No subject

needs fixing on Windows - I unfortunately don't have a Windows machine
around (I really need to get one ...), but I think disabling it for Windows
is fine for now - as Stephen points out in a later email clang tooling for
Windows is not that useful yet due to clang restrictions. If somebody
expresses a strong need for the Windows fixes, I can prioritize them;
please let me know in that case.

Cheers,
/Manuel

--20cf302d4c5c6eafe404c1b5fbcc
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div class=3D"gmail_quote">On Mon, Jun 4, 2012 at 4:41 PM, Brad King <span =
dir=3D"ltr">&lt;<a href=3D"mailto:brad.king at kitware.com" target=3D"_blank">=
brad.king at kitware.com</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_q=
uote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1e=
x">
<div class=3D"im">On 06/04/2012 10:25 AM, Bill Hoffman wrote:<br>
&gt; On 6/1/2012 1:55 PM, Brad King wrote:<br>
&gt;&gt; The generator is supposed to produce a file that lists all the<br>
&gt;&gt; compiler invocation command lines in a JSON format. =A0The test<br=
>
&gt;&gt; runs a build with the option and then runs a program that reads<br=
>
&gt;&gt; the generated file and re-plays the commands. =A0The test and its<=
br>
&gt;&gt; C++ program should not need modification because it works for the<=
br>
&gt;&gt; Makefile generator. =A0The Ninja generator will need to generate<b=
r>
&gt;&gt; the properly encoded commands like the Makefile generator does.<br=
>
&gt;<br>
&gt; Seems like the json file is reasonable:<br>
&gt;<br>
</div><div class=3D"im">&gt; =A0 =A0&quot;command&quot;: &quot;C:\\PROGRA~2=
\\MICROS~1.0\\VC\\bin\\cl.exe =A0 /nologo<br>
&gt; /DWIN32 /D_WINDOWS /W3 /Zm1000 /EHsc /GR /D_DEBUG /MDd /Zi /Ob0 /Od<br=
>
&gt; /RTC1 -I\&quot;c:\\Users\\hoffman\\Work\\My<br>
&gt; Builds\\NinjaCMake\\Tests\\CompileCommandOutput\\..\\..\\Source\&quot;=
<br>
&gt; /TP<br>
&gt; /FoCMakeFiles/CompileCommandOutput.dir/compile_command_output.cxx.obj<=
br>
&gt; /FdTARGET_PDB -c \&quot;c:\\Users\\hoffman\\Work\\My<br>
&gt; Builds\\NinjaCMake\\Tests\\CompileCommandOutput\\compile_command_outpu=
t.cxx\&quot;&quot;,<br>
&gt; =A0 =A0&quot;file&quot;: &quot;c:\\Users\\hoffman\\Work\\My<br>
&gt; Builds\\NinjaCMake\\Tests\\CompileCommandOutput\\compile_command_outpu=
t.cxx&quot;<br>
&gt; },<br>
&gt;<br>
</div><div class=3D"im">&gt; the test driver should be able to read a file =
with \\ in it,<br>
&gt; and the generator output is fine.<br>
<br>
</div>The topic was first brought up here:<br>
<br>
=A0<a href=3D"http://thread.gmane.org/gmane.comp.programming.tools.cmake.de=
vel/3678" target=3D"_blank">http://thread.gmane.org/gmane.comp.programming.=
tools.cmake.devel/3678</a><br>
<br>
Context of this discussion:<br>
<br>
=A0<a href=3D"http://thread.gmane.org/gmane.comp.programming.tools.cmake.de=
vel/3774" target=3D"_blank">http://thread.gmane.org/gmane.comp.programming.=
tools.cmake.devel/3774</a><br>
<br>
Origin of parser in question:<br>
<br>
=A0<a href=3D"http://cmake.org/gitweb?p=3Dcmake.git;a=3Dcommitdiff;h=3D0e6b=
05fc" target=3D"_blank">http://cmake.org/gitweb?p=3Dcmake.git;a=3Dcommitdif=
f;h=3D0e6b05fc</a><br>
<br>
Manuel, as author of this parser can you comment please?</blockquote><div><=
br></div><div>From taking a very high level look it seems to me like the pa=
rser probably needs fixing on Windows - I unfortunately don&#39;t have a Wi=
ndows machine around (I really need to get one ...), but I think disabling =
it for Windows is fine for now - as Stephen points out in a later email cla=
ng tooling for Windows is not that useful yet due to clang restrictions. If=
 somebody expresses a strong need for the Windows fixes, I can prioritize t=
hem; please let me know in that case.</div>
<div><br></div><div>Cheers,</div><div>/Manuel</div></div>

--20cf302d4c5c6eafe404c1b5fbcc--