| 1 | /* |
| 2 | * Copyright 2006-2007 the original author or authors. |
| 3 | * |
| 4 | * Licensed under the Apache License, Version 2.0 (the "License"); |
| 5 | * you may not use this file except in compliance with the License. |
| 6 | * You may obtain a copy of the License at |
| 7 | * |
| 8 | * http://www.apache.org/licenses/LICENSE-2.0 |
| 9 | * |
| 10 | * Unless required by applicable law or agreed to in writing, software |
| 11 | * distributed under the License is distributed on an "AS IS" BASIS, |
| 12 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 13 | * See the License for the specific language governing permissions and |
| 14 | * limitations under the License. |
| 15 | */ |
| 16 | |
| 17 | package org.springframework.batch.item.file.transform; |
| 18 | |
| 19 | import java.util.ArrayList; |
| 20 | import java.util.Arrays; |
| 21 | import java.util.List; |
| 22 | |
| 23 | /** |
| 24 | * Tokenizer used to process data obtained from files with fixed-length format. |
| 25 | * Columns are specified by array of Range objects ({@link #setColumns(Range[])} |
| 26 | * ). |
| 27 | * |
| 28 | * @author tomas.slanina |
| 29 | * @author peter.zozom |
| 30 | * @author Dave Syer |
| 31 | * @author Lucas Ward |
| 32 | */ |
| 33 | public class FixedLengthTokenizer extends AbstractLineTokenizer { |
| 34 | |
| 35 | private Range[] ranges; |
| 36 | |
| 37 | private int maxRange = 0; |
| 38 | |
| 39 | boolean open = false; |
| 40 | |
| 41 | /** |
| 42 | * Set the column ranges. Used in conjunction with the |
| 43 | * {@link RangeArrayPropertyEditor} this property can be set in the form of |
| 44 | * a String describing the range boundaries, e.g. "1,4,7" or "1-3,4-6,7" or |
| 45 | * "1-2,4-5,7-10". If the last range is open then the rest of the line is |
| 46 | * read into that column (irrespective of the strict flag setting). |
| 47 | * |
| 48 | * @see #setStrict(boolean) |
| 49 | * |
| 50 | * @param ranges the column ranges expected in the input |
| 51 | */ |
| 52 | public void setColumns(Range[] ranges) { |
| 53 | this.ranges = Arrays.asList(ranges).toArray(new Range[ranges.length]); |
| 54 | calculateMaxRange(ranges); |
| 55 | } |
| 56 | |
| 57 | /* |
| 58 | * Calculate the highest value within an array of ranges. The ranges aren't |
| 59 | * necessarily in order. For example: "5-10, 1-4,11-15". Furthermore, there |
| 60 | * isn't always a min and max, such as: "1,4-20, 22" |
| 61 | */ |
| 62 | private void calculateMaxRange(Range[] ranges) { |
| 63 | if (ranges == null || ranges.length == 0) { |
| 64 | maxRange = 0; |
| 65 | return; |
| 66 | } |
| 67 | |
| 68 | open = false; |
| 69 | maxRange = ranges[0].getMin(); |
| 70 | |
| 71 | for (int i = 0; i < ranges.length; i++) { |
| 72 | int upperBound; |
| 73 | if (ranges[i].hasMaxValue()) { |
| 74 | upperBound = ranges[i].getMax(); |
| 75 | } |
| 76 | else { |
| 77 | upperBound = ranges[i].getMin(); |
| 78 | if (upperBound > maxRange) { |
| 79 | open = true; |
| 80 | } |
| 81 | } |
| 82 | |
| 83 | if (upperBound > maxRange) { |
| 84 | maxRange = upperBound; |
| 85 | } |
| 86 | } |
| 87 | } |
| 88 | |
| 89 | /** |
| 90 | * Yields the tokens resulting from the splitting of the supplied |
| 91 | * <code>line</code>. |
| 92 | * |
| 93 | * @param line the line to be tokenised (can be <code>null</code>) |
| 94 | * |
| 95 | * @return the resulting tokens (empty if the line is null) |
| 96 | * @throws IncorrectLineLengthException if line length is greater than or |
| 97 | * less than the max range set. |
| 98 | */ |
| 99 | protected List<String> doTokenize(String line) { |
| 100 | List<String> tokens = new ArrayList<String>(ranges.length); |
| 101 | int lineLength; |
| 102 | String token; |
| 103 | |
| 104 | lineLength = line.length(); |
| 105 | |
| 106 | if (lineLength < maxRange && isStrict()) { |
| 107 | throw new IncorrectLineLengthException("Line is shorter than max range " + maxRange, maxRange, lineLength); |
| 108 | } |
| 109 | |
| 110 | if (!open && lineLength > maxRange && isStrict()) { |
| 111 | throw new IncorrectLineLengthException("Line is longer than max range " + maxRange, maxRange, lineLength); |
| 112 | } |
| 113 | |
| 114 | for (int i = 0; i < ranges.length; i++) { |
| 115 | |
| 116 | int startPos = ranges[i].getMin() - 1; |
| 117 | int endPos = ranges[i].getMax(); |
| 118 | |
| 119 | if (lineLength >= endPos) { |
| 120 | token = line.substring(startPos, endPos); |
| 121 | } |
| 122 | else if (lineLength >= startPos) { |
| 123 | token = line.substring(startPos); |
| 124 | } |
| 125 | else { |
| 126 | token = ""; |
| 127 | } |
| 128 | |
| 129 | tokens.add(token); |
| 130 | } |
| 131 | |
| 132 | return tokens; |
| 133 | } |
| 134 | } |