krickert commented on code in PR #1138: URL: https://github.com/apache/opennlp/pull/1138#discussion_r3569100439
########## opennlp-core/opennlp-runtime/src/main/java/opennlp/tools/util/normalizer/FullCaseFoldCharSequenceNormalizer.java: ########## @@ -0,0 +1,159 @@ +/* + * 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.util.normalizer; + +import java.io.BufferedReader; +import java.io.IOException; +import java.io.InputStream; +import java.io.InputStreamReader; +import java.io.UncheckedIOException; +import java.nio.charset.StandardCharsets; +import java.util.HashMap; +import java.util.Map; + +/** + * A {@link CharSequenceNormalizer} that applies Unicode full case folding (UTS #21) for + * case-insensitive matching, using the bundled {@code CaseFolding.txt} data of the Unicode Character + * Database. + * + * <p>Unlike {@link CaseFoldCharSequenceNormalizer}, which lower cases with + * {@link java.util.Locale#ROOT}, this applies the full case foldings (the {@code C} common and + * {@code F} full status mappings), including the expanding folds that plain lower casing does not + * perform: the sharp s (U+00DF) to {@code ss}, the Latin ligatures (for example U+FB00 to + * {@code ff}), and the Greek and Armenian multi-character folds. It is therefore an expanding, + * offset-changing transform, so it is offset-aware: {@link #normalizeAligned(CharSequence)} reports + * the {@link Alignment} from the folded text back to the input. A single cursor pass with no regular + * expression.</p> + * + * <p>The {@code S} simple and {@code T} Turkic mappings are deliberately excluded, so the fold is the + * language-neutral one (the Turkish dotless-i rule is not applied here). The fold matches precomposed + * code points, so apply NFC composition first if the input may contain decomposed forms. Inside an + * offset-aware pipeline that precondition cannot be established by a preceding rung, because NFC + * reports no offsets and {@code TextNormalizer.Builder.buildAligned()} rejects it; feed the aligned + * pipeline composed text, or accept that decomposed sequences pass through this fold unchanged.</p> + */ +public final class FullCaseFoldCharSequenceNormalizer implements OffsetAwareNormalizer { + + private static final long serialVersionUID = 4520210518330612934L; + + private static final String RESOURCE = "CaseFolding.txt"; + + // Maps a source code point to its full case folding (one or more code points), for the C and F + // status rows of CaseFolding.txt. + private static volatile Map<Integer, String> foldings; + + private static final FullCaseFoldCharSequenceNormalizer INSTANCE = + new FullCaseFoldCharSequenceNormalizer(); + + private FullCaseFoldCharSequenceNormalizer() { + } + + /** {@return the shared, stateless instance} */ + public static FullCaseFoldCharSequenceNormalizer getInstance() { + return INSTANCE; + } + + @Override + public CharSequence normalize(CharSequence text) { Review Comment: Switched to IllegalArgumentException to match the convention used elsewhere. ########## opennlp-core/opennlp-runtime/src/main/java/opennlp/tools/util/normalizer/FullCaseFoldCharSequenceNormalizer.java: ########## @@ -0,0 +1,159 @@ +/* + * 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.util.normalizer; + +import java.io.BufferedReader; +import java.io.IOException; +import java.io.InputStream; +import java.io.InputStreamReader; +import java.io.UncheckedIOException; +import java.nio.charset.StandardCharsets; +import java.util.HashMap; +import java.util.Map; + +/** + * A {@link CharSequenceNormalizer} that applies Unicode full case folding (UTS #21) for + * case-insensitive matching, using the bundled {@code CaseFolding.txt} data of the Unicode Character + * Database. + * + * <p>Unlike {@link CaseFoldCharSequenceNormalizer}, which lower cases with + * {@link java.util.Locale#ROOT}, this applies the full case foldings (the {@code C} common and + * {@code F} full status mappings), including the expanding folds that plain lower casing does not + * perform: the sharp s (U+00DF) to {@code ss}, the Latin ligatures (for example U+FB00 to + * {@code ff}), and the Greek and Armenian multi-character folds. It is therefore an expanding, + * offset-changing transform, so it is offset-aware: {@link #normalizeAligned(CharSequence)} reports + * the {@link Alignment} from the folded text back to the input. A single cursor pass with no regular + * expression.</p> + * + * <p>The {@code S} simple and {@code T} Turkic mappings are deliberately excluded, so the fold is the + * language-neutral one (the Turkish dotless-i rule is not applied here). The fold matches precomposed + * code points, so apply NFC composition first if the input may contain decomposed forms. Inside an + * offset-aware pipeline that precondition cannot be established by a preceding rung, because NFC + * reports no offsets and {@code TextNormalizer.Builder.buildAligned()} rejects it; feed the aligned + * pipeline composed text, or accept that decomposed sequences pass through this fold unchanged.</p> + */ +public final class FullCaseFoldCharSequenceNormalizer implements OffsetAwareNormalizer { + + private static final long serialVersionUID = 4520210518330612934L; + + private static final String RESOURCE = "CaseFolding.txt"; + + // Maps a source code point to its full case folding (one or more code points), for the C and F + // status rows of CaseFolding.txt. + private static volatile Map<Integer, String> foldings; + + private static final FullCaseFoldCharSequenceNormalizer INSTANCE = + new FullCaseFoldCharSequenceNormalizer(); + + private FullCaseFoldCharSequenceNormalizer() { + } + + /** {@return the shared, stateless instance} */ + public static FullCaseFoldCharSequenceNormalizer getInstance() { + return INSTANCE; + } + + @Override + public CharSequence normalize(CharSequence text) { + // Resolve the table once per call rather than once per code point, so the volatile read and + // double-checked-lock guard is not repeated in the hot per-character loop (CharClass.substitute + // calls the substitution function once per code point in text). + final Map<Integer, String> table = foldings(); + return CharClass.substitute(text, table::get); + } + + @Override + public AlignedText normalizeAligned(CharSequence text) { Review Comment: Switched to IllegalArgumentException to match the convention used elsewhere. -- 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]
