pieterraform 0.0.1a5

A Python Demo

PieTerraform

This is a wrapper for terraform to facilitate python developer to call terraform. It is designed to invoke terraform commands in "Functional Chaining" style

Note: Only python>=3.7 is supported.

Features

• Builder pattern. One line code can call multiple terraform commands
• Terraform commands' arguments are encapsulated as functions so code completion supported
• No need to worry about the sequence of arguments

Usage

Make sure have terraform in $PATH

terraform version

Install

pip install pieterraform

Quick start

from pieterraform import Terraform

# suppose you have terraform files in ./tf
Terraform().workdir('./tf')
    .init().run()           # 'terraform init'
    .plan().run()           # 'terraform plan'
    .apply().run()          # 'terraform apply'

# suppose you have terraform files in ./tf/prod
Terraform().workdir('./tf')
    .init().dir('prod').run()
    .plan().dir('prod').run()
    .apply().dir('prod').run()

# suppose you have terraform files in ./tf/prod
Terraform().workdir()
    .init().dir('tf/prod').run()
    .plan().dir('tf/prod').run()
    .apply().dir('tf/prod').run()

Just ONE LINE code!

With Paramers

from pieterraform import Terraform

# suppose you have terraform files in ./tf
Terraform().workdir('./tf')
    # 'terraform init -no-color -upgrade=false'
    .init().no_upgrade().no_color().run()
    # 'terraform plan -state mystate.json -no-color'
    .plan().state_file('mystate.json').no_color().out('myplan').run()
    # 'terraform apply myplan'
    .apply().use_plan('myplan').run()
    # 'terraform destroy -auto-approve -state mystate.json'
    .destroy().auto_approve().state('mystate.json').run()

With Custome Log

By default it prints log in screen. But you can cusomize it to use any logger

import logging
from pieterraform import Terraform

# output log to file
logFormatter = logging.Formatter('%(asctime)s [%(levelname)-5.5s] %(message)s')
log_file = 'log.txt'
f_handler = logging.FileHandler(log_file)
f_handler.setFormatter(logFormatter)
f_handler.setLevel(logging.DEBUG)
logger = logging.getLogger(__name__)
logger.setLevel(logging.DEBUG)
logger.addHandler(f_handler)

# suppose you have terraform files in ./tf
Terraform(logger=logger) # this will output log to file 'log.txt'
    .workdir('./tf')
    .init().run()
    .plan().run()
    .apply().run()

Traditional calls rather than functions chain

from pieterraform import Terraform

tf = Terraform().workdir('./tf')
initer = tf.init()
initer.no_upgrade()
initer.no_color().run()
planer = tf.plan()
planer.state_file('mystate.json')
planer.no_color()
planer.out('myplan').run()
applyer = tf.apply()
applyer.use_plan('myplan')
applyer.run()

Check result

from pieterraform import Terraform

a_run = Terraform().workdir('./tf')
    .init().run()
    .plan().run()
    .apply().run()

for r in a_run.results:
    print(r.output)
    print(r.command)

Source Code

This project is fully using docker as dev environment.

Prerequisition

• docker: ">= 17.06"
• docker-compose: ">= 1.26"

No python needed.

Build

make

Install to local

make install

Run test

make test

Distribution

make dist

Development

Start dev docker

make docker-dev

this will start a container named pieterraform-devenv

Use VSCode

Open your vscode, attach to above container to do remote development