neoremind commented on code in PR #16371: URL: https://github.com/apache/lucene/pull/16371#discussion_r3681147749
########## lucene/core/src/java/org/apache/lucene/analysis/CaseFoldingFilter.java: ########## @@ -0,0 +1,49 @@ +/* + * 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 org.apache.lucene.analysis; + +import java.io.IOException; +import org.apache.lucene.analysis.tokenattributes.CharTermAttribute; + +/** + * Normalizes token text by applying Unicode simple case folding, which erases case distinctions + * more thoroughly than {@link LowerCaseFilter} for non-ASCII scripts. Review Comment: Interesting PR! Happen to look at `SimpleCaseFolding`'s first lines, my first instinct was that 0x0130 -> 0x0130 maybe incorrect? But it turns out to be correct, by looking up the mapping in https://www.unicode.org/Public/17.0.0/ucd/CaseFolding.txt, U+0130 has no Common/Simple mapping (only Full and Turkic ones), so the default simple case fold makes it unchanged. This is different from `Character.toLowerCase()`, which turns "İ" into "i". I do a quick scan over all codepoints, this is the only char where `Character.toLowerCase` is different from input char but simple folding is the same to the input char. So, two minor comments: 1. would it be better to modify the javadoc "erases case distinctions more thoroughly than LowerCaseFilter" to be more precise? Something like: Normalizes token text by applying Unicode simple case folding, which erases case distinctions that LowerCaseFilter misses for non-ASCII characters. One notable difference, U+0130 is unchanged, this mapping is Turkic-specific. OR simply claim - Normalizes token text by applying Unicode simple case folding (removing the ambiguous part) 2. maybe add https://www.unicode.org/Public/17.0.0/ucd/CaseFolding.txt as a @ see in javadoc, either here or in SimpleCaseFolding? since it's the ground truth for these mappings. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
