Contributions
Re: Javascript parameter issue following fail in TestExecution
We are using TC 15.54, run on a virtual machine. Script B is not dependent on Script A. There are not defaults provided for the variables. The two most common instances of this issue we've encountered, to provide better detail, are the following. Script A accepts a nameMapping Alias as a parameter, evaluates the Alias properly and fails on a property check. Script B, which also accepts a nameMapping Alias, throws exception evaluating a null object, despite the fact that it is passed the alias in the same manner as Script A. Two independent calls, subsequently to the same Script A that accepts a string and a project var as parameters. The first call properly evaluates the string and project var and fails on a property check. The second call properly evaluates the string, but receives a null for the projectVar, despite the fact that the same project var is passed to each call and it has not been modified between the 2 invoking of the scripts.15Views0likes1CommentJavascript parameter issue following fail in TestExecution
Hello! We are observing a scenario when executing tests composed of multiple KeywordTests and Script calls assembled in Test Execution. Typically the flow of Keyword A -> Keyword B -> Script A -> Script B succeeds and all is well. However, we are seeing that when Script A fails (Log.Error, anticipated fail), Script B throws an exception for a null value. When reviewing the call stack, it indicates that after the failure in Script A, Script B runs with some parameters not defined. If the previous script succeeds, there is no issue. In the multiple cases where we observe this, the passed parameters that go missing following a failure are project variables, aliases. Strings, integers passed are unaffected. Any idea why this might be occurring? Are there steps to take on failures earlier in Test Execution to prevent this? The script calls we observe this in are Javascript functions stored in the same script file.25Views0likes4CommentsTestExecute reports failure when Log Indicates full success
Hello, I am running a long set of overnight Testcomplete tests via GITLab CI. In some cases, the test is fully executing, logging success end-to-end, and then terminating with: The TestExecute process has exited the the following code: -1 Failed to open the error log file. C:\Users\user\AppData\Local\Temp\{AA###!@$#}.log ERROR: Job failed: exit status 0xffffffff This causes gitlab to report failure instead of pass as the run log would indicate. In these cases, inspection of the exported log AND the log in the testcomplete session show green end to end. I've read that -1 indicates a license issue, but could that be accurate if my test is fully executing? command line call: .\SessionCreator.exe RunTest /UserName:$TCUSERNAME /password:$TCPASSWORD /UA /ScreenResolution:"3840*2160" /ProjectPath:"*projectPath*" /project:"Project" /test:"test" /Timeout:9000 /ExportLog:"*filepath*" /arg:"/e"174Views0likes3CommentsRe: Catch fail in keyword Execution
Is there anything that would allow me to control this more granularly than at the project level? Or can that only be applied to everything? I'd like to continue to stop on error in most scenarios.11 months agoPlace TestComplete QuestionsTestComplete Questions277Views0likes0CommentsCatch fail in keyword Execution
I am attempting to take a list tests that have been created, loop through it, catching any that fail to be logged in a report at the end. When I execute the loop: Try{ run KeywordTest} catch{ error} Most errors are not caught. For example, a tested object does not exist on the page, so it errors out and stops execution, rather than catching and continuing. Is there any method that would allow me to catch such errors and proceed?Solved294Views0likes3CommentsRelative file path for readyAPI projects in version-controlled testcomplete Project
Hello I have a version controlled Testcomplete project that is run by different users across different hosts. This project leverages the ReadyAPI plugin to execute some soap requests from the test execution page. The SOAP project elements seem to require an absolute file path either to the .xml or the projects directory. Is there any way to configure this so that it works for multiple people who are cloning the repository down to different locations? Is there any support for relative file paths? Or are we required to manually manipulate that file path in each person's local repository?294Views0likes1Comment