propathink

propathink

rethinkdb client

rethinkdb client focused on maintainable code organization. Store rethinkdb request logic in modular way by labelling and storing requests in plugins/modules. In theory, propathink's architecture is extendable allowing you to add plugins at will in order to better maintain and manage all db request logic.

configuration centric plugin design

propathink utilizes a manifest file for each project. The manifest configures the database connection and all request plugins to be loaded in a project. Those familiar with hapijs/glue will be comfortable with this design.

provisions rethink connection

Every request built using a propathink plugin is provisioned with the propathink object this and pthinkInternals object. See below for example showing how to use these objects to make a db request. Note: every rethinkdb connection is stored in the this.conn. And, pthinkInternals.db contains the configured database name declared in the manifest.

pthinkInternals.db      // configured database name from the manifest file.
this.conn               // generated rethinkdb connection object

Sample rethinkdb request made below:

return Rethinkdb.dbCreate(pthinkInternals.db).run(this.conn, (err, result) => {

    ....... code handling results .......
});

joi validation

propathink imitates hapi's use of joi to validate parameters set in a request. Sample request below:

{
    name: 'testTwo',
    comment: 'testTwo documentation here.',
    validate: {
        0: { // parameter at arguments[0] to be validated.
            username: Joi.string().min(3).max(6),
            password: Joi.string().min(3).max(8)
        }
    },
    handler: function (user, next, callback) {

             return callback(null, 'test two result', next);
    }
}

Note: The object key determines which parameter will be evaluated by joi in the request. The above valdation configuration applies the joi test to the first parameter in the handler.

request lifecycle

propathink again imitates hapijs by utilizing a request lifecyle for all request logic. The lifecyle has four steps: pre, validate, request, after.

• pre generates the connection for the request.
• validate applies joi validations to appropriate parameters in requests.
• request executes the request built in the handle.
• after runs after everything else is completed.

constructing the handle

The handle element in a propathink request object must have the last two parameters in order for the request lifecycle to work (next & callback): next is executed at the end of each step in the life cycle and needs to be passed to the callback in the handle. It must be executed in the callback your write to be consumed in the handler. This ensures the after step of the lifecyle is really executed "after" the request is completed. If this is not done correctly, the after step in the lifecyce will execute before your database read and writes are completed. Plus, executing next this way allows for monitoring performance and quantity of code execution.

database request monitoring

Excitingly, utilizing the request lifecyle design to construct database requests allows for monitoring of all requests' performance and usage.

influence of hapijs design

propathink is highly influenced by hapijs architecture. After studying hapi's architecture decided to build propathink using a plugin architecture in order to better maintain database request logic and all the other good stuff propathink does.

Test

project requires 100% test coverage. hapijs/lab testing framework and hapijs/code assertian library. Travis is used to enfore and prove 100% test coverage exists in github repo.

Style Guide

Project abides by the hapijs style guide. Style guide is enforced using hapijs/lab testing framework and The framework's linting standards enforce style guide. Travis will fail is linting tests do not pass.