Merge pull request #4415 from seleniumbase/socks5h-and-pdf-download-updates

Updates for SOCKS5h support and PDF downloads

Author: mdmintz

README.md

@@ -763,6 +763,7 @@ pytest test_coffee_cart.py --trace
 --brave  # (Shortcut for "--browser=brave".)
 --comet  # (Shortcut for "--browser=comet".)
 --atlas  # (Shortcut for "--browser=atlas".)
+--chromium  # (Shortcut for using base `Chromium`)
 --settings-file=FILE  # (Override default SeleniumBase settings.)
 --env=ENV  # (Set the test env. Access with "self.env" in tests.)
 --account=STR  # (Set account. Access with "self.account" in tests.)

examples/cdp_mode/ReadMe.md

@@ -297,7 +297,7 @@ with SB(uc=True, test=True, ad_block=True) as sb:
     sb.click_if_visible('[data-automation-id="sb-btn-close-mark"]')
     items = sb.find_elements('[data-item-id]')
     for item in items:
-        if required_text in item.text:
+        if required_text.lower() in item.text.lower():
             description = item.querySelector(
                 '[data-automation-id="product-title"]'
             )

examples/cdp_mode/playwright/raw_pixelscan_async.py

@@ -0,0 +1,29 @@
+import asyncio
+import time
+from playwright.async_api import async_playwright
+from playwright.async_api import expect
+from seleniumbase import cdp_driver
+
+
+async def main():
+    driver = await cdp_driver.start_async()
+    endpoint_url = driver.get_endpoint_url()
+
+    async with async_playwright() as p:
+        browser = await p.chromium.connect_over_cdp(endpoint_url)
+        page = browser.contexts[0].pages[0]
+        await page.goto("https://pixelscan.net/fingerprint-check")
+        time.sleep(4)
+        await expect(
+            page.locator("pxlscn-bot-detection")
+        ).to_contain_text("No automated behavior", timeout=4000)
+        await page.wait_for_selector("span.status-success", timeout=4000)
+        await expect(
+            page.locator("pxlscn-fingerprint-masking")
+        ).to_contain_text("No masking detected", timeout=4000)
+        time.sleep(2)
+        print("Bot Not Detected")
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

examples/cdp_mode/playwright/raw_pixelscan_sync.py

@@ -0,0 +1,21 @@
+from playwright.sync_api import sync_playwright
+from playwright.sync_api import expect
+from seleniumbase import sb_cdp
+
+sb = sb_cdp.Chrome()
+endpoint_url = sb.get_endpoint_url()
+
+with sync_playwright() as p:
+    browser = p.chromium.connect_over_cdp(endpoint_url)
+    page = browser.contexts[0].pages[0]
+    page.goto("https://pixelscan.net/fingerprint-check")
+    sb.sleep(4)
+    expect(
+        page.locator("pxlscn-bot-detection")
+    ).to_contain_text("No automated behavior", timeout=4000)
+    page.wait_for_selector("span.status-success", timeout=4000)
+    expect(
+        page.locator("pxlscn-fingerprint-masking")
+    ).to_contain_text("No masking detected", timeout=4000)
+    sb.sleep(2)
+    print("Bot Not Detected")

examples/cdp_mode/playwright/raw_walmart_sync.py

@@ -33,7 +33,7 @@
     items = page.locator('[data-item-id]')
     for i in range(items.count()):
         item = items.nth(i)
-        if required_text in item.inner_text():
+        if required_text.lower() in item.inner_text().lower():
             description = item.locator('[data-automation-id="product-title"]')
             if (
                 description

examples/cdp_mode/raw_cdp_target.py

@@ -0,0 +1,27 @@
+from seleniumbase import sb_cdp
+
+sb = sb_cdp.Chrome()
+sb.goto("https://www.target.com/")
+sb.sleep(1.5)
+sb.click("input#search")
+sb.sleep(0.5)
+search = "Settlers of Catan Board Game"
+required_text = "Catan"
+sb.type("input#search", search)
+sb.sleep(0.5)
+sb.click('button[aria-label="search"]')
+sb.sleep(2.5)
+print('*** Target Search for "%s":' % search)
+print('    (Results must contain "%s".)' % required_text)
+unique_item_text = []
+items = sb.find_elements('[data-test="product-details"]')
+for item in items:
+    if required_text.lower() in item.text.lower():
+        description = item.querySelector('a[data-test*="Card/title"]')
+        if description and description.text not in unique_item_text:
+            unique_item_text.append(description.text)
+            print("* " + description.text)
+            price = item.querySelector('[data-test="current-price"]')
+            if price:
+                print("  (" + price.text + ")")
+                item.scroll_into_view()

examples/cdp_mode/raw_cdp_tavus.py

@@ -21,7 +21,7 @@
 sb.click_if_visible('[data-automation-id="sb-btn-close-mark"]')
 items = sb.find_elements('[data-item-id]')
 for item in items:
-    if required_text in item.text:
+    if required_text.lower() in item.text.lower():
         description = item.querySelector(
             '[data-automation-id="product-title"]'
         )

examples/cdp_mode/raw_cdp_walmart.py

@@ -0,0 +1,24 @@
+from seleniumbase import sb_cdp
+
+sb = sb_cdp.Chrome(use_chromium=True, ad_block=True)
+sb.goto("https://www.zillow.com/")
+sb.sleep(2)
+search = "Bar Harbor ME homes on the waterfront"
+sb.type('input[aria-label="Search"]', search)
+sb.sleep(1)
+sb.click('button[type="submit"]')
+sb.sleep(2)
+sb.click_if_visible('span:contains("Cancel")')
+sb.sleep(0.5)
+sb.click_if_visible('button[title="Close"]')
+items = sb.find_visible_elements('[data-testid="property-card"]')
+print('*** %s:' % search)
+for i, item in enumerate(items):
+    print("<----- %s ----->" % (i + 1))
+    print(item.query_selector('[data-testid*="price"]').text)
+    print(item.query_selector('[data-testid*="details"]').text)
+    print(item.query_selector('[data-testid*="address"]').text)
+    item.scroll_into_view()
+print("<-------------->")
+sb.sleep(2)
+sb.quit()

examples/cdp_mode/raw_cdp_zillow.py

@@ -3,13 +3,13 @@
 with SB(uc=True, test=True) as sb:
     sb.activate_cdp_mode()
     sb.goto("https://www.indeed.com/companies/search")
-    sb.sleep(2)
+    sb.sleep(0.8)
     sb.solve_captcha()
-    sb.sleep(1)
+    sb.sleep(0.8)
     search_box = "input#company-search"
     if not sb.is_element_present(search_box):
         sb.solve_captcha()
-        sb.sleep(1)
+        sb.sleep(0.8)
     company = "NASA Jet Propulsion Laboratory"
     sb.click(search_box)
     sb.sleep(0.1)