Lammps: fix species for atomstyle charge

[jan-janssen]

@Leimeroth I gave it a first try, but unfortunately I do not have an example to test this. Can you provide a test for this function?

[coveralls]

Pull Request Test Coverage Report for Build 8384624071

Details

• 12 of 12 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.002%) to 93.176%

---

Totals
Change from base Build 8345641982:	0.002%
Covered Lines:	14214
Relevant Lines:	15255

---

💛 - Coveralls

[Leimeroth]

This correctly orders the elements which are actually present in the structure. However, when using atom_style charge no species can be set in the potential, which are not actually present in the structure, because the wrong number of atoms is written to structure.inp than:

Start File for LAMMPS
9 atoms
2 atom types

0. 4.921000003800000 xlo xhi
0. 4.261711015300000 ylo yhi
0. 5.400000095400000 zlo zhi
2.460500001800000 0E-15 0E-15 xy xz yz

Masses

  1 28.085000
  2 15.999000
  3 1.000000

Also the mass of C is not correct. Without atom_style charge it sets the correct N atom types and masses and than just never assigns the type. This brings me back to the question why there are so many different structure writing functions that all use different ways of ordering? I think extracting the ordering functionality from the well tested atom style atomic writer into a separate function that is used by all writes would be the best option

[Leimeroth]

I drafted a version refactoring the sorting in #1361. This is only working for charge and atom styles right now, because I am not sure how it works for bond and full styles.

[jan-janssen]

Also the mass of C is not correct.

Atoms which are not used in the simulation cell but defined in the potential are set to 1.0. In this case there are no C atoms in the simulation cell, correct?