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: 1783203 $
075 * @since Validator 1.4
076 */
077public class UrlValidator implements Serializable {
078
079    private static final long serialVersionUID = 7557161713937335013L;
080
081    private static final int MAX_UNSIGNED_16_BIT_INT = 0xFFFF; // port max
082
083    /**
084     * Allows all validly formatted schemes to pass validation instead of
085     * supplying a set of valid schemes.
086     */
087    public static final long ALLOW_ALL_SCHEMES = 1 << 0;
088
089    /**
090     * Allow two slashes in the path component of the URL.
091     */
092    public static final long ALLOW_2_SLASHES = 1 << 1;
093
094    /**
095     * Enabling this options disallows any URL fragments.
096     */
097    public static final long NO_FRAGMENTS = 1 << 2;
098
099    /**
100     * Allow local URLs, such as http://localhost/ or http://machine/ .
101     * This enables a broad-brush check, for complex local machine name
102     *  validation requirements you should create your validator with
103     *  a {@link RegexValidator} instead ({@link #UrlValidator(RegexValidator, long)})
104     */
105    public static final long ALLOW_LOCAL_URLS = 1 << 3; // CHECKSTYLE IGNORE MagicNumber
106
107    /**
108     * This expression derived/taken from the BNF for URI (RFC2396).
109     */
110    private static final String URL_REGEX =
111            "^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\\?([^#]*))?(#(.*))?";
112    //        12            3  4          5       6   7        8 9
113    private static final Pattern URL_PATTERN = Pattern.compile(URL_REGEX);
114
115    /**
116     * Schema/Protocol (ie. http:, ftp:, file:, etc).
117     */
118    private static final int PARSE_URL_SCHEME = 2;
119
120    /**
121     * Includes hostname/ip and port number.
122     */
123    private static final int PARSE_URL_AUTHORITY = 4;
124
125    private static final int PARSE_URL_PATH = 5;
126
127    private static final int PARSE_URL_QUERY = 7;
128
129    private static final int PARSE_URL_FRAGMENT = 9;
130
131    /**
132     * Protocol scheme (e.g. http, ftp, https).
133     */
134    private static final String SCHEME_REGEX = "^\\p{Alpha}[\\p{Alnum}\\+\\-\\.]*";
135    private static final Pattern SCHEME_PATTERN = Pattern.compile(SCHEME_REGEX);
136
137    // Drop numeric, and  "+-." for now
138    // TODO does not allow for optional userinfo. 
139    // Validation of character set is done by isValidAuthority
140    private static final String AUTHORITY_CHARS_REGEX = "\\p{Alnum}\\-\\."; // allows for IPV4 but not IPV6
141    private static final String IPV6_REGEX = "[0-9a-fA-F:]+"; // do this as separate match because : could cause ambiguity with port prefix
142
143    // userinfo    = *( unreserved / pct-encoded / sub-delims / ":" )
144    // unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~"
145    // sub-delims    = "!" / "$" / "&" / "'" / "(" / ")" / "*" / "+" / "," / ";" / "="
146    // We assume that password has the same valid chars as user info
147    private static final String USERINFO_CHARS_REGEX = "[a-zA-Z0-9%-._~!$&'()*+,;=]";
148    // since neither ':' nor '@' are allowed chars, we don't need to use non-greedy matching
149    private static final String USERINFO_FIELD_REGEX =
150            USERINFO_CHARS_REGEX + "+" + // At least one character for the name
151            "(?::" + USERINFO_CHARS_REGEX + "*)?@"; // colon and password may be absent
152    private static final String AUTHORITY_REGEX =
153            "(?:\\[("+IPV6_REGEX+")\\]|(?:(?:"+USERINFO_FIELD_REGEX+")?([" + AUTHORITY_CHARS_REGEX + "]*)))(?::(\\d*))?(.*)?";
154    //             1                          e.g. user:pass@          2                                         3       4
155    private static final Pattern AUTHORITY_PATTERN = Pattern.compile(AUTHORITY_REGEX);
156
157    private static final int PARSE_AUTHORITY_IPV6 = 1;
158
159    private static final int PARSE_AUTHORITY_HOST_IP = 2; // excludes userinfo, if present
160
161    private static final int PARSE_AUTHORITY_PORT = 3; // excludes leading colon
162
163    /**
164     * Should always be empty. The code currently allows spaces.
165     */
166    private static final int PARSE_AUTHORITY_EXTRA = 4;
167
168    private static final String PATH_REGEX = "^(/[-\\w:@&?=+,.!/~*'%$_;\\(\\)]*)?$";
169    private static final Pattern PATH_PATTERN = Pattern.compile(PATH_REGEX);
170
171    private static final String QUERY_REGEX = "^(\\S*)$";
172    private static final Pattern QUERY_PATTERN = Pattern.compile(QUERY_REGEX);
173
174    /**
175     * Holds the set of current validation options.
176     */
177    private final long options;
178
179    /**
180     * The set of schemes that are allowed to be in a URL.
181     */
182    private final Set<String> allowedSchemes; // Must be lower-case
183
184    /**
185     * Regular expressions used to manually validate authorities if IANA
186     * domain name validation isn't desired.
187     */
188    private final RegexValidator authorityValidator;
189
190    /**
191     * If no schemes are provided, default to this set.
192     */
193    private static final String[] DEFAULT_SCHEMES = {"http", "https", "ftp"}; // Must be lower-case
194
195    /**
196     * Singleton instance of this class with default schemes and options.
197     */
198    private static final UrlValidator DEFAULT_URL_VALIDATOR = new UrlValidator();
199
200    /**
201     * Returns the singleton instance of this class with default schemes and options.
202     * @return singleton instance with default schemes and options
203     */
204    public static UrlValidator getInstance() {
205        return DEFAULT_URL_VALIDATOR;
206    }
207
208    /**
209     * Create a UrlValidator with default properties.
210     */
211    public UrlValidator() {
212        this(null);
213    }
214
215    /**
216     * Behavior of validation is modified by passing in several strings options:
217     * @param schemes Pass in one or more url schemes to consider valid, passing in
218     *        a null will default to "http,https,ftp" being valid.
219     *        If a non-null schemes is specified then all valid schemes must
220     *        be specified. Setting the ALLOW_ALL_SCHEMES option will
221     *        ignore the contents of schemes.
222     */
223    public UrlValidator(String[] schemes) {
224        this(schemes, 0L);
225    }
226
227    /**
228     * Initialize a UrlValidator with the given validation options.
229     * @param options The options should be set using the public constants declared in
230     * this class.  To set multiple options you simply add them together.  For example,
231     * ALLOW_2_SLASHES + NO_FRAGMENTS enables both of those options.
232     */
233    public UrlValidator(long options) {
234        this(null, null, options);
235    }
236
237    /**
238     * Behavior of validation is modified by passing in options:
239     * @param schemes The set of valid schemes. Ignored if the ALLOW_ALL_SCHEMES option is set.
240     * @param options The options should be set using the public constants declared in
241     * this class.  To set multiple options you simply add them together.  For example,
242     * ALLOW_2_SLASHES + NO_FRAGMENTS enables both of those options.
243     */
244    public UrlValidator(String[] schemes, long options) {
245        this(schemes, null, options);
246    }
247
248    /**
249     * Initialize a UrlValidator with the given validation options.
250     * @param authorityValidator Regular expression validator used to validate the authority part
251     * This allows the user to override the standard set of domains.
252     * @param options Validation options. Set using the public constants of this class.
253     * To set multiple options, simply add them together:
254     * <p><code>ALLOW_2_SLASHES + NO_FRAGMENTS</code></p>
255     * enables both of those options.
256     */
257    public UrlValidator(RegexValidator authorityValidator, long options) {
258        this(null, authorityValidator, options);
259    }
260
261    /**
262     * Customizable constructor. Validation behavior is modifed by passing in options.
263     * @param schemes the set of valid schemes. Ignored if the ALLOW_ALL_SCHEMES option is set.
264     * @param authorityValidator Regular expression validator used to validate the authority part
265     * @param options Validation options. Set using the public constants of this class.
266     * To set multiple options, simply add them together:
267     * <p><code>ALLOW_2_SLASHES + NO_FRAGMENTS</code></p>
268     * enables both of those options.
269     */
270    public UrlValidator(String[] schemes, RegexValidator authorityValidator, long options) {
271        this.options = options;
272
273        if (isOn(ALLOW_ALL_SCHEMES)) {
274            allowedSchemes = Collections.emptySet();
275        } else {
276            if (schemes == null) {
277                schemes = DEFAULT_SCHEMES;
278            }
279            allowedSchemes = new HashSet<String>(schemes.length);
280            for(int i=0; i < schemes.length; i++) {
281                allowedSchemes.add(schemes[i].toLowerCase(Locale.ENGLISH));
282            }
283        }
284
285        this.authorityValidator = authorityValidator;
286    }
287
288    /**
289     * <p>Checks if a field has a valid url address.</p>
290     *
291     * Note that the method calls #isValidAuthority()
292     * which checks that the domain is valid.
293     *
294     * @param value The value validation is being performed on.  A <code>null</code>
295     * value is considered invalid.
296     * @return true if the url is valid.
297     */
298    public boolean isValid(String value) {
299        if (value == null) {
300            return false;
301        }
302
303        // Check the whole url address structure
304        Matcher urlMatcher = URL_PATTERN.matcher(value);
305        if (!urlMatcher.matches()) {
306            return false;
307        }
308
309        String scheme = urlMatcher.group(PARSE_URL_SCHEME);
310        if (!isValidScheme(scheme)) {
311            return false;
312        }
313
314        String authority = urlMatcher.group(PARSE_URL_AUTHORITY);
315        if ("file".equals(scheme)) {// Special case - file: allows an empty authority
316            if (authority != null) {
317                if (authority.contains(":")) { // but cannot allow trailing :
318                    return false;
319                }
320            }
321            // drop through to continue validation
322        } else { // not file:
323            // Validate the authority
324            if (!isValidAuthority(authority)) {
325                return false;
326            }
327        }
328
329        if (!isValidPath(urlMatcher.group(PARSE_URL_PATH))) {
330            return false;
331        }
332
333        if (!isValidQuery(urlMatcher.group(PARSE_URL_QUERY))) {
334            return false;
335        }
336
337        if (!isValidFragment(urlMatcher.group(PARSE_URL_FRAGMENT))) {
338            return false;
339        }
340
341        return true;
342    }
343
344    /**
345     * Validate scheme. If schemes[] was initialized to a non null,
346     * then only those schemes are allowed.
347     * Otherwise the default schemes are "http", "https", "ftp".
348     * Matching is case-blind.
349     * @param scheme The scheme to validate.  A <code>null</code> value is considered
350     * invalid.
351     * @return true if valid.
352     */
353    protected boolean isValidScheme(String scheme) {
354        if (scheme == null) {
355            return false;
356        }
357
358        // TODO could be removed if external schemes were checked in the ctor before being stored
359        if (!SCHEME_PATTERN.matcher(scheme).matches()) {
360            return false;
361        }
362
363        if (isOff(ALLOW_ALL_SCHEMES) && !allowedSchemes.contains(scheme.toLowerCase(Locale.ENGLISH))) {
364            return false;
365        }
366
367        return true;
368    }
369
370    /**
371     * Returns true if the authority is properly formatted.  An authority is the combination
372     * of hostname and port.  A <code>null</code> authority value is considered invalid.
373     * Note: this implementation validates the domain unless a RegexValidator was provided.
374     * If a RegexValidator was supplied and it matches, then the authority is regarded
375     * as valid with no further checks, otherwise the method checks against the
376     * AUTHORITY_PATTERN and the DomainValidator (ALLOW_LOCAL_URLS)
377     * @param authority Authority value to validate, alllows IDN
378     * @return true if authority (hostname and port) is valid.
379     */
380    protected boolean isValidAuthority(String authority) {
381        if (authority == null) {
382            return false;
383        }
384
385        // check manual authority validation if specified
386        if (authorityValidator != null && authorityValidator.isValid(authority)) {
387            return true;
388        }
389        // convert to ASCII if possible
390        final String authorityASCII = DomainValidator.unicodeToASCII(authority);
391
392        Matcher authorityMatcher = AUTHORITY_PATTERN.matcher(authorityASCII);
393        if (!authorityMatcher.matches()) {
394            return false;
395        }
396
397        // We have to process IPV6 separately because that is parsed in a different group
398        String ipv6 = authorityMatcher.group(PARSE_AUTHORITY_IPV6);
399        if (ipv6 != null) {
400            InetAddressValidator inetAddressValidator = InetAddressValidator.getInstance();
401                if (!inetAddressValidator.isValidInet6Address(ipv6)) {
402                    return false;
403                }
404        } else {
405            String hostLocation = authorityMatcher.group(PARSE_AUTHORITY_HOST_IP);
406            // check if authority is hostname or IP address:
407            // try a hostname first since that's much more likely
408            DomainValidator domainValidator = DomainValidator.getInstance(isOn(ALLOW_LOCAL_URLS));
409            if (!domainValidator.isValid(hostLocation)) {
410                // try an IPv4 address
411                InetAddressValidator inetAddressValidator = InetAddressValidator.getInstance();
412                if (!inetAddressValidator.isValidInet4Address(hostLocation)) {
413                    // isn't IPv4, so the URL is invalid
414                    return false;
415                }
416            }
417            String port = authorityMatcher.group(PARSE_AUTHORITY_PORT);
418            if (port != null && port.length() > 0) {
419                try {
420                    int iPort = Integer.parseInt(port);
421                    if (iPort < 0 || iPort > MAX_UNSIGNED_16_BIT_INT) {
422                        return false;
423                    }
424                } catch (NumberFormatException nfe) {
425                    return false; // this can happen for big numbers
426                }
427            }
428        }
429
430        String extra = authorityMatcher.group(PARSE_AUTHORITY_EXTRA);
431        if (extra != null && extra.trim().length() > 0){
432            return false;
433        }
434
435        return true;
436    }
437
438    /**
439     * Returns true if the path is valid.  A <code>null</code> value is considered invalid.
440     * @param path Path value to validate.
441     * @return true if path is valid.
442     */
443    protected boolean isValidPath(String path) {
444        if (path == null) {
445            return false;
446        }
447
448        if (!PATH_PATTERN.matcher(path).matches()) {
449            return false;
450        }
451
452        try {
453            URI uri = new URI(null,null,path,null);
454            String norm = uri.normalize().getPath();
455            if (norm.startsWith("/../") // Trying to go via the parent dir 
456             || norm.equals("/..")) {   // Trying to go to the parent dir
457                return false;
458            }
459        } catch (URISyntaxException e) {
460            return false;
461        }
462        
463        int slash2Count = countToken("//", path);
464        if (isOff(ALLOW_2_SLASHES) && (slash2Count > 0)) {
465            return false;
466        }
467
468        return true;
469    }
470
471    /**
472     * Returns true if the query is null or it's a properly formatted query string.
473     * @param query Query value to validate.
474     * @return true if query is valid.
475     */
476    protected boolean isValidQuery(String query) {
477        if (query == null) {
478            return true;
479        }
480
481        return QUERY_PATTERN.matcher(query).matches();
482    }
483
484    /**
485     * Returns true if the given fragment is null or fragments are allowed.
486     * @param fragment Fragment value to validate.
487     * @return true if fragment is valid.
488     */
489    protected boolean isValidFragment(String fragment) {
490        if (fragment == null) {
491            return true;
492        }
493
494        return isOff(NO_FRAGMENTS);
495    }
496
497    /**
498     * Returns the number of times the token appears in the target.
499     * @param token Token value to be counted.
500     * @param target Target value to count tokens in.
501     * @return the number of tokens.
502     */
503    protected int countToken(String token, String target) {
504        int tokenIndex = 0;
505        int count = 0;
506        while (tokenIndex != -1) {
507            tokenIndex = target.indexOf(token, tokenIndex);
508            if (tokenIndex > -1) {
509                tokenIndex++;
510                count++;
511            }
512        }
513        return count;
514    }
515
516    /**
517     * Tests whether the given flag is on.  If the flag is not a power of 2
518     * (ie. 3) this tests whether the combination of flags is on.
519     *
520     * @param flag Flag value to check.
521     *
522     * @return whether the specified flag value is on.
523     */
524    private boolean isOn(long flag) {
525        return (options & flag) > 0;
526    }
527
528    /**
529     * Tests whether the given flag is off.  If the flag is not a power of 2
530     * (ie. 3) this tests whether the combination of flags is off.
531     *
532     * @param flag Flag value to check.
533     *
534     * @return whether the specified flag value is off.
535     */
536    private boolean isOff(long flag) {
537        return (options & flag) == 0;
538    }
539
540    // Unit test access to pattern matcher
541    Matcher matchURL(String value) {
542        return URL_PATTERN.matcher(value);
543    }
544}