-----Original Message-----
From: Robert Dubner <rdub...@symas.com>
Sent: Saturday, March 15, 2025 22:15
To: Jakub Jelinek <ja...@redhat.com>
Subject: RE: [PATCH]cobol: create new
gcc/testsuite/cobol.dg/group1/check_88.cob test
> -----Original Message-----
> From: Jakub Jelinek <ja...@redhat.com>
> Sent: Saturday, March 15, 2025 20:57
> To: Robert Dubner <rdub...@symas.com>; gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH]cobol: create new
> gcc/testsuite/cobol.dg/group1/check_88.cob test
>
> Hi!
>
> Sorry, my mail got somehow truncated.
>
> On Sun, Mar 16, 2025 at 01:42:52AM +0100, Jakub Jelinek wrote:
> > On Sat, Mar 15, 2025 at 04:14:48PM -0500, Robert Dubner wrote:
> > > --- /dev/null
> > > +++ b/gcc/testsuite/cobol.dg/group1/check_88.cob
> > > @@ -0,0 +1,116 @@
> > > +*> { dg-do run }
> > > +*> { dg-output "-><-\n" }
> > > +*> { dg-output "-> <-\n" }
> > > +*> { dg-output " \n" }
> > > +*> { dg-output "There should be no garbage after character 32\n"
> > > +} *> { dg-output
> > > "-------------------------------\[*\]-----------------------------
> > > --
> > > -\n" }
>
> The reason why I was suggesting using
> *> { dg-output
> {-------------------------------\*-------------------------
> -------(\n|\r\n|\r)} } rather than this is that the quoting of special
> chars is then much simpler in TCL compared to "...".
> In "" string you need to use "\\\*" to match literal *.
> The reason for using (\n|\r\n|\r) is to be prepared for Darwin and
> Windows line endings, so that you don't have to rewrite all the tests
later.
>
> > > +*> { dg-output "CheckBinary Properly True\n" } *> { dg-output
> > > +"CheckBinary Properly False\n" }
>
> And the suggestion was not to require \n (or better (\n|\r\n|\r) at
> the end of last dg-output, with some dejagnu boards and e.g. remote
> testing the last newline is sometimes lost.
>
> > > + IDENTIFICATION DIVISION.
> > > + PROGRAM-ID. check88.
> > > +
> > > + DATA DIVISION.
> > > + WORKING-STORAGE SECTION.
> > > + 01 Check88 PIC XXX VALUE SPACE.
> > > + 88 CheckSpace VALUE SPACE.
> > > + 88 CheckHi VALUE HIGH-VALUES.
> > > + 88 CheckLo VALUE LOW-VALUES.
> > > + 88 CheckZero VALUE ZERO.
> > > + 88 CheckQuotes VALUE QUOTE.
> > > + 88 CheckBob VALUE "bob".
> > > + 88 CheckBinary VALUE X"000102". *> { dg-warning
> > > + .*embedded.* }
>
> I'd suggest *> { dg-warning "embedded" } here (or if there is
> complicated quoting, *> { dg-warning {embedded} }), it isn't needed to
> add .* at the start and end, the regexp doesn't have ^ and $ implicitly
added around it.
> Note, if you want to use .* somewhere in the middle, better use
> \[^\n\r\]* in the "" strings (I think just [^\n\r] in the {} ones).
> I don't know what the line length limits in COBOL are (and if it
> applies to comments too), it is also possible to put the directive on
> a separate line and refer for diagnostics on previous (or next line):
> 88 CheckBinary VALUE X"000102".
> *> { dg-warning "embedded" "" { target *-*-* } .-1 }
>
> > > --- /dev/null
> > > +++ b/gcc/testsuite/cobol.dg/group1/dg.exp
>
> I'd strongly advise against putting hundreds of copies of the same
> *.exp driver everywhere.
> It is expensive, shows up visibly in the *.log/*.sum files and
> complicates running individual tests.
>
> You can run individual tests through e.g.
> make check-cobol RUNTESTFLAGS="dg.exp='fail.cob pa*.cob'"
> If every directory has its own *.exp file without a reason (reason can
> be different rules on what to do in the directory), then you'd need to
> remember or search to find out that to run your check_88.cob test
> alone you need to use make check-cobol
> RUNTESTFLAGS=group1.exp=check_88.cob
>
> What you want instead is modify the existing dg.exp so that it handles
> subdirectories.
> E.g. the g++.dg/dg.exp has
> # Recursively find files in $dir and subdirs, do not walk into subdirs
> # that contain their own .exp file.
> proc find-cxx-tests { dir suffix } {
> set tests [lsort [glob -nocomplain -directory $dir "*.$suffix" ]]
> foreach subdir [lsort [glob -nocomplain -type d -directory $dir *]]
{
> if { [glob -nocomplain -directory $subdir *.exp] eq "" } {
> eval lappend tests [find-cxx-tests $subdir $suffix]
> }
> }
> return $tests
> }
>
> set tests [find-cxx-tests $srcdir/$subdir {C}]
>
> # Main loop.
> g++-dg-runtest $tests "" $DEFAULT_CXXFLAGS
>
> So, that would be for COBOL
>
> # Recursively find files in $dir and subdirs, do not walk into subdirs
> # that contain their own .exp file.
> proc find-cob-tests { dir suffix } {
> set tests [lsort [glob -nocomplain -directory $dir "*.$suffix" ]]
> foreach subdir [lsort [glob -nocomplain -type d -directory $dir *]]
{
> if { [glob -nocomplain -directory $subdir *.exp] eq "" } {
> eval lappend tests [find-cob-tests $subdir $suffix]
> }
> }
> return $tests
> }
>
> set tests [find-cob-tests $srcdir/$subdir {cob}]
>
> # Main loop.
> if [check_effective_target_cobol] {
> cobol-dg-runtest $tests "" $all_flags }
>
> Jakub
I will rework in the light of what you are saying here.
Thank you so very much.
Bob D.
p.s. Oh, and I finally found my way to "git gcc-verify", so it's
conceivable that I'll have more luck with commit messages. Or at least
possibly conceivable.
