- Correctly handles unicode characters that have a number of ascii alternates
not equal to 0.
 - Adds a --utf8 command line argument to output raw utf-8 in text output mode.

diff -u unrtf-0.21.0/src/convert.c unrtf-0.21.0-patch/src/convert.c
--- unrtf-0.21.0/src/convert.c	2008-12-13 08:04:59.000000000 +1000
+++ unrtf-0.21.0-patch/src/convert.c	2009-09-11 17:27:22.000000000 +1000
@@ -96,6 +96,7 @@
 #include "word.h"
 #include "hash.h"
 #include "convert.h"
+#include "unicode.h"
 
 #ifndef HAVE_ATTR_H
 #include "attr.h"
@@ -527,6 +528,7 @@
 
 
 extern int nopict_mode;
+extern int utf8_text_mode;
 
 /*
 #define BINARY_ATTRS
@@ -603,6 +605,11 @@
 
 static int banner_printed=FALSE;
 
+/* Number of words to skip after a unicode character,
+   set by \uc.
+*/
+static size_t unicode_word_skip_count = 1;
+
 
 /*========================================================================
  * Name:	print_banner
@@ -2139,14 +2146,18 @@
 		printf("%s", alias);
 		done++;
 	}
-	else
-		if(!done && op->unisymbol_print)
+	else if(!done){
+		if(op->unisymbol_print)
 		{
 			sprintf(tmp, "%d", param);
 
 			if (safe_printf(1, op->unisymbol_print, tmp)) fprintf(stderr, TOO_MANY_ARGS, "unisymbol_print");
 			done++;
+		} else if(utf8_text_mode){
+			printf(unicode_to_static_string(param));
+			done++;
 		}
+	}
 	
 	/*
 	** if we know how to represent the unicode character in the
@@ -2154,11 +2165,25 @@
 	** we will output that alternative.
 	*/
 	if (done)
-		return(SKIP_ONE_WORD);
+		return(SKIP_UNICODE_ALTERNATE);
 	return(FALSE);
 }
 
 /*========================================================================
+ * Name:	cmd_uc
+ * Purpose:	Executes the \uc command, setting the unicode skip count.
+ * Args:	Word, paragraph align info, and numeric param if any.
+ * Returns:	Flag, true only if rest of Words on line should be ignored.
+ *=======================================================================*/
+
+static int cmd_uc (Word *w, int align, char has_param, int param) {
+	if (has_param && param >= 0) {
+		unicode_word_skip_count = param;		
+	}
+	return FALSE;
+}
+
+/*========================================================================
  * Name:	cmd_dn
  * Purpose:	Executes the \dn command.
  * Args:	Word, paragraph align info, and numeric param if any.
@@ -2933,6 +2958,7 @@
 };
 static HashItem hashArray_u [] = {
 	{ "u", &cmd_u, NULL },
+	{ "uc", &cmd_uc, NULL },
 	{ "ul", &cmd_ul, NULL },
 	{ "up", &cmd_up, NULL },
 	{ "uld", &cmd_uld, NULL },
@@ -3214,6 +3240,7 @@
 	int is_cell_group=FALSE;
 	int paragraph_begined=FALSE;
 	int paragraph_align=ALIGN_LEFT;
+	int i;
 
 	CHECK_PARAM_NOT_NULL(w);
 
@@ -3304,7 +3331,6 @@
 
 					if (op->word_begin)
 						if (safe_printf(0, op->word_begin)) fprintf(stderr, TOO_MANY_ARGS, "word_begin");
-
 					print_with_special_exprs (s);
 
 					if (op->word_end)
@@ -3465,13 +3491,12 @@
 //unicode
 									terminate_group = hip[index].func (w,paragraph_align, have_param, param);
 	/* daved - 0.19.4 - unicode support may need to skip only one word */
-									if (terminate_group == SKIP_ONE_WORD)
-											w=w->next;
-											else
-									if (terminate_group)
+									if (terminate_group == SKIP_UNICODE_ALTERNATE) {
+										for(i=0;i < unicode_word_skip_count && w;i++,w=w->next);
+									} else if (terminate_group) {
 										while(w) w=w->next;
+									}
 								}
-
 								debug=hip[index].debug_print;
 
 								if (debug && debug_mode) {
diff -u unrtf-0.21.0/src/defs.h unrtf-0.21.0-patch/src/defs.h
--- unrtf-0.21.0/src/defs.h	2008-12-13 10:29:57.000000000 +1000
+++ unrtf-0.21.0-patch/src/defs.h	2009-09-11 17:27:22.000000000 +1000
@@ -61,10 +61,10 @@
 #define FALSE (0)
 #endif
 #if 1 /* daved - 0.19.4 */
-#define SKIP_ONE_WORD	2
+#define SKIP_UNICODE_ALTERNATE	2
 #endif
 
-#define USAGE "unrtf [--version] [--verbose] [--help] [--nopict|-n] [--noremap] [--html] [--text] [--vt] [--latex] [--rtf] [-P config_search_path]  [-t <file_with_tags>)] <filename>"
+#define USAGE "unrtf [--version] [--verbose] [--help] [--nopict|-n] [--utf8] [--noremap] [--html] [--text] [--vt] [--latex] [--rtf] [-P config_search_path]  [-t <file_with_tags>)] <filename>"
 
 
 /* Default names for RTF's default fonts */
diff -u unrtf-0.21.0/src/main.c unrtf-0.21.0-patch/src/main.c
--- unrtf-0.21.0/src/main.c	2008-12-13 10:21:36.000000000 +1000
+++ unrtf-0.21.0-patch/src/main.c	2009-09-11 17:27:22.000000000 +1000
@@ -94,6 +94,9 @@
 int verbose_mode;  /* TRUE => Output additional informations about unrtf */
 int no_remap_mode; /* don't remap codepoints */
 
+int utf8_text_mode; /* TRUE => Write unicode characters as UTF-8 when unisymbol_print
+					isn't defined. */
+
 
 OutputPersonality *op = NULL;
 
@@ -189,6 +192,7 @@
 		else if (!strcmp("--verbose", argv[i])) verbose_mode = TRUE;
 		else if (!strcmp("--simple", argv[i])) simple_mode = TRUE;
 		else if (!strcmp("--noremap", argv[i])) no_remap_mode = TRUE;
+		else if (!strcmp("--utf8", argv[i])) utf8_text_mode = TRUE;
 		else if (!strcmp("-t", argv[i]))
 		{
 			if ((i + 1) < argc && *argv[i + 1] != '-')
diff -u unrtf-0.21.0/src/unicode.c unrtf-0.21.0-patch/src/unicode.c
--- unrtf-0.21.0/src/unicode.c	2008-11-09 20:42:33.000000000 +1000
+++ unrtf-0.21.0-patch/src/unicode.c	2009-09-11 18:28:39.000000000 +1000
@@ -6,9 +6,12 @@
  *----------------------------------------------------------------------
  */
 #include <stdio.h>
+#include <my_iconv.h>
 
 #include "malloc.h"
 
+my_iconv_t utf16_desc = MY_ICONV_T_CLEAR;
+
 /*========================================================================
  * Name		get_unicode
  * Purpose:	Translates unicode character to number.
@@ -38,6 +41,37 @@
 }
 
 /*========================================================================
+ * Name		unicode_to_static_string
+ * Purpose:	Translates a unicode number to string using iconv.
+ * Args:	Unicode number.
+ * Returns:	Unicode character as string.
+ *=======================================================================*/
+char *
+unicode_to_static_string(int uc)
+{
+	iconv_t ic;
+	static char out[7];
+	size_t outremaining = 7;
+	char in[2];
+	size_t inremaining = 2;
+
+	char *outpos =  out;
+	char *inpos = in;
+
+	in[0] = ((char*)&uc)[0];
+	in[1] = ((char*)&uc)[1];
+
+	if (!my_iconv_is_valid(utf16_desc)) {
+		utf16_desc = my_iconv_open("utf8", "utf16");
+	}
+
+	my_iconv(utf16_desc, &inpos, &inremaining, &outpos, &outremaining);
+
+	out[7-inremaining] = 0;
+	return out;
+}
+
+/*========================================================================
  * Name		unicode_to_string
  * Purpose:	Translates unicode number to character.
  * Args:	Unicode number.