Skip to content

Optimize to_string #1615

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Yexuan Xiao (YexuanXiao) wants to merge 8 commits into
base: master
Choose a base branch
from
+10 −0
Open

Optimize to_string #1615

Changes from 6 commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
d9c702c
Optimize `to_string`
YexuanXiao Aug 3, 2026
a81cb51
Replace tab with space
YexuanXiao Aug 3, 2026
daa9e9f
Improve buffer size verification
YexuanXiao Aug 3, 2026
d790563
Fix WINRT_VERIFY_ check for bytes_written
YexuanXiao Aug 3, 2026
4120584
Update base_string.h
YexuanXiao Aug 7, 2026
fa9d953
Update base_string.h
YexuanXiao Aug 7, 2026
f7e67bf
Update base_string.h
YexuanXiao Aug 20, 2026
33fbb32
Merge branch 'master' into patch-7
YexuanXiao Aug 20, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions strings/base_string.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -719,8 +719,18 @@ WINRT_EXPORT namespace winrt
return{};
}

#if defined(__cpp_lib_string_resize_and_overwrite) && __cpp_lib_string_resize_and_overwrite >= 202110L
std::string result;
result.resize_and_overwrite(size, [&](char* buffer, std::size_t) -> std::size_t

@DefaultRyan Ryan Shepherd (DefaultRyan) Aug 19, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nit: Feels slightly dirty to capture size instead of using the lambda's param. Perhaps this is simplified if we are selective about our lambda captures and capture only &value.

{
auto bytes_written = WINRT_IMPL_WideCharToMultiByte(65001 /*CP_UTF8*/, 0, value.data(), static_cast<std::int32_t>(value.size()), buffer, size, nullptr, nullptr);
WINRT_VERIFY_(size, bytes_written);
return bytes_written == size ? size : 0;

@DefaultRyan Ryan Shepherd (DefaultRyan) Aug 19, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nit: This slightly changes the specified behavior from the original. Now we truncate to zero if written differs from size, whereas before we would simply accept the result.

In practice, assuming WideCharToMultiByte behaves sanely, written should always equal the value of size (obtained from the first call to WCTMB). I think we are committing to assuming this sane behavior, so let's keep the defensive guardrails consistent, and go ahead and return bytes_written directly. The WINRT_VERIFY_ sanity check is fine to keep as is to warn if something weird happened.

});
Comment thread
YexuanXiao marked this conversation as resolved.
#else
std::string result(size, '?');
WINRT_VERIFY_(size, WINRT_IMPL_WideCharToMultiByte(65001 /*CP_UTF8*/, 0, value.data(), static_cast<std::int32_t>(value.size()), result.data(), size, nullptr, nullptr));
#endif
return result;
}
}