Smirky replaces initial VdW types with generic

[bannanc]

I'm running into a weird issue in the most recent tests with smirky on VdW parameters. I noticed while I was testing for pull request #237

For example, my initial VdW file has the following:

% SMIRKS    TypeName
[*:1]       0
[1:1]       H
[5:1]       B
[6:1]       C
[7:1]       N
[8:1]       O
[9:1]       F
[13:1]      Al
[14:1]      Si
[15:1]      P
[16:1]      S
[17:1]      Cl
[33:1]      As
[34:1]      Se
[35:1]      Br
[53:1]      I

When the initial types are processed they all print as generic VdWs:

INDEX          VDWS  MOLECULES   TYPE NAME: SMIRKS                                           
    1 :          0          0 | 0: [*:1]
    2 :          0          0 | H: [*:1]
    3 :          0          0 | B: [*:1]
    4 :          0          0 | C: [*:1]
    5 :          0          0 | N: [*:1]
    6 :          0          0 | O: [*:1]
    7 :          0          0 | F: [*:1]
    8 :          0          0 | Al: [*:1]
    9 :          0          0 | Si: [*:1]
   10 :          0          0 | P: [*:1]
   11 :          0          0 | S: [*:1]
   12 :          0          0 | Cl: [*:1]
   13 :          0          0 | As: [*:1]
   14 :          0          0 | Se: [*:1]
   15 :          0          0 | Br: [*:1]
   16 :       7185        200 | I: [*:1]
TOTAL :       7185        200

This doesn't happen with Bonds, Angles, or Torsions. I am just now looking into this issue. It doesn't look like it happened in my older smirky tests so I'm not sure what we would have changed that would have caused this.

[bannanc]

Well this is my fault, the initial types are missing the # sign in front of the numbers so when they're read into the ChemicalEnvironments It doesn't catch them atomic numbers.