def build(self, shape):
dense = L.Dense(units=self.units)
dense.build(shape)
k_shape = dense.kernel.shape
b_shape = dense.bias.shape
k_size = get_size(k_shape)
b_size = get_size(b_shape)
self.hyper_layer = N.Layer(self.ai, units=k_size + b_size)
self.reshape_kernel = L.Reshape(k_shape)
self.reshape_bias = L.Reshape(b_shape)
self.split = L.Lambda(lambda x: tf.split(x, [k_size, b_size]))
super().build(shape)
def auto_index(elastic, in_file_path, out_path):
with open(in_file_path, "r") as ini:
best_index = ""
best_index_bytes = "a".encode()
biggest_size = 0
for line in ini:
if not line.startswith("{\"index") and not line.startswith("\n"):
json_line = json.loads(line)
# print(json_line)
auto_index_elastic.bulk_data(elastic, json_line)
auto_index = auto_index_elastic.get_mapping(elastic)
auto_index_bytes = json.dumps(auto_index).encode()
auto_index_elastic.clear_test_mapping(elastic)
if auto_index_bytes != best_index_bytes:
size_obj = tools.get_size(auto_index)
if size_obj.exact_size > biggest_size:
best_index = auto_index
biggest_size = size_obj.exact_size
best_index_bytes = auto_index_bytes
print(best_index)
print(size_obj.size, size_obj.unit)
with open(out_path, "w") as out:
json.dump(best_index["tmp_index"], out)
return best_index["tmp_index"]
def build(self, shape):
dense = L.Dense(self.units)
dense.build([shape[0], get_size(shape[1:])])
self.kernel = self.add_weight(
"kernel", dense.kernel.shape, initializer=self.init,
trainable=self.trainable)
self.flatten = L.Flatten()
super().build(shape)
def __init__(self, AI, out_shape, key=KEY):
super(Resizer, self).__init__()
size = get_size(out_shape)
self.resize = nature.FC(units=size)
self.reshape = L.Reshape(out_shape)
self.fn = nature.Fn(AI, key=key)
self.out_shape = out_shape
self.flatten = L.Flatten()
self.built = True
def createPathUi():
top = Toplevel()
top.title('图片目录选择')
top.geometry(tools.get_size(width=240, height=100))
top.resizable(0, 0)
pathVar = StringVar()
lable_path = Label(top, text='路径选择：')
entry_path = Entry(top, textvariable=pathVar)
button = Button(top,
text="选择目录",
width=19,
command=lambda: selectPath(pathVar, top))
# 放置部件
lable_path.grid(row=0, column=0)
entry_path.grid(row=0, column=1)
button.grid(row=1, column=1)
def init_Window(self):
# GUI标题
self.master.title("OCR批量识别身份证")
# 窗口大小
self.master.geometry(tools.get_size(width=300, height=400))
self.master.resizable(0, 0)
MenuBar = Menu(self.master)
MenuBar.add_cascade(label="开始运行", command=lambda: ocr.main(text))
MenuBar.add_cascade(label="百度API设置", command=MenuBarUi.createAPiUi)
MenuBar.add_cascade(label="选择图片目录", command=MenuBarUi.createPathUi)
self.master.config(menu=MenuBar)
scroll = Scrollbar()
text = Text(self.master, width=40, height=400)
scroll.pack(side=RIGHT, fill=Y)
text.pack(side=RIGHT, fill=Y)
scroll.config(command=text.yview)
text.config(yscrollcommand=scroll.set)
def createAPiUi():
top = Toplevel()
top.title('API设置')
top.geometry(tools.get_size(width=240, height=100))
top.resizable(0, 0)
akVar = StringVar()
skVar = StringVar()
lable_ak = Label(top, text='Api Key：')
lable_sk = Label(top, text='Secret Key：')
entry_ak = Entry(top, textvariable=akVar)
entry_sk = Entry(top, textvariable=skVar)
button = Button(top,
text="确定",
width=19,
command=lambda: submit(akVar=akVar, skVar=skVar, top=top))
# 放置部件
lable_ak.grid(row=0, column=0)
lable_sk.grid(row=1, column=0)
entry_ak.grid(row=0, column=1)
entry_sk.grid(row=1, column=1)
button.grid(row=2, column=1)
def __init__(self, agent, task_id, in_spec, out_spec, in_number=None):
super(Interface, self).__init__()
self.pull_numbers = agent.pull_numbers
self.pull_choices = agent.pull_choices
self.agent = agent
self.probabilistic = agent.probabilistic
self.task_id, self.in_spec, self.out_spec = task_id, in_spec, out_spec
self.brick_id = f"{task_id}{'_'+str(in_number) if in_number else ''}_{in_spec.format}_to_{out_spec.format}_{generate()}"
self.debug = 1 if "predictor" in self.brick_id else 0
log("", debug=self.debug)
log(f"Interface.__init__ -- {self.brick_id}", debug=self.debug)
log("in_spec", in_spec, debug=self.debug)
log("out_spec", out_spec, debug=self.debug)
self.input_layer_number = in_number
self.shape_variable_key = None
if self.in_spec.format == "image":
self.channels_before_concat_coords = self.in_spec.shape[-1]
self.size_to_resize_to = get_hw(self.in_spec.shape)
encoder_shape = self.add_coords_to_shape(self.in_spec.shape)
self.in_spec = get_spec(format="image", shape=encoder_shape)
self.in_spec.size = get_size(encoder_shape)
super(Interface, self).__init__()
self.build()
import sys
from trie import Trie
from bloom_filter import BloomFilter
from tools import get_size
if __name__ == '__main__':
bf_dups = 0
tr = Trie()
bf = BloomFilter(capacity=700000, error_rate=0.001)
with open("words.txt") as file:
for line in file:
tr.put(line.strip())
if bf.put(line.strip()):
print("Duplicate in bloom filter: {0}".format(line.strip()))
bf_dups += 1
print("Trie. number of objects put: {0}".format(len(tr)))
print("Bloom filter. number of objects put: {0}".format(len(bf)))
print()
print("Trie. Size of the object: {0}".format(sys.getsizeof(tr)))
print("Bloom filter. Size of the object: {0}".format(sys.getsizeof(bf)))
print()
print("Trie. Size of the object(full): {0}".format(get_size(tr)))
print("Bloom filter. Size of the object(full): {0}".format(get_size(bf)))
print()
print("Bloom filter errors: {0}".format(bf_dups))
print("----------------------------------------------------------")
def get_units(shape):
d_in = get_size(shape)
return d_in * d_in + 1