GitLab

the file name (not the path) for "top" and "par" to determine file type.  Often
enough, these files are stored in a "toppar" directory, so all of them will be
interpreted as RTF files.

correction maps in their calculations.

patched up by me).

boxVectors directly (which is how you would do it if you set up a PDB with a
CRYST1 record, for instance).

My test with a periodic system now works, but the interface to adding box
vectors should be improved.

Traceback (most recent call last):
  File "/path/to/python/tests/TestCharmmFiles.py", line 91, in test_HydrogenMass
    topology = self.psf_v.topology
  File "/path/to/lib/python2.7/site-packages/simtk/openmm/app/charmmpsffile.py", line 782, in topology
    topology.addAtom(atom.name, elem, residue)
UnboundLocalError: local variable 'elem' referenced before assignment

ERROR: Test that altering the mass of hydrogens works correctly.
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/travis/build/SimTk/openmm/python/tests/TestCharmmFiles.py", line 91, in test_HydrogenMass
    topology = self.psf_v.topology
  File "/usr/local/lib/python2.7/dist-packages/simtk/openmm/app/charmmpsffile.py", line 777, in topology
    masselem = Element.getByMass(mass)
NameError: global name 'Element' is not defined

pickleability.

- Add a docstring to element.Element.getByMass describing what it does
- Add a new element virtualsite with atomic number 0 (and alias it to extrapoint
  and lonepair)

constructor _should_ be able to take all of files in the CHARMM parameter set
distribution and correctly determine file type from just the name and/or
filename extension.

parameters to a ProteinStructure before calling createSystem.

really used CMake in-depth before).

Hopefully this fixes the Travis-CI build failures.