Document generator from cfn template files.


Keywords
aws, cloudformation
License
MIT
Install
pip install cfn-docgen==1.0.1

Documentation

cfn-docgen - AWS CloudFormation Document Generator

AWS CodeBuild status PyPI - Version PyPI - Python Version Visual Studio Marketplace Version (including pre-releases) Docker Image Version (latest semver) PyPI - License

Generate human-readable documents from AWS CloudFormation yaml/json templates.


Notice

We have made breaking changes from v0.7 to current versions.


Example

Given that you created some cfn template yaml file. When you use cfn-docgen cli. Then, you can generate markdown document.

$ cfn-docgen docgen \
    --format markdown \
    --source docs/sample-template.yaml \
    --dest ./docs/
[INFO] successfully generate document [./docs/sample-template.md] from template [docs/sample-template.yaml]

The left image is a source cfn template file and the right image is a generated document markdown.

For full example, see docs folder

template-source-and-document-dest


Install and How to use

CLI

$ pip install cfn-docgen

# you can also geenrate a document from a template at S3 bucket and upload it directory.
$ cfn-docgen docgen \
  --source s3://bucket/prefix/sample-template.yaml \
  --dest s3://bucket/prefix/sample-template.md
[INFO] successfully generate document [s3://bucket/prefix/sample-template.md] from template [s3://bucket/prefix/sample-template.yaml]

# and you can generate multiple documents from templates in direcotry(or s3 bucket prefix) at once
$ tree ./templates/
./templates/
├── sample-template-1.yaml
└── subfolder
    └── sample-template-2.yaml

1 directory, 2 files
$ cfn-docgen docgen \
    --source ./templates/ \
    --dest s3://bucket/documents/
[INFO] successfully generate document [s3://bucket/documents/sample-template-1.md] from template [./templates/sample-template-1.yaml]
[INFO] successfully generate document [s3://bucket/documents/subfolder/sample-template-2.md] from template [./templates/subfolder/sample-template-2.yaml]

Visual Studio Code Extension

You can use cfn-docgen as Visual Studio Code Extension

cfn-docgen-vsc-extension-docgen-sample

cfn-docgen-vsc-extension-skeleton-sample


Docker Image

# pull image from DockerHub
$ docker pull horietakehiro/cfn-docgen:latest

# local directory(before)
$ tree /tmp/sample/
/tmp/sample/
└── sample-template.json

0 directories, 1 files

# run as command
$ docker run \
  -v /tmp/sample/:/tmp/ \
  horietakehiro/cfn-docgen:latest docgen \
    --source /tmp/sample-template.json \
    --dest /tmp/
[INFO] successfully generate document [/tmp/sample-template.md] from template [/tmp/sample-template.json]

# local directory(after)
$ tree /tmp/sample/
/tmp/sample/
├── sample-template.json
└── sample-template.md

0 directories, 2 files

API

from cfn_docgen import (
    CfnDocgenService, CfnDocgenServiceCommandInput,
    CfnTemplateSource, CfnDocumentDestination
)
service = CfnDocgenService.with_default()
service.main(
    command_input=CfnDocgenServiceCommandInput(
        template_source=CfnTemplateSource("s3://bucket/template.yaml", service.context),
        document_dest=CfnDocumentDestination("s3://bucket/document.md", service.context),
        fmt="markdown",
    )
)

Serverless

You can also use cfn-docgen on AWS Cloud as serverless application.

You can deploy resources at AWS Serverless Application Repository.

Once deployed, tha S3 bucket named cfn-docgen-${AWS::AccountId}-${AWS::Region} is created on your account.

When you upload cfn template json/yaml files at templates/ folder of the bucket, cfn-docgen-serverless automatically will be triggered and generates markdown docments for them at documents/ folder.


Features


Embedding Descirptions


Top level descriptions

You can embed description for the template at top level Metadata section like this, then markdown document will be generated from it.

Metadata:
  CfnDocgen:
    Description: |
      このテンプレートファイル東京リージョン上で以下のリソースを作成します
      - VPC
      - パブリックサブネット(2AZに1つずつ)

      ![Archtecture](./images/sample-template.drawio.png)


      **注意点**
      - このテンプレートファイルは**東京リージョン**上でのみの使用に制限しています
      - このテンプレートファイルを使用する前に、[東京リージョン上に作成可能なVPCの最大数の設定](https://ap-northeast-1.console.aws.amazon.com/servicequotas/home/services/vpc/quotas/L-F678F1CE)を確認することを推奨します(デフォルトは5VPC)**

Then, the generated description will be like below.

top-level-description

You can also embed descriptions for each sections - Mappings, Conditions, Rules.

Metadata:
  CfnDocgen:
    Mappings:
      CidrBlockMap: CidrBlocks for each environment
    Conditions:
      EnvCondition: Check if the value of parameter `EnvType` is `prod`
    Rules:
      RegionRule: This template is available only in ap-northeast-1

Then, the generated description will be like below.

each-section-description

Resources and Properties description

You can embed descriptions for resources and their properties in Metadata section in each resources.

Resources: 
  VPC:
    Metadata:
      CfnDocgen:
        Description: アプリケーションサーバを稼働させるために使用するVPC
        Properties:
          EnableDnsHostnames: アプリケーションサーバのホスト名でパブリックIPを名前解決できるように有効化する
    Type: AWS::EC2::VPC
    Properties: 
      CidrBlock: ...

Then, the generated description will be like below.

resource-level-description


Integration with AWS CDK

cfn-docgen can generate documents from AWS-CDK-generated templates, and you can also embed descriptions in cdk codes like below.

from aws_cdk import (
    Stack,
    aws_ec2 as ec2,
    
)
from constructs import Construct
from typing import Any
from cfn_docgen.domain.model.cfn_template import (
    CfnTemplateMetadataCfnDocgenDefinition as Metadata,
    CfnTemplateResourceMetadataDefinition as ResourceMetadata,
    CfnTemplateResourceMetadataCfnDocgenDefinition as CfnDocgen
)
class CfnDocgenSampleCdkStack(Stack):

    def __init__(self, scope: Construct, construct_id: str, **kwargs:Any) -> None:
        super().__init__(scope, construct_id, **kwargs)
        # top-level description for the stack
        self.add_metadata(
            "CfnDocgen", Metadata(
                Description="top-level-description"
            ).model_dump(),
        )
        self.vpc_construct = ec2.Vpc(self, "VpcConstruct", max_azs=1)
        # resource-level descriptions
        self.vpc_construct.node.default_child.cfn_options.metadata = ResourceMetadata(
            CfnDocgen=CfnDocgen(Description="resource-level-description")
        ).model_dump()

Then, the table of contents of generated document will be like below.


Integration with custom resource specification

You can define custom resource specification like this and generate documents for them.

$ cfn-docgen docgen \
  -s docs/sample-template.yaml \
  -s docs/sample-template.md \
  -c docs/custom-specification.json

Generate skeletons

You can generate definition skeletons for each resource types.

$ cfn-docgen skeleton --type AWS::EC2::VPC --format yaml
Type: AWS::EC2::VPC
Metadata:
  CfnDocgen:
    Description: ''
    Properties: {}
Properties:
  InstanceTenancy: String
  Ipv4NetmaskLength: Integer
  CidrBlock: String
  Ipv4IpamPoolId: String
  EnableDnsSupport: Boolean
  EnableDnsHostnames: Boolean
  Tags:
    - Key: String
      Value: String

$ cfn-docgen skeleton --type AWS::EC2::VPC --format json
{
  "Type": "AWS::EC2::VPC",
  "Metadata": {
    "CfnDocgen": {
      "Description": "",
      "Properties": {}
    }
  },
  "Properties": {
    "InstanceTenancy": "String",
    "Ipv4NetmaskLength": "Integer",
    "CidrBlock": "String",
    "Ipv4IpamPoolId": "String",
    "EnableDnsSupport": "Boolean",
    "EnableDnsHostnames": "Boolean",
    "Tags": [
      {
        "Key": "String",
        "Value": "String"
      }
    ]
  }
}