forust/telegram-scraper
1
0
Fork 0
Code Issues Pull Requests Actions Packages Activity

Enchance weblog

Browse Source
This commit is contained in:
mr-forust
2026-04-26 23:21:26 +02:00
parent cc99f752b3
commit ff000e88d4
2 changed files with 14 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files
telegram_scraper_with_forwarding.py
+8 -7
View File
@@ -829,7 +829,7 @@ class OptimizedTelegramScraper:
self.state['channels'][channel] = last_message_id self.state['channels'][channel] = last_message_id
self.save_state() self.save_state()
print(f"\nCompleted scraping channel {channel}") print(f"Completed scraping channel {channel}")
except Exception as e: except Exception as e:
print(f"Error with channel {channel}: {e}") print(f"Error with channel {channel}: {e}")
@@ -843,10 +843,10 @@ class OptimizedTelegramScraper:
channel_name = self.state.get('channel_names', {}).get(channel, 'Unknown') channel_name = self.state.get('channel_names', {}).get(channel, 'Unknown')
if not message_ids: if not message_ids:
print(f"No media files to reprocess for {channel_name} (ID: {channel})") # print(f"No media files to reprocess for {channel_name} (ID: {channel})")
return return
print(f"📥 Reprocessing {len(message_ids)} media files for {channel_name} (ID: {channel})") # print(f"📥 Reprocessing {len(message_ids)} media files for {channel_name} (ID: {channel})")
try: try:
entity = await self._resolve_entity(channel) entity = await self._resolve_entity(channel)
@@ -881,10 +881,10 @@ class OptimizedTelegramScraper:
filled_length = int(bar_length * completed_media // len(message_ids)) filled_length = int(bar_length * completed_media // len(message_ids))
bar = '█' * filled_length + '░' * (bar_length - filled_length) bar = '█' * filled_length + '░' * (bar_length - filled_length)
sys.stdout.write(f"\r🔄 Rescrape: [{bar}] {progress:.1f}% ({completed_media}/{len(message_ids)})") # sys.stdout.write(f"\r🔄 Rescrape: [{bar}] {progress:.1f}% ({completed_media}/{len(message_ids)})")
sys.stdout.flush() # sys.stdout.flush()
print(f"\n✅ Media reprocessing complete! ({successful_downloads}/{len(message_ids)} successful)") # print(f"\n✅ Media reprocessing complete! ({successful_downloads}/{len(message_ids)} successful)")
except Exception as e: except Exception as e:
print(f"Error reprocessing media: {e}") print(f"Error reprocessing media: {e}")
@@ -1306,7 +1306,8 @@ class OptimizedTelegramScraper:
await self.client.disconnect() await self.client.disconnect()
return False return False
else: else:
print("✅ Already authenticated!") pass
# print("✅ Already authenticated!")
return True return True
webui_server.py
+6 -3
View File
@@ -589,7 +589,9 @@ class ContinuousScrapeManager:
if not channels: if not channels:
self._log("No channels configured for continuous scraping.") self._log("No channels configured for continuous scraping.")
else: else:
self._log(f"Starting iteration for {len(channels)} channel(s).") self._log("=" * 50)
self._log(f"🚀 Starting iteration for {len(channels)} channel(s).")
self._log("-" * 50)
buffer = io.StringIO() buffer = io.StringIO()
try: try:
with ( with (
@@ -605,7 +607,8 @@ class ContinuousScrapeManager:
self.status["last_iteration_at"] = utc_now_iso() self.status["last_iteration_at"] = utc_now_iso()
self.status["last_finished_at"] = utc_now_iso() self.status["last_finished_at"] = utc_now_iso()
self.status["last_error"] = None self.status["last_error"] = None
self._log("Iteration finished.") self._log("-" * 50)
self._log("✅ Iteration finished.")
except Exception as exc: except Exception as exc:
output = buffer.getvalue().strip() output = buffer.getvalue().strip()
if output: if output:
@@ -616,7 +619,7 @@ class ContinuousScrapeManager:
self.status["last_error"] = str(exc) self.status["last_error"] = str(exc)
sleep_seconds = max(5, interval_minutes * 60) sleep_seconds = max(5, interval_minutes * 60)
self._log(f"Sleeping for {interval_minutes} minute(s).") self._log(f"💤 Sleeping for {interval_minutes} minute(s).")
interrupted = self.stop_event.wait(timeout=sleep_seconds) interrupted = self.stop_event.wait(timeout=sleep_seconds)
if interrupted: if interrupted:
break break