Notion automated text recognition

This project lets you add text recognition to your Notion through the use of the Microsoft Vision API and Docker.

Preconditions

To get started, you need the following:

Notion account: https://www.notion.so/
Notion integration and API key: https://developers.notion.com/docs/getting-started
The ID of a Notion database that has the Created time or OCR Parsing (checkbox) field (see Properties > Advanced)
Microsoft Azure account: https://azure.microsoft.com/en-gb/free/cognitive-services/
Microsoft Computer Vision API key (see below)

Installation

This project can installed using Docker or Docker Compose.

Docker example

$ docker run --name some-name-for-your-container -e NOTION_TOKEN=secret_12345678 -e DATABASE_ID=d82973h2kwldj20239e1 -e MICROSOFT_API_KEY=9834023jdsadlawdkwn -e MICROSOFT_ENDPOINT=https://[YOUR_NAME].cognitiveservices.azure.com/ -e SCAN_METHOD=checkbox -e SCAN_FREQUENCY=15 michielvanbeers/notion-auto-ocr

Docker compose example

version: '3'

services:
  notion-auto-ocr:
    image: michielvanbeers/notion-auto-ocr
    restart: unless-stopped
    environment:
      - NOTION_TOKEN=secret_12345678
      - DATABASE_ID=d82973h2kwldj20239e1
      - MICROSOFT_API_KEY=9834023jdsadlawdkwn
      - MICROSOFT_ENDPOINT=https://[YOUR_NAME].cognitiveservices.azure.com/
      - SCAN_METHOD=checkbox 
      - SCAN_FREQUENCY=15 # Optional

Environment variables

NOTION_TOKEN: API token to integrate with Notion. Don't forget to allow your intergration access to your database
DATABASE_ID: ID of the database that you want to scan
MICROSOFT_API_KEY: Key to access the Microsofy Computer Vision API
MICROSOFT_ENDPOINT: Url of your API end-point. Can be retrieved from the Azure Portal
SCAN_METHOD: Method used to scan page including new image to parse. Valid value is checkbox or createtime
SCAN_FREQUENCY: Optional parameter to set the frequency of scanning for new images to parse. Leave out to do a single run (recommended for testing)

Usage

The script checks if it can find the 'ocr_text' text in image caption or under an image. If it does, it will send the content of the image to Microsoft OCR API and replace the 'ocr_text' by the result in caption or add the result at the end of the current document. When the 'SCAN_FREQUENCY' environment variable is set, it will check if there are any new pages added since the last scan (through the use of setting timestamps).

Creating a Microsoft API key

This section assumes that you already have a Microsoft Azure account. To create your (free) API resource, take the following steps:

Go to https://portal.azure.com/
Click Create a resource
Search for Computer Vision
Click on Computer Vision
Click Create
Set the following:
- Subscription
- Resource group (or create new one)
- Region (nearest region near you)
- Name (name of your liking)
- Pricing tier (Free F0)
- Click the terms checkbox
Click Review + create

Acknowledgements

This project is heavily inspired by yannick-cw/notion-ocr
Addition of SCAN_METHOD and all replacement in image caption has been developed by Marc GUYARD

To-do

Refactor to OOB setup and different files for easier maintenance
Research possibility into making standalone executable using PyInstaller
Research possibility to integrate into Notion-Enhancer

MichielvanBeers/notion-auto-ocr