|
|
import gradio as gr |
|
|
import requests |
|
|
from bs4 import BeautifulSoup |
|
|
import os |
|
|
import zipfile |
|
|
import tempfile |
|
|
from urllib.parse import urljoin |
|
|
|
|
|
def process_url(url): |
|
|
try: |
|
|
response = requests.get(url) |
|
|
response.raise_for_status() |
|
|
except requests.RequestException as e: |
|
|
return None, f"Erreur lors de la récupération de la page : {e}" |
|
|
|
|
|
soup = BeautifulSoup(response.text, 'html.parser') |
|
|
|
|
|
mp3_links = [] |
|
|
for link in soup.find_all('a', href=True): |
|
|
href = link['href'] |
|
|
if href.lower().endswith('.mp3'): |
|
|
absolute_url = urljoin(response.url, href) |
|
|
mp3_links.append(absolute_url) |
|
|
|
|
|
|
|
|
seen = set() |
|
|
mp3_links = [x for x in mp3_links if not (x in seen or seen.add(x))] |
|
|
|
|
|
if not mp3_links: |
|
|
return None, "Aucun lien MP3 trouvé sur la page." |
|
|
|
|
|
temp_dir = tempfile.mkdtemp() |
|
|
filenames = [] |
|
|
for idx, mp3_url in enumerate(mp3_links, start=1): |
|
|
try: |
|
|
mp3_response = requests.get(mp3_url) |
|
|
mp3_response.raise_for_status() |
|
|
filename = os.path.join(temp_dir, f"{idx:02d}_{os.path.basename(mp3_url)}") |
|
|
with open(filename, 'wb') as f: |
|
|
f.write(mp3_response.content) |
|
|
filenames.append(filename) |
|
|
except requests.RequestException as e: |
|
|
print(f"Erreur de téléchargement {mp3_url}: {e}") |
|
|
continue |
|
|
|
|
|
if not filenames: |
|
|
return None, "Aucun épisode téléchargé." |
|
|
|
|
|
zip_filename = os.path.join(temp_dir, 'podcast_episodes.zip') |
|
|
with zipfile.ZipFile(zip_filename, 'w') as zipf: |
|
|
for file in filenames: |
|
|
zipf.write(file, arcname=os.path.basename(file)) |
|
|
|
|
|
return zip_filename, None |
|
|
|
|
|
def download_podcast(url): |
|
|
zip_path, error = process_url(url) |
|
|
if error: |
|
|
raise gr.Error(error) |
|
|
return zip_path |
|
|
|
|
|
iface = gr.Interface( |
|
|
fn=download_podcast, |
|
|
inputs=gr.Textbox(label="URL de la page du podcast", placeholder="https://www.radiofrance.fr/..."), |
|
|
outputs=gr.File(label="Télécharger le ZIP des épisodes"), |
|
|
title="Téléchargeur de Podcast", |
|
|
description="Entrez l'URL d'une page contenant des épisodes de podcast pour télécharger tous les MP3 dans un ZIP ordonné." |
|
|
) |
|
|
|
|
|
iface.launch() |
