/pidgin/main: 22dca763eeef: Tab completion
Ankit Vani
a at nevitus.org
Thu Jan 2 10:58:27 EST 2014
Changeset: 22dca763eeef8ff4ebf4f2daa48fc3267c57c231
Author: Ankit Vani <a at nevitus.org>
Date: 2014-01-02 20:10 +0530
Branch: default
URL: https://hg.pidgin.im/pidgin/main/rev/22dca763eeef
Description:
Tab completion
diffstat:
pidgin/gtkconv.c | 218 +++++++++++++++++++++++++++++++--------------------
pidgin/gtkwebview.c | 54 ++++++++++++
pidgin/gtkwebview.h | 25 +++++
3 files changed, 211 insertions(+), 86 deletions(-)
diffs (truncated from 450 to 300 lines):
diff --git a/pidgin/gtkconv.c b/pidgin/gtkconv.c
--- a/pidgin/gtkconv.c
+++ b/pidgin/gtkconv.c
@@ -4350,20 +4350,17 @@ add_chat_user_common(PurpleChatConversat
g_free(alias_key);
}
-#if 0
/**
* @param most_matched Used internally by this function.
* @param entered The partial string that the user types before hitting the
* tab key.
- * @param entered_bytes The length of entered.
+ * @param entered_chars The length of entered.
* @param partial This is a return variable. This will be set to a string
* containing the largest common string between all matches. This will
* be inserted into the input box at the start of the word that the
* user is tab completing. For example, if a chat room contains
* "AlfFan" and "AlfHater" and the user types "a<TAB>" then this will
* contain "Alf"
- * @param nick_partial Used internally by this function. Shoudl be a
- * temporary buffer that is entered_bytes+1 bytes long.
* @param matches This is a return variable. If the given name is a potential
* match for the entered string, then add a copy of the name to this
* list. The caller is responsible for g_free'ing the data in this
@@ -4372,13 +4369,22 @@ add_chat_user_common(PurpleChatConversat
* checking for a match.
*/
static void
-tab_complete_process_item(int *most_matched, const char *entered, gsize entered_bytes, char **partial, char *nick_partial,
+tab_complete_process_item(int *most_matched, const char *entered, gsize entered_chars, char **partial,
GList **matches, const char *name)
{
- memcpy(nick_partial, name, entered_bytes);
- if (purple_utf8_strcasecmp(nick_partial, entered))
+ char *nick_partial;
+ gsize name_len = g_utf8_strlen(name, -1);
+
+ if ((glong)entered_chars > name_len)
return;
+ nick_partial = g_utf8_substring(name, 0, entered_chars);
+ if (purple_utf8_strcasecmp(nick_partial, entered)) {
+ g_free(nick_partial);
+ return;
+ }
+ g_free(nick_partial);
+
/* if we're here, it's a possible completion */
if (*most_matched == -1) {
@@ -4386,16 +4392,16 @@ tab_complete_process_item(int *most_matc
* this will only get called once, since from now
* on *most_matched is >= 0
*/
- *most_matched = strlen(name);
+ *most_matched = name_len;
*partial = g_strdup(name);
}
else if (*most_matched) {
char *tmp = g_strdup(name);
while (purple_utf8_strcasecmp(tmp, *partial)) {
- (*partial)[*most_matched] = '\0';
- if (*most_matched < (goffset)strlen(tmp))
- tmp[*most_matched] = '\0';
+ *(g_utf8_offset_to_pointer(*partial, *most_matched)) = '\0';
+ if (*most_matched < (goffset)g_utf8_strlen(tmp, -1))
+ *(g_utf8_offset_to_pointer(tmp, *most_matched)) = '\0';
(*most_matched)--;
}
(*most_matched)++;
@@ -4406,102 +4412,123 @@ tab_complete_process_item(int *most_matc
*matches = g_list_insert_sorted(*matches, g_strdup(name),
(GCompareFunc)purple_utf8_strcasecmp);
}
-#endif
+
+static gboolean
+is_first_container(WebKitDOMNode *container)
+{
+ gchar *name;
+ WebKitDOMNode *parent;
+
+ while (container) {
+ parent = webkit_dom_node_get_parent_node(container);
+ if (parent) {
+ name = webkit_dom_node_get_node_name(parent);
+
+ if (!strcmp(name, "BODY")) {
+ g_free(name);
+
+ if (webkit_dom_node_get_previous_sibling(container) == NULL)
+ return TRUE;
+ else
+ return FALSE;
+ }
+ g_free(name);
+ }
+ else
+ break;
+
+ container = parent;
+ }
+
+ return FALSE;
+}
static gboolean
tab_complete(PurpleConversation *conv)
{
-#if 0
- /* TODO WebKit */
PidginConversation *gtkconv;
- GtkTextIter cursor, word_start, start_buffer;
- int start;
- int most_matched = -1;
+ WebKitDOMNode *container;
+ glong caret, word_start, content_len;
+ int most_matched = -1, colon = 0;
+ char *ch, *ch2 = NULL;
char *entered, *partial = NULL;
- char *text;
- char *nick_partial;
+ char *content, *sub1, *sub2, *modified;
const char *prefix;
GList *matches = NULL;
gboolean command = FALSE;
- gsize entered_bytes = 0;
+ gsize entered_chars = 0;
gtkconv = PIDGIN_CONVERSATION(conv);
-
- gtk_text_buffer_get_start_iter(gtkconv->entry_buffer, &start_buffer);
- gtk_text_buffer_get_iter_at_mark(gtkconv->entry_buffer, &cursor,
- gtk_text_buffer_get_insert(gtkconv->entry_buffer));
-
- word_start = cursor;
+ gtk_webview_get_caret(GTK_WEBVIEW(gtkconv->entry), &container, &caret);
/* if there's nothing there just return */
- if (!gtk_text_iter_compare(&cursor, &start_buffer))
+ if (caret <= 0)
return PURPLE_IS_CHAT_CONVERSATION(conv);
- text = gtk_text_buffer_get_text(gtkconv->entry_buffer, &start_buffer,
- &cursor, FALSE);
-
- /* if we're at the end of ": " we need to move back 2 spaces */
- start = strlen(text) - 1;
-
- if (start >= 1 && !strncmp(&text[start-1], ": ", 2)) {
- gtk_text_iter_backward_chars(&word_start, 2);
- }
-
- /* find the start of the word that we're tabbing.
- * Using gtk_text_iter_backward_word_start won't work, because a nick can contain
- * characters (e.g. '.', '/' etc.) that Pango may think are word separators. */
- while (gtk_text_iter_backward_char(&word_start)) {
- if (gtk_text_iter_get_char(&word_start) == ' ') {
- /* Reached the whitespace before the start of the word. Move forward once */
- gtk_text_iter_forward_char(&word_start);
+ content = webkit_dom_node_get_node_value(container);
+ content_len = g_utf8_strlen(content, -1);
+
+ /* if we're at the end of ":" or ": " we need to move back 1 or 2 spaces */
+ if (caret >= 2) {
+ ch = g_utf8_offset_to_pointer(content, caret - 2);
+ ch2 = g_utf8_find_next_char(ch, NULL);
+ }
+
+ if (caret >= 2 && *ch == ':' && (*ch2 == ' ' || g_utf8_get_char(ch2) == 0xA0))
+ colon = 2;
+ else if (caret >= 1 && content[caret - 1] == ':')
+ colon = 1;
+
+ caret -= colon;
+ word_start = caret;
+
+ /* find the start of the word that we're tabbing. */
+ ch = g_utf8_offset_to_pointer(content, caret);
+ while ((ch = g_utf8_find_prev_char(content, ch))) {
+ if (*ch != ' ' && g_utf8_get_char(ch) != 0xA0)
+ --word_start;
+ else
break;
- }
}
prefix = pidgin_get_cmd_prefix();
- if (gtk_text_iter_get_offset(&word_start) == 0 &&
- (strlen(text) >= strlen(prefix)) && !strncmp(text, prefix, strlen(prefix))) {
+ if (word_start == 0 &&
+ ((gsize)caret >= strlen(prefix)) && !strncmp(content, prefix, strlen(prefix))) {
command = TRUE;
- gtk_text_iter_forward_chars(&word_start, strlen(prefix));
- }
-
- g_free(text);
-
- entered = gtk_text_buffer_get_text(gtkconv->entry_buffer, &word_start,
- &cursor, FALSE);
- entered_bytes = strlen(entered);
-
- if (!g_utf8_strlen(entered, -1)) {
+ word_start += strlen(prefix);
+ }
+
+ entered = g_utf8_substring(content, word_start, caret);
+ entered_chars = g_utf8_strlen(entered, -1);
+
+ if (!entered_chars) {
+ g_free(content);
g_free(entered);
return PURPLE_IS_CHAT_CONVERSATION(conv);
}
- nick_partial = g_malloc0(entered_bytes + 1);
-
if (command) {
GList *list = purple_cmd_list(conv);
GList *l;
/* Commands */
for (l = list; l != NULL; l = l->next) {
- tab_complete_process_item(&most_matched, entered, entered_bytes, &partial, nick_partial,
+ tab_complete_process_item(&most_matched, entered, entered_chars, &partial,
&matches, l->data);
}
g_list_free(list);
} else if (PURPLE_IS_CHAT_CONVERSATION(conv)) {
- PurpleChatConversation *chat = PURPLE_CONV_CHAT(conv);
- GList *l = purple_chat_conversation_get_users(chat);
+ GList *l = purple_chat_conversation_get_users(PURPLE_CHAT_CONVERSATION(conv));
GtkTreeModel *model = gtk_tree_view_get_model(GTK_TREE_VIEW(PIDGIN_CONVERSATION(conv)->u.chat->list));
GtkTreeIter iter;
int f;
/* Users */
for (; l != NULL; l = l->next) {
- tab_complete_process_item(&most_matched, entered, entered_bytes, &partial, nick_partial,
+ tab_complete_process_item(&most_matched, entered, entered_chars, &partial,
&matches, purple_chat_user_get_name((PurpleChatUser *)l->data));
}
-
/* Aliases */
if (gtk_tree_model_get_iter_first(GTK_TREE_MODEL(model), &iter))
{
@@ -4515,7 +4542,7 @@ tab_complete(PurpleConversation *conv)
-1);
if (name && alias && strcmp(name, alias))
- tab_complete_process_item(&most_matched, entered, entered_bytes, &partial, nick_partial,
+ tab_complete_process_item(&most_matched, entered, entered_chars, &partial,
&matches, alias);
g_free(name);
g_free(alias);
@@ -4524,38 +4551,49 @@ tab_complete(PurpleConversation *conv)
} while (f != 0);
}
} else {
- g_free(nick_partial);
+ g_free(content);
g_free(entered);
return FALSE;
}
- g_free(nick_partial);
-
- /* we're only here if we're doing new style */
-
/* if there weren't any matches, return */
if (!matches) {
/* if matches isn't set partials won't be either */
+ g_free(content);
g_free(entered);
return PURPLE_IS_CHAT_CONVERSATION(conv);
}
- gtk_text_buffer_delete(gtkconv->entry_buffer, &word_start, &cursor);
+ sub1 = g_utf8_substring(content, 0, word_start);
+ sub2 = g_utf8_substring(content, caret, content_len);
if (!matches->next) {
/* there was only one match. fill it in. */
- gtk_text_buffer_get_start_iter(gtkconv->entry_buffer, &start_buffer);
- gtk_text_buffer_get_iter_at_mark(gtkconv->entry_buffer, &cursor,
- gtk_text_buffer_get_insert(gtkconv->entry_buffer));
-
- if (!gtk_text_iter_compare(&cursor, &start_buffer)) {
- char *tmp = g_strdup_printf("%s: ", (char *)matches->data);
- gtk_text_buffer_insert_at_cursor(gtkconv->entry_buffer, tmp, -1);
+
+ if (!colon && !word_start && is_first_container(container)) {
+ char *tmp = NULL;
+ if (caret < content_len) {
+ tmp = g_strdup_printf("%s: ", (char *)matches->data);
+ } else {
+ char utf[6] = {0};
More information about the Commits
mailing list