X-Git-Url: http://git.chise.org/gitweb/?p=chise%2Fxemacs-chise.git.1;a=blobdiff_plain;f=src%2Ffile-coding.h;h=6d673fa7a92a882543f6695243f9f81603620790;hp=69345b3a33f0d31bc17ffab9b9fe367516c9d948;hb=566b3d194e2d5c783808ac39437bd7e1a28b1c5c;hpb=716cfba952c1dc0d2cf5c968971f3780ba728a89

diff --git a/src/file-coding.h b/src/file-coding.h
index 69345b3..6d673fa 100644
--- a/src/file-coding.h
+++ b/src/file-coding.h
@@ -1,6 +1,7 @@
 /* Header for code conversion stuff
    Copyright (C) 1991, 1995 Free Software Foundation, Inc.
    Copyright (C) 1995 Sun Microsystems, Inc.
+   Copyright (C) 1999,2000,2002,2005,2012 MORIOKA Tomohiko
 
 This file is part of XEmacs.
 
@@ -44,6 +45,9 @@ enum coding_system_type
   CODESYS_BIG5,		/* BIG5 (used for Taiwanese). */
   CODESYS_UCS4,		/* ISO 10646 UCS-4 */
   CODESYS_UTF8,		/* ISO 10646 UTF-8 */
+#ifdef UTF2000
+  CODESYS_UTF16,	/* ISO 10646 UTF-16 */
+#endif
   CODESYS_CCL,		/* Converter written in CCL. */
 #endif
   CODESYS_NO_CONVERSION	/* "No conversion"; used for binary files.
@@ -138,6 +142,12 @@ struct Lisp_Coding_System
     Lisp_Object encode;
   } ccl;
 #endif
+#ifdef UTF2000
+  unsigned int disable_composition	:1;
+  unsigned int enable_decomposition	:1;
+  unsigned int use_entity_reference	:1;
+  Lisp_Object ccs_priority_list;
+#endif
 };
 typedef struct Lisp_Coding_System Lisp_Coding_System;
 
@@ -181,6 +191,16 @@ DECLARE_LRECORD (coding_system, Lisp_Coding_System);
 #define CODING_SYSTEM_CCL_DECODE(codesys) ((codesys)->ccl.decode)
 #define CODING_SYSTEM_CCL_ENCODE(codesys) ((codesys)->ccl.encode)
 #endif /* MULE */
+#ifdef UTF2000
+#define CODING_SYSTEM_DISABLE_COMPOSITION(codesys) \
+  ((codesys)->disable_composition)
+#define CODING_SYSTEM_ENABLE_DECOMPOSITION(codesys) \
+  ((codesys)->enable_decomposition)
+#define CODING_SYSTEM_USE_ENTITY_REFERENCE(codesys) \
+  ((codesys)->use_entity_reference)
+#define CODING_SYSTEM_CCS_PRIORITY_LIST(codesys) \
+  ((codesys)->ccs_priority_list)
+#endif
 
 #define XCODING_SYSTEM_NAME(codesys) \
   CODING_SYSTEM_NAME (XCODING_SYSTEM (codesys))
@@ -227,6 +247,16 @@ DECLARE_LRECORD (coding_system, Lisp_Coding_System);
 #define XCODING_SYSTEM_CCL_ENCODE(codesys) \
   CODING_SYSTEM_CCL_ENCODE (XCODING_SYSTEM (codesys))
 #endif /* MULE */
+#ifdef UTF2000
+#define XCODING_SYSTEM_DISABLE_COMPOSITION(codesys) \
+  CODING_SYSTEM_DISABLE_COMPOSITION (XCODING_SYSTEM (codesys))
+#define XCODING_SYSTEM_ENABLE_DECOMPOSITION(codesys) \
+  CODING_SYSTEM_ENABLE_DECOMPOSITION (XCODING_SYSTEM (codesys))
+#define XCODING_SYSTEM_USE_ENTITY_REFERENCE(codesys) \
+  CODING_SYSTEM_USE_ENTITY_REFERENCE (XCODING_SYSTEM (codesys))
+#define XCODING_SYSTEM_CCS_PRIORITY_LIST(codesys) \
+  CODING_SYSTEM_CCS_PRIORITY_LIST (XCODING_SYSTEM (codesys))
+#endif
 
 EXFUN (Fcoding_category_list, 0);
 EXFUN (Fcoding_category_system, 1);
@@ -408,32 +438,51 @@ enum iso_esc_flag
 #define ISO_CODE_CSI	0x9B		/* control-sequence-introduce */
 #endif /* MULE */
 
-/* For detecting the encoding of text */
+/* Distinguishable categories of encodings.
+
+   This list determines the initial priority of the categories.
+
+   For better or worse, currently Mule files are encoded in 7-bit ISO 2022.
+   For this reason, under Mule ISO_7 gets highest priority.
+
+   Putting NO_CONVERSION second prevents "binary corruption" in the
+   default case in all but the (presumably) extremely rare case of a
+   binary file which contains redundant escape sequences but no 8-bit
+   characters.
+
+   The remaining priorities are based on perceived "internationalization
+   political correctness."  An exception is UCS-4 at the bottom, since
+   basically everything is compatible with UCS-4, but it is likely to
+   be very rare as an external encoding. */
+
 enum coding_category_type
 {
+  /* must be a contiguous range of values 0 -- CODING_CATEGORY_LAST - 1 */
 #ifdef MULE
-  CODING_CATEGORY_SHIFT_JIS,
   CODING_CATEGORY_ISO_7, /* ISO2022 system using only seven-bit bytes,
 			    no locking shift */
-  CODING_CATEGORY_ISO_8_DESIGNATE, /* ISO2022 system using eight-bit bytes,
-				      no locking shift, no single shift,
-				      using designation to switch charsets */
+  CODING_CATEGORY_NO_CONVERSION,
+  CODING_CATEGORY_UTF8,
   CODING_CATEGORY_ISO_8_1, /* ISO2022 system using eight-bit bytes,
 			      no locking shift, no designation sequences,
 			      one-dimension characters in the upper half. */
   CODING_CATEGORY_ISO_8_2, /* ISO2022 system using eight-bit bytes,
 			      no locking shift, no designation sequences,
 			      two-dimension characters in the upper half. */
+  CODING_CATEGORY_ISO_8_DESIGNATE, /* ISO2022 system using eight-bit bytes,
+				      no locking shift, no single shift,
+				      using designation to switch charsets */
   CODING_CATEGORY_ISO_LOCK_SHIFT, /* ISO2022 system using locking shift */
+  CODING_CATEGORY_SHIFT_JIS,
   CODING_CATEGORY_BIG5,
+  CODING_CATEGORY_UTF16,
   CODING_CATEGORY_UCS4,
-  CODING_CATEGORY_UTF8,
+#else /* not MULE */
+  CODING_CATEGORY_NO_CONVERSION,
 #endif /* MULE */
-  CODING_CATEGORY_NO_CONVERSION
+  CODING_CATEGORY_LAST		/* not a real coding category */
 };
 
-#define CODING_CATEGORY_LAST CODING_CATEGORY_NO_CONVERSION
-
 #ifdef MULE
 #define CODING_CATEGORY_SHIFT_JIS_MASK	\
   (1 << CODING_CATEGORY_SHIFT_JIS)
@@ -451,6 +500,8 @@ enum coding_category_type
   (1 << CODING_CATEGORY_BIG5)
 #define CODING_CATEGORY_UCS4_MASK \
   (1 << CODING_CATEGORY_UCS4)
+#define CODING_CATEGORY_UTF16_MASK \
+  (1 << CODING_CATEGORY_UTF16)
 #define CODING_CATEGORY_UTF8_MASK \
   (1 << CODING_CATEGORY_UTF8)
 #endif