Wikipedia Summary (API-first) Actor

Overview

This actor fetches structured summaries from Wikipedia using only its official public APIs. You provide a topic, and it returns the article's title, a short extract, and the canonical URL. It's designed for reliability in automated workflows: if a page isn't found, the run still succeeds with a NOT_FOUND status instead of failing.

Key Features

• API-Only & Reliable: Uses Wikipedia's official REST API (/api/rest_v1/page/summary/) and search API. No HTML scraping or bypassing of protections.
• Deterministic Output: Every run finishes with SUCCEEDED. Missing pages return a clear NOT_FOUND status.
• Fast & Lightweight: Typical runtime is 0.2–1.2 seconds with minimal memory (256–512 MB).
• Safe for Concurrency: Can be run in parallel; it doesn't use sessions or cookies.
• Input Flexibility: Handles topic redirects and performs a search fallback if the exact page isn't found.

How to Use

1. Run the actor and provide a JSON input object.
2. The main results are in:
   • The Dataset as a list of records.
   • The Key-Value Store as OUTPUT.json.
3. Check the logs for resolution details or warnings.

Input

The actor requires a single topic parameter.

{
  "topic": "Web scraping"
}

• topic (string, required): The article subject or title. Wikipedia handles redirects and title normalization.

Output

The actor returns a consistent JSON object. Key fields are status, resolved_title, url, title, extract, and lang.

Example Output (FOUND):

{
  "input_topic": "Web scraping",
  "status": "FOUND",
  "resolved_title": "Web scraping",
  "url": "https://en.wikipedia.org/wiki/Web_scraping",
  "title": "Web scraping",
  "extract": "Web scraping is data scraping used for extracting data from websites...",
  "lang": "en",
  "timestamp": "2025-09-11T03:00:00.000Z"
}

Example Output (NOT_FOUND):

{
  "input_topic": "Apify",
  "status": "NOT_FOUND",
  "resolved_title": null,
  "url": null,
  "title": null,
  "extract": null,
  "lang": "en",
  "timestamp": "2025-09-11T03:00:00.000Z"
}

Use Cases

• Enriching data with short descriptions (e.g., for technologies, cities, companies).
• Generating preview cards or snippets for search UIs.
• Normalizing topic names before further processing.

Notes

• Fair Use: Only public endpoints are used. It does not bypass authentication, captchas, or paywalls.
• Data Licensing: Summaries are from Wikipedia; comply with its licensing terms for downstream use.
• Support: Provide feedback via comments on finished runs.