Spaces:
Running
Running
Update conver.py
Browse files
conver.py
CHANGED
@@ -10,8 +10,7 @@ import tempfile
|
|
10 |
from pydub import AudioSegment
|
11 |
import base64
|
12 |
from pathlib import Path
|
13 |
-
import
|
14 |
-
import asyncio
|
15 |
|
16 |
@dataclass
|
17 |
class ConversationConfig:
|
@@ -28,43 +27,71 @@ class URLToAudioConverter:
|
|
28 |
def fetch_text(self, url: str) -> str:
|
29 |
if not url:
|
30 |
raise ValueError("URL cannot be empty")
|
31 |
-
|
32 |
-
|
33 |
-
|
|
|
|
|
|
|
|
|
|
|
34 |
|
35 |
def extract_conversation(self, text: str) -> Dict:
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
53 |
|
54 |
async def text_to_speech(self, conversation_json: Dict, voice_1: str, voice_2: str) -> Tuple[List[str], str]:
|
55 |
-
output_dir =
|
56 |
filenames = []
|
57 |
-
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
|
|
|
|
|
|
66 |
|
67 |
async def _generate_audio(self, text: str, voice: str, rate: int = 0, pitch: int = 0) -> Tuple[str, str]:
|
|
|
|
|
|
|
|
|
68 |
voice_short_name = voice.split(" - ")[0]
|
69 |
rate_str = f"{rate:+d}%"
|
70 |
pitch_str = f"{pitch:+d}Hz"
|
@@ -75,80 +102,65 @@ class URLToAudioConverter:
|
|
75 |
return tmp_path, None
|
76 |
|
77 |
def _create_output_directory(self) -> str:
|
78 |
-
|
79 |
-
|
80 |
-
|
81 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
|
83 |
def combine_audio_files(self, filenames: List[str], output_file: str) -> None:
|
84 |
-
|
85 |
-
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
|
|
92 |
|
93 |
async def url_to_audio(self, url: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
|
|
|
94 |
text = self.fetch_text(url)
|
95 |
words = text.split()
|
96 |
if len(words) > self.config.max_words:
|
97 |
-
text = " ".join(words[:
|
98 |
conversation_json = self.extract_conversation(text)
|
99 |
-
conversation_text = "\n".join(
|
|
|
|
|
100 |
self.llm_out = conversation_json
|
101 |
-
audio_files,
|
102 |
-
|
103 |
-
self.combine_audio_files(audio_files,
|
104 |
-
return
|
105 |
|
106 |
async def text_to_audio(self, text: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
|
|
|
107 |
conversation_json = self.extract_conversation(text)
|
108 |
-
conversation_text = "\n".join(
|
109 |
-
|
110 |
-
|
111 |
-
self.
|
112 |
-
|
|
|
|
|
113 |
|
114 |
async def raw_text_to_audio(self, text: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
|
115 |
-
|
116 |
-
|
117 |
-
|
118 |
-
|
119 |
-
|
120 |
-
|
121 |
-
|
122 |
-
|
123 |
-
|
124 |
-
|
125 |
-
|
126 |
-
# Verifica voces disponibles (DEBUG)
|
127 |
-
voices = await edge_tts.list_voices()
|
128 |
-
voice_names = [v['Name'] for v in voices]
|
129 |
-
print(f"Voces disponibles (primeras 5): {voice_names[:5]}...")
|
130 |
-
|
131 |
-
# Extrae el nombre corto de la voz (ej: "en-US-AvaMultilingualNeural")
|
132 |
-
voice_short = voice_1.split(" - ")[0] if " - " in voice_1 else voice_1
|
133 |
-
print(f"Voz a usar: {voice_short}")
|
134 |
-
|
135 |
-
# Genera el audio
|
136 |
-
communicate = edge_tts.Communicate(text, voice_short)
|
137 |
-
print("Generando audio...")
|
138 |
-
await communicate.save(output_file)
|
139 |
-
print("Audio generado.")
|
140 |
-
|
141 |
-
# Verifica que el archivo existe y no está vacío
|
142 |
-
if not os.path.exists(output_file):
|
143 |
-
print("ERROR: Archivo no creado.")
|
144 |
-
return "Error: Archivo no generado", None
|
145 |
-
elif os.path.getsize(output_file) == 0:
|
146 |
-
print("ERROR: Archivo vacío.")
|
147 |
-
return "Error: Archivo de audio vacío", None
|
148 |
-
|
149 |
-
print(f"=== DEBUG FIN (Archivo válido: {output_file}) ===")
|
150 |
-
return text, output_file
|
151 |
-
|
152 |
-
except Exception as e:
|
153 |
-
print(f"ERROR CRÍTICO: {str(e)}")
|
154 |
-
return f"Error: {str(e)}", None
|
|
|
10 |
from pydub import AudioSegment
|
11 |
import base64
|
12 |
from pathlib import Path
|
13 |
+
import time
|
|
|
14 |
|
15 |
@dataclass
|
16 |
class ConversationConfig:
|
|
|
27 |
def fetch_text(self, url: str) -> str:
|
28 |
if not url:
|
29 |
raise ValueError("URL cannot be empty")
|
30 |
+
|
31 |
+
full_url = f"{self.config.prefix_url}{url}"
|
32 |
+
try:
|
33 |
+
response = httpx.get(full_url, timeout=60.0)
|
34 |
+
response.raise_for_status()
|
35 |
+
return response.text
|
36 |
+
except httpx.HTTPError as e:
|
37 |
+
raise RuntimeError(f"Failed to fetch URL: {e}")
|
38 |
|
39 |
def extract_conversation(self, text: str) -> Dict:
|
40 |
+
if not text:
|
41 |
+
raise ValueError("Input text cannot be empty")
|
42 |
+
|
43 |
+
try:
|
44 |
+
prompt = (
|
45 |
+
f"{text}\nConvert the provided text into a short informative podcast conversation "
|
46 |
+
f"between two experts. Return ONLY a JSON object with the following structure:\n"
|
47 |
+
'{"conversation": [{"speaker": "Speaker1", "text": "..."}, {"speaker": "Speaker2", "text": "..."}]}'
|
48 |
+
)
|
49 |
+
|
50 |
+
chat_completion = self.llm_client.chat.completions.create(
|
51 |
+
messages=[{"role": "user", "content": prompt}],
|
52 |
+
model=self.config.model_name,
|
53 |
+
response_format={"type": "json_object"}
|
54 |
+
)
|
55 |
+
|
56 |
+
response_content = chat_completion.choices[0].message.content
|
57 |
+
json_str = response_content.strip()
|
58 |
+
|
59 |
+
if not json_str.startswith('{'):
|
60 |
+
start = json_str.find('{')
|
61 |
+
if start != -1:
|
62 |
+
json_str = json_str[start:]
|
63 |
+
|
64 |
+
if not json_str.endswith('}'):
|
65 |
+
end = json_str.rfind('}')
|
66 |
+
if end != -1:
|
67 |
+
json_str = json_str[:end+1]
|
68 |
+
|
69 |
+
return json.loads(json_str)
|
70 |
+
except Exception as e:
|
71 |
+
print(f"Error en extract_conversation: {str(e)}")
|
72 |
+
raise RuntimeError(f"Failed to extract conversation: {str(e)}")
|
73 |
|
74 |
async def text_to_speech(self, conversation_json: Dict, voice_1: str, voice_2: str) -> Tuple[List[str], str]:
|
75 |
+
output_dir = self._create_output_directory()
|
76 |
filenames = []
|
77 |
+
try:
|
78 |
+
for i, turn in enumerate(conversation_json["conversation"]):
|
79 |
+
filename = os.path.join(output_dir, f"output_{i}.mp3")
|
80 |
+
voice = voice_1 if i % 2 == 0 else voice_2
|
81 |
+
tmp_path, error = await self._generate_audio(turn["text"], voice)
|
82 |
+
if error:
|
83 |
+
raise RuntimeError(f"Text-to-speech failed: {error}")
|
84 |
+
os.rename(tmp_path, filename)
|
85 |
+
filenames.append(filename)
|
86 |
+
return filenames, output_dir
|
87 |
+
except Exception as e:
|
88 |
+
raise RuntimeError(f"Failed to convert text to speech: {e}")
|
89 |
|
90 |
async def _generate_audio(self, text: str, voice: str, rate: int = 0, pitch: int = 0) -> Tuple[str, str]:
|
91 |
+
if not text.strip():
|
92 |
+
return None, "Text cannot be empty"
|
93 |
+
if not voice:
|
94 |
+
return None, "Voice cannot be empty"
|
95 |
voice_short_name = voice.split(" - ")[0]
|
96 |
rate_str = f"{rate:+d}%"
|
97 |
pitch_str = f"{pitch:+d}Hz"
|
|
|
102 |
return tmp_path, None
|
103 |
|
104 |
def _create_output_directory(self) -> str:
|
105 |
+
os.makedirs("outputs", exist_ok=True)
|
106 |
+
return "outputs"
|
107 |
+
|
108 |
+
def clean_old_files(self, directory: str = "outputs", max_age_seconds: int = 86400):
|
109 |
+
now = time.time()
|
110 |
+
for filename in os.listdir(directory):
|
111 |
+
file_path = os.path.join(directory, filename)
|
112 |
+
if file_path.endswith(".mp3"):
|
113 |
+
file_age = now - os.path.getmtime(file_path)
|
114 |
+
if file_age > max_age_seconds:
|
115 |
+
os.remove(file_path)
|
116 |
|
117 |
def combine_audio_files(self, filenames: List[str], output_file: str) -> None:
|
118 |
+
if not filenames:
|
119 |
+
raise ValueError("No input files provided")
|
120 |
+
try:
|
121 |
+
combined = AudioSegment.empty()
|
122 |
+
for filename in filenames:
|
123 |
+
combined += AudioSegment.from_file(filename, format="mp3")
|
124 |
+
combined.export(output_file, format="mp3")
|
125 |
+
except Exception as e:
|
126 |
+
raise RuntimeError(f"Failed to combine audio files: {e}")
|
127 |
|
128 |
async def url_to_audio(self, url: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
|
129 |
+
self.clean_old_files()
|
130 |
text = self.fetch_text(url)
|
131 |
words = text.split()
|
132 |
if len(words) > self.config.max_words:
|
133 |
+
text = " ".join(words[:self.config.max_words])
|
134 |
conversation_json = self.extract_conversation(text)
|
135 |
+
conversation_text = "\n".join(
|
136 |
+
f"{turn['speaker']}: {turn['text']}" for turn in conversation_json["conversation"]
|
137 |
+
)
|
138 |
self.llm_out = conversation_json
|
139 |
+
audio_files, output_dir = await self.text_to_speech(conversation_json, voice_1, voice_2)
|
140 |
+
output_file = os.path.join(output_dir, f"combined_{int(time.time())}.mp3")
|
141 |
+
self.combine_audio_files(audio_files, output_file)
|
142 |
+
return output_file, conversation_text
|
143 |
|
144 |
async def text_to_audio(self, text: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
|
145 |
+
self.clean_old_files()
|
146 |
conversation_json = self.extract_conversation(text)
|
147 |
+
conversation_text = "\n".join(
|
148 |
+
f"{turn['speaker']}: {turn['text']}" for turn in conversation_json["conversation"]
|
149 |
+
)
|
150 |
+
audio_files, output_dir = await self.text_to_speech(conversation_json, voice_1, voice_2)
|
151 |
+
output_file = os.path.join(output_dir, f"combined_{int(time.time())}.mp3")
|
152 |
+
self.combine_audio_files(audio_files, output_file)
|
153 |
+
return output_file, conversation_text
|
154 |
|
155 |
async def raw_text_to_audio(self, text: str, voice_1: str, voice_2: str) -> Tuple[str, str]:
|
156 |
+
self.clean_old_files()
|
157 |
+
conversation = {
|
158 |
+
"conversation": [
|
159 |
+
{"speaker": "Host", "text": text},
|
160 |
+
{"speaker": "Co-host", "text": "(Continuación del tema)"}
|
161 |
+
]
|
162 |
+
}
|
163 |
+
audio_files, output_dir = await self.text_to_speech(conversation, voice_1, voice_2)
|
164 |
+
output_file = os.path.join(output_dir, f"raw_podcast_{int(time.time())}.mp3")
|
165 |
+
self.combine_audio_files(audio_files, output_file)
|
166 |
+
return text, output_file
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|