Menu
▾
▴
Re: [E-devel] [Patch][elementary] using textblocks in the layout as an accessibility object (with focus chain)
|
From: Carsten H. (T. R. <ra...@ra...> - 2012-06-20 09:40:51
|
On Wed, 20 Jun 2012 16:59:13 +0900 Kim Shinwoo <kim...@gm...> said:
now its good cedric-> put it n! :)
> You always ENLIGHTEN me. Thanks.
> I have revised patch with EAPI. :-]
>
> 2012/6/20 Carsten Haitzler <ra...@ra...>
>
> > On Mon, 18 Jun 2012 17:51:59 +0900 Kim Shinwoo <kim...@gm...>
> > said:
> >
> > _elm_access_highlight_set()
> > and
> > _elm_access_edje_object_part_object_register()
> >
> > need EAPI. thery may be accessed from a MODULE (thus why the other elm
> > widget
> > api's are EAPI too). this means its an exposed symbol - but instead for
> > tightly-coupled external widgets and modules, so this api is not stable and
> > guaranteed (at this stage) as per the docs in that header file.
> >
> > > oops. the patch is coming..
> > >
> > > 2012/6/18 Kim Shinwoo <kim...@gm...>
> > >
> > > > Dear All, Hello
> > > >
> > > > The attached patch has following features. Please review and give some
> > > > feedbacks.
> > > >
> > > > 1. I removed all EAPIs from my previous patch.
> > > > 2. The accessibility supports all layout on the ELM_ACCESS_MODE_ON.
> > > > sense.
> > > > 3. The default string of text part is not supported by the
> > > > accessibility. Because there is no way to get the default string.
> > > > (Raster also said that the default string is for the design
> > factor.)
> > > > 4. The previously set access_object will not be removed but be
> > reused.
> > > > The text is retrieved by using callback of the access_obj.
> > > >
> > > > Sincerely,
> > > > Shinwoo Kim.
> > > >
> > > >
> > > >
> > > > 2012/6/12 Kim Shinwoo <kim...@gm...>
> > > >
> > > >> Dear Mr. Gustavo Lima Chaves
> > > >>
> > > >> hey, sorry for below query. please ignore it.
> > > >>
> > > >>
> > > >> "Yeap, EAPIs.. if application developer would use this APIs, then
> > those
> > > >> APIs should be in the elm_access.h?
> > > >> and the elm_access.h would be included in the elm_widget.h. right?"
> > > >>
> > > >> then, thanks.
> > > >>
> > > >>
> > > >> 2012/6/12 Kim Shinwoo <kim...@gm...>
> > > >>
> > > >>> Dear Mr. Gustavo Lima Chaves
> > > >>> Hello, Thanks for your response.
> > > >>>
> > > >>>
> > > >>> > Besides you had previous internal access calls EAPId there, remove
> > > >>> > them, as I said :)
> > > >>>
> > > >>> Yeap, EAPIs.. if application developer would use this APIs, then
> > those
> > > >>> APIs should be in the elm_access.h?
> > > >>> and the elm_access.h would be included in the elm_widget.h. right?
> > > >>>
> > > >>>
> > > >>> > Do we really want to be selective on each layout should get
> > > >>> > accessibility support on it's text??
> > > >>>
> > > >>> No, not really want to..
> > > >>> All layouts should be support the accessibility that makes sense.
> > > >>>
> > > >>>
> > > >>> > It seems we can't list an Edje object's parts, so that we apply to
> > > >>> > each textual ones the accessibility thing on the layout creation,
> > not
> > > >>> > just the text setting of it :/.
> > > >>> >
> > > >>>
> > > >>> yeap, you right. the creation time should be considered.
> > > >>>
> > > >>>
> > > >>> > Don't forget to delete a previously set access_object on
> > sub_d->obj,
> > > >>> > here.
> > > >>> >
> > > >>>
> > > >>> good point, the revised patch will take care of this.
> > > >>>
> > > >>>
> > > >>> > What about _elm_access_text_register(Evas_Object *parent,
> > Evas_Object
> > > >>> > *text_obj), and we fetch the text string from the object
> > > >>> > internally. The item version of it I dislike, as it just highlights
> > > >>> > things on mouse in, with apparently no text attached.
> > > >>> >
> > > >>>
> > > >>> This _elm_access_text_set() is for the ELM_ACCESS_TYPE. Accessibility
> > > >>> works as following.
> > > >>> If there is a button which has label "OK" and the button is disabled,
> > > >>> then accessibility says...
> > > >>> "Button, OK, Disabled" (ordered by ELM_ACCESS_TYPE, ELM_ACCESS_INFO,
> > > >>> ELM_ACCESS_STATE)
> > > >>>
> > > >>> catch you later with revised patch. thanks.
> > > >>>
> > > >>> Sincerly,
> > > >>> Shinwoo Kim.
> > > >>>
> > > >>>
> > > >>>
> > > >>> 2012/6/11 Gustavo Lima Chaves <gl...@pr...>:
> > > >>>
> > > >>> > * Kim Shinwoo <kim...@gm...> [2012-06-08 19:55:40
> > +0900]:
> > > >>> >
> > > >>> >> Dear all,
> > > >>> >>
> > > >>> >> Revised patch has followings
> > > >>> >> 1. The TEXT part is contemplated also
> > > >>> >> 2. The Elm_Widget_Smart_Class is parent class of elm_access not
> > the
> > > >>> >> Elm_Layout_Smart_Class
> > > >>> >> 3. The elm_access_edje_object_part_object_register() is added
> > > >>> >> - Why I does not integrated with _elm_access_add is there
> > are
> > > >>> >> following APIs for the consistency
> > > >>> >> a. _elm_access_object_register(): is for the widget
> > object
> > > >>> >> b. _elm_access_item_register(): is for the widget item
> > > >>> >>
> > > >>> >> I have attached EXAMPLE code also. Thanks.
> > > >>> >>
> > > >>> >> Sincerely,
> > > >>> >> Shinwoo Kim.
> > > >>> >>
> > > >>> >
> > > >>> > Hi, Kim Shinwoo.
> > > >>> >
> > > >>> > Here go my thoughts on it, then.
> > > >>> >
> > > >>> >> Index: src/lib/elm_widget.h
> > > >>> >>
> > ===================================================================
> > > >>> >> --- src/lib/elm_widget.h (리비전 71695)
> > > >>> >> +++ src/lib/elm_widget.h (작업 사본)
> > > >>> >> @@ -478,6 +478,7 @@ typedef struct _Elm_Widget_Smart_Data
> > > >>> >> * TRUE
> > by
> > > >>> >> *
> > default
> > > >>> */
> > > >>> >> Eina_Bool still_in : 1;
> > > >>> >> + Eina_Bool can_access : 1;
> > > >>> >> } Elm_Widget_Smart_Data;
> > > >>> >>
> > > >>> >> /**
> > > >>> >> @@ -544,6 +545,8 @@ EAPI void
> > _elm_access_object_register(
> > > >>> >> EAPI void _elm_access_item_unregister(Elm_Widget_Item
> > > >>> *item);
> > > >>> >> EAPI void _elm_access_item_register(Elm_Widget_Item
> > > >>> *item, Evas_Object *hoverobj);
> > > >>> >> EAPI Eina_Bool _elm_access_2nd_click_timeout(Evas_Object
> > > >>> *obj);
> > > >>> >> +EAPI Evas_Object*
> > > >>> _elm_access_edje_object_part_object_register(Evas_Object *obj, const
> > > >>> Evas_Object *partobj, const char *text);
> > > >>> >> +EAPI void _elm_access_highlight_set(Evas_Object*
> > obj);
> > > >>> >
> > > >>> > Besides you had previous internal access calls EAPId there, remove
> > > >>> > them, as I said :)
> > > >>>
> > > >>> Yeap, EAPIs.. if application developer would use this APIs, then
> > those
> > > >>> APIs should be in the elm_access.h?
> > > >>> and the elm_access.h would be included in the elm_widget.h. right?
> > > >>>
> > > >>>
> > > >>> >
> > > >>> >>
> > > >>> >> /**< put this as the first member in your widget item struct */
> > > >>> >> #define ELM_WIDGET_ITEM Elm_Widget_Item base
> > > >>> >> Index: src/lib/elm_widget.c
> > > >>> >>
> > ===================================================================
> > > >>> >> --- src/lib/elm_widget.c (리비전 71695)
> > > >>> >> +++ src/lib/elm_widget.c (작업 사본)
> > > >>> >> @@ -839,6 +839,9 @@ _parent_focus(Evas_Object *obj)
> > > >>> >> _elm_widget_focus_region_show(obj);
> > > >>> >> }
> > > >>> >> sd->focus_order_on_calc = EINA_FALSE;
> > > >>> >> +
> > > >>> >> + if (_elm_config->access_mode == ELM_ACCESS_MODE_ON)
> > > >>> >> + _elm_access_highlight_set(obj);
> > > >>> >> }
> > > >>> >>
> > > >>> >> static void
> > > >>> >> Index: src/lib/elm_layout.h
> > > >>> >>
> > ===================================================================
> > > >>> >> --- src/lib/elm_layout.h (리비전 71695)
> > > >>> >> +++ src/lib/elm_layout.h (작업 사본)
> > > >>> >> @@ -646,6 +646,22 @@ EAPI Eina_Bool
> > > >>> elm_layout_part_
> > > >>> >> EAPI Eina_Bool
> > > >>> elm_layout_part_cursor_engine_only_get(const Evas_Object *obj,
> > const char
> > > >>> *part_name);
> > > >>> >>
> > > >>> >> /**
> > > >>> >> + *
> > > >>> >> + * Set accessibility to all texblock(text) parts in the layout
> > object
> > > >>> >> + *
> > > >>> >> + * @param obj The layout object.
> > > >>> >> + * @param can_access makes all textblock(text) parts in the
> > layout
> > > >>> @p obj possible
> > > >>> >> + * to have accessibility. @c EINA_TRUE means textblock(text)
> > parts
> > > >>> can be accessible
> > > >>> >> + *
> > > >>> >> + * @return EINA_TRUE on success or EINA_FALSE on failure. If @p
> > obj
> > > >>> is not
> > > >>> >> + * a proper layout object, @c EINA_FALSE is returned.
> > > >>> >> + *
> > > >>> >> + * @ingroup Layout
> > > >>> >> + */
> > > >>> >> +EAPI Eina_Bool
> > > >>> >> +elm_layout_edje_object_access_set(Evas_Object *obj, Eina_Bool
> > > >>> can_access);
> > > >>> >> +
> > > >>> >> +/**
> > > >>> >> * Set the layout content.
> > > >>> >> *
> > > >>> >> * @param obj The layout object
> > > >>> >
> > > >>> > Do we really want to be selective on each layout should get
> > > >>> > accessibility support on it's text??
> > > >>>
> > > >>> No, not really want to..
> > > >>> All layouts should be support the accessibility that makes sense.
> > > >>>
> > > >>>
> > > >>> >
> > > >>> >> Index: src/lib/elm_layout.c
> > > >>> >>
> > ===================================================================
> > > >>> >> --- src/lib/elm_layout.c (리비전 71695)
> > > >>> >> +++ src/lib/elm_layout.c (작업 사본)
> > > >>> >> @@ -655,6 +655,11 @@ _elm_layout_smart_text_set(Evas_Object *obj,
> > > >>> >>
> > > >>> >> ELM_LAYOUT_CLASS(ELM_WIDGET_DATA(sd)->api)->sizing_eval(obj);
> > > >>> >>
> > > >>> >> + if (_elm_config->access_mode == ELM_ACCESS_MODE_ON &&
> > > >>> >> + ELM_WIDGET_DATA(sd)->can_access)
> > > >>> >> + sub_d->obj = _elm_access_edje_object_part_object_register
> > > >>> >> + (obj,
> > edje_object_part_object_get(elm_layout_edje_get(obj),
> > > >>> part), text);
> > > >>> >> +
> > > >>> >
> > > >>> > It seems we can't list an Edje object's parts, so that we apply to
> > > >>> > each textual ones the accessibility thing on the layout creation,
> > not
> > > >>> > just the text setting of it :/.
> > > >>> >
> > > >>>
> > > >>> yeap, you right. the creation time should be considered.
> > > >>>
> > > >>>
> > > >>> > Don't forget to delete a previously set access_object on
> > sub_d->obj,
> > > >>> > here.
> > > >>> >
> > > >>>
> > > >>> good point, the revised patch will take care of this.
> > > >>>
> > > >>>
> > > >>> >> return EINA_TRUE;
> > > >>> >> }
> > > >>> >>
> > > >>> >> @@ -1772,6 +1777,16 @@
> > > >>> _elm_layout_widget_smart_set_user(Elm_Layout_Smart
> > > >>> >> sc->text_aliases = _text_aliases;
> > > >>> >> }
> > > >>> >>
> > > >>> >> +EAPI Eina_Bool
> > > >>> >> +elm_layout_edje_object_access_set(Evas_Object *obj, Eina_Bool
> > > >>> can_access)
> > > >>> >> +{
> > > >>> >> + ELM_LAYOUT_CHECK(obj) EINA_FALSE;
> > > >>> >> + ELM_LAYOUT_DATA_GET_OR_RETURN_VAL(obj, sd, EINA_FALSE);
> > > >>> >> +
> > > >>> >> + ELM_WIDGET_DATA(sd)->can_access = !!can_access;
> > > >>> >> + return EINA_TRUE;
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> /* And now the basic layout widget itself */
> > > >>> >> EAPI Evas_Object *
> > > >>> >> elm_layout_add(Evas_Object *parent)
> > > >>> >> Index: src/lib/elm_access.c
> > > >>> >>
> > ===================================================================
> > > >>> >> --- src/lib/elm_access.c (리비전 71695)
> > > >>> >> +++ src/lib/elm_access.c (작업 사본)
> > > >>> >> @@ -1,6 +1,45 @@
> > > >>> >> #include <Elementary.h>
> > > >>> >> #include "elm_priv.h"
> > > >>> >> +//#include "elm_widget_layout.h"
> > > >>> >
> > > >>> > Again, none of the EAPI on this file should be here: all internal
> > > >>> > calls, here.
> > > >>> >
> > > >>>
> > > >>> Yeap.
> > > >>>
> > > >>>
> > > >>> >>
> > > >>> >> +static const char ACCESS_SMART_NAME[] = "elm_access";
> > > >>> >> +
> > > >>> >> +EVAS_SMART_SUBCLASS_NEW
> > > >>> >> + (ACCESS_SMART_NAME, _elm_access, Elm_Widget_Smart_Class,
> > > >>> >> + Elm_Widget_Smart_Class, elm_widget_smart_class_get, NULL);
> > > >>> >> +
> > > >>> >> +static Evas_Object * _elm_access_add(Evas_Object *parent);
> > > >>> >> +
> > > >>> >> +static void
> > > >>> >> +_elm_access_smart_add(Evas_Object *obj)
> > > >>> >> +{
> > > >>> >> + EVAS_SMART_DATA_ALLOC(obj, Elm_Widget_Smart_Data);
> > > >>> >> + ELM_WIDGET_CLASS(_elm_access_parent_sc)->base.add(obj);
> > > >>> >> +
> > > >>> >> + elm_widget_can_focus_set(obj, EINA_TRUE);
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> +static Eina_Bool
> > > >>> >> +_elm_access_smart_on_focus(Evas_Object *obj)
> > > >>> >> +{
> > > >>> >> + evas_object_focus_set(obj, elm_widget_focus_get(obj));
> > > >>> >> +
> > > >>> >> + return EINA_TRUE;
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> +static void
> > > >>> >> +_elm_access_smart_set_user(Elm_Widget_Smart_Class *sc)
> > > >>> >> +{
> > > >>> >> + sc->base.add = _elm_access_smart_add;
> > > >>> >> +
> > > >>> >> + /* not a 'focus chain manager' */
> > > >>> >> + sc->focus_next = NULL;
> > > >>> >> + sc->focus_direction = NULL;
> > > >>> >> + sc->on_focus = _elm_access_smart_on_focus;
> > > >>> >> +
> > > >>> >> + return;
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> typedef struct _Mod_Api Mod_Api;
> > > >>> >>
> > > >>> >> struct _Mod_Api
> > > >>> >> @@ -194,9 +233,12 @@ _access_obj_hilight_resize_cb(void *data
> > > >>> __UNUSED_
> > > >>> >> evas_object_resize(o, w, h);
> > > >>> >> }
> > > >>> >>
> > > >>> >> -
> > > >>> >> -
> > > >>> >>
> > > >>>
> > //-------------------------------------------------------------------------//
> > > >>> >> +EAPI void
> > > >>> >> +_elm_access_highlight_set(Evas_Object* obj)
> > > >>> >> +{
> > > >>> >> + _access_obj_over_timeout_cb(obj);
> > > >>> >> +}
> > > >>> >>
> > > >>> >> EAPI void
> > > >>> >> _elm_access_clear(Elm_Access_Info *ac)
> > > >>> >> @@ -383,6 +425,72 @@ _elm_access_object_unhilight(Evas_Object
> > *obj)
> > > >>> >> }
> > > >>> >> }
> > > >>> >>
> > > >>> >> +static void
> > > >>> >> +_content_resize(void *data, Evas *e __UNUSED__, Evas_Object *obj,
> > > >>> >> + void *event_info __UNUSED__)
> > > >>> >> +{
> > > >>> >> + Evas_Object *accessobj;
> > > >>> >> + Evas_Coord w, h;
> > > >>> >> +
> > > >>> >> + accessobj = data;
> > > >>> >> + if (!accessobj) return;
> > > >>> >> +
> > > >>> >> + evas_object_geometry_get(obj, NULL, NULL, &w, &h);
> > > >>> >> + evas_object_resize(accessobj, w, h);
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> +static void
> > > >>> >> +_content_move(void *data, Evas *e __UNUSED__, Evas_Object *obj,
> > > >>> >> + void *event_info __UNUSED__)
> > > >>> >> +{
> > > >>> >> + Evas_Object *accessobj;
> > > >>> >> + Evas_Coord x, y;
> > > >>> >> +
> > > >>> >> + accessobj = data;
> > > >>> >> + if (!accessobj) return;
> > > >>> >> +
> > > >>> >> + evas_object_geometry_get(obj, &x, &y, NULL, NULL);
> > > >>> >> + evas_object_move(accessobj, x, y);
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> +static char *
> > > >>> >> +_access_info_cb(void *data __UNUSED__, Evas_Object *obj,
> > > >>> >> + Elm_Widget_Item *item __UNUSED__)
> > > >>> >> +{
> > > >>> >> + const char *txt = elm_widget_access_info_get(obj);
> > > >>> >> + if (txt) return strdup(txt);
> > > >>> >> + return NULL;
> > > >>> >> +}
> > > >>> >> +
> > > >>> >> +EAPI Evas_Object *
> > > >>> >> +_elm_access_edje_object_part_object_register(Evas_Object* obj,
> > > >>> >> + const Evas_Object
> > > >>> *partobj,
> > > >>> >> + const char* text)
> > > >>> >
> > > >>> > What about _elm_access_text_register(Evas_Object *parent,
> > Evas_Object
> > > >>> > *text_obj), and we fetch the text string from the object
> > > >>> > internally. The item version of it I dislike, as it just highlights
> > > >>> > things on mouse in, with apparently no text attached.
> > > >>> >
> > > >>>
> > > >>> This _elm_access_text_set() is for the ELM_ACCESS_TYPE. Accessibility
> > > >>> works as following.
> > > >>> If there is a button which has label "OK" and the button is disabled,
> > > >>> then accessibility says...
> > > >>> "Button, OK, Disabled" (ordered by ELM_ACCESS_TYPE, ELM_ACCESS_INFO,
> > > >>> ELM_ACCESS_STATE)
> > > >>>
> > > >>>
> > > >>> >> +{
> > > >>> >> + Evas_Object *ao;
> > > >>> >> + Evas_Object *po = (Evas_Object *)partobj;
> > > >>> >> +
> > > >>> >> + if (!obj || !po) return NULL;
> > > >>> >> +
> > > >>> >> + // create access object
> > > >>> >> + ao = _elm_access_add(obj);
> > > >>> >> + evas_object_event_callback_add(po, EVAS_CALLBACK_RESIZE,
> > > >>> >> + _content_resize, ao);
> > > >>> >> + evas_object_event_callback_add(po, EVAS_CALLBACK_MOVE,
> > > >>> >> + _content_move, ao);
> > > >>> >> + evas_object_show(ao);
> > > >>> >> +
> > > >>> >> + // register access object
> > > >>> >> + _elm_access_object_register(ao, po);
> > > >>> >> + _elm_access_text_set(_elm_access_object_get(ao),
> > > >>> >> + ELM_ACCESS_TYPE,
> > evas_object_type_get(po));
> > > >>> >> + _elm_access_callback_set(_elm_access_object_get(ao),
> > > >>> >> + ELM_ACCESS_INFO,
> > > >>> >> + _access_info_cb, NULL);
> > > >>> >> + elm_widget_access_info_set(ao, text);
> > > >>> >> +
> > > >>> >> + return ao;
> > > >>> >> +}
> > > >>> >> EAPI void
> > > >>> >> _elm_access_object_hilight_disable(Evas *e)
> > > >>> >> {
> > > >>> >> @@ -534,3 +642,22 @@ _elm_access_2nd_click_timeout(Evas_Object
> > *obj)
> > > >>> >> _access_2nd_click_del_cb,
> > NULL);
> > > >>> >> return EINA_FALSE;
> > > >>> >> }
> > > >>> >> +
> > > >>> >> +static Evas_Object *
> > > >>> >> +_elm_access_add(Evas_Object *parent)
> > > >>> >> +{
> > > >>> >> + Evas *e;
> > > >>> >> + Evas_Object *obj;
> > > >>> >> +
> > > >>> >> + EINA_SAFETY_ON_NULL_RETURN_VAL(parent, NULL);
> > > >>> >> +
> > > >>> >> + e = evas_object_evas_get(parent);
> > > >>> >> + if (!e) return NULL;
> > > >>> >> +
> > > >>> >> + obj = evas_object_smart_add(e, _elm_access_smart_class_new());
> > > >>> >> +
> > > >>> >> + if (!elm_widget_sub_object_add(parent, obj))
> > > >>> >> + ERR("could not add %p as sub object of %p", obj, parent);
> > > >>> >> +
> > > >>> >> + return obj;
> > > >>> >> +}
> > > >>> >
> > > >>> > BR,
> > > >>> >
> > > >>> >
> > > >>> >>
> > > >>>
> > ------------------------------------------------------------------------------
> > > >>> >> Live Security Virtual Conference
> > > >>> >> Exclusive live event will cover all the ways today's security and
> > > >>> >> threat landscape has changed and how IT managers can respond.
> > > >>> Discussions
> > > >>> >> will include endpoint security, mobile security and the latest in
> > > >>> malware
> > > >>> >> threats.
> > http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> > > >>> >
> > > >>> >> _______________________________________________
> > > >>> >> enlightenment-devel mailing list
> > > >>> >> enl...@li...
> > > >>> >> https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> > > >>> >
> > > >>> >
> > > >>> > --
> > > >>> > Gustavo Lima Chaves
> > > >>> > Computer Engineer @ ProFUSION Embedded Systems
> > > >>> >
> > > >>> >
> > > >>>
> > ------------------------------------------------------------------------------
> > > >>> > Live Security Virtual Conference
> > > >>> > Exclusive live event will cover all the ways today's security and
> > > >>> > threat landscape has changed and how IT managers can respond.
> > > >>> Discussions
> > > >>> > will include endpoint security, mobile security and the latest in
> > > >>> malware
> > > >>> > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> > > >>> > _______________________________________________
> > > >>> > enlightenment-devel mailing list
> > > >>> > enl...@li...
> > > >>> > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> > > >>>
> > > >>>
> > > >>> Sincerely,
> > > >>> Shinwoo Kim.
> > > >>>
> > > >>>
> > > >>
> > > >
> >
> >
> > --
> > ------------- Codito, ergo sum - "I code, therefore I am" --------------
> > The Rasterman (Carsten Haitzler) ra...@ra...
> >
> >
--
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler) ra...@ra...
|