Asa bgp

[thekyr]

Description

The PR is adding on ASA two more parsers:

• BGP

  • show bgp summary
  • show bgp { address_family } unicast summary

• Route

  • show route bgp

Motivation and Context

ASA were missing a parser for the BGP and we needed to create one to be able to run some checks on our monitoring system

Impact (If any)

No negative effect. The code was tested and

Screenshots:

• script to test the collection of show bgp ipv4 unicast summary

• json output of shown bgp ipv4 unicast summary

• json output of show route bgp

Checklist:

• I have updated the changelog.
• I have updated the documentation (If applicable).
• I have added tests to cover my changes (If applicable).
• All new and existing tests passed.
• All new code passed compilation.

[ThomasJRyan]

Hey there! Thank you for the contribution and sorry it's taken so long to get someone on this.

I've left you a few comments to go over. Additionally, all of our parsers need accompanying unittests. You can find examples under the various tests folders in the repo.

You create a folder with the parser class name, create sub equal and empty folders, then add appropriate py and txt files

[ThomasJRyan]

This is good, though in the wrong place. Create a file in https://github.com/CiscoTestAutomation/genieparser/tree/master/changelog/undistributed named asa_bgp_parser_<unix_timestamp>.rst and put this in there

--------------------------------------------------------------------------------
                                      New                                       
--------------------------------------------------------------------------------

* asa
    * Added ShowBgpSummary
      * show bgp summary
      * show bgp {address_family} unicat summary
    * Added ShowRouteBgp
      * show route bgp

[ThomasJRyan]

So, while I definitely see what you're going for here, and it's fair to assume we would be using some sort of standardized schema engine, Genie actually uses its own in-house engine that looks nothing like this.

I would suggest reading over this documentation and revising your schema
https://pubhub.devnetcloud.com/media/pyats-development-guide/docs/writeparser/writeparser.html#writing-the-schema-class

[ThomasJRyan]

Shouldn't need to subclass from ShowBgpSummarySchema since ShowBgpSummarySuperParser has it already

Suggested change

	class ShowBgpSummary(ShowBgpSummarySuperParser, ShowBgpSummarySchema):
	class ShowBgpSummary(ShowBgpSummarySuperParser):

[ThomasJRyan]

I don't think cmd is actually used anywhere here. I'm thinking we can remove it

[ThomasJRyan]

And actually, it looks like address_family gets overwritten on line 237. Unless p1 is never matched when an address family is passed into the command, we can probably remove this argument from the super parser as well

[ThomasJRyan]

This is generally considered bad practice with our parsers. It's typically best to assign the value right as it's found

[ThomasJRyan]

So actually, I get that you're stripping this from the output, but if you write regex that doesn't match it then this can be removed entirely. If we don't match it we don't have to worry about it

[ThomasJRyan]

While I don't necessarily disagree with cleaning up the output before parsing, I can't help but feel this makes it more difficult of a process for both yourself and anyone else who might need to make changes to this parser in the future. It also introduces extra failure points should the output change in any way

[ThomasJRyan]

The regular expressions you have defined here kind of make the entire last 50-ish lines obsolete. They'll never match the other lines, so cleaning them out seems like a waste of effort :/

[ThomasJRyan]

Should still return ret_dict rather than None

[ThomasJRyan]

It's something I probably should have asked myself before I reviewed the rest of this parser, but couldn't you have just added this to the ShowRoute class instead of creating a new one?