* fix(style): correct the comment and alignment
* fix(style): remove the unused macro and inclusion
* fix(style): correct _lv_style_get_xxx with the right pointer type
* fix(style): handle the out of memory gracefully
* fix(style): unify the term usage(class->style)
* fix(style): remove the unnecessary mask in _lv_style_get_xxx
since get_property_index ensure the returned state is a subset of required
* fix(style): centralize the style validation in one place
* fix(style): reuse the new_prop_size in _lv_style_set_xxx
* fix(style): correct the return value of get_style_prop in big endian machine
* fix(image decoder): replace lv_style.h with lv_color.h
since image decoder doesn't depends on style subsystem
and correct the related comment and some error handling
* fix(image decoder): fix memory leak when no decoder can be used
and handle the out of memory gracefully
* fix(image decoder): zero the output fields after each iteration
to clean the stale info saved by try failed decoder and remove img_data
zero from lv_img_decoder_built_in_open since it has be done in the loop
* fix(image decoder): check .bin suffix in lv_img_decoder_built_in_info
align with what has done in lv_img_decoder_built_in_open
* fix(image decoder): embed lv_fs_file_t directly to save the memory
and handle the abnormal case gracefully(avoid crash or leak)
* fix(image decoder): simplify decode builtin alpha/index bitmap
* fix(anim): relayout lv_anim_t to save 4 bytes
and fix the minor comment issue
* fix(anim): unify the type of time and animation to int32_t or uint32_t
since lv_anim_t use these types internally
* fix(anim): Don't need initialize last_task_run
since lv_anim_start will do this if the list is empty
and remove -1 to improve the accuracy
* fix(anim): remove the return value from anim_ready_handler
since no caller really use the return value
* fix(math): _lv_map check x >= max_in first
_lv_map return max_out when min_in equals max_in,
since the animation module require this behaviour
* fix(anim): normalize the time elapse by lv_map
* fix(anim): avoid iterate the list twice in anim_task
* fix(theme): the minor typo error
* fix(style): make lv_style_reset work with zerod memory
* fix(theme): fix the memory leak when lv_theme_xxx_init is called twice
due to inited variable forget to set to 1
* fix(theme): empty theme should call lv_obj_refresh_style too
and remove lv_obj_clean_style_list since the same thing is
already done in the common code(clear_styles)
