DOCSEARCHbyAlgolia
Submit Docs

metascraper

repository·master·Indexed 25 days ago

https://github.com/microlinkhq/metascraper

A modular library for extracting unified metadata—such as Open Graph, Twitter Cards, and JSON-LD—from websites. It allows developers to compose custom extraction pipelines using specific rule packages for properties like author, date, description, image, and publisher, as well as integrations for platforms like Amazon, Bluesky, Instagram, and Dribbble.

Tokens
14K
Snippets
44
Records
130
Agent score
83%

What's inside metascraper

  1. Compare Metascraper accuracy with other libraries

    master

    Metascraper achieves higher accuracy (95.54% correct) compared to libraries like html-metadata, node-metainspector, open-graph-scraper, and unfluff because it uses a series of fallbacks for each piece of metadata rather than relying solely on spec-compliant tags like Open Graph. It is specifically tuned for parsing article information.

    Metadata Accuracy Definitions

    When evaluating accuracy, the following criteria are used:

    • author: Correct if in First Last format without extra junk information.
    • date: Correct if it is the correct date (regardless of time) and follows the ISO 8601 format.
    • description: Correct if it is the publisher's chosen description or the first paragraph of the article.
    • image: Correct if it is the publisher's chosen image or the first image on the page.
    • publisher: Correct if it is the publisher's proper name or their domain name.
    • title: Correct if it is the title of the article or the title of the page.
    • url: Correct if it resolves back to the original article.