0
0
mirror of https://github.com/nodejs/node.git synced 2024-11-21 21:19:50 +01:00
nodejs/lib/buffer.js

1370 lines
39 KiB
JavaScript
Raw Permalink Normal View History

// Copyright Joyent, Inc. and other Node contributors.
//
// Permission is hereby granted, free of charge, to any person obtaining a
// copy of this software and associated documentation files (the
// "Software"), to deal in the Software without restriction, including
// without limitation the rights to use, copy, modify, merge, publish,
// distribute, sublicense, and/or sell copies of the Software, and to permit
// persons to whom the Software is furnished to do so, subject to the
// following conditions:
//
// The above copyright notice and this permission notice shall be included
// in all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
// OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
// NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
// DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
// OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
// USE OR OTHER DEALINGS IN THE SOFTWARE.
'use strict';
const {
Array,
ArrayBufferIsView,
ArrayIsArray,
ArrayPrototypeForEach,
MathFloor,
MathMin,
MathTrunc,
NumberIsInteger,
NumberIsNaN,
NumberMAX_SAFE_INTEGER,
NumberMIN_SAFE_INTEGER,
ObjectDefineProperties,
ObjectDefineProperty,
ObjectSetPrototypeOf,
RegExpPrototypeSymbolReplace,
StringPrototypeCharCodeAt,
StringPrototypeSlice,
StringPrototypeToLowerCase,
StringPrototypeTrim,
SymbolSpecies,
SymbolToPrimitive,
TypedArrayPrototypeFill,
TypedArrayPrototypeGetBuffer,
TypedArrayPrototypeGetByteLength,
TypedArrayPrototypeGetByteOffset,
TypedArrayPrototypeGetLength,
TypedArrayPrototypeSet,
TypedArrayPrototypeSlice,
Uint8Array,
Uint8ArrayPrototype,
} = primordials;
const {
byteLengthUtf8,
compare: _compare,
compareOffset,
copy: _copy,
fill: bindingFill,
isAscii: bindingIsAscii,
isUtf8: bindingIsUtf8,
indexOfBuffer,
indexOfNumber,
indexOfString,
swap16: _swap16,
swap32: _swap32,
swap64: _swap64,
kMaxLength,
kStringMaxLength,
atob: _atob,
btoa: _btoa,
} = internalBinding('buffer');
const {
constants: {
ALL_PROPERTIES,
ONLY_ENUMERABLE,
},
getOwnNonIndexProperties,
isInsideNodeModules,
} = internalBinding('util');
const {
customInspectSymbol,
lazyDOMException,
normalizeEncoding,
kIsEncodingSymbol,
defineLazyProperties,
encodingsMap,
deprecate,
} = require('internal/util');
const {
isAnyArrayBuffer,
isArrayBufferView,
isUint8Array,
isTypedArray,
} = require('internal/util/types');
const {
inspect: utilInspect,
} = require('internal/util/inspect');
const {
codes: {
ERR_BUFFER_OUT_OF_BOUNDS,
ERR_INVALID_ARG_TYPE,
ERR_INVALID_ARG_VALUE,
ERR_INVALID_BUFFER_SIZE,
ERR_MISSING_ARGS,
ERR_OUT_OF_RANGE,
ERR_UNKNOWN_ENCODING,
},
genericNodeError,
} = require('internal/errors');
const {
validateArray,
validateBuffer,
validateInteger,
validateNumber,
validateString,
} = require('internal/validators');
// Provide validateInteger() but with kMaxLength as the default maximum value.
const validateOffset = (value, name, min = 0, max = kMaxLength) =>
validateInteger(value, name, min, max);
const {
FastBuffer,
markAsUntransferable,
addBufferPrototypeMethods,
createUnsafeBuffer,
} = require('internal/buffer');
FastBuffer.prototype.constructor = Buffer;
Buffer.prototype = FastBuffer.prototype;
addBufferPrototypeMethods(Buffer.prototype);
const constants = ObjectDefineProperties({}, {
MAX_LENGTH: {
__proto__: null,
value: kMaxLength,
writable: false,
enumerable: true,
},
MAX_STRING_LENGTH: {
__proto__: null,
value: kStringMaxLength,
writable: false,
enumerable: true,
},
});
Buffer.poolSize = 8 * 1024;
let poolSize, poolOffset, allocPool, allocBuffer;
function createPool() {
poolSize = Buffer.poolSize;
allocBuffer = createUnsafeBuffer(poolSize);
allocPool = allocBuffer.buffer;
markAsUntransferable(allocPool);
poolOffset = 0;
}
createPool();
function alignPool() {
// Ensure aligned slices
if (poolOffset & 0x7) {
poolOffset |= 0x7;
poolOffset++;
}
}
let bufferWarningAlreadyEmitted = false;
let nodeModulesCheckCounter = 0;
const bufferWarning = 'Buffer() is deprecated due to security and usability ' +
'issues. Please use the Buffer.alloc(), ' +
'Buffer.allocUnsafe(), or Buffer.from() methods instead.';
function showFlaggedDeprecation() {
if (bufferWarningAlreadyEmitted ||
++nodeModulesCheckCounter > 10000 ||
(!require('internal/options').getOptionValue('--pending-deprecation') &&
isInsideNodeModules(100, true))) {
// We don't emit a warning, because we either:
// - Already did so, or
// - Already checked too many times whether a call is coming
// from node_modules and want to stop slowing down things, or
// - We aren't running with `--pending-deprecation` enabled,
// and the code is inside `node_modules`.
// - We found node_modules in up to the topmost 100 frames, or
// there are more than 100 frames and we don't want to search anymore.
return;
}
process.emitWarning(bufferWarning, 'DeprecationWarning', 'DEP0005');
bufferWarningAlreadyEmitted = true;
}
function toInteger(n, defaultVal) {
n = +n;
if (!NumberIsNaN(n) &&
n >= NumberMIN_SAFE_INTEGER &&
n <= NumberMAX_SAFE_INTEGER) {
return ((n % 1) === 0 ? n : MathFloor(n));
}
return defaultVal;
}
function copyImpl(source, target, targetStart, sourceStart, sourceEnd) {
if (!ArrayBufferIsView(source))
throw new ERR_INVALID_ARG_TYPE('source', ['Buffer', 'Uint8Array'], source);
if (!ArrayBufferIsView(target))
throw new ERR_INVALID_ARG_TYPE('target', ['Buffer', 'Uint8Array'], target);
if (targetStart === undefined) {
targetStart = 0;
} else {
targetStart = NumberIsInteger(targetStart) ? targetStart : toInteger(targetStart, 0);
if (targetStart < 0)
throw new ERR_OUT_OF_RANGE('targetStart', '>= 0', targetStart);
}
if (sourceStart === undefined) {
sourceStart = 0;
} else {
sourceStart = NumberIsInteger(sourceStart) ? sourceStart : toInteger(sourceStart, 0);
if (sourceStart < 0 || sourceStart > source.byteLength)
throw new ERR_OUT_OF_RANGE('sourceStart', `>= 0 && <= ${source.byteLength}`, sourceStart);
}
if (sourceEnd === undefined) {
sourceEnd = source.byteLength;
} else {
sourceEnd = NumberIsInteger(sourceEnd) ? sourceEnd : toInteger(sourceEnd, 0);
if (sourceEnd < 0)
throw new ERR_OUT_OF_RANGE('sourceEnd', '>= 0', sourceEnd);
}
if (targetStart >= target.byteLength || sourceStart >= sourceEnd)
return 0;
return _copyActual(source, target, targetStart, sourceStart, sourceEnd);
}
function _copyActual(source, target, targetStart, sourceStart, sourceEnd) {
if (sourceEnd - sourceStart > target.byteLength - targetStart)
sourceEnd = sourceStart + target.byteLength - targetStart;
let nb = sourceEnd - sourceStart;
const sourceLen = source.byteLength - sourceStart;
if (nb > sourceLen)
nb = sourceLen;
if (nb <= 0)
return 0;
_copy(source, target, targetStart, sourceStart, nb);
return nb;
}
/**
* The Buffer() constructor is deprecated in documentation and should not be
* used moving forward. Rather, developers should use one of the three new
* factory APIs: Buffer.from(), Buffer.allocUnsafe() or Buffer.alloc() based on
* their specific needs. There is no runtime deprecation because of the extent
* to which the Buffer constructor is used in the ecosystem currently -- a
* runtime deprecation would introduce too much breakage at this time. It's not
* likely that the Buffer constructors would ever actually be removed.
* Deprecation Code: DEP0005
*/
function Buffer(arg, encodingOrOffset, length) {
showFlaggedDeprecation();
// Common case.
if (typeof arg === 'number') {
if (typeof encodingOrOffset === 'string') {
throw new ERR_INVALID_ARG_TYPE('string', 'string', arg);
}
return Buffer.alloc(arg);
}
return Buffer.from(arg, encodingOrOffset, length);
}
ObjectDefineProperty(Buffer, SymbolSpecies, {
__proto__: null,
enumerable: false,
configurable: true,
get() { return FastBuffer; },
});
/**
* Functionally equivalent to Buffer(arg, encoding) but throws a TypeError
* if value is a number.
* Buffer.from(str[, encoding])
* Buffer.from(array)
* Buffer.from(buffer)
* Buffer.from(arrayBuffer[, byteOffset[, length]])
*/
Buffer.from = function from(value, encodingOrOffset, length) {
if (typeof value === 'string')
return fromString(value, encodingOrOffset);
if (typeof value === 'object' && value !== null) {
if (isAnyArrayBuffer(value))
return fromArrayBuffer(value, encodingOrOffset, length);
const valueOf = value.valueOf && value.valueOf();
if (valueOf != null &&
valueOf !== value &&
(typeof valueOf === 'string' || typeof valueOf === 'object')) {
return from(valueOf, encodingOrOffset, length);
}
const b = fromObject(value);
if (b)
return b;
if (typeof value[SymbolToPrimitive] === 'function') {
const primitive = value[SymbolToPrimitive]('string');
if (typeof primitive === 'string') {
return fromString(primitive, encodingOrOffset);
}
}
}
throw new ERR_INVALID_ARG_TYPE(
'first argument',
['string', 'Buffer', 'ArrayBuffer', 'Array', 'Array-like Object'],
value,
);
};
/**
* Creates the Buffer as a copy of the underlying ArrayBuffer of the view
* rather than the contents of the view.
* @param {TypedArray} view
* @param {number} [offset]
* @param {number} [length]
* @returns {Buffer}
*/
Buffer.copyBytesFrom = function copyBytesFrom(view, offset, length) {
if (!isTypedArray(view)) {
throw new ERR_INVALID_ARG_TYPE('view', [ 'TypedArray' ], view);
}
const viewLength = TypedArrayPrototypeGetLength(view);
if (viewLength === 0) {
return Buffer.alloc(0);
}
if (offset !== undefined || length !== undefined) {
if (offset !== undefined) {
validateInteger(offset, 'offset', 0);
if (offset >= viewLength) return Buffer.alloc(0);
} else {
offset = 0;
}
let end;
if (length !== undefined) {
validateInteger(length, 'length', 0);
end = offset + length;
} else {
end = viewLength;
}
view = TypedArrayPrototypeSlice(view, offset, end);
}
return fromArrayLike(new Uint8Array(
TypedArrayPrototypeGetBuffer(view),
TypedArrayPrototypeGetByteOffset(view),
TypedArrayPrototypeGetByteLength(view)));
};
// Identical to the built-in %TypedArray%.of(), but avoids using the deprecated
// Buffer() constructor. Must use arrow function syntax to avoid automatically
// adding a `prototype` property and making the function a constructor.
//
// Refs: https://tc39.github.io/ecma262/#sec-%typedarray%.of
// Refs: https://esdiscuss.org/topic/isconstructor#content-11
const of = (...items) => {
const newObj = createUnsafeBuffer(items.length);
for (let k = 0; k < items.length; k++)
newObj[k] = items[k];
return newObj;
};
Buffer.of = of;
ObjectSetPrototypeOf(Buffer, Uint8Array);
/**
* Creates a new filled Buffer instance.
* alloc(size[, fill[, encoding]])
*/
Buffer.alloc = function alloc(size, fill, encoding) {
validateNumber(size, 'size', 0, kMaxLength);
if (fill !== undefined && fill !== 0 && size > 0) {
const buf = createUnsafeBuffer(size);
return _fill(buf, fill, 0, buf.length, encoding);
}
return new FastBuffer(size);
};
/**
* Equivalent to Buffer(num), by default creates a non-zero-filled Buffer
* instance. If `--zero-fill-buffers` is set, will zero-fill the buffer.
*/
Buffer.allocUnsafe = function allocUnsafe(size) {
validateNumber(size, 'size', 0, kMaxLength);
return allocate(size);
};
/**
* Equivalent to SlowBuffer(num), by default creates a non-zero-filled
* Buffer instance that is not allocated off the pre-initialized pool.
* If `--zero-fill-buffers` is set, will zero-fill the buffer.
*/
Buffer.allocUnsafeSlow = function allocUnsafeSlow(size) {
validateNumber(size, 'size', 0, kMaxLength);
return createUnsafeBuffer(size);
};
// If --zero-fill-buffers command line argument is set, a zero-filled
// buffer is returned.
function SlowBuffer(size) {
validateNumber(size, 'size', 0, kMaxLength);
return createUnsafeBuffer(size);
}
ObjectSetPrototypeOf(SlowBuffer.prototype, Uint8ArrayPrototype);
ObjectSetPrototypeOf(SlowBuffer, Uint8Array);
function allocate(size) {
if (size <= 0) {
return new FastBuffer();
}
if (size < (Buffer.poolSize >>> 1)) {
if (size > (poolSize - poolOffset))
createPool();
const b = new FastBuffer(allocPool, poolOffset, size);
poolOffset += size;
alignPool();
return b;
}
return createUnsafeBuffer(size);
}
function fromStringFast(string, ops) {
const maxLength = Buffer.poolSize >>> 1;
let length = string.length; // Min length
if (length >= maxLength)
return createFromString(string, ops);
length *= 4; // Max length (4 bytes per character)
if (length >= maxLength)
length = ops.byteLength(string); // Actual length
if (length >= maxLength)
return createFromString(string, ops, length);
if (length > (poolSize - poolOffset))
createPool();
const actual = ops.write(allocBuffer, string, poolOffset, length);
const b = new FastBuffer(allocPool, poolOffset, actual);
poolOffset += actual;
alignPool();
return b;
}
function createFromString(string, ops, length = ops.byteLength(string)) {
const buf = Buffer.allocUnsafeSlow(length);
const actual = ops.write(buf, string, 0, length);
return actual < length ? new FastBuffer(buf.buffer, 0, actual) : buf;
}
function fromString(string, encoding) {
let ops;
if (!encoding || encoding === 'utf8' || typeof encoding !== 'string') {
ops = encodingOps.utf8;
} else {
ops = getEncodingOps(encoding);
if (ops === undefined)
throw new ERR_UNKNOWN_ENCODING(encoding);
}
return string.length === 0 ? new FastBuffer() : fromStringFast(string, ops);
}
function fromArrayBuffer(obj, byteOffset, length) {
// Convert byteOffset to integer
if (byteOffset === undefined) {
byteOffset = 0;
} else {
byteOffset = +byteOffset;
if (NumberIsNaN(byteOffset))
byteOffset = 0;
}
const maxLength = obj.byteLength - byteOffset;
if (maxLength < 0)
throw new ERR_BUFFER_OUT_OF_BOUNDS('offset');
if (length !== undefined) {
// Convert length to non-negative integer.
length = +length;
if (length > 0) {
if (length > maxLength)
throw new ERR_BUFFER_OUT_OF_BOUNDS('length');
} else {
length = 0;
}
}
return new FastBuffer(obj, byteOffset, length);
}
function fromArrayLike(obj) {
if (obj.length <= 0)
return new FastBuffer();
if (obj.length < (Buffer.poolSize >>> 1)) {
if (obj.length > (poolSize - poolOffset))
createPool();
const b = new FastBuffer(allocPool, poolOffset, obj.length);
TypedArrayPrototypeSet(b, obj, 0);
poolOffset += obj.length;
alignPool();
return b;
}
return new FastBuffer(obj);
}
function fromObject(obj) {
if (obj.length !== undefined || isAnyArrayBuffer(obj.buffer)) {
if (typeof obj.length !== 'number') {
return new FastBuffer();
}
return fromArrayLike(obj);
}
if (obj.type === 'Buffer' && ArrayIsArray(obj.data)) {
return fromArrayLike(obj.data);
}
}
// Static methods
Buffer.isBuffer = function isBuffer(b) {
return b instanceof Buffer;
};
Buffer.compare = function compare(buf1, buf2) {
if (!isUint8Array(buf1)) {
throw new ERR_INVALID_ARG_TYPE('buf1', ['Buffer', 'Uint8Array'], buf1);
}
if (!isUint8Array(buf2)) {
throw new ERR_INVALID_ARG_TYPE('buf2', ['Buffer', 'Uint8Array'], buf2);
}
if (buf1 === buf2) {
return 0;
}
return _compare(buf1, buf2);
};
Buffer.isEncoding = function isEncoding(encoding) {
return typeof encoding === 'string' && encoding.length !== 0 &&
normalizeEncoding(encoding) !== undefined;
};
Buffer[kIsEncodingSymbol] = Buffer.isEncoding;
Buffer.concat = function concat(list, length) {
validateArray(list, 'list');
if (list.length === 0)
return new FastBuffer();
if (length === undefined) {
length = 0;
for (let i = 0; i < list.length; i++) {
if (list[i].length) {
length += list[i].length;
}
}
} else {
validateOffset(length, 'length');
}
const buffer = Buffer.allocUnsafe(length);
let pos = 0;
for (let i = 0; i < list.length; i++) {
const buf = list[i];
if (!isUint8Array(buf)) {
// TODO(BridgeAR): This should not be of type ERR_INVALID_ARG_TYPE.
// Instead, find the proper error code for this.
throw new ERR_INVALID_ARG_TYPE(
`list[${i}]`, ['Buffer', 'Uint8Array'], list[i]);
}
pos += _copyActual(buf, buffer, pos, 0, buf.length);
}
// Note: `length` is always equal to `buffer.length` at this point
if (pos < length) {
// Zero-fill the remaining bytes if the specified `length` was more than
// the actual total length, i.e. if we have some remaining allocated bytes
// there were not initialized.
TypedArrayPrototypeFill(buffer, 0, pos, length);
}
return buffer;
};
function base64ByteLength(str, bytes) {
// Handle padding
if (StringPrototypeCharCodeAt(str, bytes - 1) === 0x3D)
bytes--;
if (bytes > 1 && StringPrototypeCharCodeAt(str, bytes - 1) === 0x3D)
bytes--;
// Base64 ratio: 3/4
return (bytes * 3) >>> 2;
}
const encodingOps = {
utf8: {
encoding: 'utf8',
encodingVal: encodingsMap.utf8,
byteLength: byteLengthUtf8,
write: (buf, string, offset, len) => buf.utf8Write(string, offset, len),
slice: (buf, start, end) => buf.utf8Slice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfString(buf, val, byteOffset, encodingsMap.utf8, dir),
},
ucs2: {
encoding: 'ucs2',
encodingVal: encodingsMap.utf16le,
byteLength: (string) => string.length * 2,
write: (buf, string, offset, len) => buf.ucs2Write(string, offset, len),
slice: (buf, start, end) => buf.ucs2Slice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfString(buf, val, byteOffset, encodingsMap.utf16le, dir),
},
utf16le: {
encoding: 'utf16le',
encodingVal: encodingsMap.utf16le,
byteLength: (string) => string.length * 2,
write: (buf, string, offset, len) => buf.ucs2Write(string, offset, len),
slice: (buf, start, end) => buf.ucs2Slice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfString(buf, val, byteOffset, encodingsMap.utf16le, dir),
},
latin1: {
encoding: 'latin1',
encodingVal: encodingsMap.latin1,
byteLength: (string) => string.length,
write: (buf, string, offset, len) => buf.latin1Write(string, offset, len),
slice: (buf, start, end) => buf.latin1Slice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfString(buf, val, byteOffset, encodingsMap.latin1, dir),
},
ascii: {
encoding: 'ascii',
encodingVal: encodingsMap.ascii,
byteLength: (string) => string.length,
write: (buf, string, offset, len) => buf.asciiWrite(string, offset, len),
slice: (buf, start, end) => buf.asciiSlice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfBuffer(buf,
fromStringFast(val, encodingOps.ascii),
byteOffset,
encodingsMap.ascii,
dir),
},
base64: {
encoding: 'base64',
encodingVal: encodingsMap.base64,
byteLength: (string) => base64ByteLength(string, string.length),
write: (buf, string, offset, len) => buf.base64Write(string, offset, len),
slice: (buf, start, end) => buf.base64Slice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfBuffer(buf,
fromStringFast(val, encodingOps.base64),
byteOffset,
encodingsMap.base64,
dir),
},
base64url: {
encoding: 'base64url',
encodingVal: encodingsMap.base64url,
byteLength: (string) => base64ByteLength(string, string.length),
write: (buf, string, offset, len) =>
buf.base64urlWrite(string, offset, len),
slice: (buf, start, end) => buf.base64urlSlice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfBuffer(buf,
fromStringFast(val, encodingOps.base64url),
byteOffset,
encodingsMap.base64url,
dir),
},
hex: {
encoding: 'hex',
encodingVal: encodingsMap.hex,
byteLength: (string) => string.length >>> 1,
write: (buf, string, offset, len) => buf.hexWrite(string, offset, len),
slice: (buf, start, end) => buf.hexSlice(start, end),
indexOf: (buf, val, byteOffset, dir) =>
indexOfBuffer(buf,
fromStringFast(val, encodingOps.hex),
byteOffset,
encodingsMap.hex,
dir),
},
};
function getEncodingOps(encoding) {
encoding += '';
switch (encoding.length) {
case 4:
if (encoding === 'utf8') return encodingOps.utf8;
if (encoding === 'ucs2') return encodingOps.ucs2;
encoding = StringPrototypeToLowerCase(encoding);
if (encoding === 'utf8') return encodingOps.utf8;
if (encoding === 'ucs2') return encodingOps.ucs2;
break;
case 5:
if (encoding === 'utf-8') return encodingOps.utf8;
if (encoding === 'ascii') return encodingOps.ascii;
if (encoding === 'ucs-2') return encodingOps.ucs2;
encoding = StringPrototypeToLowerCase(encoding);
if (encoding === 'utf-8') return encodingOps.utf8;
if (encoding === 'ascii') return encodingOps.ascii;
if (encoding === 'ucs-2') return encodingOps.ucs2;
break;
case 7:
if (encoding === 'utf16le' ||
StringPrototypeToLowerCase(encoding) === 'utf16le')
return encodingOps.utf16le;
break;
case 8:
if (encoding === 'utf-16le' ||
StringPrototypeToLowerCase(encoding) === 'utf-16le')
return encodingOps.utf16le;
break;
case 6:
if (encoding === 'latin1' || encoding === 'binary')
return encodingOps.latin1;
if (encoding === 'base64') return encodingOps.base64;
encoding = StringPrototypeToLowerCase(encoding);
if (encoding === 'latin1' || encoding === 'binary')
return encodingOps.latin1;
if (encoding === 'base64') return encodingOps.base64;
break;
case 3:
if (encoding === 'hex' || StringPrototypeToLowerCase(encoding) === 'hex')
return encodingOps.hex;
break;
case 9:
if (encoding === 'base64url' ||
StringPrototypeToLowerCase(encoding) === 'base64url')
return encodingOps.base64url;
break;
}
}
function byteLength(string, encoding) {
if (typeof string !== 'string') {
if (isArrayBufferView(string) || isAnyArrayBuffer(string)) {
return string.byteLength;
}
throw new ERR_INVALID_ARG_TYPE(
'string', ['string', 'Buffer', 'ArrayBuffer'], string,
);
}
const len = string.length;
if (len === 0)
return 0;
if (!encoding || encoding === 'utf8') {
return byteLengthUtf8(string);
}
if (encoding === 'ascii') {
return len;
}
const ops = getEncodingOps(encoding);
if (ops === undefined) {
// TODO (ronag): Makes more sense to throw here.
// throw new ERR_UNKNOWN_ENCODING(encoding);
return byteLengthUtf8(string);
}
return ops.byteLength(string);
}
Buffer.byteLength = byteLength;
// For backwards compatibility.
ObjectDefineProperty(Buffer.prototype, 'parent', {
__proto__: null,
enumerable: true,
get() {
if (!(this instanceof Buffer))
return undefined;
return this.buffer;
},
});
ObjectDefineProperty(Buffer.prototype, 'offset', {
__proto__: null,
enumerable: true,
get() {
if (!(this instanceof Buffer))
return undefined;
return this.byteOffset;
},
});
Buffer.prototype.copy =
function copy(target, targetStart, sourceStart, sourceEnd) {
return copyImpl(this, target, targetStart, sourceStart, sourceEnd);
};
// No need to verify that "buf.length <= MAX_UINT32" since it's a read-only
// property of a typed array.
// This behaves neither like String nor Uint8Array in that we set start/end
// to their upper/lower bounds if the value passed is out of range.
Buffer.prototype.toString = function toString(encoding, start, end) {
if (arguments.length === 0) {
return this.utf8Slice(0, this.length);
}
const len = this.length;
if (start <= 0)
start = 0;
else if (start >= len)
return '';
else
start = MathTrunc(start) || 0;
if (end === undefined || end > len)
end = len;
else
end = MathTrunc(end) || 0;
if (end <= start)
return '';
if (encoding === undefined)
return this.utf8Slice(start, end);
const ops = getEncodingOps(encoding);
if (ops === undefined)
throw new ERR_UNKNOWN_ENCODING(encoding);
return ops.slice(this, start, end);
};
Buffer.prototype.equals = function equals(otherBuffer) {
if (!isUint8Array(otherBuffer)) {
throw new ERR_INVALID_ARG_TYPE(
'otherBuffer', ['Buffer', 'Uint8Array'], otherBuffer);
}
if (this === otherBuffer)
return true;
const len = TypedArrayPrototypeGetByteLength(this);
if (len !== TypedArrayPrototypeGetByteLength(otherBuffer))
return false;
return len === 0 || _compare(this, otherBuffer) === 0;
};
let INSPECT_MAX_BYTES = 50;
// Override how buffers are presented by util.inspect().
Buffer.prototype[customInspectSymbol] = function inspect(recurseTimes, ctx) {
const max = INSPECT_MAX_BYTES;
const actualMax = MathMin(max, this.length);
const remaining = this.length - max;
let str = StringPrototypeTrim(RegExpPrototypeSymbolReplace(
/(.{2})/g, this.hexSlice(0, actualMax), '$1 '));
if (remaining > 0)
str += ` ... ${remaining} more byte${remaining > 1 ? 's' : ''}`;
// Inspect special properties as well, if possible.
if (ctx) {
let extras = false;
const filter = ctx.showHidden ? ALL_PROPERTIES : ONLY_ENUMERABLE;
const obj = { __proto__: null };
ArrayPrototypeForEach(getOwnNonIndexProperties(this, filter),
(key) => {
extras = true;
obj[key] = this[key];
});
if (extras) {
if (this.length !== 0)
str += ', ';
// '[Object: null prototype] {'.length === 26
// This is guarded with a test.
str += StringPrototypeSlice(utilInspect(obj, {
...ctx,
breakLength: Infinity,
compact: true,
}), 27, -2);
}
}
return `<${this.constructor.name} ${str}>`;
};
Buffer.prototype.inspect = Buffer.prototype[customInspectSymbol];
Buffer.prototype.compare = function compare(target,
targetStart,
targetEnd,
sourceStart,
sourceEnd) {
if (!isUint8Array(target)) {
throw new ERR_INVALID_ARG_TYPE('target', ['Buffer', 'Uint8Array'], target);
}
if (arguments.length === 1)
return _compare(this, target);
if (targetStart === undefined)
targetStart = 0;
else
validateOffset(targetStart, 'targetStart');
if (targetEnd === undefined)
targetEnd = target.length;
else
validateOffset(targetEnd, 'targetEnd', 0, target.length);
if (sourceStart === undefined)
sourceStart = 0;
else
validateOffset(sourceStart, 'sourceStart');
if (sourceEnd === undefined)
sourceEnd = this.length;
else
validateOffset(sourceEnd, 'sourceEnd', 0, this.length);
if (sourceStart >= sourceEnd)
return (targetStart >= targetEnd ? 0 : -1);
if (targetStart >= targetEnd)
return 1;
return compareOffset(this, target, targetStart, sourceStart, targetEnd,
sourceEnd);
};
// Finds either the first index of `val` in `buffer` at offset >= `byteOffset`,
// OR the last index of `val` in `buffer` at offset <= `byteOffset`.
//
// Arguments:
// - buffer - a Buffer to search
// - val - a string, Buffer, or number
// - byteOffset - an index into `buffer`; will be clamped to an int32
// - encoding - an optional encoding, relevant if val is a string
// - dir - true for indexOf, false for lastIndexOf
function bidirectionalIndexOf(buffer, val, byteOffset, encoding, dir) {
validateBuffer(buffer);
if (typeof byteOffset === 'string') {
encoding = byteOffset;
byteOffset = undefined;
} else if (byteOffset > 0x7fffffff) {
byteOffset = 0x7fffffff;
} else if (byteOffset < -0x80000000) {
byteOffset = -0x80000000;
}
// Coerce to Number. Values like null and [] become 0.
byteOffset = +byteOffset;
// If the offset is undefined, "foo", {}, coerces to NaN, search whole buffer.
if (NumberIsNaN(byteOffset)) {
byteOffset = dir ? 0 : (buffer.length || buffer.byteLength);
}
dir = !!dir; // Cast to bool.
if (typeof val === 'number')
return indexOfNumber(buffer, val >>> 0, byteOffset, dir);
let ops;
if (encoding === undefined)
ops = encodingOps.utf8;
else
ops = getEncodingOps(encoding);
if (typeof val === 'string') {
if (ops === undefined)
throw new ERR_UNKNOWN_ENCODING(encoding);
return ops.indexOf(buffer, val, byteOffset, dir);
}
if (isUint8Array(val)) {
const encodingVal =
(ops === undefined ? encodingsMap.utf8 : ops.encodingVal);
return indexOfBuffer(buffer, val, byteOffset, encodingVal, dir);
}
throw new ERR_INVALID_ARG_TYPE(
'value', ['number', 'string', 'Buffer', 'Uint8Array'], val,
);
}
Buffer.prototype.indexOf = function indexOf(val, byteOffset, encoding) {
return bidirectionalIndexOf(this, val, byteOffset, encoding, true);
};
Buffer.prototype.lastIndexOf = function lastIndexOf(val, byteOffset, encoding) {
return bidirectionalIndexOf(this, val, byteOffset, encoding, false);
};
Buffer.prototype.includes = function includes(val, byteOffset, encoding) {
return this.indexOf(val, byteOffset, encoding) !== -1;
};
// Usage:
// buffer.fill(number[, offset[, end]])
// buffer.fill(buffer[, offset[, end]])
// buffer.fill(string[, offset[, end]][, encoding])
Buffer.prototype.fill = function fill(value, offset, end, encoding) {
return _fill(this, value, offset, end, encoding);
};
function _fill(buf, value, offset, end, encoding) {
if (typeof value === 'string') {
if (offset === undefined || typeof offset === 'string') {
encoding = offset;
offset = 0;
end = buf.length;
} else if (typeof end === 'string') {
encoding = end;
end = buf.length;
}
const normalizedEncoding = normalizeEncoding(encoding);
if (normalizedEncoding === undefined) {
validateString(encoding, 'encoding');
throw new ERR_UNKNOWN_ENCODING(encoding);
}
if (value.length === 0) {
// If value === '' default to zero.
value = 0;
} else if (value.length === 1) {
// Fast path: If `value` fits into a single byte, use that numeric value.
if (normalizedEncoding === 'utf8') {
const code = StringPrototypeCharCodeAt(value, 0);
if (code < 128) {
value = code;
}
} else if (normalizedEncoding === 'latin1') {
value = StringPrototypeCharCodeAt(value, 0);
}
}
} else {
encoding = undefined;
}
if (offset === undefined) {
offset = 0;
end = buf.length;
} else {
validateOffset(offset, 'offset');
// Invalid ranges are not set to a default, so can range check early.
if (end === undefined) {
end = buf.length;
} else {
validateOffset(end, 'end', 0, buf.length);
}
if (offset >= end)
return buf;
}
if (typeof value === 'number') {
// OOB check
const byteLen = TypedArrayPrototypeGetByteLength(buf);
const fillLength = end - offset;
if (offset > end || fillLength + offset > byteLen)
throw new ERR_BUFFER_OUT_OF_BOUNDS();
TypedArrayPrototypeFill(buf, value, offset, end);
} else {
const res = bindingFill(buf, value, offset, end, encoding);
if (res < 0) {
if (res === -1)
throw new ERR_INVALID_ARG_VALUE('value', value);
throw new ERR_BUFFER_OUT_OF_BOUNDS();
}
}
return buf;
}
Buffer.prototype.write = function write(string, offset, length, encoding) {
// Buffer#write(string);
if (offset === undefined) {
return this.utf8Write(string, 0, this.length);
}
// Buffer#write(string, encoding)
if (length === undefined && typeof offset === 'string') {
encoding = offset;
length = this.length;
offset = 0;
// Buffer#write(string, offset[, length][, encoding])
} else {
validateOffset(offset, 'offset', 0, this.length);
const remaining = this.length - offset;
if (length === undefined) {
length = remaining;
} else if (typeof length === 'string') {
encoding = length;
length = remaining;
} else {
validateOffset(length, 'length', 0, this.length);
if (length > remaining)
length = remaining;
}
}
if (!encoding || encoding === 'utf8')
return this.utf8Write(string, offset, length);
if (encoding === 'ascii')
return this.asciiWrite(string, offset, length);
const ops = getEncodingOps(encoding);
if (ops === undefined)
throw new ERR_UNKNOWN_ENCODING(encoding);
return ops.write(this, string, offset, length);
};
Buffer.prototype.toJSON = function toJSON() {
if (this.length > 0) {
const data = new Array(this.length);
for (let i = 0; i < this.length; ++i)
data[i] = this[i];
return { type: 'Buffer', data };
}
return { type: 'Buffer', data: [] };
};
function adjustOffset(offset, length) {
// Use Math.trunc() to convert offset to an integer value that can be larger
// than an Int32. Hence, don't use offset | 0 or similar techniques.
offset = MathTrunc(offset);
if (offset === 0) {
return 0;
}
if (offset < 0) {
offset += length;
return offset > 0 ? offset : 0;
}
if (offset < length) {
return offset;
}
return NumberIsNaN(offset) ? 0 : length;
}
Buffer.prototype.subarray = function subarray(start, end) {
const srcLength = this.length;
start = adjustOffset(start, srcLength);
end = end !== undefined ? adjustOffset(end, srcLength) : srcLength;
const newLength = end > start ? end - start : 0;
return new FastBuffer(this.buffer, this.byteOffset + start, newLength);
};
Buffer.prototype.slice = function slice(start, end) {
return this.subarray(start, end);
};
function swap(b, n, m) {
const i = b[n];
b[n] = b[m];
b[m] = i;
}
Buffer.prototype.swap16 = function swap16() {
// For Buffer.length < 128, it's generally faster to
// do the swap in javascript. For larger buffers,
// dropping down to the native code is faster.
const len = this.length;
if (len % 2 !== 0)
throw new ERR_INVALID_BUFFER_SIZE('16-bits');
if (len < 128) {
for (let i = 0; i < len; i += 2)
swap(this, i, i + 1);
return this;
}
return _swap16(this);
};
Buffer.prototype.swap32 = function swap32() {
// For Buffer.length < 192, it's generally faster to
// do the swap in javascript. For larger buffers,
// dropping down to the native code is faster.
const len = this.length;
if (len % 4 !== 0)
throw new ERR_INVALID_BUFFER_SIZE('32-bits');
if (len < 192) {
for (let i = 0; i < len; i += 4) {
swap(this, i, i + 3);
swap(this, i + 1, i + 2);
}
return this;
}
return _swap32(this);
};
Buffer.prototype.swap64 = function swap64() {
// For Buffer.length < 192, it's generally faster to
// do the swap in javascript. For larger buffers,
// dropping down to the native code is faster.
const len = this.length;
if (len % 8 !== 0)
throw new ERR_INVALID_BUFFER_SIZE('64-bits');
if (len < 192) {
for (let i = 0; i < len; i += 8) {
swap(this, i, i + 7);
swap(this, i + 1, i + 6);
swap(this, i + 2, i + 5);
swap(this, i + 3, i + 4);
}
return this;
}
return _swap64(this);
};
Buffer.prototype.toLocaleString = Buffer.prototype.toString;
let transcode;
if (internalBinding('config').hasIntl) {
const {
icuErrName,
transcode: _transcode,
} = internalBinding('icu');
// Transcodes the Buffer from one encoding to another, returning a new
// Buffer instance.
transcode = function transcode(source, fromEncoding, toEncoding) {
if (!isUint8Array(source)) {
throw new ERR_INVALID_ARG_TYPE('source',
['Buffer', 'Uint8Array'], source);
}
if (source.length === 0) return Buffer.alloc(0);
fromEncoding = normalizeEncoding(fromEncoding) || fromEncoding;
toEncoding = normalizeEncoding(toEncoding) || toEncoding;
const result = _transcode(source, fromEncoding, toEncoding);
if (typeof result !== 'number')
return result;
const code = icuErrName(result);
const err = genericNodeError(
`Unable to transcode Buffer [${code}]`,
{ code: code, errno: result },
);
throw err;
};
}
function btoa(input) {
// The implementation here has not been performance optimized in any way and
// should not be.
// Refs: https://github.com/nodejs/node/pull/38433#issuecomment-828426932
if (arguments.length === 0) {
throw new ERR_MISSING_ARGS('input');
}
const result = _btoa(`${input}`);
if (result === -1) {
throw lazyDOMException('Invalid character', 'InvalidCharacterError');
}
return result;
}
function atob(input) {
if (arguments.length === 0) {
throw new ERR_MISSING_ARGS('input');
}
const result = _atob(`${input}`);
switch (result) {
case -2: // Invalid character
throw lazyDOMException('Invalid character', 'InvalidCharacterError');
case -1: // Single character remained
throw lazyDOMException(
'The string to be decoded is not correctly encoded.',
'InvalidCharacterError');
case -3: // Possible overflow
// TODO(@anonrig): Throw correct error in here.
throw lazyDOMException('The input causes overflow.', 'InvalidCharacterError');
default:
return result;
}
}
function isUtf8(input) {
if (isTypedArray(input) || isAnyArrayBuffer(input)) {
return bindingIsUtf8(input);
}
throw new ERR_INVALID_ARG_TYPE('input', ['ArrayBuffer', 'Buffer', 'TypedArray'], input);
}
function isAscii(input) {
if (isTypedArray(input) || isAnyArrayBuffer(input)) {
return bindingIsAscii(input);
}
throw new ERR_INVALID_ARG_TYPE('input', ['ArrayBuffer', 'Buffer', 'TypedArray'], input);
}
module.exports = {
Buffer,
SlowBuffer: deprecate(
SlowBuffer,
'SlowBuffer() is deprecated. Please use Buffer.allocUnsafeSlow()',
'DEP0030'),
transcode,
isUtf8,
isAscii,
// Legacy
kMaxLength,
kStringMaxLength,
btoa,
atob,
};
ObjectDefineProperties(module.exports, {
constants: {
__proto__: null,
configurable: false,
enumerable: true,
value: constants,
},
INSPECT_MAX_BYTES: {
__proto__: null,
configurable: true,
enumerable: true,
get() { return INSPECT_MAX_BYTES; },
set(val) {
validateNumber(val, 'INSPECT_MAX_BYTES', 0);
INSPECT_MAX_BYTES = val;
},
},
});
defineLazyProperties(
module.exports,
'internal/blob',
['Blob', 'resolveObjectURL'],
);
defineLazyProperties(
module.exports,
'internal/file',
['File'],
);