Reimplementation of keichi/binary-parser, supporting both streams and serialization.
The API is mainly the same as binary-parser's, with a couple of additions and one removed method.
Due to undefined behavior, .skip is not supported.
As a workaround the .buffer method can be used.
The .stream method returns a transform stream that reads incoming data and emits parser results.
The parser instance is looped until the end of the input is reached.
As a counterpart to binary-parser's .parse method, .serialize takes an object and returns a buffer representation of it.
Optionally, an already allowcated buffer can be passed as a second argument.
.fixedSizeNest is a variation of nest that ensures the nested parser reads a specific amount of bytes.
Fixed-size nested parsers are the only source of ambiguity remaining while serializing, as missing bytes are simply skipped.
And error is thrown during serialization if the nested parser attempts to write more bytes than the size permits.
Properties with a specified formatting function need to also provide a deformatter function that restores the original value.
As a very simple extension, the flatten option allows nested parsers to write to the current object.
Its motivation is the aim for a flat output structure after .choice calls.
A relaxation from the 32 bit limit of binary-parser.
Also, a bug of the original project that leads to the bit32 method always parsing a value of 0 is not present.
License: MIT