Fix single-character POSIX class bug in UCP mode.

git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@440 6239d852-aaf2-0410-a92c-79f79f948069

6 files changed, 74 insertions, 11 deletions

diff --git a/ChangeLog b/ChangeLog
index 2472fa0..4e3164f 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -314,6 +314,10 @@ with JIT (possibly caused by SSE2?).
 
 94. Support offset_limit in JIT.
 
+95. A sequence such as [[:punct:]b] that is, a POSIX character class followed 
+by a single ASCII character in a class item, was incorrectly compiled in UCP 
+mode. The POSIX class got lost, but only if the single character followed it.
+
 
 Version 10.20 30-June-2015
 --------------------------
diff --git a/src/pcre2_compile.c b/src/pcre2_compile.c
index c6e84ce..ff7bebd 100644
--- a/src/pcre2_compile.c
+++ b/src/pcre2_compile.c
@@ -1352,7 +1352,7 @@ if ((cb->external_flags & PCRE2_DUPCAPUSED) == 0 &&
 
 /* A large and/or complex regex can take too long to process. We have to assume
 it can match an empty string. This can happen more often when (?| groups are
-present in the pattern and the caching is disabled. Setting the cap at 1100 
+present in the pattern and the caching is disabled. Setting the cap at 1100
 allows the test for more than 1023 capturing patterns to work. */
 
 if ((*countptr)++ > 1100) return CBE_TOOCOMPLICATED;
@@ -4729,16 +4729,20 @@ for (;; ptr++)
       CLASS_SINGLE_CHARACTER:
       if (class_one_char < 2) class_one_char++;
 
-      /* If class_one_char is 1, we have the first single character in the
-      class, and there have been no prior ranges, or XCLASS items generated by
-      escapes. If this is the final character in the class, we can optimize by
-      turning the item into a 1-character OP_CHAR[I] if it's positive, or
-      OP_NOT[I] if it's negative. In the positive case, it can cause firstcu
-      to be set. Otherwise, there can be no first char if this item is first,
-      whatever repeat count may follow. In the case of reqcu, save the
-      previous value for reinstating. */
+      /* If class_one_char is 1 and xclass_has_prop is false, we have the first
+      single character in the class, and there have been no prior ranges, or
+      XCLASS items generated by escapes. If this is the final character in the
+      class, we can optimize by turning the item into a 1-character OP_CHAR[I]
+      if it's positive, or OP_NOT[I] if it's negative. In the positive case, it
+      can cause firstcu to be set. Otherwise, there can be no first char if
+      this item is first, whatever repeat count may follow. In the case of
+      reqcu, save the previous value for reinstating. */
 
-      if (!inescq && class_one_char == 1 && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
+      if (!inescq &&
+#ifdef SUPPORT_UNICODE
+          !xclass_has_prop &&
+#endif
+          class_one_char == 1 && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
         {
         ptr++;
         zeroreqcu = reqcu;
@@ -7287,7 +7291,7 @@ for (;; ptr++)
 
       else
         {
-        if (escape == ESC_C) cb->external_flags |= PCRE2_HASBKC; /* Record */ 
+        if (escape == ESC_C) cb->external_flags |= PCRE2_HASBKC; /* Record */
         if ((escape == ESC_b || escape == ESC_B || escape == ESC_A) &&
              cb->max_lookbehind == 0)
           cb->max_lookbehind = 1;
diff --git a/testdata/testinput2 b/testdata/testinput2
index a6cfd85..dfd1aa8 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -4685,4 +4685,8 @@ a)"xI
 "(*ANYCRLF)(?m)^(.*[^0-9\r\n].*|)$"g,replace=NaN
     15\r\nfoo\r\n20\r\nbar\r\nbaz\r\n\r\n20
 
+/a[[:punct:]b]/bincode
+
+/a[b[:punct:]]/bincode
+
 # End of testinput2 
diff --git a/testdata/testinput5 b/testdata/testinput5
index a288459..2432d19 100644
--- a/testdata/testinput5
+++ b/testdata/testinput5
@@ -1691,4 +1691,10 @@
 
 /abc\Cdef/info,utf
 
+/a[[:punct:]b]/ucp,bincode
+
+/a[[:punct:]b]/utf,ucp,bincode
+
+/a[b[:punct:]]/utf,ucp,bincode
+
 # End of testinput5 
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index 9408756..e5350d5 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -14888,4 +14888,22 @@ Subject length lower bound = 0
     15\r\nfoo\r\n20\r\nbar\r\nbaz\r\n\r\n20
  4: 15\x0d\x0aNaN\x0d\x0a20\x0d\x0aNaN\x0d\x0aNaN\x0d\x0aNaN\x0d\x0a20
 
+/a[[:punct:]b]/bincode
+------------------------------------------------------------------
+        Bra
+        a
+        [!-/:-@[-`b{-~]
+        Ket
+        End
+------------------------------------------------------------------
+
+/a[b[:punct:]]/bincode
+------------------------------------------------------------------
+        Bra
+        a
+        [!-/:-@[-`b{-~]
+        Ket
+        End
+------------------------------------------------------------------
+
 # End of testinput2 
diff --git a/testdata/testoutput5 b/testdata/testoutput5
index ef93100..7eb9df3 100644
--- a/testdata/testoutput5
+++ b/testdata/testoutput5
@@ -4070,4 +4070,31 @@ First code unit = 'a'
 Last code unit = 'f'
 Subject length lower bound = 0
 
+/a[[:punct:]b]/ucp,bincode
+------------------------------------------------------------------
+        Bra
+        a
+        [b[:punct:]]
+        Ket
+        End
+------------------------------------------------------------------
+
+/a[[:punct:]b]/utf,ucp,bincode
+------------------------------------------------------------------
+        Bra
+        a
+        [b[:punct:]]
+        Ket
+        End
+------------------------------------------------------------------
+
+/a[b[:punct:]]/utf,ucp,bincode
+------------------------------------------------------------------
+        Bra
+        a
+        [b[:punct:]]
+        Ket
+        End
+------------------------------------------------------------------
+
 # End of testinput5