Gallerist-AzureFunctions

Azure Functions project to process images for the web, using Python 3, Pillow, Gallerist; reading source pictures from an Azure Storage Blob Service, writing resized pictures inside the same storage.

Tested and proved to work with Python 3.6.8 and 3.7.0. Version 3.6.8 is currently the version recommended by Microsoft for Azure Functions 2 (as of today, the 22nd of September 2019).

How to run locally

1. install Azure Functions CLI.
2. create a Python virtual environment
3. install dependencies described in requirements.txt
4. configure storage account name and key in YAML files, or environmental variables as described in gallerist.yaml comments; and also the blob container name (a container must exist already in the blob service)
5. start the functions host from the gallerist folder:

$ func start

## ...
/gallerist (master)$ func start

                  %%%%%%
                 %%%%%%
            @   %%%%%%    @
          @@   %%%%%%      @@
       @@@    %%%%%%%%%%%    @@@
     @@      %%%%%%%%%%        @@
       @@         %%%%       @@
         @@      %%%       @@
           @@    %%      @@
                %%
                %

Seeing it in practice:

To make a manual verification and understand what's offered by this Azure Functions, do the following:

1. configure an Azure Storage account and settings for the function, as described above
2. create a container inside the Blob Service (can be private)
3. upload a picture to the container, for example using Azure Storage Explorer: Gallerist library handles by default JPG, PNG, GIF (including animated GIFs), MPO - supporting more formats require to edit the provided code
4. start the Azure Function as described above
5. hit the endpoint with a picture query string parameter, for example: http://localhost:7071/api/prepare?picture=01.jpg
6. the function generates a medium size picture, and a thumbnail picture, stores them in the same container, and returns useful metadata that can be stored and used by a client to display a front-end, like in the box below

{
  "width": 2071,
  "height": 2895,
  "extension": ".jpg",
  "ratio": 0.7153713298791019,
  "mime": "image/jpeg",
  "versions": [
    {
      "id": "a7ed03d7d3b64dc98bee094d1ade12e5",
      "size_name": "medium",
      "max_side": 1200,
      "file_name": "m-a7ed03d7d3b64dc98bee094d1ade12e5.jpg"
    },
    {
      "id": "ffc5f80e1033424dbf681ad5768a56cf",
      "size_name": "thumbnail",
      "max_side": 200,
      "file_name": "t-ffc5f80e1033424dbf681ad5768a56cf.jpg"
    }
  ]
}

How to run in production

Refer to Microsoft documentation, choose one of the several ways to run Azure Functions 2 (e.g. in Azure Functions service, in Kubernetes using KEDA, or Knative, through Docker).