feat: BCBS MA persistent session, Expand All fix, address/gender extraction
- Persistent Chrome profile (bcbs_ma_browser_manager) reused across patients; login probes /myportal/ to skip login + OTP while the session is active - Expand All: JS-click the expandCollapseViews span, verify sections opened, and unclip fixed-height result tables so the PDF includes every row - Extract patient address and gender from the results page and save to patient - Uppercase member IDs (XXA...), matching older lowercase records - Add /clear-bcbs-ma-session endpoint Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Claude Opus 5.5
@@ -0,0 +1,229 @@
|
||||
"""
|
||||
Browser manager for BCBS MA (Provider Central) - persistent profile with session management.
|
||||
- Uses --user-data-dir for persistent profile (login cookies + device trust)
|
||||
- One Chrome is reused across patients, so an active login skips OTP entirely
|
||||
- Tracks credentials to detect changes mid-session (triggers logout)
|
||||
- Anti-detection options to avoid bot detection
|
||||
|
||||
Unlike DDMA, cookies are NOT cleared on startup: BCBS MA (IBM Security Verify)
|
||||
keeps its login / device-trust state in cookies, so clearing them would force
|
||||
an OTP on the first check after every restart.
|
||||
"""
|
||||
import os
|
||||
import hashlib
|
||||
import threading
|
||||
from selenium import webdriver
|
||||
from selenium.webdriver.chrome.service import Service
|
||||
from webdriver_manager.chrome import ChromeDriverManager
|
||||
|
||||
# Ensure DISPLAY is set for Chrome to work (needed when running from SSH/background)
|
||||
if not os.environ.get("DISPLAY"):
|
||||
os.environ["DISPLAY"] = ":0"
|
||||
|
||||
|
||||
class BCBSMABrowserManager:
|
||||
"""
|
||||
Singleton that manages a persistent Chrome browser instance.
|
||||
- Uses --user-data-dir for persistent profile
|
||||
- Tracks credentials to detect changes mid-session
|
||||
"""
|
||||
_instance = None
|
||||
_lock = threading.Lock()
|
||||
|
||||
def __new__(cls):
|
||||
with cls._lock:
|
||||
if cls._instance is None:
|
||||
base = os.path.dirname(os.path.abspath(__file__))
|
||||
cls._instance = super().__new__(cls)
|
||||
cls._instance._driver = None
|
||||
cls._instance.profile_dir = os.path.join(base, "chrome_profile_bcbs_ma")
|
||||
cls._instance.download_dir = os.path.join(base, "downloads", "bcbs_ma")
|
||||
cls._instance._credentials_file = os.path.join(cls._instance.profile_dir, ".last_credentials")
|
||||
os.makedirs(cls._instance.profile_dir, exist_ok=True)
|
||||
os.makedirs(cls._instance.download_dir, exist_ok=True)
|
||||
return cls._instance
|
||||
|
||||
def clear_session(self):
|
||||
"""
|
||||
Log out by removing the profile's cookie / login-data files.
|
||||
Use when BCBS MA credentials are deleted or changed. Quits the browser
|
||||
first so Chrome doesn't rewrite the files on exit.
|
||||
"""
|
||||
print("[BCBS MA BrowserManager] Clearing session...")
|
||||
self.quit_driver()
|
||||
self._kill_existing_chrome_for_profile()
|
||||
self.clear_credentials_hash()
|
||||
|
||||
for filename in ["Cookies", "Cookies-journal", "Login Data", "Login Data-journal"]:
|
||||
for base in [os.path.join(self.profile_dir, "Default"), self.profile_dir]:
|
||||
filepath = os.path.join(base, filename)
|
||||
if os.path.exists(filepath):
|
||||
try:
|
||||
os.remove(filepath)
|
||||
print(f"[BCBS MA BrowserManager] Removed {filename}")
|
||||
except Exception as e:
|
||||
print(f"[BCBS MA BrowserManager] Could not remove {filename}: {e}")
|
||||
|
||||
# ── Credential hash tracking ──────────────────────────────────────────────
|
||||
|
||||
def _hash_credentials(self, username: str) -> str:
|
||||
return hashlib.sha256(username.encode()).hexdigest()[:16]
|
||||
|
||||
def get_last_credentials_hash(self) -> str | None:
|
||||
try:
|
||||
if os.path.exists(self._credentials_file):
|
||||
with open(self._credentials_file, 'r') as f:
|
||||
return f.read().strip()
|
||||
except Exception:
|
||||
pass
|
||||
return None
|
||||
|
||||
def save_credentials_hash(self, username: str):
|
||||
try:
|
||||
cred_hash = self._hash_credentials(username)
|
||||
with open(self._credentials_file, 'w') as f:
|
||||
f.write(cred_hash)
|
||||
except Exception as e:
|
||||
print(f"[BCBS MA BrowserManager] Failed to save credentials hash: {e}")
|
||||
|
||||
def credentials_changed(self, username: str) -> bool:
|
||||
last_hash = self.get_last_credentials_hash()
|
||||
if last_hash is None:
|
||||
return False # No previous credentials stored — not a change
|
||||
changed = last_hash != self._hash_credentials(username)
|
||||
if changed:
|
||||
print("[BCBS MA BrowserManager] Credentials changed — logout required")
|
||||
return changed
|
||||
|
||||
def clear_credentials_hash(self):
|
||||
try:
|
||||
if os.path.exists(self._credentials_file):
|
||||
os.remove(self._credentials_file)
|
||||
except Exception as e:
|
||||
print(f"[BCBS MA BrowserManager] Failed to clear credentials hash: {e}")
|
||||
|
||||
# ── Chrome process management ─────────────────────────────────────────────
|
||||
|
||||
def _kill_existing_chrome_for_profile(self):
|
||||
"""Kill any existing Chrome processes using this profile and remove lock files."""
|
||||
import subprocess
|
||||
import time as time_module
|
||||
try:
|
||||
result = subprocess.run(
|
||||
["pgrep", "-f", f"user-data-dir={self.profile_dir}"],
|
||||
capture_output=True, text=True
|
||||
)
|
||||
if result.stdout.strip():
|
||||
for pid in result.stdout.strip().split('\n'):
|
||||
try:
|
||||
subprocess.run(["kill", "-9", pid], check=False)
|
||||
except Exception:
|
||||
pass
|
||||
time_module.sleep(1)
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
for lock_file in ["SingletonLock", "SingletonSocket", "SingletonCookie"]:
|
||||
lock_path = os.path.join(self.profile_dir, lock_file)
|
||||
try:
|
||||
if os.path.islink(lock_path) or os.path.exists(lock_path):
|
||||
os.remove(lock_path)
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
# ── Driver lifecycle ──────────────────────────────────────────────────────
|
||||
|
||||
def get_driver(self, headless=False):
|
||||
"""Get or create the persistent browser instance."""
|
||||
with self._lock:
|
||||
if self._driver is None:
|
||||
print("[BCBS MA BrowserManager] Driver is None, creating new driver")
|
||||
self._kill_existing_chrome_for_profile()
|
||||
self._create_driver(headless)
|
||||
elif not self._is_alive():
|
||||
print("[BCBS MA BrowserManager] Driver not alive, recreating")
|
||||
self._kill_existing_chrome_for_profile()
|
||||
self._create_driver(headless)
|
||||
else:
|
||||
print("[BCBS MA BrowserManager] Reusing existing driver")
|
||||
return self._driver
|
||||
|
||||
def _is_alive(self):
|
||||
try:
|
||||
if self._driver is None:
|
||||
return False
|
||||
url = self._driver.current_url
|
||||
print(f"[BCBS MA BrowserManager] Driver alive, current URL: {url[:50]}...")
|
||||
return True
|
||||
except Exception as e:
|
||||
print(f"[BCBS MA BrowserManager] Driver not alive: {e}")
|
||||
return False
|
||||
|
||||
def _create_driver(self, headless=False):
|
||||
"""Create browser with persistent profile and anti-detection options."""
|
||||
if self._driver:
|
||||
try:
|
||||
self._driver.quit()
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
options = webdriver.ChromeOptions()
|
||||
if headless:
|
||||
options.add_argument("--headless=new")
|
||||
|
||||
# Persistent profile — keeps login cookies / device trust between runs
|
||||
options.add_argument(f"--user-data-dir={self.profile_dir}")
|
||||
options.add_argument("--no-sandbox")
|
||||
options.add_argument("--disable-dev-shm-usage")
|
||||
|
||||
# Anti-detection
|
||||
options.add_argument("--disable-blink-features=AutomationControlled")
|
||||
options.add_experimental_option("excludeSwitches", ["enable-automation"])
|
||||
options.add_experimental_option("useAutomationExtension", False)
|
||||
options.add_argument("--disable-infobars")
|
||||
|
||||
prefs = {
|
||||
"download.default_directory": self.download_dir,
|
||||
"plugins.always_open_pdf_externally": True,
|
||||
"download.prompt_for_download": False,
|
||||
"download.directory_upgrade": True,
|
||||
}
|
||||
options.add_experimental_option("prefs", prefs)
|
||||
|
||||
service = Service(ChromeDriverManager().install())
|
||||
self._driver = webdriver.Chrome(service=service, options=options)
|
||||
self._driver.maximize_window()
|
||||
|
||||
# Remove webdriver property to avoid detection
|
||||
try:
|
||||
self._driver.execute_script(
|
||||
"Object.defineProperty(navigator, 'webdriver', {get: () => undefined})"
|
||||
)
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
def quit_driver(self):
|
||||
"""Quit browser (only call on shutdown / logout — NOT between patients)."""
|
||||
with self._lock:
|
||||
if self._driver:
|
||||
try:
|
||||
self._driver.quit()
|
||||
except Exception:
|
||||
pass
|
||||
self._driver = None
|
||||
|
||||
|
||||
# ── Singleton accessor ────────────────────────────────────────────────────────
|
||||
|
||||
_manager = None
|
||||
|
||||
def get_browser_manager() -> BCBSMABrowserManager:
|
||||
global _manager
|
||||
if _manager is None:
|
||||
_manager = BCBSMABrowserManager()
|
||||
return _manager
|
||||
|
||||
|
||||
def clear_bcbs_ma_session():
|
||||
"""Log out of BCBS MA. Call when credentials are deleted or changed."""
|
||||
get_browser_manager().clear_session()
|
||||
Reference in New Issue
Block a user