[DEV-Snehal] Json file changes

Desktop/Telegram_Bot/services/telegram_listener.py

 # python
 import asyncio
 import csv
-import os
 import json
+import os
 
 from pathlib import Path
-from dotenv import load_dotenv,find_dotenv
+from dotenv import load_dotenv, find_dotenv
 from datetime import datetime
 
 from telethon import TelegramClient, events
@@ -17,6 +17,7 @@ from telethon.tl.types import InputChannel, Channel, ReactionEmoji, ReactionCust
 from transformers import pipeline
 
 load_dotenv(find_dotenv())
+# load_dotenv(dotenv_path=r"{Path for the env file}")
 
 # Please create a .env file. Add the environment variables over there. And never the env file on github or gitlab.
 API_ID = int(os.getenv("API_ID"))
@@ -24,20 +25,19 @@ API_HASH = os.getenv("API_HASH")
 SESSION = os.getenv("SESSION")
 INVITE_LINK = os.getenv("INVITE_LINK")
 
-
-DATA_DIR = Path(os.getenv("DATA_DIR", "/app/data"))
-DATA_DIR.mkdir(parents=True, exist_ok=True)
-
-timestamp = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
-JSON_PATH = DATA_DIR / f"messages_{timestamp}.json"
-
 sentiment_analyzer = pipeline("sentiment-analysis")
 
-# These are the headers for the Json file. 
+# These are the headers for the csv file. 
 HEADER = ["chat_id", "message_id", "sender_id", "timestamp", "text",
           "has_media", "media_type", "file_name", "file_size_kb",
           "views", "reactions", "sentiment"]
 
+timestamp = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
+
+JSON_PATH = Path(f"output/messages_{timestamp}.json")
+
+# JSON_PATH = Path(r"messages_.json")  # e.g., "output/messages.json"
+
 async def list_joined_channels(client):
     channels = []
     async for dialog in client.iter_dialogs():
@@ -65,7 +65,7 @@ async def process_last_messages_for_all(client, process_message, limit):
     async def handle_all(event):
         await process_message(event.message)  
 
-    return channels  # optional, if you need the list
+    return channels
 
 
 async def process_last_message_for_one_channel(client, process_message, target, limit):
@@ -95,24 +95,49 @@ def write_csv_row(row_dict):
             writer.writeheader()
         writer.writerow(row_dict)
 
-
-def write_json_entry(entry):
-    """Append a single JSON entry to a file."""
+# def write_json_entry(entry):
+#     """Append a single JSON entry to a file."""
+#     try:
+#         existing_data = []
+#         if os.path.exists(JSON_PATH):
+#             with open(JSON_PATH, "r", encoding="utf-8-sig") as f:
+#                 try:
+#                     existing_data = json.load(f)
+#                 except json.JSONDecodeError:
+#                     existing_data = []
+
+#         existing_data.append(entry)
+#         with open(JSON_PATH, "w", encoding="utf-8") as f:
+#             json.dump(existing_data, f, ensure_ascii=False, indent=4)
+#     except Exception as e:
+#         print(f"JSON write error: {e}")
+
+def write_json_entry_per_post(entry, message_id, sender_id):
+    """Create a separate JSON file for each post."""
     try:
-        existing_data = []
-        if JSON_PATH.exists():
-            with open(JSON_PATH, "r", encoding="utf-8-sig") as f:
-                try:
-                    existing_data = json.load(f)
-                except json.JSONDecodeError:
-                    existing_data = []
-
-        existing_data.append(entry)
-        with open(JSON_PATH, "w", encoding="utf-8") as f:
-            json.dump(existing_data, f, ensure_ascii=False, indent=4)
+        os.makedirs("output", exist_ok=True)
+        json_path = Path(f"output/post_{message_id}_{sender_id}.json")
+        with open(json_path, "w", encoding="utf-8-sig") as f:
+            json.dump(entry, f, ensure_ascii=False, indent=4)
+        print(f"JSON file saved: {json_path}")
     except Exception as e:
-        print(f"JSON write error: {e}")
+        print(f"Error writing JSON for post {message_id}: {e}")
 
+def write_text_file_per_post(message,sender_id):
+    """Save only the text of a message to a separate .txt file."""
+    try:
+        os.makedirs("output", exist_ok=True)
+        text_content = message.text or ""
+        if not text_content.strip():
+            return  # skip empty or non-text messages
+
+        file_path = Path(f"output/post_{message.id}_{sender_id}.txt")
+        with open(file_path, "w", encoding="utf-8-sig") as f:
+            f.write(text_content.strip())
+
+        print(f"Text file saved: {file_path}")
+    except Exception as e:
+        print(f"Error writing text file for post {message.id}: {e}")
 
 def safe_serialize_reactions(message):
     r = getattr(message, "reactions", None)
@@ -127,6 +152,7 @@ def safe_serialize_reactions(message):
 
 async def process_message_for_csv(message):
     sentiment = await asyncio.to_thread(sentiment_analyzer, message.text or "")
+    sender_id = getattr(message, "sender_id", None)
     row = {
         "chat_id": message.chat_id,
         "message_id": message.id,
@@ -142,10 +168,11 @@ async def process_message_for_csv(message):
         "sentiment": sentiment[0]["label"] if sentiment else ""
     }
     try:
-        #write_csv_row(row)
-        write_json_entry(row)
+        # write_csv_row(row)
+        write_json_entry_per_post(row, message.id, sender_id)
+        write_text_file_per_post(message,sender_id)
     except Exception as e:
-        print(f"CSV write error: {e}")
+        print(f"JSON write error: {e}")
 
 def render_reaction_label(r):
         if isinstance(r, ReactionEmoji):
@@ -226,7 +253,8 @@ async def main():
 
     if target:
         print("Listening to single channel ")
-        process_last_message_for_one_channel(client, process_message, target, limit= 5)
+        await process_last_message_for_one_channel(client, process_message, target, limit= 5)
+        # After collecting
         
     else:
         print("No INVITE_URL provided. Listening to all chats.")