Make pcretest generate a single byte for \x{} escapes in non-UTF-8 mode.

git-svn-id: svn://vcs.exim.org/pcre/code/trunk@355 2f5784b3-3f2a-0410-8824-cb99058d5e15

8 files changed, 41 insertions, 13 deletions

diff --git a/ChangeLog b/ChangeLog
index 3e5ccdf..a2dfe8f 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -21,6 +21,11 @@ Version 8.0 02 Jul-08
 4.  Caseless matching was not working for non-ASCII characters in back 
     references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
     It now works when Unicode Property Support is available. 
+    
+5.  In pcretest, an escape such as \x{de} in the data was always generating
+    a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
+    non-UTF-8 mode. If the value is greater than 255, it gives a warning about
+    truncation.   
 
 
 Version 7.7 07-May-08
diff --git a/pcretest.c b/pcretest.c
index ac8290d..31d0357 100644
--- a/pcretest.c
+++ b/pcretest.c
@@ -1806,9 +1806,19 @@ while (!done)
             {
             unsigned char buff8[8];
             int ii, utn;
-            utn = ord2utf8(c, buff8);
-            for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
-            c = buff8[ii];   /* Last byte */
+            if (use_utf8)
+              { 
+              utn = ord2utf8(c, buff8);
+              for (ii = 0; ii < utn - 1; ii++) *q++ = buff8[ii];
+              c = buff8[ii];   /* Last byte */
+              }
+            else
+             {
+             if (c > 255) 
+               fprintf(outfile, "** Character \\x{%x} is greater than 255 and "
+                 "UTF-8 mode is not enabled.\n"
+                 "** Truncation will probably give the wrong result.\n", c);
+             }      
             p = pt + 1;
             break;
             }
diff --git a/testdata/testinput2 b/testdata/testinput2
index 8536d66..2b64546 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -1988,10 +1988,10 @@ a random value. /Ix
     a\rb\<anycrlf>
 
 /^abc./mgx<any>
-    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 \x{2028}abc8 \x{2029}abc9 JUNK
+    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 JUNK
 
 /abc.$/mgx<any>
-    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7\x{2028} abc8\x{2029} abc9
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7 abc9
 
 /a/<cr><any>
 
@@ -2147,7 +2147,7 @@ a random value. /Ix
     abc\r\n\r\n
 
 /abc.$/mgx<anycrlf>
-    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7\x{2028} abc8\x{2029} abc9
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc9
 
 /^X/m
     XABC
diff --git a/testdata/testinput5 b/testdata/testinput5
index 8a8e499..62e9d1e 100644
--- a/testdata/testinput5
+++ b/testdata/testinput5
@@ -473,4 +473,8 @@ can't tell the difference.) --/
     ** Failers
     ab  
 
+/(\x{de})\1/
+    \x{de}\x{de}
+    \x{123} 
+
 / End of testinput5 /
diff --git a/testdata/testinput7 b/testdata/testinput7
index 5d593ee..5ec4827 100644
--- a/testdata/testinput7
+++ b/testdata/testinput7
@@ -4151,10 +4151,10 @@
     a\rb\<any>   
 
 /^abc./mgx<any>
-    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 \x{2028}abc8 \x{2029}abc9 JUNK
+    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 JUNK
 
 /abc.$/mgx<any>
-    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7\x{2028} abc8\x{2029} abc9
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc9
 
 /^a\Rb/<bsr_unicode>
     a\nb
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index 5dd2b4c..2ac018b 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -7851,7 +7851,7 @@ No match
 No match
 
 /^abc./mgx<any>
-    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 \x{2028}abc8 \x{2029}abc9 JUNK
+    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 JUNK
  0: abc1
  0: abc2
  0: abc3
@@ -7861,7 +7861,7 @@ No match
  0: abc7
 
 /abc.$/mgx<any>
-    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7\x{2028} abc8\x{2029} abc9
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7 abc9
  0: abc1
  0: abc2
  0: abc3
@@ -8163,7 +8163,7 @@ No match
  0+ 
 
 /abc.$/mgx<anycrlf>
-    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7\x{2028} abc8\x{2029} abc9
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc9
  0: abc1
  0: abc4
  0: abc5
diff --git a/testdata/testoutput5 b/testdata/testoutput5
index 1c745f4..d36a246 100644
--- a/testdata/testoutput5
+++ b/testdata/testoutput5
@@ -1628,4 +1628,13 @@ No match
     ab  
 No match
 
+/(\x{de})\1/
+    \x{de}\x{de}
+ 0: \xde\xde
+ 1: \xde
+    \x{123} 
+** Character \x{123} is greater than 255 and UTF-8 mode is not enabled.
+** Truncation will probably give the wrong result.
+No match
+
 / End of testinput5 /
diff --git a/testdata/testoutput7 b/testdata/testoutput7
index 9ded29d..aef4b6c 100644
--- a/testdata/testoutput7
+++ b/testdata/testoutput7
@@ -6805,7 +6805,7 @@ No match
 No match
 
 /^abc./mgx<any>
-    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 \x{2028}abc8 \x{2029}abc9 JUNK
+    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x85abc7 JUNK
  0: abc1
  0: abc2
  0: abc3
@@ -6815,7 +6815,7 @@ No match
  0: abc7
 
 /abc.$/mgx<any>
-    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc7\x{2028} abc8\x{2029} abc9
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x85 abc9
  0: abc1
  0: abc2
  0: abc3