Fix \C bug with repeated character classes in UTF-8 mode.

git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@918 6239d852-aaf2-0410-a92c-79f79f948069

6 files changed, 32 insertions, 4 deletions

diff --git a/ChangeLog b/ChangeLog
index 8651e6a..17e7966 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -20,6 +20,11 @@ Unicode newlines" in the default case when --enable-bsr-anycrlf has not been
 specified. Similarly, running "pcfre2test -C bsr" never produced the result 
 ANY.
 
+4. Matching the pattern /(*UTF)\C[^\v]+\x80/ against an 8-bit string containing 
+multi-code-unit characters caused bad behaviour and possibly a crash. This 
+issue was fixed for other kinds of repeat in release 10.20 by change 19, but
+repeating character classes were overlooked.
+
 
 Version 10.31 12-February-2018
 ------------------------------
diff --git a/src/pcre2_match.c b/src/pcre2_match.c
index 79cc93f..ce96016 100644
--- a/src/pcre2_match.c
+++ b/src/pcre2_match.c
@@ -1962,11 +1962,15 @@ fprintf(stderr, "++ op=%d\n", *Fecode);
 
           if (reptype == REPTYPE_POS) continue;    /* No backtracking */
 
+          /* After \C in UTF mode, Lstart_eptr might be in the middle of a
+          Unicode character. Use <= Lstart_eptr to ensure backtracking doesn't
+          go too far. */
+
           for (;;)
             {
             RMATCH(Fecode, RM201);
             if (rrc != MATCH_NOMATCH) RRETURN(rrc);
-            if (Feptr-- == Lstart_eptr) break;  /* Tried at original position */
+            if (Feptr-- <= Lstart_eptr) break;  /* Tried at original position */
             BACKCHAR(Feptr);
             }
           }
@@ -2126,11 +2130,15 @@ fprintf(stderr, "++ op=%d\n", *Fecode);
 
         if (reptype == REPTYPE_POS) continue;    /* No backtracking */
 
+        /* After \C in UTF mode, Lstart_eptr might be in the middle of a
+        Unicode character. Use <= Lstart_eptr to ensure backtracking doesn't
+        go too far. */
+
         for(;;)
           {
           RMATCH(Fecode, RM101);
           if (rrc != MATCH_NOMATCH) RRETURN(rrc);
-          if (Feptr-- == Lstart_eptr) break;  /* Tried at original position */
+          if (Feptr-- <= Lstart_eptr) break;  /* Tried at original position */
 #ifdef SUPPORT_UNICODE
           if (utf) BACKCHAR(Feptr);
 #endif
@@ -4002,8 +4010,8 @@ fprintf(stderr, "++ op=%d\n", *Fecode);
         if (reptype == REPTYPE_POS) continue;    /* No backtracking */
 
         /* After \C in UTF mode, Lstart_eptr might be in the middle of a
-        Unicode character. Use <= pp to ensure backtracking doesn't go too far.
-        */
+        Unicode character. Use <= Lstart_eptr to ensure backtracking doesn't
+        go too far. */
 
         for(;;)
           {
diff --git a/testdata/testinput22 b/testdata/testinput22
index e6d4053..c218ea6 100644
--- a/testdata/testinput22
+++ b/testdata/testinput22
@@ -98,4 +98,7 @@
 \= Expect no match - tests \C at end of subject
     ab
 
+/\C[^\v]+\x80/utf
+    [AΏBŀC]
+
 # End of testinput22
diff --git a/testdata/testoutput22-16 b/testdata/testoutput22-16
index 88f827c..5e23611 100644
--- a/testdata/testoutput22-16
+++ b/testdata/testoutput22-16
@@ -171,4 +171,8 @@ No match
     ab
 No match
 
+/\C[^\v]+\x80/utf
+    [AΏBŀC]
+No match
+
 # End of testinput22
diff --git a/testdata/testoutput22-32 b/testdata/testoutput22-32
index ac485fc..8576f31 100644
--- a/testdata/testoutput22-32
+++ b/testdata/testoutput22-32
@@ -169,4 +169,8 @@ No match
     ab
 No match
 
+/\C[^\v]+\x80/utf
+    [AΏBŀC]
+No match
+
 # End of testinput22
diff --git a/testdata/testoutput22-8 b/testdata/testoutput22-8
index 3d31fbc..8543652 100644
--- a/testdata/testoutput22-8
+++ b/testdata/testoutput22-8
@@ -173,4 +173,8 @@ No match
     ab
 No match
 
+/\C[^\v]+\x80/utf
+    [AΏBŀC]
+No match
+
 # End of testinput22