Skip to content

Latest commit

 

History

History
49 lines (28 loc) · 1.63 KB

README.md

File metadata and controls

49 lines (28 loc) · 1.63 KB

Wave Chunk Parser

Parses wave files chunks in order to extract metadata and audio. Also provides an option to write a bunch of given chunks back into a coherent file.

Installation

pip install wave-chunk-parser

Reading a file / IO stream

from wave_chunk_parser.chunks import RiffChunk

with open("file.wav", "rb") as file:
    riff_chunk = RiffChunk.from_file(file)

From there you can access the sub chunks from riff_chunk.sub_chunks. The data chunk uses a numpy array to hold the vectors of audio samples.

Format (fmt) and data chunks are critical. Cart chunk is optional but provides those markers we need for handling broadcast audio.

Writing a file / IO stream

You will need to build the chunks individually and supply them in a list of:

chunks = [FormatChunk, DataChunk, CartChunk]

The format chunk must come before the data chunk (we need to know how to en/decode it). Cart chunk can appear anywhere in the list or even not exist at all.

To get a blob for writing to disk (or wherever), simply:

riff_chunk = RiffChunk(chunks)
blob = riff_chunk.to_bytes()

The to_bytes function exists on every chunk type. So, if you particularly desire, you can turn a standalone format chunk into a blob.

Numpy on macOS

We need openblas to make Numpy work on macOS:

pip uninstall numpy
brew install openblas
OPENBLAS="$(brew --prefix openblas)" pip install numpy --no-cache-dir

Building the package for deployment

To build the package for deployment:

poetry run python setup.py bdist_wheel
poetry run twine upload --repository testpypi dist/*
poetry run twine upload --repository pypi dist/*