@m-kuhn and @dg0yt would be thankful if one of you could confirm this is an issue. At least it fixed the problem for me.

I can't confirm this is an issue (I was able to build using emscripten recently, I'm not sure what I'd need to change to trigger this behavior), but the \n in the quotes of the original version looks very suspicious.

This is what we get iiuc, but it's unlikely to be what we want:

set(__qt_chainload_toolchain_file "${VCPKG_CHAINLOAD_TOOLCHAIN_FILE}
")

set(__qt_chainload_toolchain_file "${VCPKG_CHAINLOAD_TOOLCHAIN_FILE}
")

This is also what I see for x64-linux native. Ugly, but not broken. Emscripten should be no different.

Thanks for the feedback

@dg0yt just tested again with emsdk 4.0.22 / CMake 3.30.4:

Error: CMake Error at build/vcpkg/packages/qtbase_wasm32-emscripten-qt-threads/share/Qt6/qt.toolchain.cmake:36:
  Parse error.  Expected a newline, got identifier with text "endif".
Call Stack (most recent call first):
  CMakeLists.txt:51 (include)

So the fix is definitely needed...

But I can confirm that your suggestion works and will just accept it.

Ugly, but not broken. Emscripten should be no different.

🤦 Of course broken, with a line break in a real filepath, as soon as __qt_chainload_toolchain_file is actually used.

@microsoft-github-policy-service agree company="OPENGIS.ch"

Tested again with ffc071e

I'm seeing there are two PRs that are touching qtbase -> #49191. I'm asking the contributor if they would like to absorb this change.