https://github.com/phongcao/pdf-azure-blob-extracter

This notebook reads a PDF file, extracts its images and content and upload them to Azure Blob storage.
https://github.com/phongcao/pdf-azure-blob-extracter

azure blob openai parser pdf

Last synced: about 1 month ago
JSON representation

This notebook reads a PDF file, extracts its images and content and upload them to Azure Blob storage.

• Host: GitHub
• URL: https://github.com/phongcao/pdf-azure-blob-extracter
• Owner: phongcao
• License: agpl-3.0
• Created: 2023-07-19T15:46:33.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-07-27T19:31:57.000Z (over 1 year ago)
• Last Synced: 2024-10-27T12:32:08.427Z (3 months ago)
• Topics: azure, blob, openai, parser, pdf
• Language: Jupyter Notebook
• Homepage:
• Size: 20.5 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# pdf-azure-blob-extracter

This notebook reads a PDF file, extracts its images and content and upload them to

Azure Blob storage.

After running, you can find the following blobs on Azure Blob storage:

```

[Container]/[PDF file name without ext]/content.md

[Container]/[PDF file name without ext]/img_1.png

[Container]/[PDF file name without ext]/img_2.png

[Container]/[PDF file name without ext]/img_3.png

...

```

In the content.md file, the embedded images are extracted and replaced by file names:

```

Text 1

![img_1.png](img_1.png)

Text 2

![img_2.png](img_2.png)

Text 3

![img_3.png](img_3.png)

```

It's easy to manipulate those image links to inject

[SAS tokens](https://learn.microsoft.com/en-us/azure/storage/common/storage-sas-overview)

so that they can be fully rendered.

## How to run

1. Copy the `.env.template` file to `.env` and fill in the required info.

2. Upgrade pip: `pip install --upgrade pip`.

3. Install dependencies: `pip install -r requirements.txt`.

4. Modify the input file name, default is `userguide.pdf`.

5. Modify the output file name, default is `userguide.md`.

6. Modify footer and header font size under `Special setttings` if needed.

7. Run the notebook and check the blob storage for the extracted images and content.