nbmediasplit

nbmediasplit is a script to extract base64 encoded image and audio PCM embedded in .ipynb file and save them into specified directories. nbmediasplit also converts ipynb file to a new one which refers to stored image and audio files.

install

pip install nbmediasplit

usage

extract image files from ipynb

nbmediasplit input.ipynb -i image_out_dir

or

nbmediasplit input.ipynb --imgdir image_out_dir

The above command extract image files from input.ipynb and store them to image_out_dir. -i or --imgdir specifies a directory to store image files. Filenames of the stored image are numbered in sequential order(0.png, ...).

extract audio files from ipynb

nbmediasplit input.ipynb -w wav_out_dir

or

nbmediasplit input.ipynb --wavdir wav_out_dir

The above command extract audio files from input.ipynb and store them to wav_out_dir. -w or --wavdir specifies a directory to store audio files. Filenames of the stored audio are numbered in sequential order(0.wav, ...).

extract image and audio files from ipynb

nbmediasplit input.ipynb -i image_out_dir -w wav_out_dir

or

nbmediasplit input.ipynb --imgdir image_out_dir --wavdir wav_out_dir

The above command does below things.

• extract image files from input.ipynb and store them to image_out_dir
• extract audio files from input.ipynb and store them to wav_out_dir.

-i or --imgdir specifies a directory to store image files. -w or --wavdir specifies a directory to store audio files. Filenames of the stored image are numbered in sequential order(0.png, ...). Filenames of the stored audio are numbered in sequential order(0.wav, ...).

extract image and audio files from ipynb and convert ipynb

If you use -o or --output option like below command, you can convert input.ipynb to new ipynb file which refers to stored image files and audio files directly.

nbmediasplit input.ipynb -i image_out_dir -w wav_out_dir -o converted.ipynb

or

nbmediasplit input.ipynb --imgdir image_out_dir --wavdir wav_out_dir --output converted.ipynb

The above command extract image files and audio files, and store them to specified directories, and generate new ipynb file converted.ipynb. converted.ipynb includes same content as input.ipynb, but base64 encoded image and audio data are replaced to HTML tag refers to stored files directly like below.

• image tag
  • <img src="${image_out_dir}/${n}.png" />
• audio tag
  • <audio controls preload="none"><source src="${wav_out_dir}/${n}.wav" type="audio/wav" /></audio>

Also, you can use --img-prefix and --wav-prefix options. These options can change the path embeded in src attribute of output html like below(actual files are stored image_out_dir and wav_out_dir).

• image tag
  • <img src="${img-prefix}/${n}.png" />
• audio tag
  • <audio controls preload="none"><source src="${wav-prefix}/${n}.wav" type="audio/wav" /></audio>

show help

nbmediasplit --help

note

Unless you trust the notebook converted by nbmediasplit in jupyter, you can't load audio source because of html sanitization. To trust notebook in jupyterlab, go to command pallet in the left sidebar(on osx, type shift+cmd+c) and execute trust notebook, then you'll load audio source if the source path is correct.