interface ICoreWebView2

interface ICoreWebView2
  : public IUnknown

WebView2 enables you to host web content using the latest Edge web browser technology.

Summary

Members Descriptions
add_ContainsFullScreenElementChanged Add an event handler for the ContainsFullScreenElementChanged event.
add_ContentLoading Add an event handler for the ContentLoading event.
add_DocumentTitleChanged Add an event handler for the DocumentTitleChanged event.
add_FrameNavigationCompleted Add an event handler for the FrameNavigationCompleted event.
add_FrameNavigationStarting Add an event handler for the FrameNavigationStarting event.
add_HistoryChanged Add an event handler for the HistoryChanged event.
add_NavigationCompleted Add an event handler for the NavigationCompleted event.
add_NavigationStarting Add an event handler for the NavigationStarting event.
add_NewWindowRequested Add an event handler for the NewWindowRequested event.
add_PermissionRequested Add an event handler for the PermissionRequested event.
add_ProcessFailed Add an event handler for the ProcessFailed event.
add_ScriptDialogOpening Add an event handler for the ScriptDialogOpening event.
add_SourceChanged Add an event handler for the SourceChanged event.
add_WebMessageReceived Add an event handler for the WebMessageReceived event.
add_WebResourceRequested Add an event handler for the WebResourceRequested event.
add_WindowCloseRequested Add an event handler for the WindowCloseRequested event.
AddHostObjectToScript Add the provided host object to script running in the WebView with the specified name.
AddScriptToExecuteOnDocumentCreated Add the provided JavaScript to a list of scripts that should be executed after the global object has been created, but before the HTML document has been parsed and before any other script included by the HTML document is executed.
AddWebResourceRequestedFilter Adds a URI and resource context filter to the WebResourceRequested event.
CallDevToolsProtocolMethod Call an asynchronous DevToolsProtocol method.
CapturePreview Capture an image of what WebView is displaying.
ExecuteScript Execute JavaScript code from the javascript parameter in the current top level document rendered in the WebView.
get_BrowserProcessId The process id of the browser process that hosts the WebView.
get_CanGoBack Returns true if the WebView can navigate to a previous page in the navigation history.
get_CanGoForward Returns true if the WebView can navigate to a next page in the navigation history.
get_ContainsFullScreenElement Indicates if the WebView contains a fullscreen HTML element.
get_DocumentTitle The title for the current top level document.
get_Settings The ICoreWebView2Settings object contains various modifiable settings for the running WebView.
get_Source The URI of the current top level document.
GetDevToolsProtocolEventReceiver Get a DevTools Protocol event receiver that allows you to subscribe to a DevTools Protocol event.
GoBack Navigates the WebView to the previous page in the navigation history.
GoForward Navigates the WebView to the next page in the navigation history.
Navigate Cause a navigation of the top level document to the specified URI.
NavigateToString Initiates a navigation to htmlContent as source HTML of a new document.
OpenDevToolsWindow Opens the DevTools window for the current document in the WebView.
PostWebMessageAsJson Post the specified webMessage to the top level document in this WebView.
PostWebMessageAsString This is a helper for posting a message that is a simple string rather than a JSON string representation of a JavaScript object.
Reload Reload the current page.
remove_ContainsFullScreenElementChanged Remove an event handler previously added with add_ContainsFullScreenElementChanged.
remove_ContentLoading Remove an event handler previously added with add_ContentLoading.
remove_DocumentTitleChanged Remove an event handler previously added with add_DocumentTitleChanged.
remove_FrameNavigationCompleted Remove an event handler previously added with add_FrameNavigationCompleted.
remove_FrameNavigationStarting Remove an event handler previously added with add_FrameNavigationStarting.
remove_HistoryChanged Remove an event handler previously added with add_HistoryChanged.
remove_NavigationCompleted Remove an event handler previously added with add_NavigationCompleted.
remove_NavigationStarting Remove an event handler previously added with add_NavigationStarting.
remove_NewWindowRequested Remove an event handler previously added with add_NewWindowRequested.
remove_PermissionRequested Remove an event handler previously added with add_PermissionRequested.
remove_ProcessFailed Remove an event handler previously added with add_ProcessFailed.
remove_ScriptDialogOpening Remove an event handler previously added with add_ScriptDialogOpening.
remove_SourceChanged Remove an event handler previously added with add_SourceChanged.
remove_WebMessageReceived Remove an event handler previously added with add_WebMessageReceived.
remove_WebResourceRequested Remove an event handler previously added with add_WebResourceRequested.
remove_WindowCloseRequested Remove an event handler previously added with add_WindowCloseRequested.
RemoveHostObjectFromScript Remove the host object specified by the name so that it is no longer accessible from JavaScript code in the WebView.
RemoveScriptToExecuteOnDocumentCreated Remove the corresponding JavaScript added using AddScriptToExecuteOnDocumentCreated with the specified script id.
RemoveWebResourceRequestedFilter Removes a matching WebResource filter that was previously added for the WebResourceRequested event.
Stop Stop all navigations and pending resource fetches.
COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT Image format used by the ICoreWebView2::CapturePreview method.
COREWEBVIEW2_KEY_EVENT_KIND The type of key event that triggered an AcceleratorKeyPressed event.
COREWEBVIEW2_MOVE_FOCUS_REASON Reason for moving focus.
COREWEBVIEW2_PERMISSION_KIND The type of a permission request.
COREWEBVIEW2_PERMISSION_STATE Response to a permission request.
COREWEBVIEW2_PHYSICAL_KEY_STATUS A structure representing the information packed into the LPARAM given to a Win32 key event.
COREWEBVIEW2_PROCESS_FAILED_KIND Kind of process failure used in the ICoreWebView2ProcessFailedEventHandler interface.
COREWEBVIEW2_SCRIPT_DIALOG_KIND Kind of JavaScript dialog used in the ICoreWebView2ScriptDialogOpeningEventHandler interface.
COREWEBVIEW2_WEB_ERROR_STATUS Error status values for web navigations.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT Enum for web resource request contexts.

Members

add_ContainsFullScreenElementChanged

Add an event handler for the ContainsFullScreenElementChanged event.

public HRESULT add_ContainsFullScreenElementChanged(ICoreWebView2ContainsFullScreenElementChangedEventHandler * eventHandler, EventRegistrationToken * token)

ContainsFullScreenElementChanged fires when the ContainsFullScreenElement property changes. This means that an HTML element inside the WebView is entering fullscreen to the size of the WebView or leaving fullscreen. This event is useful when, for example, a video element requests to go fullscreen. The listener of ContainsFullScreenElementChanged can then resize the WebView in response.

    // Register a handler for the ContainsFullScreenChanged event.
    CHECK_FAILURE(m_webView->add_ContainsFullScreenElementChanged(
        Callback<ICoreWebView2ContainsFullScreenElementChangedEventHandler>(
            [this](ICoreWebView2* sender, IUnknown* args) -> HRESULT {
                if (m_fullScreenAllowed)
                {
                    CHECK_FAILURE(
                        sender->get_ContainsFullScreenElement(&m_containsFullscreenElement));
                    if (m_containsFullscreenElement)
                    {
                        EnterFullScreen();
                    }
                    else
                    {
                        ExitFullScreen();
                    }
                }
                return S_OK;
            })
            .Get(),
        nullptr));

add_ContentLoading

Add an event handler for the ContentLoading event.

public HRESULT add_ContentLoading(ICoreWebView2ContentLoadingEventHandler * eventHandler, EventRegistrationToken * token)

ContentLoading fires before any content is loaded, including scripts added with AddScriptToExecuteOnDocumentCreated. ContentLoading will not fire if a same page navigation occurs (such as through fragment navigations or history.pushState navigations). This follows the NavigationStarting and SourceChanged events and precedes the HistoryChanged and NavigationCompleted events.

add_DocumentTitleChanged

Add an event handler for the DocumentTitleChanged event.

public HRESULT add_DocumentTitleChanged(ICoreWebView2DocumentTitleChangedEventHandler * eventHandler, EventRegistrationToken * token)

DocumentTitleChanged fires when the DocumentTitle property of the WebView changes and may fire before or after the NavigationCompleted event.

    // Register a handler for the DocumentTitleChanged event.
    // This handler just announces the new title on the window's title bar.
    CHECK_FAILURE(m_webView->add_DocumentTitleChanged(
        Callback<ICoreWebView2DocumentTitleChangedEventHandler>(
            [this](ICoreWebView2* sender, IUnknown* args) -> HRESULT {
                wil::unique_cotaskmem_string title;
                CHECK_FAILURE(sender->get_DocumentTitle(&title));
                SetWindowText(m_appWindow->GetMainWindow(), title.get());
                return S_OK;
            })
            .Get(),
        &m_documentTitleChangedToken));

add_FrameNavigationCompleted

Add an event handler for the FrameNavigationCompleted event.

public HRESULT add_FrameNavigationCompleted(ICoreWebView2NavigationCompletedEventHandler * eventHandler, EventRegistrationToken * token)

FrameNavigationCompleted fires when a child frame has completely loaded (body.onload has fired) or loading stopped with error.

    // Register a handler for the FrameNavigationCompleted event.
    // Check whether the navigation succeeded, and if not, do something.
    CHECK_FAILURE(m_webView->add_FrameNavigationCompleted(
        Callback<ICoreWebView2NavigationCompletedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2NavigationCompletedEventArgs* args)
                -> HRESULT {
                BOOL success;
                CHECK_FAILURE(args->get_IsSuccess(&success));
                if (!success)
                {
                    COREWEBVIEW2_WEB_ERROR_STATUS webErrorStatus;
                    CHECK_FAILURE(args->get_WebErrorStatus(&webErrorStatus));
                    std::wstring error_msg = WebErrorStatusToString(webErrorStatus);
                    MessageBox(nullptr,
                       (std::wstring(L"IFrame navigation failed with the ") +
                         L"give in error " + error_msg).c_str(),
                       L"Navigation Failure", MB_OK);
                    if (webErrorStatus == COREWEBVIEW2_WEB_ERROR_STATUS_DISCONNECTED)
                    {
                        // Do something here if you want to handle a specific error case.
                        // In most cases this isn't necessary, because the WebView will
                        // display its own error page automatically.
                    }
                }
                return S_OK;
            })
            .Get(),
        &m_frameNavigationCompletedToken));

add_FrameNavigationStarting

Add an event handler for the FrameNavigationStarting event.

public HRESULT add_FrameNavigationStarting(ICoreWebView2NavigationStartingEventHandler * eventHandler, EventRegistrationToken * token)

FrameNavigationStarting fires when a child frame in the WebView requests permission to navigate to a different URI. This will fire for redirects as well.

Corresponding navigations can be blocked until the event handler returns.

    // Register a handler for the FrameNavigationStarting event.
    // This handler will prevent a frame from navigating to a blocked domain.
    CHECK_FAILURE(m_webView->add_FrameNavigationStarting(
        Callback<ICoreWebView2NavigationStartingEventHandler>(
            [this](ICoreWebView2* sender,
                   ICoreWebView2NavigationStartingEventArgs* args) -> HRESULT
    {
        wil::unique_cotaskmem_string uri;
        CHECK_FAILURE(args->get_Uri(&uri));

        if (ShouldBlockUri(uri.get()))
        {
            CHECK_FAILURE(args->put_Cancel(true));
        }
        return S_OK;
    }).Get(), &m_frameNavigationStartingToken));

add_HistoryChanged

Add an event handler for the HistoryChanged event.

public HRESULT add_HistoryChanged(ICoreWebView2HistoryChangedEventHandler * eventHandler, EventRegistrationToken * token)

HistoryChanged listens to the change of navigation history for the top level document. Use HistoryChanged to check if CanGoBack/CanGoForward value has changed. HistoryChanged also fires for using GoBack/GoForward. HistoryChanged fires after SourceChanged and ContentLoading.

    // Register a handler for the HistoryChanged event.
    // Update the Back, Forward buttons.
    CHECK_FAILURE(m_webView->add_HistoryChanged(
        Callback<ICoreWebView2HistoryChangedEventHandler>(
            [this](ICoreWebView2* sender, IUnknown* args) -> HRESULT {
                BOOL canGoBack;
                BOOL canGoForward;
                sender->get_CanGoBack(&canGoBack);
                sender->get_CanGoForward(&canGoForward);
                m_toolbar->SetItemEnabled(Toolbar::Item_BackButton, canGoBack);
                m_toolbar->SetItemEnabled(Toolbar::Item_ForwardButton, canGoForward);

                return S_OK;
            })
            .Get(),
        &m_historyChangedToken));

add_NavigationCompleted

Add an event handler for the NavigationCompleted event.

public HRESULT add_NavigationCompleted(ICoreWebView2NavigationCompletedEventHandler * eventHandler, EventRegistrationToken * token)

NavigationCompleted fires when the WebView has completely loaded (body.onload has fired) or loading stopped with error.

    // Register a handler for the NavigationCompleted event.
    // Check whether the navigation succeeded, and if not, do something.
    // Also update the Cancel buttons.
    CHECK_FAILURE(m_webView->add_NavigationCompleted(
        Callback<ICoreWebView2NavigationCompletedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2NavigationCompletedEventArgs* args)
                -> HRESULT {
                BOOL success;
                CHECK_FAILURE(args->get_IsSuccess(&success));
                if (!success)
                {
                    COREWEBVIEW2_WEB_ERROR_STATUS webErrorStatus;
                    CHECK_FAILURE(args->get_WebErrorStatus(&webErrorStatus));
                    if (webErrorStatus == COREWEBVIEW2_WEB_ERROR_STATUS_DISCONNECTED)
                    {
                        // Do something here if you want to handle a specific error case.
                        // In most cases this isn't necessary, because the WebView will
                        // display its own error page automatically.
                    }
                }
                m_toolbar->SetItemEnabled(Toolbar::Item_CancelButton, false);
                m_toolbar->SetItemEnabled(Toolbar::Item_ReloadButton, true);
                return S_OK;
            })
            .Get(),
        &m_navigationCompletedToken));

add_NavigationStarting

Add an event handler for the NavigationStarting event.

public HRESULT add_NavigationStarting(ICoreWebView2NavigationStartingEventHandler * eventHandler, EventRegistrationToken * token)

NavigationStarting fires when the WebView main frame is requesting permission to navigate to a different URI. This will fire for redirects as well.

Corresponding navigations can be blocked until the event handler returns.

    // Register a handler for the NavigationStarting event.
    // This handler will check the domain being navigated to, and if the domain
    // matches a list of blocked sites, it will cancel the navigation and
    // possibly display a warning page.  It will also disable JavaScript on
    // selected websites.
    CHECK_FAILURE(m_webView->add_NavigationStarting(
        Callback<ICoreWebView2NavigationStartingEventHandler>(
            [this](ICoreWebView2* sender,
                   ICoreWebView2NavigationStartingEventArgs* args) -> HRESULT
    {
        wil::unique_cotaskmem_string uri;
        CHECK_FAILURE(args->get_Uri(&uri));

        if (ShouldBlockUri(uri.get()))
        {
            CHECK_FAILURE(args->put_Cancel(true));

            // If the user clicked a link to navigate, show a warning page.
            BOOL userInitiated;
            CHECK_FAILURE(args->get_IsUserInitiated(&userInitiated));
            static const PCWSTR htmlContent =
              L"<h1>Domain Blocked</h1>"
              L"<p>You've attempted to navigate to a domain in the blocked "
              L"sites list. Press back to return to the previous page.</p>";
            CHECK_FAILURE(sender->NavigateToString(htmlContent));
        }
        // Changes to settings will apply at the next navigation, which includes the
        // navigation after a NavigationStarting event.  We can use this to change
        // settings according to what site we're visiting.
        if (ShouldBlockScriptForUri(uri.get()))
        {
            m_settings->put_IsScriptEnabled(FALSE);
        }
        else
        {
            m_settings->put_IsScriptEnabled(m_isScriptEnabled);
        }
        return S_OK;
    }).Get(), &m_navigationStartingToken));

add_NewWindowRequested

Add an event handler for the NewWindowRequested event.

public HRESULT add_NewWindowRequested(ICoreWebView2NewWindowRequestedEventHandler * eventHandler, EventRegistrationToken * token)

NewWindowRequested fires when content inside the WebView requests to open a new window, such as through window.open. The app can pass a target WebView that will be considered the opened window.

Scripts resulted in the new window requested can be blocked until the event handler returns if a deferral is not taken on the event args. If a deferral is taken, then scripts are blocked until the deferral is completed.

    // Register a handler for the NewWindowRequested event.
    // This handler will defer the event, create a new app window, and then once the
    // new window is ready, it'll provide that new window's WebView as the response to
    // the request.
    CHECK_FAILURE(m_webView->add_NewWindowRequested(
        Callback<ICoreWebView2NewWindowRequestedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2NewWindowRequestedEventArgs* args) {
                wil::com_ptr<ICoreWebView2Deferral> deferral;
                CHECK_FAILURE(args->GetDeferral(&deferral));
                AppWindow* newAppWindow;

                wil::com_ptr<ICoreWebView2WindowFeatures> windowFeatures;
                CHECK_FAILURE(args->get_WindowFeatures(&windowFeatures));

                RECT windowRect = {0};
                UINT32 left = 0;
                UINT32 top = 0;
                UINT32 height = 0;
                UINT32 width = 0;
                BOOL shouldHaveToolbar = true;

                BOOL hasPosition = FALSE;
                BOOL hasSize = FALSE;
                CHECK_FAILURE(windowFeatures->HasPosition(&hasPosition));
                CHECK_FAILURE(windowFeatures->HasSize(&hasSize));

                bool useDefaultWindow = true;

                if (!!hasPosition && !!hasSize)
                {
                    CHECK_FAILURE(windowFeatures->get_Left(&left));
                    CHECK_FAILURE(windowFeatures->get_Top(&top));
                    CHECK_FAILURE(windowFeatures->get_Height(&height));
                    CHECK_FAILURE(windowFeatures->get_Width(&width));
                    useDefaultWindow = false;
                }
                CHECK_FAILURE(windowFeatures->get_Toolbar(&shouldHaveToolbar));

                windowRect.left = left;
                windowRect.right = left + (width < s_minNewWindowSize ? s_minNewWindowSize : width);
                windowRect.top = top;
                windowRect.bottom = top + (height < s_minNewWindowSize ? s_minNewWindowSize : height);

                if (!useDefaultWindow)
                {
                  newAppWindow = new AppWindow(m_creationModeId, L"", false, nullptr, true, windowRect, !!shouldHaveToolbar);
                }
                else
                {
                  newAppWindow = new AppWindow(m_creationModeId, L"");
                }
                newAppWindow->m_isPopupWindow = true;
                newAppWindow->m_onWebViewFirstInitialized = [args, deferral, newAppWindow]() {
                    CHECK_FAILURE(args->put_NewWindow(newAppWindow->m_webView.get()));
                    CHECK_FAILURE(args->put_Handled(TRUE));
                    CHECK_FAILURE(deferral->Complete());
                };

                return S_OK;
            })
            .Get(),
        nullptr));

add_PermissionRequested

Add an event handler for the PermissionRequested event.

public HRESULT add_PermissionRequested(ICoreWebView2PermissionRequestedEventHandler * eventHandler, EventRegistrationToken * token)

PermissionRequested fires when content in a WebView requests permission to access some privileged resources.

If a deferral is not taken on the event args, the subsequent scripts can be blocked until the event handler returns. If a deferral is taken, then the scripts are blocked until the deferral is completed.

    // Register a handler for the PermissionRequested event.
    // This handler prompts the user to allow or deny the request.
    CHECK_FAILURE(m_webView->add_PermissionRequested(
        Callback<ICoreWebView2PermissionRequestedEventHandler>(
            [this](
                ICoreWebView2* sender,
                ICoreWebView2PermissionRequestedEventArgs* args) -> HRESULT
    {
        wil::unique_cotaskmem_string uri;
        COREWEBVIEW2_PERMISSION_KIND kind = COREWEBVIEW2_PERMISSION_KIND_UNKNOWN_PERMISSION;
        BOOL userInitiated = FALSE;

        CHECK_FAILURE(args->get_Uri(&uri));
        CHECK_FAILURE(args->get_PermissionKind(&kind));
        CHECK_FAILURE(args->get_IsUserInitiated(&userInitiated));

        std::wstring message = L"Do you want to grant permission for ";
        message += NameOfPermissionKind(kind);
        message += L" to the website at ";
        message += uri.get();
        message += L"?\n\n";
        message += (userInitiated
            ? L"This request came from a user gesture."
            : L"This request did not come from a user gesture.");

        int response = MessageBox(nullptr, message.c_str(), L"Permission Request",
                                   MB_YESNOCANCEL | MB_ICONWARNING);

        COREWEBVIEW2_PERMISSION_STATE state =
              response == IDYES ? COREWEBVIEW2_PERMISSION_STATE_ALLOW
            : response == IDNO  ? COREWEBVIEW2_PERMISSION_STATE_DENY
            :                     COREWEBVIEW2_PERMISSION_STATE_DEFAULT;
        CHECK_FAILURE(args->put_State(state));

        return S_OK;
    }).Get(), &m_permissionRequestedToken));

add_ProcessFailed

Add an event handler for the ProcessFailed event.

public HRESULT add_ProcessFailed(ICoreWebView2ProcessFailedEventHandler * eventHandler, EventRegistrationToken * token)

ProcessFailed fires when a WebView process is terminated unexpectedly or becomes unresponsive.

    // Register a handler for the ProcessFailed event.
    // This handler checks if the browser process failed, and asks the user if
    // they want to recreate the webview.
    CHECK_FAILURE(m_webView->add_ProcessFailed(
        Callback<ICoreWebView2ProcessFailedEventHandler>(
            [this](ICoreWebView2* sender,
                ICoreWebView2ProcessFailedEventArgs* args) -> HRESULT
    {
        COREWEBVIEW2_PROCESS_FAILED_KIND failureType;
        CHECK_FAILURE(args->get_ProcessFailedKind(&failureType));
        if (failureType == COREWEBVIEW2_PROCESS_FAILED_KIND_BROWSER_PROCESS_EXITED)
        {
            int button = MessageBox(
                m_appWindow->GetMainWindow(),
                L"Browser process exited unexpectedly.  Recreate webview?",
                L"Browser process exited",
                MB_YESNO);
            if (button == IDYES)
            {
                m_appWindow->ReinitializeWebView();
            }
        }
        else if (failureType == COREWEBVIEW2_PROCESS_FAILED_KIND_RENDER_PROCESS_UNRESPONSIVE)
        {
            int button = MessageBox(
                m_appWindow->GetMainWindow(),
                L"Browser render process has stopped responding.  Recreate webview?",
                L"Web page unresponsive", MB_YESNO);
            if (button == IDYES)
            {
                m_appWindow->ReinitializeWebView();
            }
        }
        else if (failureType == COREWEBVIEW2_PROCESS_FAILED_KIND_RENDER_PROCESS_EXITED)
        {
            int button = MessageBox(
                m_appWindow->GetMainWindow(),
                L"Browser render process exited unexpectedly. Reload page?",
                L"Web page unresponsive", MB_YESNO);
            if (button == IDYES)
            {
                CHECK_FAILURE(m_webView->Reload());
            }
        }
        return S_OK;
    }).Get(), &m_processFailedToken));

add_ScriptDialogOpening

Add an event handler for the ScriptDialogOpening event.

public HRESULT add_ScriptDialogOpening(ICoreWebView2ScriptDialogOpeningEventHandler * eventHandler, EventRegistrationToken * token)

ScriptDialogOpening fires when a JavaScript dialog (alert, confirm, prompt, or beforeunload) will show for the webview. This event only fires if the ICoreWebView2Settings::AreDefaultScriptDialogsEnabled property is set to false. The ScriptDialogOpening event can be used to suppress dialogs or replace default dialogs with custom dialogs.

If a deferral is not taken on the event args, the subsequent scripts can be blocked until the event handler returns. If a deferral is taken, then the scripts are blocked until the deferral is completed.

    // Register a handler for the ScriptDialogOpening event.
    // This handler will set up a custom prompt dialog for the user,
    // and may defer the event if the setting to defer dialogs is enabled.
    CHECK_FAILURE(m_webView->add_ScriptDialogOpening(
        Callback<ICoreWebView2ScriptDialogOpeningEventHandler>(
            [this](
                ICoreWebView2* sender,
                ICoreWebView2ScriptDialogOpeningEventArgs* args) -> HRESULT
    {
        wil::com_ptr<ICoreWebView2ScriptDialogOpeningEventArgs> eventArgs = args;
        auto showDialog = [this, eventArgs]
        {
            wil::unique_cotaskmem_string uri;
            COREWEBVIEW2_SCRIPT_DIALOG_KIND type;
            wil::unique_cotaskmem_string message;
            wil::unique_cotaskmem_string defaultText;

            CHECK_FAILURE(eventArgs->get_Uri(&uri));
            CHECK_FAILURE(eventArgs->get_Kind(&type));
            CHECK_FAILURE(eventArgs->get_Message(&message));
            CHECK_FAILURE(eventArgs->get_DefaultText(&defaultText));

            std::wstring promptString = std::wstring(L"The page at '")
                + uri.get() + L"' says:";
            TextInputDialog dialog(
                m_appWindow->GetMainWindow(),
                L"Script Dialog",
                promptString.c_str(),
                message.get(),
                defaultText.get(),
                /* readonly */ type != COREWEBVIEW2_SCRIPT_DIALOG_KIND_PROMPT);
            if (dialog.confirmed)
            {
                CHECK_FAILURE(eventArgs->put_ResultText(dialog.input.c_str()));
                CHECK_FAILURE(eventArgs->Accept());
            }
        };

        if (m_deferScriptDialogs)
        {
            wil::com_ptr<ICoreWebView2Deferral> deferral;
            CHECK_FAILURE(args->GetDeferral(&deferral));
            m_completeDeferredDialog = [showDialog, deferral]
            {
                showDialog();
                CHECK_FAILURE(deferral->Complete());
            };
        }
        else
        {
            showDialog();
        }

        return S_OK;
    }).Get(), &m_scriptDialogOpeningToken));

add_SourceChanged

Add an event handler for the SourceChanged event.

public HRESULT add_SourceChanged(ICoreWebView2SourceChangedEventHandler * eventHandler, EventRegistrationToken * token)

SourceChanged fires when the Source property changes. SourceChanged fires for navigating to a different site or fragment navigations. It will not fire for other types of navigations such as page reloads or history.pushState with the same URL as the current page. SourceChanged fires before ContentLoading for navigation to a new document.

    // Register a handler for the SourceChanged event.
    // This handler will read the webview's source URI and update
    // the app's address bar.
    CHECK_FAILURE(m_webView->add_SourceChanged(
        Callback<ICoreWebView2SourceChangedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2SourceChangedEventArgs* args)
                -> HRESULT {
                wil::unique_cotaskmem_string uri;
                sender->get_Source(&uri);
                if (wcscmp(uri.get(), L"about:blank") == 0)
                {
                    uri = wil::make_cotaskmem_string(L"");
                }
                SetWindowText(GetAddressBar(), uri.get());

                return S_OK;
            })
            .Get(),
        &m_sourceChangedToken));

add_WebMessageReceived

Add an event handler for the WebMessageReceived event.

public HRESULT add_WebMessageReceived(ICoreWebView2WebMessageReceivedEventHandler * handler, EventRegistrationToken * token)

WebMessageReceived fires when the ICoreWebView2Settings::IsWebMessageEnabled setting is set and the top level document of the WebView calls window.chrome.webview.postMessage. The postMessage function is void postMessage(object) where object is any object supported by JSON conversion.

        window.chrome.webview.addEventListener('message', arg => {
            if ("SetColor" in arg.data) {
                document.getElementById("colorable").style.color = arg.data.SetColor;
            }
            if ("WindowBounds" in arg.data) {
                document.getElementById("window-bounds").value = arg.data.WindowBounds;
            }
        });

        function SetTitleText() {
            let titleText = document.getElementById("title-text");
            window.chrome.webview.postMessage(`SetTitleText ${titleText.value}`);
        }
        function GetWindowBounds() {
            window.chrome.webview.postMessage("GetWindowBounds");
        }

When postMessage is called, the handler's Invoke method will be called with the postMessage's object parameter converted to a JSON string.

    // Setup the web message received event handler before navigating to
    // ensure we don't miss any messages.
    CHECK_FAILURE(m_webView->add_WebMessageReceived(
        Microsoft::WRL::Callback<ICoreWebView2WebMessageReceivedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2WebMessageReceivedEventArgs* args)
    {
        wil::unique_cotaskmem_string uri;
        CHECK_FAILURE(args->get_Source(&uri));

        // Always validate that the origin of the message is what you expect.
        if (uri.get() != m_sampleUri)
        {
            return S_OK;
        }
        wil::unique_cotaskmem_string messageRaw;
        CHECK_FAILURE(args->TryGetWebMessageAsString(&messageRaw));
        std::wstring message = messageRaw.get();

        if (message.compare(0, 13, L"SetTitleText ") == 0)
        {
            m_appWindow->SetTitleText(message.substr(13).c_str());
        }
        else if (message.compare(L"GetWindowBounds") == 0)
        {
            RECT bounds = m_appWindow->GetWindowBounds();
            std::wstring reply =
                L"{\"WindowBounds\":\"Left:" + std::to_wstring(bounds.left)
                + L"\\nTop:" + std::to_wstring(bounds.top)
                + L"\\nRight:" + std::to_wstring(bounds.right)
                + L"\\nBottom:" + std::to_wstring(bounds.bottom)
                + L"\"}";
            CHECK_FAILURE(sender->PostWebMessageAsJson(reply.c_str()));
        }
        return S_OK;
    }).Get(), &m_webMessageReceivedToken));

add_WebResourceRequested

Add an event handler for the WebResourceRequested event.

public HRESULT add_WebResourceRequested(ICoreWebView2WebResourceRequestedEventHandler * eventHandler, EventRegistrationToken * token)

WebResourceRequested fires when the WebView is performing a URL request to a matching URL and resource context filter that was added with AddWebResourceRequestedFilter. At least one filter must be added for the event to fire.

The web resource requested can be blocked until the event handler returns if a deferral is not taken on the event args. If a deferral is taken, then the web resource requested is blocked until the deferral is completed.

        if (m_blockImages)
        {
            m_webView->AddWebResourceRequestedFilter(L"*", COREWEBVIEW2_WEB_RESOURCE_CONTEXT_IMAGE);
            CHECK_FAILURE(m_webView->add_WebResourceRequested(
                Callback<ICoreWebView2WebResourceRequestedEventHandler>(
                    [this](
                        ICoreWebView2* sender,
                        ICoreWebView2WebResourceRequestedEventArgs* args) {
                        COREWEBVIEW2_WEB_RESOURCE_CONTEXT resourceContext;
                        CHECK_FAILURE(
                            args->get_ResourceContext(&resourceContext));
                        // Ensure that the type is image
                        if (resourceContext != COREWEBVIEW2_WEB_RESOURCE_CONTEXT_IMAGE)
                        {
                            return E_INVALIDARG;
                        }
                        // Override the response with an empty one to block the image.
                        // If put_Response is not called, the request will continue as normal.
                        wil::com_ptr<ICoreWebView2WebResourceResponse> response;
                        CHECK_FAILURE(m_webViewEnvironment->CreateWebResourceResponse(
                            nullptr, 403 /*NoContent*/, L"Blocked", L"", &response));
                        CHECK_FAILURE(args->put_Response(response.get()));
                        return S_OK;
                    })
                    .Get(),
                &m_webResourceRequestedTokenForImageBlocking));
        }
        else
        {
            CHECK_FAILURE(m_webView->remove_WebResourceRequested(
                m_webResourceRequestedTokenForImageBlocking));
        }

add_WindowCloseRequested

Add an event handler for the WindowCloseRequested event.

public HRESULT add_WindowCloseRequested(ICoreWebView2WindowCloseRequestedEventHandler * eventHandler, EventRegistrationToken * token)

WindowCloseRequested fires when content inside the WebView requested to close the window, such as after window.close is called. The app should close the WebView and related app window if that makes sense to the app.

    // Register a handler for the WindowCloseRequested event.
    // This handler will close the app window if it is not the main window.
    CHECK_FAILURE(m_webView->add_WindowCloseRequested(
        Callback<ICoreWebView2WindowCloseRequestedEventHandler>([this](
                                                                    ICoreWebView2* sender,
                                                                    IUnknown* args) {
            if (m_isPopupWindow)
            {
                CloseAppWindow();
            }
            return S_OK;
        }).Get(),
        nullptr));

AddHostObjectToScript

Add the provided host object to script running in the WebView with the specified name.

public HRESULT AddHostObjectToScript(LPCWSTR name, VARIANT * object)

Host objects are exposed as host object proxies via window.chrome.webview.hostObjects.<name>. Host object proxies are promises and will resolve to an object representing the host object. The promise is rejected if the app has not added an object with the name. When JavaScript code access a property or method of the object, a promise is return, which will resolve to the value returned from the host for the property or method, or rejected in case of error such as there is no such property or method on the object or parameters are invalid. For example, when the application code does the following:

VARIANT object;
object.vt = VT_DISPATCH;
object.pdispVal = appObject;
webview->AddHostObjectToScript(L"host_object", &host);

JavaScript code in the WebView will be able to access appObject as following and then access attributes and methods of appObject:

let app_object = await window.chrome.webview.hostObjects.host_object;
let attr1 = await app_object.attr1;
let result = await app_object.method1(parameters);

Note that while simple types, IDispatch and array are supported, generic IUnknown, VT_DECIMAL, or VT_RECORD variant is not supported. Remote JavaScript objects like callback functions are represented as an VT_DISPATCH VARIANT with the object implementing IDispatch. The JavaScript callback method may be invoked using DISPID_VALUE for the DISPID. Nested arrays are supported up to a depth of 3. Arrays of by reference types are not supported. VT_EMPTY and VT_NULL are mapped into JavaScript as null. In JavaScript null and undefined are mapped to VT_EMPTY.

Additionally, all host objects are exposed as window.chrome.webview.hostObjects.sync.<name>. Here the host objects are exposed as synchronous host object proxies. These are not promises and calls to functions or property access synchronously block running script waiting to communicate cross process for the host code to run. Accordingly this can result in reliability issues and it is recommended that you use the promise based asynchronous window.chrome.webview.hostObjects.<name> API described above.

Synchronous host object proxies and asynchronous host object proxies can both proxy the same host object. Remote changes made by one proxy will be reflected in any other proxy of that same host object whether the other proxies and synchronous or asynchronous.

While JavaScript is blocked on a synchronous call to native code, that native code is unable to call back to JavaScript. Attempts to do so will fail with HRESULT_FROM_WIN32(ERROR_POSSIBLE_DEADLOCK).

Host object proxies are JavaScript Proxy objects that intercept all property get, property set, and method invocations. Properties or methods that are a part of the Function or Object prototype are run locally. Additionally any property or method in the array chrome.webview.hostObjects.options.forceLocalProperties will also be run locally. This defaults to including optional methods that have meaning in JavaScript like toJSON and Symbol.toPrimitive. You can add more to this array as required.

There's a method chrome.webview.hostObjects.cleanupSome that will best effort garbage collect host object proxies.

Host object proxies additionally have the following methods which run locally:

  • applyHostFunction, getHostProperty, setHostProperty: Perform a method invocation, property get, or property set on the host object. You can use these to explicitly force a method or property to run remotely if there is a conflicting local method or property. For instance, proxy.toString() will run the local toString method on the proxy object. But proxy.applyHostFunction('toString') runs toString on the host proxied object instead.

  • getLocalProperty, setLocalProperty: Perform property get, or property set locally. You can use these methods to force getting or setting a property on the host object proxy itself rather than on the host object it represents. For instance, proxy.unknownProperty will get the property named unknownProperty from the host proxied object. But proxy.getLocalProperty('unknownProperty') will get the value of the property unknownProperty on the proxy object itself.

  • sync: Asynchronous host object proxies expose a sync method which returns a promise for a synchronous host object proxy for the same host object. For example, chrome.webview.hostObjects.sample.methodCall() returns an asynchronous host object proxy. You can use the sync method to obtain a synchronous host object proxy instead: const syncProxy = await chrome.webview.hostObjects.sample.methodCall().sync()

  • async: Synchronous host object proxies expose an async method which blocks and returns an asynchronous host object proxy for the same host object. For example, chrome.webview.hostObjects.sync.sample.methodCall() returns a synchronous host object proxy. Calling the async method on this blocks and then returns an asynchronous host object proxy for the same host object: const asyncProxy = chrome.webview.hostObjects.sync.sample.methodCall().async()

  • then: Asynchronous host object proxies have a then method. This allows them to be awaitable. then will return a promise that resolves with a representation of the host object. If the proxy represents a JavaScript literal then a copy of that is returned locally. If the proxy represents a function then a non-awaitable proxy is returned. If the proxy represents a JavaScript object with a mix of literal properties and function properties, then the a copy of the object is returned with some properties as host object proxies.

All other property and method invocations (other than the above Remote object proxy methods, forceLocalProperties list, and properties on Function and Object prototypes) are run remotely. Asynchronous host object proxies return a promise representing asynchronous completion of remotely invoking the method, or getting the property. The promise resolves after the remote operations complete and the promises resolve to the resulting value of the operation. Synchronous host object proxies work similarly but block JavaScript execution and wait for the remote operation to complete.

Setting a property on an asynchronous host object proxy works slightly differently. The set returns immediately and the return value is the value that will be set. This is a requirement of the JavaScript Proxy object. If you need to asynchronously wait for the property set to complete, use the setHostProperty method which returns a promise as described above. Synchronous object property set property synchronously blocks until the property is set.

For example, suppose you have a COM object with the following interface

    [uuid(3a14c9c0-bc3e-453f-a314-4ce4a0ec81d8), object, local]
    interface IHostObjectSample : IUnknown
    {
        // Demonstrate basic method call with some parameters and a return value.
        HRESULT MethodWithParametersAndReturnValue([in] BSTR stringParameter, [in] INT integerParameter, [out, retval] BSTR* stringResult);

        // Demonstrate getting and setting a property.
        [propget] HRESULT Property([out, retval] BSTR* stringResult);
        [propput] HRESULT Property([in] BSTR stringValue);

        [propget] HRESULT IndexedProperty(INT index, [out, retval] BSTR * stringResult);
        [propput] HRESULT IndexedProperty(INT index, [in] BSTR stringValue);

        // Demonstrate native calling back into JavaScript.
        HRESULT CallCallbackAsynchronously([in] IDispatch* callbackParameter);

    };

We can add an instance of this interface into our JavaScript with AddHostObjectToScript. In this case we name it sample:

            VARIANT remoteObjectAsVariant = {};
            m_hostObject.query_to<IDispatch>(&remoteObjectAsVariant.pdispVal);
            remoteObjectAsVariant.vt = VT_DISPATCH;

            // We can call AddHostObjectToScript multiple times in a row without
            // calling RemoveHostObject first. This will replace the previous object
            // with the new object. In our case this is the same object and everything
            // is fine.
            CHECK_FAILURE(
                m_webView->AddHostObjectToScript(L"sample", &remoteObjectAsVariant));
            remoteObjectAsVariant.pdispVal->Release();

Then in the HTML document we can use this COM object via chrome.webview.hostObjects.sample:

        document.getElementById("getPropertyAsyncButton").addEventListener("click", async () => {
        const propertyValue = await chrome.webview.hostObjects.sample.property;
        document.getElementById("getPropertyAsyncOutput").textContent = propertyValue;
        });

        document.getElementById("getPropertySyncButton").addEventListener("click", () => {
        const propertyValue = chrome.webview.hostObjects.sync.sample.property;
        document.getElementById("getPropertySyncOutput").textContent = propertyValue;
        });

        document.getElementById("setPropertyAsyncButton").addEventListener("click", async () => {
        const propertyValue = document.getElementById("setPropertyAsyncInput").value;
        // The following line will work but it will return immediately before the property value has actually been set.
        // If you need to set the property and wait for the property to change value, use the setHostProperty function.
        chrome.webview.hostObjects.sample.property = propertyValue;
        document.getElementById("setPropertyAsyncOutput").textContent = "Set";
        });

        document.getElementById("setPropertyExplicitAsyncButton").addEventListener("click", async () => {
        const propertyValue = document.getElementById("setPropertyExplicitAsyncInput").value;
        // If you care about waiting until the property has actually changed value use the setHostProperty function.
        await chrome.webview.hostObjects.sample.setHostProperty("property", propertyValue);
        document.getElementById("setPropertyExplicitAsyncOutput").textContent = "Set";
        });

        document.getElementById("setPropertySyncButton").addEventListener("click", () => {
        const propertyValue = document.getElementById("setPropertySyncInput").value;
        chrome.webview.hostObjects.sync.sample.property = propertyValue;
        document.getElementById("setPropertySyncOutput").textContent = "Set";
        });

        document.getElementById("getIndexedPropertyAsyncButton").addEventListener("click", async () => {
        const index = parseInt(document.getElementById("getIndexedPropertyAsyncParam").value);
        const resultValue = await chrome.webview.hostObjects.sample.IndexedProperty[index];
        document.getElementById("getIndexedPropertyAsyncOutput").textContent = resultValue;
        });
        document.getElementById("setIndexedPropertyAsyncButton").addEventListener("click", async () => {
        const index = parseInt(document.getElementById("setIndexedPropertyAsyncParam1").value);
        const value = document.getElementById("setIndexedPropertyAsyncParam2").value;;
        chrome.webview.hostObjects.sample.IndexedProperty[index] = value;
        document.getElementById("setIndexedPropertyAsyncOutput").textContent = "Set";
        });
        document.getElementById("invokeMethodAsyncButton").addEventListener("click", async () => {
        const paramValue1 = document.getElementById("invokeMethodAsyncParam1").value;
        const paramValue2 = parseInt(document.getElementById("invokeMethodAsyncParam2").value);
        const resultValue = await chrome.webview.hostObjects.sample.MethodWithParametersAndReturnValue(paramValue1, paramValue2);
        document.getElementById("invokeMethodAsyncOutput").textContent = resultValue;
        });

        document.getElementById("invokeMethodSyncButton").addEventListener("click", () => {
        const paramValue1 = document.getElementById("invokeMethodSyncParam1").value;
        const paramValue2 = parseInt(document.getElementById("invokeMethodSyncParam2").value);
        const resultValue = chrome.webview.hostObjects.sync.sample.MethodWithParametersAndReturnValue(paramValue1, paramValue2);
        document.getElementById("invokeMethodSyncOutput").textContent = resultValue;
        });

        let callbackCount = 0;
        document.getElementById("invokeCallbackButton").addEventListener("click", async () => {
        chrome.webview.hostObjects.sample.CallCallbackAsynchronously(() => {
        document.getElementById("invokeCallbackOutput").textContent = "Native object called the callback " + (++callbackCount) + " time(s).";
        });
        });

Exposing host objects to script has security risk. Please follow best practices.

AddScriptToExecuteOnDocumentCreated

Add the provided JavaScript to a list of scripts that should be executed after the global object has been created, but before the HTML document has been parsed and before any other script included by the HTML document is executed.

public HRESULT AddScriptToExecuteOnDocumentCreated(LPCWSTR javaScript, ICoreWebView2AddScriptToExecuteOnDocumentCreatedCompletedHandler * handler)

This method injects a script that runs on all top-level document and child frame page navigations. This method runs asynchronously, and you must wait for the completion handler to finish before the injected script is ready to run. When this method completes, the handler's Invoke method is called with the id of the injected script. id is a string. To remove the injected script, use RemoveScriptToExecuteOnDocumentCreated.

Note that if an HTML document has sandboxing of some kind via sandbox properties or the Content-Security-Policy HTTP header this will affect the script run here. So, for example, if the 'allow-modals' keyword is not set then calls to the alert function will be ignored.

// Prompt the user for some script and register it to execute whenever a new page loads.
void ScriptComponent::AddInitializeScript()
{
    TextInputDialog dialog(
        m_appWindow->GetMainWindow(),
        L"Add Initialize Script",
        L"Initialization Script:",
        L"Enter the JavaScript code to run as the initialization script that "
            L"runs before any script in the HTML document.",
    // This example script stops child frames from opening new windows.  Because
    // the initialization script runs before any script in the HTML document, we
    // can trust the results of our checks on window.parent and window.top.
        L"if (window.parent !== window.top) {\r\n"
        L"    delete window.open;\r\n"
        L"}");
    if (dialog.confirmed)
    {
        m_webView->AddScriptToExecuteOnDocumentCreated(
            dialog.input.c_str(),
            Callback<ICoreWebView2AddScriptToExecuteOnDocumentCreatedCompletedHandler>(
                [this](HRESULT error, PCWSTR id) -> HRESULT
        {
            m_lastInitializeScriptId = id;
            MessageBox(nullptr, id, L"AddScriptToExecuteOnDocumentCreated Id", MB_OK);
            return S_OK;
        }).Get());

    }
}

AddWebResourceRequestedFilter

Adds a URI and resource context filter to the WebResourceRequested event.

public HRESULT AddWebResourceRequestedFilter(LPCWSTR const uri, COREWEBVIEW2_WEB_RESOURCE_CONTEXT const resourceContext)

The URI parameter can be a wildcard string ('*': zero or more, '?': exactly one). nullptr is equivalent to L"". See COREWEBVIEW2_WEB_RESOURCE_CONTEXT enum for description of resource context filters.

CallDevToolsProtocolMethod

Call an asynchronous DevToolsProtocol method.

public HRESULT CallDevToolsProtocolMethod(LPCWSTR methodName, LPCWSTR parametersAsJson, ICoreWebView2CallDevToolsProtocolMethodCompletedHandler * handler)

See the DevTools Protocol Viewer for a list and description of available methods. The methodName parameter is the full name of the method in the format {domain}.{method}. The parametersAsJson parameter is a JSON formatted string containing the parameters for the corresponding method. The handler's Invoke method will be called when the method asynchronously completes. Invoke will be called with the method's return object as a JSON string.

// Prompt the user for the name and parameters of a CDP method, then call it.
void ScriptComponent::CallCdpMethod()
{
    TextInputDialog dialog(
        m_appWindow->GetMainWindow(),
        L"Call CDP Method",
        L"CDP method name:",
        L"Enter the CDP method name to call, followed by a space,\r\n"
            L"followed by the parameters in JSON format.",
        L"Runtime.evaluate {\"expression\":\"alert(\\\"test\\\")\"}");
    if (dialog.confirmed)
    {
        size_t delimiterPos = dialog.input.find(L' ');
        std::wstring methodName = dialog.input.substr(0, delimiterPos);
        std::wstring methodParams =
            (delimiterPos < dialog.input.size()
                ? dialog.input.substr(delimiterPos + 1)
                : L"{}");

        m_webView->CallDevToolsProtocolMethod(
            methodName.c_str(),
            methodParams.c_str(),
            Callback<ICoreWebView2CallDevToolsProtocolMethodCompletedHandler>(
                [](HRESULT error, PCWSTR resultJson) -> HRESULT
                {
                    MessageBox(nullptr, resultJson, L"CDP Method Result", MB_OK);
                    return S_OK;
                }).Get());
    }
}

CapturePreview

Capture an image of what WebView is displaying.

public HRESULT CapturePreview(COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT imageFormat, IStream * imageStream, ICoreWebView2CapturePreviewCompletedHandler * handler)

Specify the format of the image with the imageFormat parameter. The resulting image binary data is written to the provided imageStream parameter. When CapturePreview finishes writing to the stream, the Invoke method on the provided handler parameter is called.

// Show the user a file selection dialog, then save a screenshot of the WebView
// to the selected file.
void FileComponent::SaveScreenshot()
{
    OPENFILENAME openFileName = {};
    openFileName.lStructSize = sizeof(openFileName);
    openFileName.hwndOwner = nullptr;
    openFileName.hInstance = nullptr;
    WCHAR fileName[MAX_PATH] = L"WebView2_Screenshot.png";
    openFileName.lpstrFile = fileName;
    openFileName.lpstrFilter = L"PNG File\0*.png\0";
    openFileName.nMaxFile = ARRAYSIZE(fileName);
    openFileName.Flags = OFN_OVERWRITEPROMPT;

    if (GetSaveFileName(&openFileName))
    {
        wil::com_ptr<IStream> stream;
        CHECK_FAILURE(SHCreateStreamOnFileEx(
            fileName, STGM_READWRITE | STGM_CREATE, FILE_ATTRIBUTE_NORMAL, TRUE, nullptr,
            &stream));

        HWND mainWindow = m_appWindow->GetMainWindow();

        CHECK_FAILURE(m_webView->CapturePreview(
            COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT_PNG, stream.get(),
            Callback<ICoreWebView2CapturePreviewCompletedHandler>(
                [mainWindow](HRESULT error_code) -> HRESULT {
                    CHECK_FAILURE(error_code);

                    MessageBox(mainWindow, L"Preview Captured", L"Preview Captured", MB_OK);
                    return S_OK;
                })
                .Get()));
    }
}

ExecuteScript

Execute JavaScript code from the javascript parameter in the current top level document rendered in the WebView.

public HRESULT ExecuteScript(LPCWSTR javaScript, ICoreWebView2ExecuteScriptCompletedHandler * handler)

This will execute asynchronously and when complete, if a handler is provided in the ExecuteScriptCompletedHandler parameter, its Invoke method will be called with the result of evaluating the provided JavaScript. The result value is a JSON encoded string. If the result is undefined, contains a reference cycle, or otherwise cannot be encoded into JSON, the JSON null value will be returned as the string 'null'. Note that a function that has no explicit return value returns undefined. If the executed script throws an unhandled exception, then the result is also 'null'. This method is applied asynchronously. If the method is called after NavigationStarting event during a navigation, the script will be executed in the new document when loading it, around the time ContentLoading is fired. ExecuteScript will work even if ICoreWebView2Settings::IsScriptEnabled is set to FALSE.

// Prompt the user for some script and then execute it.
void ScriptComponent::InjectScript()
{
    TextInputDialog dialog(
        m_appWindow->GetMainWindow(),
        L"Inject Script",
        L"Enter script code:",
        L"Enter the JavaScript code to run in the webview.",
        L"window.getComputedStyle(document.body).backgroundColor");
    if (dialog.confirmed)
    {
        m_webView->ExecuteScript(dialog.input.c_str(),
            Callback<ICoreWebView2ExecuteScriptCompletedHandler>(
                [](HRESULT error, PCWSTR result) -> HRESULT
        {
            if (error != S_OK) {
                ShowFailure(error, L"ExecuteScript failed");
            }
            MessageBox(nullptr, result, L"ExecuteScript Result", MB_OK);
            return S_OK;
        }).Get());
    }
}

get_BrowserProcessId

The process id of the browser process that hosts the WebView.

public HRESULT get_BrowserProcessId(UINT32 * value)

get_CanGoBack

Returns true if the WebView can navigate to a previous page in the navigation history.

public HRESULT get_CanGoBack(BOOL * canGoBack)

The HistoryChanged event will fire if CanGoBack changes value.

get_CanGoForward

Returns true if the WebView can navigate to a next page in the navigation history.

public HRESULT get_CanGoForward(BOOL * canGoForward)

The HistoryChanged event will fire if CanGoForward changes value.

get_ContainsFullScreenElement

Indicates if the WebView contains a fullscreen HTML element.

public HRESULT get_ContainsFullScreenElement(BOOL * containsFullScreenElement)

get_DocumentTitle

The title for the current top level document.

public HRESULT get_DocumentTitle(LPWSTR * title)

If the document has no explicit title or is otherwise empty, a default that may or may not match the URI of the document will be used.

get_Settings

The ICoreWebView2Settings object contains various modifiable settings for the running WebView.

public HRESULT get_Settings(ICoreWebView2Settings ** settings)

get_Source

The URI of the current top level document.

public HRESULT get_Source(LPWSTR * uri)

This value potentially changes as a part of the SourceChanged event firing for some cases such as navigating to a different site or fragment navigations. It will remain the same for other types of navigations such as page reloads or history.pushState with the same URL as the current page.

    // Register a handler for the SourceChanged event.
    // This handler will read the webview's source URI and update
    // the app's address bar.
    CHECK_FAILURE(m_webView->add_SourceChanged(
        Callback<ICoreWebView2SourceChangedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2SourceChangedEventArgs* args)
                -> HRESULT {
                wil::unique_cotaskmem_string uri;
                sender->get_Source(&uri);
                if (wcscmp(uri.get(), L"about:blank") == 0)
                {
                    uri = wil::make_cotaskmem_string(L"");
                }
                SetWindowText(GetAddressBar(), uri.get());

                return S_OK;
            })
            .Get(),
        &m_sourceChangedToken));

GetDevToolsProtocolEventReceiver

Get a DevTools Protocol event receiver that allows you to subscribe to a DevTools Protocol event.

public HRESULT GetDevToolsProtocolEventReceiver(LPCWSTR eventName, ICoreWebView2DevToolsProtocolEventReceiver ** receiver)

The eventName parameter is the full name of the event in the format {domain}.{event}. See the DevTools Protocol Viewer for a list of DevTools Protocol events description, and event args.

// Prompt the user to name a CDP event, and then subscribe to that event.
void ScriptComponent::SubscribeToCdpEvent()
{
    TextInputDialog dialog(
        m_appWindow->GetMainWindow(),
        L"Subscribe to CDP Event",
        L"CDP event name:",
        L"Enter the name of the CDP event to subscribe to.\r\n"
            L"You may also have to call the \"enable\" method of the\r\n"
            L"event's domain to receive events (for example \"Log.enable\").\r\n",
        L"Log.entryAdded");
    if (dialog.confirmed)
    {
        std::wstring eventName = dialog.input;
        wil::com_ptr<ICoreWebView2DevToolsProtocolEventReceiver> receiver;
        CHECK_FAILURE(
            m_webView->GetDevToolsProtocolEventReceiver(eventName.c_str(), &receiver));

        // If we are already subscribed to this event, unsubscribe first.
        auto preexistingToken = m_devToolsProtocolEventReceivedTokenMap.find(eventName);
        if (preexistingToken != m_devToolsProtocolEventReceivedTokenMap.end())
        {
            CHECK_FAILURE(receiver->remove_DevToolsProtocolEventReceived(
                preexistingToken->second));
        }

        CHECK_FAILURE(receiver->add_DevToolsProtocolEventReceived(
            Callback<ICoreWebView2DevToolsProtocolEventReceivedEventHandler>(
                [eventName](
                    ICoreWebView2* sender,
                    ICoreWebView2DevToolsProtocolEventReceivedEventArgs* args) -> HRESULT {
                    wil::unique_cotaskmem_string parameterObjectAsJson;
                    CHECK_FAILURE(args->get_ParameterObjectAsJson(&parameterObjectAsJson));
                    MessageBox(
                        nullptr, parameterObjectAsJson.get(),
                        (L"CDP Event Fired: " + eventName).c_str(), MB_OK);
                    return S_OK;
                })
                .Get(),
            &m_devToolsProtocolEventReceivedTokenMap[eventName]));
    }
}

GoBack

Navigates the WebView to the previous page in the navigation history.

public HRESULT GoBack()

GoForward

Navigates the WebView to the next page in the navigation history.

public HRESULT GoForward()

Cause a navigation of the top level document to the specified URI.

public HRESULT Navigate(LPCWSTR uri)

See the navigation events for more information. Note that this starts a navigation and the corresponding NavigationStarting event will fire sometime after this Navigate call completes.

void ControlComponent::NavigateToAddressBar()
{
    int length = GetWindowTextLength(GetAddressBar());
    std::wstring uri(length, 0);
    PWSTR buffer = const_cast<PWSTR>(uri.data());
    GetWindowText(GetAddressBar(), buffer, length + 1);

    HRESULT hr = m_webView->Navigate(uri.c_str());
    if (hr == E_INVALIDARG)
    {
        // An invalid URI was provided.
        if (uri.find(L' ') == std::wstring::npos
            && uri.find(L'.') != std::wstring::npos)
        {
            // If it contains a dot and no spaces, try tacking http:// on the front.
            hr = m_webView->Navigate((L"http://" + uri).c_str());
        }
        else
        {
            // Otherwise treat it as a web search. We aren't bothering to escape
            // URL syntax characters such as & and #
            hr = m_webView->Navigate((L"https://bing.com/search?q=" + uri).c_str());
        }
    }
    if (hr != E_INVALIDARG) {
        CHECK_FAILURE(hr);
    }
}

Initiates a navigation to htmlContent as source HTML of a new document.

public HRESULT NavigateToString(LPCWSTR htmlContent)

The htmlContent parameter may not be larger than 2 MB in total size. The origin of the new page will be about:blank.

            static const PCWSTR htmlContent =
              L"<h1>Domain Blocked</h1>"
              L"<p>You've attempted to navigate to a domain in the blocked "
              L"sites list. Press back to return to the previous page.</p>";
            CHECK_FAILURE(sender->NavigateToString(htmlContent));

OpenDevToolsWindow

Opens the DevTools window for the current document in the WebView.

public HRESULT OpenDevToolsWindow()

Does nothing if called when the DevTools window is already open.

PostWebMessageAsJson

Post the specified webMessage to the top level document in this WebView.

public HRESULT PostWebMessageAsJson(LPCWSTR webMessageAsJson)

The top level document's window.chrome.webview's message event fires. JavaScript in that document may subscribe and unsubscribe to the event via the following:

window.chrome.webview.addEventListener('message', handler)
window.chrome.webview.removeEventListener('message', handler)

The event args is an instance of MessageEvent. The ICoreWebView2Settings::IsWebMessageEnabled setting must be true or this method will fail with E_INVALIDARG. The event arg's data property is the webMessage string parameter parsed as a JSON string into a JavaScript object. The event arg's source property is a reference to the window.chrome.webview object. See add_WebMessageReceived for information on sending messages from the HTML document in the WebView to the host. This message is sent asynchronously. If a navigation occurs before the message is posted to the page, then the message will not be sent.

    // Setup the web message received event handler before navigating to
    // ensure we don't miss any messages.
    CHECK_FAILURE(m_webView->add_WebMessageReceived(
        Microsoft::WRL::Callback<ICoreWebView2WebMessageReceivedEventHandler>(
            [this](ICoreWebView2* sender, ICoreWebView2WebMessageReceivedEventArgs* args)
    {
        wil::unique_cotaskmem_string uri;
        CHECK_FAILURE(args->get_Source(&uri));

        // Always validate that the origin of the message is what you expect.
        if (uri.get() != m_sampleUri)
        {
            return S_OK;
        }
        wil::unique_cotaskmem_string messageRaw;
        CHECK_FAILURE(args->TryGetWebMessageAsString(&messageRaw));
        std::wstring message = messageRaw.get();

        if (message.compare(0, 13, L"SetTitleText ") == 0)
        {
            m_appWindow->SetTitleText(message.substr(13).c_str());
        }
        else if (message.compare(L"GetWindowBounds") == 0)
        {
            RECT bounds = m_appWindow->GetWindowBounds();
            std::wstring reply =
                L"{\"WindowBounds\":\"Left:" + std::to_wstring(bounds.left)
                + L"\\nTop:" + std::to_wstring(bounds.top)
                + L"\\nRight:" + std::to_wstring(bounds.right)
                + L"\\nBottom:" + std::to_wstring(bounds.bottom)
                + L"\"}";
            CHECK_FAILURE(sender->PostWebMessageAsJson(reply.c_str()));
        }
        return S_OK;
    }).Get(), &m_webMessageReceivedToken));

PostWebMessageAsString

This is a helper for posting a message that is a simple string rather than a JSON string representation of a JavaScript object.

public HRESULT PostWebMessageAsString(LPCWSTR webMessageAsString)

This behaves in exactly the same manner as PostWebMessageAsJson but the window.chrome.webview message event arg's data property will be a string with the same value as webMessageAsString. Use this instead of PostWebMessageAsJson if you want to communicate via simple strings rather than JSON objects.

Reload

Reload the current page.

public HRESULT Reload()

This is similar to navigating to the URI of current top level document including all navigation events firing and respecting any entries in the HTTP cache. But, the back/forward history will not be modified.

remove_ContainsFullScreenElementChanged

Remove an event handler previously added with add_ContainsFullScreenElementChanged.

public HRESULT remove_ContainsFullScreenElementChanged(EventRegistrationToken token)

remove_ContentLoading

Remove an event handler previously added with add_ContentLoading.

public HRESULT remove_ContentLoading(EventRegistrationToken token)

remove_DocumentTitleChanged

Remove an event handler previously added with add_DocumentTitleChanged.

public HRESULT remove_DocumentTitleChanged(EventRegistrationToken token)

remove_FrameNavigationCompleted

Remove an event handler previously added with add_FrameNavigationCompleted.

public HRESULT remove_FrameNavigationCompleted(EventRegistrationToken token)

remove_FrameNavigationStarting

Remove an event handler previously added with add_FrameNavigationStarting.

public HRESULT remove_FrameNavigationStarting(EventRegistrationToken token)

remove_HistoryChanged

Remove an event handler previously added with add_HistoryChanged.

public HRESULT remove_HistoryChanged(EventRegistrationToken token)

remove_NavigationCompleted

Remove an event handler previously added with add_NavigationCompleted.

public HRESULT remove_NavigationCompleted(EventRegistrationToken token)

remove_NavigationStarting

Remove an event handler previously added with add_NavigationStarting.

public HRESULT remove_NavigationStarting(EventRegistrationToken token)

remove_NewWindowRequested

Remove an event handler previously added with add_NewWindowRequested.

public HRESULT remove_NewWindowRequested(EventRegistrationToken token)

remove_PermissionRequested

Remove an event handler previously added with add_PermissionRequested.

public HRESULT remove_PermissionRequested(EventRegistrationToken token)

remove_ProcessFailed

Remove an event handler previously added with add_ProcessFailed.

public HRESULT remove_ProcessFailed(EventRegistrationToken token)

remove_ScriptDialogOpening

Remove an event handler previously added with add_ScriptDialogOpening.

public HRESULT remove_ScriptDialogOpening(EventRegistrationToken token)

remove_SourceChanged

Remove an event handler previously added with add_SourceChanged.

public HRESULT remove_SourceChanged(EventRegistrationToken token)

remove_WebMessageReceived

Remove an event handler previously added with add_WebMessageReceived.

public HRESULT remove_WebMessageReceived(EventRegistrationToken token)

remove_WebResourceRequested

Remove an event handler previously added with add_WebResourceRequested.

public HRESULT remove_WebResourceRequested(EventRegistrationToken token)

remove_WindowCloseRequested

Remove an event handler previously added with add_WindowCloseRequested.

public HRESULT remove_WindowCloseRequested(EventRegistrationToken token)

RemoveHostObjectFromScript

Remove the host object specified by the name so that it is no longer accessible from JavaScript code in the WebView.

public HRESULT RemoveHostObjectFromScript(LPCWSTR name)

While new access attempts will be denied, if the object is already obtained by JavaScript code in the WebView, the JavaScript code will continue to have access to that object. Calling this method for a name that is already removed or never added will fail.

RemoveScriptToExecuteOnDocumentCreated

Remove the corresponding JavaScript added using AddScriptToExecuteOnDocumentCreated with the specified script id.

public HRESULT RemoveScriptToExecuteOnDocumentCreated(LPCWSTR id)

RemoveWebResourceRequestedFilter

Removes a matching WebResource filter that was previously added for the WebResourceRequested event.

public HRESULT RemoveWebResourceRequestedFilter(LPCWSTR const uri, COREWEBVIEW2_WEB_RESOURCE_CONTEXT const resourceContext)

If the same filter was added multiple times, then it will need to be removed as many times as it was added for the removal to be effective. Returns E_INVALIDARG for a filter that was never added.

Stop

Stop all navigations and pending resource fetches.

public HRESULT Stop()

Does not stop scripts.

COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT

Image format used by the ICoreWebView2::CapturePreview method.

enum COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT

Values Descriptions
COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT_PNG PNG image format.
COREWEBVIEW2_CAPTURE_PREVIEW_IMAGE_FORMAT_JPEG JPEG image format.

COREWEBVIEW2_KEY_EVENT_KIND

The type of key event that triggered an AcceleratorKeyPressed event.

enum COREWEBVIEW2_KEY_EVENT_KIND

Values Descriptions
COREWEBVIEW2_KEY_EVENT_KIND_KEY_DOWN Correspond to window message WM_KEYDOWN.
COREWEBVIEW2_KEY_EVENT_KIND_KEY_UP Correspond to window message WM_KEYUP.
COREWEBVIEW2_KEY_EVENT_KIND_SYSTEM_KEY_DOWN Correspond to window message WM_SYSKEYDOWN.
COREWEBVIEW2_KEY_EVENT_KIND_SYSTEM_KEY_UP Correspond to window message WM_SYSKEYUP.

COREWEBVIEW2_MOVE_FOCUS_REASON

Reason for moving focus.

enum COREWEBVIEW2_MOVE_FOCUS_REASON

Values Descriptions
COREWEBVIEW2_MOVE_FOCUS_REASON_PROGRAMMATIC Code setting focus into WebView.
COREWEBVIEW2_MOVE_FOCUS_REASON_NEXT Moving focus due to Tab traversal forward.
COREWEBVIEW2_MOVE_FOCUS_REASON_PREVIOUS Moving focus due to Tab traversal backward.

COREWEBVIEW2_PERMISSION_KIND

The type of a permission request.

enum COREWEBVIEW2_PERMISSION_KIND

Values Descriptions
COREWEBVIEW2_PERMISSION_KIND_UNKNOWN_PERMISSION Unknown permission.
COREWEBVIEW2_PERMISSION_KIND_MICROPHONE Permission to capture audio.
COREWEBVIEW2_PERMISSION_KIND_CAMERA Permission to capture video.
COREWEBVIEW2_PERMISSION_KIND_GEOLOCATION Permission to access geolocation.
COREWEBVIEW2_PERMISSION_KIND_NOTIFICATIONS Permission to send web notifications.
COREWEBVIEW2_PERMISSION_KIND_OTHER_SENSORS Permission to access generic sensor.
COREWEBVIEW2_PERMISSION_KIND_CLIPBOARD_READ Permission to read system clipboard without a user gesture.

COREWEBVIEW2_PERMISSION_STATE

Response to a permission request.

enum COREWEBVIEW2_PERMISSION_STATE

Values Descriptions
COREWEBVIEW2_PERMISSION_STATE_DEFAULT Use default browser behavior, which normally prompt users for decision.
COREWEBVIEW2_PERMISSION_STATE_ALLOW Grant the permission request.
COREWEBVIEW2_PERMISSION_STATE_DENY Deny the permission request.

COREWEBVIEW2_PHYSICAL_KEY_STATUS

A structure representing the information packed into the LPARAM given to a Win32 key event.

typedef COREWEBVIEW2_PHYSICAL_KEY_STATUS

See the documentation for WM_KEYDOWN for details.

COREWEBVIEW2_PROCESS_FAILED_KIND

Kind of process failure used in the ICoreWebView2ProcessFailedEventHandler interface.

enum COREWEBVIEW2_PROCESS_FAILED_KIND

Values Descriptions
COREWEBVIEW2_PROCESS_FAILED_KIND_BROWSER_PROCESS_EXITED Indicates the browser process terminated unexpectedly.
COREWEBVIEW2_PROCESS_FAILED_KIND_RENDER_PROCESS_EXITED Indicates the render process terminated unexpectedly.
COREWEBVIEW2_PROCESS_FAILED_KIND_RENDER_PROCESS_UNRESPONSIVE Indicates the render process becomes unresponsive.

COREWEBVIEW2_SCRIPT_DIALOG_KIND

Kind of JavaScript dialog used in the ICoreWebView2ScriptDialogOpeningEventHandler interface.

enum COREWEBVIEW2_SCRIPT_DIALOG_KIND

Values Descriptions
COREWEBVIEW2_SCRIPT_DIALOG_KIND_ALERT A dialog invoked via the window.alert JavaScript function.
COREWEBVIEW2_SCRIPT_DIALOG_KIND_CONFIRM A dialog invoked via the window.confirm JavaScript function.
COREWEBVIEW2_SCRIPT_DIALOG_KIND_PROMPT A dialog invoked via the window.prompt JavaScript function.
COREWEBVIEW2_SCRIPT_DIALOG_KIND_BEFOREUNLOAD A dialog invoked via the beforeunload JavaScript event.

COREWEBVIEW2_WEB_ERROR_STATUS

Error status values for web navigations.

enum COREWEBVIEW2_WEB_ERROR_STATUS

Values Descriptions
COREWEBVIEW2_WEB_ERROR_STATUS_UNKNOWN An unknown error occurred.
COREWEBVIEW2_WEB_ERROR_STATUS_CERTIFICATE_COMMON_NAME_IS_INCORRECT The SSL certificate common name does not match the web address.
COREWEBVIEW2_WEB_ERROR_STATUS_CERTIFICATE_EXPIRED The SSL certificate has expired.
COREWEBVIEW2_WEB_ERROR_STATUS_CLIENT_CERTIFICATE_CONTAINS_ERRORS The SSL client certificate contains errors.
COREWEBVIEW2_WEB_ERROR_STATUS_CERTIFICATE_REVOKED The SSL certificate has been revoked.
COREWEBVIEW2_WEB_ERROR_STATUS_CERTIFICATE_IS_INVALID The SSL certificate is invalid – this could mean the certificate did not match the public key pins for the host name, the certificate is signed by an untrusted authority or using a weak sign algorithm, the certificate claimed DNS names violate name constraints, the certificate contains a weak key, the certificate's validity period is too long, lack of revocation information or revocation mechanism, non-unique host name, lack of certificate transparency information, or the certificate is chained to a legacy Symantec root.
COREWEBVIEW2_WEB_ERROR_STATUS_SERVER_UNREACHABLE The host is unreachable.
COREWEBVIEW2_WEB_ERROR_STATUS_TIMEOUT The connection has timed out.
COREWEBVIEW2_WEB_ERROR_STATUS_ERROR_HTTP_INVALID_SERVER_RESPONSE The server returned an invalid or unrecognized response.
COREWEBVIEW2_WEB_ERROR_STATUS_CONNECTION_ABORTED The connection was aborted.
COREWEBVIEW2_WEB_ERROR_STATUS_CONNECTION_RESET The connection was reset.
COREWEBVIEW2_WEB_ERROR_STATUS_DISCONNECTED The Internet connection has been lost.
COREWEBVIEW2_WEB_ERROR_STATUS_CANNOT_CONNECT Cannot connect to destination.
COREWEBVIEW2_WEB_ERROR_STATUS_HOST_NAME_NOT_RESOLVED Could not resolve provided host name.
COREWEBVIEW2_WEB_ERROR_STATUS_OPERATION_CANCELED The operation was canceled.
COREWEBVIEW2_WEB_ERROR_STATUS_REDIRECT_FAILED The request redirect failed.
COREWEBVIEW2_WEB_ERROR_STATUS_UNEXPECTED_ERROR An unexpected error occurred.

COREWEBVIEW2_WEB_RESOURCE_CONTEXT

Enum for web resource request contexts.

enum COREWEBVIEW2_WEB_RESOURCE_CONTEXT

Values Descriptions
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_ALL All resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_DOCUMENT Document resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_STYLESHEET CSS resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_IMAGE Image resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_MEDIA Other media resources such as videos.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_FONT Font resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_SCRIPT Script resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_XML_HTTP_REQUEST XML HTTP requests.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_FETCH Fetch API communication.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_TEXT_TRACK TextTrack resources.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_EVENT_SOURCE EventSource API communication.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_WEBSOCKET WebSocket API communication.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_MANIFEST Web App Manifests.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_SIGNED_EXCHANGE Signed HTTP Exchanges.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_PING Ping requests.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_CSP_VIOLATION_REPORT CSP Violation Reports.
COREWEBVIEW2_WEB_RESOURCE_CONTEXT_OTHER Other resources.