test_run

Arguments

module

A String array. This input argument must be

• the name of an internal Scilab module ("core", "time", ...)

• the name of an ATOMS module ("module_lycee", "nisp", ...). To be taken into account, the module must be loaded when test_run() is called.

• the absolute path of a module

test_name

A string array

options

A string array

• no_check_ref: does not check if the .dia and .dia.ref are equal

• no_check_error_output: The error output stream is not checked. This option can be used when Scilab complains about the localization being not available.

• create_ref: create the .dia.ref file and does not check if the .dia and .dia.ref are equal

• list: does not perform the tests but displays a list of available tests

• help: display some examples about how to use this command

• mode_nw: add the "-nw" option to the launch

• mode_nwni: add the "-nwni" option to the launch

• nonreg_test: runs only the non-regression tests, skipping unit tests

• unit_test: runs only the unit tests, skipping non-regression tests

• skip_tests

• enable_lt: Enable long-time execution tests

Description

Search for .tst files in the unit test and non-regression test library execute them, and display a report about success of failures. The .tst files are searched in directories SCI+"/modules/*/tests/unit_tests" and SCI+"/modules/*/tests/nonreg_tests". Whenever a test is executed, a .dia file is generated which contains the full list of commands executed along with message which appears in the console. When the script is done, the .dia file is compared with the .dia.ref file which is expected to be in the same directory as the .tst file. If the two file are different, the test fails.

Special tags may be inserted in the .tst file, which help to control the processing of the corresponding test. These tags are expected to be found in Scilab comments.

These are the available tags:

• <-- INTERACTIVE TEST --> This test will be skipped because it is interactive.

• <-- LONG TIME EXECUTION --> This test will be skipped because it needs long-time duration. To enable the test, call test_run with the following option: "enable_lt"

• <-- NOT FIXED --> This test will be skipped because it is a known, but unfixed bug.

• <-- TEST WITH GRAPHIC --> This test will not be executed if the option "mode_nwni" is used.

• <-- NO TRY CATCH -->

• <-- NO CHECK ERROR OUTPUT --> The error output file is not checked

• <-- NO CHECK REF --> The .dia and the .dia.ref files are not compared.

• <-- ENGLISH IMPOSED --> This test will be executed with the -l en_US option.

• <-- FRENCH IMPOSED --> This test will be executed with the -l fr_FR option.

• <-- JVM NOT MANDATORY --> This test will be executed with the nwni mode by default.

• <-- WINDOWS ONLY --> If the operating system isn't Windows, the test is skipped.

• <-- UNIX ONLY --> If the operating system isn't an Unix OS, the test is skipped.

• <-- LINUX ONLY --> If the operating system isn't GNU/Linux, the test is skipped.

• <-- MACOSX ONLY --> If the operating system isn't Mac OS X, the test is skipped.

• <-- XCOS TEST --> This test will launch all the necessary Xcos libs. This test will be launched in nw mode.

Each test is executed in a separated process, created with the "host" command. That enables the current command to continue, even if the test as created an unstable environment. It also enables the tests to be independent from one another.

Platform-specific tests

It may happen that the output of a test depends on the platform on which it is executed. In this case, the .ref file cannot be correct for all platforms and unit tests may fail for some platform. In this case, we can create a default .ref and create additionnal .ref file for each platform.

The various platform-specific .ref files must have one of the following extensions.

• .unix.dia.ref for Unix platform,

• .linux.dia.ref for GNU/Linux platform,

• .win.dia.ref for Windows platform,

• .macosx.dia.ref for Mac OS X platform.

The algorithm is the following. First, the .ref is considered. If this file does not exist, the platform-specific .ref file is examined depending on the current platform.

• on Windows platforms: .win.dia.ref,

• on Max OS X platforms: .unix.dia.ref, .macosx.dia.ref,

• on GNU/Linux platforms: .unix.dia.ref, .linux.dia.ref.

Examples

// Launch all tests
// =============================================

test_run();
test_run([]);
test_run([],[]);

// Test one or several module
// =============================================

// Test one module
test_run('time');

// Test several modules
test_run(['time','string']);

// Test a submodule
test_run(['optimization|neldermead']);

// Refer to a module by its path
test_run(SCI+'/modules/core');

// Launch a specific test
// =============================================

// One specific test
test_run('time','datenum');

// Several tests
test_run('time',['datenum';'calendar']);

// Skip some tests
// =============================================

test_run('time',['datenum';'calendar'],'skip_tests');

// Options
// =============================================

// does not check if the .dia and .dia.ref are equal
test_run('time','datenum','no_check_ref');

// Create the .dia.ref file and does not check if the .dia and .dia.ref are equal
test_run([],[],'create_ref');

// Does not perform the tests but displays a list of available tests
test_run([],[],'list');

// Display some examples about how to use this command
test_run([],[],'help');

// Runs only the non-regression tests, skipping unit tests
test_run([],[],'nonreg_test');

// Runs only the unit tests, skipping non-regression tests
test_run([],[],'unit_test');

// Do not check the error output (std err)
test_run('boolean','bug_2799','no_check_error_output');

// Combine several options
test_run([],[],['no_check_ref','mode_nw']);

Internal Design

The tests are performed in the temporary directory, not in the directory which originaly contain the tests files. The .tst file is copied into the temporary directory, the test is performed and the .dia.ref is copied back into the original location.

The .tst script is not run as is. Instead, a header and a footer are inserted at the beginning and at the end of the .tst at the time the script is copied into the temporary directory. The role of this modification is to redirect the output messages into the .dia file, so that the user can have a log file once the test is performed.

Authors

• Pierre Maréchal
• Michael Baudin