Merge branch 'issue/maxima_script' into 'master'

### Summary

Resolves a bug in gausslobatto.mac script where coefficients are generated
only with double precision instead of high-precision.

Closes [#18]

See merge request [!104]

  [#18]: gitlab.dune-project.org/core/dune-geometry/issues/18
  [!104]: gitlab.dune-project.org/core/dune-geometry/merge_requests/104


Closes #18

issue related to generation of gauss-lobatto quadrature rule from maxima script resolved. Closes #18

Allow `ReferenceElement::size(i,c,cc)` for `cc<c`

See merge request !93

See merge request !101

do not catch polymorphic type by value

See merge request !99

This silences a warning from GCC 8:

    dune/geometry/virtualrefinement.cc:496:5: warning:
    catching polymorphic type ‘class std::bad_cast’ by value [-Wcatch-value=]

[c++] Remove bind2nd

See merge request !97

std::bind2nd has been removed in C++17, and libc++ as shipped in LLVM 6 does not contain it anymore,
causing compile errors.

Instead of replacing it with std::bind, this fix removes it entirely and goes down the much simpler
path of using a generic lambda.

[test-nontype] Ignore warning when testing deprecated functionality.

See merge request !96

doc/refinement/*: remove executable bit from file permissions [ci skip]

See merge request !95

While subEntity(i,c,j,cc) is surely not allowed for
cc<c because there's no viable j, there's no reason
to not allow this case for size(i,c,cc). Actually the
implementation works nicely in this case and returns
the correct number size zero. Any grid-specific
ReferenceElement implementation should simply return
zero in this case.

Allowing this case gives the possibility to seamlessly test
for(int j=0; j < re.size(i,c,cc); ++j)
for all 0<=c<=dim and 0<=cc<=dim.

[cleanup] Canonicalize SFINAE in referenceElements()

See merge request !91

[doc] Fix documentation of ReferenceElement::subEntity

See merge request !92

The codimension of the sub-subentity wrt to the ReferenceElement
was not correct.

The used version of SFINAE by enable_if has the advantage
of not adding additional template-parameters. However, it
has the disadvantage of making the generated doxygen documentation
hard to read.

This commit changes this to SFINAE/enable_if by defaulted
int template parameter. This additional parameter also
shows up in doxygen but its much easier to identify as
SFINAE dummy because it follows a simple idomatic pattern.
More importantly it does not mess up the documentation
for the other parameters.

[GeometryType] Clean up storage

See merge request !86

Steffen Müthing authored 7 years ago and Martin Nolte committed 7 years ago

This patch removes the packed bitfield used to store both the dimension and the none flag in a
single byte. This optimization does not change anything about the size of GeometryType as a whole,
because it would be 5 bytes even with the optimization and the C++ ABI (at least on 64-bit
platforms) automatically rounds that up to 8 bytes, so we can just as easily let both the flag and
the dimension have a full byte of their own.

This also improves the code generated by both GCC and Clang quite a bit. In order to further improve
matters, I reordered the member variables and put the topologyId_ variable last.

Revert deprecation of GeometryType::BasicType for 2.6

See merge request !88

As there are still some use cases that cannot be easily switched from using GeometryType::Basictype
to plain GeometryType, revert its deprecation for the 2.6 release.

This does **not** revert the deprecation of the GeometryType constructor from BasicType.

Enforce use of correct floating point type in AxisAlignedCubeGeometry

See merge request !85

used dune macros to suppress deprecated warnings

See merge request !84

YaspGrid can use non-double floating point type, but AxisAlignedCubeGeometry
sneaked in some doubles...

[cmake] Require CMake 3.1 similar to dune-common

See merge request !82

[bugfix][test] return 1 on failure

See merge request !81

add changelogs from release to master

See merge request !80

added empty changelog

See merge request !77

[bugfix] `TopologySingletonFactory` not uses `std::unqiue_ptr` internally

See merge request !74

overhaul `TopologySingltonFactory`

See merge request !72

This patch modernizes the code for the `TopologySingletonFactory`. The
main improvement is using `std::unique_ptr` to get rid of `fill_array`,
which is going to be deprecated in the 2.6 release.

bump version to 2.7-git [ci skip]

See merge request !71