Extract URLs from Text

The Extract URLs tool pulls every URL out of any pasted text — HTML source, log files, email threads, documents, code files, chat transcripts. Every developer and content worker hits this problem: an HTML source has dozens of links you want to audit, a log file contains request URLs for traffic analysis, an email thread has shared links you want to collect, a content export has URLs embedded in free text fields. Manual copying is slow; this tool extracts all of them in milliseconds.

Output is deduplicated and optionally validated. The regex catches http://, https://, ftp://, and other common schemes, plus www. links that omit the protocol. Duplicate URLs across the text are collapsed into unique entries (with a count of how many times each appeared). Optional cleanup removes URL parameters, fragments, and trailing punctuation that often accompanies URLs in sentences (\"visit https://example.com.\" → the URL is https://example.com, not https://example.com.). Export formats include plain list, CSV with counts, JSON array, or a sitemap-style XML. All processing runs locally in your browser — no request is made to any URL, no data is sent to a server.

URL extraction regex needs to handle several edge cases correctly.

The core pattern: (https?|ftp):\\/\\/[^\\s<>\"\\(\\)]+

Schemes: http, https, ftp are the common ones. mailto, tel, and other custom schemes can be added via options. The tool also catches www.-prefixed URLs (lacking explicit protocol) by default for common web text.

Path and query: everything up to whitespace, angle brackets, quotes, or parentheses is treated as part of the URL. Punctuation at the very end is typically trimmed (periods, commas, semicolons, colons) because URLs in prose often end right before sentence punctuation.

Trailing punctuation heuristic: URLs rarely end in a period or comma, so the tool strips these by default. But some URLs legitimately contain trailing slashes or punctuation-like characters. The heuristic is configurable.

Deduplication: normalize trailing slashes (https://example.com and https://example.com/ treated as same) and case on scheme/host (HTTPS://Example.COM and https://example.com same). Case-sensitivity of the path is preserved because URLs are case-sensitive below the host level per RFC 3986.

Fragment and query handling: by default, fragments (#section) and query strings (?utm_source=...) are preserved. Options strip them for canonical URL extraction where you want to see unique resources rather than unique parameterized views.

Anchor tag extraction: for HTML input, parsing <a href=\"...\"> with a DOM parser is more robust than regex. The tool offers HTML mode that uses DOMParser for accurate anchor extraction (including relative URLs resolved to absolute if a base URL is provided).

Shortened URLs: the tool extracts but does not resolve short URLs (bit.ly, t.co, etc.). For resolution you need HTTP requests which are out of scope for a client-side extractor.

Performance: regex extraction on multi-megabyte text runs in milliseconds. HTML parsing is slower but still fast for normal documents.

Compared to Unix grep with a URL regex, this tool has a visual UI, automatic deduplication, and multiple export formats. grep wins for scripting and log processing; this tool wins for interactive extraction from pasted text.

Compared to a custom script in Python or JavaScript, this tool is faster when you just need a one-time extraction. For automated pipelines, write code; for right-now extraction, this tool is faster to use.

Compared to specialized SEO crawlers (Screaming Frog, Sitebulb), this tool extracts URLs from text you already have rather than crawling a live site. Crawlers work at a different level of the stack; this tool is for working with captured content.