Also reactivate operator& for minimal test classes. Apparently I
disabled them because of a problem in a type trait, but I'm not seeing
that now. Maybe it will appear on other compilers.
There used to be a std::move test here, but it was changed to
boost::test for a wider range of testing, but I'd quite like to test
with std::move regardless.
I'm hoping that these templated initializer lists will be considered a better
overload than the others. I have no idea if it will actually work, this is a
real shot in the dark.
The enable_if checks should probably be for implicit conversion, there might
be a chance this could override a valid call when there's an explicit
conversion.
[SVN r86419]
It looks the odd result in unnecessary_copy_tests on Visual C++ 11 is not a
bug, but add some extra tests just to make sure. Also some extra rehash and
reserve testing thrown in.
[SVN r80705]
Assigning a container requires that its elements can be assignable. Could split
the tests up so that other tests aren't assignable, but it doesn't seem worth
the hassle.
[SVN r80228]
In case there are elements in buckets that aren't in the element list. Required
because the data structure has changed since the original implementation.
[SVN r80226]
Mainly because I want to get more info on the odd test failure for Visual C++
11. I expect that with move emulation these results could vary considerably,
and since I've only tested with gcc and clang so far, it's quite likely that
this test will now fail for other compilers.
[SVN r80200]
Getting a memory deallocation error from Sandia's linux c++11 clang. Add some
trace output to see if it gives a clue where it's going wrong.
[SVN r79793]
Compilers with rvalue references can avoid creating a node, as they can use the
value from the rvalue reference to check if the value is already in the
container (in this case it is) before creating the node. Could possibly do the
same for compilers without rvalue references, if it can get a value out of
Boost.Move's rvalue reference emulation.
[SVN r79503]
Some of this was there for older compilers, some is just premature
generalization. There's still too much metaprogramming, but these are things
that are relatively easy to remove.
[SVN r79356]
It was originally introduced because of some issues with Boost.Test and older
compilers, neither of which I'm using now. Simplifies a few things.
[SVN r79352]
The propagate on move tests were failing because the values were being
copied not moved - so the container was check propagate on copy instead.
Because those tests are now weaker, and some new ones that will
hopefully work.
[SVN r74914]
I'm getting a warning from another library. I don't really need to check
debug libs anymore - that was for container_fwd which is now tested in
details.
[SVN r74840]
Helps allocators which can't use incomplete pointers, and avoid using
base pointers where that might not be possible. And some other
reorganization. Storing arguments to emplace in a structure when
variadic template parameters aren't available. Changed some of the odd
design for working with older compilers.
[SVN r74742]
Will need to deprecate the old variadic style pair construction, also
should look into extract_key, was written for compatibility with older
compilers that are no longer supported.
[SVN r74119]
Only for compilers with SFINAE expressions and recent versions of Visual
C++. Also fix Visual C++ 8, and use BOOST_UNORDERED_ prefix for all
macros.
[SVN r74106]
- Simplify mechanism for detecting traits of test allocators. There were
some portability issues, but rather than fix them I've just gone for a
simpler mechanism. Does mean that the relevant tests can't be run for
other allocators.
- Fix a couple of unnecessary_copy_tests, whose results were the wrong
way round.
- It appears that Visual C++ only implements RVO for implicitly defined
copy constructors in debug mode, so adjust a move_test to account for
the extra copies now that the copy constructors are explicitly
defined.
[SVN r73798]
It's pretty messy because I'm trying to avoid swapping allocators in
these cases. I'm also not sure of the exception requirements of
allocator swap and assignment.
[SVN r73756]
Seems to be better in some cases. For example, better forwarding if
rvalue references are supported but variadic template parameters aren't.
Also can use `boost::move` when inserting.
Older versions of gcc still fail this test. They perform more moves than
expected. It might be a limitation of Boost.Move, or maybe just poor
copy optimization.
[SVN r73539]
Store nodes in a single linked list, with hash values so that their
buckets can be found when needed. Iterators now only have to store a
pointer to the node and don't have to iterate over empty buckets to
reach the next node. This allows the container to meet the iterator
requirements - fixing the speed issues with `equal_range` and `erase`.
Also, define iterators in their own namespace, so that they don't
accidentally pull in detail functions via ADL.
I've simplified the code slightly by removing some of the special
cases for empty containers. Renamed a few things as well and other
minor changes that were made as I went along.
[SVN r71327]
In the draft standard the std::pair constructor from another pair is
only considered when overloading if both members are implicitly
constructible to their corresponding members. This breaks one of the
unordered tests which required an explicit conversion, so change it to
only require an implicit conversion.
[SVN r70028]
Borland is having some issues with the existing tests, since they call
inline functions before they're defined. It might be right to, although
all the other compilers are fine with it. But the test isn't really what
the forward headers are intended for. Try to make the tests better
represent the intent, and possibly work on Borland.
[SVN r66428]
It seems that the problem is calling sizeof on a dependent type when the
containers have only been used by reference. So by putting in these
dummy structures with the containers as members, it helps the compiler
instantiate the class to the level where sizeof works. I hope.
[SVN r58130]
- More template use, less preprocessor use.
- Removed some of the Visual C++ 6 workarounds.
- Reduced memory use of the main object.
- Split into smaller headers.
[SVN r55878]
Merged revisions 46470-46592 via svnmerge from
https://svn.boost.org/svn/boost/branches/unordered/trunk
................
r46589 | danieljames | 2008-06-21 21:37:42 +0100 (Sat, 21 Jun 2008) | 2 lines
Fix some inspect errors (tabs and missing copyright/license).
................
r46591 | danieljames | 2008-06-21 21:47:51 +0100 (Sat, 21 Jun 2008) | 1 line
Move memory.hpp into the helpers subdirectory.
................
r46592 | danieljames | 2008-06-21 22:08:53 +0100 (Sat, 21 Jun 2008) | 1 line
Prevent inspect errors for unnamed namespaces in some of the test header files.
................
[SVN r46594]
the member classes.
With some standard libraries I was getting two copies of the object after
creating a default pair, probably because it was creating an instance for its
default parameter. So only test after creating the pair object - since it isn't
our concern how many instances that creates.
[SVN r46587]
tweaks.
Merged revisions 45747-46030 via svnmerge from
https://svn.boost.org/svn/boost/branches/unordered/trunk
................
r45747 | danieljames | 2008-05-25 18:03:30 +0100 (Sun, 25 May 2008) | 1 line
Remove /branches/unordered/dev from svnmerge tracking.
................
r45874 | danieljames | 2008-05-28 18:54:53 +0100 (Wed, 28 May 2008) | 1 line
Merge from trunk.
................
r45881 | danieljames | 2008-05-29 00:10:37 +0100 (Thu, 29 May 2008) | 2 lines
Make it possible to use a custom generator in more places.
................
r45882 | danieljames | 2008-05-29 00:10:56 +0100 (Thu, 29 May 2008) | 2 lines
Remove several unused parameters.
................
r45883 | danieljames | 2008-05-29 00:11:06 +0100 (Thu, 29 May 2008) | 2 lines
No need to run move_construct_tests1 from move_construct_tests2, already calling it directly.
................
r45884 | danieljames | 2008-05-29 00:11:15 +0100 (Thu, 29 May 2008) | 2 lines
Remove unused 'construct' as in trunk.
................
r45885 | danieljames | 2008-05-29 00:11:25 +0100 (Thu, 29 May 2008) | 2 lines
More consistent implementations of insert_impl.
................
r45886 | danieljames | 2008-05-29 00:11:34 +0100 (Thu, 29 May 2008) | 2 lines
No need to use the single argument version of methods when the variadic version is available. Frees up the non-variadic version for overloading to emulate variadic methods.
................
r45887 | danieljames | 2008-05-29 00:11:44 +0100 (Thu, 29 May 2008) | 2 lines
Add another test, mainly to avoid an unused parameter warning.
................
r45888 | danieljames | 2008-05-29 00:11:54 +0100 (Thu, 29 May 2008) | 2 lines
Adjust the pixels per inch of the buckets diagram so it'll fit on the page.
................
r45889 | danieljames | 2008-05-29 00:15:49 +0100 (Thu, 29 May 2008) | 1 line
Improve pdf output, thanks to John Maddock.
................
r45890 | danieljames | 2008-05-29 00:17:45 +0100 (Thu, 29 May 2008) | 1 line
Add an svg version of the bucket diagram.
................
r45891 | danieljames | 2008-05-29 00:18:11 +0100 (Thu, 29 May 2008) | 1 line
Remove the original dia buckets diagram, I'm not using it anymore.
................
r46025 | danieljames | 2008-06-01 18:39:51 +0100 (Sun, 01 Jun 2008) | 1 line
Initialise merging from doc.
................
r46030 | danieljames | 2008-06-01 18:54:36 +0100 (Sun, 01 Jun 2008) | 29 lines
Merge in documentation improvements from the doc branch.
Merged revisions 45892-46020 via svnmerge from
https://svn.boost.org/svn/boost/branches/doc
........
r45892 | danieljames | 2008-05-29 00:30:07 +0100 (Thu, 29 May 2008) | 1 line
Merge latest changes from unordered.
........
r46017 | danieljames | 2008-06-01 18:18:15 +0100 (Sun, 01 Jun 2008) | 2 lines
Use both PNG and SVG versions of the buckets diagram.
........
r46018 | danieljames | 2008-06-01 18:18:27 +0100 (Sun, 01 Jun 2008) | 2 lines
Improved function summary tables in unordered documenations.
........
r46019 | danieljames | 2008-06-01 18:18:39 +0100 (Sun, 01 Jun 2008) | 3 lines
I'd hacked the bibliography to avoid showing the title twice, but this was
causing problems when generating PDFs, so show the title twice.
........
r46020 | danieljames | 2008-06-01 18:18:52 +0100 (Sun, 01 Jun 2008) | 3 lines
Don't use the compact boostbook refernce style in the standalone documentation
as it is too wide for printing out.
........
................
[SVN r46031]
