Silences a Sphinx warning.

Avoid confusion, as the editors don't come with context managers.

The breadcrumbs now correctly indicate the module hierarchy and can be
used to determine imports. All the documentation for ost.* is now rooted
on the base `ost` module for more clarity.

ost.mol.alg submodules are now documented in separate pages. This should
make it clearer what to import.

The trace directionality for the internal coordinate system was
established in a recursive fashion. This likely caused filling up
stack memory for very large polymers with large recursive depth.
The fix implements the same functionality in a non-recursive
fashion.

Numpy support in C++ enabled features at two points:
Efficient way of setting positions in entity handles or setting meshes
in gfx.PrimList.

Support for newer numpy versions would've required an effort to implement.
As far as I know, no one uses these features (fingers crossed) so let's just
dump numpy alltogether to simplify the build system.

Removing atoms from huge structures was observed to be super slow.
Reason was the removal from the spatial organizer which helps to find atoms
based on spatial proximity. The organizer is organized in buckets which
occupy a certain volume in space. Removing an element meant to iterate over
all buckets and all their items until the atom is found. However, we know
the position of the atom and thus can pinpoint the bucket in which its
expected to be. The SpatialOrganizer::Remove was therefore overloaded with
a version that accepts a position as hint in which bucket to look.
If the atom is there, delete and return. If not, call Remove without position
hint.

Reduces insane amount of compiler warnings. Inheritance from std::iterator
template was used to define traits of an iterator. Cool people define these
traits directly nowadays.

This allows the mmCIF writer to save the entity_poly.type as "other"
rater than error on an unknown type.
We can do this because there is no "other" in _entity_poly.type.
Added a comment about the risks of chain_type for reading/writing mmCIF
files.

Assignment of ChemClass values happens inside the ChemClass definition
in the header file. Let's say you're using something like:
ChemClass::WATER. If this is resolved at compile time, you won't have
any problem. If not, the linker searches for it in the mol shared object
file. BUT ITS NOT THERE. The hope is that defining the types as an
enum inside the ChemClass definition helps.

Changes in chem_class.hh are a temporary solution for optimization oddities.
constexpr enforces evaluation at compile time. The observed problem were odd linking
problems (undefined reference) to things like ost::mol::ChemClass::XYZ.

This unit test has been observed to fail (EXTREMELY RARELY)
It tests the edge case of 3 atoms that are all on a line.
From an external coordinate system perspective, the dihedral
is not defined in this case.
From an internal coordinate perspective this should not matter.
The last line is the failing line. The expected behaviour is
that a rotation of x around t1 (The "linear" dihedral) should
propagate and affect t2 accordingly. Observed failure:
difference{4} between s.t2.GetAngle(){-2.3561945} and Real(3.14159265358979323846/4){0.785398185} exceeds 0.0001%

In the meantime we propose to use internal coordinates with caution and
open a task in JIRA for someone brave enough to deep dive into the code.

This reverts commit 9153648f.

The intention of this commit was to fix potential issues when operating on a
dihedral connecting 4 atoms where at least 3 atoms are exactly on a line.
Dihedrals are ill defined in this case but "expected behaviour"
(whatever that means) is tested as a unit test.
A fail was observed in two cases. First case that triggered commit
9153648f, second case that showed that the
problem is not solved. Therefore reverting to original state in order to find
different solution.

This allows IsValid() to work as intended without throwing an exception.
Because the validity of the underlying entity is now checked in CheckValidity,
it should no longer be a problem.

In both cases we check if we have a valid implementation AND if the
underlying entity is still there.

This checks whether the underlying handles are still valid before
returning pointers to them. Runs the check each time CheckValidity
is called on atom, residue and chain handles.

SetElement sets default properties for mass and radius. Previously
set charge and anisou got wiped in the process. Charge and anisou
are now transferred and only mass and radius are reset.

Following fixes in cd701347, setting seqres=True is no longer required
for these things to work.