Skip to main content

Hypothesis extension to allow generating strings based on regex

Project description

License: MIT Build Status PyPI

Hypothesis extension to allow generating strings based on regex. Useful in case you have some schema (e.g. JSON Schema) which already has regular expressions validating data.


from hypothesis_regex import regex
import requests
import json

EMAIL_REGEX = r"^[a-zA-Z0-9_.+-]+@[a-zA-Z0-9-]{2,}\.[a-zA-Z0-9-.]{2,}$"

def test_registering_user(email):
    response ='/signup', json.dumps({'email': email}))
    assert response.status_code == 201


Regex strategy returns strings that always match given regex (this check is enforced by a filter) and it tries to do that in an effective way so that less generated examples are filtered out. However, some regex constructs may decrease strategy efficiency and should be used with caution:

  • “^” and “$” in the middle of a string - do not do anything.

  • “\b” and “\B” (word boundary and not a word boundary) - do not do anything and instead just rely on top-level regex match filter to filter out non-matching examples.

  • positive lookaheads and lookbehinds just generate data they should match (as if it was part of preceeding/following parts).

  • negative lookaheads and lookbehinds do not do anything so it relies on preceeding/following parts to generate correct strings (otherwise the example will be filtered out).

  • “(?(id)yes-pattern|no-pattern)” does not actually check if group with given id was actually used and instead just generates either yes- or no-pattern.

Regex strategy tries to go all crazy about generated data (e.g. “$” at the end of a string either does not generate anything or generate a newline). The idea is not to generate a nicely looking strings but instead any craze unexpected combination that will still match your given regex so you can prepare for those and handle them in most apropriate way.

You can use regex flags to get more control on strategy:

  • re.IGNORECASE - literals or literal ranges generate both lowercase and uppercase letters. E.g. r’a’ will generate both “a” and “A”, or ‘[a-z]’ will generate both lowercase and uppercase english characters.

  • re.DOTALL - “.” char will be able to generate newlines

  • re.UNICODE - character categories (”\w”, “\d” or “\s” and their negations) will generate unicode characters. This is default for Python 3, see re.ASCII to reverse it.

There are two ways to pass regex flags:

  1. By passing compiled regex with that flags: regex(re.compile(‘abc’, re.IGNORECASE))

  2. By using inline flags syntax: regex(‘(?i)abc’)


$ pip install hypothesis-regex



MIT licensed. See the bundled LICENSE file for more details.

Project details

Download files

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

Source Distribution

hypothesis-regex-0.3.1.tar.gz (9.2 kB view hashes)

Uploaded Source

Built Distribution

hypothesis_regex-0.3.1-py2.py3-none-any.whl (9.0 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page