def get(self, field="image"):
# type: (Text) -> Tuple[ImageInfo, ...]
"""Get imageinfo used on the field.
Args:
field (six.text_type): Defaults to 'image', Server defined field name,
Returns:
tuple[ImageInfo]: Image information.
"""
select = self.select
ret = set()
rows = cast.list_(select[field], (str, six.text_type))
for row in rows:
try:
for data in cast.list_(json.loads(row), dict):
info = ImageInfo(
max=cast.list_(data["max"], (six.text_type, str))[0],
min=cast.list_(data["min"], (six.text_type, str))[0],
path=data.get("path"),
attachment_id=data.get("att_id"),
)
ret.add(info)
except (TypeError, KeyError, ValueError):
continue
ret = tuple(sorted(ret))
return ret
def test_type_not_match():
with pytest.raises(
cast.CastError,
match=
"(u?'can not cast object to instance', None, <(class|type) 'int'>)"
):
cast.list_([1, None, 3], int)
def from_training_scene(cls, img: Image) -> Training:
if g.image_path:
image_id = imagetools.md5(
imagetools.cv_image(img.convert("RGB")),
save_path=g.image_path,
save_mode="RGB",
)
LOGGER.debug("from_training_scene: image=%s", image_id)
rp = mathtools.ResizeProxy(img.width)
self = cls.new()
self.confirm_position = next(
template.match(
img,
template.Specification(templates.SINGLE_MODE_TRAINING_CONFIRM,
threshold=0.8),
))[1]
radius = rp.vector(30, 540)
for t, center in zip(
Training.ALL_TYPES,
(
rp.vector2((78, 850), 540),
rp.vector2((171, 850), 540),
rp.vector2((268, 850), 540),
rp.vector2((367, 850), 540),
rp.vector2((461, 850), 540),
),
):
if mathtools.distance(self.confirm_position, center) < radius:
self.type = t
break
else:
raise ValueError("unknown type for confirm position: %s" %
self.confirm_position)
self.level = _recognize_level(
tuple(cast.list_(img.getpixel(rp.vector2((10, 200), 540)), int)))
t, b = 503, 532
self.speed = _ocr_training_effect(
img.crop(rp.vector4((18, t, 91, b), 466)))
self.stamina = _ocr_training_effect(
img.crop(rp.vector4((91, t, 163, b), 466)))
self.power = _ocr_training_effect(
img.crop(rp.vector4((163, t, 237, b), 466)))
self.guts = _ocr_training_effect(
img.crop(rp.vector4((237, t, 309, b), 466)))
self.wisdom = _ocr_training_effect(
img.crop(rp.vector4((309, t, 382, b), 466)))
self.skill = _ocr_training_effect(
img.crop(rp.vector4((387, t, 450, b), 466)))
return self
def _find_by_race_menu_item(ctx: Context, img: PIL.Image.Image) -> Race:
rp = mathtools.ResizeProxy(img.width)
spec_bbox = rp.vector4((221, 12, 478, 32), 492)
no1_fan_count_bbox = rp.vector4((207, 54, 360, 72), 492)
grade_color_pos = rp.vector2((182, 14), 492)
stadium, ground, distance, turn, track = _recognize_spec(img.crop(spec_bbox))
no1_fan_count = _recognize_fan_count(img.crop(no1_fan_count_bbox))
grades = _recognize_grade(tuple(cast.list_(img.getpixel(grade_color_pos), int)))
full_spec = (
ctx.date,
stadium,
ground,
distance,
turn,
track,
no1_fan_count,
grades,
)
for i in _find_by_spec(*full_spec):
LOGGER.info("image match: %s", i)
return i
raise ValueError("_find_by_race_menu_item: no race match spec: %s", full_spec)
def find_by_race_detail_image(ctx: Context, screenshot: PIL.Image.Image) -> Race:
rp = mathtools.ResizeProxy(screenshot.width)
grade_color_pos = rp.vector2((10, 75), 466)
spec_bbox = rp.vector4((27, 260, 302, 279), 466)
_, no1_fan_count_pos = next(
template.match(screenshot, templates.SINGLE_MODE_RACE_DETAIL_NO1_FAN_COUNT)
)
no1_fan_count_bbox = (
rp.vector(150, 466),
no1_fan_count_pos[1],
rp.vector(400, 466),
no1_fan_count_pos[1] + rp.vector(18, 466),
)
grades = _recognize_grade(
tuple(cast.list_(screenshot.getpixel(grade_color_pos), int))
)
stadium, ground, distance, turn, track = _recognize_spec(screenshot.crop(spec_bbox))
no1_fan_count = _recognize_fan_count(screenshot.crop(no1_fan_count_bbox))
full_spec = (
ctx.date,
stadium,
ground,
distance,
turn,
track,
no1_fan_count,
grades,
)
for i in _find_by_spec(*full_spec):
LOGGER.info("image match: %s", i)
return i
raise ValueError("find_by_race_details_image: no race match spec: %s", full_spec)
def test_iterator():
assert cast.list_(six.moves.range(3)) == [0, 1, 2]
def test_text_with_type():
assert cast.list_("abc", six.text_type) == ["abc"]
def test_text():
assert cast.list_("abc") == ["a", "b", "c"]
def test_none():
assert cast.list_(None) == []
def _is_race_list_scroll_to_top() -> bool:
rp = action.resize_proxy()
color = template.screenshot(max_age=0).getpixel(rp.vector2((525, 525),
540))
return (imagetools.compare_color(
(123, 121, 140), tuple(cast.list_(color, int))) > 0.9)
