From 9efc661384b2020fd71be93137522b6f071a0f04 Mon Sep 17 00:00:00 2001
From: Kalle Olavi Niemitalo <kon@iki.fi>
Date: Sat, 9 Dec 2006 21:13:13 +0200
Subject: [PATCH] Bug 827: Force UTF-8 I/O on if the UTF-8 charset is selected.

---
 src/config/options.inc  |  8 ++++++--
 src/terminal/event.c    | 15 +++++++++------
 src/terminal/screen.c   |  8 ++++++--
 src/terminal/terminal.h |  3 ++-
 4 files changed, 23 insertions(+), 11 deletions(-)

diff --git a/src/config/options.inc b/src/config/options.inc
index a78888bc..c2fae4c7 100644
--- a/src/config/options.inc
+++ b/src/config/options.inc
@@ -887,8 +887,12 @@ static struct option_info config_options_info[] = {
 
 	INIT_OPT_BOOL("terminal._template_", N_("UTF-8 I/O"),
 		"utf_8_io", 0, 0,
-		N_("Enable I/O in UTF8 for Unicode terminals. Note that currently,\n"
-		"only the subset of UTF8 according to terminal codepage is used.")),
+		N_("Enable I/O in UTF-8 for Unicode terminals. Note that currently,\n"
+		"only the subset of UTF-8 according to terminal codepage is used.\n"
+		"ELinks ignores this option if the terminal codepage is UTF-8.")),
+		/* When CONFIG_UTF8 is defined, any code that reads the "utf_8_io"
+		 * option should also check whether the "codepage" option is UTF-8,
+		 * and if so, behave as if "utf_8_io" were 1.  */
 
 	INIT_OPT_BOOL("terminal._template_", N_("Restrict frames in cp850/852"),
 		"restrict_852", 0, 0,
diff --git a/src/terminal/event.c b/src/terminal/event.c
index adbd2195..49913aab 100644
--- a/src/terminal/event.c
+++ b/src/terminal/event.c
@@ -182,7 +182,11 @@ check_terminal_name(struct terminal *term, struct terminal_info *info)
 	/* Probably not best place for set this. But now we finally have
 	 * term->spec and term->utf8 should be set before decode session info.
 	 * --Scrool */
-	term->utf8 = get_opt_bool_tree(term->spec, "utf_8_io");
+	/* Force UTF-8 I/O if the UTF-8 charset is selected.  Various
+	 * places assume that the terminal's charset is unibyte if
+	 * UTF-8 I/O is disabled.  (bug 827) */
+	term->utf8 = get_opt_bool_tree(term->spec, "utf_8_io")
+		|| is_cp_utf8(get_opt_codepage_tree(term->spec, "charset"));
 #endif /* CONFIG_UTF8 */
 }
 
@@ -297,15 +301,14 @@ handle_interlink_event(struct terminal *term, struct interlink_event *ilev)
 		/* Character Conversions.  */
 #ifdef CONFIG_UTF8
 		/* struct term_event_keyboard carries UCS-4.
-		 * - If the "utf_8_io" option (i.e. term->utf8) is
-		 *   true or the "charset" option refers to UTF-8,
-		 *   then handle_interlink_event() converts from UTF-8
+		 * - If the "utf_8_io" option is true or the "charset"
+		 *   option refers to UTF-8, then term->utf8 is true,
+		 *   and handle_interlink_event() converts from UTF-8
 		 *   to UCS-4.
 		 * - Otherwise, handle_interlink_event() converts from
 		 *   the codepage specified with the "charset" option
 		 *   to UCS-4.  */
-		utf8_io = term->utf8
-			|| is_cp_utf8(get_opt_codepage_tree(term->spec, "charset"));
+		utf8_io = term->utf8;
 #else
 		/* struct term_event_keyboard carries bytes in the
 		 * charset of the terminal.
diff --git a/src/terminal/screen.c b/src/terminal/screen.c
index 58181cff..1f5e0ca0 100644
--- a/src/terminal/screen.c
+++ b/src/terminal/screen.c
@@ -132,7 +132,7 @@ struct screen_driver {
 	unsigned int transparent:1;
 
 #ifdef CONFIG_UTF8
-	/* UTF-8 I/O */
+	/* UTF-8 I/O.  Forced on if the UTF-8 charset is selected.  (bug 827) */
 	unsigned int utf8:1;
 #endif /* CONFIG_UTF8 */
 
@@ -241,7 +241,11 @@ static void
 update_screen_driver(struct screen_driver *driver, struct option *term_spec)
 {
 #ifdef CONFIG_UTF8
-	driver->utf8 = get_opt_bool_tree(term_spec, "utf_8_io");
+	/* Force UTF-8 I/O if the UTF-8 charset is selected.  Various
+	 * places assume that the terminal's charset is unibyte if
+	 * UTF-8 I/O is disabled.  (bug 827) */
+	driver->utf8 = get_opt_bool_tree(term_spec, "utf_8_io")
+		|| is_cp_utf8(get_opt_codepage_tree(term_spec, "charset"));
 #else
 	int utf8_io = get_opt_bool_tree(term_spec, "utf_8_io");
 #endif /* CONFIG_UTF8 */
diff --git a/src/terminal/terminal.h b/src/terminal/terminal.h
index e04fb1e7..9a609aac 100644
--- a/src/terminal/terminal.h
+++ b/src/terminal/terminal.h
@@ -127,7 +127,8 @@ struct terminal {
 	unsigned int master:1;
 
 #ifdef CONFIG_UTF8
-	/* Indicates whether UTF-8 I/O is used */
+	/* Indicates whether UTF-8 I/O is used.  Forced on if the
+	 * UTF-8 charset is selected.  (bug 827) */
 	unsigned int utf8:1;
 #endif /* CONFIG_UTF8 */