Enable pdb_splitmodel to handel more than 9999 models

[ntxxt]

To solve this #170 issue

[joaomcteixeira]

I personally disagree with this for the same reason expressed by @JoaoRodrigues here:

#170 (comment)

However, I understand what @amjjbonvin argued after. Still, I disagree with this PR as it deviates from the official PDB format. Users can quickly create their branch to apply this change locally if needed.

In any case, if this PR is to be accepted it should be as @JoaoRodrigues said afterwards in the issue and not in the current form of line[6:].

Good point, I'd still keep then the model number starting at position 11 but open-ended. Otherwise you will indeed parse bad files with numbers where they should not be and they might be interpreted wrong by different tools, e.g.

Cheers!

[amjjbonvin]

However, I understand what @amjjbonvin <https://github.com/amjjbonvin> argued after. Still, I disagree with this PR as it deviates from the official PDB format. Users can quickly create their branch to apply this change locally if needed.

I do not agree - this is a simple solution, provide we indeed keep the correct column alignment. But this model number is not very critical. And in addition the PDB only accepts mmCIF format these days.