001 /*
002 Copyright (c) 2009 Olivier Chafik, All Rights Reserved
003
004 This file is part of JNAerator (http://jnaerator.googlecode.com/).
005
006 JNAerator is free software: you can redistribute it and/or modify
007 it under the terms of the GNU General Public License as published by
008 the Free Software Foundation, either version 3 of the License, or
009 (at your option) any later version.
010
011 JNAerator is distributed in the hope that it will be useful,
012 but WITHOUT ANY WARRANTY; without even the implied warranty of
013 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
014 GNU General Public License for more details.
015
016 You should have received a copy of the GNU General Public License
017 along with JNAerator. If not, see <http://www.gnu.org/licenses/>.
018 */
019 package com.ochafik.util.string;
020
021 import java.io.IOException;
022 import java.text.MessageFormat;
023 import java.util.ArrayList;
024 import java.util.Collection;
025 import java.util.LinkedList;
026 import java.util.List;
027 import java.util.regex.Matcher;
028 import java.util.regex.Pattern;
029
030 import com.ochafik.util.listenable.Adapter;
031 import com.ochafik.util.listenable.Pair;
032
033 public class RegexUtils {
034 static final String[][] simplePatternReplacements=new String[][] {
035 {"(\\.|\\+|\\[|\\]|\\{|\\})","\\\\."},
036 {"(\\*|\\?)",".$1"},
037 /*{"[a\u00e0\u00e2\u00e4]","[a\u00e0\u00e2\u00e4]"},
038 {"[e\u00e9\u00e8\u00ea\u00eb]","[e\u00e9\u00e8\u00ea\u00eb]"},
039 {"[i\u00ee\u00ef\u00ec]","[i\u00ee\u00ef\u00ec]"},
040 {"[o\u00f4\u00f6\u00f2]","[o\u00f4\u00f6\u00f2]"},
041 {"[u\u00fb\u00fc\u00f9]","[u\u00fb\u00fc\u00f9]"},
042 {"[y\u00ff]","[y\u00ff]"},
043 {"[c\u00e7]","[c\u00e7]"},*/
044 {",","|"}
045 };
046 static final Pattern[] simplePatternReplacementPatterns;
047 static {
048 simplePatternReplacementPatterns=new Pattern[simplePatternReplacements.length];
049 for (int i=simplePatternReplacements.length;i--!=0;) {
050 simplePatternReplacementPatterns[i]=Pattern.compile(simplePatternReplacements[i][0]);
051 }
052 }
053 private static final String replaceSimplePatterByRegex(String string) {
054 for (int i=simplePatternReplacements.length;i--!=0;) {
055 string=simplePatternReplacementPatterns[i].matcher(string).replaceAll(simplePatternReplacements[i][1]);
056 }
057 return string;
058 }
059 public static final Pattern simplePatternToRegex(String pattern) {
060 try {
061 return Pattern.compile(replaceSimplePatterByRegex(pattern),Pattern.CASE_INSENSITIVE|Pattern.UNICODE_CASE);
062 } catch (Exception ex) {
063 return Pattern.compile(pattern,Pattern.CASE_INSENSITIVE|Pattern.UNICODE_CASE|Pattern.LITERAL);
064 }
065 }
066 public static String regexReplace(Pattern pattern, String string, MessageFormat replacement) {
067 StringBuffer b=new StringBuffer(string.length());
068 int iLastCommitted=0;
069 Matcher matcher=pattern.matcher(string);
070 while (matcher.find(iLastCommitted)) {
071 int start=matcher.start();
072 String s = string.substring(iLastCommitted,start);
073 String g = matcher.group(0);
074 b.append(s);
075 if (replacement!=null)
076 b.append(replacement.format(getGroups(matcher),b,null));
077 else
078 b.append(g);
079
080 iLastCommitted = start + g.length();
081 }
082 b.append(string.substring(iLastCommitted));
083 return b.toString();
084 }
085 public static final boolean findLast(Matcher matcher) {
086 matcher.reset();
087 int n = 0;
088 while (matcher.find()) {
089 n++;
090 }
091 matcher.reset();
092 for (int i = 0; i < n; i++) matcher.find();
093 return n > 0;
094 }
095
096 public static final List<Pair<String, String[]>> grep(Collection<String> list, Pattern p) {
097 List<Pair<String, String[]>> ret = new ArrayList<Pair<String, String[]>>(list.size());
098 for (String s : list) {
099 Matcher matcher = p.matcher(s);
100 if (matcher.find())
101 ret.add(new Pair<String, String[]>(s, getGroups(matcher)));
102 }
103 return ret;
104 }
105
106 public static String[] getGroups(Matcher matcher) {
107 String[] groups = new String[matcher.groupCount() + 1];
108 for (int i = groups.length; i-- != 0;) {
109 groups[i] = matcher.group(i);
110 }
111 return groups;
112 }
113 public static final List<String[]> grep(String s, Pattern p) {
114 List<String[]> ret = new ArrayList<String[]>();
115 Matcher matcher = p.matcher(s);
116 while (matcher.find())
117 ret.add(getGroups(matcher));
118
119 return ret;
120 }
121 public static final String[] match(String s, Pattern p) {
122 Matcher matcher = p.matcher(s);
123 if (matcher.matches())
124 return getGroups(matcher);
125
126 return null;
127 }
128 public static List<String[]> find(String string, String pattern) {
129 return find(string, Pattern.compile(pattern));
130 }
131
132 public static List<String[]> find(String string, Pattern pattern) {
133 List<String[]> ret = new LinkedList<String[]>();
134
135 Matcher matcher = pattern.matcher(string);
136 while (matcher.find())
137 ret.add(getGroups(matcher));
138
139 return ret;
140 }
141 public static Collection<String> find(String string, String pattern, final int iGroup) {
142 return find(string, Pattern.compile(pattern), iGroup);
143 }
144 public static Collection<String> find(String string, Pattern pattern, final int iGroup) {
145 List<String> ret= new ArrayList<String>();
146 Matcher matcher = pattern.matcher(string);
147 while (matcher.find()) {
148 ret.add(matcher.group(iGroup));
149 }
150 return ret;
151 }
152 public static String findFirst(String string, Pattern pattern, int iGroup) {
153 if (string == null)
154 return null;
155
156 Matcher matcher = pattern.matcher(string);
157 return matcher.find() ? matcher.group(iGroup) : null;
158 }
159 public static String regexReplace(String pat, String text, String rep) {
160 return regexReplace(Pattern.compile(pat), text, new MessageFormat(rep));
161 }
162 public static String regexReplace(Pattern pattern, String string, Adapter<String[], String> adapter) {
163 StringBuffer b=new StringBuffer(string.length());
164 int iLastCommitted=0;
165 Matcher matcher=pattern.matcher(string);
166 while (matcher.find(iLastCommitted)) {
167 int start=matcher.start();
168 String s = string.substring(iLastCommitted,start);
169 String g = matcher.group(0);
170 b.append(s);
171 if (adapter!=null)
172 b.append(adapter.adapt(getGroups(matcher)));
173 else
174 b.append(g);
175 iLastCommitted = start + g.length();
176 }
177 b.append(string.substring(iLastCommitted));
178 return b.toString();
179 }
180
181 }