pdb_splitmodel can only handel up to 9999 number of models

[ntxxt]

[JoaoRodrigues]

Hey @ntxxt , the PDB format specifies that the MODEL record cannot have more than 4 characters (11-14), so supporting this is supporting non-standard structures. Also, the PDB format is really not what you'd want for storing more than 10k models. For these two reasons, I'd be opposed to merging this, although I understand how annoying this might be.

[amjjbonvin]

As it is not generating PDB files, allowing more than 9999 models should not really be an issue here. If we were to also modify `pdb_mkensemble` then this would violate the official format.

[JoaoRodrigues]

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.

MODEL    10  # correct, read model 10
MODEL   10   # incorrect, read model 0

[amjjbonvin]

We are running again into this issue, not limiting haddock3 to ensembles of 9999 models max. We should fix it.
Forget about the "official" format. PDB format is anyway kind of outdated.

[amjjbonvin]

Note that the current pull request about this one also has a problem in case there would be another field after the model number.

[amjjbonvin]

@ntxxt - please check this one and if ready create a pull request

[ntxxt]

A pull request is created.

[rvhonorato]

done in #180