Terraform State

In this example you are going to create a terraformResource and tarraformOutput blueprint that ingests all resources and outputs in your terraform.tfstate file using a combination of Port's API and webhook functionality.

To ingest the resources to Port, a script that sends information about resources and outputs according to the webhook configuration is used.

Prerequisites

Create the following blueprint definition and webhook configuration:

Terraform resources blueprint

{
  "identifier": "terraformResource",
  "description": "This blueprint represents a Terraform resource in our software catalog",
  "title": "Terraform Resources",
  "icon": "Terraform",
  "schema": {
    "properties": {
      "mode": {
        "type": "string",
        "title": "Mode",
        "default": "data",
        "enum": ["data", "managed", "module", "import", "state"]
      },
      "module": {
        "type": "string",
        "title": "Module"
      },
      "type": {
        "title": "Type",
        "type": "string"
      },
      "provider": {
        "title": "Provider",
        "type": "string"
      },
      "instances": {
        "title": "Instances",
        "type": "array"
      },
      "lineage": {
        "title": "Lineage",
        "type": "string"
      }
    },
    "required": []
  },
  "mirrorProperties": {},
  "calculationProperties": {},
  "relations": {}
}

Terraform outputs blueprint

{
  "identifier": "terraformOutput",
  "title": "Terraform Output",
  "description": "This blueprint represents a Terraform output in our software catalog",
  "icon": "Terraform",
  "schema": {
    "properties": {
      "type": {
        "title": "Type",
        "type": "string"
      },
      "value": {
        "title": "Value",
        "type": "string"
      },
      "sensitive": {
        "title": "Sensitive",
        "type": "boolean",
        "default": false
      },
      "description": {
        "title": "Description",
        "type": "string"
      },
      "lineage": {
        "title": "Lineage",
        "type": "string"
      }
    },
    "required": []
  },
  "mirrorProperties": {},
  "calculationProperties": {},
  "aggregationProperties": {},
  "relations": {}
}

Terraform resources webhook configuration

{
  "identifier": "terraformResourceMapper",
  "title": "Terraform Resources Mapper",
  "description": "A webhook configuration to map Terraform tfstate file",
  "icon": "Terraform",
  "mappings": [
    {
      "blueprint": "terraformOutput",
      "filter": ".body | has(\"outputs\")",
      "itemsToParse": ".body.outputs",
      "entity": {
        "identifier": ".item.name | tostring",
        "title": ".item.name | tostring",
        "properties": {
          "description": ".item.description",
          "sensitive": ".item.sensitive",
          "type": ".item.type",
          "value": ".item.value",
          "lineage": ".body.lineage"
        }
      }
    },
    {
      "blueprint": "terraformResource",
      "filter": ".body | has(\"resources\")",
      "itemsToParse": ".body.resources",
      "entity": {
        "identifier": ".item.id",
        "title": ".item.name",
        "properties": {
          "mode": ".item.mode",
          "module": ".item.module",
          "type": ".item.type",
          "provider": ".item.provider",
          "instances": ".item.instances",
          "lineage": ".body.lineage"
        }
      }
    }
  ],
  "enabled": true,
  "security": {}
}

Working with Port's API and Bash script

Here is an example snippet showing how to integrate Port's API and Webhook with your existing pipelines using Python and Bash:

Python
Bash

Create the following Python script in your repository to create or update Port entities as part of your pipeline:

Python script example

import requests
import json
import os

# Get environment variables using the config object or os.environ["KEY"]
WEBHOOK_URL = os.environ['WEBHOOK_URL'] ## the value of the URL you receive after creating the Port webhook
PATH_TO_TERRAFORM_TFSTATE_FILE = os.environ['PATH_TO_TERRAFORM_TFSTATE_FILE']


def add_entity_to_port(entity_object):
    """A function to create the passed entity in Port using the webhook URL

    Params
    --------------
    entity_object: dict
        The entity to add in your Port catalog

    Returns
    --------------
    response: dict
        The response object after calling the webhook
    """
    headers = {"Accept": "application/json"}
    response = requests.post(WEBHOOK_URL, json=entity_object, headers=headers)
    return response.json()

def parse_tf_outputs(output_data):
    tf_outputs = []
    for output_name, output_info in output_data.items():
        output_type = type(output_info.get("value")).__name__
        tf_outputs.append({
            'name': output_name,
            'description': output_info.get("description"),
            'type': output_type,
            'sensitive': output_info.get('sensitive'),
            'value': str(output_info.get('value'))
        })
    return tf_outputs

def parse_tf_resources(resources):
    tf_resources = []
    index = 1
    for resource in resources:
        resource_id = f"tf-rs-{index}"
        tf_resources.append({
            'name': resource.get('name'),
            'mode': resource.get('mode'),
            'module': resource.get('module'),
            'type': resource.get('type'),
            'provider': resource.get('provider'),
            'instances': resource.get('instances'),
            'id': resource_id
        })
        index+=1
    return tf_resources


def read_tfstate_file(tfstate_json_path):
    """This function takes a tfstate_json_path file path, converts the resources and outputs property into a
    JSON array and then sends the data to Port

    Params
    --------------
    tfstate_json_path: str
        The path to the terraform.tfstate file relative to the project's root folder

    Returns
    --------------
    response: dict
        The response object after calling the webhook
    """
    with open(tfstate_json_path) as file:
        data = json.load(file)

    resources = data.get('resources', [])
    outputs = data.get('outputs', {})
    lineage = data.get('lineage')

    tf_resources = parse_tf_resources(resources)
    tf_outputs = parse_tf_outputs(outputs)

    entity_object = {
        "resources": tf_resources,
        "outputs": tf_outputs,
        "lineage": lineage
    }
    webhook_response = add_entity_to_port(entity_object)
    return webhook_response

response = read_tfstate_file(PATH_TO_TERRAFORM_TFSTATE_FILE)
print(response)

Create the following Bash script in your repository to create or update Port entities as part of your pipeline:

Bash script example

#!/bin/bash

# Set environment variables
WEBHOOK_URL="$WEBHOOK_URL"
PATH_TO_TERRAFORM_TFSTATE_FILE="$PATH_TO_TERRAFORM_TFSTATE_FILE"

# A function to create the passed entity in Port using the webhook URL
add_entity_to_port() {
  local entity_object_file="$1"
  local headers="Accept: application/json"
  local response=$(curl -X POST -H "$headers" -H "Content-Type: application/json" --data-binary "@$entity_object_file" "$WEBHOOK_URL")
  echo "$response"
}

# This function takes a tfstate_json_path file path, converts the "resources" property into a
# JSON array and then sends the data to Port
read_tfstate_file() {
    local package_json_path="$1"
    local data=$(cat "$package_json_path")
    local resources=$(echo "$data" | jq -c '.resources[]')
    local lineage=$(echo "$data" | jq -r '.lineage')

    index=1
    tf_resources=()

    while IFS= read -r resource; do
      resource_id="tf-rs-$index"
      resource_name=$(jq -r '.name' <<< "$resource")
      resource_mode=$(jq -r '.mode' <<< "$resource")
      resource_module=$(jq -r '.module' <<< "$resource")
      resource_type=$(jq -r '.type' <<< "$resource")
      resource_provider=$(jq -r '.provider' <<< "$resource")
      resource_instances=$(jq -r '.instances' <<< "$resource")

      tf_resources+=("{\"name\":\"$resource_name\",\"mode\":\"$resource_mode\",\"module\":\"$resource_module\",\"type\":\"$resource_type\",\"provider\":\"$resource_provider\",\"instances\":$resource_instances,\"id\":\"$resource_id\"}")
      ((index++))
    done <<< "$resources"

    local entity_object="{\"resources\":[${tf_resources%,}],\"lineage\":\"$lineage\"}"

    # since some tfstate may be quite large, we can write the data unto a temporary file
    local entity_object_file=$(mktemp)
    echo "$entity_object" > "$entity_object_file"
    local webhook_response=$(add_entity_to_port "$entity_object_file")
    echo "$webhook_response"

    # Clean up the temporary file
    rm "$entity_object_file"
}

response=$(read_tfstate_file "$PATH_TO_TERRAFORM_TFSTATE_FILE")
echo "$response"

Prerequisites​

Working with Port's API and Bash script​

Prerequisites

Working with Port's API and Bash script