apitrace-tests
9 years agoAdd --exact, --no-deps, or --no-prune to trim tests as needed master
Carl Worth [Thu, 11 Apr 2013 19:07:10 +0000 (12:07 -0700)]
Add --exact, --no-deps, or --no-prune to trim tests as needed

By explicitly specifying --exact, --no-deps, or --no-prune these tests
are now immune from any change in the default behavior of "apitrace
trim". Whether --exact or --auto is the default, these tests will get
exactly the behavior they want.

9 years agoRework testing of trimming of operations with no side effects
Carl Worth [Sat, 16 Feb 2013 01:19:57 +0000 (17:19 -0800)]
Rework testing of trimming of operations with no side effects

The previous test was bogus. The only call it trimmed was glxSwapBuffers
which is actually invalid, (this call definitely does have side effects,
but can easily be trimmed if desired with --trim-spec=drawing).

The new testing here uses a trace from a new program, (source
included), called few-side-effects which simply makes various calls
that do not have side effects, (such as glGetIntegerv and
glIsEnabled).

Then, the testing exercises the two independent mechanisms available
for trimming calls without side effects:

    trim --prune

Trims calls within the user-specified range without side
effects.

    trim --deps --trim-spec=no-side-effects

Trims calls outside the user-specific range without side
effects.

Three new tests are added here, one for each of the above cases and
one for both combined, (prune, trim, and prune-and-trim).

9 years agocli: Add test for "apitrace trim" with both --calls and --frames
Carl Worth [Mon, 11 Mar 2013 22:45:05 +0000 (15:45 -0700)]
cli: Add test for "apitrace trim" with both --calls and --frames

This case was triggering a bug in "apitrace trim" where it would
incorrectly bail out after going past the last specified call (or
frame) without noticing that there were still further frames (or
calls) to be processed.

9 years agoFix test to correctly use cmake feature to find python executable
Carl Worth [Mon, 28 Jan 2013 10:55:12 +0000 (21:55 +1100)]
Fix test to correctly use cmake feature to find python executable

On my system, at least, there was no ${python} variable set, so the
tests for PIL were incorrectly failing, (such that many tests were not
being run at all).

9 years agotrim-range: Stress the call-set parser just a tiny bit more.
Carl Worth [Tue, 21 Aug 2012 22:55:16 +0000 (15:55 -0700)]
trim-range: Stress the call-set parser just a tiny bit more.

We specify an equivalent range, but in a slightly more complex way for
better testing.

9 years agoAdd a tri glsl sample.
José Fonseca [Mon, 8 Apr 2013 18:46:02 +0000 (19:46 +0100)]
Add a tri glsl sample.

9 years agoTrack rename of "retrace" to "replay" within apitrace.
Carl Worth [Sat, 26 Jan 2013 23:45:16 +0000 (10:45 +1100)]
Track rename of "retrace" to "replay" within apitrace.

Verifying that everything still passes after the rename.

Signed-off-by: José Fonseca <jose.r.fonseca@gmail.com>
9 years agoBe more lenient with shader matching.
José Fonseca [Fri, 22 Feb 2013 15:04:20 +0000 (15:04 +0000)]
Be more lenient with shader matching.

9 years agod3d*: Don't set the BufferDesc dimmension.
José Fonseca [Fri, 22 Feb 2013 15:01:10 +0000 (15:01 +0000)]
d3d*: Don't set the BufferDesc dimmension.

9 years agocli: Add testing for recently-added "apitrace trim --trim-spec"
Carl Worth [Tue, 29 Jan 2013 04:07:34 +0000 (15:07 +1100)]
cli: Add testing for recently-added "apitrace trim --trim-spec"

Each test here is quite simple. Without doing any image-based
verification, simply ensure that the expected calls are trimmed with
each different invocation of --trim-spec.

9 years agotrim_stress: Only look at files matching [0-9]*.png when counting references
Carl Worth [Mon, 28 Jan 2013 12:05:26 +0000 (23:05 +1100)]
trim_stress: Only look at files matching [0-9]*.png when counting references

If there are any differences from a previous run, there will be
several additional image files left around, (.thumb.png, etc.). And we
don't want to count those.

9 years agotrim_stress: Also test any traces in the directory trim_stress/local
Carl Worth [Mon, 28 Jan 2013 11:04:56 +0000 (22:04 +1100)]
trim_stress: Also test any traces in the directory trim_stress/local

This is useful for performing testing on traces that are too large to
be considered for include in git, or for traces that cannot be shared
due to proprietary contents.

9 years agotrim_stress: Fail test at first frame which doesn't match reference
Carl Worth [Mon, 28 Jan 2013 11:04:01 +0000 (22:04 +1100)]
trim_stress: Fail test at first frame which doesn't match reference

When stress-testing large traces, it's inefficient to work through the
entire trace before reporting an error that occurred in one of the
first frames.

Instead, check each frame one at a time immediately as each rendered
frame becomes available. Then fail the test at the first mismatched
frame.

9 years agoFix test to correctly use cmake feature to find python executable
Carl Worth [Mon, 28 Jan 2013 20:22:31 +0000 (07:22 +1100)]
Fix test to correctly use cmake feature to find python executable

On my system, at least, there was no ${python} variable set, so the
tests for PIL were incorrectly failing, (such that many tests were not
being run at all).

Signed-off-by: José Fonseca <jfonseca@vmware.com>
9 years agoMerge branch 'trim-auto'
José Fonseca [Mon, 28 Jan 2013 19:48:41 +0000 (19:48 +0000)]
Merge branch 'trim-auto'

9 years agoAdd d3d8 scnapshot. Be consistent with swapeffect.
José Fonseca [Wed, 12 Dec 2012 09:30:09 +0000 (09:30 +0000)]
Add d3d8 scnapshot. Be consistent with swapeffect.

9 years agoDon't run tests that depend on PIL.
José Fonseca [Mon, 10 Dec 2012 18:54:24 +0000 (18:54 +0000)]
Don't run tests that depend on PIL.

Unfortunately it is non-standard and binaries hard to get by on MacOSX.

9 years agoapps: Check d3dretrace snapshots.
José Fonseca [Fri, 7 Dec 2012 10:49:13 +0000 (10:49 +0000)]
apps: Check d3dretrace snapshots.

9 years agoapps/d3d8: Be precise with window size and format.
José Fonseca [Fri, 7 Dec 2012 10:48:41 +0000 (10:48 +0000)]
apps/d3d8: Be precise with window size and format.

9 years agocli: Use python differ. Fix failures.
José Fonseca [Thu, 6 Dec 2012 07:09:48 +0000 (07:09 +0000)]
cli: Use python differ. Fix failures.

Tests were failure due to off by one difference in line nos.

9 years agoUpdate examples to use DXGI.
José Fonseca [Wed, 5 Dec 2012 19:42:07 +0000 (19:42 +0000)]
Update examples to use DXGI.

That's more common and challenging for apitrace.

9 years agoapps: Be verbose when tracing.
José Fonseca [Tue, 4 Dec 2012 13:05:21 +0000 (13:05 +0000)]
apps: Be verbose when tracing.

9 years agocli: Use difflib to compare expected vs output.
José Fonseca [Tue, 4 Dec 2012 06:31:58 +0000 (06:31 +0000)]
cli: Use difflib to compare expected vs output.

9 years agocli: Don't rely on image generation for diff-image tests.
José Fonseca [Tue, 4 Dec 2012 03:19:16 +0000 (03:19 +0000)]
cli: Don't rely on image generation for diff-image tests.

Use static images instead.

9 years agotrim: Update for --exact -> --auto
José Fonseca [Tue, 4 Dec 2012 03:17:55 +0000 (03:17 +0000)]
trim: Update for --exact -> --auto

9 years agoTrack updates to "apitrace trim" which no longer drops glBindTexture calls
Carl Worth [Thu, 16 Aug 2012 20:39:10 +0000 (13:39 -0700)]
Track updates to "apitrace trim" which no longer drops glBindTexture calls

We still aren't sure *why* apitrace trim should have to leave these
calls around, but we have a trace which only passes the
trim-stress-test if these calls are untrimmed. While that mystery
waits to be solved, make the test here compatible with the current
behavior.

9 years agoglxsimple: Introduce an inter-frame texture dependency to exercise trim code
Carl Worth [Wed, 15 Aug 2012 18:16:52 +0000 (11:16 -0700)]
glxsimple: Introduce an inter-frame texture dependency to exercise trim code

The previous glxsimple program was too simple as each frame was
entirely self-contained, (textures were created, initialized, and
referenced only within single frames).

For more realistic trimming, we create and initialize a texture in one
frame, then reference that texture in a later frame. This is a better
test to ensure that when trimming the first frame the texture creation
and initialization is not discarded.

9 years agoAdd stress tests from "apitrace trim".
Carl Worth [Wed, 15 Aug 2012 18:05:13 +0000 (11:05 -0700)]
Add stress tests from "apitrace trim".

The trim_stress directory makes it simple to stress test "apitrace
trim" by simply dropping new trace files into this directory. The
stress testing ensures that creating a new trace by trimming to a
single frame still generates the same image as that frame from the
original trace. It performs this test exhaustively for each frame of
each trace.

This testing ensures that trimming doesn't break the rendered results
of any frame. It does not do anything to test that trimming actually
discards anything. So the existing cli tests are still
independendently useful for verifying that expected content is
discarded.

9 years agoAdd a couple of tests for the --frames option of "apitrace trim"
Carl Worth [Wed, 15 Aug 2012 05:40:35 +0000 (22:40 -0700)]
Add a couple of tests for the --frames option of "apitrace trim"

Recent versions of apitrace added this new command-line option, so
it's useful to test that it actually works here.

9 years agoAdd test to ensure that unused shaders are dropped from a trace.
Carl Worth [Tue, 14 Aug 2012 23:53:21 +0000 (16:53 -0700)]
Add test to ensure that unused shaders are dropped from a trace.

This tests functionality which was added even more recently to
"apitrace trim".

9 years agocli_driver: Rework driver script to preserve json-embedded whitespace
Carl Worth [Tue, 14 Aug 2012 23:50:00 +0000 (16:50 -0700)]
cli_driver: Rework driver script to preserve json-embedded whitespace

Previously, the line read from the script was being passed through
python's string.split() and string.join() functions. This was causing
sequences of successive whitespace characters to be collapsed down to
a single space, (in turn causing spurious failures when comparing
against expected output with successive whitespace).

Rework the driver to avoid munging the line past the separation of the
first command word.

9 years agoAdd test to ensure that unused textures are dropped from a trace.
Carl Worth [Tue, 14 Aug 2012 20:16:05 +0000 (13:16 -0700)]
Add test to ensure that unused textures are dropped from a trace.

This tests functionality recently added to "apitrace trim".

9 years agotrim-head: New test to ensure "apitrace trim" drops all calls after range
Carl Worth [Mon, 13 Aug 2012 02:42:35 +0000 (19:42 -0700)]
trim-head: New test to ensure "apitrace trim" drops all calls after range

This test ensures that all calls beyond the user-specified range are
correctly dropped. It also ensures that uninteresting calls are
trimmed, (since the glxsimple trace includes many of those).

9 years agoMake glxsimple loop over all drawing methods multiple times
Carl Worth [Mon, 13 Aug 2012 00:50:04 +0000 (17:50 -0700)]
Make glxsimple loop over all drawing methods multiple times

We now have 3 different drawing methods, each performed two times, and
in a total of 6 different colors. Looping over theese things twice
means that we can trim out a single frame of any one method and ensure
that "apitrace trim" has some non-trivial trimming to do both before
and after the frame of interest.

9 years agoAdd a third frame, this time drawing with a texture.
Carl Worth [Mon, 13 Aug 2012 00:16:57 +0000 (17:16 -0700)]
Add a third frame, this time drawing with a texture.

This will allow us to test that "apitrace trim" can trim out unused
textures.

9 years agoglxsimple: Add a second frame drawing a solid color with GLSL.
Carl Worth [Mon, 13 Aug 2012 00:16:04 +0000 (17:16 -0700)]
glxsimple: Add a second frame drawing a solid color with GLSL.

This will allow us to test that we can trim out unused shaders.

9 years agoAdd the source of a very simple program using OpenGL through GLX.
Carl Worth [Mon, 13 Aug 2012 00:09:15 +0000 (17:09 -0700)]
Add the source of a very simple program using OpenGL through GLX.

We'll use this program to generate traces for use in the test suite,
(to exercise using "apitrace trim" to trim away unused textures,
shader programs, etc.).

There is a Makefile here, for convenience to any test writers wanting
to expand this program. Note that this Makefile us intentionally not
integrated with the cmake setup of the apitrace-tests project. That's
because it is not expected that this test program should be built or
run as a standard part of the test suite. Instead, useful trace files
generated from glxsimple will be directly committed to the
apitrace-tests repository.

9 years agocli: Use new --call-nos=no call to "apitrace dump-images" for consistent results
Carl Worth [Mon, 13 Aug 2012 00:00:38 +0000 (17:00 -0700)]
cli: Use new --call-nos=no call to "apitrace dump-images" for consistent results

Recent "apitrace trim" is now dropping uninteresting calls by
default. This means that the call numbers of rendering operations are
changing. Consequently, image comparisons were failing not due to
image-content differences, but simply because the filenames (with
embedded call numbers in them) no longer matched.

We can avoid this problem by using the --call-nos=no option which was
added to "apitrace dump-images" in very recent versions.

9 years agoPrint commands to be executed by cli_driver.py
Carl Worth [Sat, 11 Aug 2012 18:28:59 +0000 (11:28 -0700)]
Print commands to be executed by cli_driver.py

These aren't made visible unless CTEST_OUTPUT_ON_FAILURE in set, and
in that case it's quite convenient to see what the test was doing when
it failed.

9 years agoAdd a test to ensure "apitrace trim" performs proper dependency checking
Carl Worth [Thu, 9 Aug 2012 19:46:20 +0000 (12:46 -0700)]
Add a test to ensure "apitrace trim" performs proper dependency checking

Here we trim a trace by specifying (with --calls) only the final
glXSwapBuffers and then test that we can still produce a matching
image from the trace. This will only succeed if "apitrace trim"
performs dependency checking and includes many previous calls.

Of course, historically "apitrace trim" performed no such dependency
checks, so older versions of apitrace are expected to fail this test.

9 years agoAdd some comments to the recently-added cli tests.
Carl Worth [Thu, 9 Aug 2012 00:18:32 +0000 (17:18 -0700)]
Add some comments to the recently-added cli tests.

It's easy enough to ignore lines starting with a '#' and that lets us
make these tests more self-documenting.

9 years agoAdd a test that "apitrace diff-images" flags mismatching images.
Carl Worth [Wed, 8 Aug 2012 23:18:21 +0000 (16:18 -0700)]
Add a test that "apitrace diff-images" flags mismatching images.

With this test we add a new EXPECT_FAILURE: attribute to the script
syntax to indicate that a particular apitrace command in the script is
expected to fail. This requires manually invoking the Popen
constructor rather than using subprocess.check_output, (which is
*close* to what we need but fails to assign the output when raising an
exception).

We also remove the globbing from the CMakeLists.txt file to get a
manual ordering of the tests (rather than executing in alphabetic
order).

9 years agoAdd a new 'cli' subdirectory for higher-level testing of apitrace CLI.
Carl Worth [Wed, 8 Aug 2012 22:18:23 +0000 (15:18 -0700)]
Add a new 'cli' subdirectory for higher-level testing of apitrace CLI.

As described in the README, this new directory is for tests that
perform higher-level testing of the apitrace command-line interface
than either of the existing traces or apps directories.

Compared to traces, these new tests can do higher-level things like
dumping images from traces, comparing images, etc.

Compared to apps, these tests are intended to be portable across
multiple targets. (This isn't actually the case yet because commands
like "apitrace dump-images" currently work only on OpenGL. Also, even
when that is fixed, we will also need to then extend these tests to
have traces for each supported target for testing.)

In this commit, the only new test exercises "apitrace dump-images" and
"apitrace diff-images" on a simple trace file with one frame,
(tri.trace grabbed from apps/gl).

9 years agoAdd more README documentation on how existing test drivers are structured.
Carl Worth [Tue, 7 Aug 2012 22:47:53 +0000 (15:47 -0700)]
Add more README documentation on how existing test drivers are structured.

More than once, I've reverse-engineered how the test suite uses test
drivers (app_driver.py and tool_driver.py) to invoke scripts within
the apps/ and traces/ directory. In order to save my future-self some
work in doing this yet again, I've written some documentation in
README.markdown in each directory.

Hopefully this documentation will be useful for anyone wanting to
write additional tests.

9 years agoapps/d3d11: Comprehensive test.
José Fonseca [Tue, 27 Nov 2012 20:07:08 +0000 (20:07 +0000)]
apps/d3d11: Comprehensive test.

9 years agotracematch: Fix array parsing.
José Fonseca [Tue, 27 Nov 2012 20:06:40 +0000 (20:06 +0000)]
tracematch: Fix array parsing.

9 years agoUpdate readme.
José Fonseca [Tue, 27 Nov 2012 16:21:52 +0000 (16:21 +0000)]
Update readme.

9 years agoRemove left over debug statement.
José Fonseca [Tue, 27 Nov 2012 14:24:23 +0000 (14:24 +0000)]
Remove left over debug statement.

9 years agoAdd a shader d3d9 tri test.
José Fonseca [Tue, 27 Nov 2012 14:22:57 +0000 (14:22 +0000)]
Add a shader d3d9 tri test.

9 years agoPrecise d3d8-10.1 reference traces.
José Fonseca [Mon, 26 Nov 2012 19:50:35 +0000 (19:50 +0000)]
Precise d3d8-10.1 reference traces.

9 years agoFix parsing of floats without '.'
José Fonseca [Mon, 26 Nov 2012 19:49:17 +0000 (19:49 +0000)]
Fix parsing of floats without '.'

For example "1e+10"

9 years agoAccept comments in reference traces.
José Fonseca [Mon, 26 Nov 2012 19:48:54 +0000 (19:48 +0000)]
Accept comments in reference traces.

9 years agoHandle DOS line endings.
José Fonseca [Mon, 26 Nov 2012 19:48:37 +0000 (19:48 +0000)]
Handle DOS line endings.

9 years agoRetrace d3d8.
José Fonseca [Mon, 26 Nov 2012 19:48:04 +0000 (19:48 +0000)]
Retrace d3d8.

9 years agogl/map_buffer: Exit 1 when GL_ARB_map_buffer_range is not found.
José Fonseca [Mon, 26 Nov 2012 13:52:50 +0000 (13:52 +0000)]
gl/map_buffer: Exit 1 when GL_ARB_map_buffer_range is not found.

9 years agoOnly invoke pkg_check_modules when PKG_CONFIG_FOUND is true.
José Fonseca [Mon, 26 Nov 2012 13:51:51 +0000 (13:51 +0000)]
Only invoke pkg_check_modules when PKG_CONFIG_FOUND is true.

9 years agoUse clang on MacOSX.
José Fonseca [Mon, 26 Nov 2012 13:51:32 +0000 (13:51 +0000)]
Use clang on MacOSX.

9 years agoDrop pragmas.
José Fonseca [Sat, 24 Nov 2012 09:31:11 +0000 (09:31 +0000)]
Drop pragmas.

9 years agoRename checker.py -> tracematch.py
José Fonseca [Fri, 23 Nov 2012 22:15:01 +0000 (22:15 +0000)]
Rename checker.py -> tracematch.py

9 years agoImprove gl map buffer test.
José Fonseca [Fri, 23 Nov 2012 08:34:44 +0000 (08:34 +0000)]
Improve gl map buffer test.

9 years agoUse the new trace checker.
José Fonseca [Fri, 23 Nov 2012 07:52:17 +0000 (07:52 +0000)]
Use the new trace checker.

9 years agoFurther improvements to checker.
José Fonseca [Thu, 22 Nov 2012 22:03:17 +0000 (22:03 +0000)]
Further improvements to checker.

9 years agoImprove checker.
José Fonseca [Thu, 22 Nov 2012 09:55:15 +0000 (09:55 +0000)]
Improve checker.

9 years agoUse 'apitrace retrace'
José Fonseca [Thu, 22 Nov 2012 08:49:01 +0000 (08:49 +0000)]
Use 'apitrace retrace'

9 years agoSpecify rendertarget view descriptions.
José Fonseca [Tue, 20 Nov 2012 17:06:43 +0000 (17:06 +0000)]
Specify rendertarget view descriptions.

To exercise unions.

9 years agoAdd new retrace apis.
José Fonseca [Wed, 14 Nov 2012 09:13:25 +0000 (09:13 +0000)]
Add new retrace apis.

9 years agoAdd a d3d10 level 9 test case.
José Fonseca [Thu, 25 Oct 2012 10:45:59 +0000 (11:45 +0100)]
Add a d3d10 level 9 test case.

9 years agoUpdate state workarounds.
José Fonseca [Sun, 20 May 2012 10:16:46 +0000 (11:16 +0100)]
Update state workarounds.

10 years agoTest trimming.
José Fonseca [Sat, 19 May 2012 19:53:37 +0000 (20:53 +0100)]
Test trimming.

10 years agoMake tri_glsl test robust against attribute location changes.
José Fonseca [Mon, 14 May 2012 19:22:28 +0000 (20:22 +0100)]
Make tri_glsl test robust against attribute location changes.

10 years agoUpdate for floating point precision changes.
José Fonseca [Mon, 14 May 2012 19:21:59 +0000 (20:21 +0100)]
Update for floating point precision changes.

10 years agoPropagate apitrace trace error codes.
José Fonseca [Wed, 9 May 2012 13:18:22 +0000 (14:18 +0100)]
Propagate apitrace trace error codes.

10 years agoEnsure feature level 10.0 at minimum.
José Fonseca [Wed, 9 May 2012 11:01:33 +0000 (12:01 +0100)]
Ensure feature level 10.0 at minimum.

10 years agoAllow to build samples without apitrace source tree.
José Fonseca [Wed, 9 May 2012 10:14:08 +0000 (11:14 +0100)]
Allow to build samples without apitrace source tree.

10 years agoAllow to build d3d samples independently of apitrace source.
José Fonseca [Wed, 9 May 2012 10:10:17 +0000 (11:10 +0100)]
Allow to build d3d samples independently of apitrace source.

10 years agoRename hlsl d3dcommon.
José Fonseca [Wed, 9 May 2012 10:06:12 +0000 (11:06 +0100)]
Rename hlsl d3dcommon.

10 years agoAttempt of a d3d11.1 test.
José Fonseca [Wed, 9 May 2012 09:56:29 +0000 (10:56 +0100)]
Attempt of a d3d11.1 test.

10 years agod3d10.1 sample.
José Fonseca [Wed, 9 May 2012 09:28:25 +0000 (10:28 +0100)]
d3d10.1 sample.

10 years agoPut HLSL shaders into their own directory.
José Fonseca [Wed, 9 May 2012 08:26:35 +0000 (09:26 +0100)]
Put HLSL shaders into their own directory.

10 years agoMake OpenGL/GLUT/GLEW optional.
José Fonseca [Wed, 9 May 2012 08:25:19 +0000 (09:25 +0100)]
Make OpenGL/GLUT/GLEW optional.

10 years agod3d11 test.
José Fonseca [Tue, 8 May 2012 21:26:17 +0000 (22:26 +0100)]
d3d11 test.

10 years agoCleanup d3d10 test.
José Fonseca [Tue, 8 May 2012 18:26:03 +0000 (19:26 +0100)]
Cleanup d3d10 test.

10 years agoPrecompile d3d10 shader.
José Fonseca [Fri, 4 May 2012 10:35:48 +0000 (11:35 +0100)]
Precompile d3d10 shader.

10 years agoFixup d3d10 test.
José Fonseca [Fri, 4 May 2012 10:28:05 +0000 (11:28 +0100)]
Fixup d3d10 test.

10 years agoPrototype of d3d10 sample app.
José Fonseca [Thu, 3 May 2012 23:52:13 +0000 (00:52 +0100)]
Prototype of d3d10 sample app.

Compiles but untested.

10 years agoD3D7 sample.
José Fonseca [Thu, 3 May 2012 13:21:55 +0000 (14:21 +0100)]
D3D7 sample.

10 years agoD3D8 sample app.
José Fonseca [Thu, 3 May 2012 12:31:37 +0000 (13:31 +0100)]
D3D8 sample app.

10 years agoSimple D3D9 app.
José Fonseca [Thu, 3 May 2012 10:20:41 +0000 (11:20 +0100)]
Simple D3D9 app.

10 years agoHandle CRLF lines in tool_driver.py
José Fonseca [Thu, 3 May 2012 10:15:12 +0000 (11:15 +0100)]
Handle CRLF lines in tool_driver.py

10 years agoUpdate reference for GLboolean update.
José Fonseca [Mon, 16 Apr 2012 19:10:19 +0000 (20:10 +0100)]
Update reference for GLboolean update.

10 years agoPragmas.
José Fonseca [Sun, 18 Mar 2012 00:35:46 +0000 (00:35 +0000)]
Pragmas.

10 years agoBit more work on the checker.
José Fonseca [Sun, 18 Mar 2012 00:13:13 +0000 (00:13 +0000)]
Bit more work on the checker.

10 years agoAdd .gitignore.
José Fonseca [Sat, 17 Mar 2012 15:49:00 +0000 (15:49 +0000)]
Add .gitignore.

10 years agoTest callsets.
José Fonseca [Sat, 17 Mar 2012 15:44:02 +0000 (15:44 +0000)]
Test callsets.

10 years agoDump verbosely.
José Fonseca [Sat, 17 Mar 2012 15:11:28 +0000 (15:11 +0000)]
Dump verbosely.

10 years agoMake it easier to add more tests.
José Fonseca [Sat, 17 Mar 2012 14:50:04 +0000 (14:50 +0000)]
Make it easier to add more tests.

10 years agoBetter code organization.
José Fonseca [Sat, 17 Mar 2012 14:24:29 +0000 (14:24 +0000)]
Better code organization.

10 years agoSplit the driver code.
José Fonseca [Sat, 17 Mar 2012 14:13:33 +0000 (14:13 +0000)]
Split the driver code.

10 years agoBetter checking (WIP).
José Fonseca [Sun, 11 Mar 2012 15:48:38 +0000 (15:48 +0000)]
Better checking (WIP).

10 years agoMake PIL optional.
José Fonseca [Sun, 11 Mar 2012 14:21:10 +0000 (14:21 +0000)]
Make PIL optional.