spellchk.c

Linux下的多协议即时通讯程序源代码

/*
 * Purple - Replace certain misspelled words with their correct form.
 *
 * Signification changes were made by Benjamin Kahn ("xkahn") and
 * Richard Laager ("rlaager") in April 2005--you may want to contact
 * them if you have questions.
 *
 * Pidgin is the legal property of its developers, whose names are too numerous
 * to list here.  Please refer to the COPYRIGHT file distributed with this
 * source distribution.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 *
 */

/*
 * A lot of this code (especially the config code) was taken directly
 * or nearly directly from xchat, version 1.4.2 by Peter Zelezny and others.
 */

#include "internal.h"
#include "pidgin.h"

#include "debug.h"
#include "notify.h"
#include "signals.h"
#include "util.h"
#include "version.h"

#include "gtkplugin.h"
#include "gtkprefs.h"
#include "gtkutils.h"

#include <stdio.h>
#include <string.h>
#ifndef _WIN32
#include <strings.h>
#endif
#include <sys/types.h>
#include <sys/stat.h>

#define SPELLCHECK_PLUGIN_ID "gtk-spellcheck"
#define SPELLCHK_OBJECT_KEY "spellchk"

enum {
	BAD_COLUMN,
	GOOD_COLUMN,
	WORD_ONLY_COLUMN,
	CASE_SENSITIVE_COLUMN,
	N_COLUMNS
};

struct _spellchk {
	GtkTextView *view;
	GtkTextMark *mark_insert_start;
	GtkTextMark *mark_insert_end;

	const gchar *word;
	gboolean inserting;
	gboolean ignore_correction;
	gboolean ignore_correction_on_send;
	gint pos;
};

typedef struct _spellchk spellchk;

static GtkListStore *model;

static gboolean
is_word_uppercase(const gchar *word)
{
	for (; word[0] != '\0'; word = g_utf8_find_next_char (word, NULL)) {
		gunichar c = g_utf8_get_char(word);

		if (!(g_unichar_isupper(c) ||
		      g_unichar_ispunct(c) ||
		      g_unichar_isspace(c)))
			return FALSE;
	}

	return TRUE;
}

static gboolean
is_word_lowercase(const gchar *word)
{
	for (; word[0] != '\0'; word = g_utf8_find_next_char(word, NULL)) {
		gunichar c = g_utf8_get_char(word);

		if (!(g_unichar_islower(c) ||
		      g_unichar_ispunct(c) ||
		      g_unichar_isspace(c)))
			return FALSE;
	}

	return TRUE;
}

static gboolean
is_word_proper(const gchar *word)
{
	if (word[0] == '\0')
		return FALSE;

	if (!g_unichar_isupper(g_utf8_get_char_validated(word, -1)))
		return FALSE;

	return is_word_lowercase(g_utf8_offset_to_pointer(word, 1));
}

static gchar *
make_word_proper(const gchar *word)
{
	char buf[7];
	gchar *lower = g_utf8_strdown(word, -1);
	gint bytes;
	gchar *ret;

	bytes = g_unichar_to_utf8(g_unichar_toupper(g_utf8_get_char(word)), buf);
	buf[MIN(bytes, sizeof(buf) - 1)] = '\0';

	ret = g_strconcat(buf, g_utf8_offset_to_pointer(lower, 1), NULL);
	g_free(lower);

	return ret;
}

static gboolean
substitute_simple_buffer(GtkTextBuffer *buffer)
{
	GtkTextIter start;
	GtkTextIter end;
	GtkTreeIter treeiter;
	gchar *text = NULL;

	gtk_text_buffer_get_iter_at_offset(buffer, &start, 0);
	gtk_text_buffer_get_iter_at_offset(buffer, &end, 0);
	gtk_text_iter_forward_to_end(&end);

	text = gtk_text_buffer_get_text(buffer, &start, &end, FALSE);

	if (gtk_tree_model_get_iter_first(GTK_TREE_MODEL(model), &treeiter) && text) {
		do {
			GValue val1;
			const gchar *bad;
			gchar *cursor;
			glong char_pos;

			val1.g_type = 0;
			gtk_tree_model_get_value(GTK_TREE_MODEL(model), &treeiter, WORD_ONLY_COLUMN, &val1);
			if (g_value_get_boolean(&val1))
			{
				g_value_unset(&val1);
				continue;
			}
			g_value_unset(&val1);

			gtk_tree_model_get_value(GTK_TREE_MODEL(model), &treeiter, BAD_COLUMN, &val1);
			bad = g_value_get_string(&val1);

			/* using g_utf8_* to get /character/ offsets instead of byte offsets for buffer */
			if ((cursor = g_strrstr(text, bad)))
			{
				GValue val2;
				const gchar *good;

				val2.g_type = 0;
				gtk_tree_model_get_value(GTK_TREE_MODEL(model), &treeiter, GOOD_COLUMN, &val2);
				good = g_value_get_string(&val2);

				char_pos = g_utf8_pointer_to_offset(text, cursor);
				gtk_text_buffer_get_iter_at_offset(buffer, &start, char_pos);
				gtk_text_buffer_get_iter_at_offset(buffer, &end, char_pos + g_utf8_strlen(bad, -1));
				gtk_text_buffer_delete(buffer, &start, &end);

				gtk_text_buffer_get_iter_at_offset(buffer, &start, char_pos);
				gtk_text_buffer_insert(buffer, &start, good, -1);

				g_value_unset(&val2);
				g_free(text);

				g_value_unset(&val1);
				return TRUE;
			}

			g_value_unset(&val1);
		} while (gtk_tree_model_iter_next(GTK_TREE_MODEL(model), &treeiter));
	}

	g_free(text);
	return FALSE;
}

static gchar *
substitute_word(gchar *word)
{
	GtkTreeIter iter;
	gchar *outword;
	gchar *lowerword;
	gchar *foldedword;

	if (word == NULL)
		return NULL;

	lowerword = g_utf8_strdown(word, -1);
	foldedword = g_utf8_casefold(word, -1);

	if (gtk_tree_model_get_iter_first(GTK_TREE_MODEL(model), &iter)) {
		do {
			GValue val1;
			gboolean case_sensitive;
			const char *bad;
			gchar *tmpbad = NULL;

			val1.g_type = 0;
			gtk_tree_model_get_value(GTK_TREE_MODEL(model), &iter, WORD_ONLY_COLUMN, &val1);
			if (!g_value_get_boolean(&val1)) {
				g_value_unset(&val1);
				continue;
			}
			g_value_unset(&val1);

			gtk_tree_model_get_value(GTK_TREE_MODEL(model), &iter, CASE_SENSITIVE_COLUMN, &val1);
			case_sensitive = g_value_get_boolean(&val1);
			g_value_unset(&val1);

			gtk_tree_model_get_value(GTK_TREE_MODEL(model), &iter, BAD_COLUMN, &val1);
			bad = g_value_get_string(&val1);

			if ((case_sensitive && !strcmp(bad, word)) ||
			    (!case_sensitive && (!strcmp(bad, lowerword) ||
			                        (!is_word_lowercase(bad) &&
			                         !strcmp((tmpbad = g_utf8_casefold(bad, -1)), foldedword)))))
			{
				GValue val2;
				const char *good;

				g_free(tmpbad);

				val2.g_type = 0;
				gtk_tree_model_get_value(GTK_TREE_MODEL(model), &iter, GOOD_COLUMN, &val2);
				good = g_value_get_string(&val2);

				if (!case_sensitive && is_word_lowercase(bad) && is_word_lowercase(good))
				{
					if (is_word_uppercase(word))
						outword = g_utf8_strup(good, -1);
					else if (is_word_proper(word))
						outword = make_word_proper(good);
					else
						outword = g_strdup(good);
				}
				else
					outword = g_strdup(good);

				g_value_unset(&val1);
				g_value_unset(&val2);

				g_free(foldedword);
				return outword;
			}

			g_value_unset(&val1);
			g_free(tmpbad);

		} while (gtk_tree_model_iter_next(GTK_TREE_MODEL(model), &iter));
	}
	g_free(foldedword);

	return NULL;
}

static void
spellchk_free(spellchk *spell)
{
	GtkTextBuffer *buffer;

	g_return_if_fail(spell != NULL);

	buffer = gtk_text_view_get_buffer(spell->view);

	g_signal_handlers_disconnect_matched(buffer,
			G_SIGNAL_MATCH_DATA,
			0, 0, NULL, NULL,
			spell);
	g_free(spell);
}

/* Pango doesn't know about the "'" character.  Let's fix that. */
static gboolean
spellchk_inside_word(GtkTextIter *iter)
{
	gunichar ucs4_char = gtk_text_iter_get_char(iter);
	gchar *utf8_str;
	gchar c = 0;

	utf8_str = g_ucs4_to_utf8(&ucs4_char, 1, NULL, NULL, NULL);
	if (utf8_str != NULL)
	{
		c = utf8_str[0];
		g_free(utf8_str);
	}

	/* Hack because otherwise typing things like U.S. gets difficult
	 * if you have 'u' -> 'you' set as a correction...
	 *
	 * Part 1 of 2: This marks . as being an inside-word character. */
	if (c == '.')
		return TRUE;

	/* Avoid problems with \r, for example (SF #1289031). */
	if (c == '\\')
		return TRUE;

	if (gtk_text_iter_inside_word (iter) == TRUE)
		return TRUE;

	if (c == '\'') {
		gboolean result = gtk_text_iter_backward_char(iter);
		gboolean output = gtk_text_iter_inside_word(iter);

		if (result)
		{
			/*
			 * Hack so that "u'll" will correct correctly.
			 */
			ucs4_char = gtk_text_iter_get_char(iter);
			utf8_str = g_ucs4_to_utf8(&ucs4_char, 1, NULL, NULL, NULL);
			if (utf8_str != NULL)
			{
				c = utf8_str[0];
				g_free(utf8_str);

				if (c == 'u' || c == 'U')
				{
					gtk_text_iter_forward_char(iter);
					return FALSE;
				}
			}

			gtk_text_iter_forward_char(iter);
		}

		return output;
	}
	else if (c == '&')
		return TRUE;

	return FALSE;
}

static gboolean
spellchk_backward_word_start(GtkTextIter *iter)
{
	int output;
	int result;

	output = gtk_text_iter_backward_word_start(iter);

	/* It didn't work...  */
	if (!output)
		return FALSE;

	while (spellchk_inside_word(iter)) {
		result = gtk_text_iter_backward_char(iter);

		/* We can't go backwards anymore?  We're at the beginning of the word. */
		if (!result)
			return TRUE;

		if (!spellchk_inside_word(iter)) {
			gtk_text_iter_forward_char(iter);
			return TRUE;
		}

		output = gtk_text_iter_backward_word_start(iter);
		if (!output)
			return FALSE;
	}

	return TRUE;
}

static gboolean
check_range(spellchk *spell, GtkTextBuffer *buffer,
				GtkTextIter start, GtkTextIter end, gboolean sending)
{
	gboolean replaced;
	gboolean result;
	gchar *tmp;
	int period_count = 0;
	gchar *word;
	GtkTextMark *mark;
	GtkTextIter pos;

	if ((replaced = substitute_simple_buffer(buffer)))
	{
		mark = gtk_text_buffer_get_insert(buffer);
		gtk_text_buffer_get_iter_at_mark(buffer, &pos, mark);
		spell->pos = gtk_text_iter_get_offset(&pos);

		gtk_text_buffer_get_iter_at_mark(buffer, &start, mark);
		gtk_text_buffer_get_iter_at_mark(buffer, &end, mark);
	}

	if (!sending)
	{
		/* We need to go backwords to find out if we are inside a word or not. */
		gtk_text_iter_backward_char(&end);

		if (spellchk_inside_word(&end))
		{
			gtk_text_iter_forward_char(&end);
			return replaced;  /* We only pay attention to whole words. */
		}
	}

	/* We could be in the middle of a whitespace block.  Check for that. */
	result = gtk_text_iter_backward_char(&end);

	if (!spellchk_inside_word(&end))
	{
		if (result)
			gtk_text_iter_forward_char(&end);
		return replaced;