1 package org.apache.maven.shared.utils.xml;
2
3 /*
4 * Licensed to the Apache Software Foundation (ASF) under one
5 * or more contributor license agreements. See the NOTICE file
6 * distributed with this work for additional information
7 * regarding copyright ownership. The ASF licenses this file
8 * to you under the Apache License, Version 2.0 (the
9 * "License"); you may not use this file except in compliance
10 * with the License. You may obtain a copy of the License at
11 *
12 * http://www.apache.org/licenses/LICENSE-2.0
13 *
14 * Unless required by applicable law or agreed to in writing,
15 * software distributed under the License is distributed on an
16 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
17 * KIND, either express or implied. See the License for the
18 * specific language governing permissions and limitations
19 * under the License.
20 */
21
22 /**
23 * Collection of XML encoding/decoding helpers. <br>
24 * This is all about the special characters & and <, and for attributes
25 * " and '. These must be encoded/decoded from/to XML.
26 */
27 final class XMLEncode
28 {
29
30 private static final int CDATA_BLOCK_THRESHOLD_LENGTH = 12;
31 private static final char DEFAULT_QUOTE_CHAR = '"';
32
33 /**
34 * Checks if this text purely consists of the white space characters
35 * ' ', TAB, NEWLINE.
36 */
37 public static boolean isWhiteSpace( String text )
38 {
39 for ( int i = 0; i < text.length(); i++ )
40 {
41 char c = text.charAt( i );
42 if ( !Character.isWhitespace( c ) )
43 {
44 return false;
45 }
46 }
47 return true;
48 }
49
50 /**
51 * Makes any text fit into XML attributes.
52 */
53 public static String xmlEncodeTextForAttribute( String text, char quoteChar )
54 {
55 if ( text == null )
56 {
57 return null;
58 }
59 return xmlEncodeTextAsPCDATA( text, true, quoteChar );
60 }
61
62 /**
63 * Encodes text as XML in the most suitable way, either CDATA block or PCDATA.
64 */
65 public static String xmlEncodeText( String text )
66 {
67 if ( text == null )
68 {
69 return null;
70 }
71 if ( !needsEncoding( text ) )
72 {
73 return text;
74 }
75 else
76 {
77 // only encode as cdata if is is longer than CDATA block overhead:
78 if ( text.length() > CDATA_BLOCK_THRESHOLD_LENGTH )
79 {
80 String cdata = xmlEncodeTextAsCDATABlock( text );
81 if ( cdata != null )
82 {
83 return cdata;
84 }
85 }
86 }
87 // if every thing else fails, do it the save way...
88 return xmlEncodeTextAsPCDATA( text );
89 }
90
91 /**
92 * Encodes any text as PCDATA.
93 */
94 public static String xmlEncodeTextAsPCDATA( String text )
95 {
96 if ( text == null )
97 {
98 return null;
99 }
100 return xmlEncodeTextAsPCDATA( text, false );
101 }
102
103 /**
104 * Encodes any text as PCDATA.
105 *
106 * @param forAttribute if you want
107 * quotes and apostrophes specially treated for attributes
108 */
109 public static String xmlEncodeTextAsPCDATA( String text, boolean forAttribute )
110 {
111 return xmlEncodeTextAsPCDATA( text, forAttribute, DEFAULT_QUOTE_CHAR );
112 }
113
114 /**
115 * Encodes any text as PCDATA.
116 *
117 * @param forAttribute if you want
118 * quotes and apostrophes specially treated for attributes
119 * @param quoteChar if this is for attributes this <code>char</code> is used to quote the attribute value
120 */
121 public static String xmlEncodeTextAsPCDATA( String text, boolean forAttribute, char quoteChar )
122 {
123 if ( text == null )
124 {
125 return null;
126 }
127 char c;
128 int length = text.length();
129 StringBuilder n = new StringBuilder( length * 2 );
130 if ( forAttribute )
131 {
132 n.append( quoteChar );
133 }
134
135 for ( int i = 0; i < length; i++ )
136 {
137 c = text.charAt( i );
138 switch ( c )
139 {
140 case '&':
141 n.append( "&" );
142 break;
143 case '<':
144 n.append( "<" );
145 break;
146 case '>': // FIX for sourceforge bug #802520 ("]]>" needs encoding)
147 n.append( ">" );
148 break;
149 case '"':
150 if ( forAttribute )
151 {
152 n.append( """ );
153 }
154 else
155 {
156 n.append( c );
157 }
158 break;
159 case '\'':
160 if ( forAttribute )
161 {
162 n.append( "'" );
163 }
164 else
165 {
166 n.append( c );
167 }
168 break;
169 case '\r':
170 if ( forAttribute )
171 {
172 if ( i == ( length - 1 ) || text.charAt( i + 1 ) != '\n' )
173 {
174 n.append( " " );
175 }
176 }
177 else
178 {
179 n.append( c );
180 }
181 // but skip the \r in \r\n
182
183
184 break;
185 case '\n':
186 if ( forAttribute )
187 {
188 n.append( " " );
189 }
190 break;
191
192 default:
193 n.append( c );
194 break;
195 }
196 }
197
198 if ( forAttribute )
199 {
200 n.append( quoteChar );
201 }
202
203 return n.toString();
204 }
205
206 /**
207 * Returns string as CDATA block if possible, otherwise null.
208 */
209 public static String xmlEncodeTextAsCDATABlock( String text )
210 {
211 if ( text == null )
212 {
213 return null;
214 }
215 if ( isCompatibleWithCDATABlock( text ) )
216 {
217 return "<![CDATA[" + text + "]]>";
218 }
219 else
220 {
221 return null;
222 }
223 }
224
225 /**
226 * Checks if this text needs encoding in order to be represented in XML.
227 */
228 public static boolean needsEncoding( String text )
229 {
230 return needsEncoding( text, false );
231 }
232
233 /**
234 * Checks if this text needs encoding in order to be represented in XML.
235 * <p/>
236 * Set <code>checkForAttr</code> if you want to check for storability in
237 * an attribute.
238 */
239 public static boolean needsEncoding( String data, boolean checkForAttr )
240 {
241 if ( data == null )
242 {
243 return false;
244 }
245 char c;
246 for ( int i = 0; i < data.length(); i++ )
247 {
248 c = data.charAt( i );
249 if ( c == '&' || c == '<' || ( checkForAttr && ( c == '"' || c == '\'' ) ) )
250 {
251 return true;
252 }
253 }
254 return false;
255 }
256
257 /**
258 * Can this text be stored into a CDATA block?
259 */
260 public static boolean isCompatibleWithCDATABlock( String text )
261 {
262 return text != null && ( !text.contains( "]]>" ) );
263 }
264
265 /**
266 * Make CDATA out of possibly encoded PCDATA. <br>
267 * E.g. make '&' out of '&amp;'
268 */
269 public static String xmlDecodeTextToCDATA( String pcdata )
270 {
271 if ( pcdata == null )
272 {
273 return null;
274 }
275 char c, c1, c2, c3, c4, c5;
276 StringBuilder n = new StringBuilder( pcdata.length() );
277 for ( int i = 0; i < pcdata.length(); i++ )
278 {
279 c = pcdata.charAt( i );
280 if ( c == '&' )
281 {
282 c1 = lookAhead( 1, i, pcdata );
283 c2 = lookAhead( 2, i, pcdata );
284 c3 = lookAhead( 3, i, pcdata );
285 c4 = lookAhead( 4, i, pcdata );
286 c5 = lookAhead( 5, i, pcdata );
287
288 if ( c1 == 'a' && c2 == 'm' && c3 == 'p' && c4 == ';' )
289 {
290 n.append( "&" );
291 i += 4;
292 }
293 else if ( c1 == 'l' && c2 == 't' && c3 == ';' )
294 {
295 n.append( "<" );
296 i += 3;
297 }
298 else if ( c1 == 'g' && c2 == 't' && c3 == ';' )
299 {
300 n.append( ">" );
301 i += 3;
302 }
303 else if ( c1 == 'q' && c2 == 'u' && c3 == 'o' && c4 == 't' && c5 == ';' )
304 {
305 n.append( "\"" );
306 i += 5;
307 }
308 else if ( c1 == 'a' && c2 == 'p' && c3 == 'o' && c4 == 's' && c5 == ';' )
309 {
310 n.append( "'" );
311 i += 5;
312 }
313 else
314 {
315 n.append( "&" );
316 }
317 }
318 else
319 {
320 n.append( c );
321 }
322 }
323 return n.toString();
324 }
325
326 private static char lookAhead( int la, int offset, String data )
327 {
328 try
329 {
330 return data.charAt( offset + la );
331 }
332 catch ( StringIndexOutOfBoundsException e )
333 {
334 return 0x0;
335 }
336 }
337
338 // combine multiple checks in one methods for speed
339 private static boolean contains( String text, char[] chars )
340 {
341 if ( text == null || chars == null || chars.length == 0 )
342 {
343 return false;
344 }
345 for ( int i = 0; i < text.length(); i++ )
346 {
347 char c = text.charAt( i );
348 for ( char aChar : chars )
349 {
350 if ( aChar == c )
351 {
352 return true;
353 }
354 }
355 }
356 return false;
357 }
358
359 }