Skip to main content
Join the official 2020 Python Developers SurveyStart the survey!

Python package for finding and parsing indicators of compromise from text.

Project description

Observable Finder (Formerly named "IOC Finder")

PyPi PyPI - Downloads Codacy Badge Build Status codecov live demo

Parse indicators of compromise from text. You can test this project here: http://ioc-finder.hightower.space/.

Capabilities

Currently, this package can the following items in a given text:

  • Autonomous System Numbers (ASNs) (in multiple formats such as asn1234 and as 1234)
  • Bitcoin addresses (P2PKH, P2SH, and Bech32)
  • CIDR ranges (currently ipv4 ranges; ipv6 ranges coming soon)
  • CVEs (e.g. CVE-2014-1234)
  • Domain names (support for Unicode domain names (e.g. ȩxample.com) is coming soon)
  • Email addresses (both standard format (e.g. test@example.com) and an email with an IP address as the domain (e.g. test@[192.168.0.1]))
  • File hashes (md5, sha1, sha256, sha512, and import hashes, and authentihashes)
  • File paths (beta)
  • Google Adsense Publisher IDs
  • Google Analytics Tracker IDs
  • IP address (IPv4 and IPv6)
  • MAC addresses (beta)
  • Phone numbers (beta)
  • Registry key paths (e.g. "HKEY_LOCAL_MACHINE\Software\Microsoft\Windows)
  • SSDeep hashes (beta)
  • URLs (URLs with and without schemes)
  • User agents (beta)
  • XMPP addresses (basically, this captures email addresses whose domain names contain "jabber" or "xmpp")
  • MITRE ATT&CK data (see more info):
  • TLP labels
  • Others... if you have any requests, let me know (or you can contact me here to make private suggestions)!

Also provides some helpful features like:

  • Option to parse domain name from a URL
  • Option to parse domain name from an email address
  • Option to parse IP address from a CIDR range
  • Option to parse URLs without a scheme (e.g. without https://)
  • Option to parse import hashes and authentihashes

Known Limitations

  • When parsing registry key paths, this library will NOT properly parse a registry key path where the last section contains a space. For example, <HKCU>\software\microsoft\windows\currentversion\explorer\advanced on will be parsed as <HKCU>\software\microsoft\windows\currentversion\explorer\advanced (the space in the final section is removed).
  • The items listed above (in the "Capabilities" section) that are postceded by "(beta)" are not very robust and may still have major issues. Any feedback or issues related to these items are much appreciated.
  • When parsing markdown, if there is a domain name that is surrounded by underscores (which would make the domain name italic in some flavours of markdown - e.g. _google.com_), the domain will be parsed including the leading underscore (e.g. _google.com_ would be parsed as _google.com).

Python2 Support

All versions of the IOC Finder package before version 2.x are compatible with python2.7 . Beware: version 1.x of the IOC Finder package is no longer maintained. This means that any bug fixes or improvements will not be back-ported to previous versions.

To install a specific version of the IOC Finder package (or any other one) via pip, use the following formula:

pip install ioc-finder==<VERSION NUMBER>

For example:

pip install ioc-finder==1.2.18

Installation

To install this package:

pip install ioc-finder

Usage

This package can be used in python or via a command-line interface.

Python

The primary function in this package is the ioc_finder.find_iocs() function. A simple usage looks like:

from ioc_finder import find_iocs
text = "This is just an example.com https://example.org/test/bingo.php"
iocs = find_iocs(text)
print('Domains: {}'.format(iocs['domains']))
print('URLs: {}'.format(iocs['urls']))

Inputs

You must pass some text into the find_iocs() function as string (the iocs will be parsed from this text). You can also provide the options detailed below.

Options

The find_iocs takes the following keywords (all of them default to True):

  • parse_domain_from_url (default=True): Whether or not to parse domain names from URLs (e.g. example.com from https://example.com/test)
  • parse_from_url_path (default=True): Whether or not to parse observables from URL paths (e.g. 2f3ec0e4998909bb0efab13c82d30708ca9f88679e42b75ef13ea0466951d862 from https://www.virustotal.com/gui/file/2f3ec0e4998909bb0efab13c82d30708ca9f88679e42b75ef13ea0466951d862/detection)
  • parse_domain_from_email_address (default=True): Whether or not to parse domain names from email addresses (e.g. example.com from foo@example.com)
  • parse_address_from_cidr (default=True): Whether or not to parse IP addresses from CIDR ranges (e.g. 0.0.0.1 from 0.0.0.1/24)
  • parse_urls_without_scheme (default=True): Whether or not to parse URLs without a scheme (see https://en.wikipedia.org/wiki/Uniform_Resource_Identifier#Generic_syntax) (e.g. hightower.space/projects)
  • parse_imphashes (default=True): Parse import hashes (which look like md5s, but are preceded by 'imphash' or 'import hash')
  • parse_authentihashes (default=True): Parse authentihashes (which look like sha256s, but are preceded with 'authentihash')

See test_ioc_finder.py for more examples.

Output

The find_iocs() returns a dictionary in the following structure:

{
    "asns": [],
    "attack_mitigations": {
        "enterprise": [],
        "mobile": []
    },
    "attack_tactics": {
        "enterprise": [],
        "mobile": [],
        "pre_attack": []
    },
    "attack_techniques": {
        "enterprise": [],
        "mobile": [],
        "pre_attack": []
    },
    "authentihashes": [],
    "bitcoin_addresses": [],
    "cves": [],
    "domains": [],
    "email_addresses": [],
    "email_addresses_complete": [],
    "file_paths": [],
    "google_adsense_publisher_ids": [],
    "google_analytics_tracker_ids": [],
    "imphashes": [],
    "ipv4_cidrs": [],
    "ipv4s": [],
    "ipv6s": [],
    "mac_addresses": [],
    "md5s": [],
    "phone_numbers": [],
    "registry_key_paths": [],
    "sha1s": [],
    "sha256s": [],
    "sha512s": [],
    "ssdeeps": [],
    "tlp_labels": [],
    "urls": [],
    "user_agents": [],
    "xmpp_addresses": []
}

For example, running the example code shown at the start of the usage section above produces the following output:

{
    "asns": [],
    "attack_mitigations": {
        "enterprise": [],
        "mobile": []
    },
    "attack_tactics": {
        "enterprise": [],
        "mobile": [],
        "pre_attack": []
    },
    "attack_techniques": {
        "enterprise": [],
        "mobile": [],
        "pre_attack": []
    },
    "authentihashes": [],
    "bitcoin_addresses": [],
    "cves": [],
    "domains": ["example.org", "example.com"],
    "email_addresses": [],
    "email_addresses_complete": [],
    "file_paths": [],
    "google_adsense_publisher_ids": [],
    "google_analytics_tracker_ids": [],
    "imphashes": [],
    "ipv4_cidrs": [],
    "ipv4s": [],
    "ipv6s": [],
    "mac_addresses": [],
    "md5s": [],
    "phone_numbers": [],
    "registry_key_paths": [],
    "sha1s": [],
    "sha256s": [],
    "sha512s": [],
    "ssdeeps": [],
    "tlp_labels": [],
    "urls": ["https://example.org/test/bingo.php"],
    "user_agents": [],
    "xmpp_addresses": []
}
Output Details

There are two grammars for email addresses. There is a fairly complete grammar to find email addresses matching the spec (which is very broad). Any of these complete email addresses (e.g. foo"bar@gmail.com) will be sent as output to in email_addresses_complete key.

Email addresses in the simple form we are familiar with (e.g. bar@gmail.com) will be sent as output in the email_addresses key.

Command-Line Interface

The ioc-finder package can be used from a command line like:

ioc-finder "This is just an example.com https://example.org/test/bingo.php"

This will return:

{
    "asns": [],
    "attack_mitigations": {
        "enterprise": [],
        "mobile": []
    },
    "attack_tactics": {
        "enterprise": [],
        "mobile": [],
        "pre_attack": []
    },
    "attack_techniques": {
        "enterprise": [],
        "mobile": [],
        "pre_attack": []
    },
    "authentihashes": [],
    "bitcoin_addresses": [],
    "cves": [],
    "domains": [
        "example.com",
        "example.org"
    ],
    "email_addresses": [],
    "email_addresses_complete": [],
    "file_paths": [],
    "google_adsense_publisher_ids": [],
    "google_analytics_tracker_ids": [],
    "imphashes": [],
    "ipv4_cidrs": [],
    "ipv4s": [],
    "ipv6s": [],
    "mac_addresses": [],
    "md5s": [],
    "phone_numbers": [],
    "registry_key_paths": [],
    "sha1s": [],
    "sha256s": [],
    "sha512s": [],
    "ssdeeps": [],
    "tlp_labels": [],
    "urls": [
        "https://example.org/test/bingo.php"
    ],
    "user_agents": [],
    "xmpp_addresses": []
}

Here are the usage instructions for the CLI:

Usage: ioc-finder [OPTIONS] TEXT

  CLI interface for parsing indicators of compromise.

Options:
  --no_url_domain_parsing         Using this flag will not parse domain names
                                  from URLs
  --no_email_addr_domain_parsing  Using this flag will not parse domain names
                                  from email addresses
  --no_cidr_address_parsing       Using this flag will not parse IP addresses
                                  from CIDR ranges
  --no_xmpp_addr_domain_parsing   Using this flag will not parse domain names
                                  from XMPP addresses
  --help                          Show this message and exit.

Parsing Specific Indicator Types

If you need to parse a specific indicator type, you can do this using one of the parse functions that start with parse_. For example, the code below will parse URLs:

from ioc_finder import parse_urls
results = parse_urls('https://google.com')
print(results)

Warning: It is recommended that you use the ioc_finder.find_iocs function rather than a parse function as ioc_finder.find_iocs handles some of the nuances of parsing indicator data. If you use specific functions, the text you provide will not be fanged and may not return the data you expect.

Credits

This project uses the ioc_fanger package to make sure that all indicators in the text are properly fanged.

This package was created with Cookiecutter and Floyd Hightower's python-project-template project template.

Previous iterations of this package were inspired by https://github.com/mosesschwartz/extract_iocs.

Other Helpful Projects

You may also be interested in https://github.com/ioc-fang/ioc_fanger, a project to fang and defang indicators of compromise. For example,

defanging:

example.com => example[.]com
https://example.com => hXXps://example[.]com

and fanging:

example[.]com => example.com
example(.)com => example.com
me AT example(.)com => me@example.com

Similar Projects

There are a number of projects available to find Indicators of Compromise. Your mileage may vary with them. If there are things that another package can do that you would like to see in this package, let me know (or contact me). Here are a few other ones:

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for ioc-finder, version 4.0.1
Filename, size File type Python version Upload date Hashes
Filename, size ioc_finder-4.0.1-py2.py3-none-any.whl (25.5 kB) File type Wheel Python version py2.py3 Upload date Hashes View
Filename, size ioc_finder-4.0.1.tar.gz (50.0 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page