First draft of Unicode support in pterm. It's pretty complete: it

does UTF-8 copy and paste (falling back to normal strings if
necessary), it understands X font encodings and translates things
accordingly so that if you have a Unicode font you can ask for
virtually any single-byte encoding and get it (Mac-Roman pterm,
anyone?), and so on. There's work left to be done (wide fonts for
CJK spring to mind), but I reckon this is a pretty good start.

[originally from svn r2395]

charset/internal.h

@@ -0,0 +1,89 @@
+/*
+ * internal.h - internal header stuff for the charset library.
+ */
+
+#ifndef charset_internal_h
+#define charset_internal_h
+
+/* This invariably comes in handy */
+#define lenof(x) ( sizeof((x)) / sizeof(*(x)) )
+
+/* This is an invalid Unicode value used to indicate an error. */
+#define ERROR 0xFFFFL		       /* Unicode value representing error */
+
+typedef struct charset_spec charset_spec;
+typedef struct sbcs_data sbcs_data;
+
+struct charset_spec {
+    int charset;		       /* numeric identifier */
+
+    /*
+     * A function to read the character set and output Unicode
+     * characters. The `emit' function expects to get Unicode chars
+     * passed to it; it should be sent ERROR for any encoding error
+     * on the input.
+     */
+    void (*read)(charset_spec const *charset, long int input_chr,
+		 charset_state *state,
+		 void (*emit)(void *ctx, long int output), void *emitctx);
+    /*
+     * A function to read Unicode characters and output in this
+     * character set. The `emit' function expects to get byte
+     * values passed to it; it should be sent ERROR for any
+     * non-representable characters on the input.
+     */
+    void (*write)(charset_spec const *charset, long int input_chr,
+		  charset_state *state,
+		  void (*emit)(void *ctx, long int output), void *emitctx);
+    void const *data;
+};
+
+/*
+ * This is the format of `data' used by the SBCS read and write
+ * functions; so it's the format used in all SBCS definitions.
+ */
+struct sbcs_data {
+    /*
+     * This is a simple mapping table converting each SBCS position
+     * to a Unicode code point. Some positions may contain ERROR,
+     * indicating that that byte value is not defined in the SBCS
+     * in question and its occurrence in input is an error.
+     */
+    unsigned long sbcs2ucs[256];
+
+    /*
+     * This lookup table is used to convert Unicode back to the
+     * SBCS. It consists of the valid byte values in the SBCS,
+     * sorted in order of their Unicode translation. So given a
+     * Unicode value U, you can do a binary search on this table
+     * using the above table as a lookup: when testing the Xth
+     * position in this table, you branch according to whether
+     * sbcs2ucs[ucs2sbcs[X]] is less than, greater than, or equal
+     * to U.
+     * 
+     * Note that since there may be fewer than 256 valid byte
+     * values in a particular SBCS, we must supply the length of
+     * this table as well as the contents.
+     */
+    unsigned char ucs2sbcs[256];
+    int nvalid;
+};
+
+/*
+ * Prototypes for internal library functions.
+ */
+charset_spec const *charset_find_spec(int charset);
+void read_sbcs(charset_spec const *charset, long int input_chr,
+	       charset_state *state,
+	       void (*emit)(void *ctx, long int output), void *emitctx);
+void write_sbcs(charset_spec const *charset, long int input_chr,
+		charset_state *state,
+		void (*emit)(void *ctx, long int output), void *emitctx);
+
+/*
+ * Placate compiler warning about unused parameters, of which we
+ * expect to have some in this library.
+ */
+#define UNUSEDARG(x) ( (x) = (x) )
+
+#endif /* charset_internal_h */