0.1.0-beta.3487.0 • Published 7 months ago

@or-sdk/markitdowner v0.1.0-beta.3487.0

Weekly downloads
-
License
Apache-2.0
Repository
-
Last release
7 months ago

Markitdowner

The Markitdowner class provides methods for converting PDF documents to Markdown format and generating descriptions for images using various providers and models. It is designed to interact with external services for document and image processing.

Installation

To use the Markitdowner class, install the package:

npm install @or-sdk/markitdowner

Usage

Importing and Instantiating

To use the Markitdowner class, import it into your project:

import { Markitdowner } from '@or-sdk/markitdowner';

const markitdowner = new Markitdowner({
  token: 'your-token',
  discoveryUrl: 'https://api.example.com',
  accountId: 'your-account-id',
  serviceUrl: 'https://service.example.com',
});

Methods

pdfToMarkdown

Converts a PDF document to Markdown format.

Parameters:

  • params (object):
    • url (string): The URL of the PDF document to be converted.
    • include_image_base64 (boolean, optional): Flag to include images as base64 strings in the output.
    • provider (PDFProviders): The provider to use for the PDF processing.
  • options (CallOptions, optional): Additional call options for the request.

Returns:

A promise that resolves to an object containing:

  • pages (array of objects):
    • images (array of strings): Array of image URLs or paths for the page.
    • index (number): The index of the page.
    • markdown (string): The Markdown content of the page.
  • usage_info (object):
    • doc_size_bytes (number): The size of the document in bytes.
    • pages_processed (number): The number of pages processed.

Example:

const markdownResult = await markitdowner.pdfToMarkdown({
  url: 'https://example.com/document.pdf',
  provider: 'pdfProvider',
});
console.log(markdownResult);

imageDescription

Generates a description for an image using the provided data and options.

Parameters:

  • data (object):
    • url (string): The URL of the image to be described.
    • provider (string): The name of the provider for the image processing service.
    • model (string): The model used for generating the image description.
  • options (CallOptions, optional): Additional call options for the request.

Returns:

A promise that resolves to an object containing:

  • content (string): The description of the image.

Example:

const imageDescriptionResult = await markitdowner.imageDescription({
  url: 'https://example.com/image.jpg',
  provider: 'imageProvider',
  model: 'imageModel',
});
console.log(imageDescriptionResult);

Error Handling

The Markitdowner class includes an errorParser to handle errors from API requests. Use the parseError method to parse and handle errors:

try {
  const result = await markitdowner.pdfToMarkdown({
    url: 'https://example.com/document.pdf',
    provider: 'pdfProvider',
  });
} catch (err) {
  const parsedError = markitdowner.parseError(err);
  console.error(parsedError);
}
0.1.0-beta.3487.0

7 months ago