azure-functions-python-worker/docs/index.rst

467 строки
9.6 KiB
ReStructuredText

.. image:: https://travis-ci.org/Azure/azure-functions-python-worker.svg?branch=master
:target: https://travis-ci.org/Azure/azure-functions-python-worker
==========================
Azure Functions for Python
==========================
Requirements
============
Azure Functions for Python support Python 3.6 or later.
Programming Model
=================
An Azure function is implemented as a global Python function ``main()`` in the
file called ``__init__.py``. The name of the Python function can be changed by
specifying the ``entryPoint`` attribute in ``function.json``, and the name of
the file can be changed by specifying the ``scriptFile`` attribute in
``function.json``.
Currently, the Azure function and its bindings must be declared in the
``function.json`` file. Optionally, the function parameters and the
return type may also be declared as Python type annotations. The annotations
must match the types expected by the bindings declared in ``function.json``.
Below is an example of a simple function triggerred by an HTTP request.
``function.json``:
.. code-block:: json
{
"scriptFile": "__init__.py",
"bindings": [
{
"authLevel": "anonymous",
"type": "httpTrigger",
"direction": "in",
"name": "req"
},
{
"type": "http",
"direction": "out",
"name": "$return"
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions
def main(req: azure.functions.HttpRequest) -> str:
user = req.params.get('user', 'User')
return f'Hello, {user}!'
The annotations are optional, so the function may also be written as:
.. code-block:: python
def main(req):
user = req.params.get('user', 'User')
return f'Hello, {user}!'
Logging
=======
Azure Functions adds a root :mod:`logging <python:logging>` handler
automatically, and any log output produced using the standard logging output
is captured by the Functions runtime.
Context
=======
A function can obtain the invocation context by including the special
``context`` argument in its signature. The context is passed as a
:class:`Context <azure.functions.Context>` instance:
.. code-block:: python
import azure.functions
def main(req: azure.functions.HttpRequest,
context: azure.functions.Context) -> str:
return f'{context.invocation_id}'
Bindings
========
Azure Functions for Python supports the following binding types:
* :ref:`HTTP and webhooks <azure-bindings-http>`: trigger, output;
* :ref:`Blob storage <azure-bindings-blob>`: trigger, input, output;
* :ref:`Queue <azure-bindings-queue>`: trigger, output;
* :ref:`Timers <azure-bindings-timer>`: trigger.
.. _azure-bindings-http:
HTTP and webhook bindings
-------------------------
The trigger binding is passed as a
:class:`HttpRequest <azure.functions.HttpRequest>` object. Output bindings
can be returned as a ``str`` or an
:class:`HttpResponse <azure.functions.HttpResponse>` object.
Example
~~~~~~~
``function.json``:
.. code-block:: json
{
"scriptFile": "__init__.py",
"bindings": [
{
"authLevel": "anonymous",
"type": "httpTrigger",
"direction": "in",
"name": "req"
},
{
"type": "http",
"direction": "out",
"name": "$return"
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions
def main(req: azure.functions.HttpRequest) -> str:
user = req.params.get('user', 'User')
return f'Hello, {user}!'
.. _azure-bindings-blob:
Blob storage bindings
---------------------
The trigger and input bindings are passed as
:class:`InputStream <azure.functions.InputStream>` instances. Output can
be a ``bytes``, ``str`` or a :term:`file-like object <python:file object>`.
Blob storage trigger example
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
``function.json``:
.. code-block:: json
{
"bindings": [
{
"type": "blobTrigger",
"direction": "in",
"name": "file",
"connection": "AzureWebJobsStorage",
"path": "file.txt"
},
{
"type": "http",
"direction": "out",
"name": "$return",
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions
def main(file: azure.functions.InputStream) -> bytes:
return file.read()
Blob storage output example
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
``function.json``:
.. code-block:: json
{
"bindings": [
{
"authLevel": "anonymous",
"type": "httpTrigger",
"direction": "in",
"name": "req"
},
{
"type": "blob",
"direction": "out",
"name": "file",
"connection": "AzureWebJobsStorage",
"path": "test-file.txt"
},
{
"type": "http",
"direction": "out",
"name": "$return",
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions
def main(req: azure.functions.HttpRequest,
file: azure.functions.Out[bytes]) -> azure.functions.HttpResponse:
# write the request body into the output blob
file.set(req.get_body())
return azure.functions.HttpResponse(
content_type='application/json',
body='{"status": "OK"}'
)
Note that in the above example we use the :class:`Out <azure.functions.Out>`
interface to set the value of the output binding.
.. _azure-bindings-queue:
Queue storage bindings
----------------------
Queue storage trigger bindings are passed as
:class:`QueueMessage <azure.functions.QueueMessage>` instances. Output
bindings can be returned as a ``str``, ``bytes`` or a
:class:`QueueMessage <azure.functions.QueueMessage>` instance.
Example
~~~~~~~
``function.json``:
.. code-block:: json
{
"scriptFile": "__init__.py",
"bindings": [
{
"type": "queueTrigger",
"direction": "in",
"name": "msg",
"queueName": "inputqueue",
"connection": "AzureWebJobsStorage",
},
{
"type": "blob",
"direction": "out",
"name": "$return",
"queueName": "outputqueue",
"connection": "AzureWebJobsStorage",
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions
def main(
msg: azure.functions.QueueMessage) -> azure.functions.QueueMessage:
body = msg.get_body()
# ... process message
# Put a message into the output queue signaling that this message
# was processed.
return azure.functions.QueueMessage(
body=f'Processed: {msg.id}'
)
.. _azure-bindings-timer:
Timer bindings
--------------
Timer trigger bindings are passwd as
:class:`TimerRequest <azure.functions.TimerRequest>` instances.
Example
~~~~~~~
``function.json``:
.. code-block:: json
{
"scriptFile": "__init__.py",
"bindings": [
{
"type": "timerTrigger",
"direction": "in",
"name": "timer",
"schedule": "*/5 * * * * *"
},
{
"type": "blob",
"direction": "out",
"name": "$return",
"queueName": "outputqueue",
"connection": "AzureWebJobsStorage",
}
]
}
``__init__.py``:
.. code-block:: python
import datetime
import azure.functions
def main(timer: azure.functions.TimerRequest) -> str:
# process timer event...
# put the current timestamp into the output queue.
return f'{datetime.datetime.now().timestamp()}'
.. _azure-bindings-cosmosdb:
CosmosDB Bindings
-----------------
The trigger and input CosmosDB bindings are passed as
:class:`DocumentList <azure.functions.DocumentList>` instances. Output can
be a :class:`Document <azure.functions.Document>` instance, a
:class:`DocumentList <azure.functions.DocumentList>` instance or an iterable
containing ``Document`` instances.
CosmosDB Trigger Example
~~~~~~~~~~~~~~~~~~~~~~~~
``function.json``:
.. code-block:: json
{
"bindings": [
{
"direction": "in",
"type": "cosmosDBTrigger",
"name": "docs",
"databaseName": "test",
"collectionName": "items",
"leaseCollectionName": "leases",
},
{
"type": "http",
"direction": "out",
"name": "$return",
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions as func
def main(docs: func.DocumentList) -> str:
return docs[0].to_json()
CosmosDB Output Example
~~~~~~~~~~~~~~~~~~~~~~~
``function.json``:
.. code-block:: json
{
"scriptFile": "__init__.py",
"bindings": [
{
"authLevel": "anonymous",
"type": "httpTrigger",
"direction": "in",
"name": "req"
},
{
"direction": "out",
"type": "cosmosDB",
"name": "doc",
"databaseName": "test",
"collectionName": "items",
"leaseCollectionName": "leases",
"createIfNotExists": true
},
{
"direction": "out",
"name": "$return",
"type": "http"
}
]
}
``__init__.py``:
.. code-block:: python
import azure.functions as func
def main(req: func.HttpRequest, doc: func.Out[func.Document]):
doc.set(func.Document.from_json(req.get_body()))
return 'OK'
Reference
---------
:ref:`Azure Functions for Python Reference <azure-functions-reference>`.
.. toctree::
:maxdepth: 2
:hidden:
usage
api