[cfe-commits] [PATCH] Simplifying lit script for remove-cstr-calls

[Jordan Rose]

On Oct 31, 2012, at 11:17 , Matthieu Monrocq <matthieu.monrocq at gmail.com> wrote:

> 
> 
> On Wed, Oct 31, 2012 at 4:28 PM, Edwin Vane <edwin.vane at intel.com> wrote:
> Hi chandlerc,
> 
> Following the example from loop-convert tests, simplifying the remove-cstr-calls lit script.
> 
> http://llvm-reviews.chandlerc.com/D92
> 
> Files:
>   test/remove-cstr-calls/basic.cpp
> 
> Index: test/remove-cstr-calls/basic.cpp
> ===================================================================
> --- test/remove-cstr-calls/basic.cpp
> +++ test/remove-cstr-calls/basic.cpp
> @@ -1,12 +1,7 @@
> -// RUN: rm -rf %t
> -// RUN: mkdir %t
> -// RUN: echo "[{\"directory\":\".\",\"command\":\"clang++ -c %t/test.cpp\",\"file\":\"%t/test.cpp\"}]" | sed -e 's/\\/\//g' > %t/compile_commands.json
> -// RUN: cp "%s" "%t/test.cpp"
> -// RUN: remove-cstr-calls "%t" "%t/test.cpp"
> -// RUN: cat "%t/test.cpp" | FileCheck %s
> +// RUN: grep -Ev "// *[A-Z-]+:" %s > %t.cpp
> +// RUN: remove-cstr-calls . %t.cpp --
> +// RUN: FileCheck -input-file=%t.cpp %s
> 
> FileCheck can accept arguments on the standard input (and thus it's possible to pipe into it)
> 
> Would it be simpler to have remove-cstr-calls emit output on the standard output and just pipe it into FileCheck ?

I have seen this done (and done it myself) but then you get the disadvantage of not checking the error code of remove-cstr-calls. We've hidden some crashes this way that occurred after all the checked lines have been printed.

Jordan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20121031/c064f3de/attachment.html>