This commit adds a new script called `wiki_downloader.py` which allows downloading images from Wikipedia. The script uses Selenium and undetected_chromedriver to automate the process of searching for an image on Google and downloading it.
The `download_image` function takes two parameters: `query` and `download_path`. It opens a headless Chrome browser using undetected_chromedriver and navigates to the Google Images search page with the specified query. After a short delay, it clicks on the "tout refuser" button to accept the cookies prompt. Then, it finds the first image in the search results and extracts its source URL. The image is then downloaded either by decoding a base64-encoded image or by making a request to the image URL. The downloaded image is saved in the specified `download_path`.
The script also includes a test case that demonstrates how to use the `download_image
feat(video.py): use wav format instead of mp3 for generated audio files
feat(montage.py): use Bark TTS instead of 🐸TTS
feat(speak.py): add support for Bark TTS
fix(speak.py): remove unused 🐸TTS import and variable
fix(main.py): fix asyncio.run() call placement
docs: update requirements.txt with new dependencies
fix(script.py): add prompt parameter to generate_script function
feat(main.py): add loop to run main function indefinitely and add option to quit with ctrl+c
fix(main.py): print video object instead of video url
feat(uploader.py): add retry mechanism to handle token expiration and refresh token if necessary. Also, pass youtube object to upload_thumbnail function.
refactor(montage.py): use os.path.join instead of string concatenation
refactor(thumbnail.py): use os.path.join instead of string concatenation
refactor(uploader.py): use os.path.join instead of string concatenation
refactor(uploader.py): add support for client_secret.json file with different names in credentials folder
docs(readme.md): update instructions for openai and unsplash keys
docs(readme.md): update instructions for google oauth client id json files
docs(readme.md): add information about channel.yaml file in each channel's folder
feat(generators/montage.py): add check to skip slide if it already exists
feat(generators/montage.py): add support for DEEPL_ACCESS_KEY and UNSPLASH_ACCESS_KEY environment variables
feat(generators/speak.py): add support for Johanne voice
feat(generators/speak.py): add emotion parameter to generate_voice function
feat(generators/uploader.py): add success message and authorization prompt message to run_local_server method
fix(main.py): check if credits is None before writing to meta.txt file
feat(prompts/marp.md): change theme to gaia and add lead and invert classes
refactor(montage.py): remove unused srt_fr variable
refactor(montage.py): remove saving of french subtitle file
refactor(main.py): remove unused translate import and function call
