Unable to switch to new tab using chromedriver

The application that I’m testing using HSAC library opens a new tab when clicking on a certain button. The test script is supposed to take a screenshot of that new tab and save its content (it’s a PDF embedded in html, so its content is not easily testable). The test script uses fixture switch to next tab, but this does not work: The script still stays on the the same tab. I added some high level debugging with a scenario

|scenario|dump tabs        |
|show    |tab count        |
|show    |current tab index|

|scenario              |Show form|
|dump tabs                                       |
|click                 |Vertragsbedingungen      |
|wait                  |45    |seconds           |
|dump tabs                                       |
|seconds before timeout|20                       |
|switch to next tab                              |
|note                  |show  |save page source  |
|note                  |close tab                |
|dump tabs                                       |
|switch to previous tab                          |
|dump tabs                                       |
|seconds before timeout|5                        |

which confirms my observation: Before clicking the button the output is (1, 1), Tab count increases after clicking the button which leads to a (2, 1). But switch to next tabstill shows (2, 1):

In order to analyse this I copied the java-code of switchToNextTab() and added some logging:

Custom switchToNextTab

    public boolean switchToMyNextTab() {
        boolean result = false;
        List<String> tabs = getTabHandles();
        int currentTab = getCurrentTabIndex(tabs);

        logger.error("tabs: " + Arrays.toString(tabs.toArray()));
        logger.error("tabs(0): size " + tabs.size() + " current " + currentTab);

        if (tabs.size() > 1 || currentTab < 0) {
            int nextTab = currentTab + 1;
            if (nextTab == tabs.size() || nextTab < 0) {
                nextTab = 0;
            }
            logger.error("tabs: goto " + nextTab);
            goToTab(tabs, nextTab);
            logger.error("tabs(2): size " + tabs.size() + " current " + getCurrentTabIndex(tabs));
            result = true;
        }
        return result;
    }

To my surprise the value of current tab indexchanges (2nd pic in attachments), but I run into timeouts for switchToMyNextTab() and every fixture method that is called afterwards.

FitNesse output on calling custom switchToNextTab

switch to my next tab 
org.openqa.selenium.TimeoutException: timeout: Timed out receiving message from renderer: 20.000
  (Session info: chrome=85.0.4183.102)
Build info: version: '3.141.59', revision: 'e82be7d358', time: '2018-11-14T08:17:03'
System info: host: 'Holgers-MacBook-Pro.local', ip: 'fe80:0:0:0:10b0:2b5d:511d:13e3%en0', os.name: 'Mac OS X', os.arch: 'x86_64', os.version: '10.15.6', java.version: '13.0.2'
Driver info: org.openqa.selenium.chrome.ChromeDriver
Capabilities {acceptInsecureCerts: false, browserName: chrome, browserVersion: 85.0.4183.102, chrome: {chromedriverVersion: 85.0.4183.87 (cd6713ebf92fa..., userDataDir: /var/folders/2c/_cmd0tyx2s1...}, goog:chromeOptions: {debuggerAddress: localhost:51122}, javascriptEnabled: true, networkConnectionEnabled: false, pageLoadStrategy: normal, platform: MAC, platformName: MAC, proxy: Proxy(), setWindowRect: true, strictFileInteractability: false, timeouts: {implicit: 0, pageLoad: 300000, script: 30000}, unhandledPromptBehavior: ignore, webauthn:virtualAuthenticators: true}
Session ID: f25506a1d25913a63c5cc22c2df58787
	at java.base/jdk.internal.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
	at java.base/jdk.internal.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
	at java.base/jdk.internal.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
	at java.base/java.lang.reflect.Constructor.newInstanceWithCaller(Constructor.java:500)
	at java.base/java.lang.reflect.Constructor.newInstance(Constructor.java:481)
	at org.openqa.selenium.remote.http.W3CHttpResponseCodec.createException(W3CHttpResponseCodec.java:187)
	at org.openqa.selenium.remote.http.W3CHttpResponseCodec.decode(W3CHttpResponseCodec.java:122)
	at org.openqa.selenium.remote.http.W3CHttpResponseCodec.decode(W3CHttpResponseCodec.java:49)
	at org.openqa.selenium.remote.HttpCommandExecutor.execute(HttpCommandExecutor.java:158)
	at org.openqa.selenium.remote.service.DriverCommandExecutor.execute(DriverCommandExecutor.java:83)
	at org.openqa.selenium.remote.RemoteWebDriver.execute(RemoteWebDriver.java:552)
	at org.openqa.selenium.remote.RemoteWebDriver$RemoteTargetLocator.defaultContent(RemoteWebDriver.java:922)
	at nl.hsac.fitnesse.fixture.util.selenium.SeleniumHelper.switchToDefaultContent(SeleniumHelper.java:1308)
	at nl.hsac.fitnesse.fixture.util.selenium.SeleniumHelper.goToTab(SeleniumHelper.java:1296)
	at nl.hsac.fitnesse.fixture.slim.web.BrowserTest.goToTab(BrowserTest.java:445)
	at de.fiduciagad.kfinfi.fitnesse.Sofortkredit.switchToMyNextTab(Sofortkredit.java:47)
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
	at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.base/java.lang.reflect.Method.invoke(Method.java:567)
    ...

The execution log (the truncated part shows only a series of backtraces, none of the custom loggings):

Execution log

Please see https://chromedriver.chromium.org/security-considerations for suggestions on keeping ChromeDriver safe.

ChromeDriver was started successfully.

Sept. 15, 2020 4:52:00 PM org.openqa.selenium.remote.ProtocolHandshake createSession
INFORMATION: Detected dialect: W3C

Sept. 15, 2020 4:52:26 PM de.fiduciagad.kfinfi.fitnesse.Sofortkredit showTabInfo
SCHWERWIEGEND: tabs: size 1 current 0

Sept. 15, 2020 4:55:26 PM de.fiduciagad.kfinfi.fitnesse.Sofortkredit switchToMyNextTab
SCHWERWIEGEND: tabs: [CDwindow-7269F2F1E351583739FF39A979105790, CDwindow-F610A691675FF3818DE082E59009C576]

Sept. 15, 2020 4:55:26 PM de.fiduciagad.kfinfi.fitnesse.Sofortkredit switchToMyNextTab
SCHWERWIEGEND: tabs(0): size 2 current 0

Sept. 15, 2020 4:55:26 PM de.fiduciagad.kfinfi.fitnesse.Sofortkredit switchToMyNextTab
SCHWERWIEGEND: tabs: goto 1

[1600181746.332][SEVERE]: Timed out receiving message from renderer: 20.000

[1600181746.333][SEVERE]: Timed out receiving message from renderer: 20.000

Unable to take screenshot for exception: null

org.openqa.selenium.TimeoutException: timeout: Timed out receiving message from renderer: 20.000
  (Session info: chrome=85.0.4183.102)
Build info: version: '3.141.59', revision: 'e82be7d358', time: '2018-11-14T08:17:03'
System info: host: 'Holgers-MacBook-Pro.local', ip: 'fe80:0:0:0:10b0:2b5d:511d:13e3%en0', os.name: 'Mac OS X', os.arch: 'x86_64', os.version: '10.15.6', java.version: '13.0.2'
Driver info: org.openqa.selenium.chrome.ChromeDriver
Capabilities {acceptInsecureCerts: false, browserName: chrome, browserVersion: 85.0.4183.102, chrome: {chromedriverVersion: 85.0.4183.87 (cd6713ebf92fa..., userDataDir: /var/folders/2c/_cmd0tyx2s1...}, goog:chromeOptions: {debuggerAddress: localhost:51122}, javascriptEnabled: true, networkConnectionEnabled: false, pageLoadStrategy: normal, platform: MAC, platformName: MAC, proxy: Proxy(), setWindowRect: true, strictFileInteractability: false, timeouts: {implicit: 0, pageLoad: 300000, script: 30000}, unhandledPromptBehavior: ignore, webauthn:virtualAuthenticators: true}
Session ID: f25506a1d25913a63c5cc22c2df58787

	at java.base/jdk.internal.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
    ...

So my questions are:

• Why is there a difference in behaviour of custom and library version of switchToNextTab()?
• Am I doing something wrong, or can the behaviour be fixed somehow?

Versions: HSAC 4.19.0 FitNesse 20200501

Many thanks in advance, reators

[fhoeben]

Downloading content from a ‘data-url’ instead of ‘normal’ link is actually quite a nice feature. I created an issue to add support for that. (Maybe someone can implement a PR for it before I’ll get around to it.)

[reators]

Unfortunately did the JsonHttpTest-version work only in one of our test environments, but in not the others. As the headers are quite different, and copy browser cookies didn’t help, and the internal processing on Angular side is different (using an “http interceptor”, whatever this means), I decided to go the other way you proposed:

Decoding the content you extracted is certainly possible, but I hadn’t seen the need before to expose that as a fixture method. JsonHttpTest can do it when such a value is present in a json response (see

https://github.com/fhoeben/hsac-fitnesse-fixtures/blob/e09c9702209eda9be089be12b32776eea3948cb2/src/main/java/nl/hsac/fitnesse/fixture/slim/JsonHttpTest.java#L82 ). The class DataUrlHelper strips the prefix and then Base64Fixture can save it to disk. At the moment I fear you have to write your own custom fixture to get this to work directly from a string.

It was roughly a copy’n’pasting of existing code. I’ll provide my solution here. It’s probably not as elegant as proposed in #347, but it works for me:

    protected Base64Fixture getBase64Fixture() {
        return new Base64Fixture();
    }

    public String createFileFromUrlEncoded(final String baseName, final String urlEncodedContent) {

        String base64Content = java.net.URLDecoder.decode(urlEncodedContent, StandardCharsets.UTF_8);
        if (DataUrlHelper.isDataUrl(base64Content)) {
            base64Content = DataUrlHelper.getData(base64Content);
        }
        Base64Fixture base64Fixture = getBase64Fixture();
        return base64Fixture.createFrom(baseName, base64Content);
    }

Calling this from a scenario:

|scenario    |download pdf as   |filename                              |
|$pdfencoded=|value of attribute|src      |on              |css=iframe |
|$pdf=       |create file       |@filename|from Url encoded|$pdfencoded|