Update conver.py

conver.py

@@ -10,6 +10,7 @@ import tempfile
 from pydub import AudioSegment
 import base64
 from pathlib import Path
+import hashlib
 
 @dataclass
 class ConversationConfig:
@@ -26,86 +27,50 @@ class URLToAudioConverter:
     def fetch_text(self, url: str) -> str:
         if not url:
             raise ValueError("URL cannot be empty")
-
-        full_url = f"{self.config.prefix_url}{url}"
-        try:
-            response = httpx.get(full_url, timeout=60.0)
-            response.raise_for_status()
-            return response.text
-        except httpx.HTTPError as e:
-            raise RuntimeError(f"Failed to fetch URL: {e}")
+        response = httpx.get(f"{self.config.prefix_url}{url}", timeout=60.0)
+        response.raise_for_status()
+        return response.text
 
     def extract_conversation(self, text: str) -> Dict:
-        if not text:
-            raise ValueError("Input text cannot be empty")
-
-        try:
-            prompt = (
-                f"{text}\nConvert the provided text into a short informative podcast conversation "
-                f"between two experts. Return ONLY a JSON object with the following structure:\n"
-                '{"conversation": [{"speaker": "Speaker1", "text": "..."}, {"speaker": "Speaker2", "text": "..."}]}'
-            )
-            
-            chat_completion = self.llm_client.chat.completions.create(
-                messages=[{"role": "user", "content": prompt}],
-                model=self.config.model_name,
-                response_format={"type": "json_object"}
-            )
-            
-            response_content = chat_completion.choices[0].message.content
-            json_str = response_content.strip()
-            
-            if not json_str.startswith('{'):
-                start = json_str.find('{')
-                if start != -1:
-                    json_str = json_str[start:]
-            
-            if not json_str.endswith('}'):
-                end = json_str.rfind('}')
-                if end != -1:
-                    json_str = json_str[:end+1]
-            
-            return json.loads(json_str)
-        except Exception as e:
-            print(f"Error en extract_conversation: {str(e)}")
-            print(f"Respuesta del modelo: {response_content}")
-            raise RuntimeError(f"Failed to extract conversation: {str(e)}")
+        prompt = (
+            f"{text}\nConvert the provided text into a short informative podcast conversation "
+            f"between two experts. Return ONLY a JSON object with the following structure:\n"
+            '{"conversation": [{"speaker": "Speaker1", "text": "..."}, {"speaker": "Speaker2", "text": "..."}]}'
+        )
+        chat_completion = self.llm_client.chat.completions.create(
+            messages=[{"role": "user", "content": prompt}],
+            model=self.config.model_name,
+            response_format={"type": "json_object"}
+        )
+        response_content = chat_completion.choices[0].message.content
+        json_str = response_content.strip()
+        if not json_str.startswith("{"):
+            json_str = json_str[json_str.find("{"):]
+        if not json_str.endswith("}"):
+            json_str = json_str[: json_str.rfind("}") + 1]
+        return json.loads(json_str)
 
     async def text_to_speech(self, conversation_json: Dict, voice_1: str, voice_2: str) -> Tuple[List[str], str]:
         output_dir = Path(self._create_output_directory())
         filenames = []
-        
-        try:
-            for i, turn in enumerate(conversation_json["conversation"]):
-                filename = output_dir / f"output_{i}.mp3"
-                voice = voice_1 if i % 2 == 0 else voice_2
-
-                tmp_path, error = await self._generate_audio(turn["text"], voice)
-                if error:
-                    raise RuntimeError(f"Text-to-speech failed: {error}")
-
-                os.rename(tmp_path, filename)
-                filenames.append(str(filename))
-
-            return filenames, str(output_dir)
-        except Exception as e:
-            raise RuntimeError(f"Failed to convert text to speech: {e}")
+        for i, turn in enumerate(conversation_json["conversation"]):
+            voice = voice_1 if i % 2 == 0 else voice_2
+            tmp_path, error = await self._generate_audio(turn["text"], voice)
+            if error:
+                raise RuntimeError(f"Text-to-speech failed: {error}")
+            filename = output_dir / f"output_{i}.mp3"
+            os.rename(tmp_path, filename)
+            filenames.append(str(filename))
+        return filenames, str(output_dir)
 
     async def _generate_audio(self, text: str, voice: str, rate: int = 0, pitch: int = 0) -> Tuple[str, str]:
-        if not text.strip():
-            return None, "Text cannot be empty"
-        if not voice:
-            return None, "Voice cannot be empty"
-
         voice_short_name = voice.split(" - ")[0]
         rate_str = f"{rate:+d}%"
         pitch_str = f"{pitch:+d}Hz"
         communicate = edge_tts.Communicate(text, voice_short_name, rate=rate_str, pitch=pitch_str)
-
         with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp_file:
             tmp_path = tmp_file.name
             await communicate.save(tmp_path)
-
         return tmp_path, None
 
     def _create_output_directory(self) -> str:
@@ -115,70 +80,39 @@ class URLToAudioConverter:
         return folder_name
 
     def combine_audio_files(self, filenames: List[str], output_file: str) -> None:
-        if not filenames:
-            raise ValueError("No input files provided")
-
-        try:
-            combined = AudioSegment.empty()
-            for filename in filenames:
-                audio_segment = AudioSegment.from_file(filename, format="mp3")
-                combined += audio_segment
-
-            combined.export(output_file, format="mp3")
-
-            # Limpieza
-            dir_path = os.path.dirname(filenames[0])
-            for file in os.listdir(dir_path):
-                file_path = os.path.join(dir_path, file)
-                if os.path.isfile(file_path):
-                    os.remove(file_path)
-            os.rmdir(dir_path)
-                
-        except Exception as e:
-            raise RuntimeError(f"Failed to combine audio files: {e}")
+        combined = AudioSegment.empty()
+        for filename in filenames:
+            combined += AudioSegment.from_file(filename, format="mp3")
+        combined.export(output_file, format="mp3")
+        dir_path = os.path.dirname(filenames[0])
+        for file in os.listdir(dir_path):
+            os.remove(os.path.join(dir_path, file))
+        os.rmdir(dir_path)
 
     async def url_to_audio(self, url: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
         text = self.fetch_text(url)
-
         words = text.split()
         if len(words) > self.config.max_words:
-            text = " ".join(words[:self.config.max_words])
-
+            text = " ".join(words[: self.config.max_words])
         conversation_json = self.extract_conversation(text)
-        conversation_text = "\n".join(
-            f"{turn['speaker']}: {turn['text']}" for turn in conversation_json["conversation"]
-        )
+        conversation_text = "\n".join(f"{t['speaker']}: {t['text']}" for t in conversation_json["conversation"])
         self.llm_out = conversation_json
-        audio_files, folder_name = await self.text_to_speech(
-            conversation_json, voice_1, voice_2
-        )
-
+        audio_files, folder_name = await self.text_to_speech(conversation_json, voice_1, voice_2)
         final_output = os.path.join(folder_name, "combined_output.mp3")
         self.combine_audio_files(audio_files, final_output)
         return final_output, conversation_text
 
     async def text_to_audio(self, text: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
-        """Procesamiento normal con LLM"""
         conversation_json = self.extract_conversation(text)
-        conversation_text = "\n".join(
-            f"{turn['speaker']}: {turn['text']}" for turn in conversation_json["conversation"]
-        )
-        audio_files, folder_name = await self.text_to_speech(
-            conversation_json, voice_1, voice_2
-        )
+        conversation_text = "\n".join(f"{t['speaker']}: {t['text']}" for t in conversation_json["conversation"])
+        audio_files, folder_name = await self.text_to_speech(conversation_json, voice_1, voice_2)
         final_output = os.path.join(folder_name, "combined_output.mp3")
         self.combine_audio_files(audio_files, final_output)
         return final_output, conversation_text
 
     async def raw_text_to_audio(self, text: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
-        """NUEVO: Modo sin LLM (texto directo)"""
-        conversation = {
-            "conversation": [
-                {"speaker": "Host", "text": text},
-                {"speaker": "Co-host", "text": "(Continuación del tema)"}
-            ]
-        }
-        audio_files, folder_name = await self.text_to_speech(conversation, voice_1, voice_2)
-        output_file = os.path.join(folder_name, "raw_podcast.mp3")
-        self.combine_audio_files(audio_files, output_file)
-        return text, output_file
+        hash_name = hashlib.md5(text.encode()).hexdigest()[:8]
+        output_file = f"podcast_{hash_name}.mp3"
+        communicate = edge_tts.Communicate(text, voice_1.split(" - ")[0])
+        await communicate.save(output_file)
+        return text, output_file