IA2 dump tree implementation #75
Conversation
spectranaut
changed the title
spectranaut
force-pushed
the
iaccessible2
branch
2 times, most recently
from
2c9b723 to
f5adbb0
Compare
spectranaut
force-pushed
the
iaccessible2
branch
from
f5adbb0 to
15d3751
Compare
spectranaut
changed the title
| // We will need these for windows related things, see the formatter_win.cc in | ||
| // chromium | ||
| // #include <oleacc.h> | ||
| // #include <wrl/client.h> | ||
|
|
||
| // TODO: we will probably need to get these from the IAccessible2 git repo, like | ||
| // in Chromium | ||
| // #include "third_party/iaccessible2/ia2_api_all.h" |
There was a problem hiding this comment.
Again maybe not in this PR, but it might be nice to have a similar third_party convention.
There was a problem hiding this comment.
true, it seems weird to just have these generated files here. made an issue: #96
| HWND hCurWnd = nullptr; | ||
| // TODO: This is not the right API -- probably we should copy chrome code | ||
| // and find a hwnd by titla alone. Right now this only returns a Google Chrome | ||
| // HWND. But if I don't include PID, then I get other processess with google |
There was a problem hiding this comment.
Why do we need to filter on name here at all? Do we get multiple processes from the same ProcessID?
There was a problem hiding this comment.
We get multiple windows for a single process id. And when I got all the windows (not scoped to a single process id) some of them also had "Google Chrome" in the in the name, I think it was the Google Chrome installer. But this will be handled in a follow up PR I think?
There was a problem hiding this comment.
Totally fine, was just trying to understand what the underlying issue even is.
|
|
||
| Microsoft::WRL::ComPtr<IServiceProvider> service_provider; | ||
| HRESULT hr = root_->QueryInterface(IID_PPV_ARGS(&service_provider)); | ||
| hr = service_provider->QueryService(IID_IAccessible, IID_PPV_ARGS(&ia2)); |
There was a problem hiding this comment.
perhaps it should be moved to an utility method
There was a problem hiding this comment.
Yeah I agree this should be done, but I was thinking we'd do it in a follow up with #94 94 -- I added TODO note here!
In this PR:
ia2_api_all.handia2_api_all_i.ccthat was created by running concat.sh in LinuxA11y/IAccessible2 then MIDLbuild/bin/Release/dump_tree_ia2.exeFollow up tasks:
CoInitialize(nullptr);, because you have to callCoUninitialize()when there are no longer any interface pointers on the stack (otherwise you get a seg fault). you can call multiple times -- as long as its the same amount of times. possibly we can to call every time we create or destroy a "IA2Node" ?