src/net/fortuna/ical4j/util/Uris.java

changeset 0
fb9019fb1bf7
     1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/src/net/fortuna/ical4j/util/Uris.java	Tue Feb 10 18:12:00 2015 +0100
     1.3 @@ -0,0 +1,144 @@
     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.net.URI;
    1.38 +import java.net.URISyntaxException;
    1.39 +import java.util.regex.Pattern;
    1.40 +
    1.41 +/**
    1.42 + * $Id$
    1.43 + *
    1.44 + * Created on 11/09/2005
    1.45 + *
    1.46 + * Utility methods for working with URIs.
    1.47 + * @author Ben Fortuna
    1.48 + */
    1.49 +public final class Uris {
    1.50 +
    1.51 +    /**
    1.52 +     * URI Scheme used when relaxed parsing is enabled and the given input stream lead to an invalid URI.
    1.53 +     */
    1.54 +    public static final String INVALID_SCHEME = "net.fortunal.ical4j.invalid";
    1.55 +    
    1.56 +    private static final Pattern CID_PATTERN = Pattern.compile("(?i)^cid:.*");
    1.57 +    private static final Pattern NOTES_CID_REPLACEMENT_PATTERN = Pattern.compile("[<>]");
    1.58 +
    1.59 +    /**
    1.60 +     * Constructor made private to enforce static nature.
    1.61 +     */
    1.62 +    private Uris() {
    1.63 +    }
    1.64 +
    1.65 +    /**
    1.66 +     * Encodes the specified URI string using the UTF-8 charset. In the event that an exception is thrown, the specifed
    1.67 +     * URI string is returned unmodified.
    1.68 +     * @param s a URI string
    1.69 +     * @return an encoded URI string
    1.70 +     */
    1.71 +    public static String encode(final String s) {
    1.72 +        /*
    1.73 +         * try { return URLEncoder.encode(s, ENCODING_CHARSET); } catch (UnsupportedEncodingException use) {
    1.74 +         * log.error("Error ocurred encoding URI [" + s + "]", use); }
    1.75 +         */
    1.76 +
    1.77 +        /*
    1.78 +         * Lotus Notes does not correctly strip angle brackets from cid uris. From RFC2392: A "cid" URL is converted to
    1.79 +         * the corresponding Content-ID message header [MIME] by removing the "cid:" prefix, converting the % encoded
    1.80 +         * character to their equivalent US-ASCII characters, and enclosing the remaining parts with an angle bracket
    1.81 +         * pair, "<" and ">". For example, "cid:foo4%25foo1@bar.net" corresponds to Content-ID: <foo4%25foo1@bar.net>
    1.82 +         * Reversing the process and converting URL special characters to their % encodings produces the original cid. A
    1.83 +         * "mid" URL is converted to a Message-ID or Message-ID/Content-ID pair in a similar fashion.
    1.84 +         */
    1.85 +        if (CompatibilityHints.isHintEnabled(CompatibilityHints.KEY_NOTES_COMPATIBILITY)
    1.86 +        		&& CID_PATTERN.matcher(s).matches()) {
    1.87 +        	
    1.88 +            return NOTES_CID_REPLACEMENT_PATTERN.matcher(s).replaceAll("");
    1.89 +        }
    1.90 +        return s;
    1.91 +    }
    1.92 +
    1.93 +    /**
    1.94 +     * Decodes the specified URI string using the UTF-8 charset. In the event that an exception is thrown, the specifed
    1.95 +     * URI string is returned unmodified.
    1.96 +     * @param s a URI string
    1.97 +     * @return an encoded URI string
    1.98 +     */
    1.99 +    public static String decode(final String s) {
   1.100 +        /*
   1.101 +         * try { return URLDecoder.decode(s, ENCODING_CHARSET); } catch (UnsupportedEncodingException use) {
   1.102 +         * log.error("Error ocurred decoding URI [" + s + "]", use); }
   1.103 +         */
   1.104 +        return s;
   1.105 +    }
   1.106 +
   1.107 +    /**
   1.108 +     * Attempts to create a URI instance and will optionally swallow any resulting URISyntaxException depending on
   1.109 +     * configured {@link CompatibilityHints}. Will also automatically attempt encoding of the string representation for
   1.110 +     * greater compatibility.
   1.111 +     * <p>When relaxed parsing is enabled and if the string representation is not valid, a second URI creation attempt is made
   1.112 +     *  by extracting the scheme from the scheme specific part and URI encoding that later part. For example,
   1.113 +     *  "mailto: joe smith@example.com" becomes "mailto:joe%20smith@example.com".<p>
   1.114 +     * <p>If the second attempts also leads to a {@code URISyntaxException}, an opaque URI is constructed with a scheme
   1.115 +     *  of {@code Uris.INVALID_SCHEME} and a value corresponding to the initial representation.</p>
   1.116 +     * 
   1.117 +     * @param s a string representation of a URI.
   1.118 +     * @return a URI instance, which may not correspond to the URI string if a valid
   1.119 +     *  URI string is not specified and relaxed parsing is enabled.
   1.120 +     * @throws URISyntaxException if a valid URI string is not specified and relaxed parsing is disabled
   1.121 +     */
   1.122 +    public static URI create(final String s) throws URISyntaxException {
   1.123 +        try {
   1.124 +            return new URI(encode(s));
   1.125 +        }
   1.126 +        catch (URISyntaxException use) {
   1.127 +            if (CompatibilityHints
   1.128 +                    .isHintEnabled(CompatibilityHints.KEY_RELAXED_PARSING)) {
   1.129 +                String encoded = encode(s);
   1.130 +                int index = encoded.indexOf(':');
   1.131 +                if (index != -1 && index < encoded.length() -1) {
   1.132 +                    try {
   1.133 +                        return new URI(encoded.substring(0, index), encoded.substring(index + 1), null);
   1.134 +                    } catch (URISyntaxException use2) {
   1.135 +                    }
   1.136 +                }
   1.137 +                try {
   1.138 +                    return new URI(INVALID_SCHEME, s, null);
   1.139 +                } catch (URISyntaxException use2) {
   1.140 +                    // should not happen as we are building an opaque URI
   1.141 +                    throw new IllegalArgumentException("Could not build URI from " + s);
   1.142 +                }
   1.143 +            }
   1.144 +            throw use;
   1.145 +        }
   1.146 +    }
   1.147 +}

mercurial