 @@ Failed: subpattern name expected at offset 4
 /(?P>)/
 Failed: subpattern name expected at offset 4
 
+/(?!\w)(?R)/
+Failed: recursive call could loop indefinitely at offset 9
+
+/(?=\w)(?R)/
+Failed: recursive call could loop indefinitely at offset 9
+
+/(?<!\w)(?R)/
+Failed: recursive call could loop indefinitely at offset 10
+
+/(?<=\w)(?R)/
+Failed: recursive call could loop indefinitely at offset 10
+
 / End of testinput2 /