View Javadoc

1   /*
2    * $Header: /home/cvs/jakarta-commons/validator/src/share/org/apache/commons/validator/UrlValidator.java,v 1.19 2004/02/21 17:10:29 rleland Exp $
3    * $Revision: 1.19 $
4    * $Date: 2004/02/21 17:10:29 $
5    *
6    * ====================================================================
7    * Copyright 2001-2004 The Apache Software Foundation
8    *
9    * Licensed under the Apache License, Version 2.0 (the "License");
10   * you may not use this file except in compliance with the License.
11   * You may obtain a copy of the License at
12   *
13   *     http://www.apache.org/licenses/LICENSE-2.0
14   *
15   * Unless required by applicable law or agreed to in writing, software
16   * distributed under the License is distributed on an "AS IS" BASIS,
17   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
18   * See the License for the specific language governing permissions and
19   * limitations under the License.
20   */
21  
22  package org.apache.commons.validator;
23  
24  import java.io.Serializable;
25  import java.util.Arrays;
26  import java.util.HashSet;
27  import java.util.Set;
28  
29  import org.apache.commons.validator.util.Flags;
30  import org.apache.oro.text.perl.Perl5Util;
31  
32  /***
33   * <p>Validates URLs.</p>
34   * Behavour of validation is modified by passing in options:
35   * <li>ALLOW_2_SLASHES - [FALSE]  Allows double '/' characters in the path
36   * component.</li>
37   * <li>NO_FRAGMENT- [FALSE]  By default fragments are allowed, if this option is
38   * included then fragments are flagged as illegal.</li>
39   * <li>ALLOW_ALL_SCHEMES - [FALSE] By default only http, https, and ftp are
40   * considered valid schemes.  Enabling this option will let any scheme pass validation.</li>
41   *
42   * <p>Originally based in on php script by Debbie Dyer, validation.php v1.2b, Date: 03/07/02,
43   * http://javascript.internet.com. However, this validation now bears little resemblance
44   * to the php original.</p>
45   * <pre>
46   *   Example of usage:
47   *   Construct a UrlValidator with valid schemes of "http", and "https".
48   *
49   *    String[] schemes = {"http","https"}.
50   *    Urlvalidator urlValidator = new Urlvalidator(schemes);
51   *    if (urlValidator.isValid("ftp")) {
52   *       System.out.println("url is valid");
53   *    } else {
54   *       System.out.println("url is invalid");
55   *    }
56   *
57   *    prints "url is invalid"
58   *   If instead the default constructor is used.
59   *
60   *    Urlvalidator urlValidator = new Urlvalidator();
61   *    if (urlValidator.isValid("ftp")) {
62   *       System.out.println("url is valid");
63   *    } else {
64   *       System.out.println("url is invalid");
65   *    }
66   *
67   *   prints out "url is valid"
68   *  </pre>
69   *
70   * @see
71   * <a href='http://www.ietf.org/rfc/rfc2396.txt' >
72   *  Uniform Resource Identifiers (URI): Generic Syntax
73   * </a>
74   *
75   * @since Validator 1.1
76   */
77  public class UrlValidator implements Serializable {
78  
79      /***
80       * Allows all validly formatted schemes to pass validation instead of supplying a
81       * set of valid schemes.
82       */
83      public static final int ALLOW_ALL_SCHEMES = 1 << 0;
84  
85      /***
86       * Allow two slashes in the path component of the URL.
87       */
88      public static final int ALLOW_2_SLASHES = 1 << 1;
89  
90      /***
91       * Enabling this options disallows any URL fragments.
92       */
93      public static final int NO_FRAGMENTS = 1 << 2;
94  
95      private static final String ALPHA_CHARS = "a-zA-Z";
96  
97      private static final String ALPHA_NUMERIC_CHARS = ALPHA_CHARS + "//d";
98  
99      private static final String SPECIAL_CHARS = ";/@&=,.?:+$";
100 
101     private static final String VALID_CHARS = "[^//s" + SPECIAL_CHARS + "]";
102 
103     private static final String SCHEME_CHARS = ALPHA_CHARS;
104 
105     // Drop numeric, and  "+-." for now
106     private static final String AUTHORITY_CHARS = ALPHA_NUMERIC_CHARS + "//-//.";
107 
108     private static final String ATOM = VALID_CHARS + '+';
109 
110     /***
111      * This expression derived/taken from the BNF for URI (RFC2396).
112      */
113     private static final String URL_PATTERN =
114             "/^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(//?([^#]*))?(#(.*))?/";
115     //                                                                      12            3  4          5       6   7        8 9
116 
117     /***
118      * Schema/Protocol (ie. http:, ftp:, file:, etc).
119      */
120     private static final int PARSE_URL_SCHEME = 2;
121 
122     /***
123      * Includes hostname/ip and port number.
124      */
125     private static final int PARSE_URL_AUTHORITY = 4;
126 
127     private static final int PARSE_URL_PATH = 5;
128 
129     private static final int PARSE_URL_QUERY = 7;
130 
131     private static final int PARSE_URL_FRAGMENT = 9;
132 
133     /***
134      * Protocol (ie. http:, ftp:,https:).
135      */
136     private static final String SCHEME_PATTERN = "/^[" + SCHEME_CHARS + "]/";
137 
138     private static final String AUTHORITY_PATTERN =
139             "/^([" + AUTHORITY_CHARS + "]*)(://d*)?(.*)?/";
140     //                                                                            1                          2  3       4
141 
142     private static final int PARSE_AUTHORITY_HOST_IP = 1;
143 
144     private static final int PARSE_AUTHORITY_PORT = 2;
145 
146     /***
147      * Should always be empty.
148      */
149     private static final int PARSE_AUTHORITY_EXTRA = 3;
150 
151     private static final String PATH_PATTERN =
152             "/^(/[-a-zA-Z0-9_:@&?=+,.!/~*'%$]*)$/";
153 
154     private static final String QUERY_PATTERN = "/^(.*)$/";
155 
156     private static final String LEGAL_ASCII_PATTERN = "/^[//000-//177]+$/";
157 
158     private static final String IP_V4_DOMAIN_PATTERN =
159             "/^(//d{1,3})[.](//d{1,3})[.](//d{1,3})[.](//d{1,3})$/";
160 
161     private static final String DOMAIN_PATTERN =
162             "/^" + ATOM + "(//." + ATOM + ")*$/";
163 
164     private static final String PORT_PATTERN = "/^:(//d{1,5})$/";
165 
166     private static final String ATOM_PATTERN = "/(" + ATOM + ")/";
167 
168     private static final String ALPHA_PATTERN = "/^[" + ALPHA_CHARS + "]/";
169 
170     /***
171      * Holds the set of current validation options.
172      */
173     private Flags options = null;
174 
175     /***
176      * The set of schemes that are allowed to be in a URL.
177      */
178     private Set allowedSchemes = new HashSet();
179 
180     /***
181      * If no schemes are provided, default to this set.
182      */
183     protected String[] defaultSchemes = {"http", "https", "ftp"};
184 
185     /***
186      * Create a UrlValidator with default properties.
187      */
188     public UrlValidator() {
189         this(null);
190     }
191 
192     /***
193      * Behavior of validation is modified by passing in several strings options:
194      * @param schemes Pass in one or more url schemes to consider valid, passing in
195      *        a null will default to "http,https,ftp" being valid.
196      *        If a non-null schemes is specified then all valid schemes must
197      *        be specified. Setting the ALLOW_ALL_SCHEMES option will
198      *        ignore the contents of schemes.
199      */
200     public UrlValidator(String[] schemes) {
201         this(schemes, 0);
202     }
203 
204     /***
205      * Initialize a UrlValidator with the given validation options.
206      * @param options The options should be set using the public constants declared in
207      * this class.  To set multiple options you simply add them together.  For example,
208      * ALLOW_2_SLASHES + NO_FRAGMENTS enables both of those options.
209      */
210     public UrlValidator(int options) {
211         this(null, options);
212     }
213 
214     /***
215      * Behavour of validation is modified by passing in options:
216      * @param schemes The set of valid schemes.
217      * @param options The options should be set using the public constants declared in
218      * this class.  To set multiple options you simply add them together.  For example,
219      * ALLOW_2_SLASHES + NO_FRAGMENTS enables both of those options.
220      */
221     public UrlValidator(String[] schemes, int options) {
222         this.options = new Flags(options);
223 
224         if (this.options.isOn(ALLOW_ALL_SCHEMES)) {
225             return;
226         }
227 
228         if (schemes == null) {
229             schemes = this.defaultSchemes;
230         }
231 
232         this.allowedSchemes.addAll(Arrays.asList(schemes));
233     }
234 
235     /***
236      * <p>Checks if a field has a valid url address.</p>
237      *
238      * @param value The value validation is being performed on.  A <code>null</code>
239      * value is considered invalid.
240      * @return true if the url is valid.
241      */
242     public boolean isValid(String value) {
243         if (value == null) {
244             return false;
245         }
246 
247         Perl5Util matchUrlPat = new Perl5Util();
248         Perl5Util matchAsciiPat = new Perl5Util();
249 
250         if (!matchAsciiPat.match(LEGAL_ASCII_PATTERN, value)) {
251             return false;
252         }
253 
254         // Check the whole url address structure
255         if (!matchUrlPat.match(URL_PATTERN, value)) {
256             return false;
257         }
258 
259         if (!isValidScheme(matchUrlPat.group(PARSE_URL_SCHEME))) {
260             return false;
261         }
262 
263         if (!isValidAuthority(matchUrlPat.group(PARSE_URL_AUTHORITY))) {
264             return false;
265         }
266 
267         if (!isValidPath(matchUrlPat.group(PARSE_URL_PATH))) {
268             return false;
269         }
270 
271         if (!isValidQuery(matchUrlPat.group(PARSE_URL_QUERY))) {
272             return false;
273         }
274 
275         if (!isValidFragment(matchUrlPat.group(PARSE_URL_FRAGMENT))) {
276             return false;
277         }
278 
279         return true;
280     }
281 
282     /***
283      * Validate scheme. If schemes[] was initialized to a non null,
284      * then only those scheme's are allowed.  Note this is slightly different
285      * than for the constructor.
286      * @param scheme The scheme to validate.  A <code>null</code> value is considered
287      * invalid.
288      * @return true if valid.
289      */
290     protected boolean isValidScheme(String scheme) {
291         if (scheme == null) {
292             return false;
293         }
294 
295         Perl5Util schemeMatcher = new Perl5Util();
296         if (!schemeMatcher.match(SCHEME_PATTERN, scheme)) {
297             return false;
298         }
299 
300         if (this.options.isOff(ALLOW_ALL_SCHEMES)) {
301 
302             if (!this.allowedSchemes.contains(scheme)) {
303                 return false;
304             }
305         }
306 
307         return true;
308     }
309 
310     /***
311      * Returns true if the authority is properly formatted.  An authority is the combination
312      * of hostname and port.  A <code>null</code> authority value is considered invalid.
313      */
314     protected boolean isValidAuthority(String authority) {
315         if (authority == null) {
316             return false;
317         }
318 
319         Perl5Util authorityMatcher = new Perl5Util();
320         Perl5Util matchIPV4Pat = new Perl5Util();
321 
322         if (!authorityMatcher.match(AUTHORITY_PATTERN, authority)) {
323             return false;
324         }
325 
326         boolean ipV4Address = false;
327         boolean hostname = false;
328 // check if authority is IP address or hostname
329         String hostIP = authorityMatcher.group(PARSE_AUTHORITY_HOST_IP);
330         ipV4Address = matchIPV4Pat.match(IP_V4_DOMAIN_PATTERN, hostIP);
331 
332         if (ipV4Address) {
333             // this is an IP address so check components
334             for (int i = 1; i <= 4; i++) {
335                 String ipSegment = matchIPV4Pat.group(i);
336                 if (ipSegment == null || ipSegment.length() <= 0) {
337                     return false;
338                 }
339 
340                 try {
341                     if (Integer.parseInt(ipSegment) > 255) {
342                         return false;
343                     }
344                 } catch(NumberFormatException e) {
345                     return false;
346                 }
347 
348             }
349         } else {
350             // Domain is hostname name
351             Perl5Util domainMatcher = new Perl5Util();
352             hostname = domainMatcher.match(DOMAIN_PATTERN, hostIP);
353         }
354 
355 //rightmost hostname will never start with a digit.
356         if (hostname) {
357             String[] domainSegment = new String[10];
358             boolean match = true;
359             int segmentCount = 0;
360             int segmentLength = 0;
361             Perl5Util atomMatcher = new Perl5Util();
362 
363             while (match) {
364                 match = atomMatcher.match(ATOM_PATTERN, hostIP);
365                 if (match) {
366                     domainSegment[segmentCount] = atomMatcher.group(1);
367                     segmentLength = domainSegment[segmentCount].length() + 1;
368                     hostIP =
369                             (segmentLength >= hostIP.length())
370                             ? ""
371                             : hostIP.substring(segmentLength);
372 
373                     segmentCount++;
374                 }
375             }
376             String topLevel = domainSegment[segmentCount - 1];
377             if (topLevel.length() < 2 || topLevel.length() > 4) {
378                 return false;
379             }
380 
381             // First letter of top level must be a alpha
382             Perl5Util alphaMatcher = new Perl5Util();
383             if (!alphaMatcher.match(ALPHA_PATTERN, topLevel.substring(0, 1))) {
384                 return false;
385             }
386 
387             // Make sure there's a host name preceding the authority.
388             if (segmentCount < 2) {
389                 return false;
390             }
391         }
392 
393         if (!hostname && !ipV4Address) {
394             return false;
395         }
396 
397         String port = authorityMatcher.group(PARSE_AUTHORITY_PORT);
398         if (port != null) {
399             Perl5Util portMatcher = new Perl5Util();
400             if (!portMatcher.match(PORT_PATTERN, port)) {
401                 return false;
402             }
403         }
404 
405         String extra = authorityMatcher.group(PARSE_AUTHORITY_EXTRA);
406         if (!GenericValidator.isBlankOrNull(extra)) {
407             return false;
408         }
409 
410         return true;
411     }
412 
413     /***
414      * Returns true if the path is valid.  A <code>null</code> value is considered invalid.
415      */
416     protected boolean isValidPath(String path) {
417         if (path == null) {
418             return false;
419         }
420 
421         Perl5Util pathMatcher = new Perl5Util();
422 
423         if (!pathMatcher.match(PATH_PATTERN, path)) {
424             return false;
425         }
426 
427         if (path.endsWith("/")) {
428             return false;
429         }
430 
431         int slash2Count = countToken("//", path);
432         if (this.options.isOff(ALLOW_2_SLASHES) && (slash2Count > 0)) {
433             return false;
434         }
435 
436         int slashCount = countToken("/", path);
437         int dot2Count = countToken("..", path);
438         if (dot2Count > 0) {
439             if ((slashCount - slash2Count - 1) <= dot2Count) {
440                 return false;
441             }
442         }
443 
444         return true;
445     }
446 
447     /***
448      * Returns true if the query is null or it's a properly formatted query string.
449      */
450     protected boolean isValidQuery(String query) {
451         if (query == null) {
452             return true;
453         }
454 
455         Perl5Util queryMatcher = new Perl5Util();
456         return queryMatcher.match(QUERY_PATTERN, query);
457     }
458 
459     /***
460      * Returns true if the given fragment is null or fragments are allowed.
461      */
462     protected boolean isValidFragment(String fragment) {
463         if (fragment == null) {
464             return true;
465         }
466 
467         return this.options.isOff(NO_FRAGMENTS);
468     }
469 
470     /***
471      * Returns the number of times the token appears in the target.
472      */
473     protected int countToken(String token, String target) {
474         int tokenIndex = 0;
475         int count = 0;
476         while (tokenIndex != -1) {
477             tokenIndex = target.indexOf(token, tokenIndex);
478             if (tokenIndex > -1) {
479                 tokenIndex++;
480                 count++;
481             }
482         }
483         return count;
484     }
485 }