Curly braces not removed from the doi field

[deinal]

Issue Description

The current BibTeX parser fails to remove curly braces { and } from the doi field in the parsed output.

E.g. entry DCTRGAN: Improving the Precision of Generative Models with Reweighting gets the following doi address https://doi.org/%7B10.1088/1748-0221/15/11/p11004 because it has doi with curly braces in HEPML.bib. The %7 character suggests that the first curly brace is not removed.

For entry Jet Energy Calibration with Deep Learning as a Kubeflow Pipeline both opening and closing curly braces are left in the url produced by make_md.py: https://www.doi.org/%7B10.1007/s41781-023-00103-y%7D. The difference as we can see from here is that this entry doesn't have a trailing comma at the end of the line like DCTRGAN does.

Suggested Fix

I believe a fix could be introduced here in the parser code to remove the curly braces from the doi field.

if "{" in myentry_dict["doi"]:
    myentry_dict["doi"] = myentry_dict["doi"][1:]
if "}" in myentry_dict["doi"]:
    myentry_dict["doi"] = myentry_dict["doi"][:-1]

A similar thing is done for the opening bracket in myentry_dict["eprint"] a few lines earlier.