Edit

Run Gauge specifications

Setup

macOS

JavaScript

VS Code

Modify

This page provides information about how to run a specification, multiple arguments that can be used with the gauge run command when a spec is executed, verbose reporting, data driven execution, parallel execution of specs, how to rerun failed and flaky tests, errors that occur when a spec is run, and troubleshooting information.

For information about what a specification is and related details, see Write Gauge specifications.

Run a specification

You can run a Gauge specification by using the gauge run command. When this command is run, Gauge scans the directories and sub-directories at <project_root> (location at which the Gauge project is created) and picks up valid specification files.

Prerequisite

You must have already created specification(s) at the specs directory or configured a directory(s) of your choice. For more information about specs directory, see Specs directory.

You can use the following command at <project_root> to run a Gauge specification:

gauge run [args] [flags]
  • <project_root> - location at which a Gauge project is created

  • [args] - directories in which specifications are stored, location of specification files and scenarios

  • [flags] - options that can be used with this command such as --tags, -e, -f, and so on

For more information about gauge run command, see the Gauge manual page.

Note

Gauge specifications can also be run from within the IDE (VS Code)

Multiple arguments passed to gauge run

You can choose to run specifications at the directory, file, or scenario level or a combination of these by passing appropriate arguments to the gauge run command. The arguments are separated with a space when used with the command.

The arguments can be one of the following:

  • Directories and sub-directories that contain specifications

  • Path to specification files and scenarios

  • Combination of the arguments mentioned in this list

Run specifications by using directory(s) as an argument

Specifications can be run from a single directory, sub-directories, or multiple directories. You can specify the directory(s) or path to sub-directory(s) with the gauge run command.

To run all the specifications in the specs directory, use the following command:

gauge run specs

To run specifications at the sub-directory level, use the following command:

gauge run <path_dir>

<path_dir> - location of the sub-directory where specifications are stored

To run specifications stored in multiple directories (including sub-directories), use the following command:

gauge run <dir1> <path_dir2> <path_dir3>

<path_dir2> <path_dir3> - location of the directories where specifications are stored

Example - single directory

In the following example, specifications stored at test_specs are run:

gauge run test_specs

Example - multiple directories

In the following example, specifications stored at specs and test_specs are run:

gauge run specs test_specs

Example - sub-directories

In the following example, specifications stored at sub1_specs and sub2_specs are run:

gauge run specs/sub1_specs specs/sub2_specs

sub1_specs, sub2_specs - directories located in specs

Run specifications by using spec file path as argument

You can choose and run only certain specifications by providing the appropriate location of these specifications with the gauge run command.

To run a single specification, use the following command:

gauge run <path_to_spec>

<path_to_spec> - location of the specification

To run multiple specifications, use the following command:

gauge run <path_to_spec1> <path_to_spec2> <path_to_spec3>

Example - run a single specification file

In the following example, Gauge runs the example.spec stored in the specs directory:

gauge run specs/example.spec

Example - run multiple specification files

In the following example, Gauge runs multiple specifications stored in specs and its sub-directory, sub1_specs:

gauge run specs/example.spec specs/sub1_specs/sub1_example.spec

Run specifications by using path to scenario as argument

You can choose to run only scenarios of a specification. This can be either a single scenario or multiple scenarios. The argument to the gauge run command can be a specific scenario or a list of scenarios and the line number in the span of that scenario. Any line number which the scenario spans across can be used.

Note

For more information about scenarios, see Scenario.

To run a single scenario, use the following command:

gauge run <specification_path>:<scenario_line_number>

To run multiple scenarios, use the following command:

Note

Multiple scenarios can belong to different specifications.

gauge run <specification_path>:<scenario_line_number> <specification_path>:<scenario_line_number>...

Example

Consider the following specification, spec1.spec located at specs directory:

1# Configuration
2
3## Admin Login
4* User must login as "admin"
5* Navigate to the configuration page
6
7## User Login
8* User must login as "user1"
9* Navigation to configuration page is restricted.

Single scenario

To run only the second scenario, User Login, of spec1.spec, use the following command:

gauge run specs/spec1.spec:7

Line 7 indicates that the second scenario is run. As this scenario spans from line 7 to 9, any line number including and between 7 and 9 can be used. Hence, you can also mention line 9 in the command as follows:

gauge run specs/spec1.spec:9

In both cases, the User Login scenario is run.

Multiple scenarios

In addition to spec1 used as an example previously, let us consider another spec, test1, in the specs directory as follows:

1# Search specification
2
3## Successful search
4* Log in as "admin"
5* Open the product search page
6* Search for product "Cup Cakes"
7* "Cup Cakes" should show up in the search results

To execute scenarios from both spec1 and test1, run the following command:

gauge run specs/spec1.spec:3 specs/test1.spec:4

In this case, Admin Login scenario is run from spec1.spec and Successful search scenario is run from test1.spec.

Note

test1.spec can be located in a directory other than specs.

Filter specifications and scenarios by using tags

Tags allow filtering specs and scenarios. You can use the --tags flag with the gauge run command to filter specs and scenarios.

Prerequisite

You must have already labelled the specs and scenarios with tags. For more information about Tags, see Tags.

Use the following command to filter a Gauge specification or scenario by using tags:

gauge run --tags "Tag_Name" specs

Tag_Name - tag associated with the specs or scenarios

When this command is run, only the scenarios and specifications which are tagged with Tag_Name are executed.

Example

Consider the following spec in the specs directory, which has tags search and admin. The scenario, Successful search, is tagged with successful and the scenario, Unsuccessful search has no tags.

 1# Search Specification
 2
 3The admin user must be able to search for available
 4products on the search page.
 5
 6Tags: search,  admin
 7
 8* User must be logged in as "admin"
 9* Open the product search page
10
11## Successful search
12
13Tags: successful
14
15For an existing product name, the search result will
16contain the product name.
17
18* Search for product "Cup Cakes"
19* "Cup Cakes" should show up in the search results
20
21## Unsuccessful search
22
23On an unknown product name search, the search results
24will be empty
25
26* Search for product "unknown"
27* The search results will be empty

If the tag search is used, then the entire spec is run.

gauge run --tags "search" specs

If the tag successful is used, the scenario, Successful search is run.

gauge run --tags "successful" specs

Execution hooks can also be filtered based on tags. For more information, see Write Gauge specifications.

Tag expressions

Tags can be used with expressions. This helps you search and filter specs and scenarios effectively. The following table lists the tags and expressions and their corresponding action while selecting specs and scenarios.

Attention

In the command line, while using the not symbol (!) with tags, (!) has to be preceded by escape (\).

Tags

Selects specs/scenarios that

!TagA

do not have TagA

TagA & TagB or TagA,TagB

have both TagA and TagB.

TagA & !TagB

have TagA and not TagB.

TagA | TagB

have either TagA or TagB.

(TagA & TagB) | TagC

have either TagC or both TagA and TagB

!(TagA & TagB) | TagC

have either TagC or do not have both TagA and TagB

(TagA | TagB) & TagC

have either [TagA and TagC] or [TagB and TagC]

Example

Consider the spec of the previous example - if all the scenarios tagged with search and successful must be run, then use the following command:

gauge run --tags "search & successful" specs

Based on the Tags expressions table, Successful Search scenario is run.

Verbose reporting

By default, Gauge generates reports that display whether a spec has passed or failed when executed. To ease debugging, reports can also be generated for every step to display whether a step has been executed successfully. Such reports can be generated by using the --verbose flag. These reports are generated on the console.

Use the following command to generate reports at the step level:

gauge run --verbose specs

Data driven execution

A data table is defined in Markdown table format at the beginning of the spec prior to steps. The data table should have a header row and one or more data rows. The header names from the table can be used in the steps within angular brackets < > to refer to a particular column from the data table as a parameter.

When a spec is run, each scenario is executed for every data row of the table. Table parameters are written in Multi-Markdown table formats.

Example

In the following specification hello.spec, the data table is defined at the beginning of the spec. The step uses the name column from the data table as a dynamic parameter. When the spec is run, Scenario and Second Scenario are executed first for the first row values 1, Alice followed by the second and third row values from the table.

 1# Table driven execution
 2
 3     |id| name    |
 4     |--|---------|
 5     |1 |Alice    |
 6     |2 |Bob      |
 7     |3 |Eve      |
 8
 9## Scenario
10* Say "hello" to <name>
11
12## Second Scenario
13* Say "namaste" to <name>

Selected data table rows

By default, scenarios in a spec are run for every data table row. Scenarios can also be run against selected data table rows by using the --table-rows flag along with specifying the row numbers for which the scenarios should be run. If there are multiple row numbers, the row numbers should be separated by commas. A range of table rows can also be specified.

Important

Only a single specification can be run while using the --table-rows flag.

Examples

In the following example, the scenarios in hello.spec (see Data driven execution) are run only for the first row of the data table.

gauge run --table-rows "1" specs/hello.spec

In the following example, multiple rows are specified by separating them with commas. The scenarios from the hello.spec are run for the first and third rows of the data table.

gauge run --table-rows "1,3" specs/hello.spec

In the following example, a range of table rows is specified. The scenarios from the hello.spec are run for the first, second, and third rows of the data table.

gauge run --table-rows "1-3" specs/hello.spec

External CSV for data table

Data Tables for a specification can also be passed from an external CSV file.

For more information about external CSV files used in data tables, see Parameters.

Example

In this example, users.csv is the external CSV file that contains the following data table:

id,name
1,Alice
2,Bob
3,Eve

In the spec, the steps use the <name> column from the CSV file.

1# Table driven execution
2
3table: /system/users.csv
4
5## Scenario
6* Say "hello" to <name>
7
8## Second Scenario
9* Say "namaste" to <name>

Configuring table data per environment

New in version 1.4.0.

Sometimes there are specifications that require input that depend on the environment that they execute against. In such cases the table cannot be inline (i.e. defined in the spec). The external table referenced cannot be a single file either. To accomodate this situation, Gauge allows a property gauge_data_dir which can be set to a location where Gauge looks for data files.

This property can be set in the property files and can be overridden per environment, thus allowing the project to have data files that are specific to each environment.

Note

When set only in default.properties or as an environment variable gauge_data_dir affects all the execution uniformly.

Note

This property is optional, and defaults to the Project Root location.

Example

The following example shows an env directory structure with two additional environments defined.

├── data
├── env
  └── default
     └── default.properties
  └── qa
     └── qa.properties
  └── uat
     └── uat.properties
qa.properties
gauge_data_dir = data/qa
uat.properties
gauge_data_dir = data/uat
1# Table driven execution
2
3<table:users.csv>
4
5## Scenario
6* Say "hello" to <name>
7
8## Second Scenario
9* Say "namaste" to <name>

When the above spec is executed using --env=qa, the users.csv is picked up from <PROJECT_ROOT>/data/qa and likewise for --env=uat the users.csv is resolved from <PROJECT_ROOT>/data/uat.

Parallel execution

Specs can be executed in parallel to run the tests faster. Running tests in parallel creates a number of execution streams depending on the number of CPU cores available on your system and distributes the load among worker processes.

The number of parallel execution streams can be specified by using the -n flag.

Note

It could lead to undesirable results if the number of streams specified is more than the number of CPU cores available on your system. For optimizations, you can also use threads. See Parallel execution by using threads.

Use the following command to run specs in parallel:

gauge run --parallel specs

OR

gauge run -p specs

Example

In the following example, four parallel execution streams are created.

gauge run --parallel -n=4 specs

Parallel execution by using threads

In parallel execution, every stream starts a new worker process. This can be optimized by using multithreading instead of processes. Multithreading uses only one worker process and starts multiple threads for parallel execution.

To use the multithreading feature, the enable_multithreading environment variable must be set to true. If not already present, you can add this variable to the default.properties file. For more information about default.properties, see Local configuration of Gauge (default.properties).

Prerequisites

Use thread-safe test code. Use a language runner that supports multithreading.

Note

Currently, only the Java and .NET language runners supports parallel execution of specs by using threads.

Run specs before parallel execution

Experimental Feature

Specs can be filtered based on tags to run in parallel, so that only those will be run in parallel after executing other specs in serial.

Set allow_filtered_parallel_execution variable to true in /env/default/default.properties to enable this feature.

Execution can be done by the command:

gauge run specs --parallel --only "parallelizable"

This runs all the specs that are not tagged as parallelizable in serial first, and then runs the tagged ones in parallel.

Note

This feature is currently available in gauge >= 1.0.5

Test suite execution by using the --strategy option

The --strategy option allows you to set the strategy for parallel execution of tests. This option has two values: lazy and eager. By default, the option is set to lazy.

lazy

The lazy feature enables Gauge to dynamically allocate specs to streams during execution instead of at the beginning of execution. This allows Gauge to optimise the resources on your system or execution environment. Such optimization is useful because some specs might take more time to get executed than the others. This could be either because of the number of scenarios in the specs or the nature of the feature under test.

lazy is the default value of the --strategy option.

The following command assigns tests lazily across the specified number of streams:

gauge run -n=4 --strategy="lazy" specs

OR

gauge run -n=4 specs

Note

The lazy value cannot be used when the -g flag is used with the gauge run command. This is because the grouping of tests depends on allocation of tests before the beginning of test execution, however, lazy is used during execution of tests. Using the -g flag with --strategy=lazy has no impact on your test suite execution.

Example

If there are 100 tests, which have to be run across four streams or cores, Gauge dynamically assigns the next spec in queue to the stream that has completed its previous test execution and is waiting for more work.

eager

When the -g (grouping) flag is used, the value of the strategy option is eager. In this strategy, Gauge allocates specs to streams at the beginning of test execution.

Example

When eager is used, if 100 tests are run, these tests are equally distributed before execution in the number of streams as mentioned by the -n option.

gauge run -n=4 --strategy="eager" specs

Executing a group of specification

Gauge sorts the specifications by alphabetical order and then distributes these specifications into groups. You can use the --group | -g flag to execute a specific group of specs.

Executing specs with the -n and --g flags ensures that Gauge executes the same group of specifications in the same order regardless of the number of times the gauge run command is executed.

Use the following command to execute a group of specifications:

gauge run -n=<number_of_groups> -g=<group_number> specs
  • -n - number of groups

  • -g - group number

Example

In the following example, Gauge creates four groups of specification and selects the second group for execution.

gauge run -n=4 -g=2 specs

Even if the command is run multiple times, Gauge still executes the same group of specifications in the same order.

Rerun failed scenarios

Gauge provides the ability to rerun only the scenarios which failed in the previous execution. You can use the --failed flag with the gauge run command to rerun failed scenarios.

Prerequisite

You must have already run the specifications by using the gauge run command.

Use the following command to rerun failed scenarios:

gauge run --failed

For example, if three scenarios failed during gauge run specs, the failed scenarios can be rerun instead of executing all scenarios.

When the --failed flag is used with the gauge run command, the flags that were set during the previous execution is once again set.

Example

Consider an example, where specs are run with a --env and --verbose flags. Three scenarios fail during this run.

gauge run --env="chrome" --verbose specs

To rerun only the failed scenarios, use the following command:

gauge run --failed

When this command is run, Gauge internally sets the --env and --verbose flags to corresponding values used in the previous execution. Hence, gauge run --failed is equivalent to the following command:

gauge run --env="chrome" --verbose specs <path_to_failed_scenarios>

Rerun failed scenarios by using --max-retries-count

You can use the --max-retries-count flag to rerun failed tests for a specific number of times.

Note

--max-retries-count feature is also useful if there are flaky tests in your test suite.

Use the following command to rerun failed tests for a specific number of times:

gauge run --max-retries-count=<number of retries>

Example

In the following example, Gauge reruns a failed test for a maximum of three times and then marks the spec as failed.

gauge run --max-retries-count=3

Rerun failed scenarios by using --retry-only

You can filter scenarios that must be rerun a specific number of times when failed by using the --retry-only flag. The value of this flag is the tag used to associate the scenario(s).

Use the following command to rerun failed scenarios for a specific number of times:

gauge run --max-retries-count=<number of retries> --retry-only "<tag_name>"

<tag_name> - name of the tag used to label the scenario(s) that should be rerun when failed

Note

Tags can also be used with expressions. For more information about using tags with expressions, see Tag expressions .

Example

In the following example, Gauge reruns only those scenarios that have the should-retry tag . Gauge runs these scenarios thrice as specified by the --max-retries-count flag.

gauge run --max-retries-count=3 --retry-only="should-retry"

Note

If --retry-only flag is not specified, all scenarios are retried the number of times as specified in --max-retries-count.

Errors during execution

Parse errors

This occurs if the spec or concept file doesn’t follow the expected specifications or concepts syntax.

Example

[ParseError] hello_world.spec : line no: 25, Dynamic parameter <product> could not be resolved

List of various Parse errors:

Parse Error

Gauge Execution Behaviour

Step is not defined inside a concept heading

Stops

Circular reference found in concept

Stops

Concept heading can only have dynamic parameters

Stops

Concept should have at least one step

Stops

Duplicate concept definition found

Stops

Scenario heading is not allowed in concept file

Stops

Table doesn’t belong to any step

Ignores table,Continue

Table header cannot have repeated column values

Marks that spec as failed,Continues for others

Teardown should have at least three underscore characters

Marks that spec as failed,Continues for other

Scenario heading should have at least one character

Marks that spec as failed,Continues for other

Table header should be not blank

Marks that spec as failed,Continues for other

Multiple spec headings found in the same file

Marks that spec as failed,Continues for other

Scenario should be defined after the spec heading

Marks that spec as failed,Continues for other

Could not resolve table from file

Marks that spec as failed,Continues for other

Spec does not have any element

Marks that spec as failed,Continues for other

Spec heading not found

Marks that spec as failed,Continues for other

Spec heading should have at least one character

Marks that spec as failed,Continues for other

Dynamic param could not be resolved

Marks that spec as failed,Continues for other

Step should not be blank

Marks that spec as failed,Continues for other

Duplicate scenario definition found in the same specification

Marks that spec as failed,Continues for other

Validation Errors

These are errors for which Gauge skips executing the scenario where the error occurs.

There are two types of validation error which can occurs

  1. Step implementation not found

    If the spec file has a step that does not have an implementation in the projects programming language.

  2. Duplicate step implementation

    If the spec file has a step that is implemented multiple times in the projects.

Example

[ValidationError] login.spec:33: Step implementation not found. login with "user" and "p@ssword"
[ValidationError] foo.spec:11 Duplicate step implementation => 'Vowels in English language are <table>'