Change default PDB chain ID from blank to 'Z'

[drroe]

@dacase Related to ParmEd/ParmEd#1055.

According to the PDB standard, the chain ID should be a "non-blank alphanumeric character". This PR makes it so that for output PDBs, if no chain ID information is present in the topology a default of 'Z' is used. Note that this does not happen when writing PQR files since those do not conform to the PDB standard anyway. The old behavior (outputting a blank character as chain ID) can be obtained via the already existing chainid <char> keyword for PDB trajout, e.g.

trajout mypdb.pdb chainid ' '

This PR also adds a new keyword to the mask command, trajargs, which allows passing a comma-separated list of trajectory arguments to mask when writing out Mol2/PDB files (which allows recovering the old PDBs with blank chain ID for that command).

Updates docs (including fixing areas where text ran off the PDB page) and adds a test.

[drroe]

@swails from Jenkins:

./configure: line 355: pgc++: command not found
script returned exit code 1

Are PGI compilers no longer available or is this a config change thing?

[swails]

run jenkins

[swails]

run jenkins

[swails]

The problem was that the PGI installation was broken in Docker containers on the Jenkins master (it works everywhere else). The reason was a shortcut I took on that node -- I don't have enough space on my root disk to fit all of the compiler suites I want to install in /opt, so I installed them in /iscratch/opt and created a bind mount from /iscratch/opt/pgi to /opt/pgi, which is where it is installed in all of the other slave nodes (and where the Dockerfile tries to mount it from).

But in a docker container that only mounts /opt/pgi, links to /iscratch/opt/pgi all break (which is why PGI can't be found).

For the time being I just removed the PGI label from the master node so Jenkins won't try PGI builds on that node. Once I get the PGI installation fixed on that node I'll add the label back. But for now, Jenkins is green again :)

[drroe]

But for now, Jenkins is green again :)

@swails Thanks Jason!

I'm going to try one more appveyor build. Right now the issue is:

  CPPTRAJ: Ewald test (ortho), 10 frames
../MasterTest.sh: line 480: /dev/stderr: No such file or directory
../MasterTest.sh: line 450: /dev/stderr: No such file or directory
  5 out of 5 executions exited with an error.
  5 out of 5 comparisons failed.
../MasterTest.sh: line 82: /dev/stdout: No such file or directory
../MasterTest.sh: line 83: /dev/stdout: No such file or directory

In lines 82/83 I'm just using the standard POSIX redirects (> and >>) which have always worked on appveyor before. If they're somehow broken I may drop support for appveyor entirely. I'd be interested to know how many people are actually building cpptraj on windows anyway - my guess is a very small fraction.

[hainm]

User can get linux subsystem fro Microsoft store now: https://docs.microsoft.com/en-us/windows/wsl/install-win10

:))