Spaces:
Runtime error
Runtime error
Update scraper.py
Browse files- scraper.py +10 -12
scraper.py
CHANGED
|
@@ -1,17 +1,18 @@
|
|
| 1 |
from selenium import webdriver
|
| 2 |
-
from selenium.webdriver.
|
| 3 |
-
from webdriver_manager.chrome import ChromeDriverManager
|
| 4 |
-
from selenium.webdriver.chrome.options import Options
|
| 5 |
import zipfile
|
| 6 |
import os
|
| 7 |
|
| 8 |
def scrape_pages(base_url, start_page, end_page):
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
|
| 14 |
-
driver = webdriver.
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
if not os.path.exists('pages'):
|
| 17 |
os.makedirs('pages')
|
|
@@ -38,9 +39,6 @@ def scrape_pages(base_url, start_page, end_page):
|
|
| 38 |
if __name__ == "__main__":
|
| 39 |
base_url = "https://shahvani.com/dastans"
|
| 40 |
start_page = 1
|
| 41 |
-
end_page = 10 # ای
|
| 42 |
-
|
| 43 |
-
# مثال: اگر میخواهید تعداد صفحات را به صورت پویا مشخص کنید
|
| 44 |
-
# end_page = get_last_page_number() # تابعی که تعداد صفحات را برمیگرداند
|
| 45 |
|
| 46 |
scrape_pages(base_url, start_page, end_page)
|
|
|
|
| 1 |
from selenium import webdriver
|
| 2 |
+
from selenium.webdriver.common.desired_capabilities import DesiredCapabilities
|
|
|
|
|
|
|
| 3 |
import zipfile
|
| 4 |
import os
|
| 5 |
|
| 6 |
def scrape_pages(base_url, start_page, end_page):
|
| 7 |
+
options = webdriver.ChromeOptions()
|
| 8 |
+
options.add_argument('--headless')
|
| 9 |
+
options.add_argument('--no-sandbox')
|
| 10 |
+
options.add_argument('--disable-dev-shm-usage')
|
| 11 |
|
| 12 |
+
driver = webdriver.Remote(
|
| 13 |
+
command_executor='http://localhost:4444/wd/hub',
|
| 14 |
+
options=options
|
| 15 |
+
)
|
| 16 |
|
| 17 |
if not os.path.exists('pages'):
|
| 18 |
os.makedirs('pages')
|
|
|
|
| 39 |
if __name__ == "__main__":
|
| 40 |
base_url = "https://shahvani.com/dastans"
|
| 41 |
start_page = 1
|
| 42 |
+
end_page = 10 # برای تست، عدد کمتری انتخاب کنید
|
|
|
|
|
|
|
|
|
|
| 43 |
|
| 44 |
scrape_pages(base_url, start_page, end_page)
|