diff --git a/packages/lib/helpers.js b/packages/lib/helpers.js
index c3fe23fb..d2157d37 100644
--- a/packages/lib/helpers.js
+++ b/packages/lib/helpers.js
@@ -12,6 +12,7 @@
  *
  * @param {*} value - value to return the type of.
  * @returns {string} basic type of the value passed.
+ * @static
  */
 function basicType(value) {
 	if (value === null) { return "null"; }
@@ -24,6 +25,7 @@ function basicType(value) {
  * Asynchronously wait for the given duration
  *
  * @param {number} duration - Time to wait for in milliseconds.
+ * @static
  */
 async function wait(duration) {
 	await new Promise(resolve => { setTimeout(resolve, duration); });
@@ -36,6 +38,7 @@ async function wait(duration) {
  * @param {Promise} promise - Promise to wait for.
  * @param {number} time - Maximum time im milliseconds to wait for.
  * @param {*=} timeoutResult - Value to return if the operation timed out.
+ * @static
  */
 async function timeout(promise, time, timeoutResult) {
 	let timer;
@@ -61,6 +64,7 @@ async function timeout(promise, time, timeoutResult) {
  *
  * @param {Readable} stream - byte stream to read to the end.
  * @returns {Buffer} content of the stream.
+ * @static
  */
 async function readStream(stream) {
 	let chunks = [];
@@ -85,6 +89,7 @@ async function readStream(stream) {
  *
  * @param {string} text - Text to escape RegExp meta chars in.
  * @returns {string} escaped text.
+ * @static
  */
 function escapeRegExp(text) {
 	return text.replace(/[-[\]{}()*+?.,\\^$|#\s]/g, "\\$&");
@@ -96,6 +101,7 @@ function escapeRegExp(text) {
  * Shortens a large number of bytes using the kB/MB/GB/TB prefixes.
  * @param {number} bytes - Count of bytes to format.
  * @returns {string} formatted text.
+ * @static
  */
 function formatBytes(bytes) {
 	if (bytes === 0) {
@@ -108,6 +114,169 @@ function formatBytes(bytes) {
 	return (power > 0 ? (bytes / factor ** power).toFixed(2) : bytes) + units[power];
 }
 
+function parseSearchIdentifier(pos, input) {
+	// character = ? all characters excluding space : " ?
+	// identifier = 1*character
+	let startPos = pos;
+	while (pos < input.length && ![" ", '"', ":"].includes(input.charAt(pos))) {
+		pos += 1;
+	}
+	return [pos, input.slice(startPos, pos)];
+}
+
+function parseSearchWord(pos, input) {
+	// non quote character = ? all characters excluding " ?
+	// word op = "-"
+	// word = [ word op ], ( identifier | '"', *non quote character, '"' )
+	let word = {
+		type: "word",
+	};
+	if (input.charAt(pos) === "-") {
+		word.exclude = true;
+		pos += 1;
+	}
+	if (input.charAt(pos) === '"') {
+		pos += 1;
+		let startPos = pos;
+		let endPos = pos;
+		while (pos < input.length && input.charAt(pos) !== '"') {
+			pos += 1;
+			endPos += 1;
+		}
+		if (input.charAt(pos) === '"') {
+			pos += 1;
+		}
+		word.value = input.slice(startPos, endPos);
+	} else {
+		([pos, word.value] = parseSearchIdentifier(pos, input));
+	}
+	return [pos, word];
+}
+
+function parseSearchTerm(pos, input, attributes, issues) {
+	// attribute = identifier, ':', word
+	// term = word | attribute
+	let term;
+	if (["-", '"'].includes(input.charAt(pos))) {
+		([pos, term] = parseSearchWord(pos, input));
+	} else {
+		let identifier;
+		([pos, identifier] = parseSearchIdentifier(pos, input));
+		if (input.charAt(pos) === ":") {
+			pos += 1;
+			identifier = identifier.toLowerCase();
+			if (!Object.prototype.hasOwnProperty.call(attributes, identifier)) {
+				issues.push(`Unregonized attribute "${identifier}", use quotes to escape colons`);
+			} else if (attributes[identifier] === "word") {
+				let value;
+				([pos, value] = parseSearchWord(pos, input));
+				term = { type: "attribute", name: identifier, value };
+			} else {
+				throw new Error(`Bad attribute format ${attributes[identifier]} for ${identifier}`);
+			}
+		} else {
+			term = { type: "word", value: identifier };
+		}
+	}
+	return [pos, term];
+}
+
+/**
+ * Match a parsed word term with the given texts.
+ *
+ * Returns true if the passed word term in the search matches at least one
+ * of the passed text snippets. If the word mode is exclude then returns
+ * true if none of the passed text snippets match.
+ *
+ * @param {lib/helpers.ParsedTerm} word - Word to match.
+ * @param {...string} texts - Text to match word in.
+ * @returns {boolean} true if the word matches.
+ */
+function wordMatches(word, ...texts) {
+	if (word.type !== "word") {
+		throw Error("wordMatches: parameter is not a word");
+	}
+	let matches = false;
+	for (let text of texts) {
+		if (text.includes(word.value)) {
+			matches = true;
+			break;
+		}
+	}
+	if (word.exclude) {
+		matches = !matches;
+	}
+	return matches;
+}
+
+/**
+ * @typedef {object} module:lib/helpers~ParsedTerm
+ * @property {string} type - Type of term, either attribute or word.
+ * @property {string=} name - attribute only: Name of attribute.
+ * @property {boolean=} exclude - word only: exclude results with this word.
+ * @property {string | module:lib/helpers.ParsedTerm} value -
+ *     Parsed value of this term. Is a string if type is word, and a
+ *     ParsedTerm if type is attribute.
+ */
+
+/**
+ * @typedef {object} module:lib/helpers~ParsedSearch
+ * @property {Array<module:lib/helpers~ParsedTerm>} terms -
+ *     Parsed result of search terms.
+ * @property {Array<string>} issues -
+ *     Issues detected while parsing the seach string.
+ */
+
+/**
+ * Parse a search string with optional attributes
+ *
+ * Parses the given input as a search expression consisting of space
+ * delimited words and attributes:word pairs.  For aattributes to be
+ * recognized they need to be passed as name: "word" in the attributes
+ * parameter.
+ *
+ * Words in the search expression can optionally be prefixed by a - to
+ * search for results that does not contain that word. E.g. -author contains
+ * all results not matching author.
+ *
+ * If parsing fails an string describing the problem will be added to the
+ * issues array returned and parsing will resume at some arbitrary point.
+ * The issues should be shown to the end user so that they can correct their
+ * search.
+ *
+ * @param {string} input - Search expression to parse.
+ * @param {Object<string, string>} attributes -
+ *     Recognized attributes and their format. Currently only word is
+ *     supported.
+ * @returns {module:lib/helpers~ParsedSearch} parsed terms of the search.
+ * @static
+ */
+function parseSearchString(input, attributes = {}) {
+	// whitespace = 1*" "
+	// search = [ term, *( [ whitespace ], term ) ]
+	input = input.trim();
+	let parsed = {
+		terms: [],
+		issues: [],
+	};
+	let pos = 0;
+	while (pos < input.length) {
+		let term;
+		([pos, term] = parseSearchTerm(pos, input, attributes, parsed.issues));
+		if (term) {
+			parsed.terms.push(term);
+		}
+		while (pos < input.length && input.charAt(pos) === " ") {
+			pos += 1;
+		}
+	}
+	// istanbul ignore if (should not be possible)
+	if (pos !== input.length) {
+		throw new Error(`parse search ended at ${pos} which is not the end of the input (${input.length})`);
+	}
+	return parsed;
+}
+
 
 module.exports = {
 	basicType,
@@ -116,4 +285,6 @@ module.exports = {
 	readStream,
 	escapeRegExp,
 	formatBytes,
+	wordMatches,
+	parseSearchString,
 };
diff --git a/test/lib/helpers.js b/test/lib/helpers.js
index 5b5159f7..3d4b89e3 100644
--- a/test/lib/helpers.js
+++ b/test/lib/helpers.js
@@ -43,4 +43,113 @@ describe("lib/helpers", function() {
 			), "timeout");
 		});
 	});
+
+	function parse(input, attributes) {
+		const result = libHelpers.parseSearchString(input, attributes);
+		if (!result.issues.length) {
+			return result.terms;
+		}
+		return result;
+	}
+
+	describe("parseSearchString()", function() {
+		function word(value, opts = {}) {
+			return { type: "word", value, ...opts };
+		}
+		function attr(name, value) {
+			return { type: "attribute", name, value: typeof value === "string" ? word(value) : value };
+		}
+		it("should split words", function() {
+			assert.deepEqual(parse(""), [].map(word));
+			assert.deepEqual(parse("word"), ["word"].map(word));
+			assert.deepEqual(parse("some words here"), ["some", "words", "here"].map(word));
+		});
+		it("should trim whitespace", function() {
+			assert.deepEqual(parse("   "), [].map(word));
+			assert.deepEqual(parse("word  "), ["word"].map(word));
+			assert.deepEqual(parse("   some  words    here"), ["some", "words", "here"].map(word));
+		});
+		it("should parse negated words", function() {
+			assert.deepEqual(parse("this -that"), [word("this"), word("that", { exclude: true })]);
+			assert.deepEqual(parse("not-negated"), [word("not-negated")]);
+		});
+		it("should not split quoted words", function() {
+			assert.deepEqual(parse('"some quoted words" here'), ["some quoted words", "here"].map(word));
+			assert.deepEqual(parse('"a quoted string"'), ["a quoted string"].map(word));
+			assert.deepEqual(parse('words and "a quoted string"'), ["words", "and", "a quoted string"].map(word));
+		});
+		it("should split on quotation mark", function() {
+			assert.deepEqual(parse('"quoted"unquoted'), ["quoted", "unquoted"].map(word));
+			assert.deepEqual(parse('"quoted""double"'), ["quoted", "double"].map(word));
+			assert.deepEqual(parse('word"with"quote'), ["word", "with", "quote"].map(word));
+			assert.deepEqual(
+				parse('word-"with"-negation'),
+				[word("word-"), word("with"), word("negation", { exclude: true })]
+			);
+		});
+		it("should allow colons in quotes", function() {
+			assert.deepEqual(parse('"word:value"'), ["word:value"].map(word));
+		});
+		it("should parse attribute:value", function() {
+			assert.deepEqual(parse("author:Me", { author: "word" }), [attr("author", "Me")]);
+			assert.deepEqual(parse("word author:Me", { author: "word" }), [word("word"), attr("author", "Me")]);
+			assert.deepEqual(
+				parse("author:Me and author:You", { author: "word" }),
+				[attr("author", "Me"), word("and"), attr("author", "You")]
+			);
+			assert.deepEqual(
+				parse("author:Me"),
+				{
+					terms: [word("Me")],
+					issues: ['Unregonized attribute "author", use quotes to escape colons'],
+				}
+			);
+			assert.deepEqual(
+				parse("constructor:value"),
+				{
+					terms: [word("value")],
+					issues: ['Unregonized attribute "constructor", use quotes to escape colons'],
+				}
+			);
+		});
+		it("should throw on bad attribute format", function() {
+			assert.throws(
+				() => parse("author:Me", { author: "bad" }),
+				new Error("Bad attribute format bad for author")
+			);
+		});
+		it("should ignore attribute case", function() {
+			assert.deepEqual(parse("Author:Me", { author: "word" }), [attr("author", "Me")]);
+			assert.deepEqual(parse("AUTHOR:Me", { author: "word" }), [attr("author", "Me")]);
+			assert.deepEqual(parse("auThOR:Me", { author: "word" }), [attr("author", "Me")]);
+		});
+		it("should parse partial inputs", function() {
+			assert.deepEqual(parse("-"), [word("", { exclude: true })]);
+			assert.deepEqual(parse('"'), [""].map(word));
+			assert.deepEqual(parse('"word'), ["word"].map(word));
+			assert.deepEqual(parse("author:", { author: "word" }), [attr("author", "")]);
+			assert.deepEqual(parse("author:-", { author: "word" }), [attr("author", word("", { exclude: true }))]);
+		});
+	});
+	describe("wordMatches()", function() {
+		const match = libHelpers.wordMatches;
+		it("should match any inputs", function() {
+			assert(match(parse("word")[0], "a word in a string"), "one text");
+			assert(!match(parse("word")[0], "a string without"), "one text no match");
+			assert(match(parse("word")[0], "no match", "no things", "a word", "bad text", ""), "five text");
+			assert(!match(parse("unword")[0], "no match", "no things", "a word", "bad text", ""), "five text no match");
+		});
+		it("should invert match if using exclude", function() {
+			assert(!match(parse("-word")[0], "a word in a string"), "one text");
+			assert(match(parse("-word")[0], "a string without"), "one text no match");
+			assert(!match(parse("-word")[0], "no match", "no things", "a word", "bad text", ""), "five text");
+			assert(match(parse("-unword")[0], "no match", "no things", "a word", "bad text", ""), "five text no match");
+		});
+		it("should throw on bad input", function() {
+			assert.throws(
+				() => match(parse("attr:value", { attr: "word" })[0], "a string"),
+				new Error("wordMatches: parameter is not a word"),
+			);
+		});
+	});
 });