| 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.List; |
| 21 | |
| 22 | import org.springframework.util.Assert; |
| 23 | import org.springframework.util.StringUtils; |
| 24 | |
| 25 | /** |
| 26 | * |
| 27 | * @author Rob Harrop |
| 28 | * @author Dave Syer |
| 29 | * |
| 30 | */ |
| 31 | public class DelimitedLineTokenizer extends AbstractLineTokenizer { |
| 32 | /** |
| 33 | * Convenient constant for the common case of a tab delimiter. |
| 34 | */ |
| 35 | public static final char DELIMITER_TAB = '\t'; |
| 36 | |
| 37 | /** |
| 38 | * Convenient constant for the common case of a comma delimiter. |
| 39 | */ |
| 40 | public static final char DELIMITER_COMMA = ','; |
| 41 | |
| 42 | /** |
| 43 | * Convenient constant for the common case of a " character used to escape delimiters or line endings. |
| 44 | */ |
| 45 | public static final char DEFAULT_QUOTE_CHARACTER = '"'; |
| 46 | |
| 47 | // the delimiter character used when reading input. |
| 48 | private char delimiter; |
| 49 | |
| 50 | private char quoteCharacter = DEFAULT_QUOTE_CHARACTER; |
| 51 | |
| 52 | private String quoteString; |
| 53 | |
| 54 | /** |
| 55 | * Create a new instance of the {@link DelimitedLineTokenizer} class for the common case where the delimiter is a |
| 56 | * {@link #DELIMITER_COMMA comma}. |
| 57 | * |
| 58 | * @see #DelimitedLineTokenizer(char) |
| 59 | * @see #DELIMITER_COMMA |
| 60 | */ |
| 61 | public DelimitedLineTokenizer() { |
| 62 | this(DELIMITER_COMMA); |
| 63 | } |
| 64 | |
| 65 | /** |
| 66 | * Create a new instance of the {@link DelimitedLineTokenizer} class. |
| 67 | * |
| 68 | * @param delimiter the desired delimiter |
| 69 | */ |
| 70 | public DelimitedLineTokenizer(char delimiter) { |
| 71 | Assert.state(delimiter != DEFAULT_QUOTE_CHARACTER, "[" + DEFAULT_QUOTE_CHARACTER |
| 72 | + "] is not allowed as delimiter for tokenizers."); |
| 73 | |
| 74 | this.delimiter = delimiter; |
| 75 | setQuoteCharacter(DEFAULT_QUOTE_CHARACTER); |
| 76 | } |
| 77 | |
| 78 | /** |
| 79 | * Setter for the delimiter character. |
| 80 | * |
| 81 | * @param delimiter |
| 82 | */ |
| 83 | public void setDelimiter(char delimiter) { |
| 84 | this.delimiter = delimiter; |
| 85 | } |
| 86 | |
| 87 | /** |
| 88 | * Public setter for the quoteCharacter. The quote character can be used to extend a field across line endings or to |
| 89 | * enclose a String which contains the delimiter. Inside a quoted token the quote character can be used to escape |
| 90 | * itself, thus "a""b""c" is tokenized to a"b"c. |
| 91 | * |
| 92 | * @param quoteCharacter the quoteCharacter to set |
| 93 | * |
| 94 | * @see #DEFAULT_QUOTE_CHARACTER |
| 95 | */ |
| 96 | public void setQuoteCharacter(char quoteCharacter) { |
| 97 | this.quoteCharacter = quoteCharacter; |
| 98 | this.quoteString = "" + quoteCharacter; |
| 99 | } |
| 100 | |
| 101 | /** |
| 102 | * Yields the tokens resulting from the splitting of the supplied <code>line</code>. |
| 103 | * |
| 104 | * @param line the line to be tokenized |
| 105 | * |
| 106 | * @return the resulting tokens |
| 107 | */ |
| 108 | protected List doTokenize(String line) { |
| 109 | |
| 110 | List tokens = new ArrayList(); |
| 111 | |
| 112 | // line is never null in current implementation |
| 113 | // line is checked in parent: AbstractLineTokenizer.tokenize() |
| 114 | char[] chars = line.toCharArray(); |
| 115 | boolean inQuoted = false; |
| 116 | int lastCut = 0; |
| 117 | int length = chars.length; |
| 118 | |
| 119 | for (int i = 0; i < length; i++) { |
| 120 | |
| 121 | char currentChar = chars[i]; |
| 122 | boolean isEnd = (i == (length - 1)); |
| 123 | |
| 124 | if ((isDelimiterCharacter(currentChar) && !inQuoted) || isEnd) { |
| 125 | int endPosition = (isEnd ? (length - lastCut) : (i - lastCut)); |
| 126 | |
| 127 | if (isEnd && isDelimiterCharacter(currentChar)) { |
| 128 | endPosition--; |
| 129 | } |
| 130 | |
| 131 | String value = null; |
| 132 | |
| 133 | value = maybeStripQuotes(new String(chars, lastCut, endPosition)); |
| 134 | |
| 135 | tokens.add(value); |
| 136 | |
| 137 | if (isEnd && (isDelimiterCharacter(currentChar))) { |
| 138 | tokens.add(""); |
| 139 | } |
| 140 | |
| 141 | lastCut = i + 1; |
| 142 | } else if (isQuoteCharacter(currentChar)) { |
| 143 | inQuoted = !inQuoted; |
| 144 | } |
| 145 | |
| 146 | } |
| 147 | |
| 148 | return tokens; |
| 149 | } |
| 150 | |
| 151 | /** |
| 152 | * If the string is quoted strip (possibly with whitespace outside the quotes (which will be stripped), replace |
| 153 | * escaped quotes inside the string. Quotes are escaped with double instances of the quote character. |
| 154 | * |
| 155 | * @param string |
| 156 | * @return the same string but stripped and unescaped if necessary |
| 157 | */ |
| 158 | private String maybeStripQuotes(String string) { |
| 159 | String value = string.trim(); |
| 160 | if (isQuoted(value)) { |
| 161 | value = StringUtils.replace(value, "" + quoteCharacter + quoteCharacter, "" + quoteCharacter); |
| 162 | int endLength = value.length() - 1; |
| 163 | // used to deal with empty quoted values |
| 164 | if (endLength == 0) { |
| 165 | endLength = 1; |
| 166 | } |
| 167 | string = value.substring(1, endLength); |
| 168 | } |
| 169 | return string; |
| 170 | } |
| 171 | |
| 172 | /** |
| 173 | * Is this string surrounded by quite characters? |
| 174 | * |
| 175 | * @param value |
| 176 | * @return true if the value starts and ends with the {@link #quoteCharacter} |
| 177 | */ |
| 178 | private boolean isQuoted(String value) { |
| 179 | if (value.startsWith(quoteString) && value.endsWith(quoteString)) { |
| 180 | return true; |
| 181 | } |
| 182 | return false; |
| 183 | } |
| 184 | |
| 185 | /** |
| 186 | * Is the supplied character the delimiter character? |
| 187 | * |
| 188 | * @param c the character to be checked |
| 189 | * @return <code>true</code> if the supplied character is the delimiter character |
| 190 | * @see DelimitedLineTokenizer#DelimitedLineTokenizer(char) |
| 191 | */ |
| 192 | private boolean isDelimiterCharacter(char c) { |
| 193 | return c == this.delimiter; |
| 194 | } |
| 195 | |
| 196 | /** |
| 197 | * Is the supplied character a quote character? |
| 198 | * |
| 199 | * @param c the character to be checked |
| 200 | * @return <code>true</code> if the supplied character is an quote character |
| 201 | * @see #setQuoteCharacter(char) |
| 202 | */ |
| 203 | protected boolean isQuoteCharacter(char c) { |
| 204 | return c == quoteCharacter; |
| 205 | } |
| 206 | } |