Update app.py

app.py

@@ -7,6 +7,7 @@ import subprocess
 import base64
 import json
 import shutil
+from typing import Optional
 
 app = FastAPI(title="VoiceClips heavy video processor (Wav2Lip + FFmpeg)")
 
@@ -20,9 +21,10 @@ class TranslationJob(BaseModel):
     caption_style: str
     resolution: str
     user_token: str
-    custom_font: str = None
-    custom_size: int = None
-    custom_color: str = None
+    # Null (boş) değerlerin 422 hatası vermemesi için Optional yapıldı
+    custom_font: Optional[str] = None
+    custom_size: Optional[int] = None
+    custom_color: Optional[str] = None
 
 def load_env_local():
     """Load environment variables from local .env.local file if it exists."""
@@ -298,7 +300,15 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
     
     print(f"[{video_id}] Running in PRODUCTION MODE...")
     
-    input_video_path = f"temp_{video_id}_input.mp4"
+    # Gelen URL'den uzantıyı tespit edip kaydediyoruz (.mp4, .mov, .webm vb.)
+    url_without_params = job.original_video_url.split('?')[0]
+    file_ext = url_without_params.split('.')[-1].lower() if '.' in url_without_params else 'mp4'
+    if file_ext not in ['mp4', 'mov', 'webm']:
+        file_ext = 'mp4'
+        
+    # Geçici çalışma dosyaları
+    downloaded_video_path = f"temp_{video_id}_downloaded.{file_ext}"
+    input_video_path = f"temp_{video_id}_input.mp4"  # Standartlaştırılmış MP4 dosyası
     extracted_audio_path = f"temp_{video_id}_audio.mp3"
     synthesized_audio_path = f"temp_{video_id}_tts.mp3"
     srt_file_path = f"temp_{video_id}.srt"
@@ -306,15 +316,26 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
     lipsync_output_path = f"temp_{video_id}_lipsync.mp4"
     
     try:
-        # 1. Download original video
-        print(f"[{video_id}] Downloading original video...")
+        # 1. Orijinal dosyayı indir
+        print(f"[{video_id}] Downloading original video from: {job.original_video_url}")
         res = requests.get(job.original_video_url, stream=True)
         res.raise_for_status()
-        with open(input_video_path, "wb") as f:
+        with open(downloaded_video_path, "wb") as f:
             for chunk in res.iter_content(chunk_size=8192):
                 f.write(chunk)
                 
-        # 2. Extract audio
+        # 1.5. Video standardizasyon (H.264/AAC, 25 FPS, YUV420p)
+        # Webm ve Mov formatlarını OpenCV ve Wav2Lip ile %100 uyumlu hale getirmek için standart MP4 formatına dönüştürüyoruz.
+        print(f"[{video_id}] Standardizing downloaded video to MP4 format...")
+        cmd_standardize = [
+            "ffmpeg", "-y", "-i", downloaded_video_path,
+            "-c:v", "libx264", "-pix_fmt", "yuv420p", "-r", "25",
+            "-c:a", "aac", "-ar", "16000", "-ac", "1",
+            input_video_path
+        ]
+        subprocess.run(cmd_standardize, check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+        
+        # 2. Ses Ayıklama
         print(f"[{video_id}] Extracting audio using FFmpeg...")
         cmd_extract = [
             "ffmpeg", "-y", "-i", input_video_path,
@@ -323,7 +344,7 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
         ]
         subprocess.run(cmd_extract, check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
         
-        # 3. Gemini Transcription & Translation
+        # 3. Gemini 2.0 ile Transkript & Çeviri
         print(f"[{video_id}] Calling Gemini 2.0 Flash...")
         with open(extracted_audio_path, "rb") as audio_file:
             audio_data = base64.b64encode(audio_file.read()).decode("utf-8")
@@ -359,7 +380,7 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
         else:
             translated_text = gemini_output
         
-        # 4. ElevenLabs Voice Cloning Synthesis
+        # 4. ElevenLabs Yapay Zeka Ses Sentezleme
         print(f"[{video_id}] Calling ElevenLabs TTS Voice Cloning...")
         voice_id = "21m00Tcm4TlvDq8ikWAM" 
         elevenlabs_url = f"https://api.elevenlabs.io/v1/text-to-speech/{voice_id}"
@@ -393,7 +414,7 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
         with open(synthesized_audio_path, "wb") as f:
             f.write(tts_res.content)
             
-        # 5. Lip-Sync & FFmpeg Merge
+        # 5. Lip-Sync & Video Birleştirme
         if job.has_lip_sync:
             setup_success = setup_wav2lip()
             if not setup_success:
@@ -451,7 +472,7 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
                 ]
                 subprocess.run(cmd_merge, check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
                 
-        # 6. Upload output to Supabase Storage
+        # 6. Sonuç Videosunu Supabase Storage'a Yükle
         print(f"[{video_id}] Uploading output to Supabase Storage...")
         storage_upload_url = f"{supabase_url}/storage/v1/object/videos/{video_id}_translated.mp4"
         
@@ -475,7 +496,8 @@ def run_production_pipeline(job: TranslationJob, gemini_key: str, elevenlabs_key
         update_db_status(video_id, user_token, "failed", error_message=f"Hata: {str(e)}")
         
     finally:
-        for temp_file in [input_video_path, extracted_audio_path, synthesized_audio_path, srt_file_path, output_video_path, lipsync_output_path]:
+        # Geçici dosyaların temizlenmesi
+        for temp_file in [downloaded_video_path, input_video_path, extracted_audio_path, synthesized_audio_path, srt_file_path, output_video_path, lipsync_output_path]:
             if os.path.exists(temp_file):
                 try:
                     os.remove(temp_file)