[Scons-users] Scons-users Digest, Vol 165, Issue 8

Sat Jun 7 02:49:39 EDT 2025

Honestly.. IMHO using #define HEADER "xyz.h" and then #include HEADER is
not great practice.

        Yes, I agree with you, that is not a good practice. But the code is generated by tool, and many marcos and places. Haha God. the benefit is that logic and configuration can be seprected. so tools like to use it.

         does ninja have a balance proformance?.
         And maybe I could generate .i firstly, and then generate .o. is it a good ideal?  Haha.

---- Replied Message ----
| From | scons-users-request at scons.org |
| Date | 06/07/2025 13:35 |
| To | scons-users at scons.org |
| Cc | |
| Subject | Scons-users Digest, Vol 165, Issue 8 |
Send Scons-users mailing list submissions to
     scons-users at scons.org

To subscribe or unsubscribe via the World Wide Web, visit
     https://pairlist4.pair.net/mailman/listinfo/scons-users
or, via email, send a message with subject or body 'help' to
     scons-users-request at scons.org

You can reach the person managing the list at
     scons-users-owner at scons.org

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Scons-users digest..."

Today's Topics:

  1. Re: update dependency and signatures.dblite after biilding
     (Bill Deegan)
  2. Re: Testing ninja build generator (Bill Deegan)

----------------------------------------------------------------------

Message: 1
Date: Fri, 6 Jun 2025 11:14:00 -0700
From: Bill Deegan <bill at baddogconsulting.com>
To: SCons users mailing list <scons-users at scons.org>
Subject: Re: [Scons-users] update dependency and signatures.dblite
     after biilding
Message-ID:
     <CAEyG4CGrXp77zp9hwCYOkL8j6TSeKWB3bXXeWArTCu+0ad8Oow at mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

Of course a better scanner is slower.. as would generating the .d files
from the compiler..

That said, the scanner I pointed you at may have room for improvement in
terms of performance.

Honestly.. IMHO using #define HEADER "xyz.h" and then #include HEADER is
not great practice.

For your code how many instances of a header file defined by a macro are
there?
Is it one, which is used all over the place?
Or many, used many places?

On Fri, Jun 6, 2025 at 10:20?AM ? <yjtyzw at 126.com> wrote:

>
>  Well, this is why Bill advised trying the C Conditional Scanner, which
> *does* know how to handle preprocessor instructions of this sort.  It's
> still a bit of a work in progress (there are some questions about
> recursion, in particular), which is why it's not the default.  Were you
> able to try that out?
>
> I have try it, as Bill and Mats said, CConditionalScanner can resolve my
> problem.
> And I find that after set CConditionalScanner, the building time change
> from 5min to 8min.
> ---- Replied Message ----
> From Mats Wichmann<mats at wichmann.us> <undefined>
> Date 6/6/2025 23:54
> To ?<yjtyzw at 126.com> <yjtyzw at 126.com>
> Cc SCons users mailing list<scons-users at scons.org> <scons-users at scons.org>
> Subject Re: [Scons-users] update dependency and signatures.dblite after
> biilding
> On 6/6/25 09:46, ? wrote:
>
>
> ??My problem:
> ?
>
> In my project, there are those code in some source files,
>
> #defineFOO_HEADER "foo.h"
> #includeFOO_HEADER
>
> ??
>
> and the files are independent module, I can't modify them.
>
> the scons-user.pdf said, scons the built-in C scanner is unable to
> extract the implicit dependency on a header file.
> So I create .d file for every source file, and parse dependencies
> from .d, and then add dependency by Depends function.
>
> ??And then, there is a new problem, when there are no .d, (for
> example the first building or dependencies are changed), I need to
> build project with two times, and then I can push all the
> dependencies into signature.
>
>
> yes, this isn't ideal - if the dependencies can't be automatically
> determined, you have both a first-run problem and no detection of changes.
>
>
> ??
> So I consider that whether could I update all the dependencies into
> signature with just one building.
> that is my problem.
>
> Well, this is why Bill advised trying the C Conditional Scanner, which
> *does* know how to handle preprocessor instructions of this sort.  It's
> still a bit of a work in progress (there are some questions about
> recursion, in particular), which is why it's not the default.  Were you
> able to try that out?
>
> _______________________________________________
> Scons-users mailing list
> Scons-users at scons.org
> https://pairlist4.pair.net/mailman/listinfo/scons-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20250606/18590e02/attachment-0001.htm>

------------------------------

Message: 2
Date: Fri, 6 Jun 2025 11:26:28 -0700
From: Bill Deegan <bill at baddogconsulting.com>
To: SCons users mailing list <scons-users at scons.org>
Subject: Re: [Scons-users] Testing ninja build generator
Message-ID:
     <CAEyG4CFazPZvuNMSDqDy9QCE7wbTTXNVyUEJuL7145Fi8Ku2dw at mail.gmail.com>
Content-Type: text/plain; charset="utf-8"

There is some generic filename escaping logic in scons which should likely
be used there.

env['ESCAPE'] should have the current platform's escaping function.

On Fri, Jun 6, 2025 at 9:55?AM Mats Wichmann <mats at wichmann.us> wrote:

> On 6/6/25 10:42, Julien Pommier wrote:
> >
> >
> >> There's another layer of "quoting problem" here, beyond the URL-quoting
> that I'm trying to understand.
> >>
> >> As written, the rule in build.ninja looks like:
> >>
> >> build run$ me.sh: TEMPLATE | .ninja/scons_daemon_dirty
> >>
> >> I'm assuming the '$' is some kind of space-escape syntax for ninja.
> This fails with
> >>
> >> scons: *** Do not know how to make File target `run'
> >>
> >> so we ended up seeing the target back in SCons as being just "run".
> Adding a layer of quotes doesn't work, as those are carefully preserved and
> the target names don't match.
> >>
> >> Not familiar with this code beyond the fiddling I did in Utils.py,
> hopefully someone else has some ideas, else we'll have to dig into it.
> >
> > It seems the quoting need to be applied to the line 234 in
> ninja_scons_daemon.py. Using:
> >
> >                      input_command = "build \"" + building_node + "\" "
> >
> > instead of
> >                     input_command = "build " + building_node + " "
> >
> > makes it work for my test-case
>
> yeah, that looks good.
>
>
> _______________________________________________
> Scons-users mailing list
> Scons-users at scons.org
> https://pairlist4.pair.net/mailman/listinfo/scons-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20250606/69b37031/attachment-0001.htm>

------------------------------

Subject: Digest Footer

_______________________________________________
Scons-users mailing list
Scons-users at scons.org
https://pairlist4.pair.net/mailman/listinfo/scons-users

------------------------------

End of Scons-users Digest, Vol 165, Issue 8
*******************************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20250607/ae714f5a/attachment.htm>