ClangBuildAnalyzer VS include-what-you-use

> On another note: C++ compiler should by default keep statistics about the chain of #include's / parsing during compilation and dump it to a file at the end and also summarize how badly you're re-parsing the same .h files during build.

Not exactly that, but do you know clang's -ftime-trace and tools like https://github.com/aras-p/ClangBuildAnalyzer which help analyzing where time is actually spent? (In small repeated headers I don't see much of a problem, but they of course may contain not so small things ...)

You can also use the following when you want to inspect multiple files: https://github.com/aras-p/ClangBuildAnalyzer

ClangBuildAnalyzer reports on parsing, build, and link time, whereas IncludeGuardian only reports on parsing time.

https://github.com/aras-p/ClangBuildAnalyzer is a very useful tool to quantify the cost of different headers (and other costly parts of the compile such as template instantiations). It doesn’t help with actually fixing such problems, but it’s a pretty good ruler to measure where the time is spent.

If you can compile with Clang, I suggest you to try ClangBuildAnalyzer

Interesting...how does it compare to https://github.com/include-what-you-use/include-what-you-use ?

Build Insights in Visual Studio, include-what-you-use).

Looks like https://include-what-you-use.org/ might do that.

If you are worried about includes, use https://github.com/include-what-you-use/include-what-you-use and stop thinking about it.

Getting rid of that is not straightforard, though some tools can help with that

You can use the tool https://github.com/include-what-you-use/include-what-you-use to do this for for. It tracks included files and can give comment for what is used from each file. It also warns you when you include files that you don’t use

Invisible imports (e.g. traits). In Python, everything is fully namespaced (unless you from import * in which case all bets are off). It's always explicit where a name is coming from. C is the opposite: #include lets you refer to anything defined in the headers with no namespacing. That's why a common strategy (include what you use) has an associated code style: after every non-std #include you have a comment saying which of its definitions you are using. Of course, Rust is much less implicit, but I still sometimes struggle with traits. For example, you can use tokio::net::TcpStream, but you need to also use tokio::io::AsyncReadExt for the .read trait to be defined on TcpStream. This makes it hard (for me) to answer questions like "what traits are currently available in this scope?" and "why is this module being imported?"

Aside from being closed source and not available on all architectures, how does it compare to iwyu(https://include-what-you-use.org/) or clang's relatively recent include-fixer which is also accessible via clangd?

I set up include-what-you-use and I let it tell me which headers should be where. The IWYU rules would have put all needed headers including in the cpp file.