1
/****************************************************************
2
* Licensed to the Apache Software Foundation (ASF) under one *
3
* or more contributor license agreements. See the NOTICE file *
4
* distributed with this work for additional information *
5
* regarding copyright ownership. The ASF licenses this file *
6
* to you under the Apache License, Version 2.0 (the *
7
* "License"); you may not use this file except in compliance *
8
* with the License. You may obtain a copy of the License at *
10
* http://www.apache.org/licenses/LICENSE-2.0 *
12
* Unless required by applicable law or agreed to in writing, *
13
* software distributed under the License is distributed on an *
14
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY *
15
* KIND, either express or implied. See the License for the *
16
* specific language governing permissions and limitations *
17
* under the License. *
18
****************************************************************/
20
package org.apache.james.mime4j.parser;
22
import java.io.ByteArrayInputStream;
23
import java.io.IOException;
24
import java.nio.charset.Charset;
26
import junit.framework.TestCase;
28
import org.apache.commons.io.IOUtils;
29
import org.apache.james.mime4j.MimeException;
30
import org.apache.james.mime4j.util.CharsetUtil;
32
public class MultipartTokensTest extends TestCase {
34
private static final Charset US_ASCII = CharsetUtil.US_ASCII;
36
private static final String BODY = "A Preamble\r\n" +
38
"Simple plain text\r\n" +
40
"Content-Type: text/plain; charset=US-ASCII\r\n\r\n" +
41
"Some more text\r\n" +
43
public static final String MESSAGE = "To: Road Runner <runner@example.org>\r\n" +
44
"From: Wile E. Cayote <wile@example.org>\r\n" +
45
"Date: Tue, 12 Feb 2008 17:34:09 +0000 (GMT)\r\n" +
47
"Content-Type: multipart/mixed;boundary=1729\r\n\r\n" +
50
public static final String COMPLEX_MESSAGE = "To: Wile E. Cayote <wile@example.org>\r\n" +
51
"From: Road Runner <runner@example.org>\r\n" +
52
"Date: Tue, 19 Feb 2008 17:34:09 +0000 (GMT)\r\n" +
54
"Content-Type: multipart/mixed;boundary=42\r\n\r\n" +
55
"A little preamble\r\n" +
57
"Content-Type: text/plain; charset=US-ASCII\r\n\r\n" +
60
"Content-Type: message/rfc822\r\n\r\n" +
69
public static final String COMPLEX_QP_MESSAGE =
70
"Content-Transfer-Encoding: quoted-printable\r\n" +
71
"Content-Type: message/rfc822; charset=us-ascii\r\n" +
73
"Subject: The subject\r\n" +
74
"Content-Type: multipart/alternative;\r\n" +
75
" boundary=3D=22----=3DNextPart=22\r\n" +
77
"This is a multi-part message in MIME format.\r\n" +
79
"------=3DNextPart\r\n" +
80
"Content-Type: text/plain;\r\n" +
81
" charset=3D=22iso-8859-1=22\r\n" +
85
"------=3DNextPart\r\n" +
86
"Content-Type: text/html;\r\n" +
87
" charset=3D=22iso-8859-1=22\r\n" +
89
"<HTML><BODY>=3D Some HTML =3D</BODY></HTML>\r\n" +
90
"------=3DNextPart--\r\n" +
94
MimeTokenStream parser;
97
protected void setUp() throws Exception {
99
parser = new MimeTokenStream();
103
protected void tearDown() throws Exception {
107
public void testShouldParseSimpleMessage() throws Exception {
108
parser.parse(new ByteArrayInputStream(US_ASCII.encode(MESSAGE).array()));
109
checkState(MimeTokenStream.T_START_HEADER);
110
checkState(MimeTokenStream.T_FIELD);
111
checkState(MimeTokenStream.T_FIELD);
112
checkState(MimeTokenStream.T_FIELD);
113
checkState(MimeTokenStream.T_FIELD);
114
checkState(MimeTokenStream.T_FIELD);
115
checkState(MimeTokenStream.T_END_HEADER);
116
checkState(MimeTokenStream.T_START_MULTIPART);
117
checkState(MimeTokenStream.T_PREAMBLE);
118
checkState(MimeTokenStream.T_START_BODYPART);
119
checkState(MimeTokenStream.T_START_HEADER);
120
checkState(MimeTokenStream.T_END_HEADER);
121
checkState(MimeTokenStream.T_BODY);
122
checkState(MimeTokenStream.T_END_BODYPART);
123
checkState(MimeTokenStream.T_START_BODYPART);
124
checkState(MimeTokenStream.T_START_HEADER);
125
checkState(MimeTokenStream.T_FIELD);
126
checkState(MimeTokenStream.T_END_HEADER);
127
checkState(MimeTokenStream.T_BODY);
128
checkState(MimeTokenStream.T_END_BODYPART);
129
checkState(MimeTokenStream.T_EPILOGUE);
130
checkState(MimeTokenStream.T_END_MULTIPART);
131
checkState(MimeTokenStream.T_END_MESSAGE);
132
checkState(MimeTokenStream.T_END_OF_STREAM);
135
public void testShouldParseMoreComplexMessage() throws Exception {
137
"Content-Type: multipart/alternative; boundary=\"outer-boundary\"\r\n" +
139
"--outer-boundary\r\n" +
140
"Content-Type: multipart/alternative; boundary=\"inner-boundary\"\r\n" +
142
"--inner-boundary\r\n" +
143
"Content-Type: text/plain\r\n" +
146
"--inner-boundary--\r\n" +
149
"--outer-boundary--\r\n";
151
parser.parse(new ByteArrayInputStream(US_ASCII.encode(message).array()));
152
checkState(MimeTokenStream.T_START_HEADER);
153
checkState(MimeTokenStream.T_FIELD);
154
checkState(MimeTokenStream.T_END_HEADER);
155
checkState(MimeTokenStream.T_START_MULTIPART);
156
checkState(MimeTokenStream.T_PREAMBLE);
157
checkState(MimeTokenStream.T_START_BODYPART);
158
checkState(MimeTokenStream.T_START_HEADER);
159
checkState(MimeTokenStream.T_FIELD);
160
checkState(MimeTokenStream.T_END_HEADER);
161
checkState(MimeTokenStream.T_START_MULTIPART);
162
checkState(MimeTokenStream.T_PREAMBLE);
163
checkState(MimeTokenStream.T_START_BODYPART);
164
checkState(MimeTokenStream.T_START_HEADER);
165
checkState(MimeTokenStream.T_FIELD);
166
checkState(MimeTokenStream.T_END_HEADER);
167
checkState(MimeTokenStream.T_BODY);
168
checkState(MimeTokenStream.T_END_BODYPART);
169
checkState(MimeTokenStream.T_EPILOGUE);
170
checkState(MimeTokenStream.T_END_MULTIPART);
171
checkState(MimeTokenStream.T_END_BODYPART);
172
checkState(MimeTokenStream.T_EPILOGUE);
173
checkState(MimeTokenStream.T_END_MULTIPART);
174
checkState(MimeTokenStream.T_END_MESSAGE);
175
checkState(MimeTokenStream.T_END_OF_STREAM);
178
public void testShouldParseMessageWithEmbeddedMessage() throws Exception {
179
parser.parse(new ByteArrayInputStream(US_ASCII.encode(COMPLEX_MESSAGE).array()));
180
checkState(MimeTokenStream.T_START_HEADER);
181
checkState(MimeTokenStream.T_FIELD);
182
checkState(MimeTokenStream.T_FIELD);
183
checkState(MimeTokenStream.T_FIELD);
184
checkState(MimeTokenStream.T_FIELD);
185
checkState(MimeTokenStream.T_FIELD);
186
checkState(MimeTokenStream.T_END_HEADER);
187
checkState(MimeTokenStream.T_START_MULTIPART);
188
checkState(MimeTokenStream.T_PREAMBLE);
189
checkState(MimeTokenStream.T_START_BODYPART);
190
checkState(MimeTokenStream.T_START_HEADER);
191
checkState(MimeTokenStream.T_FIELD);
192
checkState(MimeTokenStream.T_END_HEADER);
193
checkState(MimeTokenStream.T_BODY);
194
checkState(MimeTokenStream.T_END_BODYPART);
195
checkState(MimeTokenStream.T_START_BODYPART);
196
checkState(MimeTokenStream.T_START_HEADER);
197
checkState(MimeTokenStream.T_FIELD);
198
checkState(MimeTokenStream.T_END_HEADER);
199
checkState(MimeTokenStream.T_START_MESSAGE);
200
checkState(MimeTokenStream.T_START_HEADER);
201
checkState(MimeTokenStream.T_FIELD);
202
checkState(MimeTokenStream.T_FIELD);
203
checkState(MimeTokenStream.T_FIELD);
204
checkState(MimeTokenStream.T_FIELD);
205
checkState(MimeTokenStream.T_FIELD);
206
checkState(MimeTokenStream.T_END_HEADER);
207
checkState(MimeTokenStream.T_START_MULTIPART);
208
checkState(MimeTokenStream.T_PREAMBLE);
209
checkState(MimeTokenStream.T_START_BODYPART);
210
checkState(MimeTokenStream.T_START_HEADER);
211
checkState(MimeTokenStream.T_END_HEADER);
212
checkState(MimeTokenStream.T_BODY);
213
checkState(MimeTokenStream.T_END_BODYPART);
214
checkState(MimeTokenStream.T_START_BODYPART);
215
checkState(MimeTokenStream.T_START_HEADER);
216
checkState(MimeTokenStream.T_FIELD);
217
checkState(MimeTokenStream.T_END_HEADER);
218
checkState(MimeTokenStream.T_BODY);
219
checkState(MimeTokenStream.T_END_BODYPART);
220
checkState(MimeTokenStream.T_EPILOGUE);
221
checkState(MimeTokenStream.T_END_MULTIPART);
222
checkState(MimeTokenStream.T_END_MESSAGE);
223
checkState(MimeTokenStream.T_END_BODYPART);
224
checkState(MimeTokenStream.T_START_BODYPART);
225
checkState(MimeTokenStream.T_START_HEADER);
226
checkState(MimeTokenStream.T_END_HEADER);
227
checkState(MimeTokenStream.T_BODY);
228
checkState(MimeTokenStream.T_END_BODYPART);
229
checkState(MimeTokenStream.T_EPILOGUE);
230
checkState(MimeTokenStream.T_END_MULTIPART);
231
checkState(MimeTokenStream.T_END_MESSAGE);
232
checkState(MimeTokenStream.T_END_OF_STREAM);
235
public void testShouldParseMessagesWithEmbeddedQuotedPrintableEncodedMessage() throws Exception {
236
parser.parse(new ByteArrayInputStream(US_ASCII.encode(COMPLEX_QP_MESSAGE).array()));
237
checkState(MimeTokenStream.T_START_HEADER);
238
checkState(MimeTokenStream.T_FIELD);
239
checkState(MimeTokenStream.T_FIELD);
240
checkState(MimeTokenStream.T_END_HEADER);
241
checkState(MimeTokenStream.T_START_MESSAGE);
242
checkState(MimeTokenStream.T_START_HEADER);
243
checkState(MimeTokenStream.T_FIELD);
244
checkState(MimeTokenStream.T_FIELD);
245
checkState(MimeTokenStream.T_END_HEADER);
246
checkState(MimeTokenStream.T_START_MULTIPART);
247
checkState(MimeTokenStream.T_PREAMBLE);
248
checkState(MimeTokenStream.T_START_BODYPART);
249
checkState(MimeTokenStream.T_START_HEADER);
250
checkState(MimeTokenStream.T_FIELD);
251
checkState(MimeTokenStream.T_END_HEADER);
252
checkState(MimeTokenStream.T_BODY);
253
assertEquals("text/plain", parser.getBodyDescriptor().getMimeType());
254
assertEquals("iso-8859-1", parser.getBodyDescriptor().getCharset());
255
assertEquals("Some text\r\n",
256
IOUtils.toString(parser.getInputStream()));
257
checkState(MimeTokenStream.T_END_BODYPART);
258
checkState(MimeTokenStream.T_START_BODYPART);
259
checkState(MimeTokenStream.T_START_HEADER);
260
checkState(MimeTokenStream.T_FIELD);
261
checkState(MimeTokenStream.T_END_HEADER);
262
checkState(MimeTokenStream.T_BODY);
263
assertEquals("text/html", parser.getBodyDescriptor().getMimeType());
264
assertEquals("iso-8859-1", parser.getBodyDescriptor().getCharset());
265
assertEquals("<HTML><BODY>= Some HTML =</BODY></HTML>",
266
IOUtils.toString(parser.getInputStream()));
267
checkState(MimeTokenStream.T_END_BODYPART);
268
checkState(MimeTokenStream.T_EPILOGUE);
269
checkState(MimeTokenStream.T_END_MULTIPART);
270
checkState(MimeTokenStream.T_END_MESSAGE);
271
checkState(MimeTokenStream.T_END_MESSAGE);
272
checkState(MimeTokenStream.T_END_OF_STREAM);
275
public void testMultipartMessageWithoutHeader() throws Exception {
276
parser.parseHeadless(new ByteArrayInputStream(US_ASCII.encode(BODY).array()),
277
"multipart/mixed;boundary=1729");
278
checkState(MimeTokenStream.T_END_HEADER);
279
checkState(MimeTokenStream.T_START_MULTIPART);
280
checkState(MimeTokenStream.T_PREAMBLE);
281
checkState(MimeTokenStream.T_START_BODYPART);
282
checkState(MimeTokenStream.T_START_HEADER);
283
checkState(MimeTokenStream.T_END_HEADER);
284
checkState(MimeTokenStream.T_BODY);
285
checkState(MimeTokenStream.T_END_BODYPART);
286
checkState(MimeTokenStream.T_START_BODYPART);
287
checkState(MimeTokenStream.T_START_HEADER);
288
checkState(MimeTokenStream.T_FIELD);
289
checkState(MimeTokenStream.T_END_HEADER);
290
checkState(MimeTokenStream.T_BODY);
291
checkState(MimeTokenStream.T_END_BODYPART);
292
checkState(MimeTokenStream.T_EPILOGUE);
293
checkState(MimeTokenStream.T_END_MULTIPART);
294
checkState(MimeTokenStream.T_END_MESSAGE);
295
checkState(MimeTokenStream.T_END_OF_STREAM);
298
private void checkState(final int state) throws IOException, MimeException {
299
assertEquals(MimeTokenStream.stateToString(state), MimeTokenStream.stateToString(parser.next()));