Avogadro 1.94 cannot read 1.2.0 created PDB file

[MSoegtrop]

I believe this to be a bug with Avogadro

Environment Information

Avogadro version: 1.94
Operating system and version: Windows 10 64 21H1

Expected Behavior

Avogadro can read PDB file created with Avogadro 1.2.0

Actual Behavior

Especially PDB files with crystal structures created with 1.2.0 cannot be read with 1.94.
Error is “Failed to parse chain identifier”.
Other tools can read the PDB files, although they are a bit buggy - say use the same identifier for many atoms.

Steps to Reproduce

E.g. in 1.2.0

• File / Import / Crystal C -graphite
• Build / Supercell / 2x2x1
• Save as PDB

The resul cannot be read in 1.94.

[MSoegtrop]

After some more tests I would say that Avogradro 1.94 cannot read any PDB file generated with Avogadro 1.2.0.

[ghutchis]

Huh. Can you upload the file and send me the link?

[MSoegtrop]

Here is one example - it is created in 1.2.0 with the 3 steps I gave in the initial post:

graph_cell_2x2.pdb

[ghutchis]

This patch should fixit. When this finishes building in ~2 hours, it should have a Win64 binary to test added as a comment on GitHub.

If a chain ID is missing / invalid, skip (to read non-standard PDB)

[ghutchis]

e.g. https://github.com/OpenChemistry/avogadrolibs/suites/3154880598/artifacts/72554111
(valid for ~90 days … 1.95 should certainly be released sooner. )

[MSoegtrop]

I can confirm that this works. It now asks which reader to choose (Avogadro vs OpenBabel) which I guess is expected.

[ghutchis]

Yes, at the moment, it will ask if you have a preference for different readers.