Hi Alex, Actually, Krzysz and I are thinking about moving towards GTest.

regards, Johan

On 30-10-2014 16:58, Alex Valavanis wrote:

In fact, while we're at it, it may be worth bumping the version of cxxtest in trunk. There's a much newer upstream version (4.4) at http://sourceforge.net/projects/cxxtest/files/cxxtest/

Upstream changelog

• Version 4.4 (2014-06-03)

  • Fixed compilation error on Windows (MSVC) in XmlFormatter.h (#86)
  • Fix to ensure that tearDown() is called (#89)
  • Add option run test with a user defined command in scons (#91)
  • Use a Python path relative to the cxxtestgen script (#88)
  • Add defensive guard in ErrorFormatter.h (#96)
  • Fixed bug with "None" appearing in CXXTEST_CPPATH (#99)
  • Added CXXTEST_LIBPATH to properly use shared libraries (#100)
  • Added guards when XmlFormatter.h data is not initialize (#87)

• Version 4.3 (2013-07-05)

  • Changes to assess code coverage of the cxxtestgen command
  • Standardizing C++ file formats (using astyle)
  • Bug fixes that led to the test runner hanging
  • Adding special assertions for floating point values
  • Added date to XML output
  • Added support for comparison of C strings

• Version 4.2.1 (2013-03-22)

  • Fixing documentation of LGPL version

• Version 4.2 (2013-03-16)

  • Changes to support test fixtures in namespaces
  • Adding logic to support test skipping
  • Change to create self-contained HTML documentation
  • Fixed inheritance issue in GlobalFixture (#69)
  • Update LGPL version
  • Changes to try/catch to avoid ambiguities withn catching std::exception (#53)
  • Fixed TS_ASSERT_DELTA to work on integer types (#65)
  • Changed output format to print world-name (#70)

• Version 4.1 (2012-11-30)

  • Added absolute paths to resolve bug when relative path links are provided.
  • Bug fix when files contain unicode characters
  • Fix for --no-static-init: Changed how non-static tests are created
  • Updated user guide to include SCons build system
  • Closing out Tigris and SourceForge tickets
  • Added valgrind tests.

• Version 4.0.3 (2012-01-07)

  • Adding support for Python 2.4 - 3.2
  • Various cleanup of CxxTest root directory
  • Adding patch that allows the cxxtestgen script to be used when symlinked.

• Version 4.0.2 (2012-01-02)

  • Bug fix to enable installation of cxxtestgen without the 'setuptools' package

• Version 4.0.1 (2012-01-01)

  • Documentation updates
  • Bug fix for installation of cxxtestgen script

• Version 4.0 (2011-12-28)

  • Perl is no longer used to support CxxTest scripts. Python is now

the only scripting language used by CxxTest.

• The testing scripts have been rewritten using the PyUnit framework.
• The installation process for CxxTest now leverages and integrates

with the system Python installation.

• A more comprehensive C++ parser is now available, which supports

testing of templates.

• The CxxTest GUI is no longer supported.
• The <<ts_trace,TS_TRACE>> and <<ts_warn,TS_WARN>> macros have the

same behavior now.

• CxxTest runners now have a command-line interface that facilitates

interactive use of the test runner.

• A new user guide is now available in PDF, HTML and Ebook formats.

On 30 October 2014 15:48, Alex Valavanis <valavanisalex@...400...> wrote:

...

The xfail stuff was just a hack I made to stop old tests from killing the PPA builds... they are definitely not supposed to fail, so it's a bit of an abuse of the xfail flag and should be fixed before the release. As I understand it, they're mostly failures that appeared with changes in 2geom... I suspect they're very easy to fix for people who are familiar with the code. The relevant bug reports I made are here:

https://bugs.launchpad.net/inkscape/+bug/1249324

Hope it's of some use!

AV

On 30 October 2014 15:05, Johan Engelen <jbc.engelen@...2592...> wrote:

...

On 21-10-2014 21:06, Johan Engelen wrote:

...

On 21-10-2014 2:34, Bryce Harrington wrote:

...

Johan, would you be willing to hammer on getting 'make distcheck' to be passing 100% clean? Last I checked it was pretty close but there's a few warnings and other cleanups of the sort you like to tackle.

OK :)

I was hoping to work on this during some downtime at the mentor summit, but...

It is the first time I ran 'make distcheck', on my Mac, and... it just works? :) I see that the unit tests are set to "expect to fail", so that obviously always passes (I don't know why it is set to xfail).

In short: do I need to do anything here? Or is it already good?

cheers, Johan

---

---

Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel