new func
This commit is contained in:
parent
35cc5d3bde
commit
b6ceeb44ac
8 changed files with 406 additions and 345 deletions
91
src/utils.py
91
src/utils.py
|
|
@ -1,23 +1,32 @@
|
|||
import logging
|
||||
import os
|
||||
import random
|
||||
import time
|
||||
|
||||
from selenium import webdriver
|
||||
|
||||
import logging
|
||||
log_file = "app_log.log"
|
||||
|
||||
logging.basicConfig(
|
||||
level=logging.DEBUG,
|
||||
format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
|
||||
handlers=[
|
||||
logging.FileHandler(log_file, mode='a', encoding='utf-8'),
|
||||
logging.StreamHandler()
|
||||
],
|
||||
force=True # This will reset the root logger's handlers and apply the new configuration
|
||||
)
|
||||
|
||||
# Настройка логирования
|
||||
logging.basicConfig(level=logging.DEBUG, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# Отключаем логирование для selenium и urllib3
|
||||
logging.getLogger("selenium.webdriver.remote.remote_connection").setLevel(logging.WARNING)
|
||||
logging.getLogger("urllib3").setLevel(logging.WARNING)
|
||||
logging.getLogger("httpcore").setLevel(logging.WARNING)
|
||||
|
||||
file_handler = logging.FileHandler(log_file, mode='a', encoding='utf-8')
|
||||
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
|
||||
file_handler.setFormatter(formatter)
|
||||
logger.addHandler(file_handler)
|
||||
logger.setLevel(logging.DEBUG)
|
||||
|
||||
chromeProfilePath = os.path.join(os.getcwd(), "chrome_profile", "linkedin_profile")
|
||||
|
||||
|
||||
def ensure_chrome_profile():
|
||||
logger.debug("Ensuring Chrome profile exists at path: %s", chromeProfilePath)
|
||||
profile_dir = os.path.dirname(chromeProfilePath)
|
||||
|
|
@ -29,51 +38,74 @@ def ensure_chrome_profile():
|
|||
logger.debug("Created Chrome profile directory: %s", chromeProfilePath)
|
||||
return chromeProfilePath
|
||||
|
||||
|
||||
def is_scrollable(element):
|
||||
scroll_height = element.get_attribute("scrollHeight")
|
||||
client_height = element.get_attribute("clientHeight")
|
||||
scrollable = int(scroll_height) > int(client_height)
|
||||
logger.debug("Element scrollable check: scrollHeight=%s, clientHeight=%s, scrollable=%s", scroll_height, client_height, scrollable)
|
||||
logger.debug("Element scrollable check: scrollHeight=%s, clientHeight=%s, scrollable=%s", scroll_height,
|
||||
client_height, scrollable)
|
||||
return scrollable
|
||||
|
||||
|
||||
def scroll_slow(driver, scrollable_element, start=0, end=3600, step=300, reverse=False):
|
||||
logger.debug("Starting slow scroll: start=%d, end=%d, step=%d, reverse=%s", start, end, step, reverse)
|
||||
|
||||
if reverse:
|
||||
start, end = end, start
|
||||
step = -step
|
||||
|
||||
if step == 0:
|
||||
logger.error("Step value cannot be zero.")
|
||||
raise ValueError("Step cannot be zero.")
|
||||
|
||||
max_scroll_height = int(scrollable_element.get_attribute("scrollHeight"))
|
||||
current_scroll_position = int(scrollable_element.get_attribute("scrollTop"))
|
||||
logger.debug("Max scroll height of the element: %d", max_scroll_height)
|
||||
logger.debug("Current scroll position: %d", current_scroll_position)
|
||||
|
||||
if end > max_scroll_height:
|
||||
logger.warning("End value exceeds the scroll height. Adjusting end to %d", max_scroll_height)
|
||||
end = max_scroll_height
|
||||
if reverse:
|
||||
|
||||
if current_scroll_position < start:
|
||||
start = current_scroll_position
|
||||
logger.debug("Adjusted start position for upward scroll: %d", start)
|
||||
else:
|
||||
|
||||
if end > max_scroll_height:
|
||||
logger.warning("End value exceeds the scroll height. Adjusting end to %d", max_scroll_height)
|
||||
end = max_scroll_height
|
||||
|
||||
script_scroll_to = "arguments[0].scrollTop = arguments[1];"
|
||||
|
||||
try:
|
||||
if scrollable_element.is_displayed():
|
||||
if not is_scrollable(scrollable_element):
|
||||
logger.warning("The element is not scrollable.")
|
||||
print("The element is not scrollable.")
|
||||
return
|
||||
|
||||
if (step > 0 and start >= end) or (step < 0 and start <= end):
|
||||
logger.warning("No scrolling will occur due to incorrect start/end values.")
|
||||
print("No scrolling will occur due to incorrect start/end values.")
|
||||
return
|
||||
for position in range(start, end, step):
|
||||
return
|
||||
|
||||
position = start
|
||||
while (step > 0 and position < end) or (step < 0 and position > end):
|
||||
try:
|
||||
driver.execute_script(script_scroll_to, scrollable_element, position)
|
||||
logger.debug("Scrolled to position: %d", position)
|
||||
except Exception as e:
|
||||
logger.error("Error during scrolling: %s", e)
|
||||
print(f"Error during scrolling: {e}")
|
||||
time.sleep(random.uniform(1.0, 1.6))
|
||||
|
||||
position += step
|
||||
step = max(10, abs(step) - 10) * (-1 if reverse else 1)
|
||||
|
||||
time.sleep(random.uniform(0.6, 1.5))
|
||||
|
||||
driver.execute_script(script_scroll_to, scrollable_element, end)
|
||||
logger.debug("Scrolled to final position: %d", end)
|
||||
time.sleep(1)
|
||||
time.sleep(0.5)
|
||||
else:
|
||||
logger.warning("The element is not visible.")
|
||||
print("The element is not visible.")
|
||||
|
|
@ -81,7 +113,8 @@ def scroll_slow(driver, scrollable_element, start=0, end=3600, step=300, reverse
|
|||
logger.error("Exception occurred during scrolling: %s", e)
|
||||
print(f"Exception occurred: {e}")
|
||||
|
||||
def chromeBrowserOptions():
|
||||
|
||||
def chrome_browser_options():
|
||||
logger.debug("Setting Chrome browser options")
|
||||
ensure_chrome_profile()
|
||||
options = webdriver.ChromeOptions()
|
||||
|
|
@ -112,10 +145,10 @@ def chromeBrowserOptions():
|
|||
options.add_experimental_option("prefs", prefs)
|
||||
|
||||
if len(chromeProfilePath) > 0:
|
||||
initialPath = os.path.dirname(chromeProfilePath)
|
||||
profileDir = os.path.basename(chromeProfilePath)
|
||||
options.add_argument('--user-data-dir=' + initialPath)
|
||||
options.add_argument("--profile-directory=" + profileDir)
|
||||
initial_path = os.path.dirname(chromeProfilePath)
|
||||
profile_dir = os.path.basename(chromeProfilePath)
|
||||
options.add_argument('--user-data-dir=' + initial_path)
|
||||
options.add_argument("--profile-directory=" + profile_dir)
|
||||
logger.debug("Using Chrome profile directory: %s", chromeProfilePath)
|
||||
else:
|
||||
options.add_argument("--incognito")
|
||||
|
|
@ -123,14 +156,16 @@ def chromeBrowserOptions():
|
|||
|
||||
return options
|
||||
|
||||
|
||||
def printred(text):
|
||||
RED = "\033[91m"
|
||||
RESET = "\033[0m"
|
||||
red = "\033[91m"
|
||||
reset = "\033[0m"
|
||||
logger.debug("Printing text in red: %s", text)
|
||||
print(f"{RED}{text}{RESET}")
|
||||
print(f"{red}{text}{reset}")
|
||||
|
||||
|
||||
def printyellow(text):
|
||||
YELLOW = "\033[93m"
|
||||
RESET = "\033[0m"
|
||||
yellow = "\033[93m"
|
||||
reset = "\033[0m"
|
||||
logger.debug("Printing text in yellow: %s", text)
|
||||
print(f"{YELLOW}{text}{RESET}")
|
||||
print(f"{yellow}{text}{reset}")
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue