* defs: Delete.
* configure.ac (AC_CONFIG_LINKS): No longer link it.
* t/ax/tap-setup.sh: Don't look for a file named 'defs' in
the grandparent directory of the current test directory to
decide whether we're running in the correct directory.
* syntax-check.mk (xdefs): Remove '$(srcdir)/defs'.
* (sc_tests_Exit_not_exit): Rename ...
(sc_tests_exit_not_Exit): ... like this, which is more faithful
to what the check actually does. Adjust a grammaro in comments
while at it.
(syntax_check_rules): Adjust.
tests: prefer including 'test-init.sh' rather than './defs'
This is a follow-up to today's commit v1.12.4-22-g0610fc8,
"tests: prepare to move ./defs to t/ax/test-init.sh"
* All tests: To run the common setup, use the command:
. test-init.sh
instead of the older, "historical" one:
. ./defs || exit 1
Note that the "|| exit 1" wasn't really useful, since the 'errexit'
shell flag is in effect in both './defs' and 'test-init.sh', and all
the known shells that are good enough to run the automake testsuite
do automatically exit with error when a sourced file cannot be found
(at least, they do so in non-interactive mode, which is the only
mode that concerns us in the testsuite).
* maint:
tests: merge, tweak and modernize few test scripts
tests: move coverage about BUILT_SOURCES
tests: more meaningful names for some test cases
tests: merge some grepping tests on Yacc support
tests: merge, tweak and modernize few test scripts
Basically an adjusted-and-improved cherry-pick from Automake-NG
commit v1.12.1-343-gff30f83.
* t/specflg.sh, t/specflg2.sh, t/specflg3.sh: Merged into ...
* t/per-target-flags.sh: ... this test.
* t/fo.sh: Remove, its weak grepping checks well superseded by
the semantic checks in 't/fort4.sh'.
* t/cxxo.sh: Remove, its weak grepping checks well superseded
by the semantic checks in 't/cxx-demo.sh'.
* t/cxxcpp.sh: Enhance a little.
* t/empty.sh: Renamed ...
* t/empty-data-primary.sh: ... to this. Add trailing ':' command.
* t/empty2.sh, t/empty3.sh, t/empty4.sh: Merged ...
* t/empty-sources-primary.tap: ... into this new test.
* t/no-outdir-option.sh: Remove. A test to check than an obsolete
and now deleted option ("--output-dir") stays deleted is way too
much even for the most test-infected person ;-)
* t/list-of-tests.mk: Adjust.
Basically a backport of some tests from Automake-NG.
* t/built-sources-check.sh: Sync it with the version in the ng/master
branch. Accordingly, move part of the checks out ...
* t/built-sources-install.sh: ... into this new test, synced from
ng/master as well.
* t/built-sources-subdir.sh: Minor tweaks and enhancements to sync it
with the version in ng/master.
* t/built-sources-cond.sh: New test, synced from ng/master.
* t/built-sources.sh: Likewise, with minor edits to avoid a spurious
failure.
* t/built-sources-fork-bomb.sh: Likewise.
* t/list-of-tests.mk: Update.
tests: prepare to move ./defs to t/ax/test-init.sh
We don't do this in a sweeping passage, because that would cause
endless headaches in the synchronization between the maint, master
and ng/master branches. Instead, we setup our framework to allow
test scripts to work by sourcing either './defs' or 'test-init.sh',
so that we'll be able to make the transition gradual and painless.
* t/ax/test-init.sh: New, copied from the previous ./defs file.
* defs: Simply work by sourcing the new file.
* Makefile.am (dist_noinst_DATA): List the new file.
* t/README: Adjust to mandate the sourcing of 'test-init.sh' rather
than of './defs'.
* t/c-demo.sh: Source 'test-init.sh' instead of ./defs. This is
done to verify our new setup actually works.
* t/ac-output-old.tap: Likewise.
* maint:
depcomp: avoid potential interferences from the environment
depcomp: improve comments about the 'gcc' depmode
sync: update files from upstream with "make fetch"
maintcheck: remove an obsolescent check
tests: rename some tests to more expressive names, again
tests: remove an obsolescent grepping check
tests: merge two tests on automatic remake functionality
tests: rename some test to more expressive names
news: 'compile' supports libfoo.a naming when wrapping Microsoft tools
tests: ensure generation of wrapper tests matching multiple conditions
tests: simplify a loop in gen-testsuite-part
compile: support libfoo.a naming when wrapping Microsoft tools
NEWS: fix wording and grammaros, re-wrap text accordingly
cosmetics: fix typo in 'lib/depcomp' comments
* syntax-checks.mk (sc_test_names): Remove this check, which verified
that no test name contained an m4/m4sugar builtin or macro name. Since
most tests use their own name as the first argument to AC_INIT, doing
that would have tickled a bug in Autoconf 2.62; but the bug was fixed
in Autoconf 2.63 already; and we are going to soon require Autoconf
2.65 anyway (in automake 1.13), so this check has become more annoying
than useful.
(syntax_check_rules): Don't list the removed check.
* t/remake3.sh: This one, superseded by ...
* t/remake3a.sh: ... this semantic test, whose comments have been
djusted accordingly.
* t/list-of-tests.mk: Adjust.
Jim Meyering [Mon, 22 Oct 2012 14:14:49 +0000 (16:14 +0200)]
compat: reinstate AM_PROG_MKDIR_P, for gettext
Do not remove AM_PROG_MKDIR_P just yet.
gettext (latest from git) still AC_REQUIRE's AM_PROG_MKDIR_P via its
intl.m4 and po.m4 files, which are pulled into *many* projects.
When I try to build one of those projects (coreutils) using the latest
from automake.git/master, I see this failure:
$ aclocal -I m4
configure.ac:477: warning: AM_PROG_MKDIR_P is m4_require'd \
but not m4_defun'd
m4/po.m4:23: AM_PO_SUBDIRS is expanded from...
m4/gettext.m4:57: AM_GNU_GETTEXT is expanded from...
configure.ac:477: the top level
That is because AM_PROG_MKDIR_P was removed (via commit v1.12-20-g8a1c64f) in preparation for the next release of automake.
* NEWS: Remove the paragraph that announced the removal of
AM_PROG_MKDIR_P.
* Makefile.am (dist_automake_ac_DATA): Add m4/mkdirp.m4.
* m4/mkdirp.m4: Re-add file.
* t/mkdirp-deprecation.sh: Likewise.
* t/list-of-tests.mk: Add it.
* automake.in: Restore removed code, and adjust comments, s/1.13/1.14/
to reflect new plan for removal.
* doc/automake.texi (Obsolete Macros): Restore the section, but
now with only one entry: the one for AM_PROG_MKDIR_P.
tests: ensure generation of wrapper tests matching multiple conditions
* gen-testsuite-part: Our old code to generate wrapper tests had a
severe limitation, in that if a test matched two or more conditions
calling for generation of wrapper tests, still only one wrapper test
was generated, instead of the three that would have been expected --
that is, one using the setup code triggered by the first condition,
one using the setup code triggered by the second condition, and one
using both this setup code fragments.
Admittedly, this was only a theoretical limitation for the moment,
since since so far no test exists that matches two or more conditions
for wrapping. Still, this might change in the future, and easily in
an unnoticed way, so better fix the issue now, before it might become
a real problem.
* gen-testsuite-part: No need to loop on the (key, value) entries
of the %test_generators has: we only use the value, and never the
key. So loop simply on the values.
Peter Rosin [Wed, 3 Oct 2012 22:08:26 +0000 (00:08 +0200)]
compile: support libfoo.a naming when wrapping Microsoft tools
There is a future plan to provide some means to have Automake
create static libraries that are named differently depending
on the system [1].
The background is that everyone has always named static libraries
libfoo.a, except the Redmond crowd who names them foo.lib, and
you have to jump through hoops to have Automake create libraries
named foo.lib in the land of non-GNU Windows while still creating
libfoo.a everywhere else.
However, there is probably no sane way to accomplish that system
dependent naming discussed in [1] without user intervention,
which makes it necessary to support the classic libfoo.a naming
when using Microsoft tools in the best possible way, for the
benefit of all projects today and for future projects not
opting in to whatever scheme is selected for the problem at
hand.
* lib/compile (func_cl_dashl): As a last resort, match -lfoo with
libfoo.a, if that file exist on the library search path.
* t/compile4.sh: Remove obsolescent workaround for the above.
* t/compile6.sh: Extend to check that libbaz.a is indeed found
when baz.lib and baz.dll.lib does not exist and that bar.lib
and bar.dll.lib are preferred over libbar.a.
Signed-off-by: Peter Rosin <peda@lysator.liu.se> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
* maint:
config headers: remove stale comment in makefile fragment
NEWS: wording and quoting fixlets in few older entries
config headers: don't emit rules for headers not generated by autoheader
docs: fix minor typo: s/expending/expanding/
sync: update files from upstream with "make fetch"
maint: post-release minor version bump
maint: typo fixes s/lies into/lies in/
release: stable release 1.12.4
NEWS: minor fix
config headers: don't emit rules for headers not generated by autoheader
This change fixed automake bug#12495.
Even if an AC_CONFIG_HEADERS invocation is passed a list of several files
as the first argument, only the first one of those file is considered by
autoheader for automatic generation of the corresponding '.in' template.
This is done on purpose, and is clearly documented in the Autoconf manual,
which (as of the 2.69 version) reads something like this:
The autoheader program searches for the first invocation of
AC_CONFIG_HEADERS in configure sources to determine the name of
the template. If the first call of AC_CONFIG_HEADERS specifies
more than one input file name, autoheader uses the first one.
That is, an invocation like:
AC_CONFIG_HEADERS([config.h config2.h])
should cause autoheader to generate only a 'config.h.in' template,
and not also a 'config2.h.in' one.
Accordingly, automake, when tracing AC_CONFIG_HEADERS, should generate
remake rules only for the template associated to the first input file
name passed to that macro. In some situations, however, automake failed
to properly limit itself in this way; for example, with an input like:
AC_CONFIG_HEADERS([config.h sub/foo.h])
in configure.ac, and with the 'sub' directory listed in the SUBDIRS
variable of the top-level Makefile, automake would erroneously generate
in 'sub/Makefile.in' a rule to remake the 'foo.h.in' template by
invoking autoheader.
* NEWS: Update.
* doc/automake.texi (Optional): Improve wording in the description of
hat rules automake generates in response to an 'AC_CONFIG_HEADERS'
invocation.
* lib/am/remake-hdr.am: Only emit autoheader-invoking remake rules for
the %CONFIG_HIN% template if that corresponds to the first argument of
AC_CONFIG_HEADERS, as explaned above. Do so using the automake-time
conditional %?FIRST-HDR%, that is properly passed ...
* automake.in (handle_configure): ... from a 'file_contents' invocation
in here.
* t/autohdr-subdir-pr12495.sh: New test.
* t/list-of-tests.mk: Add it.
* THANKS: Update.
Jim Meyering [Wed, 12 Sep 2012 15:31:24 +0000 (17:31 +0200)]
fix typos in mk-dirp.m4
* t/gettext-macros.sh: Fix typo in name of macro emitted into
mk-dirp.m4: s/AM_MKDIR_P/AM_PROG_MKDIR_P/,
and as Stefano Lattarini noted, also fix this typo:
s/AC_MKDIR_P/AC_PROG_MKDIR_P/.
coverage: better exposure for automake bug#12372 (tags-related)
Alas, in contrast with what is said in the commit message of previous
commit 'v1.12.3-14-g94b7b8e', that bug is still present also in the
current maint branch (which will become automake version 1.12.4); it
is just that it only triggers when a _SOURCES variable contains only
files with custom extension.
* t/tags-pr12372.sh: Extend.
* t/list-of-tests.mk: Add it.
Suggested-by: Юрий Пухальский <aikipooh@gmail.com> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
* maint:
sync: update files from upstream with "make fetch"
news: cygnus will be removed in automake 1.13
news: some changes for 1.13 has been "de-planned"
news: report that the have seen fixlets after 1.12.3
warns: enable category 'obsolete' by default
* NEWS (Future backward-incompatibilities): The planned Automake release
1.13 already has too much stuff on its plate; so we are not going to
In Automake 1.13, we are definitely not going to change the exact order
in which the directories in the aclocal macro search path are looked up.
Also, experience and user feedback have shown that the "obsolescent"
two-arguments invocation for AM_INIT_AUTOMAKE:
AM_INIT_AUTOMAKE(PACKAGE-NAME, PACKAGE-VERSION)
is still useful (until at least Autoconf is fixed to offer better support
for "dynamically" package versions), so we are not going to remove
support for that usage in Automake 1.13. For more details, see commit v1.12.2-245-g2abe183 of 2012-08-24, "AM_INIT_AUTOMAKE: allow obsolescent
two-args invocation once again".
(New in 1.12.1): Adjust accordingly.
No surprise that our users were bitten by backward-incompatible changes
especially hard: the warnings in the 'obsolete' category, that might
have informed them of the upcoming incompatibilities, and help them to
prepare for the transition, where not enabled by default!
* NEWS, doc/automake.texi: Update.
* lib/Automake/ChannelDefs.pm: Enable warnings in the category 'obsolete'
by default.
* t/warnings-obsolete-default.sh: New test.
* t/list-of-tests.mk: Add it.
* t/backcompat.sh: Use 'configure.ac' rather than 'configure.in' as
autoconf input file, to avoid spurious aclocal errors.
* t/backcompat2.sh: Likewise.
* t/backcompat3.sh: Likewise.
* t/backcompat5.sh: Add '-Wno-obsolete' when invoking aclocal. Adjust
heading comments.
* t/backcompat6.sh: Likewise.
* t/cygnus-imply-foreign.sh: Add '-Wno-obsolete' when invoking automake.
* maint:
sync: update files from upstream with "make fetch"
automake: don't define many identical 'lang_*_rewrite' subroutines
coverage: bugs #8844 and #9933 (already fixed by Akim's work on ylwrap)
coverage: bugs #8844 and #9933 (already fixed by Akim's work on ylwrap)
* t/flex-header.sh: New test, show that automake bug#8844 and bug#9933
have already been fixed by the recent-ish improvements to ylwrap (merged
with commit v1.12.2-27-gec5cb49 of 2012-07-16, "Merge branch 'yacc-work'
into maint").
* t/list-of-tests.mk: Update.
* maint:
docs: don't suggest to use recursive makefile setup
tests: fix a timestamp race in python tests
tests: fixup: make distcheck-override-infodir pass again
sync: update files from upstream with "make fetch"
maint: post-release minor version bump
release: stable release 1.12.3
maintcheck: fix spurious warnings
docs: fix typo: s/make install-info/make uninstall-info/
tests: fixup: make a couple of tests executable
AM_INIT_AUTOMAKE: allow obsolescent two-args invocation once again
This partially reverts commit 'v1.12-67-ge186355' of 2012-05-25,
"init: obsolete usages of AM_INIT_AUTOMAKE not supported anymore"
Some users still need to be able to define the version number for
their package dynamically, at configure runtime.
Their user case is that, for development snapshots, they want to be
able to base the complete version of the package on the VCS revision
ID (mostly Git or Mercurial). They could of course do so by
specifying such version dynamically in their call to AC_INIT, as is
done by several GNU packages. But then they would need to regenerate
and re-run the configure script before each snapshot, which might be
very time-consuming for complex packages, to the point of slowing
down and even somewhat impeding development.
The situation should truly be solved in Autoconf, by allowing a way
to specify the version dynamically in a way that doesn't force the
configure script to be regenerated and re-run every time the package
version changes. But until Autoconf has been improved to allow
this, Automake will have to support the obsolescent two-arguments
invocation for AM_INIT_AUTOMAKE, to avoid regressing the suboptimal
but working solution for the use case described above.
See also:
<http://lists.gnu.org/archive/html/automake/2012-08/msg00025.html>
* NEWS: Update.
* m4/init.m4 (AM_INIT_AUTOMAKE): Support once again invocation with
two or three arguments.
* t/aminit-moreargs-no-more.sh: Renamed ...
* t/aminit-moreargs-deprecated.sh: ... like this, and updated.
* t/nodef.sh: Recovered test, with minor adjustments.
* t/backcompat.sh: Likewise.
* t/backcompat2.sh: Likewise.
* t/backcompat3.sh: Likewise.
* t/backcompat6.sh: Likewise.
* t/list-of-tests.mk: Adjust.
Suggested-by: Bob Friesenhahn n<bfriesen@simple.dallas.tx.us> Signed-off-by: Stefano Lattarini <stefano.lattarini@gmail.com>
docs: don't suggest to use recursive makefile setup
* doc/automake.texi (Introduction): Here, by erroneously telling that
"there should generally be one Makefile.am per directory of a project".
For reference, see commit 'v1.12.1-25-g61dfb47' of 2012-06-12, "docs:
recursive make considered harmful".
Adam Sampson [Thu, 16 Aug 2012 16:54:41 +0000 (18:54 +0200)]
tests: fix a timestamp race in python tests
Fixes automake bug#12210.
* t/python-missing.sh: Call aclocal and autoconf with the "--force"
option. We need this because, on fast machines, it's possible for
'mypy.m4' and 'aclocal.m4' to end up with the same timestamp as configure,
so autoconf (without the "--force" options) wouldn't bother to rebuild it,
and would just rerun the previous AM_PATH_PYTHON test, succeeding rather
than failing as expected.
* t/python-am-path-iftrue.sh: Likewise.
* t/distcheck-override-infodir.sh: Be sure that valid occurences
of the "aclocal" and "automake" strings, which can confuse the
'sc_tests_plain_automake' check, are protected by leading "#"
characters.
* t/ax/test-lib.sh: Always use '$(...)' for command subtitution,
to avoid triggering the 'sc_tests_command_subst' check; there was
still once place where `...` was used. While at it, fix a related
comment.
* t/ax/test-defs.in ($sleep): Use creative quoting to avoid
spuriously triggering the 'sc_tests_plain_sleep' check.
* maint:
tests: rework tests on AM_PATH_PYTHON
cosmetics: fix typos and references in comments
typofix: in a test diagnostic
readme: fixlets to HACKING
* t/python8.sh, t/python9.sh: Merge into ...
* t/python-am-path-iftrue.sh: ... this new test, with minor adjustments.
* t/python4.sh, t/python5.sh, t/python6.sh, t/python7.sh: Merge into ...
* t/python-missing.sh: ... this new test.
* t/python5b.sh: Rename ...
* t/python-too-old.sh: ... like this, and adjust/extend.
* t/list-of-tests.mk: Adjust.
* HACKING: Use longer "=====" lines to separate different section (this
is just eye-candy admittedly, but I prefer it).
(Release procedure): Don't tell to "update NEWS"; that should be updated
throughout the normal course of development. Instead, tell to just check
it. Improve description of the re-bootstrapping and rechecking procedure,
also suggesting to use "git clean" beforehand (with all due warnings!).
* maint:
automake: remove an unused local variable
distcheck: more resilient against possible failures
cleanup: remove almost-unused global var 'am_relative_dir'
distcheck: more resilient against possible failures
* lib/am/distdir.am (distcheck): Ensure that a failure in the commands
making the just-extracted source tree read-only cause the recipe to fail.
While at it, save a fork by creating the '_build' and '_inst' subdir
with a single mkdir invocation.
* automake.in ($am_relative_dir): Delete, it was only used once ...
(generate_makefile): ... in here, so it's simpler to inline its
expansion.
(initialize_per_input): Don't reset the deleted variable.
* maint:
cleanup: remove two almost-unused global variables: {am,in}_file_name
cleanup: remove almost-unused global var 'topsrcdir'
automake: remove an unused variable
tests: make a test script more semantic
tests: remove an obsolete, no-op test script
tests: remove an obsolete test script
* automake.in ($am_file_name, $in_file_name): Delete these, which were
used only in the 'read_main_am_file' subroutine; instead ...
(read_main_am_file): ... modify it to only work from the '$makefile_am'
argument (which it was already receiving), and the new '$makefile_in'
argument, which is now passed to it ...
(generate_makefile): ... from here.
(initialize_per_input): Don't reset the two deleted variables anymore.
* automake.in ($topsrcdir): Delete, it was only used once ...
(handle_LIBOBJS_or_ALLOCA): ... in here, so it's simpler to inline
its expansion. Improve formatting of immediately surrounding code
a little while we are at it.
(initialize_per_input): Don't reset the deleted variable.
This is mostly useful for Automake-NG, that is heavily overhauling the
generated Makefiles and thus is prone to break grepping checks (which
can sometimes end up causing false negatives in the testsuite, sadly).
But this is not a reason not to strengthen the test for mainline
Automake as well.
* t/info.sh: This: it tried to operate by checking the contents of the
variable '$(INFOS)', but that is not even defined (and probably has been
obsolete for quite a long time). Since other tests already do thorough
testing of the Texinfo support, just remove this test.
* t/list-of-tests.mk: Adjust.
* t/scripts.sh: This: it used to check that the 'AC_PROG_INSTALL' macro
was not uselessly required, but today that macro is AC_REQUIRE'd by
'AM_INIT_AUTOMAKE' anyway, so that the test is no more significant.
* t/list-of-tests.mk: Adjust.
* maint:
tests: avoid spurious failures with older Texinfo
tests: avoid tons of spurious failures on NetBSD
runtest: avoid spurious failures on NetBSD
news: dependency tracking for Portland Group Compilers is now supported
* t/distcheck-override-infodir.sh (main.texi): Add explicit calls to
'@dircategory' and '@direntry', to ensure a 'dir' file will be created
also by 'install-info' coming with Texinfo 4.8.
* t/ax/am-test-lib.sh (process_requirements): Set the '$am_tool' variable
to the empty string before trying to unset it; otherwise, we might be
attempting to unset an already-unset variable, which (together with the
presence of the 'errexit' shell flag) causes spurious failures at least
with the /bin/sh shell from NetBSD 5.1. This was actually causing the
great majority of the Automake tests (all those not using a "required=..."
declaration) to fail spuriously on that platform!
* runtest.in: Use ${1+"$@"} rather than simply "$@", because the 'set -u'
setting used in the script causes the latter to trigger a spurious error
with the NetBSD 5.1 /bin/sh ("./runtest: @: parameter not set") if there
are no arguments to the test.
* elisp-work:
news: document all the recent elisp-related changes and improvements
coverage: byte-compiling elisp files in different subdirectories
elisp: honour AM_ELCFLAFS and ELCFLAGS in byte-compilation
elisp: --batch implies -q, remove -q
elisp: support elisp files in subdirectories properly
elisp: simplify suffix rules using emacs '-L' option
elisp: no need to "absolutize" $(srcdir) and $(builddir) ...
elisp: prefer $(builddir) files over $(srcdir) ones
elisp: use suffix rules, get rid of 'elisp-comp' script (mostly a rewrite)
coverage: elisp path contains $(srcdir) and $(builddir)
coverage: emacs lisp files in subdirectories
* t/dist-formats.tap: Here: OpenSolaris zip do not accept the
'--version' option, but accept the '-v' one with a similar
meaning (if no further arguments are given).
* maint:
tests: fix a spurious XPASS on OpenIndiana
tests: avoid spurious failure of 't/vala-vapi.sh' on OpenIndiana
tests: avoid spurious failure of 't/uninstall-fail.sh' on OpenIndiana
tests: reimplement wrappers for automake and aclocal in perl
tests: work around a ksh bug w.r.t. ${1+"$@"}
depcomp: style changes to Portland Group Compilers support
depcomp: initial support for Portland Group Compilers
* t/instspc.tap: Here, by isolating the $(DESTDIR) used by runs with
different "problematic strings" to prevent them to unduly interfering
with each other. With this, the Automake testsuite finally run cleanly
on the OpenIndiana and Solaris 10 systems I have access to.
tests: avoid spurious failure of 't/vala-vapi.sh' on OpenIndiana
* t/vala-vapi.sh: Use 'printf', not 'echo', to print strings containing
substrings like '\n', that can be interpreted like escape strings. That
because the /bin/sh and the /bin/bash shell from OpenIndiana actually
interpret them that way:
$ /bin/sh -c 'echo "foo\nbar"'
foo
bar
$ /bin/bash -c 'echo "foo\nbar"'
foo
bar
tests: avoid spurious failure of 't/uninstall-fail.sh' on OpenIndiana
On current OpenIndiana (based on what once was OpenSolaris 11), the shell
/bin/sh (which, differently from what happens on Solaris, is a true POSIX
shell, thus worthy of consideration) somehow manages to "eat" the
error message from 'rm' when that fails to remove a file due to lacking
permission on the parent directory:
That is probably due to an improper optimization, that is, the shell tries
to be smart and remove the file itself instead of invoking 'rm', but fails
spectacularly in the attempt.
* t/uninstall-fail.sh: The just-described bug was causing a spurious
failure in this test case. Cater to thus situation, by relaxing the
test when a faulty shell is detected. And while at it, fix and improve
an unrelated comment.
tests: reimplement wrappers for automake and aclocal in perl
This will allow us to avoid one extra shell invocation per automake
and aclocal invocation in our testsuite, and, more importantly, will
allow us not to worry about potential shell portability issues, at
least in those wrappers. For an example of such a portability issue,
refer to the recent commit v1.12.2-80-g65dadf6 "tests: work around a
ksh bug w.r.t. ${1+"$@"}".
* t/wrap/automake.in, t/wrap/aclocal.in: Rewritten in perl.
Fixes automake bug#10898. See also the older (much older) thread:
<http://lists.gnu.org/archive/html/automake-patches/2009-12/msg00036.html>
At least the AT&T and OpenSolaris versions of the Korn shell, as well
as the /bin/sh from OpenIndiana 11, have a strange bug regarding the
expansion of ${1+"$@"}: when exactly *one empty* argument is passed to
a script run by one of this shells, inside that script ${1+"$@"} will
expand to *nothing*, rather than to to the single empty string, as
one would expect (OTOH, $# will correctly expand to 1). This buggy
behaviour was causing a spurious failure in our testsuite (test 6 in
't/automake-cmdline.tap'). Work around it.
projects / thirdparty / automake.git / log
