init: scripts diversos (crawlers, conversores, scrapers)

tts-generator/gemini_tts_teste.py

@@ -0,0 +1,141 @@
+"""
+gemini_tts_teste.py
+
+Author: Descomplicar® Crescimento Digital
+Link: https://descomplicar.pt
+Copyright: 2025 Descomplicar®
+"""
+
+# To run this code you need to install the following dependencies:
+# pip install google-genai
+
+import base64
+import mimetypes
+import os
+import re
+import struct
+from google import genai
+from google.genai import types
+
+
+def save_binary_file(file_name, data):
+    f = open(file_name, "wb")
+    f.write(data)
+    f.close()
+    print(f"File saved to to: {file_name}")
+
+
+def generate_test():
+    client = genai.Client(
+        api_key=os.environ.get("GEMINI_API_KEY"),
+    )
+
+    model = "gemini-2.5-pro-preview-tts"
+    contents = [
+        types.Content(
+            role="user",
+            parts=[
+                types.Part.from_text(text="""Bem-vindo à Descomplicar, a agência de aceleração digital que transforma a sua presença online numa máquina de crescimento. 
+
+Somos especialistas em Marketing Digital, criação de websites profissionais e estratégias que geram resultados reais para o seu negócio.
+
+Na Descomplicar, a nossa filosofia é simples: tornar o complexo mais simples. Descomplicamos o marketing digital para que você se possa focar no que faz melhor - gerir o seu negócio.
+
+Marque uma reunião connosco e descubra como podemos acelerar o crescimento digital da sua empresa."""),
+            ],
+        ),
+    ]
+    generate_content_config = types.GenerateContentConfig(
+        temperature=1.1,
+        response_modalities=[
+            "audio",
+        ],
+        speech_config=types.SpeechConfig(
+            voice_config=types.VoiceConfig(
+                prebuilt_voice_config=types.PrebuiltVoiceConfig(
+                    voice_name="Autonoe"
+                )
+            )
+        ),
+    )
+
+    file_index = 0
+    for chunk in client.models.generate_content_stream(
+        model=model,
+        contents=contents,
+        config=generate_content_config,
+    ):
+        if (
+            chunk.candidates is None
+            or chunk.candidates[0].content is None
+            or chunk.candidates[0].content.parts is None
+        ):
+            continue
+        if chunk.candidates[0].content.parts[0].inline_data and chunk.candidates[0].content.parts[0].inline_data.data:
+            file_name = f"descomplicar_teste_audio_{file_index}"
+            file_index += 1
+            inline_data = chunk.candidates[0].content.parts[0].inline_data
+            data_buffer = inline_data.data
+            file_extension = mimetypes.guess_extension(inline_data.mime_type)
+            if file_extension is None:
+                file_extension = ".wav"
+                data_buffer = convert_to_wav(inline_data.data, inline_data.mime_type)
+            save_binary_file(f"{file_name}{file_extension}", data_buffer)
+        else:
+            print(chunk.text)
+
+def convert_to_wav(audio_data: bytes, mime_type: str) -> bytes:
+    """Generates a WAV file header for the given audio data and parameters."""
+    parameters = parse_audio_mime_type(mime_type)
+    bits_per_sample = parameters["bits_per_sample"]
+    sample_rate = parameters["rate"]
+    num_channels = 1
+    data_size = len(audio_data)
+    bytes_per_sample = bits_per_sample // 8
+    block_align = num_channels * bytes_per_sample
+    byte_rate = sample_rate * block_align
+    chunk_size = 36 + data_size
+
+    header = struct.pack(
+        "<4sI4s4sIHHIIHH4sI",
+        b"RIFF",          # ChunkID
+        chunk_size,       # ChunkSize (total file size - 8 bytes)
+        b"WAVE",          # Format
+        b"fmt ",          # Subchunk1ID
+        16,               # Subchunk1Size (16 for PCM)
+        1,                # AudioFormat (1 for PCM)
+        num_channels,     # NumChannels
+        sample_rate,      # SampleRate
+        byte_rate,        # ByteRate
+        block_align,      # BlockAlign
+        bits_per_sample,  # BitsPerSample
+        b"data",          # Subchunk2ID
+        data_size         # Subchunk2Size (size of audio data)
+    )
+    return header + audio_data
+
+def parse_audio_mime_type(mime_type: str) -> dict[str, int | None]:
+    """Parses bits per sample and rate from an audio MIME type string."""
+    bits_per_sample = 16
+    rate = 24000
+
+    parts = mime_type.split(";")
+    for param in parts:
+        param = param.strip()
+        if param.lower().startswith("rate="):
+            try:
+                rate_str = param.split("=", 1)[1]
+                rate = int(rate_str)
+            except (ValueError, IndexError):
+                pass
+        elif param.startswith("audio/L"):
+            try:
+                bits_per_sample = int(param.split("L", 1)[1])
+            except (ValueError, IndexError):
+                pass
+
+    return {"bits_per_sample": bits_per_sample, "rate": rate}
+
+
+if __name__ == "__main__":
+    generate_test()