feat: caché de contenido de fuentes — reutiliza URLs scrapeadas en últimos 7 días

src/db/database.py

@@ -318,6 +318,19 @@ class ResearchDB:
         row = await cursor.fetchone()
         return row[0] if row else None
 
+    async def get_cached_content(self, url: str,
+                                  max_age_days: int = 7) -> Optional[str]:
+        threshold = time.time() - (max_age_days * 86400)
+        async with self.db.execute(
+            """SELECT sc.content FROM source_contents sc
+               JOIN sources s ON s.id = sc.source_id
+               WHERE s.url = ? AND sc.created_at > ?
+               ORDER BY sc.created_at DESC LIMIT 1""",
+            (url, threshold)
+        ) as cur:
+            row = await cur.fetchone()
+            return row[0] if row else None
+
     async def get_outputs(self, session_id: int) -> list[dict]:
         cursor = await self.db.execute(
             "SELECT * FROM outputs WHERE session_id = ? ORDER BY created_at DESC",

src/scraper/exhaustive.py

@@ -414,6 +414,23 @@ class ExhaustiveScraper:
             source_id = source["id"]
 
             try:
+                try:
+                    cached = await self.db.get_cached_content(url)
+                except Exception as cache_err:
+                    logger.warning("Cache lookup failed", url=url, error=str(cache_err))
+                    cached = None
+
+                if cached:
+                    logger.debug("Cache hit", url=url)
+                    await self.db.save_source_content(source_id, cached)
+                    await self.db.update_source(
+                        source_id,
+                        status="scraped",
+                        scraped_at=time.time(),
+                        word_count=len(cached.split()),
+                    )
+                    return 0
+
                 if source_type == "youtube":
                     content, title = await fetch_with_retry(
                         lambda: self._extract_youtube(url), url