Feature Suggestion: cclib-based taskdocs

[materialsproject]

Feature Suggestion: cclib-based taskdocs

[utf]

I think this is a great idea. I'd be very happy if you added this in a PR.

I think the base cclib document could also:

1. Subclass the molecule meta task document you made (so that it has keys such as molecule and point group).
2. Contain a dir_name key that points to the calculation directory.

[Andrew-S-Rosen]

Great, I'll give that a go! That's precisely what I was thinking as well. Should be easy to set up.

[Andrew-S-Rosen]

Regarding Point 2, it'd probably have to be a logfile_path kind of key because cclib reads specific log files rather than a directory (most, if not all, of the molecular DFT codes have just one main log file to read from). It detects the code on-the-fly, but it would need to know the file path. It's possible to rely solely on a dir_name key but that would require us to make some decisions, like "search for all files with .out, .chk, .log extensions in dir_name and parse the one with the matching extension and the newest one if there are multiple". While that's super convenient (and admittedly what I have been doing locally...), there's in principle nothing stopping me from renaming my log file .loggylog for fun.

[utf]

Fixed in #64

Thanks!