]> granicus.if.org Git - check/log
check
8 years agoMerge pull request #57 from RyDroid/travis-license
Branden Archer [Thu, 7 Jul 2016 00:29:44 +0000 (20:29 -0400)]
Merge pull request #57 from RyDroid/travis-license

Adding the license header to the file for Travis CI

8 years agoAdding the license header to the file for Travis CI
Nicola Spanti (RyDroid) [Mon, 4 Jul 2016 21:00:30 +0000 (18:00 -0300)]
Adding the license header to the file for Travis CI

8 years agoMerge pull request #54 from brarcher/mem-leak-segv
Branden Archer [Wed, 29 Jun 2016 12:09:14 +0000 (08:09 -0400)]
Merge pull request #54 from brarcher/mem-leak-segv

Do not report failure line numbers if file not setup

8 years agoDo not report failure line numbers if file not setup
Branden Archer [Wed, 29 Jun 2016 02:12:23 +0000 (22:12 -0400)]
Do not report failure line numbers if file not setup

When running the master suite failure line numbers are written to
a file. Usually this is fine. However, the check_mem_leak.c test
re-uses the same tests. As a result, every time a failure is hit
a SEGV occurs because the file is never setup.

The memory leak test does not check against the line numbers,
so omitting these are fine. This fix prevents lots of SEGV
failures when running check_mem_leak.c under valgrind.

8 years agoMerge pull request #53 from brarcher/dup-tcase-ignore
Branden Archer [Sun, 19 Jun 2016 23:02:20 +0000 (19:02 -0400)]
Merge pull request #53 from brarcher/dup-tcase-ignore

Dup tcase ignore

8 years agoDisallow a test case from being added to a suite twice
Branden Archer [Sun, 19 Jun 2016 20:52:26 +0000 (16:52 -0400)]
Disallow a test case from being added to a suite twice

It was possible to add a test case to a suite more than
once. When the suite is freed later on the test case is
freed twice, causing a double free error.

Although is likely a bug from the test writer, this change
allows check to disallow double additions to prevent
double freeing a test case.

https://github.com/libcheck/check/issues/50

8 years agoRemove double addition of test suite
Branden Archer [Sun, 19 Jun 2016 20:49:20 +0000 (16:49 -0400)]
Remove double addition of test suite

The original purpose of adding the default timeout tests twice
does not seem to be relevant anymore; the timeout tests have been
split into several different types, each using either the
tcase_set_timeout or an environment variable. The tc_timeout_default
suite never has anything but the default timeout set, so the
spirit of why it was originally added no longer applies.
In addition, if the suite were formally torn down it would cause
a double free error.

8 years agoAdd call to check if list contains an item
Branden Archer [Sun, 19 Jun 2016 02:37:31 +0000 (22:37 -0400)]
Add call to check if list contains an item

8 years agoMerge pull request #47 from RyDroid/gitignore
Branden Archer [Tue, 7 Jun 2016 18:57:10 +0000 (11:57 -0700)]
Merge pull request #47 from RyDroid/gitignore

Improving .gitignore

8 years agoImproving .gitignore
Nicola Spanti (RyDroid) [Mon, 6 Jun 2016 01:15:31 +0000 (22:15 -0300)]
Improving .gitignore

8 years agoMerge pull request #45 from RyDroid/editorconfig
Branden Archer [Tue, 7 Jun 2016 01:03:07 +0000 (18:03 -0700)]
Merge pull request #45 from RyDroid/editorconfig

Adding an EditorConfig file

8 years agoAdding an EditorConfig file
Nicola Spanti (RyDroid) [Mon, 6 Jun 2016 01:08:47 +0000 (22:08 -0300)]
Adding an EditorConfig file

8 years agoMerge pull request #46 from RyDroid/readme
Branden Archer [Mon, 6 Jun 2016 03:10:05 +0000 (23:10 -0400)]
Merge pull request #46 from RyDroid/readme

Improving README

8 years agoImproving README
Nicola Spanti (RyDroid) [Mon, 6 Jun 2016 01:16:21 +0000 (22:16 -0300)]
Improving README

8 years agoMerge pull request #43 from libcheck/percent-n-check
Branden Archer [Wed, 1 Jun 2016 01:08:41 +0000 (21:08 -0400)]
Merge pull request #43 from libcheck/percent-n-check

Add test case where %n is passed into a ck_assert call

8 years agoAdd test case where %n is passed into a ck_assert call
Branden Archer [Thu, 26 May 2016 03:15:17 +0000 (23:15 -0400)]
Add test case where %n is passed into a ck_assert call

In the past it was possible to cause issues if arguments
were passed into ck_assert* macros which included %n.
Exist unit tests check for output format options, such as
%d and %f. No test existed for %n, which writes a value to
memory if used.

Ulrich Eckhardt observed that passing something with a %n in it
would result in issues running Check 0.9.10. Although the
issue is already resolved, this commit adds the test case mentioned in
Ulrich's GitHub issue: https://github.com/libcheck/check/issues/41

8 years agoMerge pull request #39 from libcheck/appveyor-unit-test-output
Branden Archer [Sun, 24 Apr 2016 14:03:35 +0000 (10:03 -0400)]
Merge pull request #39 from libcheck/appveyor-unit-test-output

Output unit test failures for MSVC on AppVeyor

8 years agoOutput unit test failures for MSVC on AppVeyor
Branden Archer [Sun, 24 Apr 2016 02:39:59 +0000 (22:39 -0400)]
Output unit test failures for MSVC on AppVeyor

8 years agoMerge pull request #37 from libcheck/rm-ac_config_aux_dir-example
Branden Archer [Wed, 20 Apr 2016 14:34:01 +0000 (10:34 -0400)]
Merge pull request #37 from libcheck/rm-ac_config_aux_dir-example

remove AC_CONFIG_AUX_DIR in example

8 years agoremove AC_CONFIG_AUX_DIR in example
Branden Archer [Wed, 20 Apr 2016 13:28:25 +0000 (09:28 -0400)]
remove AC_CONFIG_AUX_DIR in example

Potentially due to changes in automake and autoreconf

  https://www.gnu.org/software/automake/manual/html_node/Error-required-file-ltmain_002esh-not-found.html

the ltmain.sh script, generated by libtoolize, does not seem to honor the AC_CONFIG_AUX_DIR parameter. As a
result, the configure script will fail. To work around this, remove the AC_CONFIG_AUX_DIR option
to place all generated scripts into the current directory.

8 years agoMerge pull request #36 from Earnestly/master
Branden Archer [Wed, 20 Apr 2016 13:23:34 +0000 (09:23 -0400)]
Merge pull request #36 from Earnestly/master

Remove AC_CONFIG_AUX_DIR

8 years agoRemove AC_CONFIG_AUX_DIR
Earnestly [Tue, 19 Apr 2016 18:43:44 +0000 (19:43 +0100)]
Remove AC_CONFIG_AUX_DIR

This is half an issue report and half a pull request.

The reason for this is that while autoreconf -fi works for most of the
needed scripts it doesn't copy ltmain.sh into build-aux which results
the following ./configure failure:

    configure.ac:137: error: required file 'build-aux/ltmain.sh' not found

This patch, which is a similar one I've provided to cloog[0] which
allows ./configure to at least complete.

I'm not sure what you think since obviously the directory was used for
some reason, likely to reduce top-level directory clutter, but this at
least works for me.

0 https://groups.google.com/forum/?hl=en#!topic/cloog-development/28_M3Qk87-Q

Signed-off-by: Earnestly <zibeon@gmail.com>
8 years agoMerge pull request #35 from jd-boyd/readme_md
Branden Archer [Wed, 6 Apr 2016 12:39:10 +0000 (08:39 -0400)]
Merge pull request #35 from jd-boyd/readme_md

Convert README to README.md.

8 years agoAdjust README.md to be found out of tree.
Joshua D. Boyd [Wed, 6 Apr 2016 04:49:44 +0000 (00:49 -0400)]
Adjust README.md to be found out of tree.

8 years agoConvert README to README.md.
Joshua D. Boyd [Thu, 31 Mar 2016 03:48:06 +0000 (23:48 -0400)]
Convert README to README.md.

Now README is generated from README.md for packaging purposes.

Also add travis and appveyor status icons.

This addresses issue #32.

8 years agoMerge pull request #27 from jd-boyd/test_cmake
Branden Archer [Wed, 30 Mar 2016 11:58:56 +0000 (07:58 -0400)]
Merge pull request #27 from jd-boyd/test_cmake

Change travis config to test both cmake and autoconf

8 years agoAdd check_check_export to CMake test list.
Joshua D. Boyd [Wed, 30 Mar 2016 07:34:33 +0000 (03:34 -0400)]
Add check_check_export to CMake test list.

8 years agoFix running tests from cmake out of tree.
Joshua D. Boyd [Wed, 30 Mar 2016 07:13:11 +0000 (03:13 -0400)]
Fix running tests from cmake out of tree.

test_vars was being built in tree, and while the test programs were built out of tree, the CMakeLists.txt led to trying to run them in tree.

8 years agoIn travis, use make test for cmake builds.
Joshua D. Boyd [Sun, 27 Mar 2016 04:58:25 +0000 (00:58 -0400)]
In travis, use make test for cmake builds.

Keep make check in autoconf builds.

8 years agoChange travis config to test both cmake and autoconf
Joshua D. Boyd [Wed, 23 Mar 2016 06:26:52 +0000 (02:26 -0400)]
Change travis config to test both cmake and autoconf

Add Joshua Boyd to AUTHORS file.

8 years agoMerge pull request #31 from libcheck/appveyor-msvc
Branden Archer [Mon, 28 Mar 2016 14:14:37 +0000 (10:14 -0400)]
Merge pull request #31 from libcheck/appveyor-msvc

Enable AppVeyor to build for MSVC

8 years agoautotools: Do not attempt to check for (v)asprintf
Branden Archer [Mon, 28 Mar 2016 11:59:41 +0000 (07:59 -0400)]
autotools: Do not attempt to check for (v)asprintf

These functions are not used in Check

8 years agoBuild MSVC with AppVeyor
Branden Archer [Mon, 28 Mar 2016 11:58:58 +0000 (07:58 -0400)]
Build MSVC with AppVeyor

8 years agoremove (v)asprintf from snprintf.c
Branden Archer [Mon, 28 Mar 2016 11:50:33 +0000 (07:50 -0400)]
remove (v)asprintf from snprintf.c

these functions were never used, and their existence was
causing some difficulty for CMake and at least the MSVC
compiler for Windows.

8 years agoMerge pull request #30 from libcheck/appveyor-ci
Branden Archer [Mon, 28 Mar 2016 11:39:55 +0000 (07:39 -0400)]
Merge pull request #30 from libcheck/appveyor-ci

Create appveyor.yml

8 years agoCreate appveyor.yml
Branden Archer [Mon, 28 Mar 2016 03:05:58 +0000 (23:05 -0400)]
Create appveyor.yml

8 years agoMerge pull request #29 from jd-boyd/osx_cmake_itimerspec
Branden Archer [Thu, 24 Mar 2016 12:46:39 +0000 (08:46 -0400)]
Merge pull request #29 from jd-boyd/osx_cmake_itimerspec

Detect missing itimerspec on OSX.

8 years agoDetect missing itimerspec on OSX.
Joshua Boyd [Wed, 23 Mar 2016 21:54:41 +0000 (17:54 -0400)]
Detect missing itimerspec on OSX.

Set define to compiler accordingly.

This fixes cmake on osx support.

8 years agoMerge pull request #25 from jd-boyd/use_have_mkstemp
Branden Archer [Wed, 23 Mar 2016 12:30:21 +0000 (08:30 -0400)]
Merge pull request #25 from jd-boyd/use_have_mkstemp

In cmake build, use mkstemp when present

8 years agoRemove trailing whitespace from CMakeLists.txt.
Joshua D. Boyd [Wed, 23 Mar 2016 02:59:21 +0000 (22:59 -0400)]
Remove trailing whitespace from CMakeLists.txt.

8 years agoIn cmake build, use mkstemp when present
Joshua D. Boyd [Tue, 22 Mar 2016 04:50:34 +0000 (00:50 -0400)]
In cmake build, use mkstemp when present

This fixes this error, seen when building with cmake:

    check/src/check_msg.c:247: warning: the use of 'tempnam' is dangerous, better use `mkstemp'

CMakeLists.txt checks for mkstemp being present and sents a cmake
variable HAVE_MKSTEMP.  The check_msg.c file will use mkstemp if
HAVE_MKSTEMP is true, but the cmake variable wasn't causing HAVE_MKSTEMP
to be defined for the C compiler.

8 years agoMerge pull request #24 from blueness/posix-strsignal
Branden Archer [Tue, 22 Mar 2016 22:09:15 +0000 (18:09 -0400)]
Merge pull request #24 from blueness/posix-strsignal

Fix a type in configure.ac and conform to POSIX strsignal

8 years agoAUTHORS: list Anthony G. Basile as a contributor
Anthony G. Basile [Tue, 22 Mar 2016 07:12:11 +0000 (03:12 -0400)]
AUTHORS: list Anthony G. Basile as a contributor

Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
8 years agolib/strsignal.c: strsignal() should be not be declared const
Anthony G. Basile [Mon, 21 Mar 2016 12:53:31 +0000 (08:53 -0400)]
lib/strsignal.c: strsignal() should be not be declared const

POSIX requires strsignal() to return a pointer to a char, not a pointer
to a const char. [1]  On uClibc, and possibly other libcs, this causes
problems with the correct declaration in string.h.

[1] man 3 strsignal

Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
8 years agoconfigure.ac: fix typo
Anthony G. Basile [Mon, 21 Mar 2016 12:41:25 +0000 (08:41 -0400)]
configure.ac: fix typo

AC_CHECK_DECLS requires commas between declaration.  The missing comma
leads to HAVE_DECL_GETLINE_GETTIMEOFDAY which is non-sensical.

Signed-off-by: Anthony G. Basile <blueness@gentoo.org>
8 years agoMerge pull request #23 from libcheck/authors
Branden Archer [Mon, 7 Mar 2016 20:22:43 +0000 (15:22 -0500)]
Merge pull request #23 from libcheck/authors

Update AUTHORS to mention updating via pull request

8 years agoUpdate AUTHORS to mention updating via pull request
Branden Archer [Mon, 7 Mar 2016 18:17:15 +0000 (13:17 -0500)]
Update AUTHORS to mention updating via pull request

8 years agoMerge pull request #22 from jnpkrn/AUTHORS-update
Branden Archer [Mon, 7 Mar 2016 18:15:21 +0000 (13:15 -0500)]
Merge pull request #22 from jnpkrn/AUTHORS-update

AUTHORS: list Jan Pokorny as a contributor

8 years agoAUTHORS: list Jan Pokorny as a contributor
Jan Pokorný [Mon, 7 Mar 2016 09:50:22 +0000 (10:50 +0100)]
AUTHORS: list Jan Pokorny as a contributor

(as instructed)

8 years agoMerge pull request #21 from jnpkrn/fix-memmove-segfault
Branden Archer [Sat, 5 Mar 2016 03:10:48 +0000 (22:10 -0500)]
Merge pull request #21 from jnpkrn/fix-memmove-segfault

Fix segfault when memmoving with negative/enormous n

8 years agoFix segfault when memmoving with negative/enormous n
Jan Pokorný [Tue, 1 Mar 2016 17:41:58 +0000 (18:41 +0100)]
Fix segfault when memmoving with negative/enormous n

I observed a segmentation fault caused by trying to memmove by -15,
which makes 18446744073709551601 on my 64-bit platform after an argument
type promotion (from int into size_t).  In my case, this was connected
with filling up disk during the test facilitated by check, hence I derive
that the main issue was that not enough bytes for particular type of
message was actually read (and previously written, for that matter) and
because of this incompleteness, get_result happily consumed more bytes
than was read.

Additional debugging info at the point of segfault (src/check_pack.c):

> 468│         /* Move remaining data in buffer to the beginning */
> 469├>        memmove(buf, buf + n, nparse);
> 470│         /* If EOF has not been seen */
> 471│         if(nread > 0)
>
> (gdb) p nparse
> $1 = -15
> (gdb) p n
> $2 = 23
> (gdb) p nread
> $3 = 0

8 years agoMerge pull request #20 from mickbeaver/master
Branden Archer [Tue, 23 Feb 2016 01:29:58 +0000 (20:29 -0500)]
Merge pull request #20 from mickbeaver/master

Added Mick Beaver to AUTHORS.

8 years agoMerge remote-tracking branch 'upstream/master'
Mick Beaver [Mon, 22 Feb 2016 15:13:02 +0000 (09:13 -0600)]
Merge remote-tracking branch 'upstream/master'

8 years agoAdded Mick Beaver to AUTHORS.
Mick Beaver [Mon, 22 Feb 2016 14:23:15 +0000 (08:23 -0600)]
Added Mick Beaver to AUTHORS.

8 years agoMerge pull request #19 from mickbeaver/master
Branden Archer [Fri, 19 Feb 2016 21:31:17 +0000 (16:31 -0500)]
Merge pull request #19 from mickbeaver/master

Modifications for Visual Studio 2015

8 years agoModifications for Visual Studio 2015
mibeaver [Wed, 17 Feb 2016 16:31:12 +0000 (10:31 -0600)]
Modifications for Visual Studio 2015
* Added tests for snprintf and vsnprintf symbols
* Added test for time_spec in time.h.

8 years agoMerge pull request #17 from libcheck/travis-ci
Branden Archer [Sat, 6 Feb 2016 16:52:32 +0000 (11:52 -0500)]
Merge pull request #17 from libcheck/travis-ci

Create .travis.yml for travis-ci builds

8 years agoAdd texinfo package to builder on travis-ci
Branden Archer [Sat, 6 Feb 2016 07:03:19 +0000 (02:03 -0500)]
Add texinfo package to builder on travis-ci

texinfo is necessary for building documentation

8 years agoCreate .travis.yml for travis-ci builds
Branden Archer [Sat, 6 Feb 2016 06:55:57 +0000 (01:55 -0500)]
Create .travis.yml for travis-ci builds

8 years agoMerge pull request #16 from ExcaliburZero/arch-install
Branden Archer [Sun, 10 Jan 2016 22:18:38 +0000 (17:18 -0500)]
Merge pull request #16 from ExcaliburZero/arch-install

web: add Arch Linux installation information

8 years agoweb: add Arch Linux installation information
ExcaliburZero [Fri, 8 Jan 2016 22:43:39 +0000 (17:43 -0500)]
web: add Arch Linux installation information

Add information to the installation webpage on how to install the program in Arch Linux using Pacman.

Also fix a small grammar error in the Fedora installation section.

8 years agoMerge pull request #14 from libcheck/hacking-build-pull-requests
brarcher [Wed, 30 Dec 2015 02:44:39 +0000 (21:44 -0500)]
Merge pull request #14 from libcheck/hacking-build-pull-requests

Update HACKING on building pull requests

8 years agoMerge branch 'master' into hacking-build-pull-requests
brarcher [Tue, 29 Dec 2015 21:20:49 +0000 (16:20 -0500)]
Merge branch 'master' into hacking-build-pull-requests

8 years agoMerge pull request #13 from gsauthof/posix-printf
brarcher [Tue, 29 Dec 2015 21:20:37 +0000 (16:20 -0500)]
Merge pull request #13 from gsauthof/posix-printf

Use only POSIX conforming features of printf

8 years agoUpdate HACKING on building pull requests
brarcher [Tue, 29 Dec 2015 17:52:20 +0000 (12:52 -0500)]
Update HACKING on building pull requests

8 years agoUse only POSIX conforming features of printf
Georg Sauthoff [Tue, 29 Dec 2015 17:10:52 +0000 (18:10 +0100)]
Use only POSIX conforming features of printf

This fixes 3 test failures on Solaris 10.

POSIX standardized printf, but the hex style \xHH sequences
aren't included in the standard. POSIX printf does understand
octal style \NNN sequences, though.

This should also work with shells where printf is a builtin (e.g.
on Lubuntu which probably uses dash).

Tested with xpg4-sh on Solaris 10 and bash/zsh/dash on Linux.

8 years agoReplace /bin/sh with env sh in the bang line.
Georg Sauthoff [Tue, 29 Dec 2015 17:07:21 +0000 (18:07 +0100)]
Replace /bin/sh with env sh in the bang line.

The bang line is changed since /bin/sh isn't POSIX
conforming on Solaris 10.

And according to POSIX /bin/sh does not even have to be
POSIX conforming. Thus, it is more portable to rewrite
the bang line using env.

9 years agoMerge pull request #12 from brarcher/printf-fix
brarcher [Sun, 27 Dec 2015 01:14:50 +0000 (20:14 -0500)]
Merge pull request #12 from brarcher/printf-fix

use printf binary directly when forming expected test output

9 years agouse printf binary directly when forming expected test output
Branden Archer [Sat, 26 Dec 2015 16:51:45 +0000 (11:51 -0500)]
use printf binary directly when forming expected test output

It was observed on Lubuntu that the sh built-in print does not
escape hex as expected. For example, the following is expected:

$ printf "hi\x9there\n"
hi there
$

However, on Lubuntu the built-in printf command in sh produces the
following:

$ sh -c 'printf "hi\x9there\n'
hi\x9there
$

To avoid this, use the printf binary, which env will find and
invoke.

9 years agoMerge pull request #11 from brarcher/master
brarcher [Sat, 26 Dec 2015 14:07:12 +0000 (09:07 -0500)]
Merge pull request #11 from brarcher/master

add gcov information to HACKING

9 years agoadd gcov information to HACKING
Branden Archer [Sat, 26 Dec 2015 05:57:32 +0000 (00:57 -0500)]
add gcov information to HACKING

9 years agoMerge pull request #8 from brarcher/master
brarcher [Fri, 25 Dec 2015 04:57:43 +0000 (23:57 -0500)]
Merge pull request #8 from brarcher/master

More cleanup of SourceForge or SVN stuff

9 years agoRemove references to SourceForge release process
Branden Archer [Fri, 25 Dec 2015 03:37:05 +0000 (22:37 -0500)]
Remove references to SourceForge release process

This is no longer relevant for GitHub

9 years agoRemove reference to SVNChangeLog and ChangeLogOld
Branden Archer [Fri, 25 Dec 2015 03:35:33 +0000 (22:35 -0500)]
Remove reference to SVNChangeLog and ChangeLogOld

These files are no longer needed

9 years agoMerge pull request #7 from brarcher/master
brarcher [Fri, 25 Dec 2015 00:52:02 +0000 (19:52 -0500)]
Merge pull request #7 from brarcher/master

Add information about updating generated documentation for website

9 years agoAdd information about updating generated documentation for website
Branden Archer [Thu, 24 Dec 2015 23:46:14 +0000 (18:46 -0500)]
Add information about updating generated documentation for website

9 years agoMerge pull request #5 from brarcher/master
brarcher [Thu, 24 Dec 2015 21:57:58 +0000 (16:57 -0500)]
Merge pull request #5 from brarcher/master

Update HACKING with information about GitHub

9 years agoAdd information on building of pull requests to HACKING
Branden Archer [Thu, 24 Dec 2015 20:19:45 +0000 (15:19 -0500)]
Add information on building of pull requests to HACKING

9 years agoUpdate release process for GitHub
Branden Archer [Thu, 24 Dec 2015 20:19:28 +0000 (15:19 -0500)]
Update release process for GitHub

9 years agoRemove mention of SVN from HACKING
Branden Archer [Thu, 24 Dec 2015 20:14:27 +0000 (15:14 -0500)]
Remove mention of SVN from HACKING

9 years agoMerge pull request #4 from brarcher/master
brarcher [Thu, 24 Dec 2015 18:33:40 +0000 (13:33 -0500)]
Merge pull request #4 from brarcher/master

Deleting .cvsignore

9 years agoDeleting .cvsignore
brarcher [Thu, 24 Dec 2015 15:53:43 +0000 (10:53 -0500)]
Deleting .cvsignore

No longer need this, not in cvs anymore

9 years agoRemove information about packages in README
Branden Archer [Thu, 24 Dec 2015 03:39:23 +0000 (22:39 -0500)]
Remove information about packages in README

This information is likely no longer relevant or important

9 years agoMention new project page in README
Branden Archer [Thu, 24 Dec 2015 03:38:57 +0000 (22:38 -0500)]
Mention new project page in README

9 years agoweb: update webpage to point to github instead of sourceforge
Branden Archer [Thu, 24 Dec 2015 03:28:51 +0000 (22:28 -0500)]
web: update webpage to point to github instead of sourceforge

9 years agocheckmk: use double slash in regular expression
brarcher [Wed, 23 Dec 2015 16:59:04 +0000 (16:59 +0000)]
checkmk: use double slash in regular expression

For many awk implementations using both a single and double slash
escapes in regex expressions in strings works correctly. However,
it was observed on Solaris that a single slash results in the
literal value being added to the string (e.g. \n was a newline)
and only the double slash works as expected.

As there should be no harm, updating the regex expressions to use
double slashes.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1229 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agotests: fix expected strings for output tests when fork is unavailable
brarcher [Mon, 14 Sep 2015 02:11:58 +0000 (02:11 +0000)]
tests: fix expected strings for output tests when fork is unavailable

There were a few discrepancies with the expected output and the
actual output for the special characters test.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1228 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agotests: properly emit issue in log output test
brarcher [Mon, 14 Sep 2015 01:25:16 +0000 (01:25 +0000)]
tests: properly emit issue in log output test

The expected string was in ${2}, but ${expected} was used.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1227 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agotests: if xmllint exists, also run xml output through it
brarcher [Mon, 14 Sep 2015 01:25:14 +0000 (01:25 +0000)]
tests: if xmllint exists, also run xml output through it

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1226 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoEmit only valid XML characters
brarcher [Mon, 14 Sep 2015 01:25:13 +0000 (01:25 +0000)]
Emit only valid XML characters

Previously output XML files could contain illegal characters.
Some of these characters could be encoded, and others should
not be output at all.

This change will attempt to encode characters if they are not
printable or are special. Note that if a test contains a UTF-8
charater it will not be printed out accurately.

In order to get the tests for encoding XML characters to work,
the characters separating fields in a shell variable needed to
be changed. Originally white space would separate fields,
meaning shell would eat the extra whitespace in the tests.
Because of the change, the xml output test needed to be
reworked to not rely on white space field separators.
Additionally, the printf program is used to properly un-escape
strings used in the tests.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1225 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agouse filterdiff tool if available to strip timestamps from example
brarcher [Sun, 13 Sep 2015 21:55:50 +0000 (21:55 +0000)]
use filterdiff tool if available to strip timestamps from example

The diff tool will emit a timestamp along with the diff. This
timestamps prevents build output from being consistent between
builds.

If the filterdiff tool is detected, use it to strip the timestamp
and allow for reproducible build output.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1224 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoUpdate FSF address in LGPL headers
brarcher [Sun, 13 Sep 2015 16:39:09 +0000 (16:39 +0000)]
Update FSF address in LGPL headers

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1223 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoAdd LGPL header to files without the header
brarcher [Sun, 13 Sep 2015 16:39:05 +0000 (16:39 +0000)]
Add LGPL header to files without the header

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1222 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoAdd development header back to NEWS
brarcher [Sun, 2 Aug 2015 19:49:55 +0000 (19:49 +0000)]
Add development header back to NEWS

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1221 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoUpdate release information on website 0.10.0
brarcher [Sun, 2 Aug 2015 19:49:53 +0000 (19:49 +0000)]
Update release information on website

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1220 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoUpdate for release
brarcher [Sun, 2 Aug 2015 19:22:58 +0000 (19:22 +0000)]
Update for release

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1218 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoUpdate for release
brarcher [Sun, 2 Aug 2015 19:21:14 +0000 (19:21 +0000)]
Update for release

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1217 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agoconfigure: Remove duplicated use of AC_USE_SYSTEM_EXTENSIONS
brarcher [Sun, 2 Aug 2015 16:59:36 +0000 (16:59 +0000)]
configure: Remove duplicated use of AC_USE_SYSTEM_EXTENSIONS

Either the first or the second one is redundant so one should be
removed but beware, there's a bug introduced in automake 1.14
(fixed in 1.15) that would potentially cause trouble if you put it
before AM_INIT_AUTOMAKE.

More info about this automake bug here:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=15981

https://sourceforge.net/p/check/bugs/109

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1216 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agotests: adjust expected line number due to previous commit
brarcher [Mon, 6 Jul 2015 23:29:02 +0000 (23:29 +0000)]
tests: adjust expected line number due to previous commit

A header was added to the program under test, which adjusted the
line numbers.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1215 64e312b2-a51f-0410-8e61-82d0ca0eb02a

9 years agotests: include missing libcompat header
brarcher [Mon, 6 Jul 2015 13:20:31 +0000 (13:20 +0000)]
tests: include missing libcompat header

This header is necessary one some platforms where some portions
of check.h are unrecognized. For example, using the MSVC compiler
on Windows the pid_t usage in check.h results in a compiler error.

git-svn-id: svn+ssh://svn.code.sf.net/p/check/code/trunk@1214 64e312b2-a51f-0410-8e61-82d0ca0eb02a