Test Runner Syntax Options

At the prompt, enter:

test-runner [options] [test-suite] 

where:

test-runner

The name of the executable that will run the test suite, and then produce the test results. It can be one of:

• mfurun (Windows) - for use with procedural COBOL
• mfurunj (Windows) - for use with JVM COBOL
• cobmfurun32[_t](UNIX) - for use with procedural COBOL
• cobmfurun64[_t] (UNIX) - for use with procedural COBOL
• cobmfurunj (UNIX) - for use with JVM COBOL

The optional _t runs the test suite in the multi-threaded run-time system, and is required for tests that use a timeout value.

options

The options that apply to the test run. These are optional, and you can apply one or more of the following:

-testcase:testcase-name

Only run the specified test case from the test suite; all other test cases are skipped. Not set by default.

Alternatively, you can specify multiple test cases using one of the following approaches:

• List the test cases using comma as a separator:

  -testcase:testcase-name1,testcase-name2,... everything.dll

  Or:

• Specify a side file that includes a list of test cases:

  -testcase:@filelist.txt everything.dll

Where:

• everything.dll - all your source code compiled
• filelist.txt - the file that includes a list of all test cases to execute

-ignore-options-in-mfu:{true|false}

Ignores any options in the specified test fixture file, enabling you to override them with ones specified on the command line.

-ignore-return-code:{true|false}

When set to true, the return-code indicating a test pass or fail is ignored, and the test continues running. The default is false.

-verbose:{true|false}

Displays verbose output to the screen. The default is true.

-process:{single|separate}

Use a single process for the entire test run, or multiple separate processes (one for the test runner and one for each test case within the run). If you are using separate processes, the parent test runner process can log errors and continue with the test run in the event of a test case failure. The default is separate.

-isolate:{true|false}

Isolates each test case, so that any resources shared between test cases start from their correct initial state for each test case. The default is true.

-report{printfile|noprintfile|junit|nojunit}...

Outputs the test results to a .txt file (printfile option) or .xml file (junit option). The default is to produce a .txt file. The JUnit report takes the form of TEST-<package-name>-<test-case-name>.xml. Set <package-name> using the -junit-packname:package-name option.

-generate-flist:filename

Outputs a list of failed test names to filename.

Tip: You can use the generated file to rerun the failed tests by using the -testcase:@filename test-suite syntax.

-report:markdown

Outputs the test results to a GitHub style markdown format file (.md).

Note: You can use various external utilities such as the pandoc utility (see http://pandoc.org/) to convert the markdown file into an .html or a .pdf file. Here are some example commands for converting the .md file when you use pandoc:

pandoc -thtml5 -s -S --toc -c pandoc.css -fmarkdown_github mfumeta.md -o mfumeta.html

Or

pandoc -s -S --toc -c pandoc.css -fmarkdown_github mfumeta.md -o mfumeta.pdf

-report:timings-csv

Generates a .csv file containing the timings for the test run.

-reportfile:filename

Changes the name of the .txt report file. It defaults to <test suite-name>-report.txt.

-junit-packname:package-name

Packages the JUnit formatted .xml files under package-name. Not set by default.

-outdir:directory

Specifies a directory for the reports. If not specified, the reports are created in the directory in which the tests were run.

-trait:trait-name(s)

Specifies that only test cases with the specified trait name should be run (and all other test cases skipped). You can specifiy more than one trait, separated by a comma, if required. You set a trait for a test case by using the MFU-MD-TRAITS variable in your source code, or specifying the traits=trait-name in the text fixture file.

Tip: You can combine this option with the -generate-mfu option to create a test fixture file designed to only run test cases with a particular trait; for example:

Windows:

mfurun -generate-mfu -trait:smoke test-suite.dll

UNIX:

cobmfurun32|64[_t] -generate-mfu -trait:smoke test-suite.so

-jenkins-ci

Enables support for producing test results that will be used on a Jenkins CI server.

-jenkins-ci:junit-attachments

Enables support for producing test results that will be used on a Jenkins CI server, and also enables the Jenkins plugin for JUnit formatted results.

-silk-central

Outputs the results to an .xml file that is compatible with Silk Central. The output.xml file, along with any required assets (log and dump files), are placed in the location determined by the SCTM_EXEC_RESULTSFOLDER environment variable, where they can be picked up and processed by Silk Central (refer to your Silk Central Help for more information on ProcessExecuter Tests). If you specify this option when your Silk Central environment is not correctly configured, you receive an error.

-jit:{core|debug}

Produces a core dump file or invokes just-in-time debugging when a test case errors.

Note: This option has no effect when asserting a test failure.

-debugbreak

Immediately starts the debugger.

-debugstart:id

Issues a start/wait command to the debugger using the specified id.

-generate-mfu filename

Produces a skeleton test fixture file (.mfu file) from the test cases within filename (a .dll file or .so file).

-high-res-timer:{true|false}

When set to true, a native high-resolution performance counter is used to report the duration of the test run. The default is false.

Note: High-resolution timings cannot be displayed in the junit or Silk Central .xml formats; normal timings are always displayed.

test-suite

A COBOL program compiled to .dll (Windows) or .so (UNIX), or an .mfu test fixture file. For procedural managed program, specify a .jar file instead.