Download the PHP package ben-rowan/vcsv-stream without Composer

vcsv://stream

A virtual CSV stream for creating unit / integration test fixtures.

Install

Modes

You can generate CSV fixtures in two modes with vcsv://stream, online and offline.

Online

You can use vcsv://stream as a direct replacement for your standard file stream. In this mode vcsv://stream will dynamically generate fake CSV data based on the provided configuration.

pros:

• Fast test development / feedback loop.
• Each test run get's a different set of valid data. This means you have the chance to catch unexpected bugs in your code.

cons:

• Higher CPU and memory utilisation when generating large files (I'll see if I can reduce this!)

Offline

You can also use vcsv://stream to generate a CSV fixture on disk. You can then use this later in your test.

pros:

• Fast test execution with reduced CPU and memory load.

cons:

• Slower test development / feedback loop.
• Each test run get's the same data. This reduces the scope of your testing.

Usage

Online

The first step is to call vcsv://streams setup method in your test. This initialises and registers the stream:

Next you load your configuration (see [here]() for config docs):

Finally you can use the stream as you would any other file stream:

Offline

The provided generate:csv command makes this easy:

This will output a CSV file to stdout so you can easily see what's being generated. Once you're happy you can dump it to disk:

Configuration

vcsv://stream uses yaml configuration files to define your CSV fixtures. This has many benifits over simply creating CSV files and adding them to source control:

• You're defining the rules for what get's generated. This means you know you have 100 rows of this type of data followed by 100 rows of another. If you just have the CSV then you have to guess this or get the information from a separate set of documentation.
• You can add comments. This means you can describe the fixture in detail at the top as well as commenting choices you've made throughout. This is a big help.
• You can easily edit the CSV from your IDE / text editor. This reduces the overhead of making changes as you develop your test (this is particularly true for the online mode).

Example

Here's a complete example of some vcsv://stream configuration. We'll go through each section below:

header

This section defines the header for your CSV. It's required even if you choose not to output a header because it defines the default data generators for each of the columns.

include

When set to true the CSV will include a header. When set to false it wont.

columns

This is how we tell vcsv://stream how many columns we'd like to give our CSV and how to generate data for those columns by default.

The name provided for each column will become the columns header if this is enabled. The names must be unique.

vcsv://stream can generate data in three ways:

value Columns

You can provide a fixed value that you'd like this column to contain:

faker Columns

You can randomly generate data using one of the Faker libraries formatters:

By setting unique to true we're telling Faker that we don't want this column to contain any duplicate data.

text Columns

You can also tell vcsv://stream that you'd just like the column to contain lorem ipsum:

records

This defines the actual CSV data. We define the data in 'chunks' meaning we can vary the data throughout the file. Each record has a name, this should be something that helps you remember why it was created.

count

This defines the number of rows that should be created using this set of config.

columns

This defines a set of override data generators. If we don't define a column here then the previously defined default header data generator will be used instead.

Note: the column names here must match one of the header columns.