Try to split long lines on spaces

Try to split long lines on spaces to avoid words being splitted. This
can be turned off with the option `split_line_on_space'. The code
assumes that the terminal encoding has ASCII spaces.

6 files changed, 48 insertions, 19 deletions

diff --git a/NEWS b/NEWS
index de36d195..70717965 100644
--- a/NEWS
+++ b/NEWS
@@ -1,5 +1,7 @@
 v0.8.18-head 2014-XX-YY  The Irssi team <staff@irssi.org>
 	+ Disable SSLv3 due to the POODLE vulnerability.
+	+ Try to split long lines on spaces to avoid words being splitted. Adds
+	  a new option: 'split_line_on_space' which defaults to on.
 
 v0.8.17 2014-10-11  The Irssi team <staff@irssi.org>
 	+ Document that SSL connections aren't properly handled during /UPGRADE. See Github PR #39.
diff --git a/src/core/misc.c b/src/core/misc.c
index 5e6087cb..586e4f7c 100644
--- a/src/core/misc.c
+++ b/src/core/misc.c
@@ -967,19 +967,30 @@ char *ascii_strdown(char *str)
 	return str;
 }
 
-char **strsplit_len(const char *str, int len)
-{
-	char **ret;
-	size_t total_len = strlen(str);
-	int n = total_len / len;
-	int i;
-
-	if (total_len % len)
-		n++;
-
-	ret = g_new(char *, n + 1);
-	for (i = 0; i < n; i++, str += len)
-		ret[i] = g_strndup(str, len);
+char **strsplit_len(const char *str, int len, gboolean onspace)
+{
+	char **ret = g_new(char *, 1);
+	int n;
+	int offset;
+
+	for (n = 0; *str != '\0'; n++, str += MIN(len - offset, strlen(str))) {
+		offset = 0;
+		if (onspace) {
+			/*
+			 * Try to find a space to split on and leave
+			 * the space on the previous line.
+			 */
+			int i;
+			for (i = 0; i < len; i++) {
+				if (str[len-1-i] == ' ') {
+					offset = i;
+					break;
+				}
+			}
+		}
+		ret[n] = g_strndup(str, len - offset);
+		ret = g_renew(char *, ret, n + 2);
+	}
 	ret[n] = NULL;
 
 	return ret;
diff --git a/src/core/misc.h b/src/core/misc.h
index 8fb5078f..c6369489 100644
--- a/src/core/misc.h
+++ b/src/core/misc.h
@@ -116,6 +116,6 @@ uoff_t str_to_uofft(const char *str);
 int find_substr(const char *list, const char *item);
 
 /* split `str' into `len' sized substrings */
-char **strsplit_len(const char *str, int len);
+char **strsplit_len(const char *str, int len, gboolean onspace);
 
 #endif
diff --git a/src/core/recode.c b/src/core/recode.c
index 029d7ff1..d001a46a 100644
--- a/src/core/recode.c
+++ b/src/core/recode.c
@@ -183,7 +183,7 @@ char *recode_out(const SERVER_REC *server, const char *str, const char *target)
 }
 
 char **recode_split(const SERVER_REC *server, const char *str,
-		    const char *target, int len)
+		    const char *target, int len, gboolean onspace)
 {
 	GIConv cd = (GIConv)-1;
 	const char *from = translit_charset;
@@ -219,7 +219,7 @@ char **recode_split(const SERVER_REC *server, const char *str,
 	cd = g_iconv_open(to, from);
 	if (cd == (GIConv)-1) {
 		/* Fall back to splitting by byte. */
-		ret = strsplit_len(str, len);
+		ret = strsplit_len(str, len, onspace);
 		goto out;
 	}
 
@@ -235,11 +235,25 @@ char **recode_split(const SERVER_REC *server, const char *str,
 			 */
 			ret[n] = NULL;
 			g_strfreev(ret);
-			ret = strsplit_len(str, len);
+			ret = strsplit_len(str, len, onspace);
 			goto out;
 		}
 
 		/* Outbuf overflowed, split the input string. */
+		if (onspace) {
+			/*
+			 * Try to find a space to split on and leave
+			 * the space on the previous line.
+			 */
+			int i;
+			for (i = 0; i < inbuf - previnbuf; i++) {
+				if (previnbuf[inbuf-previnbuf-1-i] == ' ') {
+					inbuf -= i;
+					inbytesleft += i;
+					break;
+				}
+			}
+		}
 		ret[n++] = g_strndup(previnbuf, inbuf - previnbuf);
 		ret = g_renew(char *, ret, n + 1);
 		previnbuf = inbuf;
diff --git a/src/core/recode.h b/src/core/recode.h
index b70ec630..719e8f54 100644
--- a/src/core/recode.h
+++ b/src/core/recode.h
@@ -4,7 +4,7 @@
 char *recode_in (const SERVER_REC *server, const char *str, const char *target);
 char *recode_out (const SERVER_REC *server, const char *str, const char *target);
 char **recode_split(const SERVER_REC *server, const char *str,
-		    const char *target, int len);
+		    const char *target, int len, gboolean onspace);
 gboolean is_valid_charset(const char *charset);
 gboolean is_utf8(void);
 void recode_update_charset(void);
diff --git a/src/irc/core/irc-servers.c b/src/irc/core/irc-servers.c
index 27878989..baf8a0d2 100644
--- a/src/irc/core/irc-servers.c
+++ b/src/irc/core/irc-servers.c
@@ -85,6 +85,7 @@ static char **split_line(const SERVER_REC *server, const char *line,
 {
 	const char *start = settings_get_str("split_line_start");
 	const char *end = settings_get_str("split_line_end");
+	gboolean onspace = settings_get_bool("split_line_on_space");
 	char *recoded_start = recode_out(server, start, target);
 	char *recoded_end = recode_out(server, end, target);
 	char **lines;
@@ -103,7 +104,7 @@ static char **split_line(const SERVER_REC *server, const char *line,
 		return NULL;
 	}
 
-	lines = recode_split(server, line, target, len);
+	lines = recode_split(server, line, target, len, onspace);
 	for (i = 0; lines[i] != NULL; i++) {
 		if (i != 0 && *start != '\0') {
 			/* Not the first line. */
@@ -972,6 +973,7 @@ void irc_servers_init(void)
 	settings_add_str("misc", "usermode", DEFAULT_USER_MODE);
 	settings_add_str("misc", "split_line_start", "");
 	settings_add_str("misc", "split_line_end", "");
+	settings_add_bool("misc", "split_line_on_space", TRUE);
 	settings_add_time("flood", "cmd_queue_speed", DEFAULT_CMD_QUEUE_SPEED);
 	settings_add_int("flood", "cmds_max_at_once", DEFAULT_CMDS_MAX_AT_ONCE);