[API] Add compose() and decompose() unicode funcs, rename other ones

Add compose() and decompose() unicode funcs.  These implement
pair-wise canonical composition/decomposition.

The glib/icu implementations are lacking for now.  We are adding
API for this to glib, but I cannot find any useful API in ICU.
May end of implementing these in-house.

Changed all unicode_funcs callback names to remove the "_get" part.
Eg, hb_unicode_get_script_func_t is now hb_unicode_script_func_t,
and hb_unicode_get_script() is hb_unicode_script() now.

TODO

@@ -68,7 +68,9 @@ Tests to write:
 
 - ot-layout enumeration API (needs font)
 
-- Finish test-shape.c
+- Finish test-shape.c, grep for TODO
+
+- Finish test-unicode.c, grep for TODO
 
 
 Optimizations:

src/hb-glib.cc

@@ -231,7 +231,11 @@ hb_unicode_funcs_t _hb_glib_unicode_funcs = {
     hb_glib_get_eastasian_width,
     hb_glib_get_general_category,
     hb_glib_get_mirroring,
-    hb_glib_get_script
+    hb_glib_get_script,
+    /* TODO
+    hb_glib_compose,
+    hb_glib_decompose,
+    */
   }
 };
 

src/hb-icu.cc

@@ -174,7 +174,11 @@ hb_unicode_funcs_t _hb_icu_unicode_funcs = {
     hb_icu_get_eastasian_width,
     hb_icu_get_general_category,
     hb_icu_get_mirroring,
-    hb_icu_get_script
+    hb_icu_get_script,
+    /* TODO
+    hb_icu_compose,
+    hb_icu_decompose,
+    */
   }
 };
 

src/hb-ot-shape.cc

@@ -197,8 +197,8 @@ hb_set_unicode_props (hb_ot_shape_context_t *c)
 
   unsigned int count = c->buffer->len;
   for (unsigned int i = 1; i < count; i++) {
-    info[i].general_category() = hb_unicode_get_general_category (unicode, info[i].codepoint);
-    info[i].combining_class() = hb_unicode_get_combining_class (unicode, info[i].codepoint);
+    info[i].general_category() = hb_unicode_general_category (unicode, info[i].codepoint);
+    info[i].combining_class() = hb_unicode_combining_class (unicode, info[i].codepoint);
   }
 }
 
@@ -252,7 +252,7 @@ hb_mirror_chars (hb_ot_shape_context_t *c)
 
   unsigned int count = c->buffer->len;
   for (unsigned int i = 0; i < count; i++) {
-    hb_codepoint_t codepoint = hb_unicode_get_mirroring (unicode, c->buffer->info[i].codepoint);
+    hb_codepoint_t codepoint = hb_unicode_mirroring (unicode, c->buffer->info[i].codepoint);
     if (likely (codepoint == c->buffer->info[i].codepoint))
       c->buffer->info[i].mask |= rtlm_mask; /* XXX this should be moved to before setting user-feature masks */
     else

src/hb-shape.cc

@@ -63,7 +63,7 @@ hb_shape (hb_font_t          *font,
     hb_unicode_funcs_t *unicode = buffer->unicode;
     unsigned int count = buffer->len;
     for (unsigned int i = 0; i < count; i++) {
-      hb_script_t script = hb_unicode_get_script (unicode, buffer->info[i].codepoint);
+      hb_script_t script = hb_unicode_script (unicode, buffer->info[i].codepoint);
       if (likely (script != HB_SCRIPT_COMMON &&
 		  script != HB_SCRIPT_INHERITED &&
 		  script != HB_SCRIPT_UNKNOWN)) {

src/hb-unicode-private.hh

@@ -49,6 +49,8 @@ HB_BEGIN_DECLS
   HB_UNICODE_FUNC_IMPLEMENT (general_category) \
   HB_UNICODE_FUNC_IMPLEMENT (mirroring) \
   HB_UNICODE_FUNC_IMPLEMENT (script) \
+  HB_UNICODE_FUNC_IMPLEMENT (compose) \
+  HB_UNICODE_FUNC_IMPLEMENT (decompose) \
   /* ^--- Add new callbacks here */
 
 /* Simple callbacks are those taking a hb_codepoint_t and returning a hb_codepoint_t */
@@ -67,13 +69,13 @@ struct _hb_unicode_funcs_t {
 
   bool immutable;
 
-  /* Don't access these directly.  Call hb_unicode_get_*() instead. */
+  /* Don't access these directly.  Call hb_unicode_*() instead. */
 
   struct {
-#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_get_##name##_func_t name;
+#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_##name##_func_t name;
     HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
 #undef HB_UNICODE_FUNC_IMPLEMENT
-  } get;
+  } func;
 
   struct {
 #define HB_UNICODE_FUNC_IMPLEMENT(name) void *name;

src/hb-unicode.cc

@@ -40,45 +40,67 @@ HB_BEGIN_DECLS
  */
 
 static unsigned int
-hb_unicode_get_combining_class_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
-				    hb_codepoint_t      unicode   HB_UNUSED,
-				    void               *user_data HB_UNUSED)
+hb_unicode_combining_class_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+				hb_codepoint_t      unicode   HB_UNUSED,
+				void               *user_data HB_UNUSED)
 {
   return 0;
 }
 
 static unsigned int
-hb_unicode_get_eastasian_width_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
-				    hb_codepoint_t      unicode   HB_UNUSED,
-				    void               *user_data HB_UNUSED)
+hb_unicode_eastasian_width_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+				hb_codepoint_t      unicode   HB_UNUSED,
+				void               *user_data HB_UNUSED)
 {
   return 1;
 }
 
 static hb_unicode_general_category_t
-hb_unicode_get_general_category_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
-				     hb_codepoint_t      unicode   HB_UNUSED,
-				     void               *user_data HB_UNUSED)
+hb_unicode_general_category_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+				 hb_codepoint_t      unicode   HB_UNUSED,
+				 void               *user_data HB_UNUSED)
 {
   return HB_UNICODE_GENERAL_CATEGORY_OTHER_LETTER;
 }
 
 static hb_codepoint_t
-hb_unicode_get_mirroring_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
-			      hb_codepoint_t      unicode   HB_UNUSED,
-			      void               *user_data HB_UNUSED)
+hb_unicode_mirroring_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+			  hb_codepoint_t      unicode   HB_UNUSED,
+			  void               *user_data HB_UNUSED)
 {
   return unicode;
 }
 
 static hb_script_t
-hb_unicode_get_script_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
-			   hb_codepoint_t      unicode   HB_UNUSED,
-			   void               *user_data HB_UNUSED)
+hb_unicode_script_nil (hb_unicode_funcs_t *ufuncs    HB_UNUSED,
+		       hb_codepoint_t      unicode   HB_UNUSED,
+		       void               *user_data HB_UNUSED)
 {
   return HB_SCRIPT_UNKNOWN;
 }
 
+static hb_bool_t
+hb_unicode_compose_nil (hb_unicode_funcs_t *ufuncs,
+			hb_codepoint_t      a         HB_UNUSED,
+			hb_codepoint_t      b         HB_UNUSED,
+			hb_codepoint_t     *ab        HB_UNUSED,
+			void               *user_data HB_UNUSED)
+{
+  /* TODO handle Hangul jamo here? */
+  return FALSE;
+}
+
+static hb_bool_t
+hb_unicode_decompose_nil (hb_unicode_funcs_t *ufuncs,
+			  hb_codepoint_t      ab        HB_UNUSED,
+			  hb_codepoint_t     *a         HB_UNUSED,
+			  hb_codepoint_t     *b         HB_UNUSED,
+			  void               *user_data HB_UNUSED)
+{
+  /* TODO handle Hangul jamo here? */
+  return FALSE;
+}
+
 
 hb_unicode_funcs_t _hb_unicode_funcs_nil = {
   HB_OBJECT_HEADER_STATIC,
@@ -86,7 +108,7 @@ hb_unicode_funcs_t _hb_unicode_funcs_nil = {
   NULL, /* parent */
   TRUE, /* immutable */
   {
-#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_get_##name##_nil,
+#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_unicode_##name##_nil,
     HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
 #undef HB_UNICODE_FUNC_IMPLEMENT
   }
@@ -113,7 +135,7 @@ hb_unicode_funcs_create (hb_unicode_funcs_t *parent)
   hb_unicode_funcs_make_immutable (parent);
   ufuncs->parent = hb_unicode_funcs_reference (parent);
 
-  ufuncs->get = parent->get;
+  ufuncs->func = parent->func;
 
   /* We can safely copy user_data from parent since we hold a reference
    * onto it and it's immutable.  We should not copy the destroy notifiers
@@ -193,7 +215,7 @@ hb_unicode_funcs_get_parent (hb_unicode_funcs_t *ufuncs)
 										\
 void										\
 hb_unicode_funcs_set_##name##_func (hb_unicode_funcs_t		   *ufuncs,	\
-				    hb_unicode_get_##name##_func_t  func,	\
+				    hb_unicode_##name##_func_t	    func,	\
 				    void			   *user_data,	\
 				    hb_destroy_func_t		    destroy)	\
 {										\
@@ -204,11 +226,11 @@ hb_unicode_funcs_set_##name##_func (hb_unicode_funcs_t		   *ufuncs,	\
     ufuncs->destroy.name (ufuncs->user_data.name);				\
 										\
   if (func) {									\
-    ufuncs->get.name = func;							\
+    ufuncs->func.name = func;							\
     ufuncs->user_data.name = user_data;						\
     ufuncs->destroy.name = destroy;						\
   } else {									\
-    ufuncs->get.name = ufuncs->parent->get.name;				\
+    ufuncs->func.name = ufuncs->parent->func.name;				\
     ufuncs->user_data.name = ufuncs->parent->user_data.name;			\
     ufuncs->destroy.name = NULL;						\
   }										\
@@ -221,13 +243,30 @@ hb_unicode_funcs_set_##name##_func (hb_unicode_funcs_t		   *ufuncs,	\
 #define HB_UNICODE_FUNC_IMPLEMENT(return_type, name)				\
 										\
 return_type									\
-hb_unicode_get_##name (hb_unicode_funcs_t *ufuncs,				\
-		       hb_codepoint_t      unicode)				\
+hb_unicode_##name (hb_unicode_funcs_t *ufuncs,					\
+		   hb_codepoint_t      unicode)					\
 {										\
-  return ufuncs->get.name (ufuncs, unicode, ufuncs->user_data.name);		\
+  return ufuncs->func.name (ufuncs, unicode, ufuncs->user_data.name);		\
 }
     HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS_SIMPLE
 #undef HB_UNICODE_FUNC_IMPLEMENT
 
+hb_bool_t
+hb_unicode_compose (hb_unicode_funcs_t *ufuncs,
+		    hb_codepoint_t      a,
+		    hb_codepoint_t      b,
+		    hb_codepoint_t     *ab)
+{
+  return ufuncs->func.compose (ufuncs, a, b, ab, ufuncs->user_data.compose);
+}
+
+hb_bool_t
+hb_unicode_decompose (hb_unicode_funcs_t *ufuncs,
+		      hb_codepoint_t      ab,
+		      hb_codepoint_t     *a,
+		      hb_codepoint_t     *b)
+{
+  return ufuncs->func.decompose (ufuncs, ab, a, b, ufuncs->user_data.decompose);
+}
 
 HB_END_DECLS

src/hb-unicode.h

@@ -90,73 +90,103 @@ hb_unicode_funcs_get_parent (hb_unicode_funcs_t *ufuncs);
 
 /* typedefs */
 
-typedef unsigned int                  (*hb_unicode_get_combining_class_func_t)  (hb_unicode_funcs_t *ufuncs,
+typedef unsigned int			(*hb_unicode_combining_class_func_t)	(hb_unicode_funcs_t *ufuncs,
 										 hb_codepoint_t      unicode,
 										 void               *user_data);
-typedef unsigned int                  (*hb_unicode_get_eastasian_width_func_t)  (hb_unicode_funcs_t *ufuncs,
+typedef unsigned int			(*hb_unicode_eastasian_width_func_t)	(hb_unicode_funcs_t *ufuncs,
 										 hb_codepoint_t      unicode,
 										 void               *user_data);
-typedef hb_unicode_general_category_t (*hb_unicode_get_general_category_func_t) (hb_unicode_funcs_t *ufuncs,
+typedef hb_unicode_general_category_t	(*hb_unicode_general_category_func_t)	(hb_unicode_funcs_t *ufuncs,
 										 hb_codepoint_t      unicode,
 										 void               *user_data);
-typedef hb_codepoint_t                (*hb_unicode_get_mirroring_func_t)        (hb_unicode_funcs_t *ufuncs,
+typedef hb_codepoint_t			(*hb_unicode_mirroring_func_t)		(hb_unicode_funcs_t *ufuncs,
 										 hb_codepoint_t      unicode,
 										 void               *user_data);
-typedef hb_script_t                   (*hb_unicode_get_script_func_t)           (hb_unicode_funcs_t *ufuncs,
+typedef hb_script_t			(*hb_unicode_script_func_t)		(hb_unicode_funcs_t *ufuncs,
 										 hb_codepoint_t      unicode,
 										 void               *user_data);
 
+typedef hb_bool_t			(*hb_unicode_compose_func_t)		(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      a,
+										 hb_codepoint_t      b,
+										 hb_codepoint_t     *ab,
+										 void               *user_data);
+typedef hb_bool_t			(*hb_unicode_decompose_func_t)		(hb_unicode_funcs_t *ufuncs,
+										 hb_codepoint_t      ab,
+										 hb_codepoint_t     *a,
+										 hb_codepoint_t     *b,
+										 void               *user_data);
+
 /* setters */
 
 void
 hb_unicode_funcs_set_combining_class_func (hb_unicode_funcs_t *ufuncs,
-					   hb_unicode_get_combining_class_func_t combining_class_func,
+					   hb_unicode_combining_class_func_t combining_class_func,
 					   void *user_data, hb_destroy_func_t destroy);
 
 void
 hb_unicode_funcs_set_eastasian_width_func (hb_unicode_funcs_t *ufuncs,
-					   hb_unicode_get_eastasian_width_func_t eastasian_width_func,
+					   hb_unicode_eastasian_width_func_t eastasian_width_func,
 					   void *user_data, hb_destroy_func_t destroy);
 
 void
 hb_unicode_funcs_set_general_category_func (hb_unicode_funcs_t *ufuncs,
-					    hb_unicode_get_general_category_func_t general_category_func,
+					    hb_unicode_general_category_func_t general_category_func,
 					    void *user_data, hb_destroy_func_t destroy);
 
 void
 hb_unicode_funcs_set_mirroring_func (hb_unicode_funcs_t *ufuncs,
-				     hb_unicode_get_mirroring_func_t mirroring_func,
+				     hb_unicode_mirroring_func_t mirroring_func,
 				     void *user_data, hb_destroy_func_t destroy);
 
 void
 hb_unicode_funcs_set_script_func (hb_unicode_funcs_t *ufuncs,
-				  hb_unicode_get_script_func_t script_func,
+				  hb_unicode_script_func_t script_func,
 				  void *user_data, hb_destroy_func_t destroy);
 
+void
+hb_unicode_funcs_set_compose_func (hb_unicode_funcs_t *ufuncs,
+				   hb_unicode_compose_func_t compose_func,
+				   void *user_data, hb_destroy_func_t destroy);
+
+void
+hb_unicode_funcs_set_decompose_func (hb_unicode_funcs_t *ufuncs,
+				     hb_unicode_decompose_func_t decompose_func,
+				     void *user_data, hb_destroy_func_t destroy);
 
 
 /* accessors */
 
 unsigned int
-hb_unicode_get_combining_class (hb_unicode_funcs_t *ufuncs,
-				hb_codepoint_t unicode);
+hb_unicode_combining_class (hb_unicode_funcs_t *ufuncs,
+			    hb_codepoint_t unicode);
 
 unsigned int
-hb_unicode_get_eastasian_width (hb_unicode_funcs_t *ufuncs,
-				hb_codepoint_t unicode);
+hb_unicode_eastasian_width (hb_unicode_funcs_t *ufuncs,
+			    hb_codepoint_t unicode);
 
 hb_unicode_general_category_t
-hb_unicode_get_general_category (hb_unicode_funcs_t *ufuncs,
-				 hb_codepoint_t unicode);
+hb_unicode_general_category (hb_unicode_funcs_t *ufuncs,
+			     hb_codepoint_t unicode);
 
 hb_codepoint_t
-hb_unicode_get_mirroring (hb_unicode_funcs_t *ufuncs,
-			  hb_codepoint_t unicode);
+hb_unicode_mirroring (hb_unicode_funcs_t *ufuncs,
+		      hb_codepoint_t unicode);
 
 hb_script_t
-hb_unicode_get_script (hb_unicode_funcs_t *ufuncs,
-		       hb_codepoint_t unicode);
+hb_unicode_script (hb_unicode_funcs_t *ufuncs,
+		   hb_codepoint_t unicode);
 
+hb_bool_t
+hb_unicode_compose (hb_unicode_funcs_t *ufuncs,
+		    hb_codepoint_t      a,
+		    hb_codepoint_t      b,
+		    hb_codepoint_t     *ab);
+hb_bool_t
+hb_unicode_decompose (hb_unicode_funcs_t *ufuncs,
+		      hb_codepoint_t      ab,
+		      hb_codepoint_t     *a,
+		      hb_codepoint_t     *b);
 
 HB_END_DECLS
 

test/test-unicode.c

@@ -93,7 +93,7 @@ a_is_for_arabic_get_script (hb_unicode_funcs_t *ufuncs,
   } else {
     hb_unicode_funcs_t *parent = hb_unicode_funcs_get_parent (ufuncs);
 
-    return hb_unicode_get_script (parent, codepoint);
+    return hb_unicode_script (parent, codepoint);
   }
 }
 
@@ -457,7 +457,7 @@ typedef struct {
   { \
     #name, \
     (func_setter_func_t) hb_unicode_funcs_set_##name##_func, \
-    (getter_func_t) hb_unicode_get_##name, \
+    (getter_func_t) hb_unicode_##name, \
     name##_tests, \
     G_N_ELEMENTS (name##_tests), \
     name##_tests_more, \
@@ -667,8 +667,8 @@ test_unicode_subclassing_nil (data_fixture_t *f, gconstpointer user_data)
   hb_unicode_funcs_set_script_func (aa, a_is_for_arabic_get_script,
                                     &f->data[1], free_up);
 
-  g_assert_cmphex (hb_unicode_get_script (aa, 'a'), ==, HB_SCRIPT_ARABIC);
-  g_assert_cmphex (hb_unicode_get_script (aa, 'b'), ==, HB_SCRIPT_UNKNOWN);
+  g_assert_cmphex (hb_unicode_script (aa, 'a'), ==, HB_SCRIPT_ARABIC);
+  g_assert_cmphex (hb_unicode_script (aa, 'b'), ==, HB_SCRIPT_UNKNOWN);
 
   g_assert (!f->data[0].freed && !f->data[1].freed);
   hb_unicode_funcs_destroy (aa);
@@ -686,8 +686,8 @@ test_unicode_subclassing_default (data_fixture_t *f, gconstpointer user_data)
   hb_unicode_funcs_set_script_func (aa, a_is_for_arabic_get_script,
                                     &f->data[1], free_up);
 
-  g_assert_cmphex (hb_unicode_get_script (aa, 'a'), ==, HB_SCRIPT_ARABIC);
-  g_assert_cmphex (hb_unicode_get_script (aa, 'b'), ==, HB_SCRIPT_LATIN);
+  g_assert_cmphex (hb_unicode_script (aa, 'a'), ==, HB_SCRIPT_ARABIC);
+  g_assert_cmphex (hb_unicode_script (aa, 'b'), ==, HB_SCRIPT_LATIN);
 
   g_assert (!f->data[0].freed && !f->data[1].freed);
   hb_unicode_funcs_destroy (aa);
@@ -714,9 +714,9 @@ test_unicode_subclassing_deep (data_fixture_t *f, gconstpointer user_data)
   hb_unicode_funcs_set_script_func (aa, a_is_for_arabic_get_script,
                                     &f->data[1], free_up);
 
-  g_assert_cmphex (hb_unicode_get_script (aa, 'a'), ==, HB_SCRIPT_ARABIC);
-  g_assert_cmphex (hb_unicode_get_script (aa, 'b'), ==, HB_SCRIPT_LATIN);
-  g_assert_cmphex (hb_unicode_get_script (aa, '0'), ==, HB_SCRIPT_UNKNOWN);
+  g_assert_cmphex (hb_unicode_script (aa, 'a'), ==, HB_SCRIPT_ARABIC);
+  g_assert_cmphex (hb_unicode_script (aa, 'b'), ==, HB_SCRIPT_LATIN);
+  g_assert_cmphex (hb_unicode_script (aa, '0'), ==, HB_SCRIPT_UNKNOWN);
 
   g_assert (!f->data[0].freed && !f->data[1].freed);
   hb_unicode_funcs_destroy (aa);
@@ -779,6 +779,9 @@ test_unicode_script_roundtrip (gconstpointer user_data)
 }
 
 
+/* TODO test compose() and decompose() */
+
+
 int
 main (int argc, char **argv)
 {