Silly of me to overlook it: another obvious way you might like to

[sgt/charset] / iso2022.c

diff --git a/iso2022.c b/iso2022.c
index 0990a77..6e527dd 100644 (file)
--- a/iso2022.c
+++ b/iso2022.c
@@ -2,11 +2,10 @@
  * iso2022.c - support for ISO/IEC 2022 (alias ECMA-35).
  *
  * This isn't a complete implementation of ISO/IEC 2022, but it's
- * close.  It only handles decoding, because a fully general encoder
- * isn't really useful.  It can decode 8-bit and 7-bit versions, with
- * support for single-byte and multi-byte character sets, all four
- * containers (G0, G1, G2, and G3), using both single-shift and
- * locking-shift sequences.
+ * close.  It can decode 8-bit and 7-bit versions, with support for
+ * single-byte and multi-byte character sets, all four containers
+ * (G0, G1, G2, and G3), using both single-shift and locking-shift
+ * sequences.
  *
  * The general principle is that any valid ISO/IEC 2022 sequence
  * should either be correctly decoded or should emit an ERROR.  The
@@ -82,8 +81,7 @@ typedef int (*to_dbcs_planar_t)(long int, int *, int *, int *);
  * 
  * We are permitted to use ?:, however, and that works quite well
  * since the actual result of the sizeof expression _is_ evaluable
- * at compile time. So here's my final answer, with the unfortunate
- * remaining problem of evaluating its arguments multiple times:
+ * at compile time. So here's my final answer:
  */
 #define TYPECHECK(x,y) ( sizeof((x)) == sizeof((x)) ? (y) : (y) )
 #define DEPLANARISE(x) TYPECHECK((x) == (to_dbcs_planar_t)NULL, (to_dbcs_t)(x))
@@ -138,18 +136,18 @@ const struct iso2022_subcharset {
      * text, we'll use a preference order which matches that. So we
      * begin with the charsets defined in the compound text spec.
      */
-    { S4, 0, 'B', CCS, 0x00, &sbcsdata_CS_ASCII },
-    { S6, 0, 'A', CCS, 0x80, &sbcsdata_CS_ISO8859_1 },
-    { S6, 0, 'B', CCS, 0x80, &sbcsdata_CS_ISO8859_2 },
-    { S6, 0, 'C', CCS, 0x80, &sbcsdata_CS_ISO8859_3 },
-    { S6, 0, 'D', CCS, 0x80, &sbcsdata_CS_ISO8859_4 },
-    { S6, 0, 'F', CCS, 0x80, &sbcsdata_CS_ISO8859_7 },
-    { S6, 0, 'G', CCS, 0x80, &sbcsdata_CS_ISO8859_6 },
-    { S6, 0, 'H', CCS, 0x80, &sbcsdata_CS_ISO8859_8 },
-    { S6, 0, 'L', CCS, 0x80, &sbcsdata_CS_ISO8859_5 },
-    { S6, 0, 'M', CCS, 0x80, &sbcsdata_CS_ISO8859_9 },
-    { S4, 0, 'I', CCS, 0x80, &sbcsdata_CS_JISX0201 },
-    { S4, 0, 'J', CCS, 0x00, &sbcsdata_CS_JISX0201 },
+    { S4, 0, 'B', CCS, 0x00, &sbcsdata_CS_ASCII, NULL, NULL, 0 },
+    { S6, 0, 'A', CCS, 0x80, &sbcsdata_CS_ISO8859_1, NULL, NULL, 0 },
+    { S6, 0, 'B', CCS, 0x80, &sbcsdata_CS_ISO8859_2, NULL, NULL, 0 },
+    { S6, 0, 'C', CCS, 0x80, &sbcsdata_CS_ISO8859_3, NULL, NULL, 0 },
+    { S6, 0, 'D', CCS, 0x80, &sbcsdata_CS_ISO8859_4, NULL, NULL, 0 },
+    { S6, 0, 'F', CCS, 0x80, &sbcsdata_CS_ISO8859_7, NULL, NULL, 0 },
+    { S6, 0, 'G', CCS, 0x80, &sbcsdata_CS_ISO8859_6, NULL, NULL, 0 },
+    { S6, 0, 'H', CCS, 0x80, &sbcsdata_CS_ISO8859_8, NULL, NULL, 0 },
+    { S6, 0, 'L', CCS, 0x80, &sbcsdata_CS_ISO8859_5, NULL, NULL, 0 },
+    { S6, 0, 'M', CCS, 0x80, &sbcsdata_CS_ISO8859_9, NULL, NULL, 0 },
+    { S4, 0, 'I', CCS, 0x80, &sbcsdata_CS_JISX0201, NULL, NULL, 0 },
+    { S4, 0, 'J', CCS, 0x00, &sbcsdata_CS_JISX0201, NULL, NULL, 0 },
     { M4, 0, 'A', CCS, -0x21, 0, &gb2312_to_unicode, &unicode_to_gb2312, -1 },
     { M4, 0, 'B', CCS, -0x21, 0, &jisx0208_to_unicode, &unicode_to_jisx0208, -1 },
     { M4, 0, 'C', CCS, -0x21, 0, &ksx1001_to_unicode, &unicode_to_ksx1001, -1 },
@@ -159,13 +157,13 @@ const struct iso2022_subcharset {
      * Next, other reasonably standard things: the rest of the ISO
      * 8859 sets, UK-ASCII, and CNS 11643.
      */
-    { S6, 0, 'T', COS, 0x80, &sbcsdata_CS_ISO8859_11 },
-    { S6, 0, 'V', COS, 0x80, &sbcsdata_CS_ISO8859_10 },
-    { S6, 0, 'Y', COS, 0x80, &sbcsdata_CS_ISO8859_13 },
-    { S6, 0, '_', COS, 0x80, &sbcsdata_CS_ISO8859_14 },
-    { S6, 0, 'b', COS, 0x80, &sbcsdata_CS_ISO8859_15 },
-    { S6, 0, 'f', COS, 0x80, &sbcsdata_CS_ISO8859_16 },
-    { S4, 0, 'A', COS, 0x00, &sbcsdata_CS_BS4730 },
+    { S6, 0, 'T', COS, 0x80, &sbcsdata_CS_ISO8859_11, NULL, NULL, 0 },
+    { S6, 0, 'V', COS, 0x80, &sbcsdata_CS_ISO8859_10, NULL, NULL, 0 },
+    { S6, 0, 'Y', COS, 0x80, &sbcsdata_CS_ISO8859_13, NULL, NULL, 0 },
+    { S6, 0, '_', COS, 0x80, &sbcsdata_CS_ISO8859_14, NULL, NULL, 0 },
+    { S6, 0, 'b', COS, 0x80, &sbcsdata_CS_ISO8859_15, NULL, NULL, 0 },
+    { S6, 0, 'f', COS, 0x80, &sbcsdata_CS_ISO8859_16, NULL, NULL, 0 },
+    { S4, 0, 'A', COS, 0x00, &sbcsdata_CS_BS4730, NULL, NULL, 0 },
     { M4, 0, 'G', COS, -0x21, 0, &cns11643_1_to_unicode, DEPLANARISE(&unicode_to_cns11643), 0 },
     { M4, 0, 'H', COS, -0x21, 0, &cns11643_2_to_unicode, DEPLANARISE(&unicode_to_cns11643), 1 },
     { M4, 0, 'I', COS, -0x21, 0, &cns11643_3_to_unicode, DEPLANARISE(&unicode_to_cns11643), 2 },
@@ -178,8 +176,8 @@ const struct iso2022_subcharset {
      * Private-use designations: DEC private sets and Emacs's Big5
      * abomination.
      */
-    { S4, 0, '0', CPU, 0x00, &sbcsdata_CS_DEC_GRAPHICS },
-    { S4, 0, '<', CPU, 0x80, &sbcsdata_CS_DEC_MCS },
+    { S4, 0, '0', CPU, 0x00, &sbcsdata_CS_DEC_GRAPHICS, NULL, NULL, 0 },
+    { S4, 0, '<', CPU, 0x80, &sbcsdata_CS_DEC_MCS, NULL, NULL, 0 },
     { M4, 0, '0', CPU, -0x21, 0, &emacs_big5_1_to_unicode, DEPLANARISE(&unicode_to_emacs_big5), 1 },
     { M4, 0, '1', CPU, -0x21, 0, &emacs_big5_2_to_unicode, DEPLANARISE(&unicode_to_emacs_big5), 2 },
 
@@ -195,8 +193,8 @@ const struct iso2022_subcharset {
     /*
      * Finally, fallback entries for null character sets.
      */
-    { S4, 0, '~', CNU },
-    { S6, 0, '~', CNU }, /* empty 96-set */
+    { S4, 0, '~', CNU, 0, NULL, NULL, NULL, 0 },
+    { S6, 0, '~', CNU, 0, NULL, NULL, NULL, 0 }, /* empty 96-set */
     { M4, 0, '~', CNU, 0, 0, &null_dbcs_to_unicode, &unicode_to_null_dbcs, -1 }, /* empty 94^n-set */
     { M6, 0, '~', CNU, 0, 0, &null_dbcs_to_unicode, &unicode_to_null_dbcs, -1 }, /* empty 96^n-set */
 };