JIS0201.h: New file, generated from Unicode table.

�
	* gnu/gcj/convert/JIS0201.h:  New file, generated from Unicode table.
	* gnu/gcj/convert/Input_JavaSrc.java: New BytesToUnicode class.
	* gnu/gcj/convert/Input_SJIS.java: New BytesToUnicode class.
	* gnu/gcj/convert/Output_EUCJIS.java:  New UnicodeToBytes class.
	* gnu/gcj/convert/Output_SJIS.java:  New UnicodeToBytes class.
	* gnu/gcj/convert/natInput_EUCJIS.cc:  New file.
	* gnu/gcj/convert/natInput_SJIS.cc:  New file.
	* gnu/gcj/convert/natOutput_EUCJIS.cc:  New file.
	* gnu/gcj/convert/natOutput_SJIS.cc:  New file.
	* gnu/gcj/convert/make-trie.c:  New file: functions to make a trie.
	* gnu/gcj/convert/gen-from-JIS.c:  Invoke make-trie for output.
	* gnu/gcj/convert/Unicode_to_JIS.cc:  New generated trie table.

From-SVN: r26502

libjava/gnu/gcj/convert/Input_JavaSrc.java

@@ -0,0 +1,89 @@
+/* Copyright (C) 1999  Cygnus Solutions
+
+   This file is part of libgcj.
+
+This software is copyrighted work licensed under the terms of the
+Libgcj License.  Please consult the file "LIBGCJ_LICENSE" for
+details.  */
+
+package gnu.gcj.convert; 
+ 
+/**
+ * Convert Ascii with \ u XXXX-escapes to Unicode.
+ * @author Per Bothner <bothner@cygnus.com>
+ * @date April 1999.
+ */
+
+public class Input_JavaSrc extends BytesToUnicode
+{
+  public String getName() { return "JavaSrc"; }
+
+  // 0: normal
+  // 1: seen '\\'
+  // 2: seen '\\' and 'u'
+  // 3: seen '\\' and need to emit value.
+  // 4, 5, 6, 7:  seen '\\u', 'u' and (state-3) hex digits.
+  int state = 0;
+
+  int value;
+
+  public int read (char[] outbuffer, int outpos, int outlength)
+  {
+    int origpos = outpos;
+    for (;;)
+      {
+	if (inpos >= inlength)
+	  break;
+	if (outpos >= outlength)
+	  break;
+	char b = (char) (inbuffer[inpos++] & 0xFF);
+	switch (state)
+	  {
+	  case 0:
+	    if (b == '\\')
+	      {
+		state = 1;
+		continue;
+	      }
+	    break;
+	  case 1:
+	    if (b == 'u')
+	      {
+		state = 2;
+		continue;
+	      }
+	    if (b != '\\')
+	      {
+		value = b;
+		b = '\\';
+		state = 3;
+	      }
+	    break;
+	  case 3:
+	    b = (char) value;
+	    break;
+	  default:  //  case 4:  case 5:  case 6:  case 7:
+	    int digit = Character.digit(b, 16);
+	    if (digit < 0)
+	      {
+		b = '\uFFFD';
+		state = 0;
+	      }
+	    else
+	      {
+		value = value * 16 + digit;
+		if (state < 7)
+		  {
+		    state++;
+		    continue;
+		  }
+		b = (char) value;
+	      }
+	    state = 0;
+	  }
+	outbuffer[outpos++] = b;
+      }
+    return outpos - origpos;
+  }
+}
+

libjava/gnu/gcj/convert/Output_SJIS.java

@@ -0,0 +1,26 @@
+/* Copyright (C) 1999  Cygnus Solutions
+
+   This file is part of libgcj.
+
+This software is copyrighted work licensed under the terms of the
+Libgcj License.  Please consult the file "LIBGCJ_LICENSE" for
+details.  */
+
+package gnu.gcj.convert;
+
+/**
+ * Convert Unicode to SJIS (Shift JIS, used on Japanese MS-Windows).
+ * @author Per Bothner <bothner@cygnus.com>
+ * @date April 1999.
+ */
+
+public class Output_SJIS extends UnicodeToBytes
+{
+  public String getName() { return "SJIS"; }
+
+  public native int write (char[] inbuffer, int inpos, int inlength);
+
+  public native int write (String str, int inpos, int inlength, char[] work);
+
+  int pending = -1;
+}

libjava/gnu/gcj/convert/natInput_SJIS.cc

@@ -0,0 +1,76 @@
+/* Copyright (C) 1999  Cygnus Solutions
+
+   This file is part of libgcj.
+
+This software is copyrighted work licensed under the terms of the
+Libgcj License.  Please consult the file "LIBGCJ_LICENSE" for
+details.  */
+
+#include <config.h>
+#include <cni.h>
+#include <gnu/gcj/convert/Input_SJIS.h>
+
+#define ERROR_CHAR 0xFFFD
+
+extern unsigned short JIS0208_to_Unicode[84][94];
+extern unsigned short JIS0212_to_Unicode[76][94];
+
+jint
+gnu::gcj::convert::Input_SJIS::read(jcharArray outbuffer, jint outpos,
+				    jint outlength)
+{
+  jint start_outpos = outpos;
+  for (;;)
+    {
+      if (outpos >= outlength)
+	break;
+      if (inpos >= inlength)
+	break;
+      int b = ((unsigned char*) elements(inbuffer))[inpos++];
+      if (first_byte == 0)
+	{
+	  if (b < 128)
+	    {
+#if 1
+	      // Technically, we should translate 0x5c to Yen symbol;
+	      // in practice, it is not clear.
+	      if (b == 0x5c)
+		b = 0x00A5;  // Yen sign.
+#endif
+	      elements(outbuffer)[outpos++] = (char) b;
+	    }
+	  else if (b >= 0xA1 && b <= 0xDF)
+	    {
+	      b += 0xFF61 - 0xA1;
+	      elements(outbuffer)[outpos++] = b;
+	    }
+	  else
+	    first_byte = b;
+	}
+      else
+	{
+	  // From Lunde: "CJKV Informatio Processing", O'Reilly, 1999, p 420:
+	  bool adjust = b < 159;
+	  int rowOffset = first_byte < 160 ? 112 : 176;
+	  int cellOffset = adjust ? (b > 127 ? 32 : 31) : 126;
+	  first_byte = ((first_byte - rowOffset) << 1) - adjust;
+	  b -= cellOffset;
+
+	  first_byte -= 33;
+	  b -= 33;
+
+	  if ((unsigned) first_byte >= 84 || (unsigned) b >= 94)
+	    b = ERROR_CHAR;
+	  else
+	    {
+	      b = JIS0208_to_Unicode[first_byte][b];
+	      if (b == 0)
+		b = ERROR_CHAR;
+	    }
+	  elements(outbuffer)[outpos++] = b;
+
+	  first_byte = 0;
+	}
+    }
+  return outpos - start_outpos;
+}