What does a Genie parser do, and why is it better than parsing CLI output yourself?

device.parse("show …") runs a show command and returns its output as a structured dictionary. You don't write regexes, and you get one data structure across device types and OS versions.

* The parser turns human-oriented CLI text into a dictionary, so a check is a lookup instead of a regex. *

Raw CLI output is text meant for humans. To check it in a script you would normally write regular expressions, which break as soon as a column shifts or a new OS version changes the wording. Genie ships parsers for thousands of show commands and does that work for you:

parsed = device.parse("show ip interface brief")
for intf, d in parsed.items():                    # structured, not text
    if d["status"] == "administratively down":
        print(f"{intf} is shut down")

bgp = device.parse("show ip bgp summary")
for nbr, info in bgp["vrf"]["default"]["neighbor"].items():
    if info["state"] != "Established":
        print(f"BGP neighbour {nbr} not established")

This makes targeted checks for specific values, states or patterns easy, for example:

• is an interface administratively down?
• is a BGP neighbour not established?
• is a VLAN missing ("100" not in device.parse("show vlan brief")["vlans"])?
• is CPU load too high (five_sec_cpu_total > 80)?

Advantages:

• no manual regex parsing,
• one data structure for different devices and OS versions,
• ideal for automated tests, validations and audits, for example running these checks before and after every change.

Go deeper:

• Genie documentation — Parsers — how to call parsers and use their output.
• Genie feature browser — parser list — search which show commands have a parser per OS.