Closed zjing7 closed 3 years ago
Yes, I have very recently changed the parsing of the .key file and the .prm file. The new version worked for everything I tested. Can you post a sample keyfile that fails, and I'll fix it. It is almost certainly something very simple to fix. Thanks.
tinker.key
parameters none
atom 501 501 C "C_1" 6 12.011 3
multipole 501 0 0 0.00000
0.00000 0.00000 -0.00000
0.00000
0.00000 -0.00000
-0.00000 0.00000 -0.00000
sample.bar
5 298.00
1 -120258.6086 -120258.7365 388793.2249
2 -120281.9093 -120282.2081 388793.2249
3 -120574.6366 -120574.7243 388793.2249
4 -120726.4866 -120726.5753 388793.2249
5 -120717.6796 -120717.9625 388793.2249
5 298.00
1 -120688.3353 -120688.4099 388793.2249
2 -120559.9875 -120560.2053 388793.2249
3 -120716.9108 -120716.9504 388793.2249
4 -120796.4867 -120796.5996 388793.2249
5 -120408.1856 -120408.4933 388793.2249
I believe this is fixed now. Sorry for the silly problem.. It didn't show up on my Mac development machine since the Mac generally pads array overflows, and hides lots of problems like this. Linux is much better for testing!
The only changes are to single lines in getkey.f and getprm.f
Let me know if you still have trouble.
I am getting an error with the latest version when the key file contains multipoles (or many spaces followed by a number). This happens even when running programs like bar.x that doesn't require a key file. This seems to be related to the latest commit.
At line 107 of file getkey.f Fortran runtime error: Substring out of bounds: upper bound (21) of 'keyword' exceeds string length (20)