Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/novabyte/grizzly-protobuf
Grizzly NIO framework filters to process Google's Protocol Buffers serialized messages.
https://github.com/novabyte/grizzly-protobuf
Last synced: about 1 month ago
JSON representation
Grizzly NIO framework filters to process Google's Protocol Buffers serialized messages.
- Host: GitHub
- URL: https://github.com/novabyte/grizzly-protobuf
- Owner: novabyte
- License: apache-2.0
- Created: 2013-09-03T00:32:22.000Z (over 11 years ago)
- Default Branch: master
- Last Pushed: 2013-09-06T21:32:41.000Z (over 11 years ago)
- Last Synced: 2024-10-13T18:47:10.947Z (2 months ago)
- Language: Java
- Homepage:
- Size: 219 KB
- Stars: 4
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
Grizzly-Protobuf
================
[Grizzly NIO framework](http://grizzly.java.net/) filters to encode and decode
Google's [Protocol Buffers](https://developers.google.com/protocol-buffers/)
serialized messages.
Grizzly-Protobuf is created and maintained by Chris Molozian (@novabyte).
Code licensed under the [Apache License v2.0](http://www.apache.org/licenses/LICENSE-2.0).
Documentation licensed under [CC BY 3.0](http://creativecommons.org/licenses/by/3.0/).
## Usage ##
Grizzly-Protobuf is available on [Maven Central](http://search.maven.org/).
```xml
me.cmoz.grizzly
grizzly-protobuf
1.0.0
```
If you're not using Maven (or a dependency resolver that's compatible with Maven
repositories), you can download the JARs you need for your project from
Maven Central.
### Filter Types ###
At the moment there are two filters to choose from:
1. __FixedLengthProtobufFilter__, this uses a fixed length header to store the
size of the protobuf message being (de)serialized.
2. __Varint32ProtobufFilter__, this uses [`varint32` encoding](https://developers.google.com/protocol-buffers/docs/encoding#varints)
to store the size of the protobuf message being (de)serialized.
If you're not sure which to use, start with the `Varint32ProtobufFilter` and
move to the fixed length header alternative if necessary.
## Example ##
Both filters need a `MessageLite` type to use to parse all incoming messages.
The example below demonstrates the `Person` protocol format from the Protocol Buffers
[Java Tutorial](https://developers.google.com/protocol-buffers/docs/javatutorial)
with the `Varint32ProtobufFilter`. All messages sent/received by the filter chain
will (de)serialize to the `Person` protocol format:
```java
final MessageLite prototype = Person.newBuilder()
.setId(1234)
.setName("John Doe")
.setEmail("[email protected]")
.build();
final FilterChainBuilder serverFilterBuilder = FilterChainBuilder.stateless()
.add(new TransportFilter())
.add(new Varint32ProtobufFilter(prototype.getDefaultInstanceForType()));
```
For more detailed examples of how to integrate this filter into your code have a
look at the [test cases](https://github.com/novabyte/grizzly-protobuf/tree/master/src/test/java/me/cmoz/grizzly/protobuf).
## Developer Notes ##
The codebase requires the [Gradle](http://gradle.org) build tool at version
`1.6+` and the Java compiler at version `1.6.0` or greater.
The main external dependency for the project is [Grizzly NIO](http://grizzly.java.net/),
at `2.3.5` or greater and [Protobuf-Java](http://search.maven.org/#artifactdetails|com.google.protobuf|protobuf-java|2.5.0|bundle)
at `2.5.0` (for improved parsing performance [see here](http://protobuf.googlecode.com/svn/trunk/CHANGES.txt))
although older versions are also supported.
For a full list of dependencies see the [build script](https://github.com/novabyte/grizzly-protobuf/blob/master/build.gradle).
All dependencies are downloaded by Gradle during the build process.
### Building the codebase ###
A list of all possible build targets can be displayed by Gradle with
`gradle tasks`.
In a regular write-compile-test cycle use `gradle test`.
A list of all project dependencies can be displayed by Gradle with
`gradle dependencies`.
It is recommended to run Gradle with the
[Build Daemon](http://docs.codehaus.org/display/GRADLE/Gradle+Build+Daemon)
enabled to improve performance. e.g. `gradle --daemon` once the daemon is
running it can be stopped with `gradle --stop`.
## Contribute ##
All contributions to the documentation and the codebase are very welcome. Feel
free to open issues on the tracker wherever the documentation needs improving.
Also, pull requests are always welcome! `:)`