pyjq-ng 0.6.4

Tool used for converting jiffies from iptables xt_recent into timestamps.

pyjq

A simple Python package to Query Json Data.

Features

• Supports pure json files
• Supports multiple json objects in a file, delimited by newlines (/n)
• Supports gzipped files
• Supports customizabile filters
• Supports pure datetime range filters

Todo

The filters could be extended easily, adopting Python3 stdlib operator. See pyjq.PyJQ.filter to extend ops mapping.

Installation

pip install pyjq-ng

Example data

See example/alerts.json. pyjq works on lines by lines (splitted by \n). It have been used for Wazuh alert json files and Django dumps.

pyjq -j examples/django_dump.json -limit 2 -filter 'fields__original_url == https://google.com'
pyjq -j examples/django_dump.json -limit 2 -filter 'model == urlshortener.urlshortener'

Usage

'agent__name' it's an example of the namespace used by pyjq to access to nested childs. It other word it means json['agent']['name']. It haven't limits on number of nested elements.

Apply some custom filters with AND and OR operators on Wazuh Alert file

pyjq -j ../Scaricati/alerts.json -filter 'agent__ip == 172.16.16.102 and agent__name == telegram-gw or agent__ip == 172.16.16.108'

Contains operator

pyjq -j ../Scaricati/alerts.json -filter 'rule__description in iptables and agent__name == dev-bastion'

Convert a specified filed to a pure datetime object and filter in a specified range

pyjq -j ../Scaricati/alerts.json -start_datetime 2020-04-06T10:22:00 -end_datetime 2020-04-06T13:22:00 -datetime_field timestamp

Realtime reading, it will only takes the latter entries, delimited by newline \n

pyjq -j /var/ossec/logs/alerts/alerts.json -datetime_field timestamp -realtime

Use a gzipped json file directly

pyjq -j ../Scaricati/alerts.json.gzip

Limit results to 2

pyjq -j ../Scaricati/alerts.json  -limit 2

Realtime monitoring of a specific entity

pyjq -j /var/ossec/logs/alerts/alerts.json -realtime -filter 'agent__name == tinyurl and rule__level == 3'

Author

Giuseppe De Marco [e-mail]

Credits

Wazuh SIEM group @GarrLab