Add scripts to read go database

[vandan-revanur]

This PR adds a functionality to read geometry databases such as the one available here: GW5000

[ndattani]

I would like to make some minor changes to both of these files before merging it into the HPQC Main branch, then making a pull request into the PySCF main branch. I'll try to do these today when I get on a computer!

[ndattani]

When trying this code, I got an error, leading me to run pip install requests before running the program. Is it impossible to get this working without this dependency? The fewer dependencies, the easier it is for beginners to get started without seeing error messages :)

[vandan-revanur]

We can get it to work, by downloading the database.txt file and putting the path to the file in the script.
However, I feel it is ugly and everyone would have to edit the script to put their respective paths.
The current approach is cleaner because people can just run the script without making any changes to the script.

However, if you prefer the file approach I can fix that.

[ndattani]

When you clone PySCF, you get a database of basis sets (STO-3G, 6-31G, cc-pVDZ, def2-TZVPP, etc.) for each element in the periodic table, downloaded. Then PySCF can just read that database file directly. So the part about this being "ugly" because of needing to edit the script with a personalized "path", is not necessary.

The disadvantage of having the database shipped with PySCF is that it's huge. I won't try to include a database of this size in PySCF without prior approval (and perhaps some studies on what would be the best compression/decompression method, if any).

So for now, the URL approach is best. It is just so unfortunate that we need to import another package for this!

[ndattani]

I've changed go_db.py to get_xyz.py. Is it possible to remove this import statement entirely? I'd prefer to minimize the number of imports in the example file, and to minimize the number of total lines. Users can choose to add this line if they want to simplify the code later in the file, but ideally an "example" file would be supremely simple, like this one: https://github.com/pyscf/pyscf/blob/master/examples/scf/00-simple_hf.py

[vandan-revanur]

I think this change will make it confusing for people, especially because the module and the function are called the same name. If we would need to remove the import statement then we need to put the function in the same script.
Is that something that you would like? If yes, it would contradict your other statement that we need to minimize number of lines in total.

The example at the moment is more compact in terms of number of lines when compared to having the function in the script. Let me know how you would like to tackle this.