CODEU is a program created by Google to develop the skills of future software engineers. This project is not an offical Google Product. This project is a playground for those looking to develop their coding and software engineering skills.
All instructions here are relative to a LINUX environment. There will be some differences if you are working on a non-LINUX system. We will not support any other development environment.
This project was built using JAVA 7. It is recommended that you install JAVA 7 when working with this project.
- To build the project:
$ ./clean.sh $ ./make.sh $ ./database create
NOTE: Be sure to initialize the database before running the server, as it will
cause an error if otherwise.
-
To test the project:
$ ./test.sh -
To run the project you will need to run both the client and the server. Run the following two commands in separate shells:
$ ./run_server.sh <team_id> <team_secret> <port> <persistent-dir> $ ./run_client.sh <host> <port>
NOTE: You can run the server and admin with just the "test" parameter
($ ./run_server test) to execute with default values.
You must specify the following startup arguments for `run_server.sh:
+ `<team_id>` and `<team_secret>`: a numeric id for your team, and a secret
code, which are used to authenticate your server with the Relay server.
You can specify any integer value for `<team_id>`, and a value expressed
in hexadecimal format (using numbers `0-9` and letters in the range
`A-F`) for `<team_secret>` when you launch the server in your local setup
since it will not connect to the Relay server.
+ `<port>`: the TCP port that your Server will listen on for connections
from the Client. You can use any value between 1024 and 65535, as long as
there is no other service currently listening on that port in your
system. The server will return an error:
```
java.net.BindException: Address already in use (Bind failed)
```
if the port is already in use.
+ `<persistent-dir>`: the path where you want the server to save data between
runs.
The startup arguments for `run_client.sh` are the following:
+ `<host>`: the hostname or IP address of the computer on which the server
is listening. If you are running server and client on the same computer,
you can use `localhost` here.
+ `<port>`: the port on which your server is listening. Must be the same
port number you have specified when you launched `run_server.sh`.
All running images write informational and exceptional events to log files.
The default setting for log messages is "INFO". You may change this to get
more or fewer messages, and you are encouraged to add more LOG statements
to the code. The logging is implemented in codeu.chat.util.Logger.java,
which is built on top of java.util.logging.Logger, which you can refer to
for more information.
In addition to your team's client and server, the project also includes a
Relay Server and a script that runs it (run_relay.sh).
This is not needed to get started with the project.
All the source files (except test-related source files) are in
./src/codeu/chat. The test source files are in ./test/codeu/chat. If you
use the supplied scripts to build the project, the .class files will be placed
in ./bin. There is a ./third_party directory that holds the jar files for
JUnit (a Java testing framework). Your environment may or may not already have
this installed. The supplied scripts use the version in ./third_party.
Finally, there are some high-level design documents in the project Wiki. Please review them as they can help you find your way around the sources.
The major project components have been separated into their own packages. The
main packages/directories under src/codeu/chat are:
Classes for building the two clients (codeu.chat.ClientMain and
codeu.chat.SimpleGuiClientMain).
Classes for building the server (codeu.chat.ServerMain).
Classes for building the Relay Server (codeu.chat.RelayMain). The Relay Server
is not needed to get started.
Classes that are shared by the clients and servers.
Some basic infrastructure classes used throughout the project.
Training data for our machine learning model was gathered from the "Data For Everyone" dataset library at © Crowdflower Inc. (https://www.crowdflower.com/data-for-everyone)
https://opennlp.apache.org/ Used to tag the messages and sentiment analysis.
https://www.sqlite.org/ Used to create the database.