def handle_starttag(self, tag, attrs):
if tag == "a":
if len(attrs) == 0:
pass
else:
for (variable, value) in attrs:
if variable == "href":
self.links.append(value)
if tag == 'img':
for name,value in attrs:
if name == 'src':
util.get_img(self.urlString.rsplit('/',1)[0] + "/" + value)
def handle_starttag(self, tag, attrs):
if tag == "a":
if len(attrs) == 0:
pass
else:
for (variable, value) in attrs:
if variable == "href":
self.links.append(value)
if tag == 'img':
for name, value in attrs:
if name == 'src':
util.get_img(
self.urlString.rsplit('/', 1)[0] + "/" + value)
def test_generator(self):
test_len = self.test_len()
index = 0
while True:
data = self.data[
index * self.batch_size + self.train_size: (index + 1) * self.batch_size + self.train_size]
index += 1
index %= test_len
words = []
images = []
nums = []
for i in range(self.batch_size):
img_path = data[i][0][0]
word = data[i][1][0]
nums.append(len(word))
word = self.__word_process(word)
img = get_img("{}/{}".format(self.path, img_path), self.width, self.height)
images.append(img)
words.append(word)
words = pad_sequences(words, maxlen=self.max_string_length, truncating="post", padding="post")
label_input = np.array(nums).reshape([-1, 1])
images = np.array(images)
# images = np.transpose(images, [0, 1, 2])
yield {"img": images[:, :, :, np.newaxis], "label_input": label_input,
"y_input": words}, np.zeros([self.batch_size, 1])
def predict(base_model, file_):
img = get_img(file_, width, height)
y_pred = base_model.predict(img[np.newaxis, :, :, np.newaxis])
print(file_.center(50, "*"))
print(y_pred)
str_out = ''.join([id2char[x] for x in y_pred[0] if x != -1])
print(str_out)
plt.imshow(img, cmap="gray")
plt.show()
def get_ratio(path: str):
img = get_img(path)
width, height = img.size
return {
"width": width,
"height": height,
"value": width / height,
"aspect": "landscape" if width > height else "portrait"
}
def __getitem__(self, index):
data = self.data[index * self.batch_size: (index + 1) * self.batch_size]
words = []
images = []
nums = []
for i in range(self.batch_size):
img_path = data[i][0][0]
word = data[i][1][0]
word = self.__word_process(word)
img = get_img("{}/{}".format(self.path, img_path), self.width, self.height)
img = random_brightness(img[:, :, np.newaxis], [0.1, 1.5])
images.append(img[:, :, 0])
words.append(word)
nums.append(len(word))
words = pad_sequences(words, maxlen=self.max_string_length, truncating="post", padding="post")
label_input = np.array(nums).reshape([-1, 1])
images = np.array(images)
return {"img": images[:, :, :, np.newaxis], "label_input": label_input,
"y_input": words}, np.zeros([self.batch_size, 1])
def get_colors(path: str, nb_colors=4):
imgfile = get_img(path)
numarray = numpy.array(imgfile.getdata(), numpy.uint8)
clusters = KMeans(n_clusters=nb_colors)
clusters.fit(numarray)
npbins = numpy.arange(0, nb_colors + 1)
histogram = numpy.histogram(clusters.labels_, bins=npbins)
indexes = []
to_sort = list(histogram[0])
for i in range(nb_colors):
vMax = max(to_sort)
indexes.append({'was': i, 'to': list(histogram[0]).index(vMax)})
to_sort.remove(vMax)
# indexes.sort(key=lambda xxx: xxx["to"])
# HISTOGRAM
# values = sorted(histogram[0], reverse=True)
colors = []
for i in range(nb_colors):
centerIndex = next(x for x in indexes if x["to"] == i)["was"]
center = clusters.cluster_centers_[centerIndex]
color = '#%02x%02x%02x' % (math.ceil(center[0]), math.ceil(
center[1]), math.ceil(center[2]))
colors.append(color)
# Liste de couleurs
def convert_to_rgb(color: str):
(r, g, b) = ImageColor.getrgb(color)
return {"r": r, "g": g, "b": b}
return list(map(convert_to_rgb, colors))
from bs4 import BeautifulSoup
import util
import re
import shutil
import os
import requests
infi = 'download.html'
with open(infi, 'r') as f:
bsobj = BeautifulSoup(f.read(), 'html.parser')
del bsobj.span['class']
del bsobj.span['itemprop']
ncts = bsobj.find_all('noscript')
for n in ncts:
n.decompose()
imgs = bsobj.find_all('img')
for i in imgs:
#如果img包含src属性
if 'data-actualsrc' in i.attrs:
img_url = i['data-actualsrc']
for a in list(i.attrs):
del i[a]
i['src'] = util.get_img(img_url, 'images')
#保存新生成的html
util.save_html('test.html', str(bsobj))