blob: cdb70feef5b32f9b2ed3b9d3e79a7dab0e311c44 [file] [log] [blame]
// Copyright 2013 The Flutter Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
import 'dart:convert';
import 'src/serialization.dart';
export 'src/serialization.dart' show ReadBuffer, WriteBuffer;
const int _writeBufferStartCapacity = 64;
/// A message encoding/decoding mechanism.
///
/// Both operations throw an exception, if conversion fails. Such situations
/// should be treated as programming errors.
///
/// See also:
///
/// * [BasicMessageChannel], which use [MessageCodec]s for communication
/// between Flutter and platform plugins.
abstract class MessageCodec<T> {
/// Encodes the specified [message] in binary.
///
/// Returns null if the message is null.
ByteData? encodeMessage(T message);
/// Decodes the specified [message] from binary.
///
/// Returns null if the message is null.
T? decodeMessage(ByteData? message);
}
/// [MessageCodec] using the Flutter standard binary encoding.
///
/// Supported messages are acyclic values of these forms:
///
/// * null
/// * [bool]s
/// * [num]s
/// * [String]s
/// * [Uint8List]s, [Int32List]s, [Int64List]s, [Float64List]s
/// * [List]s of supported values
/// * [Map]s from supported values to supported values
///
/// Decoded values will use `List<Object?>` and `Map<Object?, Object?>`
/// irrespective of content.
///
/// The type returned from [decodeMessage] is `dynamic` (not `Object?`), which
/// means *no type checking is performed on its return value*. It is strongly
/// recommended that the return value be immediately cast to a known type to
/// prevent runtime errors due to typos that the type checker could otherwise
/// catch.
///
/// The codec is extensible by subclasses overriding [writeValue] and
/// [readValueOfType].
///
/// ## Android specifics
///
/// On Android, messages are represented as follows:
///
/// * null: null
/// * [bool]\: `java.lang.Boolean`
/// * [int]\: `java.lang.Integer` for values that are representable using 32-bit
/// two's complement; `java.lang.Long` otherwise
/// * [double]\: `java.lang.Double`
/// * [String]\: `java.lang.String`
/// * [Uint8List]\: `byte[]`
/// * [Int32List]\: `int[]`
/// * [Int64List]\: `long[]`
/// * [Float64List]\: `double[]`
/// * [List]\: `java.util.ArrayList`
/// * [Map]\: `java.util.HashMap`
///
/// When sending a `java.math.BigInteger` from Java, it is converted into a
/// [String] with the hexadecimal representation of the integer. (The value is
/// tagged as being a big integer; subclasses of this class could be made to
/// support it natively; see the discussion at [writeValue].) This codec does
/// not support sending big integers from Dart.
///
/// ## iOS specifics
///
/// On iOS, messages are represented as follows:
///
/// * null: nil
/// * [bool]\: `NSNumber numberWithBool:`
/// * [int]\: `NSNumber numberWithInt:` for values that are representable using
/// 32-bit two's complement; `NSNumber numberWithLong:` otherwise
/// * [double]\: `NSNumber numberWithDouble:`
/// * [String]\: `NSString`
/// * [Uint8List], [Int32List], [Int64List], [Float64List]\:
/// `FlutterStandardTypedData`
/// * [List]\: `NSArray`
/// * [Map]\: `NSDictionary`
class StandardMessageCodec implements MessageCodec<Object?> {
/// Creates a [MessageCodec] using the Flutter standard binary encoding.
const StandardMessageCodec();
// The codec serializes messages as outlined below. This format must match the
// Android and iOS counterparts and cannot change (as it's possible for
// someone to end up using this for persistent storage).
//
// * A single byte with one of the constant values below determines the
// type of the value.
// * The serialization of the value itself follows the type byte.
// * Numbers are represented using the host endianness throughout.
// * Lengths and sizes of serialized parts are encoded using an expanding
// format optimized for the common case of small non-negative integers:
// * values 0..253 inclusive using one byte with that value;
// * values 254..2^16 inclusive using three bytes, the first of which is
// 254, the next two the usual unsigned representation of the value;
// * values 2^16+1..2^32 inclusive using five bytes, the first of which is
// 255, the next four the usual unsigned representation of the value.
// * null, true, and false have empty serialization; they are encoded directly
// in the type byte (using _valueNull, _valueTrue, _valueFalse)
// * Integers representable in 32 bits are encoded using 4 bytes two's
// complement representation.
// * Larger integers are encoded using 8 bytes two's complement
// representation.
// * doubles are encoded using the IEEE 754 64-bit double-precision binary
// format. Zero bytes are added before the encoded double value to align it
// to a 64 bit boundary in the full message.
// * Strings are encoded using their UTF-8 representation. First the length
// of that in bytes is encoded using the expanding format, then follows the
// UTF-8 encoding itself.
// * Uint8Lists, Int32Lists, Int64Lists, Float32Lists, and Float64Lists are
// encoded by first encoding the list's element count in the expanding
// format, then the smallest number of zero bytes needed to align the
// position in the full message with a multiple of the number of bytes per
// element, then the encoding of the list elements themselves, end-to-end
// with no additional type information, using two's complement or IEEE 754
// as applicable.
// * Lists are encoded by first encoding their length in the expanding format,
// then follows the recursive encoding of each element value, including the
// type byte (Lists are assumed to be heterogeneous).
// * Maps are encoded by first encoding their length in the expanding format,
// then follows the recursive encoding of each key/value pair, including the
// type byte for both (Maps are assumed to be heterogeneous).
//
// The type labels below must not change, since it's possible for this interface
// to be used for persistent storage.
static const int _valueNull = 0;
static const int _valueTrue = 1;
static const int _valueFalse = 2;
static const int _valueInt32 = 3;
static const int _valueInt64 = 4;
static const int _valueLargeInt = 5;
static const int _valueFloat64 = 6;
static const int _valueString = 7;
static const int _valueUint8List = 8;
static const int _valueInt32List = 9;
static const int _valueInt64List = 10;
static const int _valueFloat64List = 11;
static const int _valueList = 12;
static const int _valueMap = 13;
static const int _valueFloat32List = 14;
@override
ByteData? encodeMessage(Object? message) {
if (message == null) {
return null;
}
final WriteBuffer buffer =
WriteBuffer(startCapacity: _writeBufferStartCapacity);
writeValue(buffer, message);
return buffer.done();
}
@override
dynamic decodeMessage(ByteData? message) {
if (message == null) {
return null;
}
final ReadBuffer buffer = ReadBuffer(message);
final Object? result = readValue(buffer);
if (buffer.hasRemaining) {
throw const FormatException('Message corrupted');
}
return result;
}
/// Writes [value] to [buffer] by first writing a type discriminator
/// byte, then the value itself.
///
/// This method may be called recursively to serialize container values.
///
/// Type discriminators 0 through 127 inclusive are reserved for use by the
/// base class, as follows:
///
/// * null = 0
/// * true = 1
/// * false = 2
/// * 32 bit integer = 3
/// * 64 bit integer = 4
/// * larger integers = 5 (see below)
/// * 64 bit floating-point number = 6
/// * String = 7
/// * Uint8List = 8
/// * Int32List = 9
/// * Int64List = 10
/// * Float64List = 11
/// * List = 12
/// * Map = 13
/// * Float32List = 14
/// * Reserved for future expansion: 15..127
///
/// The codec can be extended by overriding this method, calling super
/// for values that the extension does not handle. Type discriminators
/// used by extensions must be greater than or equal to 128 in order to avoid
/// clashes with any later extensions to the base class.
///
/// The "larger integers" type, 5, is never used by [writeValue]. A subclass
/// could represent big integers from another package using that type. The
/// format is first the type byte (0x05), then the actual number as an ASCII
/// string giving the hexadecimal representation of the integer, with the
/// string's length as encoded by [writeSize] followed by the string bytes. On
/// Android, that would get converted to a `java.math.BigInteger` object. On
/// iOS, the string representation is returned.
void writeValue(WriteBuffer buffer, Object? value) {
if (value == null) {
buffer.putUint8(_valueNull);
} else if (value is bool) {
buffer.putUint8(value ? _valueTrue : _valueFalse);
} else if (value is double) {
// Double precedes int because in JS everything is a double.
// Therefore in JS, both `is int` and `is double` always
// return `true`. If we check int first, we'll end up treating
// all numbers as ints and attempt the int32/int64 conversion,
// which is wrong. This precedence rule is irrelevant when
// decoding because we use tags to detect the type of value.
buffer.putUint8(_valueFloat64);
buffer.putFloat64(value);
// ignore: avoid_double_and_int_checks, JS code always goes through the `double` path above
} else if (value is int) {
if (-0x7fffffff - 1 <= value && value <= 0x7fffffff) {
buffer.putUint8(_valueInt32);
buffer.putInt32(value);
} else {
buffer.putUint8(_valueInt64);
buffer.putInt64(value);
}
} else if (value is String) {
buffer.putUint8(_valueString);
final Uint8List asciiBytes = Uint8List(value.length);
Uint8List? utf8Bytes;
int utf8Offset = 0;
// Only do utf8 encoding if we encounter non-ascii characters.
for (int i = 0; i < value.length; i += 1) {
final int char = value.codeUnitAt(i);
if (char <= 0x7f) {
asciiBytes[i] = char;
} else {
utf8Bytes = utf8.encoder.convert(value.substring(i));
utf8Offset = i;
break;
}
}
if (utf8Bytes != null) {
writeSize(buffer, utf8Offset + utf8Bytes.length);
buffer.putUint8List(Uint8List.sublistView(asciiBytes, 0, utf8Offset));
buffer.putUint8List(utf8Bytes);
} else {
writeSize(buffer, asciiBytes.length);
buffer.putUint8List(asciiBytes);
}
} else if (value is Uint8List) {
buffer.putUint8(_valueUint8List);
writeSize(buffer, value.length);
buffer.putUint8List(value);
} else if (value is Int32List) {
buffer.putUint8(_valueInt32List);
writeSize(buffer, value.length);
buffer.putInt32List(value);
} else if (value is Int64List) {
buffer.putUint8(_valueInt64List);
writeSize(buffer, value.length);
buffer.putInt64List(value);
} else if (value is Float32List) {
buffer.putUint8(_valueFloat32List);
writeSize(buffer, value.length);
buffer.putFloat32List(value);
} else if (value is Float64List) {
buffer.putUint8(_valueFloat64List);
writeSize(buffer, value.length);
buffer.putFloat64List(value);
} else if (value is List) {
buffer.putUint8(_valueList);
writeSize(buffer, value.length);
for (final Object? item in value) {
writeValue(buffer, item);
}
} else if (value is Map) {
buffer.putUint8(_valueMap);
writeSize(buffer, value.length);
value.forEach((Object? key, Object? value) {
writeValue(buffer, key);
writeValue(buffer, value);
});
} else {
throw ArgumentError.value(value);
}
}
/// Reads a value from [buffer] as written by [writeValue].
///
/// This method is intended for use by subclasses overriding
/// [readValueOfType].
Object? readValue(ReadBuffer buffer) {
if (!buffer.hasRemaining) {
throw const FormatException('Message corrupted');
}
final int type = buffer.getUint8();
return readValueOfType(type, buffer);
}
/// Reads a value of the indicated [type] from [buffer].
///
/// The codec can be extended by overriding this method, calling super for
/// types that the extension does not handle. See the discussion at
/// [writeValue].
Object? readValueOfType(int type, ReadBuffer buffer) {
switch (type) {
case _valueNull:
return null;
case _valueTrue:
return true;
case _valueFalse:
return false;
case _valueInt32:
return buffer.getInt32();
case _valueInt64:
return buffer.getInt64();
case _valueFloat64:
return buffer.getFloat64();
case _valueLargeInt:
case _valueString:
final int length = readSize(buffer);
return utf8.decoder.convert(buffer.getUint8List(length));
case _valueUint8List:
final int length = readSize(buffer);
return buffer.getUint8List(length);
case _valueInt32List:
final int length = readSize(buffer);
return buffer.getInt32List(length);
case _valueInt64List:
final int length = readSize(buffer);
return buffer.getInt64List(length);
case _valueFloat32List:
final int length = readSize(buffer);
return buffer.getFloat32List(length);
case _valueFloat64List:
final int length = readSize(buffer);
return buffer.getFloat64List(length);
case _valueList:
final int length = readSize(buffer);
final List<Object?> result = List<Object?>.filled(length, null);
for (int i = 0; i < length; i++) {
result[i] = readValue(buffer);
}
return result;
case _valueMap:
final int length = readSize(buffer);
final Map<Object?, Object?> result = <Object?, Object?>{};
for (int i = 0; i < length; i++) {
result[readValue(buffer)] = readValue(buffer);
}
return result;
default:
throw const FormatException('Message corrupted');
}
}
/// Writes a non-negative 32-bit integer [value] to [buffer]
/// using an expanding 1-5 byte encoding that optimizes for small values.
///
/// This method is intended for use by subclasses overriding
/// [writeValue].
void writeSize(WriteBuffer buffer, int value) {
assert(0 <= value && value <= 0xffffffff);
if (value < 254) {
buffer.putUint8(value);
} else if (value <= 0xffff) {
buffer.putUint8(254);
buffer.putUint16(value);
} else {
buffer.putUint8(255);
buffer.putUint32(value);
}
}
/// Reads a non-negative int from [buffer] as written by [writeSize].
///
/// This method is intended for use by subclasses overriding
/// [readValueOfType].
int readSize(ReadBuffer buffer) {
final int value = buffer.getUint8();
switch (value) {
case 254:
return buffer.getUint16();
case 255:
return buffer.getUint32();
default:
return value;
}
}
}