Cctk

Major changes:
- Integration with [``xtb``](https://xtb-docs.readthedocs.io/en/latest/contents.html) and [``crest``](https://xtb-docs.readthedocs.io/en/latest/crest.html) has now been added in ``cctk/optimize.py``. Calling ``molecule.optimize()`` will optimize the molecule at the GFN2-xtb level of theory, while calling ``molecule.csearch()`` will run a full conformational search and return a ``ConformationalEnsemble`` (this can be slow). These external packages must be accessible by ``bash``, or ``cctk`` will throw an error.

Minor changes:
- A bug pertaining to reading Gaussian output files for molecules with only one atom has been fixed.
- ``ConformationalEnsemble.eliminate_redundant()`` now returns the RMSD values for each conformer.

Minor hotfix relating to frequency parsing.

Major changes:

- ``cctk.quasiclassical`` now permits generation of thermally-reasonable perturbations to molecules with defined vibrational modes (based on quantum harmonic oscillator approximation). Concomitantly, a new ``VibrationalMode`` object has been defined, which can automatically be read from ``freq`` output files (Gaussian only).
- ``Molecule`` objects can now be initiated from name or SMILES if ``rdkit`` is installed.
- ``molecule.volume()`` is much slower and much more accurate. The old *qhull*-based method can be requested if accuracy is unimportant.
- Symmetric atoms can now be detected from methyl, isopropyl, and tert-butyl groups using ``molecule.get_symmetric_atoms()``.
- ``topology.find_group()`` can find a group within a given molecule.

Minor changes:

- ``.xyz`` trajectories can now be read.
- ``cctk.Molecule.are_isomorphic()`` checks if two molecules have the same connectivity, which is handy.
- NMR shifts can now be read from Orca.

Major changes:

- Significant under-the-hood optimization of ``GaussianFile.read_file()``, ``molecule.assign_connectivity()``, and ``ensemble.eliminate_redundant`` (3-10x improvement). Backend code is more complex, especially for ``read_file``, but behavior should be identical (there will probably be unforeseen consequences so email me if anything breaks).
- ``analyze.py`` now uses multiprocessing for a net 100x improvement in speed. This strategy should be broadly applicable for file I/O.
- Basis sets can now be loaded from the Basis Set Exchange (Gaussian only).
- J-couplings can now be read from NMR jobs in Gaussian.
- Periodic boundary conditions can be supported in ``molecule.assign_connectivity``

Minor changes:

- 2-letter elemental symbols are now parsed properly in ``.pdb`` files.
- Writing a Gaussian file without ``p`` in the route card now prints a warning instead of raising a job-ending ``ValueError``.


v0.2.3-hotfix
A variety of minor errors which emerged after last update (for both Gaussian and Orca files) have been fixed.

pypi`` does not permit "hotfix" in releases to upload.

This update permits reading of Orca ``.out`` files and automated extraction of properties in much the same way Gaussian ``.out`` files can be read. The properties that can be extracted are fairly basic right now -- energies, frequencies, etc.

Updated ``analyze_orca.py`` and ``monitor_orca.py`` scripts have been added to the ``scripts/`` folder.