Merge pull request #21796 from Yoast/390-fix-plurals-with-umlaut

Extend exception list for German nouns with umlaut in plural

packages/yoastseo/spec/languageProcessing/languages/de/helpers/internal/determineStemSpec.js

@@ -7,10 +7,6 @@ const morphologyDataDE = getMorphologyData( "de" ).de;
 const wordsToStem = [
 	// Default stemmer
 	[ "studenten", "student" ],
-	// Nouns: exceptionStems
-	[ "vögel", "vogel" ],
-	// Nouns: exceptionStems compound
-	[ "raubvögel", "raubvogel" ],
 	// Nouns: exceptionStems with one plural matching multiple singulars
 	[ "stadium", "stadi" ],
 	[ "stadion", "stadi" ],
@@ -55,8 +51,61 @@ const wordsToStem = [
 	[ "Kraftwerke", "Kraftwerk" ],
 ];
 
-describe( "Test for determining stems for German words", () => {
-	it( "creates stems for German words", () => {
-		wordsToStem.forEach( wordToStem => expect( determineStem( wordToStem[ 0 ], morphologyDataDE ) ).toBe( wordToStem[ 1 ] ) );
+describe.each( wordsToStem )( "Test for determining stems for German words", ( word, stem ) => {
+	it( "stems for German word " + word + " to " + stem, () => {
+		expect( determineStem( word, morphologyDataDE ) ).toBe( stem );
+	} );
+} );
+
+const umlautExceptions = [
+	// A noun that gets umlaut in plural
+	[ "vögel", "vogel" ],
+	[ "läden", "laden" ],
+	// A noun that gets umlaut and an irregular plural dative suffix
+	[ "müttern", "mutter" ],
+	[ "schwägern", "schwager" ],
+	// A noun that gets umlaut and a regular case suffix
+	[ "bädern", "bad" ],
+	[ "ängsten", "angst" ],
+	[ "hände", "hand" ],
+	// A noun that gets umlaut and -e in plural
+	[ "häuse", "haus" ],
+	[ "ängste", "angst" ],
+	// A noun that gets umlaut and -er in plural
+	[  "männer", "mann" ],
+	[ "wörter", "wort" ],
+	// compound noun that gets umlaut in plural
+	[ "raubvögel", "raubvogel" ],
+	// compound noun that gets umlaut and -e in plural
+	[ "landflüchte", "landflucht" ],
+	[ "geschwülst", "geschwulst" ],
+	[ "feuersbrünst", "feuersbrunst" ],
+	[ "hirschbrünft", "hirschbrunft" ],
+	[ "brünst", "brunst" ],
+	[ "lebensbrünst", "lebensbrunst" ],
+	[ "liebesbrünst", "liebesbrunst" ],
+	// More umlaut nouns from all groups
+	[ "schwäger", "schwager" ],
+	[ "schäden", "schaden" ],
+	[ "töchter", "tochter" ],
+	[ "brünst", "brunst" ],
+	[ "brüder", "bruder" ],
+	[ "gärten", "garten" ],
+	[ "gräben", "graben" ],
+	[ "kästen", "kasten" ],
+	[ "mütter", "mutter" ],
+	[ "läden", "laden" ],
+	[ "väter", "vater" ],
+	[ "füchs", "fuchs" ],
+	[ "ärzte", "arzt" ],
+	[ "gäns", "gans" ],
+	[ "häls", "hal" ],
+	[ "äxte", "axt" ],
+	[ "äste", "ast" ],
+];
+
+describe.each( umlautExceptions )( "Test for determining stems for German words with umlauts", ( word, stem ) => {
+	it( "stems for German word with umlaut " + word + " to " + stem, () => {
+		expect( determineStem( word, morphologyDataDE ) ).toBe( stem );
 	} );
 } );

packages/yoastseo/src/languageProcessing/languages/de/helpers/internal/determineStem.js

@@ -1,23 +1,21 @@
 import { flatten } from "lodash";
 import { languageProcessing } from "yoastseo";
-const { flattenSortLength } = languageProcessing;
-
 import { detectAndStemRegularParticiple } from "./detectAndStemRegularParticiple";
 
 import stem from "./stem";
 
+const { flattenSortLength } = languageProcessing;
+
 /**
  * Returns a stem for a word that appears on the noun exception lists.
  *
- * @param {Object}  morphologyDataNouns The German morphology data for nouns.
- * @param {string}  stemmedWord         The stem to check.
+ * @param {array[]}	exceptionList	The exception list to check.
+ * @param {string}	stemmedWord		The stem to check.
  *
  * @returns {string|null} The stemmed word or null if none was found.
  */
-const findStemOnNounExceptionList = function( morphologyDataNouns, stemmedWord ) {
-	const exceptionStems = morphologyDataNouns.exceptionStems;
-
-	for ( const exceptionStemSet of exceptionStems ) {
+const findStemOnNounExceptionList = function( exceptionList, stemmedWord ) {
+	for ( const exceptionStemSet of exceptionList ) {
 		const matchedStem = exceptionStemSet.find( exceptionStem => stemmedWord.endsWith( exceptionStem ) );
 
 		if ( matchedStem ) {
@@ -108,14 +106,21 @@ const findStemOnVerbExceptionList = function( morphologyDataVerbs, stemmedWord )
  * @returns {string} Stemmed form of the word.
  */
 export default function determineStem( word, morphologyDataGerman ) {
+	// Already return the stem here if the word contains umlaut and ends with an ending that looks like a valid suffix, e.g. "läden" stemmed to "laden".
+	const umlautException = morphologyDataGerman.nouns.umlautException || [];
+	const findUmlautException = findStemOnNounExceptionList( umlautException, word );
+	if ( findUmlautException ) {
+		return findUmlautException;
+	}
+
 	const verbData = morphologyDataGerman.verbs;
 	const stemmedWord = stem( verbData, word );
 
 	/*
 	 * Goes through the stem exception functions from left to right, returns the first stem it finds.
 	 * If no stem has been found, return the original, programmatically created, stem.
 	 */
-	return findStemOnNounExceptionList( morphologyDataGerman.nouns, stemmedWord ) ||
+	return findStemOnNounExceptionList( morphologyDataGerman.nouns.exceptionStems, stemmedWord ) ||
 		findStemOnAdjectiveExceptionList( morphologyDataGerman.adjectives, stemmedWord ) ||
 		findStemOnVerbExceptionList( verbData, stemmedWord ) ||
 		detectAndStemRegularParticiple( verbData, word ) ||