1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/src/net/fortuna/ical4j/util/Strings.java Tue Feb 10 18:12:00 2015 +0100
1.3 @@ -0,0 +1,238 @@
1.4 +/**
1.5 + * Copyright (c) 2012, Ben Fortuna
1.6 + * All rights reserved.
1.7 + *
1.8 + * Redistribution and use in source and binary forms, with or without
1.9 + * modification, are permitted provided that the following conditions
1.10 + * are met:
1.11 + *
1.12 + * o Redistributions of source code must retain the above copyright
1.13 + * notice, this list of conditions and the following disclaimer.
1.14 + *
1.15 + * o Redistributions in binary form must reproduce the above copyright
1.16 + * notice, this list of conditions and the following disclaimer in the
1.17 + * documentation and/or other materials provided with the distribution.
1.18 + *
1.19 + * o Neither the name of Ben Fortuna nor the names of any other contributors
1.20 + * may be used to endorse or promote products derived from this software
1.21 + * without specific prior written permission.
1.22 + *
1.23 + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
1.24 + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
1.25 + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
1.26 + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
1.27 + * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
1.28 + * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
1.29 + * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
1.30 + * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
1.31 + * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
1.32 + * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
1.33 + * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
1.34 + */
1.35 +package net.fortuna.ical4j.util;
1.36 +
1.37 +import java.util.regex.Pattern;
1.38 +
1.39 +/**
1.40 + * $Id$ [23-Apr-2004]
1.41 + *
1.42 + * Utility methods for working with parameters.
1.43 + * @author Ben Fortuna
1.44 + * <pre>
1.45 + * 4.3.11 Text
1.46 + *
1.47 + * Value Name: TEXT
1.48 + *
1.49 + * Purpose This value type is used to identify values that contain human
1.50 + * readable text.
1.51 + *
1.52 + * Formal Definition: The character sets supported by this revision of
1.53 + * iCalendar are UTF-8 and US ASCII thereof. The applicability to other
1.54 + * character sets is for future work. The value type is defined by the
1.55 + * following notation.
1.56 + *
1.57 + * text = *(TSAFE-CHAR / ":" / DQUOTE / ESCAPED-CHAR)
1.58 + * ; Folded according to description above
1.59 + *
1.60 + * ESCAPED-CHAR = "\\" / "\;" / "\," / "\N" / "\n")
1.61 + * ; \\ encodes \, \N or \n encodes newline
1.62 + * ; \; encodes ;, \, encodes ,
1.63 + *
1.64 + * TSAFE-CHAR = %x20-21 / %x23-2B / %x2D-39 / %x3C-5B
1.65 + * %x5D-7E / NON-US-ASCII
1.66 + * ; Any character except CTLs not needed by the current
1.67 + * ; character set, DQUOTE, ";", ":", "\", ","
1.68 + *
1.69 + * Note: Certain other character sets may require modification of the
1.70 + * above definitions, but this is beyond the scope of this document.
1.71 + *
1.72 + * Description: If the property permits, multiple "text" values are
1.73 + * specified by a COMMA character (US-ASCII decimal 44) separated list
1.74 + * of values.
1.75 + *
1.76 + * The language in which the text is represented can be controlled by
1.77 + * the "LANGUAGE" property parameter.
1.78 + *
1.79 + * An intentional formatted text line break MUST only be included in a
1.80 + * "TEXT" property value by representing the line break with the
1.81 + * character sequence of BACKSLASH (US-ASCII decimal 92), followed by a
1.82 + * LATIN SMALL LETTER N (US-ASCII decimal 110) or a LATIN CAPITAL LETTER
1.83 + * N (US-ASCII decimal 78), that is "\n" or "\N".
1.84 + *
1.85 + * The "TEXT" property values may also contain special characters that
1.86 + * are used to signify delimiters, such as a COMMA character for lists
1.87 + * of values or a SEMICOLON character for structured values. In order to
1.88 + * support the inclusion of these special characters in "TEXT" property
1.89 + * values, they MUST be escaped with a BACKSLASH character. A BACKSLASH
1.90 + * character (US-ASCII decimal 92) in a "TEXT" property value MUST be
1.91 + * escaped with another BACKSLASH character. A COMMA character in a
1.92 + * "TEXT" property value MUST be escaped with a BACKSLASH character
1.93 + * (US-ASCII decimal 92). A SEMICOLON character in a "TEXT" property
1.94 + * value MUST be escaped with a BACKSLASH character (US-ASCII decimal
1.95 + * 92). However, a COLON character in a "TEXT" property value SHALL NOT
1.96 + * be escaped with a BACKSLASH character.Example: A multiple line value
1.97 + * of:
1.98 + *
1.99 + * Project XYZ Final Review
1.100 + * Conference Room - 3B
1.101 + * Come Prepared.
1.102 + *
1.103 + * would be represented as:
1.104 + *
1.105 + * Project XYZ Final Review\nConference Room - 3B\nCome Prepared.
1.106 + * </pre>
1.107 + */
1.108 +public final class Strings {
1.109 +
1.110 + /**
1.111 + * Defines a regular expression representing all parameter strings that
1.112 + * should be quoted.
1.113 + */
1.114 + public static final Pattern PARAM_QUOTE_PATTERN = Pattern.compile("[:;,]|[^\\p{ASCII}]");
1.115 +
1.116 + private static final Pattern ESCAPE_PUNCTUATION_PATTERN = Pattern.compile("([,;])");
1.117 + private static final Pattern UNESCAPE_PUNCTUATION_PATTERN = Pattern.compile("\\\\([,;\"])");
1.118 +
1.119 + private static final Pattern ESCAPE_NEWLINE_PATTERN = Pattern.compile("\r?\n");
1.120 + private static final Pattern UNESCAPE_NEWLINE_PATTERN = Pattern.compile("(?<!\\\\)\\\\n");
1.121 +
1.122 + private static final Pattern ESCAPE_BACKSLASH_PATTERN = Pattern.compile("\\\\");
1.123 + private static final Pattern UNESCAPE_BACKSLASH_PATTERN = Pattern.compile("\\\\\\\\");
1.124 +
1.125 +
1.126 +
1.127 + /**
1.128 + * A string used to denote the start (and end) of iCalendar content lines.
1.129 + */
1.130 + public static final String LINE_SEPARATOR = "\r\n";
1.131 +
1.132 + /**
1.133 + * Constructor made private to prevent instantiation.
1.134 + */
1.135 + private Strings() {
1.136 + }
1.137 +
1.138 + /**
1.139 + * Convenience method for adding quotes. The specified
1.140 + * object is converted to a string representation by
1.141 + * calling its <code>toString()</code> method.
1.142 + * @param aValue an object to quote
1.143 + * @return a quoted string
1.144 + */
1.145 + public static String quote(final Object aValue) {
1.146 + if (aValue != null) {
1.147 + return "\"" + aValue + "\"";
1.148 + }
1.149 + return "\"\"";
1.150 + }
1.151 +
1.152 + /**
1.153 + * Convenience method for removing surrounding quotes
1.154 + * from a string value.
1.155 + * @param aValue a string to remove quotes from
1.156 + * @return an un-quoted string
1.157 + */
1.158 + public static String unquote(final String aValue) {
1.159 + if (aValue != null && aValue.startsWith("\"") && aValue.endsWith("\"")) {
1.160 + return aValue.substring(0, aValue.length() - 1).substring(1);
1.161 + }
1.162 + return aValue;
1.163 + }
1.164 +
1.165 + /**
1.166 + * Convenience method for escaping special characters.
1.167 + * @param aValue a string value to escape
1.168 + * @return an escaped representation of the specified
1.169 + * string
1.170 + */
1.171 + public static String escape(final String aValue) {
1.172 + return escapePunctuation(escapeNewline(escapeBackslash(aValue)));
1.173 + }
1.174 +
1.175 + /**
1.176 + * Convenience method for replacing escaped special characters
1.177 + * with their original form.
1.178 + * @param aValue a string value to unescape
1.179 + * @return a string representation of the specified
1.180 + * string with escaped characters replaced with their
1.181 + * original form
1.182 + */
1.183 + public static String unescape(final String aValue) {
1.184 + return unescapeBackslash(unescapeNewline(unescapePunctuation(aValue)));
1.185 + }
1.186 +
1.187 + private static String escapePunctuation(String value) {
1.188 + if (value != null) {
1.189 + return ESCAPE_PUNCTUATION_PATTERN.matcher(value).replaceAll("\\\\$1");
1.190 + }
1.191 + return value;
1.192 + }
1.193 +
1.194 + private static String unescapePunctuation(String value) {
1.195 + if (value != null) {
1.196 + return UNESCAPE_PUNCTUATION_PATTERN.matcher(value).replaceAll("$1");
1.197 + }
1.198 + return value;
1.199 + }
1.200 +
1.201 + public static String escapeNewline(String value) {
1.202 + if (value != null) {
1.203 + return ESCAPE_NEWLINE_PATTERN.matcher(value).replaceAll("\\\\n");
1.204 + }
1.205 + return value;
1.206 + }
1.207 +
1.208 + private static String unescapeNewline(String value) {
1.209 + if (value != null) {
1.210 + return UNESCAPE_NEWLINE_PATTERN.matcher(value).replaceAll("\n");
1.211 + }
1.212 + return value;
1.213 + }
1.214 +
1.215 + private static String escapeBackslash(String value) {
1.216 + if (value != null) {
1.217 + return ESCAPE_BACKSLASH_PATTERN.matcher(value).replaceAll("\\\\\\\\");
1.218 + }
1.219 + return value;
1.220 + }
1.221 +
1.222 + private static String unescapeBackslash(String value) {
1.223 + if (value != null) {
1.224 + return UNESCAPE_BACKSLASH_PATTERN.matcher(value).replaceAll("\\\\");
1.225 + }
1.226 + return value;
1.227 + }
1.228 +
1.229 + /**
1.230 + * Wraps <code>java.lang.String.valueOf()</code> to return an empty string
1.231 + * where the specified object is null.
1.232 + * @param object an object instance
1.233 + * @return a string representation of the object
1.234 + */
1.235 + public static String valueOf(final Object object) {
1.236 + if (object == null) {
1.237 + return "";
1.238 + }
1.239 + return object.toString();
1.240 + }
1.241 +}
