Work Arounds for "SystemError: intialization of md_mypyc did not return an extension module"
Hi, all. I'm having similar issues to this discussion post from a couple of years ago: importing of python module for simple-salesforce | SmartBear Community I know third-party libraries are a little iffy with TestComplete, but I was wanting to try to send out telemetry using Azure Application Insights, and I am able to successfully run my Test Script once before it throws the error from trying to import the module when I re-run the test just manually with TestComplete's Run Test Routine button. Was wondering if people have found any python script workaround if they've encountered similar errors. Thank you for your time and help!
TestComplete does not recognize my external libraries.
I have followed the documentation from following link: https://support.smartbear.com/testcomplete/docs/scripting/specifics/python.html I followed step by step; to install libraries to TestComplete python interpreter to the source folder: "...SmartBear\TestComplete 15\Bin\Extensions\Python\Python313\Lib". After closing TestComplete and opening it again, the interpreter is not being refreshed; not getting the newly added libraries I installed. It keeps throwing me an error as follows: ModuleNotFoundError: No module named 'imp'. The imported library I'm trying to use is a customized file with multiple classes. However, it does also happen for win23com.client and pillow libraries. import win32com.client from PIL import ImageGrab This seems to be an error with python 3.13, but I do not find a way to downgrade the python interpreter version from TestComplete. I do not find a way to fix this even after following documentation on how to install packages, either. Would appreciate your help on this, Thanks
Testcomplete not executing interactive tests when triggered with TestComplete plugin .
Testcomplete is not executing interactive tests when triggered with TestComplete plugin . Testcomplete plugin version used in Jenkins job is 2.9.2 . The agent us connected to Jenkins master using JNLP for interactive execution . please help on this issue . The Testcomplete plugin invocation script is as given below: testcompletetest actionOnErrors: 'NONE', credentialsId: 'UTAAdmin_User_Pass', executorType: 'TE', executorVersion: '15.0', generateMHT: true, sessionScreenResolution: '1680x1050', suite: "${env.Pjs}", useTCService: true, timeout: '21600', useTimeout: true , useActiveSession: true Please find the log below 13:58:56 [TestComplete] The test execution started (TestAuto). 13:58:57 [TestComplete] Found TestComplete/TestExecute installations: 13:58:57 Type: TC, Version: 15.40.421.7, Path: "C:\Program Files (x86)\SmartBear\TestComplete 15\x64\bin\TestComplete.exe" 13:58:57 Type: TE, Version: 15.40.421.11, Path: "C:\Program Files (x86)\SmartBear\TestExecute 15\x64\bin\TestExecute.exe" 13:58:57 Type: TELite, Version: 15.40.421.50, Path: "C:\Program Files (x86)\SmartBear\TestExecuteLite 15\x64\bin\TestExecuteLite.exe" 13:58:57 [TestComplete] Selected TestComplete/TestExecute installation: 13:58:57 Type: TE, Version: 15.40.421.11, Path: "C:\Program Files (x86)\SmartBear\TestExecute 15\x64\bin\TestExecute.exe" 13:58:57 [TestComplete] [WARNING] The node is connected via Java Web Start (JNLP). In this mode, the "Run interactive user session" property of the TestComplete Test step is ignored. TestComplete (or TestExecute) will work in the current user session. 13:58:57 [TestComplete] Launching the test runner. 13:58:57 [TestComplete] [WARNING] Unable to publish test results (xml data is empty). [Pipeline] }
Checking API Status in TestComplete
Introduction I first saw the need to verify the state of an API several years ago with an application that used an address validation API in several of it's operations. If this API was down or did not respond in a timely manor, many of the automated test cases would run and fail. In this case, and in others, I have found that doing a simple call to the API to check the returned status code allowed me to skip, fail or log a warning with a logical message instead of allowing the application to fail with another less direct error message due to the API issue. The aqHttp Object The TestComplete aqHttp object and it's methods are very useful for performing simple checks like this and are also useful for other more complex tasks like leveraging an API to return a test data set or even verifying a certain data is returned prior to running tests against the UI that depend on the data. Sending and Receiving HTTP Requests From Script Tests More Complete API Testing Most proper API testing should be done using a tools like ReadyAPI or SoapUI. Both of these tools will integrate with TestComplete or can be used alone and will provide much more capabilities and automation options. Integration With ReadyAPI and SoapUI Code Example Here I have provided a working example of how to code a Get request using 'aqHttp.CreateRequest' to confirm an API returns a status code of 200 and it will log the returned records. function sendGetRequest() { let resourcePath ="https://restcountries.com/v3.1/all" let resourceQuery = "?fields=name,capital"; let url = resourcePath + resourceQuery; try { // Send GET request let response = aqHttp.CreateRequest("GET", url, false).Send(); // Check for successful response if (response.StatusCode === 200) { // Parse JSON response let allData = JSON.parse(response.Text); Log.Message("Total records received: " + allData.length); // Process each record allData.forEach((record, index) => { Log.Message("Record " + (index + 1) + ": " + JSON.stringify(record)); }); return true; // Send a bool back to the calling function. } else { throw new Error("Failed to fetch data. Status code: " + response.StatusCode); } } catch (error) { Log.Error("Error during request or data processing: " + error.message); } } Enhancements You could accept parameters for the resourcePath and resourceQuery. Parameterize the logging loop run or remove it. Return the JSON to the calling function for use. Perform other tasks based on the return code. Conclusion With the growing use of API calls in desktop applications and the fact that APIs are almost the foundation of any web site checking an API before a test case run is almost a requirement for consistent test runs and good error logging. This small script can bring big rewards to your test runs and reports. Cheers! I hope you find it as useful as I have! If you find my posts helpful drop me a like! 👍 Leave a comment if you want to contribute or have a better solution or an improvement. 😎 Have a great day
FindAll returns only one (first) element
Hi guys, I'm experiencing a weird problem. I have a mapped element, from which I need to get an array of elements. If tested in chrome/edge dev mode, the xpath of mapped element happily returns number of elements. But FindAll returns only one (first) element. Here is the code I'm using: function CompareElementsWithRefJSON(createNewRefFiles, setElementFlag, headerInputs, tabInputs, autocompleteMenu, selectPopUpMenu, tabsData, filePath) { let headerItems = headerInputs.FindAll('type','*',5000); headerItems = headerItems.toArray(); ... } I have a workaround, using FindElements, which works OK, but I don't like it, because it's unnecessary complicated: function CompareElementsWithRefJSON(createNewRefFiles, setElementFlag, headerInputs, tabInputs, autocompleteMenu, selectPopUpMenu, tabsData, filePath) { let formInputsXpath = GetXpathFromAlias(headerInputs); // This returns xpath of mapped alias object let headerParent = headerInputs.Parent; // Get parent element let headerItems = headerParent.FindElements(formInputsXpath); // Returns array of elements. Works OK. ... } So, any idea why FindAll returns just one element? BTW, it does not matter what PropName I use as the first parameter. The result is the same. Always just one element.
