Skip to content

京
京东搜索采集自动验证
Commit 4060af6c authored by huangziyu's avatar huangziyu
Browse files
Options

1

parent 446cd288
Hide whitespace changes
Inline Side-by-side
Showing with 66 additions and 56 deletions
京东搜索_自动版.py
View file @ 4060af6c
......@@ -31,6 +31,7 @@ mysql_zdzs_proxy_user = 'zgcindex'
mysql_zdzs_proxy_password = 'zgcprice2019'
mysql_zdzs_proxy_database = 'zdzs_proxy'
import config
username = config.username
# 登陆密码
password = config.password
......@@ -188,14 +189,19 @@ def detect_circular_distortion(image_path, circle_radius=65):
return average_distance
def any_of_elements_present(*locators):
"""检查给定的任意一个元素是否出现在DOM中"""
def _predicate(driver):
for locator in locators:
if EC.presence_of_element_located(locator)(driver):
return True
return False
return _predicate
def find_best_match_angle(background_img_path, rotating_img_path):
# 加载背景图片和验证码图片
background = Image.open(background_img_path)
......@@ -340,8 +346,8 @@ def get_cookie():
element = driver.find_element_by_class_name("JDJRV-slide-btn")
# 获取元素的X和Y位置
x_position = element.location['x']+50
y_position = element.location['y']+100
x_position = element.location['x'] + 50
y_position = element.location['y'] + 100
print("元素的X位置：", x_position)
print("元素的Y位置：", y_position)
......@@ -393,7 +399,8 @@ def get_cookie():
print("网页加载完成")
text = driver.page_source
while_count = 0
while ('快速验证' in text and '验证一下，购物无忧' in text) or '前方拥挤，请刷新重试' in text or 'class="title">加载中...</span></div></div><!----></div>' in text:
while (
'快速验证' in text and '验证一下，购物无忧' in text) or '前方拥挤，请刷新重试' in text or 'class="title">加载中...</span></div></div><!----></div>' in text:
while_count += 1
if while_count > 1:
driver.get(url)
......@@ -439,8 +446,8 @@ def get_cookie():
element = driver.find_element_by_class_name("bg-blue")
# 获取元素的X和Y位置
x_position = element.location['x']+50
y_position = element.location['y']+100
x_position = element.location['x'] + 50
y_position = element.location['y'] + 100
print("元素的X位置：", x_position)
print("元素的Y位置：", y_position)
......@@ -473,62 +480,65 @@ def get_cookie():
except NoSuchElementException:
# print("未查找到验证图片背景图，可能是旋转图片，换一种方式获取div")
try:
background = driver.find_element(by=By.XPATH, value='//*[@id="img-back-div"]')
background_url = background.get_attribute("style").replace('"); height: 181px;', "").replace(
'background-image: url("', "")
# 用来找到登录图片的小滑块
slide = driver.find_element(by=By.XPATH, value='//*[@id="img-rotate-div"]/img')
slide_url = slide.get_attribute("src")
background_img = 'background_img.png'
slide_img = 'slide_img.png'
# 下载背景大图保存到本地
request.urlretrieve(background_url, background_img)
# 下载滑块保存到本地
request.urlretrieve(slide_url, slide_img)
best_angle, best_match_score = find_best_match_angle(background_img, slide_img)
print(f"最佳匹配角度为: {best_angle} 度")
# time.sleep(0.5)
# w1 = cv2.imread(background_img).shape[1]
# # 获取网页背景图宽度
# w2 = background.size['width']
x = 228 / 360 * best_angle
x = int(x)
print(f"移动: {x} ")
# 找到要获取位置的元素，比如通过其ID
element = driver.find_element_by_id("slider-div")
# 获取元素的X和Y位置
x_position = element.location['x']+50
y_position = element.location['y']+100
print("元素的X位置：", x_position)
print("元素的Y位置：", y_position)
# 其中x为屏幕左上角至滑块中心的横向像素值，y为屏幕左上角至滑块中心纵向像素值, 可根据自己屏幕配置
# time.sleep(0.5)
SlideUtils.slide_by_pyautogui2(x_position,
y_position,
x+2,
offset_increase)
try:
WebDriverWait(driver, 2).until(
EC.presence_of_element_located((By.ID, "J_main"))
)
text = driver.page_source
continue
except:
while_count_2 = 0
while while_count_2 < 5:
while_count_2 += 1
background = driver.find_element(by=By.XPATH, value='//*[@id="img-back-div"]')
background_url = background.get_attribute("style").replace('"); height: 181px;', "").replace(
'background-image: url("', "")
# 用来找到登录图片的小滑块
slide = driver.find_element(by=By.XPATH, value='//*[@id="img-rotate-div"]/img')
slide_url = slide.get_attribute("src")
background_img = 'background_img.png'
slide_img = 'slide_img.png'
# 下载背景大图保存到本地
request.urlretrieve(background_url, background_img)
# 下载滑块保存到本地
request.urlretrieve(slide_url, slide_img)
best_angle, best_match_score = find_best_match_angle(background_img, slide_img)
print(f"最佳匹配角度为: {best_angle} 度")
# time.sleep(0.5)
# w1 = cv2.imread(background_img).shape[1]
# # 获取网页背景图宽度
# w2 = background.size['width']
x = 228 / 360 * best_angle
x = int(x)
print(f"移动: {x} ")
# 找到要获取位置的元素，比如通过其ID
element = driver.find_element_by_id("slider-div")
# 获取元素的X和Y位置
x_position = element.location['x'] + 50
y_position = element.location['y'] + 100
print("元素的X位置：", x_position)
print("元素的Y位置：", y_position)
# 其中x为屏幕左上角至滑块中心的横向像素值，y为屏幕左上角至滑块中心纵向像素值, 可根据自己屏幕配置
# time.sleep(0.5)
SlideUtils.slide_by_pyautogui2(x_position,
y_position,
x + 2,
offset_increase)
try:
WebDriverWait(driver, 5).until(
EC.presence_of_element_located((By.CLASS_NAME, "verifyBtn"))
WebDriverWait(driver, 2).until(
EC.presence_of_element_located((By.ID, "J_main"))
)
except NoSuchElementException:
print("网页加载异常")
text = driver.page_source
continue
break
except:
try:
WebDriverWait(driver, 5).until(
EC.presence_of_element_located((By.CLASS_NAME, "verifyBtn"))
)
except NoSuchElementException:
print("网页加载异常")
text = driver.page_source
break
finally:
print("网页加载完成")
finally:
print("网页加载完成")
finally:
print("网页加载完成")
text = driver.page_source
except NoSuchElementException:
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment