automake-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH] New tests on obsoleted usages of automake/autoconf macros.


From: Stefano Lattarini
Subject: Re: [PATCH] New tests on obsoleted usages of automake/autoconf macros.
Date: Fri, 12 Nov 2010 20:57:05 +0100
User-agent: KMail/1.13.3 (Linux/2.6.30-2-686; KDE/4.4.4; i686; ; )

On Friday 12 November 2010, Ralf Wildenhues wrote:
> * Stefano Lattarini wrote on Fri, Nov 12, 2010 at 08:37:14PM CET:
> > On Friday 12 November 2010, Ralf Wildenhues wrote:
> > > The easiest fix would be to grep for the lines in question, I guess.
> > I agree.  What about the attached patch?  (It also makes the testcase
> > simpler and shorter, BTW).
> 
> Leave out the $ in the grep patterns, in case grep assumes LF only.
Done adding a proper comment.

> OK with that change.  But you could have left the other comparisons
> intact; however you like.
Nah, the new way is better IMO.  I've just added a `!' to the `foo' output
to act as a poor-man end-of-line indicator.
 
> Thanks,
> Ralf
> 
> > Subject: [PATCH] Fix bug in test `backcompat6.test' (MSYS portability).
> > 
> > * tests/backcompat6.test (Makefile.am): Grep the output from the
> > test program, rather than comparing it with a file generated
> > on-the-fly through the `echo' builtin.  This prevents a spurious
> > failure on MinGW/MSYS due to LF vs. CRLF line endings.
> 
> How about
> 
> * tests/backcompat6.test (Makefile.am): Grep output, to avoid newline
> encoding differences on MSYS.
How about

  * tests/backcompat6.test (Makefile.am): Grep the output from the
    test program, rather than diffing it, to avoid spurious failures
    on MinGW/MSYS due to LF vs. CRLF line endings.

-*-*-

This is what I squashed in; I will merge to master and push later (or
tomorrow) if there are no further objections.

diff --git a/ChangeLog b/ChangeLog
index 57dc968..bf2e617 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -2,9 +2,8 @@
 
        Fix bug in test `backcompat6.test' (MSYS portability).
        * tests/backcompat6.test (Makefile.am): Grep the output from the
-       test program, rather than comparing it with a file generated
-       on-the-fly through the `echo' builtin.  This prevents a spurious
-       failure on MinGW/MSYS due to LF vs. CRLF line endings.
+       test program, rather than diffing it, to avoid spurious failures
+       on MinGW/MSYS due to LF vs. CRLF line endings.
        Reported by Ralf Wildenhues.
 
 2010-11-09  Stefano Lattarini  <address@hidden>
diff --git a/tests/backcompat6.test b/tests/backcompat6.test
index ca7412b..642ffaa 100755
--- a/tests/backcompat6.test
+++ b/tests/backcompat6.test
@@ -57,8 +57,11 @@ check-local:
        test x'$(PACKAGE)' = x'nonesuch-zardoz'
        test x'$(VERSION)' = x'nonesuch-0.1'
        ./foo # for debugging
-       ./foo | grep '^PACKAGE = nonesuch-zardoz$$'
-       ./foo | echo '^VERSION = nonesuch-0\.1$$'
+        ## Do not anchor the regexps w.r.t. the end of line, because on
+        ## MinGW/MSYS, grep may assume LF line endings only, while our
+        ## `foo' program may generate CRLF line endings.
+       ./foo | grep '^PACKAGE = nonesuch-zardoz!'
+       ./foo | echo '^VERSION = nonesuch-0\.1!'
 END
 
 cat > quux.c <<'END'
@@ -66,7 +69,7 @@ cat > quux.c <<'END'
 #include <stdio.h>
 int main (void)
 {
-  printf("PACKAGE = %s\nVERSION = %s\n", PACKAGE, VERSION);
+  printf("PACKAGE = %s!\nVERSION = %s!\n", PACKAGE, VERSION);
   return 0;
 }
 END





reply via email to

[Prev in Thread] Current Thread [Next in Thread]