1 /*
2  * $Header: /home/jerenkrantz/tmp/commons/commons-convert/cvs/home/cvs/jakarta-commons//httpclient/src/java/org/apache/commons/httpclient/util/ParameterParser.java,v 1.5 2004/05/13 04:01:22 mbecke Exp $
3  * $Revision: 480424 $
4  * $Date: 2006-11-29 06:56:49 +0100 (Wed, 29 Nov 2006) $
5  *
6  * ====================================================================
7  *
8  *  Licensed to the Apache Software Foundation (ASF) under one or more
9  *  contributor license agreements.  See the NOTICE file distributed with
10  *  this work for additional information regarding copyright ownership.
11  *  The ASF licenses this file to You under the Apache License, Version 2.0
12  *  (the "License"); you may not use this file except in compliance with
13  *  the License.  You may obtain a copy of the License at
14  *
15  *      http://www.apache.org/licenses/LICENSE-2.0
16  *
17  *  Unless required by applicable law or agreed to in writing, software
18  *  distributed under the License is distributed on an "AS IS" BASIS,
19  *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
20  *  See the License for the specific language governing permissions and
21  *  limitations under the License.
22  * ====================================================================
23  *
24  * This software consists of voluntary contributions made by many
25  * individuals on behalf of the Apache Software Foundation.  For more
26  * information on the Apache Software Foundation, please see
27  * <http://www.apache.org/>.
28  *
29  */
30 
31 package org.apache.commons.httpclient.util;
32 import java.util.ArrayList;
33 import java.util.List;
34 
35 import org.apache.commons.httpclient.NameValuePair;
36 
37 /**
38  * A simple parser intended to parse sequences of name/value pairs.
39  * Parameter values are exptected to be enclosed in quotes if they
40  * contain unsafe characters, such as '=' characters or separators.
41  * Parameter values are optional and can be omitted.
42  *
43  * <p>
44  *  <code>param1 = value; param2 = "anything goes; really"; param3</code>
45  * </p>
46  *
47  * @author <a href="mailto:oleg@ural.ru">Oleg Kalnichevski</a>
48  *
49  * @since 3.0
50  */
51 public class ParameterParser {
52 
53     /** String to be parsed */
54     private char[] chars = null;
55 
56     /** Current position in the string */
57     private int pos = 0;
58 
59     /** Maximum position in the string */
60     private int len = 0;
61 
62     /** Start of a token */
63     private int i1 = 0;
64 
65     /** End of a token */
66     private int i2 = 0;
67 
68     /** Default ParameterParser constructor */
ParameterParser()69     public ParameterParser() {
70         super();
71     }
72 
73 
74     /** Are there any characters left to parse? */
hasChar()75     private boolean hasChar() {
76         return this.pos < this.len;
77     }
78 
79 
80     /** A helper method to process the parsed token. */
getToken(boolean quoted)81     private String getToken(boolean quoted) {
82         // Trim leading white spaces
83         while ((i1 < i2) && (Character.isWhitespace(chars[i1]))) {
84             i1++;
85         }
86         // Trim trailing white spaces
87         while ((i2 > i1) && (Character.isWhitespace(chars[i2 - 1]))) {
88             i2--;
89         }
90         // Strip away quotes if necessary
91         if (quoted) {
92             if (((i2 - i1) >= 2)
93                 && (chars[i1] == '"')
94                 && (chars[i2 - 1] == '"')) {
95                 i1++;
96                 i2--;
97             }
98         }
99         String result = null;
100         if (i2 >= i1) {
101             result = new String(chars, i1, i2 - i1);
102         }
103         return result;
104     }
105 
106 
107     /** Is given character present in the array of characters? */
isOneOf(char ch, char[] charray)108     private boolean isOneOf(char ch, char[] charray) {
109         boolean result = false;
110         for (int i = 0; i < charray.length; i++) {
111             if (ch == charray[i]) {
112                 result = true;
113                 break;
114             }
115         }
116         return result;
117     }
118 
119 
120     /** Parse out a token until any of the given terminators
121      * is encountered. */
parseToken(final char[] terminators)122     private String parseToken(final char[] terminators) {
123         char ch;
124         i1 = pos;
125         i2 = pos;
126         while (hasChar()) {
127             ch = chars[pos];
128             if (isOneOf(ch, terminators)) {
129                 break;
130             }
131             i2++;
132             pos++;
133         }
134         return getToken(false);
135     }
136 
137 
138     /** Parse out a token until any of the given terminators
139      * is encountered. Special characters in quoted tokens
140      * are escaped. */
parseQuotedToken(final char[] terminators)141     private String parseQuotedToken(final char[] terminators) {
142         char ch;
143         i1 = pos;
144         i2 = pos;
145         boolean quoted = false;
146         boolean charEscaped = false;
147         while (hasChar()) {
148             ch = chars[pos];
149             if (!quoted && isOneOf(ch, terminators)) {
150                 break;
151             }
152             if (!charEscaped && ch == '"') {
153                 quoted = !quoted;
154             }
155             charEscaped = (!charEscaped && ch == '\\');
156             i2++;
157             pos++;
158 
159         }
160         return getToken(true);
161     }
162 
163     /**
164      * Extracts a list of {@link NameValuePair}s from the given string.
165      *
166      * @param str the string that contains a sequence of name/value pairs
167      * @return a list of {@link NameValuePair}s
168      *
169      */
parse(final String str, char separator)170     public List parse(final String str, char separator) {
171 
172         if (str == null) {
173             return new ArrayList();
174         }
175         return parse(str.toCharArray(), separator);
176     }
177 
178     /**
179      * Extracts a list of {@link NameValuePair}s from the given array of
180      * characters.
181      *
182      * @param chars the array of characters that contains a sequence of
183      * name/value pairs
184      *
185      * @return a list of {@link NameValuePair}s
186      */
parse(final char[] chars, char separator)187     public List parse(final char[] chars, char separator) {
188 
189         if (chars == null) {
190             return new ArrayList();
191         }
192         return parse(chars, 0, chars.length, separator);
193     }
194 
195 
196     /**
197      * Extracts a list of {@link NameValuePair}s from the given array of
198      * characters.
199      *
200      * @param chars the array of characters that contains a sequence of
201      * name/value pairs
202      * @param offset - the initial offset.
203      * @param length - the length.
204      *
205      * @return a list of {@link NameValuePair}s
206      */
parse(final char[] chars, int offset, int length, char separator)207     public List parse(final char[] chars, int offset, int length, char separator) {
208 
209         if (chars == null) {
210             return new ArrayList();
211         }
212         List params = new ArrayList();
213         this.chars = chars;
214         this.pos = offset;
215         this.len = length;
216 
217         String paramName = null;
218         String paramValue = null;
219         while (hasChar()) {
220             paramName = parseToken(new char[] {'=', separator});
221             paramValue = null;
222             if (hasChar() && (chars[pos] == '=')) {
223                 pos++; // skip '='
224                 paramValue = parseQuotedToken(new char[] {separator});
225             }
226             if (hasChar() && (chars[pos] == separator)) {
227                 pos++; // skip separator
228             }
229             if (paramName != null && !(paramName.equals("") && paramValue == null)) {
230                 params.add(new NameValuePair(paramName, paramValue));
231             }
232         }
233         return params;
234     }
235 }
236