michael@0: /**
michael@0:  * Copyright (c) 2012, Ben Fortuna
michael@0:  * All rights reserved.
michael@0:  *
michael@0:  * Redistribution and use in source and binary forms, with or without
michael@0:  * modification, are permitted provided that the following conditions
michael@0:  * are met:
michael@0:  *
michael@0:  *  o Redistributions of source code must retain the above copyright
michael@0:  * notice, this list of conditions and the following disclaimer.
michael@0:  *
michael@0:  *  o Redistributions in binary form must reproduce the above copyright
michael@0:  * notice, this list of conditions and the following disclaimer in the
michael@0:  * documentation and/or other materials provided with the distribution.
michael@0:  *
michael@0:  *  o Neither the name of Ben Fortuna nor the names of any other contributors
michael@0:  * may be used to endorse or promote products derived from this software
michael@0:  * without specific prior written permission.
michael@0:  *
michael@0:  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
michael@0:  * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
michael@0:  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
michael@0:  * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
michael@0:  * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
michael@0:  * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
michael@0:  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
michael@0:  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
michael@0:  * LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
michael@0:  * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
michael@0:  * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
michael@0:  */
michael@0: package net.fortuna.ical4j.util;
michael@0: 
michael@0: import java.util.regex.Pattern;
michael@0: 
michael@0: /**
michael@0:  * $Id$ [23-Apr-2004]
michael@0:  *
michael@0:  * Utility methods for working with parameters.
michael@0:  * @author Ben Fortuna
michael@0:  * <pre>
michael@0:  * 4.3.11 Text
michael@0:  * 
michael@0:  *    Value Name: TEXT
michael@0:  * 
michael@0:  *    Purpose This value type is used to identify values that contain human
michael@0:  *    readable text.
michael@0:  * 
michael@0:  *    Formal Definition: The character sets supported by this revision of
michael@0:  *    iCalendar are UTF-8 and US ASCII thereof. The applicability to other
michael@0:  *    character sets is for future work. The value type is defined by the
michael@0:  *    following notation.
michael@0:  * 
michael@0:  *      text       = *(TSAFE-CHAR / ":" / DQUOTE / ESCAPED-CHAR)
michael@0:  *      ; Folded according to description above
michael@0:  * 
michael@0:  *      ESCAPED-CHAR = "\\" / "\;" / "\," / "\N" / "\n")
michael@0:  *         ; \\ encodes \, \N or \n encodes newline
michael@0:  *         ; \; encodes ;, \, encodes ,
michael@0:  * 
michael@0:  *      TSAFE-CHAR = %x20-21 / %x23-2B / %x2D-39 / %x3C-5B
michael@0:  *                   %x5D-7E / NON-US-ASCII
michael@0:  *         ; Any character except CTLs not needed by the current
michael@0:  *         ; character set, DQUOTE, ";", ":", "\", ","
michael@0:  * 
michael@0:  *      Note: Certain other character sets may require modification of the
michael@0:  *      above definitions, but this is beyond the scope of this document.
michael@0:  * 
michael@0:  *    Description: If the property permits, multiple "text" values are
michael@0:  *    specified by a COMMA character (US-ASCII decimal 44) separated list
michael@0:  *    of values.
michael@0:  * 
michael@0:  *    The language in which the text is represented can be controlled by
michael@0:  *    the "LANGUAGE" property parameter.
michael@0:  * 
michael@0:  *    An intentional formatted text line break MUST only be included in a
michael@0:  *    "TEXT" property value by representing the line break with the
michael@0:  *    character sequence of BACKSLASH (US-ASCII decimal 92), followed by a
michael@0:  *    LATIN SMALL LETTER N (US-ASCII decimal 110) or a LATIN CAPITAL LETTER
michael@0:  *    N (US-ASCII decimal 78), that is "\n" or "\N".
michael@0:  * 
michael@0:  *    The "TEXT" property values may also contain special characters that
michael@0:  *    are used to signify delimiters, such as a COMMA character for lists
michael@0:  *    of values or a SEMICOLON character for structured values. In order to
michael@0:  *    support the inclusion of these special characters in "TEXT" property
michael@0:  *    values, they MUST be escaped with a BACKSLASH character. A BACKSLASH
michael@0:  *    character (US-ASCII decimal 92) in a "TEXT" property value MUST be
michael@0:  *    escaped with another BACKSLASH character. A COMMA character in a
michael@0:  *    "TEXT" property value MUST be escaped with a BACKSLASH character
michael@0:  *    (US-ASCII decimal 92). A SEMICOLON character in a "TEXT" property
michael@0:  *    value MUST be escaped with a BACKSLASH character (US-ASCII decimal
michael@0:  *    92).  However, a COLON character in a "TEXT" property value SHALL NOT
michael@0:  *    be escaped with a BACKSLASH character.Example: A multiple line value
michael@0:  *    of:
michael@0:  * 
michael@0:  *      Project XYZ Final Review
michael@0:  *      Conference Room - 3B
michael@0:  *      Come Prepared.
michael@0:  * 
michael@0:  *    would be represented as:
michael@0:  * 
michael@0:  *      Project XYZ Final Review\nConference Room - 3B\nCome Prepared.
michael@0:  * </pre>
michael@0:  */
michael@0: public final class Strings {
michael@0: 
michael@0:     /**
michael@0:      * Defines a regular expression representing all parameter strings that
michael@0:      * should be quoted.
michael@0:      */
michael@0:     public static final Pattern PARAM_QUOTE_PATTERN = Pattern.compile("[:;,]|[^\\p{ASCII}]");
michael@0:     
michael@0:     private static final Pattern ESCAPE_PUNCTUATION_PATTERN = Pattern.compile("([,;])");
michael@0:     private static final Pattern UNESCAPE_PUNCTUATION_PATTERN = Pattern.compile("\\\\([,;\"])");
michael@0:     
michael@0:     private static final Pattern ESCAPE_NEWLINE_PATTERN = Pattern.compile("\r?\n");
michael@0:     private static final Pattern UNESCAPE_NEWLINE_PATTERN = Pattern.compile("(?<!\\\\)\\\\n");
michael@0:     
michael@0:     private static final Pattern ESCAPE_BACKSLASH_PATTERN = Pattern.compile("\\\\");
michael@0:     private static final Pattern UNESCAPE_BACKSLASH_PATTERN = Pattern.compile("\\\\\\\\");
michael@0:     
michael@0:     
michael@0:     
michael@0:     /**
michael@0:      * A string used to denote the start (and end) of iCalendar content lines.
michael@0:      */
michael@0:     public static final String LINE_SEPARATOR = "\r\n";
michael@0: 
michael@0:     /**
michael@0:      * Constructor made private to prevent instantiation.
michael@0:      */
michael@0:     private Strings() {
michael@0:     }
michael@0: 
michael@0:     /**
michael@0:      * Convenience method for adding quotes. The specified
michael@0:      * object is converted to a string representation by
michael@0:      * calling its <code>toString()</code> method.
michael@0:      * @param aValue an object to quote
michael@0:      * @return a quoted string
michael@0:      */
michael@0:     public static String quote(final Object aValue) {
michael@0:         if (aValue != null) {
michael@0:             return "\"" + aValue + "\"";
michael@0:         }
michael@0:         return "\"\"";
michael@0:     }
michael@0: 
michael@0:     /**
michael@0:      * Convenience method for removing surrounding quotes
michael@0:      * from a string value.
michael@0:      * @param aValue a string to remove quotes from
michael@0:      * @return an un-quoted string
michael@0:      */
michael@0:     public static String unquote(final String aValue) {
michael@0:         if (aValue != null && aValue.startsWith("\"") && aValue.endsWith("\"")) {
michael@0:             return aValue.substring(0, aValue.length() - 1).substring(1);
michael@0:         }
michael@0:         return aValue;
michael@0:     }
michael@0: 
michael@0:     /**
michael@0:      * Convenience method for escaping special characters.
michael@0:      * @param aValue a string value to escape
michael@0:      * @return an escaped representation of the specified
michael@0:      * string
michael@0:      */
michael@0:     public static String escape(final String aValue) {
michael@0:         return escapePunctuation(escapeNewline(escapeBackslash(aValue)));
michael@0:     }
michael@0:     
michael@0:     /**
michael@0:      * Convenience method for replacing escaped special characters
michael@0:      * with their original form.
michael@0:      * @param aValue a string value to unescape
michael@0:      * @return a string representation of the specified
michael@0:      * string with escaped characters replaced with their
michael@0:      * original form
michael@0:      */
michael@0:     public static String unescape(final String aValue) {
michael@0:         return unescapeBackslash(unescapeNewline(unescapePunctuation(aValue)));
michael@0:     }
michael@0: 
michael@0:     private static String escapePunctuation(String value) {
michael@0:         if (value != null) {
michael@0:             return ESCAPE_PUNCTUATION_PATTERN.matcher(value).replaceAll("\\\\$1");
michael@0:         }
michael@0:         return value;
michael@0:     }
michael@0: 
michael@0:     private static String unescapePunctuation(String value) {
michael@0:         if (value != null) {
michael@0:             return UNESCAPE_PUNCTUATION_PATTERN.matcher(value).replaceAll("$1");
michael@0:         }
michael@0:         return value;
michael@0:     }
michael@0: 
michael@0:     public static String escapeNewline(String value) {
michael@0:         if (value != null) {
michael@0:             return ESCAPE_NEWLINE_PATTERN.matcher(value).replaceAll("\\\\n");
michael@0:         }
michael@0:         return value;
michael@0:     }
michael@0: 
michael@0:     private static String unescapeNewline(String value) {
michael@0:         if (value != null) {
michael@0:             return UNESCAPE_NEWLINE_PATTERN.matcher(value).replaceAll("\n");
michael@0:         }
michael@0:         return value;
michael@0:     }
michael@0: 
michael@0:     private static String escapeBackslash(String value) {
michael@0:         if (value != null) {
michael@0:             return ESCAPE_BACKSLASH_PATTERN.matcher(value).replaceAll("\\\\\\\\");
michael@0:         }
michael@0:         return value;
michael@0:     }
michael@0: 
michael@0:     private static String unescapeBackslash(String value) {
michael@0:         if (value != null) {
michael@0:             return UNESCAPE_BACKSLASH_PATTERN.matcher(value).replaceAll("\\\\");
michael@0:         }
michael@0:         return value;
michael@0:     }
michael@0:     
michael@0:     /**
michael@0:      * Wraps <code>java.lang.String.valueOf()</code> to return an empty string
michael@0:      * where the specified object is null.
michael@0:      * @param object an object instance
michael@0:      * @return a string representation of the object
michael@0:      */
michael@0:     public static String valueOf(final Object object) {
michael@0:         if (object == null) {
michael@0:             return "";
michael@0:         }
michael@0:         return object.toString();
michael@0:     }
michael@0: }