emkornfield commented on code in PR #3390: URL: https://github.com/apache/parquet-java/pull/3390#discussion_r2719341086
########## parquet-column/src/main/java/org/apache/parquet/column/values/alp/AlpEncoderDecoder.java: ########## @@ -0,0 +1,384 @@ +/* + * 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.parquet.column.values.alp; + +import static org.apache.parquet.column.values.alp.AlpConstants.*; + +/** + * Core ALP (Adaptive Lossless floating-Point) encoding and decoding logic. + * + * <p>ALP works by converting floating-point values to integers using decimal scaling, + * then applying Frame of Reference (FOR) encoding and bit-packing. + * Values that cannot be losslessly converted are stored as exceptions. + * + * <p>Encoding formula: encoded = round(value * 10^(exponent - factor)) + * <p>Decoding formula: value = encoded / 10^(exponent - factor) + * + * <p>Exception conditions: + * <ul> + * <li>NaN values</li> + * <li>Infinity values</li> + * <li>Negative zero (-0.0)</li> + * <li>Out of integer range</li> + * <li>Round-trip failure (decode(encode(v)) != v)</li> + * </ul> + */ +public final class AlpEncoderDecoder { + + private AlpEncoderDecoder() { + // Utility class + } + + // ========== Float Encoding/Decoding ========== + + /** + * Check if a float value is an exception (cannot be losslessly encoded). + * + * @param value the float value to check + * @return true if the value is an exception + */ + public static boolean isFloatException(float value) { + // NaN check + if (Float.isNaN(value)) { + return true; + } + // Infinity check + if (Float.isInfinite(value)) { + return true; + } + // Negative zero check + if (Float.floatToRawIntBits(value) == FLOAT_NEGATIVE_ZERO_BITS) { + return true; + } + return false; + } + + /** + * Check if a float value will be an exception for the given exponent/factor. + * + * @param value the float value + * @param exponent the decimal exponent (0-10) + * @param factor the decimal factor (0 <= factor <= exponent) + * @return true if the value is an exception for this encoding + */ + public static boolean isFloatException(float value, int exponent, int factor) { + if (isFloatException(value)) { + return true; + } + + // Try encoding and check for round-trip failure + float multiplier = FLOAT_POW10[exponent]; + if (factor > 0) { Review Comment: Not sure I understand the > 0 check, need to examine encoding logic in more detail, is this just an optimization to avoid dividing by 1? -- 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]
