A persona is a meaningful name for a user or a set of data representing a particular situation. In BDD, it can describe a specific configuration, seed-data set, or entity structure—not just a UX user profile. section 17.1
A persona definition provides the data needed to create an instance for automation and lets stakeholders understand the intended behavior while reading a feature file. Record meaningful names and their relevance in the product glossary. section 17.0 section 17.1
Personas can name:
- Specific users, such as an admin user or website guest. section 17.1
- A particular data situation, such as a customer with a massive order history. section 17.1
Table 17.1 Comparison of test data definition strategies
|
Usage of test data
|
Example
|
Recommended strategy
|
|
Incidental data that is necessary to
automate the application, but not relevant to the scenario and the specified
business rule.
|
When an
order is
placed
It is
unimportant what is being ordered.
|
Manage data in the automation code,
for example using the Domain Defaults pattern.
|
|
Data that is relevant (“essential”)
for understanding the scenario
|
When an order for 5 "Large"
pizzas is placed
It is
essential to say how many large pizzas are being ordered.
|
Include the data in the scenario as
step text, data table or docstring. This can be achieved by using step
definition parameters (chapter 12) and the patterns Terminology
Promotion, Data Table Accessor (chapter
14), and Approved Outcome (chapter 16).
|
|
A small persona definition, that is
relevant in a single feature file.
|
Given the
customer has the following orders
| id | status |
| A
| Completed |
| B
| Placed |
| C | Placed
|
|
Include the data in the feature file.
If it is needed by multiple scenarios, specify it as a Background.
The In-place
Persona pattern in this chapter discusses this approach in detail.
|
|
Large persona definitions, where the
details are not required to understand the scenario. Typically, created and
validated by domain experts.
The persona definitions may contain
expected results as well as input data.
|
Given
sales traffic from "sales.csv"
|
Store the data in external sources,
like Excel, CSV, XML or JSON files. Name the external sources in a way that
describes their purpose. The scenario can explicitly load the data by name
and process it.
The External
Persona pattern in this chapter discusses this approach in detail.
|
|
Large persona definitions that
contain mostly incidental data, but that include a few data points that are
essential to understanding the scenario. The complete definition is created
and validated by domain experts.
|
Given sales traffic from
"sales.csv" with
| exclude section |
| truffle sales |
|
Store the definitions in external
sources as a template. The scenario should explicitly specify both the
template to be used and any essential data variations. The automation
generates the final definition from the template and the specified variations.
The Templated
Persona pattern in this chapter discusses this approach in detail.
|
|
Persona definitions that vary for
different executions of the same scenarios, for example verifying the
application behavior described by the scenarios for different market regions.
|
Given the
net price of the Margherita pizza is 8[CUR]
|
Use placeholders in scenarios to
refer to the varying data (e.g., [CUR]) and substitute concrete values for
the placeholders, based on the persona definition provided for each execution
(e.g., use USD if the execution verifies the scenarios using the “US market”
persona).
The Parameterized
Scenario pattern in chapter 20 discusses this approach in detail.
|
Compare test-data strategies by whether the data is incidental to automation, essential to the scenario, or suited to a persona definition. section 17.1










