X-Git-Url: http://git.maemo.org/git/?p=modest;a=blobdiff_plain;f=src%2Fmodest-text-utils.c;h=f2ca84088c2f6d9eb20451cb4fee7cc67dfffc15;hp=3a89582bdb87680172e9a296043eac068afca844;hb=d39a0b6738e86acb01327d641b6e78eb0afa8c63;hpb=a3524da28501507a8c5dc8319bd62d61448edf46

diff --git a/src/modest-text-utils.c b/src/modest-text-utils.c
index 3a89582..f2ca840 100644
--- a/src/modest-text-utils.c
+++ b/src/modest-text-utils.c
@@ -65,13 +65,6 @@
 #define HYPERLINKIFY_MAX_LENGTH (1024*50)
 
 
-/*
- * we mark the ampersand with \007 when converting text->html
- * because after text->html we do hyperlink detecting, which
- * could be screwed up by the ampersand
- */
-#define MARK_AMP '\007'
-#define MARK_AMP_STR "\007"
 
 /*
  * we need these regexps to find URLs in plain text e-mails
@@ -90,19 +83,44 @@ struct _url_match_t {
 	const gchar* prefix;
 };
 
-/* note: match MARK_AMP_STR as well, because after txt->html, a '&' will look like $(MARK_AMP_STR)"amp;" */
+
+/*
+ * we mark the ampersand with \007 when converting text->html
+ * because after text->html we do hyperlink detecting, which
+ * could be screwed up by the ampersand.
+ * ie. 1<3 ==> 1\007lt;3
+ */
+#define MARK_AMP '\007'
+#define MARK_AMP_STR "\007"
+
+/* mark &amp; separately, because they are parts of urls.
+ * ie. a&b => a\006amp;b, but a>b => a\007gt;b
+ *
+ * we need to handle '&' separately, because it can be part of URIs
+ * (as in href="http://foo.bar?a=1&b=1"), so inside those URIs
+ * we need to re-replace \006amp; with '&' again, while outside uri's
+ * it will be '&amp;'
+ * 
+ * yes, it's messy, but a consequence of doing text->html first, then hyperlinkify
+ */
+#define MARK_AMP_URI '\006'
+#define MARK_AMP_URI_STR "\006"
+
+
+/* note: match MARK_AMP_URI_STR as well, because after txt->html, a '&' will look like $(MARK_AMP_URI_STR)"amp;" */
 #define MAIL_VIEWER_URL_MATCH_PATTERNS  {				\
-	{ "^(file|rtsp|http|ftp|https)://[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_STR "]+[-a-z0-9_$%" MARK_AMP_STR "=?/~#]",\
+	{ "(file|rtsp|http|ftp|https|mms|mmsh|rtsp|rdp|lastfm)://[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_URI_STR \
+			"]+[-a-z0-9_$%" MARK_AMP_URI_STR "=?/~#]",	\
 	  NULL, NULL },\
-	{ "^www\\.[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_STR "]+[-a-z0-9_$%" MARK_AMP_STR "=?/~#]",\
+	{ "www\\.[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_URI_STR "]+[-a-z0-9_$%" MARK_AMP_URI_STR "=?/~#]",\
 			NULL, "http://" },				\
-	{ "^ftp\\.[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_STR "]+[-a-z0-9_$%" MARK_AMP_STR "=?/~#]",\
+	{ "ftp\\.[-a-z0-9_$.+!*(),;:@%=?/~#" MARK_AMP_URI_STR "]+[-a-z0-9_$%" MARK_AMP_URI_STR "=?/~#]",\
 	  NULL, "ftp://" },\
-	{ "^(voipto|callto|chatto|jabberto|xmpp):[-_a-z@0-9.+]+", \
+	{ "(jabberto|voipto|sipto|sip|chatto|xmpp):[-_a-z@0-9.+]+", \
 	   NULL, NULL},						    \
-	{ "^mailto:[-_a-z0-9.\\+]+@[-_a-z0-9.]+",		    \
+	{ "mailto:[-_a-z0-9.\\+]+@[-_a-z0-9.]+",		    \
 	  NULL, NULL},\
-	{ "^[-_a-z0-9.\\+]+@[-_a-z0-9.]+",\
+	{ "[-_a-z0-9.\\+]+@[-_a-z0-9.]+",\
 	  NULL, "mailto:"}\
 	}
 
@@ -213,6 +231,8 @@ forward_cite (const gchar *from,
 	      const gchar *to,
 	      const gchar *subject)
 {
+	g_return_val_if_fail (sent, NULL);
+	
 	return g_strdup_printf ("%s\n%s %s\n%s %s\n%s %s\n%s %s\n", 
 				FORWARD_STRING, 
 				FROM_STRING, (from)?from:"",
@@ -380,7 +400,7 @@ modest_text_utils_convert_buffer_to_html_start (GString *html, const gchar *data
 
 	/* replace with special html chars where needed*/
 	for (i = 0; i != n; ++i)  {
-		char kar = data[i];
+		guchar kar = data[i];
 		
 		if (space_seen && kar != ' ') {
 			g_string_append_c (html, ' ');
@@ -390,21 +410,25 @@ modest_text_utils_convert_buffer_to_html_start (GString *html, const gchar *data
 		/* we artificially insert a breakpoint (newline)
 		 * after 256, to make sure our lines are not so long
 		 * they will DOS the regexping later
+		 * Also, check that kar is ASCII to make sure that we
+		 * don't break a UTF8 char in two
 		 */
-		if (++break_dist == 256) {
+		if (++break_dist >= 256 && kar < 127) {
 			g_string_append_c (html, '\n');
 			break_dist = 0;
 		}
 		
 		switch (kar) {
 		case 0:
-		case MARK_AMP: /* this is a temp place holder for '&'; we can only
+		case MARK_AMP:
+		case MARK_AMP_URI:	
+			/* this is a temp place holder for '&'; we can only
 				* set the real '&' after hyperlink translation, otherwise
 				* we might screw that up */
 			break; /* ignore embedded \0s and MARK_AMP */	
 		case '<'  : g_string_append (html, MARK_AMP_STR "lt;");   break;
 		case '>'  : g_string_append (html, MARK_AMP_STR "gt;");   break;
-		case '&'  : g_string_append (html, MARK_AMP_STR "amp;");  break;
+		case '&'  : g_string_append (html, MARK_AMP_URI_STR "amp;");  break; /* special case */
 		case '"'  : g_string_append (html, MARK_AMP_STR "quot;");  break;
 
 		/* don't convert &apos; --> wpeditor will try to re-convert it... */	
@@ -433,7 +457,7 @@ modest_text_utils_convert_buffer_to_html_finish (GString *html)
 	int i;
 	/* replace all our MARK_AMPs with real ones */
 	for (i = 0; i != html->len; ++i)
-		if ((html->str)[i] == MARK_AMP)
+		if ((html->str)[i] == MARK_AMP || (html->str)[i] == MARK_AMP_URI)
 			(html->str)[i] = '&';
 }
 
@@ -552,53 +576,6 @@ modest_text_utils_get_addresses_indexes (const gchar *addresses, GSList **start_
 	return;
 }
 
-#if 0
-GSList *
-modest_text_utils_split_addresses_list (const gchar *addresses)
-{
-	gchar *current, *start, *last_blank;
-	GSList *result = NULL;
-
-	start = (gchar *) addresses;
-	current = start;
-	last_blank = start;
-
-	while (*current != '\0') {
-		if ((start == current)&&((*current == ' ')||(*current == ',')||(*current == ';'))) {
-			start = g_utf8_next_char (start);
-			last_blank = current;
-		} else if ((*current == ',')||(*current == ';')) {
-			gchar *new_address = NULL;
-			new_address = g_strndup (start, current - last_blank);
-			result = g_slist_prepend (result, new_address);
-			start = g_utf8_next_char (current);
-			last_blank = start;
-		} else if (*current == '\"') {
-			if (current == start) {
-				current = g_utf8_next_char (current);
-				start = g_utf8_next_char (start);
-			}
-			while ((*current != '\"')&&(*current != '\0'))
-				current = g_utf8_next_char (current);
-		}
-				
-		current = g_utf8_next_char (current);
-	}
-
-	if (start != current) {
-		gchar *new_address = NULL;
-		new_address = g_strndup (start, current - last_blank);
-		result = g_slist_prepend (result, new_address);
-	}
-
-	result = g_slist_reverse (result);
-	return result;
-
-}
-#endif
-
-
-
 
 GSList *
 modest_text_utils_split_addresses_list (const gchar *addresses)
@@ -607,6 +584,8 @@ modest_text_utils_split_addresses_list (const gchar *addresses)
 	const gchar *my_addrs = addresses;
 	const gchar *end;
 	gchar *addr;
+
+	g_return_val_if_fail (addresses, NULL);
 	
 	/* skip any space, ',', ';' at the start */
 	while (my_addrs && (my_addrs[0] == ' ' || my_addrs[0] == ',' || my_addrs[0] == ';'))
@@ -1110,12 +1089,12 @@ hyperlinkify_plain_text (GString *txt)
 		gchar *url  = g_strndup (txt->str + match->offset, match->len);
 		gchar *repl = NULL; /* replacement  */
 
-		/* the string still contains $(MARK_AMP_STR)"amp;" for each
+		/* the string still contains $(MARK_AMP_URI_STR)"amp;" for each
 		 * '&' in the original, because of the text->html conversion.
 		 * in the href-URL (and only there), we must convert that back to
 		 * '&'
 		 */
-		gchar *href_url = replace_string (url, MARK_AMP_STR "amp;", '&');
+		gchar *href_url = replace_string (url, MARK_AMP_URI_STR "amp;", '&');
 		
 		/* the prefix is NULL: use the one that is already there */
 		repl = g_strdup_printf ("<a href=\"%s%s\">%s</a>",
@@ -1256,7 +1235,7 @@ modest_text_utils_utf8_strcmp (const gchar* s1, const gchar *s2, gboolean insens
 	/* if it's not case sensitive */
 	if (!insensitive) {
 
-		/* optimization: short cut if first char is ascii */ 
+		/* optimization: shortcut if first char is ascii */ 
 		if (((s1[0] & 0xf0)== 0) && ((s2[0] & 0xf0) == 0)) 
 			return s1[0] - s2[0];
 		
@@ -1654,3 +1633,57 @@ modest_text_utils_is_forbidden_char (const gchar character,
 
 	return FALSE; /* it's valid! */
 }
+
+gchar *      
+modest_text_utils_label_get_selection (GtkLabel *label)
+{
+	gint start, end;
+	gchar *selection;
+
+	if (gtk_label_get_selection_bounds (GTK_LABEL (label), &start, &end)) {
+		const gchar *start_offset;
+		const gchar *end_offset;
+		start_offset = gtk_label_get_text (GTK_LABEL (label));
+		start_offset = g_utf8_offset_to_pointer (start_offset, start);
+		end_offset = gtk_label_get_text (GTK_LABEL (label));
+		end_offset = g_utf8_offset_to_pointer (end_offset, end);
+		selection = g_strndup (start_offset, end_offset - start_offset);
+		return selection;
+	} else {
+		return g_strdup ("");
+	}
+}
+
+static gboolean
+_forward_search_image_char (gunichar ch,
+			    gpointer userdata)
+{
+	return (ch == 0xFFFC);
+}
+
+gboolean
+modest_text_utils_buffer_selection_is_valid (GtkTextBuffer *buffer)
+{
+	gboolean result;
+	GtkTextIter start, end;
+
+	g_return_val_if_fail (GTK_IS_TEXT_BUFFER (buffer), FALSE);
+
+	result = gtk_text_buffer_get_has_selection (GTK_TEXT_BUFFER (buffer));
+
+	/* check there are no images in selection */
+	if (result) {
+		gtk_text_buffer_get_selection_bounds (buffer, &start, &end);
+		if (gtk_text_iter_get_char (&start)== 0xFFFC)
+			result = FALSE;
+		else {
+			gtk_text_iter_backward_char (&end);
+			if (gtk_text_iter_forward_find_char (&start, _forward_search_image_char,
+							     NULL, &end))
+				result = FALSE;
+		}
+				    
+	}
+
+	return result;
+}