* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
- * License along with this library; if not, write to the
- * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
- * Boston, MA 02111-1307, USA.
+ * License along with this library. If not, see <http://www.gnu.org/licenses/>.
*/
/*
#include "gtktextiterprivate.h"
#include "gtkintl.h"
#include "gtkdebug.h"
-#include "gtkalias.h"
+
#include <string.h>
+
+/**
+ * SECTION:gtktextiter
+ * @Short_description: Text buffer iterator
+ * @Title: GtkTextIter
+ *
+ * You may wish to begin by reading the <link linkend="TextWidget">text widget
+ * conceptual overview</link> which gives an overview of all the objects and data
+ * types related to the text widget and how they work together.
+ */
+
+
#define FIX_OVERFLOWS(varname) if ((varname) == G_MININT) (varname) = G_MININT + 1
typedef struct _GtkTextRealIter GtkTextRealIter;
static void
check_invariants (const GtkTextIter *iter)
{
- if (gtk_debug_flags & GTK_DEBUG_TEXT)
+ if (gtk_get_debug_flags () & GTK_DEBUG_TEXT)
_gtk_text_iter_check (iter);
}
#else
g_slice_free (GtkTextIter, iter);
}
-GType
-gtk_text_iter_get_type (void)
+/**
+ * gtk_text_iter_assign:
+ * @iter: a #GtkTextIter
+ * @other: another #GtkTextIter
+ *
+ * Assigns the value of @other to @iter. This function
+ * is not useful in applications, because iterators can be assigned
+ * with <literal>GtkTextIter i = j;</literal>. The
+ * function is used by language bindings.
+ *
+ * Since: 3.2
+ **/
+void
+gtk_text_iter_assign (GtkTextIter *iter,
+ const GtkTextIter *other)
{
- static GType our_type = 0;
-
- if (our_type == 0)
- our_type = g_boxed_type_register_static (I_("GtkTextIter"),
- (GBoxedCopyFunc) gtk_text_iter_copy,
- (GBoxedFreeFunc) gtk_text_iter_free);
+ g_return_if_fail (iter != NULL);
+ g_return_if_fail (other != NULL);
- return our_type;
+ *iter = *other;
}
+G_DEFINE_BOXED_TYPE (GtkTextIter, gtk_text_iter,
+ gtk_text_iter_copy,
+ gtk_text_iter_free)
+
GtkTextLineSegment*
_gtk_text_iter_get_indexable_segment (const GtkTextIter *iter)
{
* anchor is returned (with no new reference count added). Otherwise,
* %NULL is returned.
*
- * Return value: the anchor at @iter
+ * Return value: (transfer none): the anchor at @iter
**/
GtkTextChildAnchor*
gtk_text_iter_get_child_anchor (const GtkTextIter *iter)
/**
* gtk_text_iter_ends_tag:
* @iter: an iterator
- * @tag: a #GtkTextTag, or %NULL
+ * @tag: (allow-none): a #GtkTextTag, or %NULL
*
* Returns %TRUE if @tag is toggled off at exactly this point. If @tag
* is %NULL, returns %TRUE if any tag is toggled off at this point. Note
/**
* gtk_text_iter_get_attributes:
* @iter: an iterator
- * @values: a #GtkTextAttributes to be filled in
+ * @values: (out): a #GtkTextAttributes to be filled in
*
* Computes the effect of any tags applied to this spot in the
* text. The @values parameter should be initialized to the default
}
/**
- * gtk_text_iter_forward_text_chars:
+ * gtk_text_iter_backward_text_chars:
* @iter: a #GtkTextIter
* @count: number of chars to move
*
}
/**
- * gtk_text_iter_backward_word_starts
+ * gtk_text_iter_backward_word_starts:
* @iter: a #GtkTextIter
* @count: number of times to move
*
}
/**
- * gtk_text_iter_backward_visible_word_starts
+ * gtk_text_iter_backward_visible_word_starts:
* @iter: a #GtkTextIter
* @count: number of times to move
*
/**
* gtk_text_iter_forward_find_char:
* @iter: a #GtkTextIter
- * @pred: a function to be called on each character
+ * @pred: (scope call): a function to be called on each character
* @user_data: user data for @pred
- * @limit: search limit, or %NULL for none
+ * @limit: (allow-none): search limit, or %NULL for none
*
* Advances @iter, calling @pred on each character. If
* @pred returns %TRUE, returns %TRUE and stops scanning.
/**
* gtk_text_iter_backward_find_char:
* @iter: a #GtkTextIter
- * @pred: function to be called on each character
+ * @pred: (scope call): function to be called on each character
* @user_data: user data for @pred
- * @limit: search limit, or %NULL for none
+ * @limit: (allow-none): search limit, or %NULL for none
*
* Same as gtk_text_iter_forward_find_char(), but goes backward from @iter.
*
forward_chars_with_skipping (GtkTextIter *iter,
gint count,
gboolean skip_invisible,
- gboolean skip_nontext)
+ gboolean skip_nontext,
+ gboolean skip_decomp)
{
gint i;
{
gboolean ignored = FALSE;
+ /* minimal workaround to avoid the infinite loop of bug #168247. */
+ if (gtk_text_iter_is_end (iter))
+ return;
+
if (skip_nontext &&
gtk_text_iter_get_char (iter) == GTK_TEXT_UNKNOWN_CHAR)
ignored = TRUE;
_gtk_text_btree_char_is_invisible (iter))
ignored = TRUE;
+ if (!ignored && skip_decomp)
+ {
+ /* being UTF8 correct sucks: this accounts for extra
+ offsets coming from canonical decompositions of
+ UTF8 characters (e.g. accented characters) which
+ g_utf8_normalize() performs */
+ gchar *normal;
+ gchar *casefold;
+ gchar buffer[6];
+ gint buffer_len;
+
+ buffer_len = g_unichar_to_utf8 (gtk_text_iter_get_char (iter), buffer);
+ casefold = g_utf8_casefold (buffer, buffer_len);
+ normal = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ i -= (g_utf8_strlen (normal, -1) - 1);
+ g_free (normal);
+ g_free (casefold);
+ }
+
gtk_text_iter_forward_char (iter);
if (!ignored)
}
}
+static const gchar *
+pointer_from_offset_skipping_decomp (const gchar *str,
+ gint offset)
+{
+ gchar *casefold, *normal;
+ const gchar *p, *q;
+
+ p = str;
+
+ while (offset > 0)
+ {
+ q = g_utf8_next_char (p);
+ casefold = g_utf8_casefold (p, q - p);
+ normal = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ offset -= g_utf8_strlen (normal, -1);
+ g_free (casefold);
+ g_free (normal);
+ p = q;
+ }
+
+ return p;
+}
+
+static gboolean
+exact_prefix_cmp (const gchar *string,
+ const gchar *prefix,
+ guint prefix_len)
+{
+ GUnicodeType type;
+
+ if (strncmp (string, prefix, prefix_len) != 0)
+ return FALSE;
+ if (string[prefix_len] == '\0')
+ return TRUE;
+
+ type = g_unichar_type (g_utf8_get_char (string + prefix_len));
+
+ /* If string contains prefix, check that prefix is not followed
+ * by a unicode mark symbol, e.g. that trailing 'a' in prefix
+ * is not part of two-char a-with-hat symbol in string. */
+ return type != G_UNICODE_SPACING_MARK &&
+ type != G_UNICODE_ENCLOSING_MARK &&
+ type != G_UNICODE_NON_SPACING_MARK;
+}
+
+static const gchar *
+utf8_strcasestr (const gchar *haystack,
+ const gchar *needle)
+{
+ gsize needle_len;
+ gsize haystack_len;
+ const gchar *ret = NULL;
+ gchar *p;
+ gchar *casefold;
+ gchar *caseless_haystack;
+ gint i;
+
+ g_return_val_if_fail (haystack != NULL, NULL);
+ g_return_val_if_fail (needle != NULL, NULL);
+
+ casefold = g_utf8_casefold (haystack, -1);
+ caseless_haystack = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ g_free (casefold);
+
+ needle_len = g_utf8_strlen (needle, -1);
+ haystack_len = g_utf8_strlen (caseless_haystack, -1);
+
+ if (needle_len == 0)
+ {
+ ret = (gchar *)haystack;
+ goto finally;
+ }
+
+ if (haystack_len < needle_len)
+ {
+ ret = NULL;
+ goto finally;
+ }
+
+ p = (gchar *)caseless_haystack;
+ needle_len = strlen (needle);
+ i = 0;
+
+ while (*p)
+ {
+ if (exact_prefix_cmp (p, needle, needle_len))
+ {
+ ret = pointer_from_offset_skipping_decomp (haystack, i);
+ goto finally;
+ }
+
+ p = g_utf8_next_char (p);
+ i++;
+ }
+
+finally:
+ g_free (caseless_haystack);
+
+ return ret;
+}
+
+static const gchar *
+utf8_strrcasestr (const gchar *haystack,
+ const gchar *needle)
+{
+ gsize needle_len;
+ gsize haystack_len;
+ const gchar *ret = NULL;
+ gchar *p;
+ gchar *casefold;
+ gchar *caseless_haystack;
+ gint i;
+
+ g_return_val_if_fail (haystack != NULL, NULL);
+ g_return_val_if_fail (needle != NULL, NULL);
+
+ casefold = g_utf8_casefold (haystack, -1);
+ caseless_haystack = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ g_free (casefold);
+
+ needle_len = g_utf8_strlen (needle, -1);
+ haystack_len = g_utf8_strlen (caseless_haystack, -1);
+
+ if (needle_len == 0)
+ {
+ ret = (gchar *)haystack;
+ goto finally;
+ }
+
+ if (haystack_len < needle_len)
+ {
+ ret = NULL;
+ goto finally;
+ }
+
+ i = haystack_len - needle_len;
+ p = g_utf8_offset_to_pointer (caseless_haystack, i);
+ needle_len = strlen (needle);
+
+ while (p >= caseless_haystack)
+ {
+ if (exact_prefix_cmp (p, needle, needle_len))
+ {
+ ret = pointer_from_offset_skipping_decomp (haystack, i);
+ goto finally;
+ }
+
+ p = g_utf8_prev_char (p);
+ i--;
+ }
+
+finally:
+ g_free (caseless_haystack);
+
+ return ret;
+}
+
+/* normalizes caseless strings and returns true if @s2 matches
+ the start of @s1 */
+static gboolean
+utf8_caselessnmatch (const gchar *s1,
+ const gchar *s2,
+ gssize n1,
+ gssize n2)
+{
+ gchar *casefold;
+ gchar *normalized_s1;
+ gchar *normalized_s2;
+ gint len_s1;
+ gint len_s2;
+ gboolean ret = FALSE;
+
+ g_return_val_if_fail (s1 != NULL, FALSE);
+ g_return_val_if_fail (s2 != NULL, FALSE);
+ g_return_val_if_fail (n1 > 0, FALSE);
+ g_return_val_if_fail (n2 > 0, FALSE);
+
+ casefold = g_utf8_casefold (s1, n1);
+ normalized_s1 = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ g_free (casefold);
+
+ casefold = g_utf8_casefold (s2, n2);
+ normalized_s2 = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ g_free (casefold);
+
+ len_s1 = strlen (normalized_s1);
+ len_s2 = strlen (normalized_s2);
+
+ if (len_s1 >= len_s2)
+ ret = (strncmp (normalized_s1, normalized_s2, len_s2) == 0);
+
+ g_free (normalized_s1);
+ g_free (normalized_s2);
+
+ return ret;
+}
+
static gboolean
lines_match (const GtkTextIter *start,
const gchar **lines,
gboolean visible_only,
gboolean slice,
+ gboolean case_insensitive,
GtkTextIter *match_start,
GtkTextIter *match_end)
{
}
if (match_start) /* if this is the first line we're matching */
- found = strstr (line_text, *lines);
+ {
+ if (!case_insensitive)
+ found = strstr (line_text, *lines);
+ else
+ found = utf8_strcasestr (line_text, *lines);
+ }
else
{
/* If it's not the first line, we have to match from the
* start of the line.
*/
- if (strncmp (line_text, *lines, strlen (*lines)) == 0)
- found = line_text;
+ if ((!case_insensitive &&
+ (strncmp (line_text, *lines, strlen (*lines)) == 0)) ||
+ (case_insensitive &&
+ utf8_caselessnmatch (line_text, *lines, strlen (line_text),
+ strlen (*lines))))
+ {
+ found = line_text;
+ }
else
found = NULL;
}
/* If match start needs to be returned, set it to the
* start of the search string.
*/
+ forward_chars_with_skipping (&next, offset,
+ visible_only, !slice, FALSE);
if (match_start)
- {
- *match_start = next;
-
- forward_chars_with_skipping (match_start, offset,
- visible_only, !slice);
- }
+ *match_start = next;
/* Go to end of search string */
- offset += g_utf8_strlen (*lines, -1);
-
- forward_chars_with_skipping (&next, offset,
- visible_only, !slice);
+ forward_chars_with_skipping (&next, g_utf8_strlen (*lines, -1),
+ visible_only, !slice, TRUE);
g_free (line_text);
/* pass NULL for match_start, since we don't need to find the
* start again.
*/
- return lines_match (&next, lines, visible_only, slice, NULL, match_end);
+ return lines_match (&next, lines, visible_only, slice, case_insensitive, NULL, match_end);
}
/* strsplit () that retains the delimiter as part of the string. */
static gchar **
strbreakup (const char *string,
const char *delimiter,
- gint max_tokens)
+ gint max_tokens,
+ gint *num_strings,
+ gboolean case_insensitive)
{
GSList *string_list = NULL, *slist;
gchar **str_array, *s;
+ gchar *casefold, *new_string;
guint i, n = 1;
g_return_val_if_fail (string != NULL, NULL);
do
{
guint len;
- gchar *new_string;
len = s - string + delimiter_len;
new_string = g_new (gchar, len + 1);
strncpy (new_string, string, len);
new_string[len] = 0;
+
+ if (case_insensitive)
+ {
+ casefold = g_utf8_casefold (new_string, -1);
+ g_free (new_string);
+ new_string = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ g_free (casefold);
+ }
+
string_list = g_slist_prepend (string_list, new_string);
n++;
string = s + delimiter_len;
if (*string)
{
n++;
- string_list = g_slist_prepend (string_list, g_strdup (string));
+
+ if (case_insensitive)
+ {
+ casefold = g_utf8_casefold (string, -1);
+ new_string = g_utf8_normalize (casefold, -1, G_NORMALIZE_NFD);
+ g_free (casefold);
+ }
+ else
+ new_string = g_strdup (string);
+
+ string_list = g_slist_prepend (string_list, new_string);
}
str_array = g_new (gchar*, n);
g_slist_free (string_list);
+ if (num_strings != NULL)
+ *num_strings = n - 1;
+
return str_array;
}
* @iter: start of search
* @str: a search string
* @flags: flags affecting how the search is done
- * @match_start: (allow-none): return location for start of match, or %NULL
- * @match_end: (allow-none): return location for end of match, or %NULL
+ * @match_start: (out caller-allocates) (allow-none): return location for start of match, or %NULL
+ * @match_end: (out caller-allocates) (allow-none): return location for end of match, or %NULL
* @limit: (allow-none): bound for the search, or %NULL for the end of the buffer
*
* Searches forward for @str. Any match is returned by setting
* pixbufs or child widgets mixed inside the matched range. If these
* flags are not given, the match must be exact; the special 0xFFFC
* character in @str will match embedded pixbufs or child widgets.
+ * If you specify the #GTK_TEXT_SEARCH_CASE_INSENSITIVE flag, the text will
+ * be matched regardless of what case it is in.
*
* Return value: whether a match was found
**/
GtkTextIter search;
gboolean visible_only;
gboolean slice;
-
+ gboolean case_insensitive;
+
g_return_val_if_fail (iter != NULL, FALSE);
g_return_val_if_fail (str != NULL, FALSE);
visible_only = (flags & GTK_TEXT_SEARCH_VISIBLE_ONLY) != 0;
slice = (flags & GTK_TEXT_SEARCH_TEXT_ONLY) == 0;
-
+ case_insensitive = (flags & GTK_TEXT_SEARCH_CASE_INSENSITIVE) != 0;
+
/* locate all lines */
- lines = strbreakup (str, "\n", -1);
+ lines = strbreakup (str, "\n", -1, NULL, case_insensitive);
search = *iter;
break;
if (lines_match (&search, (const gchar**)lines,
- visible_only, slice, &match, &end))
+ visible_only, slice, case_insensitive, &match, &end))
{
if (limit == NULL ||
(limit &&
}
static gboolean
-vectors_equal_ignoring_trailing (gchar **vec1,
- gchar **vec2)
+vectors_equal_ignoring_trailing (gchar **vec1,
+ gchar **vec2,
+ gboolean case_insensitive)
{
/* Ignores trailing chars in vec2's last line */
while (*i1 && *i2)
{
- if (strcmp (*i1, *i2) != 0)
+ gint len1;
+ gint len2;
+
+ if (!case_insensitive)
{
- if (*(i2 + 1) == NULL) /* if this is the last line */
+ if (strcmp (*i1, *i2) != 0)
{
- gint len1 = strlen (*i1);
- gint len2 = strlen (*i2);
-
- if (len2 >= len1 &&
- strncmp (*i1, *i2, len1) == 0)
+ if (*(i2 + 1) == NULL) /* if this is the last line */
{
- /* We matched ignoring the trailing stuff in vec2 */
- return TRUE;
+ len1 = strlen (*i1);
+ len2 = strlen (*i2);
+
+ if (len2 >= len1 &&
+ strncmp (*i1, *i2, len1) == 0)
+ {
+ /* We matched ignoring the trailing stuff in vec2 */
+ return TRUE;
+ }
+ else
+ {
+ return FALSE;
+ }
}
else
{
return FALSE;
}
}
- else
+ }
+ else
+ {
+ len1 = strlen (*i1);
+ len2 = strlen (*i2);
+
+ if (!utf8_caselessnmatch (*i1, *i2, len1, len2))
{
- return FALSE;
+ if (*(i2 + 1) == NULL) /* if this is the last line */
+ {
+ if (utf8_caselessnmatch (*i2, *i1, len2, len1))
+ {
+ /* We matched ignoring the trailing stuff in vec2 */
+ return TRUE;
+ }
+ else
+ {
+ return FALSE;
+ }
+ }
+ else
+ {
+ return FALSE;
+ }
}
}
+
++i1;
++i2;
}
if (*i1 || *i2)
- {
- return FALSE;
- }
+ return FALSE;
else
return TRUE;
}
{
gint n_lines;
gchar **lines;
+
GtkTextIter first_line_start;
GtkTextIter first_line_end;
- gboolean slice;
- gboolean visible_only;
+
+ guint slice : 1;
+ guint visible_only : 1;
};
static void
}
win->lines[i] = line_text;
+ win->first_line_start = line_start;
+ win->first_line_end = line_end;
line_end = line_start;
gtk_text_iter_backward_line (&line_start);
* @iter: a #GtkTextIter where the search begins
* @str: search string
* @flags: bitmask of flags affecting the search
- * @match_start: (allow-none): return location for start of match, or %NULL
- * @match_end: (allow-none): return location for end of match, or %NULL
+ * @match_start: (out caller-allocates) (allow-none): return location for start of match, or %NULL
+ * @match_end: (out caller-allocates) (allow-none): return location for end of match, or %NULL
* @limit: (allow-none): location of last possible @match_start, or %NULL for start of buffer
*
* Same as gtk_text_iter_forward_search(), but moves backward.
gboolean retval = FALSE;
gboolean visible_only;
gboolean slice;
+ gboolean case_insensitive;
g_return_val_if_fail (iter != NULL, FALSE);
g_return_val_if_fail (str != NULL, FALSE);
visible_only = (flags & GTK_TEXT_SEARCH_VISIBLE_ONLY) != 0;
slice = (flags & GTK_TEXT_SEARCH_TEXT_ONLY) == 0;
-
+ case_insensitive = (flags & GTK_TEXT_SEARCH_CASE_INSENSITIVE) != 0;
+
/* locate all lines */
- lines = strbreakup (str, "\n", -1);
-
- l = lines;
- n_lines = 0;
- while (*l)
- {
- ++n_lines;
- ++l;
- }
+ lines = strbreakup (str, "\n", -1, &n_lines, case_insensitive);
win.n_lines = n_lines;
win.slice = slice;
do
{
- gchar *first_line_match;
+ const gchar *first_line_match;
if (limit &&
gtk_text_iter_compare (limit, &win.first_line_end) > 0)
* end in '\n', so this will only match at the
* end of the first line, which is correct.
*/
- first_line_match = g_strrstr (*win.lines, *lines);
+ if (!case_insensitive)
+ first_line_match = g_strrstr (*win.lines, *lines);
+ else
+ first_line_match = utf8_strrcasestr (*win.lines, *lines);
if (first_line_match &&
- vectors_equal_ignoring_trailing (lines + 1, win.lines + 1))
+ vectors_equal_ignoring_trailing (lines + 1, win.lines + 1,
+ case_insensitive))
{
/* Match! */
gint offset;
next = win.first_line_start;
start_tmp = next;
forward_chars_with_skipping (&start_tmp, offset,
- visible_only, !slice);
+ visible_only, !slice, FALSE);
if (limit &&
gtk_text_iter_compare (limit, &start_tmp) > 0)
}
forward_chars_with_skipping (&next, offset,
- visible_only, !slice);
+ visible_only, !slice, TRUE);
if (match_end)
*match_end = next;
if (_gtk_text_line_is_last (real->line, real->tree))
g_error ("Iterator was on last line (past the end iterator)");
}
-
-#define __GTK_TEXT_ITER_C__
-#include "gtkaliasdef.c"