before claude

pricewatch/app/tasks/scheduler.py

@@ -6,6 +6,7 @@ from __future__ import annotations
 
 from dataclasses import dataclass
 from datetime import datetime, timedelta, timezone
+import hashlib
 from typing import Optional
 
 import redis
@@ -127,11 +128,13 @@ class ScrapingScheduler:
         interval_hours: int = 24,
         use_playwright: Optional[bool] = None,
         save_db: bool = True,
+        job_id: Optional[str] = None,
     ) -> ScheduledJobInfo:
         """Planifie un scraping recurrent (intervalle en heures)."""
         interval_seconds = int(timedelta(hours=interval_hours).total_seconds())
         next_run = datetime.now(timezone.utc) + timedelta(seconds=interval_seconds)
 
+        resolved_job_id = job_id or self._job_id_for_url(url)
         job = self.scheduler.schedule(
             scheduled_time=next_run,
             func=scrape_product,
@@ -139,6 +142,13 @@ class ScrapingScheduler:
             kwargs={"use_playwright": use_playwright, "save_db": save_db},
             interval=interval_seconds,
             repeat=None,
+            id=resolved_job_id,
         )
         logger.info(f"Job planifie: {job.id}, prochaine execution: {next_run.isoformat()}")
         return ScheduledJobInfo(job_id=job.id, next_run=next_run)
+
+    @staticmethod
+    def _job_id_for_url(url: str) -> str:
+        """Genere un job_id stable pour eviter les doublons."""
+        fingerprint = hashlib.sha1(url.strip().lower().encode("utf-8")).hexdigest()
+        return f"scrape_{fingerprint}"

pricewatch/app/tasks/scrape.py

@@ -157,6 +157,36 @@ def scrape_product(
             )
             success = False
             fetch_error = str(exc)
+        # Si captcha detecte via HTTP, forcer une tentative Playwright.
+        if (
+            fetch_method == FetchMethod.HTTP
+            and use_playwright
+            and snapshot.debug.errors
+            and any("captcha" in error.lower() for error in snapshot.debug.errors)
+        ):
+            logger.info("[FETCH] Captcha detecte, tentative Playwright")
+            pw_result = fetch_playwright(
+                canonical_url,
+                headless=not headful,
+                timeout_ms=timeout_ms,
+                save_screenshot=save_screenshot,
+            )
+            if pw_result.success and pw_result.html:
+                try:
+                    snapshot = store.parse(pw_result.html, canonical_url)
+                    snapshot.debug.method = FetchMethod.PLAYWRIGHT
+                    snapshot.debug.duration_ms = pw_result.duration_ms
+                    snapshot.debug.html_size_bytes = len(pw_result.html.encode("utf-8"))
+                    snapshot.add_note("Captcha detecte via HTTP, fallback Playwright")
+                    success = snapshot.debug.status != DebugStatus.FAILED
+                except Exception as exc:
+                    snapshot.add_note(f"Fallback Playwright echoue: {exc}")
+                    logger.error(f"[PARSE] Exception fallback Playwright: {exc}")
+                    fetch_error = str(exc)
+            else:
+                error = pw_result.error or "Erreur Playwright"
+                snapshot.add_note(f"Fallback Playwright echoue: {error}")
+                fetch_error = error
     else:
         snapshot = ProductSnapshot(
             source=store.store_id,