Improvements to get_mol_kegg()

[Aariq]

• Allow usage of pathway module IDs (Allow get_mol_kegg() to work with pathway modules #82)
• Skip download if .mol file is already present (Don't re-download files with get_mol_kegg() #83)
• Improve efficiency of download by getting up to 10 compounds in one API call. (Improved download of 21 compounds from 10.4 seconds to 2.6 seconds)

The last of the three added purrr and glue as dependencies. I don't think this is a problem, but I'd be happy to try to find some base R way of doing this.