Add some tests about range types.

Add some tests about range types.
Add missing tests for context types.
Rearrange some other tests.
Make printing of constant types more useful.
Give counter-examples for properties that no longer hold.

R=rossberg@chromium.org
BUG=

Committed: https://code.google.com/p/v8/source/detail?r=24212

[neis1, 2014-09-24 12:01:01]

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc
File test/cctest/test-types.cc (left):

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ol...
test/cctest/test-types.cc:1779: */
I'm not sure why I thought this wouldn't hold anymore.

[rossberg, 2014-09-24 12:59:09]

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc
File test/cctest/test-types.cc (right):

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ne...
test/cctest/test-types.cc:520: // Subtyping: Class(M1)->Is(Class(M2)) iff M1 =
M2
Hm, I'd rather leave the subtyping tests where they were. Some layering is
useful, and tests like Class and Constant are meant to just test the basic type
constructors, not anything about relations defined on top. (Which was also the
reason why I did not use subtyping in the Equal predicate originally.)

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ne...
test/cctest/test-types.cc:716: // If b is regular numberic bitset, then b->Min()
and b->Max() are integers.
How about introducing a separate test MinMax for testing the general properties
of the Min and Max methods?

Typo: numeric

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ne...
test/cctest/test-types.cc:725: // If b is regular numberic bitset, then
Range(b->Min(), b->Max())->Is(b).
Typo

[neis1, 2014-09-24 15:12:08]

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc
File test/cctest/test-types.cc (right):

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ne...
test/cctest/test-types.cc:520: // Subtyping: Class(M1)->Is(Class(M2)) iff M1 =
M2
On 2014/09/24 12:59:09, rossberg wrote:
> Hm, I'd rather leave the subtyping tests where they were. Some layering is
> useful, and tests like Class and Constant are meant to just test the basic
type
> constructors, not anything about relations defined on top. (Which was also the
> reason why I did not use subtyping in the Equal predicate originally.)

I changed it back now.  But if you look at e.g. Constant(), you'll see that
there already are a couple dozen subtyping tests in there.

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ne...
test/cctest/test-types.cc:716: // If b is regular numberic bitset, then b->Min()
and b->Max() are integers.
On 2014/09/24 12:59:09, rossberg wrote:
> How about introducing a separate test MinMax for testing the general
properties
> of the Min and Max methods?
> 
> Typo: numeric

Done.

https://codereview.chromium.org/602613002/diff/1/test/cctest/test-types.cc#ne...
test/cctest/test-types.cc:725: // If b is regular numberic bitset, then
Range(b->Min(), b->Max())->Is(b).
On 2014/09/24 12:59:09, rossberg wrote:
> Typo

Done.

[rossberg, 2014-09-24 17:26:46]

lgtm

[neis, 2014-09-25 08:31:32]

Committed patchset #2 (id:20001) manually as 24212 (presubmit successful).