Wikimedia Commons File Talk Scraper

Scrape discussion pages from Wikimedia Commons file talk pages.

Features

• Filter file talk pages by title prefix, resume from any title, sort A to Z or Z to A
• Extract section headings, comments, timestamps (ISO 8601), and signatures from the page wikitext
• Structured threads: each section with its comments, author, and timestamp
• Linked file details: direct image URL, MIME type, dimensions, and file size
• Last revision ID, date, and editor, plus the templates used on the page (for example deletion-request notices)
• Optional raw wikitext and an option to keep only pages with signed comments
• Configurable maximum items (up to 1,000,000 for paid users)
• Paginated results via the official MediaWiki API, 50 pages per request

Usage

1. Provide a search term in the input to filter file talk pages by title.
2. Set the maximum number of items to scrape.
3. Run the actor and get structured data.

Input

Field	Type	Description
maxItems	Integer	Maximum number of items to scrape. Free users limited to 10.
search	String	Only pages whose file name starts with this text (the "File talk:" prefix is optional).
startFrom	String	Resume the alphabetical listing at this file name.
sortDirection	String	ascending (A to Z, default) or descending (Z to A).
onlyWithComments	Boolean	Skip pages that only hold banners and keep those with at least one signed comment. Default false.
includeWikitext	Boolean	Add the full wikitext of the talk page. Default true.

Output

Each output item contains:

• pageId - Wikimedia page ID
• title - Page title
• namespace - Namespace number (7 for file talk)
• discussionPageUrl - URL to the discussion page
• fileTitle / filePageUrl - The file the discussion is about
• fileExists - Whether the file still exists on Commons
• imageUrl - Direct URL of the original file (null if the file was deleted)
• fileMime, fileWidth, fileHeight, fileSizeBytes - File details
• sectionHeadings - Array of section headings
• comments - Array of comment texts
• timestamps - Array of comment timestamps (ISO 8601)
• signatures - Array of comment authors, one per signed comment
• participants - Unique comment authors
• commentCount - Number of signed comments
• threads - Sections with their comments (text, user, timestamp)
• templates - Templates used on the page
• lastRevisionId, lastEditedAt, lastEditor - Latest revision of the talk page
• pageLengthBytes - Page size
• wikitext - Raw page content (when includeWikitext is on)
• scrapedAt - ISO timestamp of when the item was scraped

Limits

• Free users: maximum 10 items
• Paid users: up to 1,000,000 items

Sign Up

Upgrade to a paid plan for higher limits.