Update app.py

app.py

@@ -7,6 +7,10 @@ import tempfile
 import os
 import uvicorn
 import logging
+import hashlib
+import json
+import sqlite3
+from datetime import datetime
 
 # تنظیم لاگ
 logging.basicConfig(level=logging.INFO)
@@ -27,9 +31,82 @@ logger.info(f"Loading model on {device}")
 model = whisper.load_model("large-v3", device=device)
 logger.info("Model loaded successfully")
 
+# ایجاد دیتابیس کش
+def init_cache_db():
+    conn = sqlite3.connect('transcription_cache.db')
+    cursor = conn.cursor()
+    cursor.execute('''
+        CREATE TABLE IF NOT EXISTS cache (
+            id INTEGER PRIMARY KEY AUTOINCREMENT,
+            file_hash TEXT UNIQUE,
+            filename TEXT,
+            file_size INTEGER,
+            transcription TEXT,
+            created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
+        )
+    ''')
+    conn.commit()
+    conn.close()
+
+# محاسبه هش فایل
+def calculate_file_hash(content, filename, file_size):
+    hash_input = f"{filename}_{file_size}_{len(content)}"
+    file_hash = hashlib.md5(content[:1024] + content[-1024:] + hash_input.encode()).hexdigest()
+    return file_hash
+
+# جستجو در کش
+def get_from_cache(file_hash):
+    try:
+        conn = sqlite3.connect('transcription_cache.db')
+        cursor = conn.cursor()
+        cursor.execute('SELECT transcription FROM cache WHERE file_hash = ?', (file_hash,))
+        result = cursor.fetchone()
+        conn.close()
+        return result[0] if result else None
+    except:
+        return None
+
+# ذخیره در کش
+def save_to_cache(file_hash, filename, file_size, transcription):
+    try:
+        conn = sqlite3.connect('transcription_cache.db')
+        cursor = conn.cursor()
+        cursor.execute('''
+            INSERT OR REPLACE INTO cache (file_hash, filename, file_size, transcription)
+            VALUES (?, ?, ?, ?)
+        ''', (file_hash, filename, file_size, transcription))
+        conn.commit()
+        conn.close()
+    except Exception as e:
+        logger.error(f"Error saving to cache: {e}")
+
+# پاک کردن کش قدیمی (بیش از 30 روز)
+def cleanup_old_cache():
+    try:
+        conn = sqlite3.connect('transcription_cache.db')
+        cursor = conn.cursor()
+        cursor.execute("DELETE FROM cache WHERE created_at < datetime('now', '-30 days')")
+        conn.commit()
+        conn.close()
+    except Exception as e:
+        logger.error(f"Error cleaning cache: {e}")
+
+# راه‌اندازی دیتابیس
+init_cache_db()
+
 @app.get("/")
 async def root():
-    return {"message": "Whisper API is running", "device": device}
+    conn = sqlite3.connect('transcription_cache.db')
+    cursor = conn.cursor()
+    cursor.execute('SELECT COUNT(*) FROM cache')
+    cache_count = cursor.fetchone()[0]
+    conn.close()
+    
+    return {
+        "message": "Whisper API is running", 
+        "device": device,
+        "cached_files": cache_count
+    }
 
 @app.post("/transcribe")
 async def transcribe_audio(file: UploadFile = File(...)):
@@ -51,6 +128,22 @@ async def transcribe_audio(file: UploadFile = File(...)):
         if file_size == 0:
             raise HTTPException(status_code=400, detail="Empty file")
         
+        # محاسبه هش فایل
+        file_hash = calculate_file_hash(contents, file.filename, file_size)
+        logger.info(f"File hash: {file_hash}")
+        
+        # جستجو در کش
+        cached_result = get_from_cache(file_hash)
+        if cached_result:
+            logger.info("Found in cache, returning cached result")
+            return JSONResponse({
+                "text": cached_result,
+                "from_cache": True,
+                "message": "نتیجه از کش بازگردانده شد"
+            })
+        
+        logger.info("Not found in cache, processing...")
+        
         # تشخیص فرمت فایل
         file_ext = os.path.splitext(file.filename)[1].lower()
         if not file_ext:
@@ -75,9 +168,20 @@ async def transcribe_audio(file: UploadFile = File(...)):
         
         text = result["text"].strip()
         if not text:
-            return JSONResponse({"text": "متن شناسایی نشد", "warning": "No speech detected"})
+            text = "متن شناسایی نشد"
+        
+        # ذخیره در کش
+        save_to_cache(file_hash, file.filename, file_size, text)
+        logger.info("Result saved to cache")
         
-        return JSONResponse({"text": text})
+        # پاک کردن کش قدیمی
+        cleanup_old_cache()
+        
+        return JSONResponse({
+            "text": text,
+            "from_cache": False,
+            "message": "پردازش جدید انجام شد و در کش ذخیره شد"
+        })
     
     except Exception as e:
         logger.error(f"Error in transcription: {str(e)}")
@@ -98,8 +202,33 @@ async def transcribe_audio(file: UploadFile = File(...)):
             except:
                 pass
 
+@app.get("/cache/stats")
+async def cache_stats():
+    try:
+        conn = sqlite3.connect('transcription_cache.db')
+        cursor = conn.cursor()
+        
+        cursor.execute('SELECT COUNT(*) FROM cache')
+        total_count = cursor.fetchone()[0]
+        
+        cursor.execute('SELECT COUNT(*) FROM cache WHERE created_at >= datetime("now", "-1 day")')
+        today_count = cursor.fetchone()[0]
+        
+        cursor.execute('SELECT AVG(LENGTH(transcription)) FROM cache')
+        avg_text_length = cursor.fetchone()[0] or 0
+        
+        conn.close()
+        
+        return {
+            "total_cached_files": total_count,
+            "cached_today": today_count,
+            "average_text_length": int(avg_text_length)
+        }
+    except Exception as e:
+        return {"error": str(e)}
+
 if __name__ == "__main__":
     uvicorn.run(app, host="0.0.0.0", port=7860, timeout_keep_alive=900)
 
 if __name__ == "__main__":
-    uvicorn.run(app, host="0.0.0.0", port=7860, timeout_keep_alive=900)
+    uvicorn.run(app, host="0.0.0.0", port=7860, timeout_keep_alive=300)