https://github.com/egberts/nft-token-first

Show what the next tokens are from anywhere in the Netfilter NFT CLI syntax tree
https://github.com/egberts/nft-token-first

bison dhparse ebnf ebnf-syntax netfilter nft nft-app nftables nftables-ui

Last synced: about 2 months ago
JSON representation

Show what the next tokens are from anywhere in the Netfilter NFT CLI syntax tree

• Host: GitHub
• URL: https://github.com/egberts/nft-token-first
• Owner: egberts
• License: mit
• Created: 2025-07-12T22:15:04.000Z (3 months ago)
• Default Branch: master
• Last Pushed: 2025-07-12T23:23:49.000Z (3 months ago)
• Last Synced: 2025-07-13T00:26:55.328Z (3 months ago)
• Topics: bison, dhparse, ebnf, ebnf-syntax, netfilter, nft, nft-app, nftables, nftables-ui
• Language: HTML
• Homepage:
• Size: 542 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          

This is about the Netfilter `nft` CLI terminal app.

Given a token, list all the next valid tokens at the NFT command prompt, given a certain point in the syntax tree.

For example, Bison `input` symbol would produce the following token:

        'include'

        'undefine'

        'error'

        'define'

        'redefine'

        '\n'

        ';'

        'add'

        'table'

        'chain'

        'rule'

        

        'last'

        'ip'

        'ip6'

        'inet'

        'netdev'

        'bridge'

        'arp'

        'accept'

        'drop'

        'continue'

        'jump'

        'goto'

        'return'

        'limit'

        'secmark'

        'synproxy'

        'replace'

        'create'

        'insert'

        'delete'

        'get'

        'list'

        'reset'

        'flush'

        'rename'

        'import'

        'export'

        'monitor'

        'describe'

        'destroy'

dhcparser\_nexus/dhparse-working.py is that starting Python point.

Changing the variable 'vim\_syntax\_group\_name\_target' gets you your different objective.

## WHY ##

I needed to study autocompletion of Netfilter 'nft' CLI tool, which is not much (at the moment).

## HOW ##

In Netfilter NFT CLI parse table, there are over 1,400 edge-state transitions.  It makes zero sense to code that up manually.

I leveraged Bison tool into reading Netfilter nftables' and outputting a rudimentary EBNF (Extended Backus-Naur Form) syntax.

And used a two-stage parser.

## MORE DEETS ##  

I designed a two-stage parser:

   1. how to read EBNF (in DHParse-format)

   2. how to read Netfilter `nft` CLI (in EBNF-format)

For the first-stage, I leveraged [DHParser](https://dhparser.readthedocs.io/en/latest/), a [PEG parser](https://egbert.net/blog/articles/parsing-in-python-compendium.html) to read a parser file.

I used a [EBNF syntax](https://github.com/egberts/nft-token-first/blob/master/dhcparser_nexus/ebnf-flexible.dhparse) file that enables reading of any EBNF file (a file that describes EBNF, ... in EBNF): That file format is for DHParse-only.  

For the second-stage, I shoehorned Netfilter NFT CLI EBNF ... AGAIN ... thru the Python DHParser("EBNF") but this time using Netfilter NFT EBNF (supplied by Bison and [bison\_parse.y](https://git.netfilter.org/nftables/tree/src/parser_bison.y)  source file).

Then I wrote a leaf-tree navigator to look for only the first following token(s) (and not to go down any deeper).

## OTHERS ##

dhparser\_project subdirectory and its development files were made by executing:

    ~/work/github/DHParser/scripts/dhparser.py

    # and selecting option 1 to create a project and entering in `dhparser_project`