Wrapper for JavaScript's ArrayBuffer/DataView maintaining index and default endianness. Supports arbitrary reading/writing, implicit growth, clipping, cloning and reversing as well as UTF-8 characters and NULL-terminated C-strings.
ByteBuffer is available via npm:
npm install byte-bufferOr for usage in the browser:
dist/byte-buffer.jsdist/byte-buffer.min.js
As an ECMAScript module:
import ByteBuffer from 'byte-buffer';
const b = new ByteBuffer();In CommonJS environments:
const ByteBuffer = require('byte-buffer');
const b = new ByteBuffer();Available in the global scope when included in browser environments:
const b = new ByteBuffer();ByteBuffer's API borrows heavily from Adobe's IDataInput and IDataOutput as well as David Flanagan's BufferView.
The concept of separate buffers and views - as outlined in MDN's JavaScript typed arrays - is not used. ByteBuffer handles this separation for you.
Use the following constants to indicate endianness:
ByteBuffer.BIG_ENDIAN
ByteBuffer.LITTLE_ENDIANnew ByteBuffer(1) // Buffer of one byte with big-endian byte order
new ByteBuffer(1, ByteBuffer.LITTLE_ENDIAN) // Little-endian byte order insteadByteBuffers may also be constructed from other byte-aware sources:
new ByteBuffer(new ArrayBuffer(2))
new ByteBuffer(new Uint8Array(3))
new ByteBuffer(new DataView(new ArrayBuffer(4)))
new ByteBuffer(new ByteBuffer(5))Or from generic sequences:
new ByteBuffer([0, 1, 2, 3])After construction a ByteBuffer's read/write index is always at the front of the
buffer. Hereafter b is assumed to be an instance of ByteBuffer.
b.buffer // Reference to internal ArrayBuffer
b.buffer = new ArrayBuffer(3) // Sets new bufferb.raw // Reference to raw buffer (read-only)b.view // Reference to internal DataView (read-only)b.length // Number of bytes in the buffer (read-only)
b.byteLength // Aliasb.order // Buffer's current default byte order
b.order = ByteBuffer.BIG_ENDIAN // Sets byte orderb.available // Number of available bytes (read-only)ByteBuffer maintains a read/write index to simplify usage.
b.index // Current read/write index
b.index = 4 // Sets indexIf the index is out of bounds, a RangeError will be thrown.
b.front() // Sets index to front of the bufferb.end() // Sets index to end of the bufferb.seek(10) // Forwards ten bytes
b.seek(-2) // Backs two bytesThese methods may be chained:
b.front().seek(2)All read methods default to the ByteBuffer's byte order if not given.
b.readByte()b.readUnsignedByte()b.readShort() // Buffer's default byte order
b.readShort(ByteBuffer.LITTLE_ENDIAN) // Explicit byte orderb.readUnsignedShort()b.readInt()b.readUnsignedInt()b.readFloat()b.readDouble()b.read(6) // Reads 6 bytes
b.read() // Reads all remaining bytesb.readString(5) // Reads 5 bytes as a string
b.readString() // Reads all remaining bytes as a string
b.readUTFChars() // Aliasb.readCString() // Reads string up to NULL-byte or end of bufferAll write methods default to the ByteBuffer's byte order if not given.
b.writeByte(10)b.writeUnsignedByte(-10)b.writeShort(-2048)
b.writeShort(-2048, ByteBuffer.LITTLE_ENDIAN) // Explicit byte orderb.writeUnsignedShort(4096)b.writeInt(-524288)b.writeUnsignedInt(1048576)b.writeFloat(13.37)b.writeDouble(1048576.89)b.write([1, 2, 3])
b.write(new ArrayBuffer(2))
b.write(new Uint8Array(3))
b.write(new ByteBuffer(5))Additionally, all the above write methods may be chained:
b.writeShort(0x2020).write([1, 2, 3])The following string related methods do not return the buffer itself, but rather provide the number of bytes that were written to it. More on this under implicit growth strategy a bit further down.
b.writeString('ByteBuffer') // Writes given string and returns number of bytes
b.writeUTFChars('ByteBuffer') // Aliasb.writeCString('ByteBuffer') // Writes given string and returns number of bytes (including NULL-byte)The buffer may be grown at the front or at the end. When prepending, the buffer's index is adjusted accordingly.
b.prepend(2) // Prepends given number of bytesb.append(2) // Appends given number of bytesThis feature allows a ByteBuffer to grow implicitly when writing arbitrary data. Since every implicit growth requires the buffer to be rebuilt from scratch, care must be taken when using this feature. Writing low byte-length pieces of data in rapid succession is not recommended.
To protect the unaware from harm, this feature needs to be explicitly enabled:
b = new ByteBuffer(2, ByteBuffer.BIG_ENDIAN, true) // Last argument indicates implicit growth strategy
b.writeUnsignedInt(2345102) // Implicitly makes room for 4 bytes - by growing with 2 - prior to writingThe implicit growth strategy can also be enabled and disabled after construction:
b.implicitGrowth = true/falseImplicit growth is a must when dealing with UTF-8 encoded strings, as dealing with arbitrary user data - e.g. names or addresses - may include various characters that require to be encoded in multiple bytes, which would be relatively verbose to calculate beforehand.
The buffer may be truncated at the front, end or both. Both arguments are
optional and may be negative in which case the offsets are calculated from the
respective boundaries of the buffer. The begin-argument defaults to the current
index, allowing efficient clipping in various scenarios, e.g. when used in
combination with network sockets to shift off read data. The end-argument
defaults to the end of the buffer.
b.clip(2, -2)
b.clip(-2, 4)b.slice(2, 4) // Independent clone of given slice of the bufferb.clone() // Independent clone of the entire bufferb.reverse() // Reverses buffer in placeb.toArray() // Changes to this array are not backedb.toHex() // Hexadecimal representation of this buffer, e.g: 42 79 74 65 42 75 66 66 65 72b.toASCII() // ASCII representation of this buffer, e.g: B y t e B u f f e rByteBuffer is written in ES2015+, modularized using ECMAScript Modules, compiled by Babel, bundled with rollup and tested through Jest.
Getting this toolchain up and running, is easy and straight-forward:
-
Get the code:
git clone git://github.com/timkurvers/byte-buffer.git
-
Download and install Node.js – including
npm– for your platform. -
Install dependencies:
npm install
-
Run
npm test:watchwhich will run tests when source files change.
When contributing, please:
- Fork the repository
- Accompany each logical unit of operation with at least one test
- Open a pull request
- Do not include any distribution files (such as
dist/byte-buffer.js)
Christopher Chedeau's jDataView
- Maintains read-index and supports seeking
- Various string/char utilities (may support UTF-8)
- Does not support writing values
- Does not support NULL-terminated C-strings
- Does not support growing, clipping, cloning and reversing
- Supports a wide range of browsers/setups
David Flanagan's BufferView
- Supports reading/writing values
- Maintains index and supports seeking
- Supports UTF-8 characters
- Does not support NULL-terminated C-strings
- Does not support growing, clipping, cloning and reversing as view and buffer are immutable
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot]")
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent