Extension 01 / kotmark-ext-autolink

Autolink

Autolinks are converted into ordinary link nodes, so the feature works with every renderer.

Artifact: io.heapy.kotmark:kotmark-ext-autolink:0.1.0
Package: io.heapy.kotmark.ext.autolink

Usage

Add the extension module and register it with the parser.

Detect links while parsing

import io.heapy.kotmark.ext.autolink.AutolinkType
import io.heapy.kotmark.ext.autolink.autolink
import io.heapy.kotmark.extensions
import io.heapy.kotmark.parser.parser
import io.heapy.kotmark.renderer.html.htmlRenderer

fun renderWithAutolinks(markdown: String): String {
    val extensions = extensions {
        autolink { linkTypes = setOf(AutolinkType.URL, AutolinkType.EMAIL) }
    }
    // Autolinks are found while parsing; the HTML renderer sees ordinary links.
    val document = parser { extensions(extensions) }.parse(markdown)
    return htmlRenderer().render(document)
}

Example

INPUT · Markdown

Visit https://example.com or write to team@example.com.

OUTPUT · HTML

<p>Visit <a href="https://example.com">https://example.com</a> or write to <a href="mailto:team@example.com">team@example.com</a>.</p>

Behavior

Code spans and code blocks are not scanned.
Only lowercase www is recognized.
The standalone autolink artifact also provides link extraction for plain text.

GFM differences

The extractor follows autolink-java rules. These differ from GFM in three examples from its Autolinks specification section:

Example	Difference
5	An unmatched closing parenthesis stops the link even when more URL text follows.
6	`&hl` remains inside the link while its final semicolon stays outside.
11	An email prefix before a trailing `-` or `_` is linked.

Extract plain-text links

The standalone autolink module extracts spans without parsing Markdown.

Extract URL, www, and email spans

import io.heapy.kotmark.autolink.LinkType
import io.heapy.kotmark.autolink.linkExtractor

fun findLinks(text: String): List<String> {
    val extractor = linkExtractor {
        linkTypes = setOf(LinkType.URL, LinkType.WWW, LinkType.EMAIL)
    }
    return extractor.extractLinks(text).map { span ->
        "${span.type}: ${text.substring(span.beginIndex, span.endIndex)}"
    }
}

INPUT · Plain text

See www.kotlinlang.org and https://example.com.

OUTPUT · Link spans

WWW: www.kotlinlang.org
URL: https://example.com