[uim-commit] r2839 - branches/r5rs/sigscheme

[yamaken at freedesktop.org]

Author: yamaken
Date: 2006-01-08 01:18:44 -0800 (Sun, 08 Jan 2006)
New Revision: 2839

Modified:
   branches/r5rs/sigscheme/encoding.c
   branches/r5rs/sigscheme/encoding.h
   branches/r5rs/sigscheme/print.c
Log:
* sigscheme/encoding.h
  - (SCM_MBS_INIT2, SCM_CHARCODEC_READ_CHAR): New macro
  - (scm_charcodec_read_char): New function decl
* sigscheme/encoding.c
  - (scm_charcodec_read_char): New function
* sigscheme/print.c
  - (print_string): Support multibyte string properly


Modified: branches/r5rs/sigscheme/encoding.c
===================================================================
--- branches/r5rs/sigscheme/encoding.c	2006-01-08 09:17:00 UTC (rev 2838)
+++ branches/r5rs/sigscheme/encoding.c	2006-01-08 09:18:44 UTC (rev 2839)
@@ -334,6 +334,35 @@
     return NULL;
 }
 
+int
+scm_charcodec_read_char(ScmCharCodec *codec, ScmMultibyteString *mbs,
+                        const char *caller)
+{
+    ScmMultibyteCharInfo mbc;
+    ScmMultibyteState state;
+    int ch;
+    DECLARE_INTERNAL_FUNCTION("scm_charcodec_read_char");
+
+    SCM_ASSERT(SCM_MBS_GET_SIZE(*mbs));
+
+    SCM_MANGLE(name) = caller;
+
+    state = SCM_MBS_GET_STATE(*mbs);
+    mbc = SCM_CHARCODEC_SCAN_CHAR(codec, *mbs);
+    if (SCM_MBCINFO_ERRORP(mbc) || SCM_MBCINFO_INCOMPLETEP(mbc))
+        ERR("invalid char sequence");
+    ch = SCM_CHARCODEC_STR2INT(codec,
+                               SCM_MBS_GET_STR(*mbs),
+                               SCM_MBCINFO_GET_SIZE(mbc),
+                               state);
+    if (ch == EOF)
+        ERR("invalid char sequence");
+
+    SCM_MBS_SKIP_CHAR(*mbs, mbc);
+
+    return ch;
+}
+
 /*=======================================
   Encoding-specific functions
 =======================================*/

Modified: branches/r5rs/sigscheme/encoding.h
===================================================================
--- branches/r5rs/sigscheme/encoding.h	2006-01-08 09:17:00 UTC (rev 2838)
+++ branches/r5rs/sigscheme/encoding.h	2006-01-08 09:18:44 UTC (rev 2839)
@@ -97,6 +97,12 @@
         SCM_MBS_SET_SIZE((mbs), 0);                                          \
         SCM_MBS_CLEAR_STATE(mbs);                                            \
     } while (/* CONSTCOND */ 0)
+#define SCM_MBS_INIT2(mbs, s, siz)                                           \
+    do {                                                                     \
+        SCM_MBS_SET_STR((mbs), (s));                                         \
+        SCM_MBS_SET_SIZE((mbs), (siz));                                      \
+        SCM_MBS_CLEAR_STATE(mbs);                                            \
+    } while (/* CONSTCOND */ 0)
 #define SCM_MBS_SKIP_CHAR(mbs, inf)                                          \
     do {                                                                     \
         SCM_MBS_SET_STR((mbs),                                               \
@@ -115,6 +121,8 @@
     ((*(codec)->str2int)((src), (len), (state)))
 #define SCM_CHARCODEC_INT2STR(codec, dst, ch, state)                         \
     ((*(codec)->int2str)((dst), (ch), (state)))
+#define SCM_CHARCODEC_READ_CHAR(codec, mbs)                                  \
+    (scm_charcodec_read_char((codec), &(mbs), SCM_MANGLE(name)))
 
 /*=======================================
   Type Definitions
@@ -198,6 +206,7 @@
 ScmMultibyteString scm_mb_substring(ScmMultibyteString str, int i, int len);
 #define scm_mb_strref(str, i) (scm_mb_substring((str), (i), 1))
 ScmCharCodec *scm_mb_find_codec(const char *encoding);
+int scm_charcodec_read_char(ScmCharCodec *codec, ScmMultibyteString *mbs,
+                            const char *caller);
 
-
 #endif /* __SCM_ENCODING_H */

Modified: branches/r5rs/sigscheme/print.c
===================================================================
--- branches/r5rs/sigscheme/print.c	2006-01-08 09:17:00 UTC (rev 2838)
+++ branches/r5rs/sigscheme/print.c	2006-01-08 09:18:44 UTC (rev 2839)
@@ -286,32 +286,45 @@
     }
 }
 
-/* FIXME: support multibyte char properly */
 static void
 print_string(ScmObj port, ScmObj obj, enum OutputType otype)
 {
+    ScmCharCodec *codec;
+    ScmMultibyteString mbs;
     const ScmSpecialCharInfo *info;
     const char *str;
-    int len, c, i;
+    int len, c;
+    DECLARE_INTERNAL_FUNCTION("print_string");
 
     str = SCM_STRING_STR(obj);
     len = strlen(str);
 
     switch (otype) {
     case AS_WRITE:
-        scm_port_put_char(port, '\"'); /* first doublequote */
-        for (i = 0; i < len; i++) {
-            c = str[i];
-            for (info = scm_special_char_table; info->esc_seq; info++) {
-                if (c == info->code) {
-                    scm_port_puts(port, info->esc_seq);
-                    break;
+        scm_port_put_char(port, '\"'); /* opening doublequote */
+        if (scm_current_char_codec != scm_port_codec(port)) {
+            /* Since the str does not have its encoding information, here
+             * assumes that scm_current_char_codec is that. And then SigScheme
+             * does not have an encoding conversion mechanism, puts it
+             * as-is. */
+            scm_port_puts(port, str);
+        } else {
+            codec = scm_port_codec(port);
+            SCM_MBS_INIT2(mbs, str, len);
+            while (SCM_MBS_GET_SIZE(mbs)) {
+                c = SCM_CHARCODEC_READ_CHAR(codec, mbs);
+                for (info = scm_special_char_table; info->esc_seq; info++) {
+                    if (c == info->code) {
+                        scm_port_puts(port, info->esc_seq);
+                        goto continue2;
+                    }
                 }
+                scm_port_put_char(port, c);
+            continue2:
+                ;
             }
-            if (!info->esc_seq)
-                scm_port_put_char(port, str[i]);
         }
-        scm_port_put_char(port, '\"'); /* last doublequote */
+        scm_port_put_char(port, '\"'); /* closing doublequote */
         break;
 
     case AS_DISPLAY: