Update app.py

app.py

@@ -2,52 +2,44 @@ import gradio as gr
 from huggingface_hub import HfApi
 import os
 import uuid
-import random
 
 # --- SYSTEM CONFIGURATION ---
+# The Bridge looks for the Secret Token we just added
 HF_TOKEN = os.getenv("HF_TOKEN")
-# We stick to ONE repo because 5TB is massive
-DATASET_ID = "Threatthriver/sys_log_dump_v4_shards" 
+
+# TARGET REPO: Your specific dataset
+DATASET_ID = "Threatthriver/sys_log_dump_v4_shards"
+HIDDEN_PATH = "data/blobs/0x44"
 
 # Initialize Connection
 api = HfApi(token=HF_TOKEN)
 
-def get_random_shard_folder():
-    """
-    Returns a random hex folder (0x00 to 0xFF).
-    This spreads files across 256 folders to prevent Git lag.
-    """
-    shard_id = random.randint(0, 255)
-    return f"data/blobs/0x{shard_id:02x}"
-
 def system_shard_allocator(packet_file):
     """
-    Allocates binary shard to persistent storage (5TB limit).
+    INTERNAL: Allocates binary shard to persistent storage.
     """
     if packet_file is None:
         return "ERROR: No packet received."
     
-    # 1. Generate Stealth ID
+    # 1. Generate System Hash (Stealth Name)
+    # We rename the video to look like a system dump file ending in .bin
     request_id = str(uuid.uuid4())
     shard_name = f"sys_core_dump_{request_id}.bin"
     
-    # 2. Pick a random subfolder to keep directories clean
-    hidden_path = get_random_shard_folder()
-    
-    print(f"⚙️ Allocating Shard: {hidden_path}/{shard_name}...")
+    print(f"⚙️ Allocating Shard: {shard_name}...")
     
     try:
-        # 3. Upload to the Dataset
+        # 2. Upload to the Hidden Folder in the Dataset
         api.upload_file(
             path_or_fileobj=packet_file.name,
-            path_in_repo=f"{hidden_path}/{shard_name}",
+            path_in_repo=f"{HIDDEN_PATH}/{shard_name}",
             repo_id=DATASET_ID,
             repo_type="dataset",
             commit_message=f"Auto-sync daemon: shard {request_id}"
         )
         
-        # 4. Generate the Stream URL
-        data_stream_url = f"https://huggingface.co/datasets/{DATASET_ID}/resolve/main/{hidden_path}/{shard_name}"
+        # 3. Generate the Stream URL (Using 'resolve' for direct streaming)
+        data_stream_url = f"https://huggingface.co/datasets/{DATASET_ID}/resolve/main/{HIDDEN_PATH}/{shard_name}"
         
         return data_stream_url
     
@@ -55,16 +47,18 @@ def system_shard_allocator(packet_file):
         return f"ALLOCATION_ERROR: {str(e)}"
 
 # --- DAEMON INTERFACE ---
-if __name__ == "__main__":
-    with gr.Blocks(title="System Sync Daemon") as system_daemon:
-        gr.Markdown("### ⚠️ SYSTEM SYNC DAEMON [5TB STORAGE NODE]")
-        
-        with gr.Row():
-            input_packet = gr.File(label="Binary Packet Input")
-            output_log = gr.Textbox(label="Allocation Path")
-        
-        btn = gr.Button("Execute Sync")
-        # api_name="predict" is required for the Next.js client
-        btn.click(fn=system_shard_allocator, inputs=input_packet, outputs=output_log, api_name="predict")
+with gr.Blocks(title="System Sync Daemon") as system_daemon:
+    gr.Markdown("### ⚠️ SYSTEM SYNC DAEMON [BACKGROUND PROCESS]")
+    
+    with gr.Row():
+        input_packet = gr.File(label="Binary Packet Input")
+        output_log = gr.Textbox(label="Allocation Path")
+    
+    btn = gr.Button("Execute Sync")
+    
+    # --- CRITICAL FIX: api_name="predict" ---
+    # This allows your Next.js client to call client.predict("/predict", ...)
+    btn.click(fn=system_shard_allocator, inputs=input_packet, outputs=output_log, api_name="predict")
 
-    system_daemon.queue().launch()
+# Launch
+system_daemon.queue().launch()