krickert commented on code in PR #1056: URL: https://github.com/apache/opennlp/pull/1056#discussion_r3281242731
########## opennlp-core/opennlp-runtime/src/main/java/opennlp/tools/stopword/StopwordLists.java: ########## @@ -0,0 +1,183 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package opennlp.tools.stopword; + +import java.io.IOException; +import java.io.InputStream; +import java.io.UncheckedIOException; +import java.nio.charset.Charset; +import java.util.Collections; +import java.util.HashMap; +import java.util.LinkedHashSet; +import java.util.Locale; +import java.util.Map; +import java.util.Set; + +import opennlp.tools.util.LanguageCodeValidator; + +/** + * Static factory for {@link StopwordFilter} instances backed by bundled + * language-specific stopword resources or caller-supplied input streams. + * <p> + * Bundled lists ship for the eleven languages enumerated in + * <a href="https://issues.apache.org/jira/browse/OPENNLP-660";>OPENNLP-660</a>: + * Bulgarian (bg), Danish (da), German (de), English (en), Spanish (es), + * Finnish (fi), French (fr), Italian (it), Dutch (nl), Portuguese (pt), + * Russian (ru). Each list is keyed by its ISO 639-1 two-letter code. + */ +public final class StopwordLists { + + private static final String RESOURCE_PATH_PREFIX = "/opennlp/tools/stopword/"; + + private static final Set<String> SUPPORTED_LANGUAGES; + + /** + * ISO 639-2 bibliographic codes that {@link Locale#getISO3Language()} does + * not produce (the JDK returns only the terminologic forms), but which + * {@link LanguageCodeValidator} accepts. Mapped to their ISO 639-1 + * equivalents here so that callers can use either form. + */ + private static final Map<String, String> BIBLIOGRAPHIC_TO_ISO6391; + + static { + final Set<String> langs = new LinkedHashSet<>(); + Collections.addAll(langs, + "bg", "da", "de", "en", "es", "fi", "fr", "it", "nl", "pt", "ru"); + SUPPORTED_LANGUAGES = Collections.unmodifiableSet(langs); + + final Map<String, String> biblio = new HashMap<>(); + biblio.put("dut", "nl"); // Dutch + biblio.put("fre", "fr"); // French + biblio.put("ger", "de"); // German + BIBLIOGRAPHIC_TO_ISO6391 = Collections.unmodifiableMap(biblio); + } + + private StopwordLists() { + // utility class + } + + /** + * Returns a case-insensitive {@link StopwordFilter} for the given ISO 639 + * language code. Three-letter codes are normalized to their two-letter + * equivalent when a bundled list exists for the latter. + * + * @param iso639Code The ISO 639-1 or ISO 639-2/3 language code. + * Must not be {@code null}. + * @return A {@link StopwordFilter} backed by the bundled resource. + * @throws IllegalArgumentException if {@code iso639Code} is {@code null}, + * is not a valid ISO 639 code, or has no bundled list for this language. + * @throws UncheckedIOException if reading the bundled resource fails. + */ + public static StopwordFilter forLanguage(final String iso639Code) { + if (iso639Code == null) { + throw new IllegalArgumentException("iso639Code must not be null"); + } + LanguageCodeValidator.validateLanguageCode(iso639Code); + + final String normalized = normalizeToIso6391(iso639Code); + + if (!SUPPORTED_LANGUAGES.contains(normalized)) { + throw new IllegalArgumentException( + "No bundled stopword list for language '" + iso639Code + + "'. Supported languages: " + SUPPORTED_LANGUAGES); + } + + final String resource = RESOURCE_PATH_PREFIX + normalized + ".txt"; + final InputStream in = StopwordLists.class.getResourceAsStream(resource); + if (in == null) { + throw new IllegalArgumentException( + "Bundled stopword resource '" + resource + "' not found on the" + + " classpath. Supported languages: " + SUPPORTED_LANGUAGES); + } + try (InputStream stream = in) { + return new DictionaryStopwordFilter(stream, Charset.forName("UTF-8"), false); + } catch (final IOException e) { + throw new UncheckedIOException( + "Failed to load bundled stopword list for '" + normalized + "'", e); + } + } + + /** + * @return An unmodifiable view of the bundled ISO 639-1 codes for which + * stopword lists are shipped. + */ + public static Set<String> supportedLanguages() { + return SUPPORTED_LANGUAGES; + } + + /** + * Loads a stopword filter from a caller-supplied input stream. + * + * @param in The input stream. Must not be {@code null}. + * @param cs The {@link Charset} to decode with. Must not be {@code null}. + * @param caseSensitive Whether the resulting filter matches case-sensitively. + * @return A {@link StopwordFilter} populated from {@code in}. + * @throws IllegalArgumentException if {@code in} or {@code cs} is + * {@code null}. + * @throws IOException Thrown if an IO error occurs while reading. + */ + public static StopwordFilter load(final InputStream in, final Charset cs, + final boolean caseSensitive) throws IOException { + if (in == null) { + throw new IllegalArgumentException("in must not be null"); + } + if (cs == null) { + throw new IllegalArgumentException("cs must not be null"); + } + return new DictionaryStopwordFilter(in, cs, caseSensitive); + } + + /** + * Normalizes an ISO 639-2/3 three-letter code to its ISO 639-1 two-letter + * equivalent when one is available, otherwise returns the (lower-cased) + * input unchanged. The caller is responsible for validating the code first + * via {@link LanguageCodeValidator#validateLanguageCode(String)}. + * <p> + * Two-letter inputs are simply lower-cased and returned. For three-letter + * inputs, ISO 639-2 bibliographic codes ({@code dut}, {@code fre}, + * {@code ger}) are mapped explicitly because {@link Locale#getISO3Language()} + * returns only the terminologic forms ({@code nld}, {@code fra}, + * {@code deu}). Other three-letter codes are resolved by enumerating + * {@link Locale#getAvailableLocales()} and matching against + * {@link Locale#getISO3Language()}. + */ + private static String normalizeToIso6391(final String code) { Review Comment: We could save some CPU cycles if unknown ISO 639-3 codes were resolved via a static lookup at class init, instead of scanning Locale.getAvailableLocales() on each call. That could live in the same cache as forLanguage, keyed by normalized language code. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
