Java > Open Source Codes > org > htmlparser > tests > lexerTests > PageTests


1 // HTMLParser Library $Name: v1_5_20050313 $ - A java-based parser for HTML
2 // http://sourceforge.org/projects/htmlparser
3 // Copyright (C) 2004 Derrick Oswald
4 //
5 // Revision Control Information
6 //
7 // $Source: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/tests/lexerTests/PageTests.java,v $
8 // $Author: derrickoswald $
9 // $Date: 2004/07/31 16:42:31 $
10 // $Revision: 1.18 $
11 //
12 // This library is free software; you can redistribute it and/or
13 // modify it under the terms of the GNU Lesser General Public
14 // License as published by the Free Software Foundation; either
15 // version 2.1 of the License, or (at your option) any later version.
16 //
17 // This library is distributed in the hope that it will be useful,
18 // but WITHOUT ANY WARRANTY; without even the implied warranty of
19 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
20 // Lesser General Public License for more details.
21 //
22 // You should have received a copy of the GNU Lesser General Public
23 // License along with this library; if not, write to the Free Software
24 // Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
25 //
26
27 package org.htmlparser.tests.lexerTests;
28
29 import java.io.IOException ;
30 import java.net.URL ;
31 import java.net.URLConnection ;
32
33 import org.htmlparser.lexer.Page;
34 import org.htmlparser.tests.ParserTestCase;
35 import org.htmlparser.util.ParserException;
36
37 public class PageTests extends ParserTestCase
38 {
39     static
40     {
41         System.setProperty ("org.htmlparser.tests.lexerTests.PageTests", "PageTests");
42     }
43
44     /**
45      * The default charset.
46      * This should be <code>ISO-8859-1</code>,
47      * see RFC 2616 (http://www.ietf.org/rfc/rfc2616.txt?number=2616) section 3.7.1
48      * Another alias is "8859_1".
49      */
50     public static final String DEFAULT_CHARSET = "ISO-8859-1";
51
52     /**
53      * Base URI for absolute URL tests.
54      */
55     static final String BASEURI = "http://a/b/c/d;p?q";
56
57     /**
58      * Page for absolute URL tests.
59      */
60     public static Page mPage;
61     static
62     {
63         mPage = new Page ();
64         mPage.setBaseUrl (BASEURI);
65     }
66         
67     /**
68      * Test the third level page class.
69      */
70     public PageTests (String name)
71     {
72         super (name);
73     }
74
75     /**
76      * Test initialization with a null value.
77      */
78     public void testNull () throws ParserException
79     {
80         try
81         {
82             new Page ((URLConnection )null);
83             assertTrue ("null value in constructor", false);
84         }
85         catch (IllegalArgumentException iae)
86         {
87             // expected outcome
88 }
89
90         try
91         {
92             new Page ((String )null);
93             assertTrue ("null value in constructor", false);
94         }
95         catch (IllegalArgumentException iae)
96         {
97             // expected outcome
98 }
99     }
100
101     /**
102      * Test initialization with a real value.
103      */
104     public void testURLConnection () throws ParserException, IOException
105     {
106         String link;
107         URL url;
108
109         link = "http://www.ibm.com/jp/";
110         url = new URL (link);
111         new Page (url.openConnection ());
112     }
113
114     /**
115      * Test initialization with non-existant URL.
116      */
117     public void testBadURLConnection () throws IOException
118     {
119         String link;
120         URL url;
121
122         link = "http://www.bigbogosity.org/";
123         url = new URL (link);
124         try
125         {
126            new Page (url.openConnection ());
127         }
128         catch (ParserException pe)
129         {
130             // expected response
131 }
132     }
133
134     //
135 // Tests from Appendix C Examples of Resolving Relative URI References
136 // RFC 2396 Uniform Resource Identifiers (URI): Generic Syntax
137 // T. Berners-Lee et al.
138 // http://www.ietf.org/rfc/rfc2396.txt
139
140     // Within an object with a well-defined base URI of
141 // http://a/b/c/d;p?q
142 // the relative URI would be resolved as follows:
143
144     // C.1. Normal Examples
145 // g:h = g:h
146 // g = http://a/b/c/g
147 // ./g = http://a/b/c/g
148 // g/ = http://a/b/c/g/
149 // /g = http://a/g
150 // //g = http://g
151 // ?y = http://a/b/c/?y
152 // g?y = http://a/b/c/g?y
153 // #s = (current document)#s
154 // g#s = http://a/b/c/g#s
155 // g?y#s = http://a/b/c/g?y#s
156 // ;x = http://a/b/c/;x
157 // g;x = http://a/b/c/g;x
158 // g;x?y#s = http://a/b/c/g;x?y#s
159 // . = http://a/b/c/
160 // ./ = http://a/b/c/
161 // .. = http://a/b/
162 // ../ = http://a/b/
163 // ../g = http://a/b/g
164 // ../.. = http://a/
165 // ../../ = http://a/
166 // ../../g = http://a/g
167
168     public void test1 () throws ParserException
169     {
170         assertEquals ("test1 failed", "https:h", mPage.getAbsoluteURL ("https:h"));
171     }
172     public void test2 () throws ParserException
173     {
174         assertEquals ("test2 failed", "http://a/b/c/g", mPage.getAbsoluteURL ("g"));
175     }
176     public void test3 () throws ParserException
177     {
178         assertEquals ("test3 failed", "http://a/b/c/g", mPage.getAbsoluteURL ("./g"));
179     }
180     public void test4 () throws ParserException
181     {
182         assertEquals ("test4 failed", "http://a/b/c/g/", mPage.getAbsoluteURL ("g/"));
183     }
184     public void test5 () throws ParserException
185     {
186         assertEquals ("test5 failed", "http://a/g", mPage.getAbsoluteURL ("/g"));
187     }
188     public void test6 () throws ParserException
189     {
190         assertEquals ("test6 failed", "http://g", mPage.getAbsoluteURL ("//g"));
191     }
192     public void test7 () throws ParserException
193     {
194         assertEquals ("test7 failed", "http://a/b/c/?y", mPage.getAbsoluteURL ("?y"));
195     }
196     public void test8 () throws ParserException
197     {
198         assertEquals ("test8 failed", "http://a/b/c/g?y", mPage.getAbsoluteURL ("g?y"));
199     }
200     public void test9 () throws ParserException
201     {
202         assertEquals ("test9 failed", "https:h", mPage.getAbsoluteURL ("https:h"));
203     }
204     public void test10 () throws ParserException
205     {
206         assertEquals ("test10 failed", "https:h", mPage.getAbsoluteURL ("https:h"));
207     }
208     // #s = (current document)#s
209 public void test11 () throws ParserException
210     {
211         assertEquals ("test11 failed", "http://a/b/c/g#s", mPage.getAbsoluteURL ("g#s"));
212     }
213     public void test12 () throws ParserException
214     {
215         assertEquals ("test12 failed", "http://a/b/c/g?y#s", mPage.getAbsoluteURL ("g?y#s"));
216     }
217     public void test13 () throws ParserException
218     {
219         assertEquals ("test13 failed", "http://a/b/c/;x", mPage.getAbsoluteURL (";x"));
220     }
221     public void test14 () throws ParserException
222     {
223         assertEquals ("test14 failed", "http://a/b/c/g;x", mPage.getAbsoluteURL ("g;x"));
224     }
225     public void test15 () throws ParserException
226     {
227         assertEquals ("test15 failed", "http://a/b/c/g;x?y#s", mPage.getAbsoluteURL ("g;x?y#s"));
228     }
229     public void test16 () throws ParserException
230     {
231         assertEquals ("test16 failed", "http://a/b/c/", mPage.getAbsoluteURL ("."));
232     }
233     public void test17 () throws ParserException
234     {
235         assertEquals ("test17 failed", "http://a/b/c/", mPage.getAbsoluteURL ("./"));
236     }
237     public void test18 () throws ParserException
238     {
239         assertEquals ("test18 failed", "http://a/b/", mPage.getAbsoluteURL (".."));
240     }
241     public void test19 () throws ParserException
242     {
243         assertEquals ("test19 failed", "http://a/b/", mPage.getAbsoluteURL ("../"));
244     }
245     public void test20 () throws ParserException
246     {
247         assertEquals ("test20 failed", "http://a/b/g", mPage.getAbsoluteURL ("../g"));
248     }
249     public void test21 () throws ParserException
250     {
251         assertEquals ("test21 failed", "http://a/", mPage.getAbsoluteURL ("../.."));
252     }
253     public void test22 () throws ParserException
254     {
255         assertEquals ("test22 failed", "http://a/g", mPage.getAbsoluteURL ("../../g"));
256     }
257
258     // C.2. Abnormal Examples
259 // Although the following abnormal examples are unlikely to occur in
260 // normal practice, all URI parsers should be capable of resolving them
261 // consistently. Each example uses the same base as above.
262 //
263 // An empty reference refers to the start of the current document.
264 //
265 // <> = (current document)
266 //
267 // Parsers must be careful in handling the case where there are more
268 // relative path ".." segments than there are hierarchical levels in the
269 // base URI's path. Note that the ".." syntax cannot be used to change
270 // the authority component of a URI.
271 //
272 // ../../../g = http://a/../g
273 // ../../../../g = http://a/../../g
274 //
275 // In practice, some implementations strip leading relative symbolic
276 // elements (".", "..") after applying a relative URI calculation, based
277 // on the theory that compensating for obvious author errors is better
278 // than allowing the request to fail. Thus, the above two references
279 // will be interpreted as "http://a/g" by some implementations.
280 //
281 // Similarly, parsers must avoid treating "." and ".." as special when
282 // they are not complete components of a relative path.
283 //
284 // /./g = http://a/./g
285 // /../g = http://a/../g
286 // g. = http://a/b/c/g.
287 // .g = http://a/b/c/.g
288 // g.. = http://a/b/c/g..
289 // ..g = http://a/b/c/..g
290 //
291 // Less likely are cases where the relative URI uses unnecessary or
292 // nonsensical forms of the "." and ".." complete path segments.
293 //
294 // ./../g = http://a/b/g
295 // ./g/. = http://a/b/c/g/
296 // g/./h = http://a/b/c/g/h
297 // g/../h = http://a/b/c/h
298 // g;x=1/./y = http://a/b/c/g;x=1/y
299 // g;x=1/../y = http://a/b/c/y
300 //
301 // All client applications remove the query component from the base URI
302 // before resolving relative URI. However, some applications fail to
303 // separate the reference's query and/or fragment components from a
304 // relative path before merging it with the base path. This error is
305 // rarely noticed, since typical usage of a fragment never includes the
306 // hierarchy ("/") character, and the query component is not normally
307 // used within relative references.
308 //
309 // g?y/./x = http://a/b/c/g?y/./x
310 // g?y/../x = http://a/b/c/g?y/../x
311 // g#s/./x = http://a/b/c/g#s/./x
312 // g#s/../x = http://a/b/c/g#s/../x
313 //
314 // Some parsers allow the scheme name to be present in a relative URI if
315 // it is the same as the base URI scheme. This is considered to be a
316 // loophole in prior specifications of partial URI [RFC1630]. Its use
317 // should be avoided.
318 //
319 // http:g = http:g ; for validating parsers
320 // | http://a/b/c/g ; for backwards compatibility
321
322 // public void test23 () throws HTMLParserException
323 // {
324 // assertEquals ("test23 failed", "http://a/../g", mPage.getAbsoluteURL ("../../../g"));
325 // }
326 // public void test24 () throws HTMLParserException
327 // {
328 // assertEquals ("test24 failed", "http://a/../../g", mPage.getAbsoluteURL ("../../../../g"));
329 // }
330 public void test23 () throws ParserException
331     {
332         assertEquals ("test23 failed", "http://a/g", mPage.getAbsoluteURL ("../../../g"));
333     }
334     public void test24 () throws ParserException
335     {
336         assertEquals ("test24 failed", "http://a/g", mPage.getAbsoluteURL ("../../../../g"));
337     }
338     public void test25 () throws ParserException
339     {
340         assertEquals ("test25 failed", "http://a/./g", mPage.getAbsoluteURL ("/./g"));
341     }
342     public void test26 () throws ParserException
343     {
344         assertEquals ("test26 failed", "http://a/../g", mPage.getAbsoluteURL ("/../g"));
345     }
346     public void test27 () throws ParserException
347     {
348         assertEquals ("test27 failed", "http://a/b/c/g.", mPage.getAbsoluteURL ("g."));
349     }
350     public void test28 () throws ParserException
351     {
352         assertEquals ("test28 failed", "http://a/b/c/.g", mPage.getAbsoluteURL (".g"));
353     }
354     public void test29 () throws ParserException
355     {
356         assertEquals ("test29 failed", "http://a/b/c/g..", mPage.getAbsoluteURL ("g.."));
357     }
358     public void test30 () throws ParserException
359     {
360         assertEquals ("test30 failed", "http://a/b/c/..g", mPage.getAbsoluteURL ("..g"));
361     }
362     public void test31 () throws ParserException
363     {
364         assertEquals ("test31 failed", "http://a/b/g", mPage.getAbsoluteURL ("./../g"));
365     }
366     public void test32 () throws ParserException
367     {
368         assertEquals ("test32 failed", "http://a/b/c/g/", mPage.getAbsoluteURL ("./g/."));
369     }
370     public void test33 () throws ParserException
371     {
372         assertEquals ("test33 failed", "http://a/b/c/g/h", mPage.getAbsoluteURL ("g/./h"));
373     }
374     public void test34 () throws ParserException
375     {
376         assertEquals ("test34 failed", "http://a/b/c/h", mPage.getAbsoluteURL ("g/../h"));
377     }
378     public void test35 () throws ParserException
379     {
380         assertEquals ("test35 failed", "http://a/b/c/g;x=1/y", mPage.getAbsoluteURL ("g;x=1/./y"));
381     }
382     public void test36 () throws ParserException
383     {
384         assertEquals ("test36 failed", "http://a/b/c/y", mPage.getAbsoluteURL ("g;x=1/../y"));
385     }
386     public void test37 () throws ParserException
387     {
388         assertEquals ("test37 failed", "http://a/b/c/g?y/./x", mPage.getAbsoluteURL ("g?y/./x"));
389     }
390     public void test38 () throws ParserException
391     {
392         assertEquals ("test38 failed", "http://a/b/c/g?y/../x", mPage.getAbsoluteURL ("g?y/../x"));
393     }
394     public void test39 () throws ParserException
395     {
396         assertEquals ("test39 failed", "http://a/b/c/g#s/./x", mPage.getAbsoluteURL ("g#s/./x"));
397     }
398     public void test40 () throws ParserException
399     {
400         assertEquals ("test40 failed", "http://a/b/c/g#s/../x", mPage.getAbsoluteURL ("g#s/../x"));
401     }
402 // public void test41 () throws HTMLParserException
403 // {
404 // assertEquals ("test41 failed", "http:g", mPage.getAbsoluteURL ("http:g"));
405 // }
406 public void test41 () throws ParserException
407     {
408         assertEquals ("test41 failed", "http://a/b/c/g", mPage.getAbsoluteURL ("http:g"));
409     }
410
411 }

A to Z: JavaDoc & Examples

---

Daily Java News & Articles

---

Open Source Projects

---

Open Source Codes

---

Free Computer Books

---

Remove Frame

---

Popular Tags