"entropy" has two meanings depending on the computational package

Hi, it’s me again.

Entropies extracted from Gaussian (09C01) or ORCA (4.0.1.2) logfiles have different meanings:

% # First a Gaussian logfile:
% ccget entropy enthalpy freeenergy gaussian_out.out
Attempting to read gaussian_out.out
entropy
0.00016293141036397934
enthalpy
-552.415496
freeenergy
-552.464074

Results from Gaussian assume $G = H - T S$ as usual (entropy means $S$ here):

In [1]: -298.15 * 0.00016293141036397934 + -552.415496                          
Out[1]: -552.464074

% # Now a ORCA logfile:
% ccget entropy enthalpy freeenergy orca_out.out
Attempting to read orca_out.out
entropy
0.04797662
enthalpy
-612.38317299
freeenergy
-612.43114961

But in results from ORCA, entropy means $T S$!

In [2]: -0.04797662 + -612.38317299                                             
Out[2]: -612.43114961

Is this wanted? I can imagine algorithms using ccclib not working across packages because of this.

[Chokerino]

From what i understand this issue involves -

1. Changing self.entropy = float(next(inputfile).split()[4]) to self.entropy = float(next(inputfile).split()[4]/self.temperature) in the orca parser.
2. Adding a comment is that file to show the same. If this needs to be odcument anywhere else please let me know.
3. The Qchem parser has a TODO comment # 'entropy' (incorrect)

[ATenderholt]

I definitely think we should have entropy be the standard definition (i.e. Gaussian, G = H - T * S).