Gmail OAuth 2.0 API Automation Example and example SubmitListener.beforeSubmit
Hi. Recently I had a use case where I had to verify that our application sends an email in a particular case. In my case, it was an email whenever a certain status is reached, but it could also be for instance the sending of an invitation (user-creation) email or whatnot. In essence: I want to check at a give moment when the status condition is reached that the email is delivered to the proper email address. In the past, there we some free email generators (like mailinator, 10minutemail,...) where you could get API calls to check the inbox, but I couldn't find any that offered that feature for free. So I decided to setup a dedicated gmail email address for my tests and talk to the gmail api to read the inbox (all messages), verify the subject and the actual content of a mail and delete the messages. More info on the gmail api here: https://developers.google.com/gmail/api/reference/rest. To get started, I created a gmail user and setup an OAuth2.0 google client Id (https://console.cloud.google.com/apis/credentials). These settings (clientId and secret) I used to setup an Authorization code grant as described here:https://support.smartbear.com/readyapi/docs/requests/auth/types/oauth2/grants/auth-code.html So far so good, the readyAPI internal browers showed me the google popup and I could manually insert the authentication that was needed to generate successfully an access token from google. BUT: When I tried to automate the flow in this popup using the example code provided in the documentation (=https://support.smartbear.com/readyapi/docs/requests/auth/types/oauth2/automate/sample.html) it did not work for me. Therefore I wanted to share the changes I made to it in order to get it working. Also the event handler SubmitListener.beforeSubmit(described on the same page) needed some rework for me to work properly. As an extra, I also have a test case setup script that deletes all emails in the gmail inbox so I have a proper starting situation for my tests. Hope this can help any other testers that would need this! Automation Scripts tab of the Auth Manager I have encrypted project properties that store my gmail username (gmailUser) and password (gmailPass). Page 1: // This function asks for permission to use OAuth. The user must be logged in to use it. Logging in is performed in the script below. function consent() { if (document.getElementById('submit_approve_access')){ document.getElementById('submit_approve_access').click(); } } // This function fills user password in when the user name is already known. It uses the project-level "pass" property. function fillpwd() { document.getElementsByName('password')[0].value = '${#Project#gmailPass}'; document.getElementById('passwordNext').click(); window.setInterval(consent, 1000); } // This script checks what page is displayed and provides the appropriate data. It uses the project-level "user" and "pass" properties. if (document.getElementById('profileIdentifier')) { document.getElementById('profileIdentifier').click(); window.setTimeout(fillpwd, 1000) }else if (document.getElementById('identifierId') && document.getElementById('identifierNext')) { document.getElementById('identifierId').value = '${#Project#gmailUser}'; document.getElementById('identifierNext').click(); window.setTimeout(fillpwd, 1000); } else if (document.getElementByType('password')) { fillpwd(); } else if(document.getElementById('submit_approve_access')){ window.setInterval(consent, 100); } Page 2: function consent() { if (document.getElementById('submit_approve_access')){ document.getElementById('submit_approve_access').click(); } } window.setInterval(consent, 100); Event handler SubmitListener.beforeSubmit: Note: There might be some redundant iteration of code in there, feel free to rewrite, main thing is: it works. I also expected that I could use the "Target" column to filter on the requests steps that start with "gmail*" but that didn't do it. So I fixed that in another way, together with providing some smart checking whether a new token generation is needed or not (gmail token is valid for 60 minutes). // Import the required classes import com.eviware.soapui.impl.rest.actions.oauth.OltuOAuth2ClientFacade; import com.eviware.soapui.support.editor.inspectors.auth.TokenType; import com.eviware.soapui.model.support.ModelSupport; import java.time.LocalDateTime import java.time.format.DateTimeFormatter def testStepName = context.getModelItem().getName() def expiresOn = context.expand('${#Project#expiresOn}') if (testStepName.toLowerCase().contains("gmail")) { // IF expiresOn == "" OR dateNow is > expiresOn then we need to get a new token. Otherwise the old should still do.... TimeZone.setDefault(TimeZone.getTimeZone('UTC')) TimeZone tz = TimeZone.getTimeZone("UTC") LocalDateTime dateNow = LocalDateTime.now() def patternUTC = "yyyy-MM-dd\'T\'HH:mm:ss.SSS\'Z\'" DateTimeFormatter dateFormatUTC = DateTimeFormatter.ofPattern(patternUTC).withLocale(Locale.US) String nowUtcFormat = dateNow.format(dateFormatUTC) LocalDateTime dateExpiresOn = dateNow.plusMinutes(55) String expiresOnUtcFormat = dateExpiresOn.format(dateFormatUTC) if (expiresOn == "") { /* log.info("Expires on is empty, so we need to run submitListener and get new token. We also write the new expiresOn to the project properties!") log.info("nowUtcFormat = " + nowUtcFormat) log.info("dateExpiresOn = " + expiresOnUtcFormat) log.info "let's run the submitListener.beforeSubmit to get a new accessToken. We know this one will be valid for 60 minutes, so we set a the expiresOn project property to now + 55 minutes" */ // Set up variables def project = ModelSupport.getModelItemProject(context.getModelItem()) project.setPropertyValue("expiresOn", expiresOnUtcFormat) def authProfile = project.getAuthRepository().getEntry("google") def oldToken = authProfile.getAccessToken() def tokenType = TokenType.ACCESS // Create a facade object def oAuthFacade = new OltuOAuth2ClientFacade(tokenType) // Request an access token in headless mode oAuthFacade.requestAccessToken(authProfile, true, true) // Wait until the access token gets updated int iteration = 0 while (oldToken == authProfile.getAccessToken() && iteration < 10) { sleep(500) iteration++ } // Post the info to the log log.info("Gmail authentication event handler: Project property \"expiresOn\" is empty! We get/set new token: " + authProfile.getAccessToken() + " with new expiresOn = " + expiresOnUtcFormat + ". The old token (with unknown expiresOn) was = " + oldToken) } else { def potentialNewExpiresOn = expiresOnUtcFormat //expiresOn retrieved from project properties dateExpiresOn = LocalDateTime.parse(expiresOn, dateFormatUTC) expiresOnUtcFormat = dateExpiresOn.format(dateFormatUTC) //log.info("dateExpiresOn = " + expiresOnUtcFormat) if (dateNow > dateExpiresOn) { // log.info "Expired! Let's get a new token..." // Set up variables def project = ModelSupport.getModelItemProject(context.getModelItem()) def authProfile = project.getAuthRepository().getEntry("google") def oldToken = authProfile.getAccessToken() def tokenType = TokenType.ACCESS // Create a facade object def oAuthFacade = new OltuOAuth2ClientFacade(tokenType) // Request an access token in headless mode oAuthFacade.requestAccessToken(authProfile, true, true) // Wait until the access token gets updated int iteration = 0 while (oldToken == authProfile.getAccessToken() && iteration < 10) { sleep(500) iteration++ } // Post the info to the log project.setPropertyValue("expiresOn", potentialNewExpiresOn) log.info("Token was expired! We get/set new token: " + authProfile.getAccessToken() + " with new expiresOn = " + potentialNewExpiresOn + ". The old token (expiresOn = $expiresOn vs now " + nowUtcFormat + ") was = " + oldToken) } else { //log.info "Not yet expired. Let's keep using the same old token (expiresOn = $expiresOn vs now " + nowUtcFormat + ")" } } } Setup script to delete all emails in the gmail inbox for proper start situation: I have a disabled test suite "WorkItem" with a test case named "GmailStartSituationCleanup" This test case has 4 steps: 1°GET gmail messagesList 2° Script "IterateOverAllGmailMessageIds" 3°DELETE gmail messageId 4°GET gmail messagesList-EmptyListCheck def testSuiteWorkItem = testRunner.testCase.testSuite.project.getTestSuiteByName("WorkItem") def testCaseGmailStartSituationCleanup = testSuiteWorkItem.getTestCaseByName("GmailStartSituationCleanup") testCaseGmailStartSituationCleanup.run(new com.eviware.soapui.support.types.StringToObjectMap(), false) The groovy test step 2°IterateOverAllGmailMessageIds = import com.eviware.soapui.support.JsonUtil def testStepAllMessages = testRunner.testCase.getTestStepAt(context.getCurrentStepIndex()-1) def teststepNameAllMessages = testStepAllMessages.getName() def testStepDeleteMessage = testRunner.testCase.getTestStepAt(context.getCurrentStepIndex()+1) testStepDeleteMessage.setDisabled(true) def testStepVerifAllDeleted = testRunner.testCase.getTestStepAt(context.getCurrentStepIndex()+2) testStepVerifAllDeleted.setDisabled(false) def responseStatus = testStepAllMessages.testRequest.response.responseHeaders["#status#"][0] if (responseStatus.contains("HTTP/1.1 2")){ def responseMessages = context.expand( '${'+teststepNameAllMessages+'#Response#$[\'messages\']}' ) if (responseMessages!= "" && responseMessages!= null){ def numberOfMessages = (JsonUtil.parseTrimmedText(responseMessages)).size() def id for (i=0;i<numberOfMessages;i++){ id = context.expand( '${'+teststepNameAllMessages+'#Response#$[\'messages\']['+i+'][\'id\']}' ) testStepDeleteMessage.setPropertyValue("messageId", id) log.info "Cleanup of gmail messages : Message "+(i+1).toString()+"/"+numberOfMessages.toString()+" with messageId $id will be deleted so we can continue with a proper starting situation..." testStepDeleteMessage.run(testRunner, context) } }else{ log.info "Cleanup of gmail messages : No messages found. We can continue with proper starting situation" testStepVerifAllDeleted.setDisabled(true) } }1.2KViews5likes2CommentsGroovy Automated DataSource Loopers
I understand that this is a SoapUI Pro forum but I think creating the loop steps through groovy manually can be a good lesson and can help you customise your loops to your heart (and/or requirement)'s content! So without further ado this is how I script my groovy loopers NOTE: in this example I am connecting to an OracleDB to fuel my requests. 1) Structure: The structure within my test case is quite simple:loopStarterto connect to the db/initialise the data - storing the current information into properties,responseto send these properties off to the API through a request,loopEnderto either propagate or exit the loop if the condition is met. 2) loopStarter Connection to/creation of your data source, be it a database, excel etc In this example I will be using the groovy.sqlclass (documentation) to connect to my db, and store the results into properties import groovy.sql.Sql sql = Sql.newInstance(<connection details>) def res = sql.rows(<sql query>) def loopProperties = testRunner.testCase.getTestStepByName("loopProperties") //will initialise count in the loopProperties step if count does not yet exist if(!loopProperties.hasProperty("count")){ loopProperties.setPropertyValue("count","0") } def count = Integer.parseInt(loopProperties.getPropertyValue("count")) //store the properties from the current result loopProperties.setPropertyValue("x",res[count].x) loopProperties.setPropertyValue("y",res[count].y) loopProperties.setPropertyValue("querySize",(String)res.size()) sql.close() sql.rows returns an array of ArrayLists, so to access the current result (more on this later) we will use res[count]. Just to reiterate this step ONLY sets the properties to be sent off in the request.querySize is set so we can continue looping over all the results from the query. 3) response To then call your properties into your request step we will do this: <soapenv:Envelope namespace:ns="namespace"> <soapenv:Header/> <soapenv:Body> <ns:x>${loopProperties#x}</ns:x> <ns:y>${loopProperties#y}</ns:y> </soapenv:Body> </soapenv:Envelope> if you choose not to store your properties in a property step and wanted to store them at the test case level you can fetch them like so: ... <ns:x>${#TestCase#x}</ns:x> ... 4) loopEnder This is the loop exit condition. It will tell the loop whether or not there are more results from your query to send through the request. This is also a pretty simple step: - check count vs querySize - increment count if true - go to loopStarter if true, to update the properties (count has been increased andbecause of this, so will the current result) def loopProperties = testRunner.testCase.getTestStepByName("loopProperties") def count = Integer.parseInt(loopProperties.getPropertyValue("count")) def querySize = Integer.parseInt(loopProperties.getPropertyValue("querySize")) if( count<querySize-1 ){ count = count+1 loopProperties.setPropertyValue("count", (String)count) testRunner.gotoStepByName("loopStarter") } 5) reset count Unnecessary step but I like getting into the habbit of clearing all properties after I'm done, and this groovy script will do it for you def loopProperties = testRunner.testCase.getTestStepByName("loopProperties") String[] removals = loopProperties.getPropertyNames() for(i=0;i<removals.size();i++){ loopProperties.removeProperty(removals[i]) } I've been thinking of doing something similar to this for the Excel DataSource scriptOlga_Tmentioned, well similar to the loopStarter step anyway - doing it completely through groovy so SoapUI free users can do it too, I just need to look into parsing exel files with groovy. Anyways I hope this helps you all, Mo3KViews3likes4CommentsTip : Use of tags in command line via bamboo build - Run all tests while sending -T tags parameters
In our CI build pipeline (bamboo) I setup two parameters tagTestSuite and tagTestCase. This way, all our automation testers can specify, if needed, which test cases need to be ran based on certain tags. This works fine if there are indeed such tags passed on. But what if a project just wants to run all of its tests? These -T parameters cannot be left empty in the command line as that fails the build... Deleting them from the bamboo variables will then lose the standardized approach for all testers to do a tag-based execution. Seeing the documentation here (https://support.smartbear.com/readyapi/docs/functional/running/automating/cli.html)I noticed you can provide logical operators. So I thought I came up with a nice solution: I provide as default value for those parameters in bamboo a tag value that nobody would use and by setting the "!" operator in front of it I'd expected thatALL test suites/cases would run. So I've set parameters and resulted in a command line that has these : "-TTestSuite !itShouldRunAllTestSuites" "-TTestCase !itShouldRunAllTestCase". Since none of the testSuite/testCases has a tag "itShouldRunAllTestSuites" or "itShouldRunAllTestCases" I thought my project would be run in full.... But nope: error is shown:[errorlog] java.lang.Exception: The tag "itShouldRunAllTestSuites" was not found. Yes, I could go and add a tag to every single test suite and test case to work around this, but that could be tedious and not so future proof... Solution: ReadyAPI seems to look first for a tag and then complain it wasn't found. So I tried this: I've added the tags "itShouldRunAllTestCases" and "itShouldRunAllTestSuites" to my project but did not tag any of my testsuites/testcases with this tag. Executed above again and now ALL of my tests get's executed nicely!836Views2likes2Comments21 Selenium startup error: on alertsite with chromebrowser
I am trying to build a selenium jar file on ubuntu wsl in an environment with chrome browser 85 using and I tried building it using java -jar command as mentioned in the document and it is working fine in my local environment. When I try to upload the JAR file built to alertsite via this code, it gives me error code 21 I did verify the main class is present in the jar file, and all the selenium libraries are installed as well. I presume its an issue due to the browser version mismatch on alertsite. Can someone let me know what the issue here is? which is the supported version on alertsite. Im not able to install older versions of chrome on my wsl as the official releases from google are removed. what can be done here? Thanks in advance PS: selenium version 3.141.59 and java version 11 i specify in my pom.xml during build. 21 Selenium startup error: on alertsite with chromebrowser5Views1like0CommentsAutomation script in ReadyAPI for Authentication
Hi, I am new to Ready API. I am trying to authenticate the user credentials and set it up as the script inside Auth Manager so that we can run the ready API test suites from the pipeline automatically. Manual set up for authentication is working fine and we are able to call the APIs and get results. We use Oauth2.0 authentication. The automation script that I have is not working. I am not sure what I am missing. This is what I tried referring to this document: https://support.smartbear.com/readyapi/docs/requests/auth/types/oauth2/automate/index.html var loginForm = document.querySelectorAll(".ab.cd"); var username = document.getElementById("username").value; var password = document.getElementById("password").value; var loginBtnForm = document.querySelector(".c2"); var loginBtn = document.querySelectorAll(".c3.c4.c5.c6.c7"); document.getElementById("username").value = "testusername"; document.getElementById("password").value = "testpassword"; loginBtn.click();137Views1like6CommentsTestEngine external JAR files - error - unable to resolve class
Background - We have test engine hosted on some linux server and we execute tests suing CURL cmd or POSTMAN. Given some external JARs When we add external JARs in /bin/ext folder of test engine server and re-start the test engine And run the tests using curl command from locally or using POSTMAN Then it throws error for groovy script test step saying unable to resolve class <className present in external JARs> on import statement itself. We have also configured .vmoptions of test engine server to read ext JARs but still it throws error. Locally, In ReadyAPI, we are placing those JARs in /bin/ext/ folder and it works perfectly fine. Any solution to this problem?199Views1like2CommentsIs it possible to integrate the Azure TestPlan Test cases and the ReadyAPI tests ?
I have created manual test cases in Azure devops Test Plan. I have automated the manual test case functionality using ReadyAPI. Now i want to link the Azure Test Plan test cases while running the ReadyAPI tests via the Azure pieplines. Is it possible to do so ? I am using the Azure devops extension named "ReadyAPI Test for Azure Devops" and "Publish Test Results" in my pipeline.206Views1like1CommentUsing smart card / CAC certificate with ReadyAPI
As of this morning, I am trying to figure out if its possible to use a CAC/Smart card certificate device with ReadyAPI. The intent would be to sign via WSS-Configuration any outgoing SOAP requests. Current work flow is we utilize a .p12 file for a test user, and set that up in WSS-Config with all of the necessary parts and set it to the endpoints we are testing. I am not sure how to achieve that same level of testing with a Smart Card. Any/all insight would be much appreciated!Solved368Views1like3CommentsWhy I dont see GraphQL tab in Endpoints& Environment Section in ReadyAPI?
Why I dont see GraphQL tab in Endpoints& Environment Section in ReadyAPI? We are using 3 different ReadyAPI licenses that supports virtualization, Functional Test, Performance Testing. When looking for GraphQL tab under Endpoints and Environment section to support Functional/Performance testing of GraphQL, we do not see GraphQL tab? do we need to have any additional license for it? can we do GraphQL endpoint validation without this tab? are there any video/webinar links that demo GraphQL endpoint functional/performance testing. Thank you!316Views1like1Comment3.44 issue with java.net.SocketTimeoutException: Read timed out
test step is failing due to "java.net.SocketTimeoutException: Read timed out", we tried increasing the socket timeout but it is still failing after the default 1 minute time. this project and test step is working fine when using 3.41 readyapi version but same test is failing in 3.44 and the latest 3.45 versions. please let us know what need to be done here as this impacting our regression testsSolved1.3KViews1like2Comments