tomoe-commits Mailing List for Tomoe

Revision: 1594
          http://tomoe.svn.sourceforge.net/tomoe/?rev=1594&view=rev
Author:   makeinu
Date:     2008-03-25 19:18:17 -0700 (Tue, 25 Mar 2008)

Log Message:
-----------
Reverse search order.

Modified Paths:
--------------
    tomoe/trunk/ChangeLog
    tomoe/trunk/module/dict/tomoe-unihan.c

Modified: tomoe/trunk/ChangeLog
===================================================================
--- tomoe/trunk/ChangeLog	2008-03-25 10:43:05 UTC (rev 1593)
+++ tomoe/trunk/ChangeLog	2008-03-26 02:18:17 UTC (rev 1594)
@@ -1,3 +1,7 @@
+2008-03-26  Takuro Ashie  <as...@ho...>
+
+	* module/dict/tomoe-unihan.c: Reverse search order.
+
 2008-03-25  Takuro Ashie  <as...@ho...>
 
 	* module/recognizer/tomoe-recognizer-simple.c

Modified: tomoe/trunk/module/dict/tomoe-unihan.c
===================================================================
--- tomoe/trunk/module/dict/tomoe-unihan.c	2008-03-25 10:43:05 UTC (rev 1593)
+++ tomoe/trunk/module/dict/tomoe-unihan.c	2008-03-26 02:18:17 UTC (rev 1594)
@@ -201,15 +201,15 @@
 _tomoe_unihan_search (TomoeQuery *query)
 {
     GList *list = NULL;
-    guint i;
+    gint i;
 
     if (tomoe_query_is_empty (query)) {
-        for (i = 0; i < G_N_ELEMENTS(unihan_array); i++)
+        for (i = (gint) G_N_ELEMENTS(unihan_array) - 1; i >= 0; i--)
             collect_all_chars(unihan_array[i].get (),
                               unihan_array[i].size (),
                               &list);
     } else {
-        for (i = 0; i < G_N_ELEMENTS(unihan_array); i++)
+        for (i = (gint) G_N_ELEMENTS(unihan_array) - 1; i >= 0; i--)
             collect_chars_by_query(unihan_array[i].get (),
                                    unihan_array[i].size (),
                                    query,


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

Revision: 1593
          http://tomoe.svn.sourceforge.net/tomoe/?rev=1593&view=rev
Author:   makeinu
Date:     2008-03-25 03:43:05 -0700 (Tue, 25 Mar 2008)

Log Message:
-----------
Make binary dictionary as default for handwriting recognition.

Modified Paths:
--------------
    tomoe/trunk/ChangeLog
    tomoe/trunk/module/recognizer/tomoe-recognizer-simple.c

Modified: tomoe/trunk/ChangeLog
===================================================================
--- tomoe/trunk/ChangeLog	2008-03-25 10:09:58 UTC (rev 1592)
+++ tomoe/trunk/ChangeLog	2008-03-25 10:43:05 UTC (rev 1593)
@@ -1,5 +1,17 @@
 2008-03-25  Takuro Ashie  <as...@ho...>
 
+	* module/recognizer/tomoe-recognizer-simple.c
+	  I'll make binary dictionary as default handwriting dictionary.
+	  If you have any objections, please show me alternative codes.
+
+	  TODO:
+	    - C version of converter to build it without ruby.
+	    - Install binary dictionaries to system wide directory.
+	    - Fall back to XML dictionary if no binary dictionary exists.
+	    - Add properties not only writings but also others (if possible).
+
+2008-03-25  Takuro Ashie  <as...@ho...>
+
 	* module/dict/tomoe-unihan.[ch], module/dict/tomoe-dict-unihan.c:
 	  Reduce heavy, very heavy wasted memories.
 	  It's not completed yet but already works with tomoe-gtk.

Modified: tomoe/trunk/module/recognizer/tomoe-recognizer-simple.c
===================================================================
--- tomoe/trunk/module/recognizer/tomoe-recognizer-simple.c	2008-03-25 10:09:58 UTC (rev 1592)
+++ tomoe/trunk/module/recognizer/tomoe-recognizer-simple.c	2008-03-25 10:43:05 UTC (rev 1593)
@@ -186,12 +186,12 @@
 
         language = tomoe_recognizer_get_language (TOMOE_RECOGNIZER (object));
         if (language)
-            dict_name = g_strconcat ("handwriting-", language, ".xml", NULL);
+            dict_name = g_strconcat ("handwriting-", language, ".bin", NULL);
         else
-            dict_name = g_strdup ("handwriting.xml");
+            dict_name = g_strdup ("handwriting.bin");
 
         filename = g_build_filename (RECOGNIZER_DATADIR, dict_name, NULL);
-        recognizer->dict = tomoe_dict_new ("xml",
+        recognizer->dict = tomoe_dict_new ("binary",
                                            "filename", filename,
                                            NULL);
         if (recognizer->dict && !tomoe_dict_is_available (recognizer->dict)) {


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.

Revision: 1592
          http://tomoe.svn.sourceforge.net/tomoe/?rev=1592&view=rev
Author:   makeinu
Date:     2008-03-25 03:09:58 -0700 (Tue, 25 Mar 2008)

Log Message:
-----------
Reduce heavy wasted memories.

Modified Paths:
--------------
    tomoe/trunk/ChangeLog
    tomoe/trunk/module/dict/tomoe-dict-unihan.c
    tomoe/trunk/module/dict/tomoe-unihan.c
    tomoe/trunk/module/dict/tomoe-unihan.h

Modified: tomoe/trunk/ChangeLog
===================================================================
--- tomoe/trunk/ChangeLog	2008-02-15 12:59:03 UTC (rev 1591)
+++ tomoe/trunk/ChangeLog	2008-03-25 10:09:58 UTC (rev 1592)
@@ -1,3 +1,9 @@
+2008-03-25  Takuro Ashie  <as...@ho...>
+
+	* module/dict/tomoe-unihan.[ch], module/dict/tomoe-dict-unihan.c:
+	  Reduce heavy, very heavy wasted memories.
+	  It's not completed yet but already works with tomoe-gtk.
+	
 2008-02-15  Kouhei Sutou  <ko...@co...>
 
 	* test/tomoe-test-utils.rb: added benchmark related paths.

Modified: tomoe/trunk/module/dict/tomoe-dict-unihan.c
===================================================================
--- tomoe/trunk/module/dict/tomoe-dict-unihan.c	2008-02-15 12:59:03 UTC (rev 1591)
+++ tomoe/trunk/module/dict/tomoe-dict-unihan.c	2008-03-25 10:09:58 UTC (rev 1592)
@@ -38,7 +38,6 @@
 #include <glib-utils.h>
 
 #include "tomoe-unihan.h"
-#include "tomoe-dict-ptr-array.h"
 
 #define TOMOE_TYPE_DICT_UNIHAN            tomoe_type_dict_unihan
 #define TOMOE_DICT_UNIHAN(obj)            (G_TYPE_CHECK_INSTANCE_CAST ((obj), TOMOE_TYPE_DICT_UNIHAN, TomoeDictUnihan))
@@ -58,17 +57,17 @@
 typedef struct _TomoeDictUnihanClass TomoeDictUnihanClass;
 struct _TomoeDictUnihan
 {
-    TomoeDictPtrArray    object;
-    gchar               *name;
+    TomoeDict    object;
+    gchar       *name;
 };
 
 struct _TomoeDictUnihanClass
 {
-    TomoeDictPtrArrayClass parent_class;
+    TomoeDictClass parent_class;
 };
 
 static GType tomoe_type_dict_unihan = 0;
-static TomoeDictPtrArray *parent_class;
+static TomoeDict *parent_class;
 static TomoeDictUnihan *the_singleton = NULL;
 
 static GObject     *constructor               (GType                  type,
@@ -89,6 +88,17 @@
 static gboolean     is_available              (TomoeDict     *dict);
 static gchar       *get_available_private_utf8 (TomoeDict    *dict);
 
+static gboolean     register_char             (TomoeDict     *dict,
+                                               TomoeChar     *chr);
+static gboolean     unregister_char           (TomoeDict     *dict,
+                                               const gchar   *utf8);
+static TomoeChar   *get_char                  (TomoeDict     *dict,
+                                               const gchar   *utf8);
+static GList       *search                    (TomoeDict     *dict,
+                                               TomoeQuery    *query);
+static gboolean     copy                      (TomoeDict     *src_dict,
+                                               TomoeDict     *dest_dict);
+
 static void
 class_init (TomoeDictUnihanClass *klass)
 {
@@ -106,6 +116,13 @@
 
     dict_class = TOMOE_DICT_CLASS (klass);
     dict_class->get_name        = get_name;
+#if 1
+    dict_class->register_char   = register_char;
+    dict_class->unregister_char = unregister_char;
+    dict_class->get_char        = get_char;
+    dict_class->search          = search;
+    dict_class->copy            = copy;
+#endif
     dict_class->flush           = flush;
     dict_class->is_editable     = is_editable;
     dict_class->is_available    = is_available;
@@ -145,7 +162,7 @@
         };
 
     tomoe_type_dict_unihan = g_type_module_register_type (type_module,
-                                                          TOMOE_TYPE_DICT_PTR_ARRAY,
+                                                          TOMOE_TYPE_DICT,
                                                           "TomoeDictUnihan",
                                                           &info, 0);
 }
@@ -189,12 +206,8 @@
     GObjectClass *klass = G_OBJECT_CLASS (parent_class);
 
     if (!the_singleton) {
-        GPtrArray *chars;
         object = klass->constructor (type, n_props, props);
         the_singleton = TOMOE_DICT_UNIHAN (object);
-
-        chars = tomoe_dict_ptr_array_get_array (TOMOE_DICT_PTR_ARRAY (object));
-        _tomoe_unihan_create (chars);
     } else {
         object = g_object_ref (G_OBJECT (the_singleton));
     }
@@ -288,6 +301,44 @@
     return NULL;
 }
 
+
+static gboolean
+register_char (TomoeDict *dict, TomoeChar *chr)
+{
+    /* FIXME: need? */
+    return FALSE;
+}
+
+static gboolean
+unregister_char (TomoeDict *dict, const gchar *utf8)
+{
+    /* FIXME: need? */
+    return FALSE;
+}
+
+static TomoeChar *
+get_char (TomoeDict *dict, const gchar *utf8)
+{
+    g_return_val_if_fail (TOMOE_IS_DICT (dict), NULL);
+
+    return _tomoe_unihan_get_char (utf8);
+}
+
+static GList *
+search (TomoeDict *dict, TomoeQuery *query)
+{
+    g_return_val_if_fail (TOMOE_IS_DICT (dict), NULL);
+
+    return _tomoe_unihan_search (query);
+}
+
+static gboolean
+copy (TomoeDict *src_dict, TomoeDict *dest_dict)
+{
+    /* FIXME */
+    return FALSE;
+}
+
 /*
 vi:ts=4:nowrap:ai:expandtab
 */

Modified: tomoe/trunk/module/dict/tomoe-unihan.c
===================================================================
--- tomoe/trunk/module/dict/tomoe-unihan.c	2008-02-15 12:59:03 UTC (rev 1591)
+++ tomoe/trunk/module/dict/tomoe-unihan.c	2008-03-25 10:09:58 UTC (rev 1592)
@@ -6,122 +6,214 @@
 */
 
 
+#include <tomoe-candidate.h>
 #include "tomoe-unihan.h"
 #include "tomoe-unihan-data.h"
 
-static gint
-_tomoe_unihan_append_infos (GPtrArray *array, TomoeUnihanInfo *infos,
-                            gint infos_size, gint offset)
+static struct UnihanArray {
+    TomoeUnihanInfo *(*get)  (void);
+    gint             (*size) (void);
+} unihan_array[] = {
+    { _tomoe_unihan_data0, _tomoe_unihan_data0_size},
+    { _tomoe_unihan_data1, _tomoe_unihan_data1_size},
+    { _tomoe_unihan_data2, _tomoe_unihan_data2_size},
+    { _tomoe_unihan_data3, _tomoe_unihan_data3_size},
+    { _tomoe_unihan_data4, _tomoe_unihan_data4_size},
+    { _tomoe_unihan_data5, _tomoe_unihan_data5_size},
+    { _tomoe_unihan_data6, _tomoe_unihan_data6_size},
+    { _tomoe_unihan_data7, _tomoe_unihan_data7_size},
+    { _tomoe_unihan_data8, _tomoe_unihan_data8_size},
+    { _tomoe_unihan_data9, _tomoe_unihan_data9_size},
+};
+
+static TomoeChar *
+_tomoe_char_from_unihan (TomoeUnihanInfo *info)
 {
-    gint i;
+    TomoeChar *chr;
+    gint j, len;
 
-    for (i = 0; i < infos_size; i++) {
-        TomoeUnihanInfo info;
-        TomoeChar *chr;
-        gint j, len;
+    g_return_val_if_fail(info, NULL);
 
-        info = infos[i];
+    chr = tomoe_char_new ();
+    tomoe_char_set_utf8 (chr, info->utf8);
 
-        chr = tomoe_char_new ();
-        tomoe_char_set_utf8 (chr, info.utf8);
+    if (info->n_strokes)
+        tomoe_char_set_n_strokes (chr, info->n_strokes);
 
-        if (info.n_strokes)
-            tomoe_char_set_n_strokes (chr, info.n_strokes);
+    len = info->readings_size;
+    for (j = 0; j < len; j++) {
+        TomoeReading *reading;
+        TomoeUnihanReading *reading_info;
 
-        len = info.readings_size;
-        for (j = 0; j < len; j++) {
-            TomoeReading *reading;
-            TomoeUnihanReading reading_info;
+        reading_info = &info->readings[j];
+        reading = tomoe_reading_new (reading_info->type,
+                                     reading_info->reading);
+        tomoe_char_add_reading (chr, reading);
+        g_object_unref (reading);
+    }
 
-            reading_info = info.readings[j];
-            reading = tomoe_reading_new (reading_info.type,
-                                         reading_info.reading);
-            tomoe_char_add_reading (chr, reading);
-            g_object_unref (reading);
-        }
+    len = info->radicals_size;
+    for (j = 0; j < len; j++) {
+        tomoe_char_add_radical (chr, info->radicals[j]);
+    }
 
-        len = info.radicals_size;
-        for (j = 0; j < len; j++) {
-            tomoe_char_add_radical (chr, info.radicals[j]);
-        }
+    len = info->variants_size;
+    for (j = 0; j < len; j++) {
+        tomoe_char_set_variant (chr, info->variants[j]);
+    }
 
-        len = info.variants_size;
-        for (j = 0; j < len; j++) {
-            tomoe_char_set_variant (chr, info.variants[j]);
-        }
+    len = info->meta_data_size;
+    for (j = 0; j < len; j++) {
+        TomoeUnihanMetaData meta_data;
 
-        len = info.meta_data_size;
-        for (j = 0; j < len; j++) {
-            TomoeUnihanMetaData meta_data;
+        meta_data = info->meta_data[j];
+        tomoe_char_register_meta_data (chr, meta_data.key, meta_data.value);
+    }
 
-            meta_data = info.meta_data[j];
-            tomoe_char_register_meta_data (chr, meta_data.key, meta_data.value);
+    return chr;
+}
+
+static gboolean
+does_match_char_with_utf8 (TomoeUnihanInfo *info, const gchar *utf8)
+{
+    if (!utf8)
+        return TRUE;
+
+    return g_utf8_collate (info->utf8, utf8) == 0;
+}
+
+static gboolean
+does_match_char_with_n_strokes (TomoeUnihanInfo *info, gint min, gint max)
+{
+    gint n_strokes;
+
+    if (min < 0 && max < 0)
+        return TRUE;
+
+    n_strokes = info->n_strokes;
+    if (n_strokes < 0)
+        return FALSE;
+
+    return ((min < 0 || min <= n_strokes) &&
+            (max < 0 || max >= n_strokes));
+}
+
+static gboolean
+does_match_char_with_reading (TomoeUnihanInfo *info, TomoeReading *reading)
+{
+    TomoeReadingType type;
+    gint j;
+
+    if (!reading)
+        return TRUE;
+    if (!tomoe_reading_get_reading (reading))
+        return TRUE;
+
+    type = tomoe_reading_get_reading_type (reading);
+
+    for (j = 0; j < info->readings_size; j++) {
+        TomoeUnihanReading *reading_info;
+
+        reading_info = &info->readings[j];
+
+        if (type != reading_info->type &&
+            (type != TOMOE_READING_UNKNOWN &&
+             reading_info->type != TOMOE_READING_UNKNOWN))
+        {
+            return FALSE;
         }
 
-        array->pdata[i + offset] = chr;
+        if (g_str_has_prefix (reading_info->reading,
+                              tomoe_reading_get_reading(reading)))
+            return TRUE;
     }
 
-    return i + offset;
+    return FALSE;
 }
 
-void
-_tomoe_unihan_create (GPtrArray *array)
+static gboolean
+does_match_char_with_readings (TomoeUnihanInfo *info, const GList *readings)
 {
-    gint offset, infos_size;
+    GList *node;
 
-    infos_size =
-        _tomoe_unihan_data0_size () +
-        _tomoe_unihan_data1_size () +
-        _tomoe_unihan_data2_size () +
-        _tomoe_unihan_data3_size () +
-        _tomoe_unihan_data4_size () +
-        _tomoe_unihan_data5_size () +
-        _tomoe_unihan_data6_size () +
-        _tomoe_unihan_data7_size () +
-        _tomoe_unihan_data8_size () +
-        _tomoe_unihan_data9_size ();
-    g_ptr_array_set_size (array, infos_size);
-    array->len = infos_size;
+    for (node = (GList *)readings; node; node = g_list_next (node)) {
+        TomoeReading *reading = node->data;
+        if (!does_match_char_with_reading (info, reading))
+            return FALSE;
+    }
 
-    offset = 0;
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data0 (),
-                                         _tomoe_unihan_data0_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data1 (),
-                                         _tomoe_unihan_data1_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data2 (),
-                                         _tomoe_unihan_data2_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data3 (),
-                                         _tomoe_unihan_data3_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data4 (),
-                                         _tomoe_unihan_data4_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data5 (),
-                                         _tomoe_unihan_data5_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data6 (),
-                                         _tomoe_unihan_data6_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data7 (),
-                                         _tomoe_unihan_data7_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data8 (),
-                                         _tomoe_unihan_data8_size (),
-                                         offset);
-    offset = _tomoe_unihan_append_infos (array,
-                                         _tomoe_unihan_data9 (),
-                                         _tomoe_unihan_data9_size (),
-                                         offset);
+    return TRUE;
 }
+
+static void
+collect_chars_by_query (TomoeUnihanInfo *infos,
+                        gint size,
+                        TomoeQuery *q,
+                        GList **list)
+{
+    gint i;
+
+    for (i = size - 1; i >= 0; i--) {
+        TomoeUnihanInfo *info = &infos[i];
+
+        /* FIXME: add remaining properties */
+        if (does_match_char_with_utf8 (
+                info, tomoe_query_get_utf8 (q)) &&
+            does_match_char_with_n_strokes (
+                info,
+                tomoe_query_get_min_n_strokes (q),
+                tomoe_query_get_max_n_strokes (q)) &&
+            does_match_char_with_readings (
+                info,
+                tomoe_query_get_readings (q)))
+        {
+            TomoeChar *chr = _tomoe_char_from_unihan(info);
+            if (!chr) continue;
+            *list = g_list_prepend (*list, tomoe_candidate_new (chr));
+            g_object_unref(G_OBJECT(chr));
+        }
+    }
+}
+
+static void
+collect_all_chars (TomoeUnihanInfo *infos,
+                   gint size,
+                   GList **list)
+{
+    gint i;
+
+    for (i = size - 1; i >= 0; i--) {
+        TomoeUnihanInfo *info = &infos[i];
+        TomoeChar *chr = _tomoe_char_from_unihan(info);
+        if (!chr) continue;
+        *list = g_list_prepend (*list, tomoe_candidate_new (chr));
+        g_object_unref(G_OBJECT(chr));
+    }
+}
+
+TomoeChar *_tomoe_unihan_get_char (const gchar *utf8)
+{
+    /* FIXME */
+    return NULL;
+}
+
+GList *
+_tomoe_unihan_search (TomoeQuery *query)
+{
+    GList *list = NULL;
+    guint i;
+
+    if (tomoe_query_is_empty (query)) {
+        for (i = 0; i < G_N_ELEMENTS(unihan_array); i++)
+            collect_all_chars(unihan_array[i].get (),
+                              unihan_array[i].size (),
+                              &list);
+    } else {
+        for (i = 0; i < G_N_ELEMENTS(unihan_array); i++)
+            collect_chars_by_query(unihan_array[i].get (),
+                                   unihan_array[i].size (),
+                                   query,
+                                   &list);
+    }
+    return list;
+}

Modified: tomoe/trunk/module/dict/tomoe-unihan.h
===================================================================
--- tomoe/trunk/module/dict/tomoe-unihan.h	2008-02-15 12:59:03 UTC (rev 1591)
+++ tomoe/trunk/module/dict/tomoe-unihan.h	2008-03-25 10:09:58 UTC (rev 1592)
@@ -24,10 +24,13 @@
 #define __TOMOE_UNIHAN_H__
 
 #include <glib.h>
+#include <tomoe-char.h>
+#include <tomoe-query.h>
 
 G_BEGIN_DECLS
 
-void _tomoe_unihan_create (GPtrArray *array);
+TomoeChar *_tomoe_unihan_get_char (const gchar *utf8);
+GList     *_tomoe_unihan_search   (TomoeQuery  *query);
 
 G_END_DECLS
 


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.