std.regex and HTML entities

htmlentities.d

import std.regex;

immutable string[string] namedEntities;

// Construct AA at startup
shared static this()
{
	namedEntities = [
		"gt": ">",
		"lt": "<",
		"acE": "\u223E\u0333"
		// Generate this table from the HTML5 JSON document of named entities
		// Probs best to put it in its own module that is auto-generated in its entirety
		// Can also use string imports to parse the JSON at compile-time, but since the
		// JSON document doesn't change, that's just way overkill
	];
}

string entityToUTF8(Captures!string capture)
{
	import std.algorithm : startsWith;
	import std.conv : text, to; // text helps convert from dchar to string (single code point to UTF-8)

	// capture[0] is whole match, capture[1] is first submatch (the part in parantheses)
	auto submatch = capture[1];

	if(submatch.startsWith('#'))
		return (cast(dchar)submatch[1 .. $].to!uint()).text();
	else
		return namedEntities[submatch];
}

unittest
{
	// Compile regex to machine code at CT
	static entityPattern = ctRegex!`&([#a-zA-Z0-9]+);`;

	auto text = "&lt;3 &#39;apostrophe&#39; &acE;";
	text = text.replaceAll!entityToUTF8(entityPattern);

	assert(text == "<3 'apostrophe' \u223E\u0333");
}