I have a NPAPI plugin running in Safari, Chrome and Firefox. I'm able to handle all key combinations with the event callback, but the problem I'm having is with certain key combinations.
In the plugin, I'm trying to use Command+O to fire the plugin's file open dialog, but the browser is also firing its open file dialog.
Supposedly, returning TRUE for event handling method should report the browser that the plugin handled the event, but I'm having no luck with that.
You neglected to specify which OS you're on; based on the fact that you said "Command+O" instead of "Ctrl-O" I'm assuming you're on Mac. The bad news is that on Mac, all events are passed to the plugin from the browser. If the browser doesn't choose to make it possible to override the handling of an event (and it sounds like this is the case) then there is probably nothing you can do about it, short of possibly using some OS hook to intercept the key event before the browser gets it, which seems risky.
On windows you might have a little more luck, since a windowed plugin should actually get events first before they are passed to the parent window and thus could intercept them. On Mac, though, it's all windowless and you get events when the browser feels like giving them to you.
Related
In any web browser and in the Windows file manager and in many other applications there is support for forward and backward navigation. This always (or at least most of the time) by default works with extra mouse buttons if your mouse has any.
I want to implement this in a C++ application I'm making based on WinAPI. However I wonder how one would do this? Are the mouse buttons captured "manually" in each application that has this forward/backward navigation or is there native support for it somewhere in WinAPI?
Manually capturing the buttons is probably always an option, but if there already is an existing functionality that handles this then it seems like that should be used instead. That's probably more reliable as well.
To sum up: I want my application to correctly handle/receive backward and forward clicks from a mouse that has such buttons.
The WM_APPCOMMAND message offers the APPCOMMAND_BROWSER_FORWARD ("Navigate forward") and APPCOMMAND_BROWSER_BACKWARD ("Navigate backward") navigation commands. You can handle them in your application, even if it's not a browser.
The documentation has information, how and when the WM_APPCOMMAND is generated:
DefWindowProc generates the WM_APPCOMMAND message when it processes the WM_XBUTTONUP or WM_NCXBUTTONUP message, or when the user types an application command key.
I'm writing a plugin for a big x64 application in C++. I want the plugin to open a dialog and show a web view of my site.
I'm been able to use WKWebView in macOS and it works well. On Windows I'm evaluating CEF https://bitbucket.org/chromiumembedded/cef (please let me know of any alternative, ideally I would like it to be Webkit-based).
Let's say the application framework that is hosting my plugin has already created a window for my plugin and has it's own message loop, so I can only receive events in a sort of WindowProc. I can also get the HWND of the window.
My implementation is inspired by cefsimple example, because cefclient is way too complicated. I've implemented the subprocess architecture with the external executable and everything works fine until it's rendering the client area of the window. Then I have problems with closing the window (it crashes) and resizing the window interactively (the window frame is resized but the web view in the client area does not resize).
I've tried all possible combinations, but I've run out of ideas. Namely:
If I use CefRunMessageLoop() the web view is rendered correctly but the main application does not process the UI events like close window button. Resize does not work.
If I call CefDoMessageLoopWork() myself once in a while (from WindowProc) the web view is rendered correctly and it processes the close button, but it crashes. Resize does not work.
If I use settings.multi_threaded_message_loop = true the web view is rendered correctly and I can close the window without crash. The destructor of the window calls CefShutdown(). But if I try to reopen the window it crashes! Are CefInitialize and CefShutdown allowed to be called only once?
And resizing still does not work. I don't understand why in the cefsimple example resizing works and in my window it does not work.
Besides message processing issues, probably I'm not closing the browser correctly, any advice? Why is so complicated? WKWebView is so straighforward!
There is no error message, no stack trace, no source code, no OS/CEF version - I doubt this question can be answered.
I can only tell you how to close browser cleanly: call CefShutdown at the right time (see cefclient/cefsimple examples) and do not keep any references to CEF objects when calling shutdown.
I'm trying to use SendMessage to post mouse clicks to a background window (Chrome), which works fine, but brings the window to front after every click. Is there any way to avoid that?
Before anyone says this is a duplicate question, please make sure that the other topic actually mentions not activating the target window, because I couldn't find any.
Update: aha, hiding the window does the trick, almost. It receives simulated mouse/keyboard events as intended, and doesn't show up on screen. However, I can just barely use my own mouse to navigate around the computer, and keyboard input is completely disrupted.
So my question is, how does sending messages to a window affect other applications? Since I'm not actually simulating mouse/keyboard events, shouldn't the other windows be completely oblivious to this?
Is it possibly related to the window calling SetCapture when it receives WM_LBUTTONDOWN? And how would I avoid that, other than hooking the API call (which would be very, very ugly for such a small task)?
The default handling provided by the system (via DefWindowProc) causes windows to come to the front (when clicked on) as a response to the WM_MOUSEACTIVATE message, not WM_LBUTTONDOWN.
The fact that Chrome comes to the front in response to WM_LBUTTONDOWN suggests that it's something Chrome is specifically doing, rather than default system behaviour that you might be able to prevent in some way.
The source code to Chrome is available; I suggest you have a look at it and see if it is indeed something Chrome is doing itself. If so, the only practical way you would be able to prevent it (short of compiling your own version of Chrome) is to inject code into Chrome's process and sub-class its main window procedure.
I've got a Windows 7 VM which runs my mandatory corporate communication systems (Lync and Outlook). What I'd like to do is run a process on this Windows machine which monitors the system tray and sends notifications to my host machine (Xubuntu 13.04) so I'm informed when I get an email or IM (I've already tried seamless RDP to do this but it's an ineffective solution).
Anything Linux or network oriented I can handle with relative ease, what I do not know how to do is to how to query the state of the Windows system tray (or attach an event listener for state changes). I'm comfortable with C++ and Python but I'll give any viable solution a go.
Detailed state information would be preferable but at the very minimum I need to be able to detect changes in the number of icons in the tray.
On Windows, if you install Visual Studio, among Visual Studio Tools there is a useful tool, named "Spy++", basically it's a tool that shows you all the windows and gives ability to see what messages particular window receives.
Using this tool, you can see that whole panel, that contains "Start button", shortcuts, tray, clock, etc. is "Shell_TrayWnd". You can use "Find" menu to search for a particular window just dragging an "aim" on any window.
The tray window itself is "SysPager" (000100D2 on attached image), you can log messages for this window and see what type of message this window receives when you receive mail in Outlook.
After that you can write a code that will listen to all messages that this window receives, and basing on what you have seen in "Spy++" determine what happened.
This is just for start. Writing a code that will get a window handle and listen to messages that window receives is another part, but I think it's already covered at MSDN or even at StackOverflow.
Searching for a particular window handle is done by FindWindowEx function and in order to listen to message you have to set a "hook" that is done by SetWindowsHookEx function. Hooking is described pretty good on MSDN.
may be you need Outlook inspector (http://msdn.microsoft.com/en-us/library/office/ff869356(v=office.15).aspx). As i understand it helps to watch events outlook processes.
I need to add ability to catch event each time there is mouse hover event both on Linux and Windows. In Linux I easily connect to hovering-over-link signal. I cannot seem to find Windows equivalent. Searched google, checked the MSDN, I could not find.
Does Such a functionality available in windows? If no is there a way to achieve it?
Can you use the StatusTextChange event? This event is fired when the mouse is hovered over a link, with sText set to the link URL.
It is also fired with additional messages such as when loading a page. In these cases, sText is set to "Connecting to (URL)", "Waiting for (URL)", etc. These messages can be simply ignored by doing a basic URL validation of sText.
A more involved method would be in the DocumentComplete event to search for all anchor elements using getElementsByTagName and attach to the HTMLAnchorEvents onmouseover. This of course wouldn't work for anchor elements dynamically created after page load.