Azure Functions (Python)

Samples / Templates of Azure Function Apps with Python

Local Setup (Debugging)

Command Line

Run the function apps from your command line locally.

1. Make sure Azure Functions Core Tools is installed.
2. Make sure python 3.6+ and required libraries are installed.
3. Make sure you are on the azure branch. Use command func host start to run the function app locally.

VS Code

VS Code IDE provides more control fpr local deployment while providing debugging, repo control, and also deployment to azure portal as integrated features.

1. Install Azure Extensions:

• Azure Account
• Azure CLI Tools
• Azure Functions
• Azure Repos
• Python

2. Open GM Folder

3. Open Command Pallette (Ctrl+Shift+P or F1), and select Azure Functions: Initialize Project for use with VS Code and follow prompts

4. Open Command Pallette (Ctrl+Shift+P or F1), and select Python: Select Interpreter and select either conda environments or relative venv created.

Sample of .vscode\settings.json setting if local folder .venv is selected and set as python interpreter:

		# .vscode\settings.json
	{
	  "azureFunctions.pythonVenv": ".venv",
	  "python.pythonPath": ".venv\\Scripts\\python.exe",
	  "azureFunctions.deploySubpath": "AzureFunction.zip",
	  "azureFunctions.scmDoBuildDuringDeployment": true,
	  "azureFunctions.projectLanguage": "Python",
	  "azureFunctions.projectRuntime": "~2",
	  "azureFunctions.preDeployTask": "func: pack --build-native-deps",
	  "debug.internalConsoleOptions": "neverOpen",
	}

Sample of .vscode\settings.json using conda environment as python interpreter. This assumes a windows environment variable CONDAPATH is defined pointing to conda installation path.
It can be replaced with absolute path of conda which in default case is as follow
CONDAPATH = 'C:\Users\<YOUR USERNAME>\AppData\Local\Continuum\miniconda3\'.

This might require a single time running of conda init powershell command in terminal window of VS Code.

		# .vscode\settings.json
	{
	  "azureFunctions.pythonVenv": "%CONDAPATH%\\envs\\azure",
	  "python.pythonPath": "%CONDAPATH%\\azure\\Scripts\\python.exe",
	  "azureFunctions.deploySubpath": "AzureFunction.zip",
	  "azureFunctions.scmDoBuildDuringDeployment": true,
	  "azureFunctions.projectLanguage": "Python",
	  "azureFunctions.projectRuntime": "~2",
	  "azureFunctions.preDeployTask": "func: pack --build-native-deps",
	  "debug.internalConsoleOptions": "neverOpen",
	}

Shared Code

The first three example listed below demonstrate various arrangements of shared/common code in Azure python functions.

• FuncApp-Http-Py-OptionA
• FuncApp-Http-Py-OptionB
• FuncApp-Http-Py-OptionC

Database Connection

The other example provides the basis for handling Azure SQL servers and databases.

• FuncApp-Http-sql-Example

Note the deployment details mentioned in following sections.

Local Settings with DB (VS Code)

For local code development and debugging, you can define an application setting variable, e.g. 'dbConnectionValue', under 'Values' element of 'local.settings.json'. Note that in example shown below, the variable 'dbConnectionValue' is visible in python code during local deployment and debugging.

  # local.settings.json

  {
    "IsEncrypted": false,
    "AzureWebJobsStorage": "UseDevelopmentStorage=true",
    "Values": {
      "FUNCTIONS_WORKER_RUNTIME": "python",
      "dbConnectionValue": "Driver={ODBC Driver 13 for SQL Server};Server=tcp:<your db server>.database.windows.net,1433;Database=<your db name>;Uid=<your user>@<your db server>;Pwd=<your pass>;Encrypt=yes;TrustServerCertificate=no;Connection Timeout=30;"
    },
    "ConnectionStrings": {
      "sqlConnectionString": "Driver={ODBC Driver 17 for SQL Server};Server=tcp:<your db server>.database.windows.net,1433;Database=<your db name>;Uid=<your user>@<your db server>;Pwd=<your pass>;Encrypt=yes;TrustServerCertificate=no;Connection Timeout=30;"
    }
  }

Environment variable 'SQLAZURECONNSTR_sqlConnectionString' is not available in local deployment (see Azure Git Issue). In order to have a smooth database access after the configuration is completed, string variable for database is available to python scripts and can be accessed as 'SQLAZURECONNSTR_'+ name of connection string (e.g. 'sqlConnectionString' in the above setup):

  # Azure Function:  Python Code
  # Database Connection String access to support both Portal and Local deployment
  try:
      # Azure Portal Deployment
      connStr = "SQLAZURECONNSTR_sqlConnectionString"
      db = pyodbc.connect( os.environ.get(connStr) )

  except Exception as e1:
      try:
          # Local Deployment
          connStr = "dbConnectionValue"
          db = pyodbc.connect( os.environ.get(connStr) )

     except Exception as e2:
            return "%s \n %s" %(str(e1),str(e2)), os.environ.get(connStr)             

  cursor = db.cursor()
  cursor.execute("SELECT * FROM table1")    

Local Deployment/Debugging Settings with DB (VS Code)

For deployment of code developed locally, if you have database connection (e.g. using pyodbc), you want to build dependencies locally to avoid pyodbc module not found error when deployed and triggered via azure portal.

Once the azure function project is initialized for use in VS code, the azureFunctions.preDeployTask parameter in .vscode\settings.json can be modified as

  # .vscode\settings.json
	"azureFunctions.preDeployTask": "func: pack --build-native-deps"

Note that this feature will require a docker daemon (e.g. docker desktop for windows) to be running. The docker needs to be running with linux containers.

Azure Portal Deployment Settings with DB

Warning: Use ODBC Driver 17 ONLY (ODBC Driver 13 is NOT working currently as of October 2010)

Standard process to set connection string for Azure functions is as below.

• Navigate to Function Apps on Azure portal
• Select your Function App named 'Function Name'
• Select 'Platform features' tab
• Select 'Configuration' item
• Press 'New connection string' Under 'Connection strings' section
• Set the string on 'Add/Edit connection string' dialogue opened

Name: "sqlConnectionString"

Value: "Driver={ODBC Driver 17 for SQL Server};Server=tcp:<your db server>.database.windows.net,1433;Database=<your db name>;Uid=<your user>@<your db server>;Pwd=<your pass>;Encrypt=yes;TrustServerCertificate=no;Connection Timeout=30;"

Type: 'SQLAzure'

Warning: The connection string can be obtained from your Azure SQL with ODBC; However, be sure to use ODBC Driver 17 as ODBC Driver 13 is NOT working on azure portal.

• Press 'OK' on 'Add/Edit connection string' dialogue opened
• Press 'Save' button on top of Configuration

After the configuration is completed, the connection string variable for database is available to python scripts and can be accessed as 'SQLAZURECONNSTR_' + name of connection string (e.g. 'sqlConnectionString' in the above setup):

  # Azure Function:  Python Code
  # Database Connection String can be accessed in python as below
  connStr = "SQLAZURECONNSTR_sqlConnectionString"
  db = pyodbc.connect( os.environ.get(connStr) )
  cursor = db.cursor()
  cursor.execute("SELECT * FROM table1")    

Alternatively as discuss below for local deployment, the connection string can be defined as a variable under 'Application settings' and used in python. The steps to add an application setting variable is quite similar to adding connection string except it is initiated with the following:

• Press 'New application setting' Under 'Application settings' section

After the configuration is completed, the application setting variable for database is available to python scripts and can be accessed as name of application setting variable (e.g. 'sqlConnectionValue'):

  # Azure Function:  Python Code
  # Database Connection String can be accessed in python as below
  connStr = "sqlConnectionValue"      # *** Your defined application setting variable ***
  db = pyodbc.connect( os.environ.get(connStr) )
  cursor = db.cursor()
  cursor.execute("SELECT * FROM table1")