Using test scripts under source control
Hi, I have the following thoughts/questions about using test scripts under SCM. Any comments?
When setting up automatic testing in RQM, using fx. RFT, you can automatically kick off RFT on a successful RTC build.
But do you know, if there is a suggested way to fetch the correct test scripts from source control?
There might be different tests for different builds (release levels) of the same code, so a common shared location might not be a good solution. I've thought about making the build deploying the tests scripts to the shared location or to the test machine.
Maybe it's also possible to fetch test scripts from a wrapper RFT test script?
I appreciate any input :-)
When setting up automatic testing in RQM, using fx. RFT, you can automatically kick off RFT on a successful RTC build.
But do you know, if there is a suggested way to fetch the correct test scripts from source control?
There might be different tests for different builds (release levels) of the same code, so a common shared location might not be a good solution. I've thought about making the build deploying the tests scripts to the shared location or to the test machine.
Maybe it's also possible to fetch test scripts from a wrapper RFT test script?
I appreciate any input :-)
Accepted answer
Hi,
If you are using RQM's Execution Schedules to kick off the RFT automation once the RTC build completes you could add a command-line adapter script to the execution schedule, that runs before the RFT script. It's purpose would be to use RTC's SCM command-line interface to download any changes to their source controlled scripts, before it starts. Or you could do the same using a wrapper RFT script.
A challenge will be, figuring out how to make the initial script smart enough to determine the build it's associated with, in order to download the build specific scripts files from SCM. You could potentially create different RQM scripts that point to the same command-line script. Each RQM script would have an execution variable that has the build definition name. When the command-line script runs, it could read this execution variable to determine which build it's associated with and extract the correct files from source control.
Helpful Links:
http://publib.boulder.ibm.com/infocenter/clmhelp/v3r0m1/topic/com.ibm.team.scm.doc/topics/m_cli_usecases.html
http://publib.boulder.ibm.com/infocenter/clmhelp/v3r0m1/topic/com.ibm.rational.test.qm.doc/topics/c_execvar_overview.html
Thanks,
Deepa
If you are using RQM's Execution Schedules to kick off the RFT automation once the RTC build completes you could add a command-line adapter script to the execution schedule, that runs before the RFT script. It's purpose would be to use RTC's SCM command-line interface to download any changes to their source controlled scripts, before it starts. Or you could do the same using a wrapper RFT script.
A challenge will be, figuring out how to make the initial script smart enough to determine the build it's associated with, in order to download the build specific scripts files from SCM. You could potentially create different RQM scripts that point to the same command-line script. Each RQM script would have an execution variable that has the build definition name. When the command-line script runs, it could read this execution variable to determine which build it's associated with and extract the correct files from source control.
Helpful Links:
http://publib.boulder.ibm.com/infocenter/clmhelp/v3r0m1/topic/com.ibm.team.scm.doc/topics/m_cli_usecases.html
http://publib.boulder.ibm.com/infocenter/clmhelp/v3r0m1/topic/com.ibm.rational.test.qm.doc/topics/c_execvar_overview.html
Thanks,
Deepa
Hi, I have the following thoughts/questions about using test scripts under SCM. Any comments?
When setting up automatic testing in RQM, using fx. RFT, you can automatically kick off RFT on a successful RTC build.
But do you know, if there is a suggested way to fetch the correct test scripts from source control?
There might be different tests for different builds (release levels) of the same code, so a common shared location might not be a good solution. I've thought about making the build deploying the tests scripts to the shared location or to the test machine.
Maybe it's also possible to fetch test scripts from a wrapper RFT test script?
I appreciate any input :-)