001/*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  You may obtain a copy of the License at
008 *
009 *      http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017package org.apache.commons.validator.routines;
018
019import java.io.Serializable;
020import java.net.URI;
021import java.net.URISyntaxException;
022import java.util.Collections;
023import java.util.HashSet;
024import java.util.Locale;
025import java.util.Set;
026import java.util.regex.Matcher;
027import java.util.regex.Pattern;
028
029/**
030 * <p><b>URL Validation</b> routines.</p>
031 * Behavior of validation is modified by passing in options:
032 * <ul>
033 * <li>ALLOW_2_SLASHES - [FALSE]  Allows double '/' characters in the path
034 * component.</li>
035 * <li>NO_FRAGMENT- [FALSE]  By default fragments are allowed, if this option is
036 * included then fragments are flagged as illegal.</li>
037 * <li>ALLOW_ALL_SCHEMES - [FALSE] By default only http, https, and ftp are
038 * considered valid schemes.  Enabling this option will let any scheme pass validation.</li>
039 * </ul>
040 *
041 * <p>Originally based in on php script by Debbie Dyer, validation.php v1.2b, Date: 03/07/02,
042 * http://javascript.internet.com. However, this validation now bears little resemblance
043 * to the php original.</p>
044 * <pre>
045 *   Example of usage:
046 *   Construct a UrlValidator with valid schemes of "http", and "https".
047 *
048 *    String[] schemes = {"http","https"}.
049 *    UrlValidator urlValidator = new UrlValidator(schemes);
050 *    if (urlValidator.isValid("ftp://foo.bar.com/")) {
051 *       System.out.println("url is valid");
052 *    } else {
053 *       System.out.println("url is invalid");
054 *    }
055 *
056 *    prints "url is invalid"
057 *   If instead the default constructor is used.
058 *
059 *    UrlValidator urlValidator = new UrlValidator();
060 *    if (urlValidator.isValid("ftp://foo.bar.com/")) {
061 *       System.out.println("url is valid");
062 *    } else {
063 *       System.out.println("url is invalid");
064 *    }
065 *
066 *   prints out "url is valid"
067 *  </pre>
068 *
069 * @see
070 * <a href="http://www.ietf.org/rfc/rfc2396.txt">
071 *  Uniform Resource Identifiers (URI): Generic Syntax
072 * </a>
073 *
074 * @version $Revision: 1715435 $
075 * @since Validator 1.4
076 */
077public class UrlValidator implements Serializable {
078
079    private static final long serialVersionUID = 7557161713937335013L;
080
081    /**
082     * Allows all validly formatted schemes to pass validation instead of
083     * supplying a set of valid schemes.
084     */
085    public static final long ALLOW_ALL_SCHEMES = 1 << 0;
086
087    /**
088     * Allow two slashes in the path component of the URL.
089     */
090    public static final long ALLOW_2_SLASHES = 1 << 1;
091
092    /**
093     * Enabling this options disallows any URL fragments.
094     */
095    public static final long NO_FRAGMENTS = 1 << 2;
096
097    /**
098     * Allow local URLs, such as http://localhost/ or http://machine/ .
099     * This enables a broad-brush check, for complex local machine name
100     *  validation requirements you should create your validator with
101     *  a {@link RegexValidator} instead ({@link #UrlValidator(RegexValidator, long)})
102     */
103    public static final long ALLOW_LOCAL_URLS = 1 << 3; // CHECKSTYLE IGNORE MagicNumber
104
105    /**
106     * This expression derived/taken from the BNF for URI (RFC2396).
107     */
108    private static final String URL_REGEX =
109            "^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\\?([^#]*))?(#(.*))?";
110    //        12            3  4          5       6   7        8 9
111    private static final Pattern URL_PATTERN = Pattern.compile(URL_REGEX);
112
113    /**
114     * Schema/Protocol (ie. http:, ftp:, file:, etc).
115     */
116    private static final int PARSE_URL_SCHEME = 2;
117
118    /**
119     * Includes hostname/ip and port number.
120     */
121    private static final int PARSE_URL_AUTHORITY = 4;
122
123    private static final int PARSE_URL_PATH = 5;
124
125    private static final int PARSE_URL_QUERY = 7;
126
127    private static final int PARSE_URL_FRAGMENT = 9;
128
129    /**
130     * Protocol scheme (e.g. http, ftp, https).
131     */
132    private static final String SCHEME_REGEX = "^\\p{Alpha}[\\p{Alnum}\\+\\-\\.]*";
133    private static final Pattern SCHEME_PATTERN = Pattern.compile(SCHEME_REGEX);
134
135    // Drop numeric, and  "+-." for now
136    // TODO does not allow for optional userinfo. 
137    // Validation of character set is done by isValidAuthority
138    private static final String AUTHORITY_CHARS_REGEX = "\\p{Alnum}\\-\\."; // allows for IPV4 but not IPV6
139    private static final String IPV6_REGEX = "[0-9a-fA-F:]+"; // do this as separate match because : could cause ambiguity with port prefix
140
141    // userinfo    = *( unreserved / pct-encoded / sub-delims / ":" )
142    // unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~"
143    // sub-delims    = "!" / "$" / "&" / "'" / "(" / ")" / "*" / "+" / "," / ";" / "="
144    // We assume that password has the same valid chars as user info
145    private static final String USERINFO_CHARS_REGEX = "[a-zA-Z0-9%-._~!$&'()*+,;=]";
146    // since neither ':' nor '@' are allowed chars, we don't need to use non-greedy matching
147    private static final String USERINFO_FIELD_REGEX =
148            USERINFO_CHARS_REGEX + "+:" + // At least one character for the name
149            USERINFO_CHARS_REGEX + "*@"; // password may be absent
150    private static final String AUTHORITY_REGEX =
151            "(?:\\[("+IPV6_REGEX+")\\]|(?:(?:"+USERINFO_FIELD_REGEX+")?([" + AUTHORITY_CHARS_REGEX + "]*)))(:\\d*)?(.*)?";
152    //             1                          e.g. user:pass@          2                                   3       4
153    private static final Pattern AUTHORITY_PATTERN = Pattern.compile(AUTHORITY_REGEX);
154
155    private static final int PARSE_AUTHORITY_IPV6 = 1;
156
157    private static final int PARSE_AUTHORITY_HOST_IP = 2; // excludes userinfo, if present
158
159    // Not needed, because it is validated by AUTHORITY_REGEX
160//    private static final int PARSE_AUTHORITY_PORT = 3;
161
162    /**
163     * Should always be empty. The code currently allows spaces.
164     */
165    private static final int PARSE_AUTHORITY_EXTRA = 4;
166
167    private static final String PATH_REGEX = "^(/[-\\w:@&?=+,.!/~*'%$_;\\(\\)]*)?$";
168    private static final Pattern PATH_PATTERN = Pattern.compile(PATH_REGEX);
169
170    private static final String QUERY_REGEX = "^(.*)$";
171    private static final Pattern QUERY_PATTERN = Pattern.compile(QUERY_REGEX);
172
173    /**
174     * Holds the set of current validation options.
175     */
176    private final long options;
177
178    /**
179     * The set of schemes that are allowed to be in a URL.
180     */
181    private final Set<String> allowedSchemes; // Must be lower-case
182
183    /**
184     * Regular expressions used to manually validate authorities if IANA
185     * domain name validation isn't desired.
186     */
187    private final RegexValidator authorityValidator;
188
189    /**
190     * If no schemes are provided, default to this set.
191     */
192    private static final String[] DEFAULT_SCHEMES = {"http", "https", "ftp"}; // Must be lower-case
193
194    /**
195     * Singleton instance of this class with default schemes and options.
196     */
197    private static final UrlValidator DEFAULT_URL_VALIDATOR = new UrlValidator();
198
199    /**
200     * Returns the singleton instance of this class with default schemes and options.
201     * @return singleton instance with default schemes and options
202     */
203    public static UrlValidator getInstance() {
204        return DEFAULT_URL_VALIDATOR;
205    }
206
207    /**
208     * Create a UrlValidator with default properties.
209     */
210    public UrlValidator() {
211        this(null);
212    }
213
214    /**
215     * Behavior of validation is modified by passing in several strings options:
216     * @param schemes Pass in one or more url schemes to consider valid, passing in
217     *        a null will default to "http,https,ftp" being valid.
218     *        If a non-null schemes is specified then all valid schemes must
219     *        be specified. Setting the ALLOW_ALL_SCHEMES option will
220     *        ignore the contents of schemes.
221     */
222    public UrlValidator(String[] schemes) {
223        this(schemes, 0L);
224    }
225
226    /**
227     * Initialize a UrlValidator with the given validation options.
228     * @param options The options should be set using the public constants declared in
229     * this class.  To set multiple options you simply add them together.  For example,
230     * ALLOW_2_SLASHES + NO_FRAGMENTS enables both of those options.
231     */
232    public UrlValidator(long options) {
233        this(null, null, options);
234    }
235
236    /**
237     * Behavior of validation is modified by passing in options:
238     * @param schemes The set of valid schemes. Ignored if the ALLOW_ALL_SCHEMES option is set.
239     * @param options The options should be set using the public constants declared in
240     * this class.  To set multiple options you simply add them together.  For example,
241     * ALLOW_2_SLASHES + NO_FRAGMENTS enables both of those options.
242     */
243    public UrlValidator(String[] schemes, long options) {
244        this(schemes, null, options);
245    }
246
247    /**
248     * Initialize a UrlValidator with the given validation options.
249     * @param authorityValidator Regular expression validator used to validate the authority part
250     * This allows the user to override the standard set of domains.
251     * @param options Validation options. Set using the public constants of this class.
252     * To set multiple options, simply add them together:
253     * <p><code>ALLOW_2_SLASHES + NO_FRAGMENTS</code></p>
254     * enables both of those options.
255     */
256    public UrlValidator(RegexValidator authorityValidator, long options) {
257        this(null, authorityValidator, options);
258    }
259
260    /**
261     * Customizable constructor. Validation behavior is modifed by passing in options.
262     * @param schemes the set of valid schemes. Ignored if the ALLOW_ALL_SCHEMES option is set.
263     * @param authorityValidator Regular expression validator used to validate the authority part
264     * @param options Validation options. Set using the public constants of this class.
265     * To set multiple options, simply add them together:
266     * <p><code>ALLOW_2_SLASHES + NO_FRAGMENTS</code></p>
267     * enables both of those options.
268     */
269    public UrlValidator(String[] schemes, RegexValidator authorityValidator, long options) {
270        this.options = options;
271
272        if (isOn(ALLOW_ALL_SCHEMES)) {
273            allowedSchemes = Collections.emptySet();
274        } else {
275            if (schemes == null) {
276                schemes = DEFAULT_SCHEMES;
277            }
278            allowedSchemes = new HashSet<String>(schemes.length);
279            for(int i=0; i < schemes.length; i++) {
280                allowedSchemes.add(schemes[i].toLowerCase(Locale.ENGLISH));
281            }
282        }
283
284        this.authorityValidator = authorityValidator;
285    }
286
287    /**
288     * <p>Checks if a field has a valid url address.</p>
289     *
290     * Note that the method calls #isValidAuthority()
291     * which checks that the domain is valid.
292     *
293     * @param value The value validation is being performed on.  A <code>null</code>
294     * value is considered invalid.
295     * @return true if the url is valid.
296     */
297    public boolean isValid(String value) {
298        if (value == null) {
299            return false;
300        }
301
302        // Check the whole url address structure
303        Matcher urlMatcher = URL_PATTERN.matcher(value);
304        if (!urlMatcher.matches()) {
305            return false;
306        }
307
308        String scheme = urlMatcher.group(PARSE_URL_SCHEME);
309        if (!isValidScheme(scheme)) {
310            return false;
311        }
312
313        String authority = urlMatcher.group(PARSE_URL_AUTHORITY);
314        if ("file".equals(scheme)) {// Special case - file: allows an empty authority
315            if (!"".equals(authority)) {
316                if (authority.contains(":")) { // but cannot allow trailing :
317                    return false;
318                }
319            }
320            // drop through to continue validation
321        } else { // not file:
322            // Validate the authority
323            if (!isValidAuthority(authority)) {
324                return false;
325            }
326        }
327
328        if (!isValidPath(urlMatcher.group(PARSE_URL_PATH))) {
329            return false;
330        }
331
332        if (!isValidQuery(urlMatcher.group(PARSE_URL_QUERY))) {
333            return false;
334        }
335
336        if (!isValidFragment(urlMatcher.group(PARSE_URL_FRAGMENT))) {
337            return false;
338        }
339
340        return true;
341    }
342
343    /**
344     * Validate scheme. If schemes[] was initialized to a non null,
345     * then only those schemes are allowed.
346     * Otherwise the default schemes are "http", "https", "ftp".
347     * Matching is case-blind.
348     * @param scheme The scheme to validate.  A <code>null</code> value is considered
349     * invalid.
350     * @return true if valid.
351     */
352    protected boolean isValidScheme(String scheme) {
353        if (scheme == null) {
354            return false;
355        }
356
357        // TODO could be removed if external schemes were checked in the ctor before being stored
358        if (!SCHEME_PATTERN.matcher(scheme).matches()) {
359            return false;
360        }
361
362        if (isOff(ALLOW_ALL_SCHEMES) && !allowedSchemes.contains(scheme.toLowerCase(Locale.ENGLISH))) {
363            return false;
364        }
365
366        return true;
367    }
368
369    /**
370     * Returns true if the authority is properly formatted.  An authority is the combination
371     * of hostname and port.  A <code>null</code> authority value is considered invalid.
372     * Note: this implementation validates the domain unless a RegexValidator was provided.
373     * If a RegexValidator was supplied and it matches, then the authority is regarded
374     * as valid with no further checks, otherwise the method checks against the
375     * AUTHORITY_PATTERN and the DomainValidator (ALLOW_LOCAL_URLS)
376     * @param authority Authority value to validate, alllows IDN
377     * @return true if authority (hostname and port) is valid.
378     */
379    protected boolean isValidAuthority(String authority) {
380        if (authority == null) {
381            return false;
382        }
383
384        // check manual authority validation if specified
385        if (authorityValidator != null && authorityValidator.isValid(authority)) {
386            return true;
387        }
388        // convert to ASCII if possible
389        final String authorityASCII = DomainValidator.unicodeToASCII(authority);
390
391        Matcher authorityMatcher = AUTHORITY_PATTERN.matcher(authorityASCII);
392        if (!authorityMatcher.matches()) {
393            return false;
394        }
395
396        // We have to process IPV6 separately because that is parsed in a different group
397        String ipv6 = authorityMatcher.group(PARSE_AUTHORITY_IPV6);
398        if (ipv6 != null) {
399            InetAddressValidator inetAddressValidator = InetAddressValidator.getInstance();
400                if (!inetAddressValidator.isValidInet6Address(ipv6)) {
401                    return false;
402                }
403        } else {
404            String hostLocation = authorityMatcher.group(PARSE_AUTHORITY_HOST_IP);
405            // check if authority is hostname or IP address:
406            // try a hostname first since that's much more likely
407            DomainValidator domainValidator = DomainValidator.getInstance(isOn(ALLOW_LOCAL_URLS));
408            if (!domainValidator.isValid(hostLocation)) {
409                // try an IPv4 address
410                InetAddressValidator inetAddressValidator = InetAddressValidator.getInstance();
411                if (!inetAddressValidator.isValidInet4Address(hostLocation)) {
412                    // isn't IPv4, so the URL is invalid
413                    return false;
414                }
415            }
416        }
417
418        String extra = authorityMatcher.group(PARSE_AUTHORITY_EXTRA);
419        if (extra != null && extra.trim().length() > 0){
420            return false;
421        }
422
423        return true;
424    }
425
426    /**
427     * Returns true if the path is valid.  A <code>null</code> value is considered invalid.
428     * @param path Path value to validate.
429     * @return true if path is valid.
430     */
431    protected boolean isValidPath(String path) {
432        if (path == null) {
433            return false;
434        }
435
436        if (!PATH_PATTERN.matcher(path).matches()) {
437            return false;
438        }
439
440        try {
441            URI uri = new URI(null,null,path,null);
442            String norm = uri.normalize().getPath();
443            if (norm.startsWith("/../") // Trying to go via the parent dir 
444             || norm.equals("/..")) {   // Trying to go to the parent dir
445                return false;
446            }
447        } catch (URISyntaxException e) {
448            return false;
449        }
450        
451        int slash2Count = countToken("//", path);
452        if (isOff(ALLOW_2_SLASHES) && (slash2Count > 0)) {
453            return false;
454        }
455
456        return true;
457    }
458
459    /**
460     * Returns true if the query is null or it's a properly formatted query string.
461     * @param query Query value to validate.
462     * @return true if query is valid.
463     */
464    protected boolean isValidQuery(String query) {
465        if (query == null) {
466            return true;
467        }
468
469        return QUERY_PATTERN.matcher(query).matches();
470    }
471
472    /**
473     * Returns true if the given fragment is null or fragments are allowed.
474     * @param fragment Fragment value to validate.
475     * @return true if fragment is valid.
476     */
477    protected boolean isValidFragment(String fragment) {
478        if (fragment == null) {
479            return true;
480        }
481
482        return isOff(NO_FRAGMENTS);
483    }
484
485    /**
486     * Returns the number of times the token appears in the target.
487     * @param token Token value to be counted.
488     * @param target Target value to count tokens in.
489     * @return the number of tokens.
490     */
491    protected int countToken(String token, String target) {
492        int tokenIndex = 0;
493        int count = 0;
494        while (tokenIndex != -1) {
495            tokenIndex = target.indexOf(token, tokenIndex);
496            if (tokenIndex > -1) {
497                tokenIndex++;
498                count++;
499            }
500        }
501        return count;
502    }
503
504    /**
505     * Tests whether the given flag is on.  If the flag is not a power of 2
506     * (ie. 3) this tests whether the combination of flags is on.
507     *
508     * @param flag Flag value to check.
509     *
510     * @return whether the specified flag value is on.
511     */
512    private boolean isOn(long flag) {
513        return (options & flag) > 0;
514    }
515
516    /**
517     * Tests whether the given flag is off.  If the flag is not a power of 2
518     * (ie. 3) this tests whether the combination of flags is off.
519     *
520     * @param flag Flag value to check.
521     *
522     * @return whether the specified flag value is off.
523     */
524    private boolean isOff(long flag) {
525        return (options & flag) == 0;
526    }
527
528    // Unit test access to pattern matcher
529    Matcher matchURL(String value) {
530        return URL_PATTERN.matcher(value);
531    }
532}