Who need to read this blog

This article is NOT ONLY for who is new to IBM DevOps Test (including IBM DevOps Loop and DevOps Test Hub), and want to understand how to start to use these tool, BUT ALSO for who has the entry level of skills with our test tool, but from time to time confused on:

• how to use this feature and that feature
• want to know why there is a feature like this and why I need this
• want to do some "real" work but got some problem
• want to know whether there is some feature I missed.

This article is NOT a marketing article because it's focus on technology side. 

What you may missed

• Technical guide: Testing using DevOps Loop (1)  -- overview and basic recording

Now let's start the journey here.

1. Test Data overview

Test Data is an important new feature brought in DevOps Loop (and Test Hub). Different from web client testing, it's a new feature not existing in thick-client IDE testing. IDE only support basic test data support, like the "static" dataset. Benefit from the central server design, the test data can be kept and shared between users. It enable Test Data to evolve powerful features which we will elaborate in this blog and the next ones.

1.1 Test data concepts

In DevOps Loop (and DevOps Test Hub), there are different test data elements as below:

• External cvs file

• Dataset

• Data definition

• Schema
• Database
• Data Fabrication
• Catalog
• Generator

 We will elaborate these elements of Test Data one by one in the following sections

1.2 Data Fabrication

Data fabrication in software testing is the process of creating synthetic test data that imitates real production data.

It helps testers validate systems safely without exposing sensitive or regulated information.

This approach improves testing coverage, privacy compliance, and efficiency across test environments.

See the details:

Test data is the core component of any application testing. It is very time-consuming to create test data manually. You can use the data fabrication feature in IBM® DevOps Test Hub (Test Hub) to create realistic datasets for comprehensive testing scenarios. The feature provides data fabrication capabilities that can help you to generate synthetic data quickly and efficiently without the risk of data leaks or privacy issues.

more

1.3 Version Reference

All the features discussed by this series of blogs is based on:

• IBM DevOps Loop v1.0.3
• IBM DevOps Test Hub v11.0.6

Note: The product features mentioned in this article are subject to change due to the rapid evolution of the product

2. Walk through test data elements

2.1 External CSV file

It means the cvs format data files stored outside of DevOps Loop, like in your laptop. They can be imported into DevOps Loop. 

2.2 Dataset

This is similar to cvs file. But it's a kind of "internal" cvs file. It defines data fields, and the data rows as well. The dataset is the most common way to provide data to test script, in a data-driven-test. 

create it:

• External CSV files
• Database Query
• Data Definition (Fabricate Data)

How to use it:

select the data source in test script level, then assign the variables to match it; 

select the data source in test suite level, then assign the suite variable to match it.

demonstration:

We will demo how to create dataset, and how to use dataset in test script.

2.3 Data definition, Catalog and Generator

2.3.1 What data definition consist of?

Data definition has 2 parts of information:

• field definition

The definition of data structure, like string, enum, number, etc

• data definition

It contains the format/restriction of the data in each fields, like the number must be 0-4 digits.

Both of these are configured and defined by Generator. 

Data definition does NOT have:

• data entry

There is no data rows inside of a data definition

2.3.2 What data definition was used for?

The only purpose the data definition used for, is to Data Fabrication. If you don't need Data Fabrication, you don't need Data Definition.

2.3.3 How to use Data Definition

create it by:

• External CSV files

You can create Data Definition by importing a CSV file. 

But you can only generate field names, not the definition on data and field. You still need to create Generator later.

• Database Query

But you can only generate field names, not the definition on data and field. You still need to create Generator later.

• Dataset

You can create Data Definition by importing a Dataset.  

But you can only generate field names, not the definition on data and field. You still need to create Generator later.

• Create from scratch using Generator

This is the most common way to create a Data Definition. We will have more discussion on this in the section for Generator.

How to use it:

• generate Fabricated data as a Dataset, then used in test script indirectly
• generate Fabricated data to use in test script directly

The difference of these 2 ways is just depend on whether you want to reuse fabricated data.

If you want to reuse the generate data, save it in a dataset. You can create a new dataset, by creating via "Fabricate data"

2.3.4 Catalog

Catalog is a simple concept. It's just the "container" of Generator. You can define different Catalog for different logic purpose.

2.3.5 Generator

Generator is used to describe what kind of data will be generated for one field, including the format, the restriction on the content value, and some other logics.

Basic Generator

Basic generators are the built-in generators of the Data fabrication feature of Loop Test that you can use to generate diverse test data. These generators provide commonly used data types and patterns that you can select and configure in a Data definition to generate realistic test data.

Custom generators

Custom generators are what you can created using the Basic Generator as template. You can configure it in the pre-defined scope of the original generator. You can't add new feature which the original generator didn't support.

3. Demo

In this video, we will demo how to use the dataset and data definition in testing. It includes:

• how to create dataset from script variable directly.
• how to create a data definition from scratch
• how to run a data-driven test using data from a dataset

4. More to come

This blog is focused on the overview of Test in Loop, and how to create web UI testing script by recording. As part of a series blogs, we have more topics to cover.In the following blogs, we will introduce:

• Enhance your test script with Data Driven testing.
• Test management
• Test fabrication
• Test data analysis and visualization 

We will update the new blogs link as below:

• Technical guide: Testing using DevOps Loop (1) -- overview and basic web client recording demo

• Technical guide: Testing using DevOps Loop (3) -- Test Data management (continued)