Regex to Extract Markdown Links — [text](url) Pattern

Extracting Markdown Links

Markdown’s inline link syntax is [text](url "optional title"). Reference links use [text][ref] with a definition elsewhere. Both can be extracted with regex when content is well-formed.

Inline Links

\[(?<text>[^\]]+)\]\((?<url>[^\s\)]+)(?:\s+"(?<title>[^"]*)")?\)

Captures three named groups: text, url, and optional title.

Tested Examples

Input	text	url	title
[Google](https://google.com)	Google	https://google.com	—
[Docs](https://x.io "Documentation")	Docs	https://x.io	Documentation
[link](/relative/path)	link	/relative/path	—
[empty]()	— (matches but url is empty)	—	—

Reference Links

\[(?<text>[^\]]+)\]\[(?<ref>[^\]]+)\]

Matches [click][ref-id].

Reference Definitions

^\s*\[(?<id>[^\]]+)\]:\s*(?<url>\S+)(?:\s+"(?<title>[^"]*)")?\s*$

(Use the m flag for multi-line documents.)

Image Syntax

Markdown images are like links with a leading !:

!\[(?<alt>[^\]]*)\]\((?<url>[^\s\)]+)(?:\s+"(?<title>[^"]*)")?\)

Avoid Matching Code Blocks

To skip Markdown links inside fenced code blocks, pre-process the document:

const stripped = md.replace(/```[\s\S]*?```/g, "");
const links = [...stripped.matchAll(/\[([^\]]+)\]\(([^\)]+)\)/g)];

Practical Note

For full CommonMark compliance, use a parser (marked, micromark, markdown-it). Regex works well for link audits, sitemap generation from Markdown docs, and bulk URL replacement.