- 修正了linkedin/goavro的bug。包括但不限于这些Pull Request，时间紧张只有quick dirty的修正，原作者也都认可这些fix。(嗯，但他也挺忙了也没空改个更好的出来。)。
- 修改了一个基于能和apache/avro python的1.7.7版本的简化ipc，去掉了握手的过程。
- 根据简化的py的ipc添加了go的ipc, 例子在 example/ipc中
Goavro is a golang library that implements encoding and decoding of
Avro data. It provides an interface to encode data directly to
io.Writer streams, and decoding data from io.Reader
streams. Goavro fully adheres to
version 1.7.7 of the Avro specification.
Documentation is available via
Please see the example programs in the examples directory for
reference.
Although the Avro specification defines the terms reader and writer as
library components which read and write Avro data, Go has particular
strong emphasis on what a Reader and Writer are. Namely, it is bad
form to define an interface which shares the same name but uses a
different method signature. In other words, all Reader interfaces
should effectively mirror an io.Reader, and all Writer interfaces
should mirror an io.Writer. Adherance to this standard is essential
to keep libraries easy to use.
An io.Reader reads data from the stream specified at object creation
time into the parameterized slice of bytes and returns both the number
of bytes read and an error. An Avro reader also reads from a stream,
but it is possible to create an Avro reader that can read from one
stream, then read from another, using the same compiled schema. In
other words, an Avro reader puts the schema first, whereas an
io.Reader puts the stream first.
To support an Avro reader being able to read from multiple streams,
it's API must be different and incompatible with io.Reader interface
from the Go standard. Instead, an Avro reader looks more like the
Unmarshal functionality provided by the Go encoding/json library.
Creating a goavro.Codec is fast, but ought to be performed exactly
once per Avro schema to process. Once a Codec is created, it may be
used multiple times to either decode or encode data.
The Codec interface exposes two methods, one to encode data and one
to decode data. They encode directly into an io.Writer, and decode
directly from an io.Reader.
A particular Codec can work with only one Avro schema. However,
there is no practical limit to how many Codecs may be created and
used in a program. Internally a goavro.codec is merely a namespace
and two function pointers to decode and encode data. Because codecs
maintain no state, the same Codec can be concurrently used on
different io streams as desired.
func (c *codec) Decode(r io.Reader) (interface{}, error)
func (c *codec) Encode(w io.Writer, datum interface{}) errorThe below is an example of creating a Codec from a provided JSON
schema. Codecs do not maintain any internal state, and may be used
multiple times on multiple io.Readers, io.Writers, concurrently if
desired.
someRecordSchemaJson := `{"type":"record","name":"Foo","fields":[{"name":"field1","type":"int"},{"name":"field2","type":"string","default":"happy"}]}`
codec, err := goavro.NewCodec(someRecordSchemaJson)
if err != nil {
return nil, err
}The below is a simplified example of decoding binary data to be read
from an io.Reader into a single datum using a previously compiled
Codec. The Decode method of the Codec interface may be called
multiple times, each time on the same or on different io.Reader
objects.
// uses codec created above, and an io.Reader, definition not shown
datum, err := codec.Decode(r)
if err != nil {
return nil, err
}The below is a simplified example of encoding a single datum into the
Avro binary format using a previously compiled Codec. The Encode
method of the Codec interface may be called multiple times, each
time on the same or on different io.Writer objects.
// uses codec created above, an io.Writer, definition not shown,
// and some data
err := codec.Encode(w, datum)
if err != nil {
return nil, err
}Another example, this time leveraging bufio.Writer:
// Encoding data using bufio.Writer to buffer the writes
// during data encoding:
func encodeWithBufferedWriter(c Codec, w io.Writer, datum interface{}) error {
bw := bufio.NewWriter(w)
err := c.Encode(bw, datum)
if err != nil {
return err
}
return bw.Flush()
}
err := encodeWithBufferedWriter(codec, w, datum)
if err != nil {
return nil, err
}The Codec interface provides means to encode and decode any Avro
data, but a number of additional helper types are provided to handle
streaming of Avro data.
See the example programs examples/file/reader.go and
examples/file/writer.go for more context:
This example wraps the provided io.Reader in a bufio.Reader and
dumps the data to standard output.
func dumpReader(r io.Reader) {
fr, err := goavro.NewReader(goavro.BufferFromReader(r))
if err != nil {
log.Fatal("cannot create Reader: ", err)
}
defer func() {
if err := fr.Close(); err != nil {
log.Fatal(err)
}
}()
for fr.Scan() {
datum, err := fr.Read()
if err != nil {
log.Println("cannot read datum: ", err)
continue
}
fmt.Println(datum)
}
}This example buffers the provided io.Writer in a bufio.Writer, and
writes some data to the stream.
func makeSomeData(w io.Writer) error {
recordSchema := `
{
"type": "record",
"name": "example",
"fields": [
{
"type": "string",
"name": "username"
},
{
"type": "string",
"name": "comment"
},
{
"type": "long",
"name": "timestamp"
}
]
}
`
fw, err := goavro.NewWriter(
goavro.BlockSize(13), // example; default is 10
goavro.Compression(goavro.CompressionSnappy), // default is CompressionNull
goavro.WriterSchema(recordSchema),
goavro.ToWriter(w))
if err != nil {
log.Fatal("cannot create Writer: ", err)
}
defer fw.Close()
// make a record instance using the same schema
someRecord, err := goavro.NewRecord(goavro.RecordSchema(recordSchema))
if err != nil {
log.Fatal(err)
}
// identify field name to set datum for
someRecord.Set("username", "Aquaman")
someRecord.Set("comment", "The Atlantic is oddly cold this morning!")
// you can fully qualify the field name
someRecord.Set("com.example.timestamp", int64(1082196484))
fw.Write(someRecord)
// make another record
someRecord, err = goavro.NewRecord(goavro.RecordSchema(recordSchema))
if err != nil {
log.Fatal(err)
}
someRecord.Set("username", "Batman")
someRecord.Set("comment", "Who are all of these crazies?")
someRecord.Set("com.example.timestamp", int64(1427383430))
fw.Write(someRecord)
}Goavro is a fully featured encoder and decoder of binary Avro data. It fully supports recursive data structures, unions, and namespacing. It does have a few limitations that have yet to be implemented.
The Avro specification allows an implementation to optionally map a writer's schema to a reader's schema using aliases. Although goavro can complile schemas with aliases, it does not yet implement this feature.
The Avro Data Serialization format describes two encodings: binary and JSON. Goavro only implements binary encoding of data streams, because that is what most applications need.
Most applications will use the binary encoding, as it is smaller and faster. But, for debugging and web-based applications, the JSON encoding may sometimes be appropriate.
Note that data schemas are always encoded using JSON, as per the specification.
Kakfa is the reason goavro was written. Similar to Avro Object Container Files being a layer of abstraction above Avro Data Serialization format, Kafka's use of Avro is a layer of abstraction that also sits above Avro Data Serialization format, but has its own schema. Like Avro Object Container Files, this has been implemented but removed until the API can be improved.
Copyright 2015 LinkedIn Corp. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0.
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.Copyright [201X] LinkedIn Corp. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0.
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
Copyright (c) 2011 The Snappy-Go Authors. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
- Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
- Neither the name of Google Inc. nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
Goavro links with Google Snappy to provide Snappy compression and decompression support.