AI Interpolator Unstructured
Last updated on
7 June 2024
The Unstructured is a context filler plugin, that takes txt, pdf, pptx, docx, jpg, md and some other file format and extracts the content structured in a text field.
Module link: www.drupal.org/project/unstructured
Installation
- Use Composer to get the module: "composer require 'drupal/unstructured"
- Install the AI Interpolator Unstructured like any other module.
- Visit /admin/config/convertapi/settings and paste the API token and/or host, depending on how your are running the Unstructured via a self hosted server or SaaS.
How to use
In this use case we will extract the content from a Word file.
- Create a new node called "Article"
- On the "Article" you generate a File field that you name "File".
- Make sure to only enable the extension docx
- On the "Article" you generate Text (formatted, long) field that you name "Contents"
- If you scroll down on the settings page you should now see a check box titled "Enable AI Interpolator". Check that one.
- In the "Choose AI Interpolato Rule" field select "Unstructured API: File to String (Unstructured)".
- In the "Interpolator Base Field" you choose "File" as the file field to scrape.
- Save the settings.
- Now try generating an Article and upload a Word file.
- After saving the Article should have a the content from the file in the text field.
Example ideas of workflows that could be done with the ConvertAPI plugin.
- For using an a PDF, Word or Excel as a context for any of the other workflows described.
- This means that you can generate a story or a podcast or an article header text or another image from the document you uploaded.
You can also find videos of OpenAI Workflows under workflows-of-ai.com
Available Rules
| Rule Name | Description | Field Types |
|---|---|---|
| Unstructured API: File to Images | This rules takes and PDF or Image document and extract all image pieces into an image field. | Image |
| Unstructured API: File to String | Takes txt, pdf, pptx, docx, jpg, md and some other file format and extracts the content structured in a text field. | Text (plain, long) Text (formatted, long) |
| Unstructured API: File to Tables | This rules takes Excel sheets or tables inside Word, Markdown, Images, Powerpoint and extracts them into a TableField | TableField |
Help improve this page
Page status: No known problems
You can:
You can:
- Log in, click Edit, and edit this page
- Log in, click Discuss, update the Page status value, and suggest an improvement
- Log in and create a Documentation issue with your suggestion
