npm install --save pixbincodec
The PixBin format is a simple way to serialize Javascript/JSON objects as well as low-level buffers into a single binary buffer file you can save on your computer. Originally, it was created so that Pixpipejs could save a piece of data and that could later be reinjected into another pipeline for further processing. In the context of Pixpipejs, this piece of data could be an Image2D
, an Image3D
, a LineString
, etc. In the end, it's just Javascript/JSON Objects and low level buffers!
The first use-case is in Pixpipejs/Javascript but the PixBin format can be created and decoded with other languages. For, example we also have Python codec.
That's right. At first, we wanted to use one of the NIfTI or MINC formats to encode Pixpipe outputs, but the browser-side Javascript context implied to write a NIfTI or MINC encoder from scratch, was cumbersome. Though MINC can store extensive metadata, multimodality data, and supports internal compression, it is built upon HDF5 making it difficult to reliably interact with in the web, and NIfTI does not have these desirable features.
To learn more about the PixBin format, read this in-depth description.
- Optimized for numerical data
- Still a generic store-what-you-want format
- Is multimodality, a bit like an archive, so you can store multiple blocks inside
- Can handle as many metadata as you need, per block and at the parent level
- Data of each block are zlib compressed (lossless) - optional
- Streamable over blocks, since each block of data is compressed independently
- Perform checksum validation on each block to guaranty data integrity - optional
- Provide an easy-to-read header with an index of all the blocks (without having to read/decode those blocks)
- Is a binary format
- Easy to write a parser for
See the examples
directory for the source, or:
- A short example of coding - decoding
- A more extensive example of coding - creating a file - decoding
In order to be serialize into the PixBin format, a JS object must contain:
- a
_data
attribute, no matter what it contains - a
_metadata
attribute, no matter what it contains
Notice: since this format is intended to encode numerical data like pixel arrays or position arrays, the encoding the _data
object will be optimized in the following cases:
_data
is a typed array. There is a single stream to encode (case 1)._data
is anArray
of typed arrays. There are several streams to encode (case 2).
In those cases, the data will directly be encoded as low level types rather than being serialized into a more descriptive language. This will also result in smaller files.
The third case, if you chose that _data
is an Object
(or {}
), then it will be serialized (see Object serialization). There is a single stream to encode (case 3).