Category: Power Apps#

Python + Dataverse Series – Post #03: Create, Update, Delete records via Web API

Hi Folks,

This is continuation in this Python with Dataverse Series, in this blog post, we will perform a full CRUD(Create, Retrieve, Update, Delete) in Dataverse using Web API.

Please use the below code for the same…to make any calls using WEB API to Dataverse.

import pyodbc
import msal
import requests
import json
import re
# Azure AD details
client_id = 'XXXX'
client_secret = 'XXXX'
tenant_id = 'XXXX'
authority = f'https://login.microsoftonline.com/{tenant_id}'
resource = 'https://XXXX.crm8.dynamics.com'
# SQL endpoint
sql_server = 'XXXX.crm8.dynamics.com'
database = 'XXXX'
# Get token with error handling
try:
print(f"Attempting to authenticate with tenant: {tenant_id}")
print(f"Authority URL: {authority}")
app = msal.ConfidentialClientApplication(client_id, authority=authority, client_credential=client_secret)
print("Acquiring token…")
token_response = app.acquire_token_for_client(scopes=[f'{resource}/.default'])
if 'error' in token_response:
print(f"Token acquisition failed: {token_response['error']}")
print(f"Error description: {token_response.get('error_description', 'No description available')}")
else:
access_token = token_response['access_token']
print("Token acquired successfully and your token is!"+access_token)
print(f"Token length: {len(access_token)} characters")
except ValueError as e:
print(f"Configuration Error: {e}")
print("\nPossible solutions:")
print("1. Verify your tenant ID is correct")
print("2. Check if the tenant exists and is active")
print("3. Ensure you're using the right Azure cloud (commercial, government, etc.)")
except Exception as e:
print(f"Unexpected error: {e}")
#Get 5 contacts from Dataverse using Web API
import requests
import json
try:
#Full CRUD Operations – Create, Read, Update, Delete a contact in Dataverse
print("Making Web API request to perform CRUD operations on contacts…")
# Dataverse Web API endpoint for contacts
web_api_url = f"{resource}/api/data/v9.2/contacts"
# Base headers with authorization token
headers = {
'Authorization': f'Bearer {access_token}',
'OData-MaxVersion': '4.0',
'OData-Version': '4.0',
'Accept': 'application/json',
'Content-Type': 'application/json'
}
# Create a new contact
new_contact = { "firstname": "John", "lastname": "Doe" }
print("Creating a new contact…")
# Request the server to return the created representation. If not supported or omitted,
# Dataverse often returns 204 No Content and provides the entity id in a response header.
create_headers = headers.copy()
create_headers['Prefer'] = 'return=representation'
response = requests.post(web_api_url, headers=create_headers, json=new_contact)
created_contact = {}
contact_id = None
# If the API returned the representation, parse the JSON
if response.status_code in (200, 201):
try:
created_contact = response.json()
except ValueError:
created_contact = {}
contact_id = created_contact.get('contactid') or created_contact.get('contactid@odata.bind')
print("New contact created successfully (body returned).")
print(f"Created Contact ID: {contact_id}")
# If the API returned 204 No Content, Dataverse includes the entity URL in 'OData-EntityId' or 'Location'
elif response.status_code == 204:
entity_url = response.headers.get('OData-EntityId') or response.headers.get('Location')
if entity_url:
# Extract GUID using regex (GUID format)
m = re.search(r"([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12})", entity_url)
if m:
contact_id = m.group(1)
created_contact = {'contactid': contact_id}
print("New contact created successfully (no body). Extracted Contact ID from headers:")
print(f"Created Contact ID: {contact_id}")
else:
print("Created but couldn't parse entity id from response headers:")
print(f"Headers: {response.headers}")
else:
print("Created but no entity location header found. Headers:")
print(response.headers)
else:
print(f"Failed to create contact. Status code: {response.status_code}")
print(f"Error details: {response.text}")
# Read the created contact
if not contact_id:
# Defensive: stop further CRUD if we don't have an id
print("No contact id available; aborting read/update/delete steps.")
else:
print("Reading the created contact…")
response = requests.get(f"{web_api_url}({contact_id})", headers=headers)
if response.status_code == 200:
print("Contact retrieved successfully!")
contact_data = response.json()
print(json.dumps(contact_data, indent=4))
else:
print(f"Failed to retrieve contact. Status code: {response.status_code}")
print(f"Error details: {response.text}")
# Update the contact's email
updated_data = { "emailaddress1": "john.doe@example.com" }
response = requests.patch(f"{web_api_url}({contact_id})", headers=headers, json=updated_data)
if response.status_code == 204:
print("Contact updated successfully!")
else:
print(f"Failed to update contact. Status code: {response.status_code}")
print(f"Error details: {response.text}")
# Delete the contact
response = requests.delete(f"{web_api_url}({contact_id})", headers=headers)
if response.status_code == 204:
print("Contact deleted successfully!")
else:
print(f"Failed to delete contact. Status code: {response.status_code}")
print(f"Error details: {response.text}")
except requests.exceptions.RequestException as e:
print(f"Request error: {e}")
except KeyError as e:
print(f"Token not available: {e}")
except Exception as e:
print(f"Unexpected error: {e}")
view raw FullCRUDWebAPI hosted with ❤ by GitHub

You can use the VS Code as IDE, copy the above code in a python file, next click on Run Python File at the top of the VS Code

Hope this helps someone making Web API Calls using Python.

Cheers,

PMDY

Building a Cloud-Native Power Apps ALM Pipeline with GitLab, Google Cloud Build and PAC CLI: Streamlining Solution Lifecycle Automation

A unique combination to achieve deployment automation of Power Platform Solutions

Hi Folks,

This post is about ALM in Power Platform integrating with a different ecosystem than usual, i.e. using Google Cloud, sounds interesting..? This approach is mainly intended for folks using Google Cloud or GitLab as part of their implementation.

Integrating Google Cloud Build with Power Platform for ALM (Application Lifecycle Management) using GitLab is feasible and beneficial. This integration combines GitLab as a unified DevOps platform with Google Cloud Build for executing CI/CD pipelines, enabling automated build, test, export, and deployment of Power Platform solutions efficiently. This was the core idea for my session on Friday 28 November, at New Zealand Business Applications Summit 2025.

Detailed Steps for this implementation

Create an access token in GitLab for API Access and Read Access

Click on Add new token, you can select at the minimum the below scopes while you were working with CI-CD using GitLab

Create a host connection for the repository in GitLab

Specify the personal access token created in the previous step

Link your repository

The created host connections in the previous step will be shown under Connec ctions drop down

Create Trigger in Google Cloud Build

Click on Create trigger above, provide a name, select a nearest region

Event:

For now, I am choosing Manual invocation for illustration

Specify where the name of the Repository where your YAML in GitLab resides

You can optionally specify the substitution variables which are nothing but parameters you can pass to your pipeline from Google Cloud Build Configuration

You can optionally give this for any approval and choose the service account tagged to your google account in the drop down.

Click on Save.

Next proceed to GitLab YAML

You can find the full code below

steps:
– id: "export_managed"
name: "mcr.microsoft.com/dotnet/sdk:9.0"
entrypoint: "bash"
args:
– "-c"
– |
echo "=== 🏁 Starting Export Process ==="
# ✅ Define solution name from substitution variable
SOLUTION_NAME="${_SOLUTION_NAME}"
# ✅ Install PAC CLI
mkdir -p "${_HOME}/.dotnet/tools"
dotnet tool install –global Microsoft.PowerApps.CLI.Tool –version 1.48.2 || true
# Add dotnet global tools dir to the shell PATH for this step/session (preserve existing PATH)
export PATH="$_PATH:${_HOME}/.dotnet/tools"
echo "=== 🔐 Authenticating to Power Platform Environment ==="
pac auth create –name "manual" –url "https://ecellorsdev.crm8.dynamics.com" –tenant "XXXXX-XXXX-XXXXX-XXXXXX-XXXXX" –applicationId "XXXXXXXXXXXXXX" –clientSecret "XXXXXXXXXXXXXXXX"
pac auth list
echo "=== 📦 Exporting Solution: ${_SOLUTION_NAME} ==="
pac solution export \
–name "${_SOLUTION_NAME}" \
–path "/tmp/${_SOLUTION_NAME}.zip" \
–managed true \
–environment "${_SOURCE_ENV_URL}"
echo "=== ✅ Solution exported to /tmp/${_SOLUTION_NAME}.zip ==="
echo "=== 🔐 Authenticating to Target Environment ==="
pac auth create \
–name "target" \
–url "https://org94bd5a39.crm.dynamics.com" \
–tenant "XXXXXXXXXXXXXXXXXXXXXXXX" \
–applicationId "XXXX-XXXXX-XXXXX-XXXXXX" \
–clientSecret "xxxxxxxxxxxxxxxxxxxx"
echo "=== 📥 Importing Solution to Target Environment ==="
pac solution import \
–path "/tmp/${_SOLUTION_NAME}.zip" \
–environment "${_TARGET_ENV_URL}" \
–activate-plugins \
–publish-changes
echo "=== 🎉 Solution imported successfully! ==="
options:
logging: CLOUD_LOGGING_ONLY
substitutions:
_SOLUTION_NAME: "PluginsForALM_GitLab"
_SOURCE_ENV_URL: "https://org.crm.dynamics.com"
_TARGET_ENV_URL: "https://org.crm.dynamics.com"
_TENANT_ID: "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
_CLIENT_ID: "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
_CLIENT_SECRET: "xxxxxxxxxxxxxxxxxxxxxxxxxxxx"
_SOLUTIONS_DIR: "/workspace/Plugins/08112025"
view raw GitLabDemo.yaml hosted with ❤ by GitHub

Solution from Source Environment

Now lets run the trigger which will export the solution from the source environment and import to the target environment….we have manual trigger, automatic trigger whenever there is an commit to the repo in GitLab etc., you may pick whatever suits your needs best.

Solution imported to the target environment using Google Cloud Build

The below table illustrates key differences between Google Cloud Build and Azure Devops….

AspectGoogle Cloud BuildAzure DevOps Build Pipelines
Pricing ModelPay-as-you-go with per-second billingPer-minute billing with tiered pricing
Cost OptimizationSustained use discounts, preemptible VMsReserved capacity and enterprise agreements
Build EnvironmentServerless, container-native, managed by Google CloudRequires self-hosted or paid hosted agents
Free TierAvailable with build minutes and creditsAvailable but more limited
Operational OverheadLow, no need to manage build agentsHigher, managing agents or paying for hosted agents
Ideal ForVariable, short, or containerized workloadsLarge Microsoft-centric organizations
Integration Cost ImpactTightly integrated with Google Cloud serverless infrastructureIntegrated with Microsoft ecosystem but may incur additional licensing costs

Conclusion:

PAC CLI is a powerful command-line tool that automates authentication, environment, and solution management within Power Platform ALM, enabling consistent and repeatable deployment workflows. It integrates smoothly with DevOps tools like GitLab and Google Cloud Build, helping teams scale ALM practices efficiently while maintaining control and visibility over Power Platform environments. Just note, my intention was showcase the power of PAC CLI with wider ecosystem, not only with Microsoft.

Cheers,

PMDY