CndScaffold >=3.0.0

Why use CndScaffold

CndScaffold is a python lib, designed to help management of GitOps process. Now base on jinja2, it allow you to merge a dict of data into a list of file.

Connected with the right provider, you can directly push your result to a git server (gitlab / github /Azure DevOps)

How it's works ? (with example)

• define in jinja format a list of template file (the file can be text, yaml or anything else)
• create a file of definition
• apply data on it

What we want as result

Data source

For example we got this data

name: Denis
customers:
  - name: my_first_customer
    email: my_first_customer@sample-mail.com
    id: c6a240ac-587a-11ee-8c99-0242ac120002
    token: ABC
  - name: my_second
    email: my_second_customer@sample-mail.com
    id: d174a3e4-587a-11ee-8c99-0242ac120002
    token: DEF
  - name: my_last
    email: my_last_customer@sample-mail.com
    id: ebe22792-587a-11ee-8c99-0242ac120002
    token: GHI

Files result

And we want to generate multiple file for a gitops process, who look like this result:

config.yml

name: Denis
type: my_type_sample

customers.yml

customers:
- name: my_first_customer
  email: my_first_customer@sample-mail.com
  id: c6a240ac-587a-11ee-8c99-0242ac120002
- name: my_second
  email: my_second_customer@sample-mail.com
  id: d174a3e4-587a-11ee-8c99-0242ac120002
- name: my_last
  email: my_last_customer@sample-mail.com
  id: ebe22792-587a-11ee-8c99-0242ac120002

customers/c6a240ac-587a-11ee-8c99-0242ac120002.yml

token: ABC

customers/d174a3e4-587a-11ee-8c99-0242ac120002.yml

token: DEF

customers/ebe22792-587a-11ee-8c99-0242ac120002.yml

token: GHI

How we do it

Definition of a list of file in jinja

A single simple file :

definition/customers/file1.yml

name: {{ name }}
type: my_type_sample

Now a file with a loop :

definition/customers/loop.yml

{% if customers |length > 1 %}
customers: 
{%- for customer in customers %}
  name: {{ customer.name -}}
  email: {{ customer.email -}}
  id: {{ customer.id -}}
{% endfor %}
{% endif %}

And finally to create multiple file with a loop (like one per customer):

definition/customers/multiple.yml

token: {{ token }}

Definition file

definition/customers.yml

source_folder: customers
files:
- name: file1.yml
  step: ['init']
  target: config.yml
- name: loop.yml
  step: ['init', 'build', 'runtime']
- name: multiple.yml
  step: ['init', 'build', 'runtime']
  repeat: customers 
  target: customers/{{ id }}.yml 

Build the code (localfile)

The first sample will generate the file into a results folder

run.py (generate data in a local folder)

import cnd_scaffold
import cndprint


data = my_data # .... load your data as you need
cnd_scaffold = cnd_scaffold.Scaffold(_print)
model = cnd_scaffold.load_model('definition', customers.yml')
final_files = cnd_scaffold.apply('init', model, data)
final_files = cnd_scaffold.push_files(final_files, 'results')

Build the code (provider gitlab)

This second example will push the result into git_lab_project_id

run.py (generate data in a local folder)

import cnd_scaffold
import cndprint
import cnd_provider


data = my_data # .... load your data as you need
provider = cnd_provider.CndProviderGitlab({"private_token": 'replace-me', "host": "https://mygitlab"}
cnd_scaffold = cnd_scaffold.Scaffold(_print)
model = cnd_scaffold.load_model('definition', customers.yml')
final_files = cnd_scaffold.apply('init', model, data)
final_files = cnd_scaffold.push_files(final_files, "git_lab_project_id", provider=provider)

That's all !

With this definition, each time you use cnd_scaffold you will generate this file structure/content from your own data.

Tips

Step

step, will allow you using different case, you can use it as you prefer, but I suggest you this:

• init: First time only, it's mean this file will never been update by the script
• build: Each time the process start, it will be update
• runtime: Perfect for Terraform secret for example, you can set this to never write result on git, but only on the current pipeline session