tweak when we decide to use hex strings vs literal strings

git-svn-id: svn+q:///qpdf/trunk@810 71b93d88-0707-0410-a8cf-f5a4172ac649

tweak when we decide to use hex strings vs literal strings
git-svn-id: svn+q:///qpdf/trunk@810 71b93d88-0707-0410-a8cf-f5a4172ac649
Jay Berkenbilt
1 parent 55e40037
Showing 7 changed files with 26 additions and 24 deletions
libqpdf/QPDF_String.cc
qpdf/qtest/qpdf/encrypted1.out
qpdf/qtest/qpdf/good13.out
qpdf/qtest/qpdf/good13.qdf
qpdf/qtest/qpdf/good14.out
qpdf/qtest/qpdf/good9.qdf
qpdf/qtest/qpdf/show-xref-by-id.out
@@ -9,6 +9,10 @@
 #include <string.h>
 // See above about ctype.
+static bool is_ascii_printable(unsigned char ch)
+{
+    return ((ch >= 32) && (ch <= 126));
+}
 static bool is_iso_latin1_printable(unsigned char ch)
 {
     return (((ch >= 32) && (ch <= 126)) || (ch >= 160));
@@ -40,12 +44,13 @@ QPDF_String::unparse(bool force_binary)
 	for (unsigned int i = 0; i < this->val.length(); ++i)
 	{
 	    char ch = this->val[i];
-	    // Note: do not use locale to determine printability.  The PDF
-	    // specification accepts arbitrary binary data.  Some locales
-	    // imply multibyte characters.  We'll consider something
-	    // printable if it is printable in ISO-Latin-1.  We'll code
-	    // this manually rather than being rude and setting locale.
-	    if ((ch == 0) || (! (is_iso_latin1_printable(ch) ||
+	    // Note: do not use locale to determine printability.  The
+	    // PDF specification accepts arbitrary binary data.  Some
+	    // locales imply multibyte characters.  We'll consider
+	    // something printable if it is printable in 7-bit ASCII.
+	    // We'll code this manually rather than being rude and
+	    // setting locale.
+	    if ((ch == 0) || (! (is_ascii_printable(ch) ||
 				 strchr("\n\r\t\b\f", ch))))
 	    {
 		++nonprintable;
@@ -64,10 +69,7 @@ QPDF_String::unparse(bool force_binary)
 	}
 	// Use hex notation if more than 20% of the characters are not
-	// printable in the current locale.  Uniformly distributed random
-	// characters will not pass this test even with ISO-Latin-1 in
-	// which 76% are either printable or in the set of standard
-	// escaped characters.
+	// printable in plain ASCII.
 	if (5 * nonprintable > val.length())
 	{
 	    use_hexstring = true;
@@ -5,5 +5,5 @@
   /nesting is direct
   /strings is direct
 unparse: 7 0 R
-unparseResolved: << /hex#20strings [ (Potato) <01020300040560> (AB) ] /indirect 8 0 R /nesting << /a [ 1 2 << /x (y) >> [ (z) ] ] /b << / (legal) /a [ 1 2 ] >> >> /strings [ (one) ($¢) () (\(\)) (\() (\)) (a\f\b\t\r\nb) <410042> (a\nb) (a b) ] >>
+unparseResolved: << /hex#20strings [ (Potato) <01020300040560> (AB) ] /indirect 8 0 R /nesting << /a [ 1 2 << /x (y) >> [ (z) ] ] /b << / (legal) /a [ 1 2 ] >> >> /strings [ (one) <24a2> () (\(\)) (\() (\)) (a\f\b\t\r\nb) <410042> (a\nb) (a b) ] >>
 test 1 done
@@ -40,7 +40,7 @@ endobj
   >>
   /strings [
     (one)
-    ($¢)
+    <24a2>
     ()
     (\(\))
     (\()
@@ -129,12 +129,12 @@ xref
 0000000000 65535 f 
 0000000052 00000 n 
 0000000133 00000 n 
-0000000576 00000 n 
-0000000685 00000 n 
-0000000927 00000 n 
-0000001026 00000 n 
-0000001072 00000 n 
-0000001217 00000 n 
+0000000578 00000 n 
+0000000687 00000 n 
+0000000929 00000 n 
+0000001028 00000 n 
+0000001074 00000 n 
+0000001219 00000 n 
 trailer <<
   /QTest 2 0 R
   /Root 1 0 R
@@ -142,5 +142,5 @@ trailer &lt;&lt;
   /ID [<31415926535897932384626433832795><31415926535897932384626433832795>]
 >>
 startxref
-1252
+1254
 %%EOF
@@ -12,7 +12,7 @@ three lines
 (\001B%DEF)<01>
 <8a8b>
 (ab)
-<8c>(Ý) ) >
+<8c><dd> ) >
 <610062> (MOO)
 -- stream 1 --
 This stream does end with a newline.
@@ -95,7 +95,7 @@ trailer &lt;&lt;
   /QTest (¡Hola!)
   /Root 1 0 R
   /Size 8
-  /Z ('\236yÔ\005\037ãjø×¸B^Q\n\2139\224\rD|kã¢JZm:½l\231\002\\)
+  /Z <279e79d4051fe36af8d7b8425e510a8b39940d447c6be3a24a5a6d3abd6c99025c>
   /ID [<31415926535897932384626433832795><31415926535897932384626433832795>]
 >>
 startxref
 Object is stream.  Dictionary:
-<< /DecodeParms << /Columns 4 /Predictor 12 >> /Encrypt 11 0 R /Filter /FlateDecode /ID [ (ª&\237þÂ\226±?>ø5ª¡:\n\b) (ím\023\021OµUt\bµ\026Ê9'\025\033) ] /Info 3 0 R /Length 52 /Root 1 0 R /Size 13 /Type /XRef /W [ 1 2 1 ] >>
+<< /DecodeParms << /Columns 4 /Predictor 12 >> /Encrypt 11 0 R /Filter /FlateDecode /ID [ <aa269ffec296b13f3ef835aaa13a0a08> <ed6d13114fb5557408b516ca3927151b> ] /Info 3 0 R /Length 52 /Root 1 0 R /Size 13 /Type /XRef /W [ 1 2 1 ] >>