add cookie session persistence

iaai_scraper/scraper.py

@@ -1,8 +1,12 @@
+import json
 import logging
+import os
 import re
 import signal
 import time
 import uuid
+from datetime import datetime, timezone
+from pathlib import Path
 
 from playwright.sync_api import Error as PlaywrightError
 from playwright.sync_api import BrowserContext, Page, sync_playwright
@@ -86,6 +90,7 @@ class IAAIScraper:
     def close(self) -> None:
         if self.context is not None:
             try:
+                self._save_storage_state()
                 self.context.close()
             except PlaywrightError:
                 pass
@@ -106,12 +111,51 @@ class IAAIScraper:
             finally:
                 self.playwright = None
 
+    def _load_storage_state(self) -> str | None:
+        """Load saved browser session (cookies + localStorage) if valid."""
+        session_cfg = self.settings.session
+        if not session_cfg.enabled:
+            return None
+        state_path = session_cfg.storage_state_path
+        if not state_path or not os.path.isfile(state_path):
+            logger.info("No saved session found at %s", state_path)
+            return None
+        try:
+            mtime = os.path.getmtime(state_path)
+            age_days = (time.time() - mtime) / 86400
+            if age_days > session_cfg.max_age_days:
+                logger.info("Session expired (%.1f days old, max %d). Starting fresh.", age_days, session_cfg.max_age_days)
+                os.remove(state_path)
+                return None
+            logger.info("Reusing saved session from %s (%.1f days old)", state_path, age_days)
+            return state_path
+        except Exception as exc:
+            logger.warning("Failed to load session state: %s", exc)
+            return None
+
+    def _save_storage_state(self) -> None:
+        """Save browser session (cookies + localStorage) to disk."""
+        session_cfg = self.settings.session
+        if not session_cfg.enabled or not session_cfg.save_on_exit:
+            return
+        if not self.context:
+            return
+        state_path = session_cfg.storage_state_path
+        try:
+            Path(state_path).parent.mkdir(parents=True, exist_ok=True)
+            self.context.storage_state(path=state_path)
+            logger.info("Session state saved to %s", state_path)
+        except Exception as exc:
+            logger.warning("Failed to save session state: %s", exc)
+
     def _new_context(self, storage_state: str | None = None) -> BrowserContext:
         if self.browser is None:
             self.__enter__()
         if self.context:
+            self._save_storage_state()
             self.context.close()
-        self.context = self.browser_factory.create_context(self.browser, storage_state=storage_state)
+        effective_state = storage_state or self._load_storage_state()
+        self.context = self.browser_factory.create_context(self.browser, storage_state=effective_state)
         return self.context
 
     def init_db(self):