Fake Data Generator

What each data type gives you

There are 21 types. A few behave in ways worth knowing before you load them into a database:

• Credit Card (test) numbers pass the Luhn check and come in several lengths and layouts, like 3721-613349-73872. Strip the dashes if your column only takes digits.
• IBAN values have correct check digits (the mod-97 rule from ISO 13616) and come from dozens of countries, for example FI8219492700530086 or IT84G22147706599864U4701ENQ.
• Password gives 15 random letters and digits, sometimes with an _. Color gives a hex value like #deecd6, and IP Address an IPv4 address like 63.157.190.32.
• Date is written as YYYY-MM-DD. Number gives whole numbers only, between your two limits.
• Phone Number uses US national format, like (579) 533-9020. Address is one string with street, city, state and ZIP code.

Why test with fake data instead of a copy of production

A production dump holds real people's names, emails and addresses, and from a development database it spreads to laptops, logs and screenshots. Privacy laws such as the EU's GDPR, Brazil's LGPD and India's DPDP Act cover personal data wherever it's stored, test systems included.

Generated data avoids that. One catch: faker emails use real domains such as gmail.com and yahoo.com. Turn off outgoing mail in test environments, or replace the domain with example.com, which RFC 2606 reserves for examples.

How the CSV and SQL output is written

The CSV follows RFC 4180, the common CSV format:

• The first line is a header with the column names. Lines end with CRLF (\r\n).
• A value that contains a comma, a double quote or a line break is wrapped in double quotes, and a double quote inside it is written twice ("").
• Booleans are the words true and false, because CSV has no data types.
• The SQL output is one multi-row INSERT. Strings are in single quotes with ' doubled, numbers have no quotes, and booleans are TRUE and FALSE.
• Table and column names made of letters, digits and underscores stay as they are. Other names get ANSI double quotes, such as "first name". MySQL reads double quotes as identifiers only in ANSI_QUOTES mode, so keep names simple there.
• TRUE and FALSE work in PostgreSQL, MySQL and SQLite. SQL Server has no boolean literals, so replace them with 1 and 0 for a bit column.

Load the CSV into a database

Create the table first, with the columns in the same order as the file. Each command skips the header row. PostgreSQL reads the text true and false into a boolean column; other databases may keep them as text, so the SQL INSERT output is often simpler there.

\copy customers FROM 'fake-data.csv' WITH (FORMAT csv, HEADER true)

LOAD DATA LOCAL INFILE 'fake-data.csv' INTO TABLE customers
FIELDS TERMINATED BY ',' OPTIONALLY ENCLOSED BY '"'
LINES TERMINATED BY '\r\n'
IGNORE 1 LINES;

.import --csv --skip 1 fake-data.csv customers

Fake data for other countries and languages

faker has data for dozens of locales, with local names, cities and phone formats. If your users are in Germany, Brazil, India or Japan, generate rows in code with the matching locale. The comments show the output for seed 10.

import { fakerDE, fakerFR, fakerPT_BR, fakerEN_IN, fakerJA } from "@faker-js/faker";

for (const f of [fakerDE, fakerFR, fakerPT_BR, fakerEN_IN, fakerJA]) {
  f.seed(10);
  console.log(f.person.fullName(), "|", f.location.city(), "|", f.phone.number({ style: "national" }));
}
// Marius Köhre | Alt Marisaburg | 05867 297513
// Liétald Le gall | Calais | 06 10 69 05 86
// Alexandre Reis | Melissa do Descoberto | (69) 05867-2975
// Aanand Patil | Khair | 06710 690 586
// 本間 海翔 | 湖節子村 | 005-867-2975

from faker import Faker

fake = Faker(["de_DE", "fr_FR", "pt_BR", "en_IN", "ja_JP"])
for locale in fake.locales:
    f = fake[locale]
    print(f.name(), "|", f.city(), "|", f.phone_number())