* transcode_data.h (TRANSCODE_ERROR): removed.

* tool/transcode-tblgen.rb: 8bit byte of ASCII-8BIT is a valid
  (but unique to ASCII-8BIT) character.

* transcode.c (rb_eConversionUndefined): new error.
  (rb_eInvalidByteSequence): ditto.


git-svn-id: http://svn.ruby-lang.org/repos/ruby/trunk@18524 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

5 files changed, 53 insertions, 38 deletions

diff --git a/ChangeLog b/ChangeLog
index ba3bc6e17..0edaf761c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,13 @@
+Tue Aug 12 16:13:45 2008  Tanaka Akira  <akr@fsij.org>
+
+	* transcode_data.h (TRANSCODE_ERROR): removed.
+
+	* tool/transcode-tblgen.rb: 8bit byte of ASCII-8BIT is a valid
+	  (but unique to ASCII-8BIT) character.
+
+	* transcode.c (rb_eConversionUndefined): new error.
+	  (rb_eInvalidByteSequence): ditto.
+
 Tue Aug 12 16:08:51 2008  Minero Aoki  <aamine@loveruby.net>
 
 	* lib/net/pop.rb: POP3Command should export @socket since POP
diff --git a/test/ruby/test_transcode.rb b/test/ruby/test_transcode.rb
index 5d3827452..6ad1b0920 100644
--- a/test/ruby/test_transcode.rb
+++ b/test/ruby/test_transcode.rb
@@ -25,9 +25,9 @@ class TestTranscode < Test::Unit::TestCase
     assert_raise(ArgumentError) { 'abc'.encode!('foo', 'bar') }
     assert_raise(ArgumentError) { 'abc'.force_encoding('utf-8').encode('foo') }
     assert_raise(ArgumentError) { 'abc'.force_encoding('utf-8').encode!('foo') }
-    assert_raise(RuntimeError) { "\x80".encode('utf-8','ASCII-8BIT') }
-    assert_raise(RuntimeError) { "\x80".encode('utf-8','US-ASCII') }
-    assert_raise(RuntimeError) { "\xA5".encode('utf-8','iso-8859-3') }
+    assert_raise(Encoding::ConversionUndefined) { "\x80".encode('utf-8','ASCII-8BIT') }
+    assert_raise(Encoding::InvalidByteSequence) { "\x80".encode('utf-8','US-ASCII') }
+    assert_raise(Encoding::ConversionUndefined) { "\xA5".encode('utf-8','iso-8859-3') }
   end
 
   def test_arguments
@@ -342,40 +342,39 @@ class TestTranscode < Test::Unit::TestCase
     check_both_ways("\u71FC", "\xE0\x9E", 'shift_jis') # 
     check_both_ways("\u71F9", "\xE0\x9F", 'shift_jis') # 
     check_both_ways("\u73F1", "\xE0\xFC", 'shift_jis') # 
-    assert_raise(RuntimeError) { "\xEF\x40".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xEF\x7E".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xEF\x80".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xEF\x9E".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xEF\x9F".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xEF\xFC".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xF0\x40".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xF0\x7E".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xF0\x80".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xF0\x9E".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xF0\x9F".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xF0\xFC".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xEF\x40".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xEF\x7E".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xEF\x80".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xEF\x9E".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xEF\x9F".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xEF\xFC".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xF0\x40".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xF0\x7E".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xF0\x80".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xF0\x9E".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xF0\x9F".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xF0\xFC".encode("utf-8", 'shift_jis') }
     check_both_ways("\u9ADC", "\xFC\x40", 'shift_jis') # @
-    assert_raise(RuntimeError) { "\xFC\x7E".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xFC\x80".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xFC\x9E".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xFC\x9F".encode("utf-8", 'shift_jis') }
-    assert_raise(RuntimeError) { "\xFC\xFC".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xFC\x7E".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xFC\x80".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xFC\x9E".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xFC\x9F".encode("utf-8", 'shift_jis') }
+    assert_raise(Encoding::ConversionUndefined) { "\xFC\xFC".encode("utf-8", 'shift_jis') }
     check_both_ways("\u677E\u672C\u884C\u5F18", "\x8f\xbc\x96\x7b\x8d\x73\x8d\x4f", 'shift_jis') # {sO
     check_both_ways("\u9752\u5C71\u5B66\u9662\u5927\u5B66", "\x90\xC2\x8E\x52\x8A\x77\x89\x40\x91\xE5\x8A\x77", 'shift_jis') # ŽRw@w
     check_both_ways("\u795E\u6797\u7FA9\u535A", "\x90\x5F\x97\xD1\x8B\x60\x94\x8E", 'shift_jis') # _ы`
   end
 
   def test_iso_2022_jp
-    assert_raise(RuntimeError) { "\x1b(A".encode("utf-8", "iso-2022-jp") }
-    assert_raise(RuntimeError) { "\x1b$(A".encode("utf-8", "iso-2022-jp") }
-    assert_raise(RuntimeError) { "\x1b$C".encode("utf-8", "iso-2022-jp") }
-    assert_raise(RuntimeError) { "\x0e".encode("utf-8", "iso-2022-jp") }
-    assert_raise(RuntimeError) { "\x80".encode("utf-8", "iso-2022-jp") }
-    assert_raise(RuntimeError) { "\x1b$(Dd!\x1b(B".encode("utf-8", "iso-2022-jp") }
-    assert_raise(RuntimeError) { "\u9299".encode("iso-2022-jp") }
-    assert_raise(RuntimeError) { "\u9299".encode("iso-2022-jp") }
-    assert_raise(RuntimeError) { "\uff71\uff72\uff73\uff74\uff75".encode("iso-2022-jp") }
-    assert_raise(RuntimeError) { "\x1b(I12345\x1b(B".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x1b(A".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x1b$(A".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x1b$C".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x0e".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x80".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x1b$(Dd!\x1b(B".encode("utf-8", "iso-2022-jp") }
+    assert_raise(Encoding::ConversionUndefined) { "\u9299".encode("iso-2022-jp") }
+    assert_raise(Encoding::ConversionUndefined) { "\uff71\uff72\uff73\uff74\uff75".encode("iso-2022-jp") }
+    assert_raise(Encoding::InvalidByteSequence) { "\x1b(I12345\x1b(B".encode("utf-8", "iso-2022-jp") }
     assert_equal("\xA1\xA1".force_encoding("euc-jp"),
                  "\e$B!!\e(B".encode("EUC-JP", "ISO-2022-JP"))
     assert_equal("\e$B!!\e(B".force_encoding("ISO-2022-JP"),
diff --git a/tool/transcode-tblgen.rb b/tool/transcode-tblgen.rb
index 493e3b11a..e4ee481f4 100644
--- a/tool/transcode-tblgen.rb
+++ b/tool/transcode-tblgen.rb
@@ -317,7 +317,8 @@ class ActionMap
       min = max = 0
     end
 
-    if n = OffsetsMemo[offsets]
+    offsets_key = [min, max, offsets[min..max]]
+    if n = OffsetsMemo[offsets_key]
       offsets_name = n
       offsets_code = ''
     else
@@ -326,7 +327,7 @@ class ActionMap
 static const unsigned char
 #{offsets_name}[#{2+max-min+1}] = #{format_offsets(min,max,offsets)};
 End
-      OffsetsMemo[offsets] = offsets_name
+      OffsetsMemo[offsets_key] = offsets_name
     end
 
     if n = InfosMemo[infos]
@@ -524,6 +525,7 @@ ValidEncoding = {
 }
 
 {
+  'ASCII-8BIT'  => '1byte',
   'ISO-8859-1'  => '1byte',
   'ISO-8859-2'  => '1byte',
   'ISO-8859-3'  => '1byte',
diff --git a/transcode.c b/transcode.c
index 1802552a4..df198b119 100644
--- a/transcode.c
+++ b/transcode.c
@@ -15,6 +15,9 @@
 #include "transcode_data.h"
 #include <ctype.h>
 
+VALUE rb_eConversionUndefined;
+VALUE rb_eInvalidByteSequence;
+
 static VALUE sym_invalid, sym_undef, sym_ignore, sym_replace;
 #define INVALID_IGNORE 0x1
 #define INVALID_REPLACE 0x2
@@ -895,7 +898,7 @@ resume:
             goto resume;
 	}
         rb_trans_close(ts);
-	rb_raise(TRANSCODE_ERROR, "invalid byte sequence");
+	rb_raise(rb_eInvalidByteSequence, "invalid byte sequence");
     }
     if (ret == transcode_undefined_conversion) {
 	/* valid character in from encoding
@@ -909,7 +912,7 @@ resume:
 	    goto resume;
 	}
         rb_trans_close(ts);
-        rb_raise(TRANSCODE_ERROR, "conversion undefined for byte sequence (maybe invalid byte sequence)");
+        rb_raise(rb_eConversionUndefined, "conversion undefined for byte sequence (maybe invalid byte sequence)");
     }
     if (ret == transcode_obuf_full) {
         more_output_buffer(destination, resize_destination, ts, &out_start, out_pos, &out_stop);
@@ -974,7 +977,7 @@ transcode_loop(const unsigned char **in_pos, unsigned char **out_pos,
                 break;
             }
             rb_trans_close(ts);
-            rb_raise(TRANSCODE_ERROR, "invalid byte sequence");
+            rb_raise(rb_eInvalidByteSequence, "invalid byte sequence");
             break;
 
           case transcode_undefined_conversion:
@@ -989,7 +992,7 @@ transcode_loop(const unsigned char **in_pos, unsigned char **out_pos,
                 break;
             }
             rb_trans_close(ts);
-            rb_raise(TRANSCODE_ERROR, "conversion undefined for byte sequence (maybe invalid byte sequence)");
+            rb_raise(rb_eConversionUndefined, "conversion undefined for byte sequence (maybe invalid byte sequence)");
             break;
 
           case transcode_obuf_full:
@@ -1204,6 +1207,9 @@ rb_str_transcode(VALUE str, VALUE to)
 void
 Init_transcode(void)
 {
+    rb_eConversionUndefined = rb_define_class_under(rb_cEncoding, "ConversionUndefined", rb_eStandardError);
+    rb_eInvalidByteSequence = rb_define_class_under(rb_cEncoding, "InvalidByteSequence", rb_eStandardError);
+
     transcoder_table = st_init_strcasetable();
 
     sym_invalid = ID2SYM(rb_intern("invalid"));
diff --git a/transcode_data.h b/transcode_data.h
index 24c88fc89..4a0199df0 100644
--- a/transcode_data.h
+++ b/transcode_data.h
@@ -125,6 +125,4 @@ typedef struct {
 void rb_declare_transcoder(const char *enc1, const char *enc2, const char *lib);
 void rb_register_transcoder(const rb_transcoder *);
 
-#define TRANSCODE_ERROR rb_eRuntimeError /*change exception*/
-
 #endif /* RUBY_TRANSCODE_DATA_H */