fixturebuilder_py

Library to generate test data to use as database fixtures and provide simple syntax to set the database to it's initial state at the start of each test.

The library provides two main components:

FixtureBuilder: a simple builder to configure the contents of a single table FixtureCollection: a builder to combine multiple FixtureBuilders to a complete database setup, including foreign key relationships.

Design goals

Immutablility

Test data should always be completely isolated. Changes to a dataset should only affect a single use, without any effects on other uses.

Easy setup of complex schemas

Example data for different test cases oftentimes only differs in a couple of values. The bulk of the example values can be random values or even identical throughout testcases. With traditional tools the developer has to do a lot of busy work to maintain example data. Fixturebuilder aims to provide a simple interface that only requires to maintain the values that actually matter.

Interoperability

It should be easy to integrate fixturebuilder into existing tool chains.

Installt

$ pip install fixturebuilder

Usage FixtureBuilder

from fixturebuilder import FixtureBuilder

# define the data structure to be worked on
start_data = {
    'prop1': 'value1',
    'prop2': 'value2',
    'dict1': {
        'dictprop1': 'dictvalue1'
    },
    'list1': [
        'listvalue1',
        'listvalue2'
    ],
    'list2': [
        {'listdictprop1': 'listdictprop1', 'listdictlist1': ['listdictlistvalue1']}
    ]
}

# initialize a new builder
builder = FixtureBulder.create(start_data)

# retrieve the data unchanged
test_data1 = builder.data

# set a new value on a property and retrieve the updated data
test_data2 = builder.set('prop1', 1).data

# do complex manipulation on the start_data and retrieve the modified data
test_data3 = builder
    .with_dict('dict1')  # descend into 'dict1' and work on it's properties
        .set('dictprop1', 12)
        .set('newlist', ['value'])
        .add('newlist', 'additional value')  # add a new value to a list
        .set('newdict', {'someprop': 'some value'})
        .with_dict('newdict')
            .set('someprop', 'change previously created value')
        .done()  # finish working on 'newdict' and return to 'dict1'
    .done()  # finish working on 'dict1' and return to the root
    .data


test_data4 = builder
    .duplicate_last_list_entry('list2')  # append a duplicated dict to the list
    .with_dict_list_element('list2')     # descend into the last list element to work on it's properties
        .add('listdictlist1', 'new value1')
        .set('listdictprop1', False)
    .done()  # finish working on the list element and return to the root
    .data

# retrieve the original unmodified data
original_data = builder.data

# setting new property raises an error
builder.set('newprop', 'newvalue')

# adding a new property adds it to the builder's data structure
builder.add('newprop', 'newvalue')

Usage value creator function

from fixturebuilder import FixtureBuilder
from faker import Faker

# define the data structure to be worked on
start_data = {
    'prop1': faker.random_number,
    'prop2': faker.address,
    'dict1': {
        'dictprop1': faker.first_name
    },
    'list1': [
        faker.date,
        faker.date
    ],
    'list2': [
        {'listdictprop1': faker.text, 'listdictlist1': [faker.]}
    ]
}

# initialize a new builder
builder = FixtureBuilder.create(start_data)

# retrieve the data unchanged
test_data1 = builder.data

# creates a new set of data
builder2 = builder.copy()

Usage FixtureCollection

from fixturebuilder import FixtureBuilder, FixtureCollection

# define the data structure to be worked on
table1 = {
    'id': 10,
    'prop1': 'value1',
    'prop2': 'value2',
}

table2 = {
    'id': 20,
    'attr1': 'attrval1',
    'attr2': 'attrval2',
}

builder1 = FixtureBuilder.create(table1)
builder2 = FixtureBuilder.create(table2)

collection = FixtureCollection.create() \
    .add_fixture('table1', builder1) \
    .add_fixture('table2', builder2) \
    .add_link('table2.table1_id', 'table1.id=1')

# get the raw data represented by this collection
print(collection.data)
# get the raw data for a single fixture. All links are resolved to their actual values
print(collection.get_fixture('table2').data)

flowpl / fixturebuilder_py Goto Github PK