README

cpp-struct-js

The intention of this module is to interace node.js with an Arduino software I am writing. For this I need to read binary structures that are generated by C++ code running on an Arduino system.

Supported types:

uint8,16,32
int8,16,32
float
strings

Only one dimensional arrays are supported. It might be possible to extend the code to support this, but I don't see a need for this right now.

Pointers are not supported, but that wouldn't make sense anyway since serialization can't deal with native memory adresses.

Example


struct = require("cpp-struct")

var player = new struct("Player", [
    "name", struct.char(12),
    "id", struct.uint32_t()
]);

var record = new struct("Record", [
    "playerIndex", struct.uint8_t(),
    "typeId", struct.uint8_t(),
    "record", struct.uint16_t(31)
]);

var EEPROMData = new struct("EEPROMData", [
    "gameName", struct.char(14),
    "version", struct.uint16_t(),
    "players", struct.type(player,32),
    "records", struct.type(record,32)
]);

var buffer = new Buffer(EEPROMData.size());
EEPROMData.encode(buffer,0, {
    gameName: "SuperTesting!!!",
    version: 1,
    players:[
        {name:"Hello",id:1},
        {id:2},
        {name:"HUHU"}
    ]
},{endian:"LE"})

Reference

Class: struct (var struct = require('cpp-struct'))

Constructor signature: name, schema, [count, [bytes]]

name: Internal name that's used when exporting and referencing other structs.
schema: Interleaved array where even elements are variable name identifiers and odd elements are struct instances
count (optional): If the type is supposed to be an array, this is the number of elements in the array
bytes (optional): Size of a single element in bytes (total size = count * bytes)

Note: count and bytes arguments are used usually only internally - you don't have to worry about these if you don't intend to add new native types next to the currently implemented ones such as int/uint/float/double - say you'd want a int24, you'll need to care about these arguments.

Instance member: struct.setEncoder

Function signature: func

func (buffer,pos,data,opt): Function that's called instead of the default encoder -- buffer: Buffer object to write to -- pos: position to write to -- data: the data value to be written (can be undefined) -- opt: optional options object that specifies encoding details (such as endianess)

Instance member: struct.setDecoder

Why not using one of the other modules?

I want to create schemas in node.js and use those for reading and writing binary buffers. Next to that I also want to output the header files for CPP which I can simply compile and use in my project.

The schema defines exact sizes, not the data - which may be incomplete and is to be filled with default data (zeros).

The other modules I looked at either focused on packing data as binary efficient or packed data directly without schema.

The closest match I could find was buffer-layout.

Usage no npm install needed!