RADE

What You Will Learn with This Article

This article is intended to help you understand how to create and to run test objects using the mkodt command. If you want to have more information and examples of test objects, see the use cases documentation.

[Top]

The mkodt Command

The mkodt command runs test objects which are delivered in test frameworks. This command gives the choice (by using different options) of running one test object or all test objects found in a given framework or in a workspace. The command is available with mkmk commands. To access to this command, you have to run the mkmk_profile or the adl_profile.

[Top]

Test File Tree

The CAA file tree is composed by framework. To each code framework correspond a test framework which contains a FunctionTests directory; In this directory, you must find everything you need to run test objects.

• File Tree

  

• File Tree Description

  Directory	Description
  TestCases	ODT shell-scripts
  SwitchTestCases	Switch ODT shell-scripts
  InputData	Input of shell-scripts. Every necessary files to run ODTs
  OutputRef	Reference output of ODTs
  Output/$OS	Output generated by mkodt on each operating system

[Top]

Options Available with mkodt

Use the concepts of "workspace" and "project"

• Help mode
  mkodt -h

• Contextual mode

mkodt -D | -s <name> | -a [-tXC]

• Descriptive mode

mkodt -W <Workspace> | -d <path> -F <Framework> | -FWList <fileFWList>  | -s <shell > | -l <File1> | -L <File2> | -a [-StXC] -NOMaxTime -NoPert -ExcludFW <fileFWExcluded> -PriorityFW <fileFWPriority> -o <Output> -p <PATH1:PATH2...> -r <RTPATH1:RTPATH2...> -UpdateInit -EnoviaEnv <Site> -html -xml

[Top]

Environment Variables Evaluated by mkodt

[Top]

Environment Variables Evaluated by the User

[Top]

Test Results: mkodt Output

The mkodt command generates two output files in the Output/$OS directory. These files are:

1. xxx.traces: contains traces generated by the shell script.
   Example: The shell script test1.sh runs two executable which contain two cout:

   SetOdtParam max_time=2 Mod Mod2 exit$?

   test1.traces generated by mkodt:

   

2. xxx.mkodt_result: contains result of the ODT replay. The file format is:

   FrameworkName -f ODTName OS ReturnCode date ? ODTType Pertinence Elapsed_time max_time

   where:

   • ODTType = S if the ODT is a Switch ODT, otherwise ODTType = .
   • Pertinence = 0 if the ODT is not pertinent, otherwise ODT is pertinent
   • max_time = . if the max_time value is not modified in the ODT

   Example:

   The file test1.mkodt_result generated by mkodt:

   

[Top]

Test Shell Script Creation

First of all, the mkodt command replays the ODTs without any Settings. But, if the license is managed by a license server, mkodt cannot silently request the nodelock licences. So, the licences to request must be specified.

To access to the license, the shell script must export the environment variable ODT_REQUESTED_PRODUCT. This variable will contain the list of needed configurations and products licenses. Each license must be separated by a semi-colon:

export ODT_REQUESTED_PRODUCT="_CFG1.slt+;PRD1.prd;...."

Examples

• MD2 configuration alone (one license):

  export ODT_REQUESTED_PRODUCT="_MD2.slt+"

• MD2 configuration and GPS shareable product (two licenses):

  export ODT_REQUESTED_PRODUCT="_MD2.slt+;GPS.prd"

• MD2 configuration and GPS add-on product (one license):

  export ODT_REQUESTED_PRODUCT="_MD2.slt+GPS"

Note that ODT_REQUESTED_PRODUCT variable is only supported for mkodt command.

Available commands in a test shell script

SetOdtParam

A shell can include the SetOdtParam instruction to control some aspects of its execution.

• SetOdtParam max_time=xx

  max_time is the maximum time allowed for test to be completed. The default value of max_time is 5 minutes. max_time can not exceed 60 minutes. If a test exceeds this max_time, the corresponding process is killed and the ODT exit with a return code 99.

• SetOdtParam Replay_xxxx=NO

  xxxx can be AIX, SunOS, HP-UX, Unix, Windows_NT, aix_a, solaris_a, hpux_b, intel_a. This instruction indicates that the current test object must not be run on the given operating system - this property is useful when all test objects are run automatically and not by hand. More than one such instruction can be written in a shell.

• SetOdtParam USER=xxx

  This instruction is used to specify the user (xxx) when the ODT will be replayed.

• SetOdtParam TYPE=xxx

  This instruction assigns a type to the ODT. Subsequently, with the release, it is possible to set the environment variable ReplayExcludType to the type that has been assigned to the ODT and thus exclude the replay of this ODT for that release.

  Example 1: If the ODT is:

  export ODT_REQUESTED_PRODUCT="_AL2.slt+;_ME2.slt+" SetOdtParam TYPE=MonTYPE LOAD exit $

  If in the Release, ReplayExcludType is set to MonTYPE, this ODT will not be replayed.

  This instruction is used in Product Release to select ODTs to replay.

  Example 2: In a product Release P1: set ReplayExcludType=TYPE1!TYPE2 (on Windows) or export ReplayExcludType=TYPE1&TYPE2 on UNIX.

  This Release P1 is not going to replay ODTs that contain:

  SetOdtParam TYPE=TYPE1

  and/or:

  SetOdtParam TYPE=TYPE2

WhichDir

This command is used to find a directory or a file in the workspace concatenation (ADL_ODT_CONCATENATION). 

ODTInit

This command can be used in ODTs that contain an initialization step; indeed, some ODTs could use the same data, for example, the same Part. In a replay, this command will factorize the initialization step….

To use this command, you have to create a shell-script, named ODTInit_logicname.sh in the ADL_ODT_IN directory of the test framework; This shell will create the output data that will be used in the others ODTs. To use the output of the ODTInit_logicname.sh file, you have to start the ODT shell-script with:

ODTInit logicname

The purpose of this command is to gain time on the ODT replay.

What's Forbidden in a Test Shell Script

[Top]

References

History

Copyright © 2000, Dassault Systèmes. All rights reserved.