1 /* 2 * Copyright (C) 2012 The Android Open Source Project 3 * 4 * Licensed under the Apache License, Version 2.0 (the "License"); 5 * you may not use this file except in compliance with the License. 6 * You may obtain a copy of the License at 7 * 8 * http://www.apache.org/licenses/LICENSE-2.0 9 * 10 * Unless required by applicable law or agreed to in writing, software 11 * distributed under the License is distributed on an "AS IS" BASIS, 12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13 * See the License for the specific language governing permissions and 14 * limitations under the License. 15 */ 16 17 import java.util.regex.Matcher; 18 import java.util.regex.Pattern; 19 20 public class Main { 21 // from android.util.Patterns 22 public static final String GOOD_IRI_CHAR = 23 "a-zA-Z0-9\u00A0-\uD7FF\uF900-\uFDCF\uFDF0-\uFFEF"; 24 25 public static final String TOP_LEVEL_DOMAIN_STR_FOR_WEB_URL = 26 "(?:" 27 + "(?:aero|arpa|asia|a[cdefgilmnoqrstuwxz])" 28 + "|(?:biz|b[abdefghijmnorstvwyz])" 29 + "|(?:cat|com|coop|c[acdfghiklmnoruvxyz])" 30 + "|d[ejkmoz]" 31 + "|(?:edu|e[cegrstu])" 32 + "|f[ijkmor]" 33 + "|(?:gov|g[abdefghilmnpqrstuwy])" 34 + "|h[kmnrtu]" 35 + "|(?:info|int|i[delmnoqrst])" 36 + "|(?:jobs|j[emop])" 37 + "|k[eghimnprwyz]" 38 + "|l[abcikrstuvy]" 39 + "|(?:mil|mobi|museum|m[acdeghklmnopqrstuvwxyz])" 40 + "|(?:name|net|n[acefgilopruz])" 41 + "|(?:org|om)" 42 + "|(?:pro|p[aefghklmnrstwy])" 43 + "|qa" 44 + "|r[eosuw]" 45 + "|s[abcdeghijklmnortuvyz]" 46 + "|(?:tel|travel|t[cdfghjklmnoprtvwz])" 47 + "|u[agksyz]" 48 + "|v[aceginu]" 49 + "|w[fs]" 50 + "|(?:\u03b4\u03bf\u03ba\u03b9\u03bc\u03ae|\u0438\u0441\u043f\u044b\u0442\u0430\u043d\u0438\u0435|\u0440\u0444|\u0441\u0440\u0431|\u05d8\u05e2\u05e1\u05d8|\u0622\u0632\u0645\u0627\u06cc\u0634\u06cc|\u0625\u062e\u062a\u0628\u0627\u0631|\u0627\u0644\u0627\u0631\u062f\u0646|\u0627\u0644\u062c\u0632\u0627\u0626\u0631|\u0627\u0644\u0633\u0639\u0648\u062f\u064a\u0629|\u0627\u0644\u0645\u063a\u0631\u0628|\u0627\u0645\u0627\u0631\u0627\u062a|\u0628\u06be\u0627\u0631\u062a|\u062a\u0648\u0646\u0633|\u0633\u0648\u0631\u064a\u0629|\u0641\u0644\u0633\u0637\u064a\u0646|\u0642\u0637\u0631|\u0645\u0635\u0631|\u092a\u0930\u0940\u0915\u094d\u0937\u093e|\u092d\u093e\u0930\u0924|\u09ad\u09be\u09b0\u09a4|\u0a2d\u0a3e\u0a30\u0a24|\u0aad\u0abe\u0ab0\u0aa4|\u0b87\u0ba8\u0bcd\u0ba4\u0bbf\u0baf\u0bbe|\u0b87\u0bb2\u0b99\u0bcd\u0b95\u0bc8|\u0b9a\u0bbf\u0b99\u0bcd\u0b95\u0baa\u0bcd\u0baa\u0bc2\u0bb0\u0bcd|\u0baa\u0bb0\u0bbf\u0b9f\u0bcd\u0b9a\u0bc8|\u0c2d\u0c3e\u0c30\u0c24\u0c4d|\u0dbd\u0d82\u0d9a\u0dcf|\u0e44\u0e17\u0e22|\u30c6\u30b9\u30c8|\u4e2d\u56fd|\u4e2d\u570b|\u53f0\u6e7e|\u53f0\u7063|\u65b0\u52a0\u5761|\u6d4b\u8bd5|\u6e2c\u8a66|\u9999\u6e2f|\ud14c\uc2a4\ud2b8|\ud55c\uad6d|xn\\-\\-0zwm56d|xn\\-\\-11b5bs3a9aj6g|xn\\-\\-3e0b707e|xn\\-\\-45brj9c|xn\\-\\-80akhbyknj4f|xn\\-\\-90a3ac|xn\\-\\-9t4b11yi5a|xn\\-\\-clchc0ea0b2g2a9gcd|xn\\-\\-deba0ad|xn\\-\\-fiqs8s|xn\\-\\-fiqz9s|xn\\-\\-fpcrj9c3d|xn\\-\\-fzc2c9e2c|xn\\-\\-g6w251d|xn\\-\\-gecrj9c|xn\\-\\-h2brj9c|xn\\-\\-hgbk6aj7f53bba|xn\\-\\-hlcj6aya9esc7a|xn\\-\\-j6w193g|xn\\-\\-jxalpdlp|xn\\-\\-kgbechtv|xn\\-\\-kprw13d|xn\\-\\-kpry57d|xn\\-\\-lgbbat1ad8j|xn\\-\\-mgbaam7a8h|xn\\-\\-mgbayh7gpa|xn\\-\\-mgbbh1a71e|xn\\-\\-mgbc0a9azcg|xn\\-\\-mgberp4a5d4ar|xn\\-\\-o3cw4h|xn\\-\\-ogbpf8fl|xn\\-\\-p1ai|xn\\-\\-pgbs0dh|xn\\-\\-s9brj9c|xn\\-\\-wgbh1c|xn\\-\\-wgbl6a|xn\\-\\-xkc2al3hye2a|xn\\-\\-xkc2dl3a5ee0h|xn\\-\\-yfro4i67o|xn\\-\\-ygbi2ammx|xn\\-\\-zckzah|xxx)" 51 + "|y[et]" 52 + "|z[amw]))"; 53 54 public static final String WEB_URL_STR = 55 "((?:(http|https|Http|Https|rtsp|Rtsp):\\/\\/(?:(?:[a-zA-Z0-9\\$\\-\\_\\.\\+\\!\\*\\'\\(\\)" 56 + "\\,\\;\\?\\&\\=]|(?:\\%[a-fA-F0-9]{2})){1,64}(?:\\:(?:[a-zA-Z0-9\\$\\-\\_" 57 + "\\.\\+\\!\\*\\'\\(\\)\\,\\;\\?\\&\\=]|(?:\\%[a-fA-F0-9]{2})){1,25})?\\@)?)?" 58 + "((?:(?:[" + GOOD_IRI_CHAR + "][" + GOOD_IRI_CHAR + "\\-]{0,64}\\.)+" // named host 59 + TOP_LEVEL_DOMAIN_STR_FOR_WEB_URL 60 + "|(?:(?:25[0-5]|2[0-4]" // or ip address 61 + "[0-9]|[0-1][0-9]{2}|[1-9][0-9]|[1-9])\\.(?:25[0-5]|2[0-4][0-9]" 62 + "|[0-1][0-9]{2}|[1-9][0-9]|[1-9]|0)\\.(?:25[0-5]|2[0-4][0-9]|[0-1]" 63 + "[0-9]{2}|[1-9][0-9]|[1-9]|0)\\.(?:25[0-5]|2[0-4][0-9]|[0-1][0-9]{2}" 64 + "|[1-9][0-9]|[0-9])))" 65 + "(?:\\:\\d{1,5})?)" // plus option port number 66 + "(\\/(?:(?:[" + GOOD_IRI_CHAR + "\\;\\/\\?\\:\\@\\&\\=\\#\\~" // plus option query params 67 + "\\-\\.\\+\\!\\*\\'\\(\\)\\,\\_])|(?:\\%[a-fA-F0-9]{2}))*)?" 68 + "(?:\\b|$)"; // and finally, a word boundary or end of 69 // input. This is to stop foo.sure from 70 // matching as foo.su 71 72 public static final Pattern WEB_URL = Pattern.compile(WEB_URL_STR); 73 74 public static final String testStr1 = 75 "http://www.google.com/blah?client=tablet-android&source=android-home"; 76 public static final String testStr2 = "http:///www.google.com/"; 77 public static final String testStr3 = 78 "http://www.google.com/search?hl=en&redir_esc=&client=tablet-android-verizon&source=android-browser-type&v=141000000&qsubts=1327020479959&action=devloc&q=cnn"; 79 80 public static void main(String[] args) { 81 System.out.println("str1 matches: " + WEB_URL.matcher(testStr1).matches()); 82 System.out.println("str2 matches: " + WEB_URL.matcher(testStr2).matches()); 83 System.out.println("str3 matches: " + WEB_URL.matcher(testStr3).matches()); 84 } 85 86 static String getStringAsHex(String text) { 87 StringBuilder sb = new StringBuilder(text.length() * 4); 88 89 for (int i = 0; i < text.length(); i++) { 90 sb.append(Integer.toHexString((int) text.charAt(i))); 91 } 92 93 return sb.toString(); 94 } 95 } 96
