Barnum is a python-based application for quickly and easily creating pseudo-random data typically used for application testing.
I am developing a shopping cart application in Django and realized that I needed a bunch of data to simulate the store's behavior under somewhat normal production usage.
I got tired of always trying to think of names and addresses for customers and so decided to automate the process a little bit. Such was born Barnum.
I created this in 2007 and hosted it on Google Code. After a while I forgot about it. The code sat in obscurity until I received an email from Google about the eventually closing of Google Code. I decided I would migrate it to github to see if it is useful to anyone else.
I was able to find some online systems for generating large amounts of test data. I could not find any application that had the breadth of data generation capabilities nor the ability to easily interface with Django in the way I wanted to.
One of the most unique aspects of Barnum is that the data is what I'll call "plausible." For example, here's an example "identity" randomly generated from Barnum:
Sid Seymour 10 Kimbrough Grove Drive Arthur ND, 58006 (701)642-6471
Who works at: Network Hardware Co as a Personnel Clerk Senior
- You should notice a couple of things about this data.
- There's a realistic first and last name
- The street names are also plausible
- Arthur, ND is a real city and the zip code is 58006
- 701 is an area code used for North Dakota
- The fictional company is somewhat reasonable.
- The job position also makes sense.
Well, I find that when testing applications, if it's just a random string of numbers of letters, it gets hard to tell if something is out of place or "looks wrong." If you'd like to just generate totally random information, then you probably don't need Barnum!
I originally created this in 2007 and it sat idle for a while. In that time there have been some other python solutions.
I think fake factory is one of the best out there and provides a lot more options than barnum.
However, Barnum still has some unique aspects, specifically having some "intelligence" as to how the data is related and the flexibility you have to easily customize the data by adding or subtracting to the data in the source-data directory.
- Here's a list of types of dummy data Barnum can create:
- First name and/or last name in either gender
- Job title
- Phone number
- Street number and name
- Zip code plus city & state
- Company name
- Credit card number and type (with valid checksum)
- Dates
- Email addresses
- Sample password
- Words (latin)
- Sentences and/or paragraphs of random latin words
Barnum should work on python 2.7.x and python 3.x
pip install barnum
There has been a change in the most recent version of barnum to make the code more pythonic.
The recommended method is to import barnum and call the functions from your own script.
If you'd like to call it from another script, here's an example or two from the interpreter:
Python 3.5.1 |Continuum Analytics, Inc.| (default, Dec 7 2015, 11:16:01)
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import barnum
>>> barnum.create_name()
('Isabel', 'Madigan')
>>> barnum.create_name()
('Jesica', 'Loveless')
>>> barnum.create_city_state_zip()
('54666', 'Warrens', 'WI')
>>> barnum.create_city_state_zip()
('55113', 'Saint Paul', 'MN')
>>> barnum.create_phone()
'(716)248-1703'
>>> barnum.create_phone()
'(603)502-8450'
>>> barnum.create_phone('38138')
'(615)564-2637'
>>> barnum.create_sentence()
'In lobortis ut te at et feugiat ipsum vel ex feugiat eros.'
>>> barnum.create_sentence()
'In erat hendrerit at odio eu tincidunt exerci.'
>>> barnum.create_cc_number()
('mastercard', ['5231056277792200'])
>>> barnum.create_cc_number()
('visa', ['4929064950922570'])
>>> barnum.create_nouns()
'place blue'
>>> barnum.create_nouns()
'Steven clarinet'
>>> barnum.create_date()
datetime.datetime(2024, 2, 4, 9, 51, 38, 971944)
>>> barnum.create_date(past=True)
datetime.datetime(2006, 9, 23, 9, 51, 46, 927690)
>>> barnum.create_email()
'Dalton.Segal@luptatumdelenitaugue.org'
>>> barnum.create_pw()
'naPTg67M'
You can see that it should be trivial to incorporate this data into any python script. The possibilities of creating CSV's, raw SQL, Python Objects, etc are practically endless!
I pulled sample data and existing scripts from a bunch of different sources. It looks like a lot of the sources are now dead links.
- The names are from 1990 US Census data
- The street names are from real us streets in a few locales.
- Company names are randomly generated by me.
- Job Titles were taken from another census site that I can't seem to find now.
- Zip Codes from another dead site.
- Random latin text came from http://www.4guysfromrolla.com/webtech/052800-1.shtml
- Credit Card generator is from Graham King - http://www.darkcoding.net/index.php/credit-card-numbers/
- Password generator is from Pradeep Kishore Gowda via the Python Cookbook
If all you'd like to do is add some more seed data to an existing source, edit the appropriate file in the source-data directory and execute the convert_data.py script to create a new pickle file.
I've moved the code to github to make it easier for others to contribute. Feel free to send pull requests or submit tickets.
I needed info for the US only. I had access to this data and knew what I wanted. If you would like to add other countries or info, feel free to contribute!
Ummm. Probably not. All of the data is random. The credit card numbers conform to the Luhn 10 checksum formula but are not necessarily valid numbers. Even if they were, you would need to know the real name, address and phone number before you could do anything illegal with the data. I think we're all pretty safe.
Choosing names for projects is kind of fun but kind of a hassle. There needs to be a name but it can't be anything too stupid. I started off thinking of an acronym and ended up with PT ("Python Testing") and immediately thought of P.T. Barnum. I really liked the name because I was using this for Satchmo and project made in Django. Single word names seemed cool. Also, I like the fact that P.T. Barnum was really a master at making people think something was real that wasn't. Which is exactly what this little script does.
I use a couple of other python scripts that were licensed under the GPL. So, I figured it was best to just release under the GPL. If you would like another license arrangement, let me know and I'll see if there's something we can do.