svn-gvsig-desktop / branches / v2_0_0_prep / libraries / libGeocoding / src-test / org / gvsig / normalization / operations / TestNormAlgorithm.java @ 29385
History | View | Annotate | Download (10.2 KB)
1 |
/* gvSIG. Geographic Information System of the Valencian Government
|
---|---|
2 |
*
|
3 |
* Copyright (C) 2007-2008 Infrastructures and Transports Department
|
4 |
* of the Valencian Government (CIT)
|
5 |
*
|
6 |
* This program is free software; you can redistribute it and/or
|
7 |
* modify it under the terms of the GNU General Public License
|
8 |
* as published by the Free Software Foundation; either version 2
|
9 |
* of the License, or (at your option) any later version.
|
10 |
*
|
11 |
* This program is distributed in the hope that it will be useful,
|
12 |
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
13 |
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
14 |
* GNU General Public License for more details.
|
15 |
*
|
16 |
* You should have received a copy of the GNU General Public License
|
17 |
* along with this program; if not, write to the Free Software
|
18 |
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
|
19 |
* MA 02110-1301, USA.
|
20 |
*
|
21 |
*/
|
22 |
|
23 |
/*
|
24 |
* AUTHORS (In addition to CIT):
|
25 |
* 2008 Prodevelop S.L main development
|
26 |
*/
|
27 |
|
28 |
package org.gvsig.normalization.operations; |
29 |
|
30 |
import java.io.File; |
31 |
import java.io.FileReader; |
32 |
import java.io.IOException; |
33 |
import java.util.List; |
34 |
|
35 |
import junit.framework.TestCase; |
36 |
|
37 |
import org.gvsig.fmap.dal.DALFileLibrary; |
38 |
import org.gvsig.fmap.dal.DALLibrary; |
39 |
import org.gvsig.fmap.dal.store.shp.SHPLibrary; |
40 |
import org.gvsig.normalization.algorithm.NormalizationAlgorithm; |
41 |
import org.gvsig.normalization.algorithm.impl.DefaultNormalizationAlgorithm; |
42 |
import org.gvsig.normalization.pattern.Patternnormalization; |
43 |
import org.gvsig.normalization.pattern.impl.DefaultPatternnormalization; |
44 |
import org.gvsig.tools.ToolsLibrary; |
45 |
import org.gvsig.tools.ToolsLocator; |
46 |
import org.gvsig.tools.locator.Library; |
47 |
import org.gvsig.tools.persistence.PersistenceException; |
48 |
import org.gvsig.tools.persistence.PersistenceManager; |
49 |
import org.gvsig.tools.persistence.PersistentState; |
50 |
import org.gvsig.tools.persistence.xmlentity.XMLEntityManager; |
51 |
import org.slf4j.Logger; |
52 |
import org.slf4j.LoggerFactory; |
53 |
|
54 |
public class TestNormAlgorithm extends TestCase { |
55 |
|
56 |
private static final Logger log = LoggerFactory |
57 |
.getLogger(TestNormAlgorithm.class); |
58 |
|
59 |
/**
|
60 |
* Constructor
|
61 |
*/
|
62 |
public TestNormAlgorithm() {
|
63 |
|
64 |
ToolsLibrary tools = new ToolsLibrary();
|
65 |
tools.initialize(); |
66 |
ToolsLocator.registerPersistenceManager(XMLEntityManager.class); |
67 |
tools.postInitialize(); |
68 |
|
69 |
Library dlib = new DALLibrary();
|
70 |
dlib.initialize(); |
71 |
dlib.postInitialize(); |
72 |
|
73 |
Library libFile = new DALFileLibrary();
|
74 |
libFile.initialize(); |
75 |
libFile.postInitialize(); |
76 |
|
77 |
Library shpLib = new SHPLibrary();
|
78 |
shpLib.initialize(); |
79 |
shpLib.postInitialize(); |
80 |
} |
81 |
|
82 |
/**
|
83 |
*
|
84 |
* @throws PersistenceException
|
85 |
* @throws IOException
|
86 |
*/
|
87 |
public void testSplitChain() throws PersistenceException, IOException { |
88 |
|
89 |
log.debug("TestNormAlgorithm. Test splits strings");
|
90 |
|
91 |
log.debug("SubTest 1");
|
92 |
String c1 = ";aaa,,bbb;;;ccc/ddd@eee##;"; |
93 |
|
94 |
File f1 = new File( |
95 |
"test-data/normalization/patSplitChain.xml");
|
96 |
|
97 |
assertTrue(f1.exists()); |
98 |
assertNotNull(f1); |
99 |
|
100 |
List<String> r1 = parser(f1, c1); |
101 |
assertNotNull(r1); |
102 |
|
103 |
assertEquals("", (String) r1.get(0)); |
104 |
assertEquals("aaa", (String) r1.get(1)); |
105 |
assertEquals("", (String) r1.get(2)); |
106 |
assertEquals("bbb", (String) r1.get(3)); |
107 |
assertEquals("", (String) r1.get(4)); |
108 |
assertEquals("", (String) r1.get(5)); |
109 |
assertEquals("ccc", (String) r1.get(6)); |
110 |
assertEquals("ddd", (String) r1.get(7)); |
111 |
assertEquals("eee", (String) r1.get(8)); |
112 |
assertEquals("", (String) r1.get(9)); |
113 |
assertEquals(";", (String) r1.get(10)); |
114 |
|
115 |
log.debug("SubTest 2.");
|
116 |
String c2 = "aaa bbb ccc ddd,76 %"; |
117 |
|
118 |
File f2 = new File( |
119 |
"test-data/normalization/patSplitChain2.xml");
|
120 |
assertTrue(f2.exists()); |
121 |
assertNotNull(f2); |
122 |
|
123 |
List<String> r2 = parser(f2, c2); |
124 |
assertNotNull(r2); |
125 |
|
126 |
assertEquals("aaa", (String) r2.get(0)); |
127 |
assertEquals("bbb", (String) r2.get(1)); |
128 |
assertEquals("ccc", (String) r2.get(2)); |
129 |
assertEquals("ddd", (String) r2.get(3)); |
130 |
assertEquals("76", (String) r2.get(4)); |
131 |
assertEquals("%", (String) r2.get(5)); |
132 |
|
133 |
log.debug("SubTest 3.");
|
134 |
String c3 = "Av;Germanias;25;3;Moncada;Valencia"; |
135 |
|
136 |
File f3 = new File( |
137 |
"test-data/normalization/testdata/patSplitChain3.xml");
|
138 |
assertTrue(f3.exists()); |
139 |
assertNotNull(f3); |
140 |
|
141 |
List<String> r3 = parser(f3, c3); |
142 |
assertNotNull(r3); |
143 |
|
144 |
assertEquals("Av", (String) r3.get(0)); |
145 |
assertEquals("Germanias", (String) r3.get(1)); |
146 |
assertEquals(25, Integer.parseInt((String) r3.get(2))); |
147 |
assertEquals(3, Integer.parseInt((String) r3.get(3))); |
148 |
assertEquals("Moncada", (String) r3.get(4)); |
149 |
assertEquals("Valencia", (String) r3.get(5)); |
150 |
|
151 |
log.debug("SubTest 4.");
|
152 |
String c4 = "Av. Germanias 15 2 Moncada Valencia "; |
153 |
|
154 |
File f4 = new File( |
155 |
"test-data/normalization/patSplitChain4.xml");
|
156 |
assertTrue(f4.exists()); |
157 |
assertNotNull(f4); |
158 |
|
159 |
List<String> r4 = parser(f4, c4); |
160 |
assertNotNull(r4); |
161 |
|
162 |
assertEquals("Av.", ((String) r4.get(0)).trim()); |
163 |
assertEquals("Germanias", ((String) r4.get(1)).trim()); |
164 |
assertEquals(15, Integer.parseInt(((String) r4.get(2)).trim())); |
165 |
assertEquals(2, Integer.parseInt(((String) r4.get(3)).trim())); |
166 |
assertEquals("Moncada", ((String) r4.get(4)).trim()); |
167 |
assertEquals("Valencia", ((String) r4.get(5)).trim()); |
168 |
|
169 |
log.debug("SubTest 5.");
|
170 |
String c5 = "Av;,;Germanias;15;;2;Moncada;Valencia"; |
171 |
|
172 |
File f5 = new File( |
173 |
"test-data/normalization/patSplitChain5.xml");
|
174 |
assertTrue(f5.exists()); |
175 |
assertNotNull(f5); |
176 |
|
177 |
List<String> r5 = parser(f5, c5); |
178 |
assertNotNull(r5); |
179 |
|
180 |
assertEquals("Av", (String) r5.get(0)); |
181 |
assertEquals("Germanias", (String) r5.get(1)); |
182 |
assertEquals(15, Integer.parseInt((String) r5.get(2))); |
183 |
assertEquals(2, Integer.parseInt((String) r5.get(3))); |
184 |
assertEquals("Moncada", (String) r5.get(4)); |
185 |
assertEquals("Valencia", (String) r5.get(5)); |
186 |
|
187 |
log.debug("SubTest 6.");
|
188 |
String c6 = "Av. Germanias 15-2 Moncada (Valencia)"; |
189 |
|
190 |
File f6 = new File( |
191 |
"test-data/normalization/patSplitChain6.xml");
|
192 |
assertTrue(f6.exists()); |
193 |
assertNotNull(f6); |
194 |
|
195 |
List<String> r6 = parser(f6, c6); |
196 |
assertNotNull(r6); |
197 |
|
198 |
assertEquals("Av.", ((String) r6.get(0)).trim()); |
199 |
assertEquals("Germanias", ((String) r6.get(1)).trim()); |
200 |
assertEquals(15, Integer.parseInt(((String) r6.get(2)).trim())); |
201 |
assertEquals(2, Integer.parseInt(((String) r6.get(3)).trim())); |
202 |
assertEquals("Moncada", ((String) r6.get(4)).trim()); |
203 |
assertEquals("Valencia", ((String) r6.get(5)).trim()); |
204 |
|
205 |
log.debug("SubTest 7.");
|
206 |
String c7 = "Juana Aguirre;Piedras;No 623;Piso2;Dto.4;C1070AAM;Capital Federal;ARGENTINA"; |
207 |
|
208 |
File f7 = new File( |
209 |
"test-data/normalization/patSplitChain7.xml");
|
210 |
assertTrue(f7.exists()); |
211 |
assertNotNull(f7); |
212 |
|
213 |
List<String> r7 = parser(f7, c7); |
214 |
assertNotNull(r7); |
215 |
|
216 |
assertEquals("Juana Aguirre", ((String) r7.get(0)).trim()); |
217 |
assertEquals("Piedras", ((String) r7.get(1)).trim()); |
218 |
assertEquals("No 623", ((String) r7.get(2)).trim()); |
219 |
assertEquals("Piso2", ((String) r7.get(3)).trim()); |
220 |
assertEquals("Dto.4", ((String) r7.get(4)).trim()); |
221 |
assertEquals("C1070AAM", ((String) r7.get(5)).trim()); |
222 |
assertEquals("Capital Federal", ((String) r7.get(6)).trim()); |
223 |
assertEquals("ARGENTINA", ((String) r7.get(7)).trim()); |
224 |
|
225 |
log.debug("SubTest 8.");
|
226 |
String c8 = "5.548\t5478.254\t0.24578457\t256.21450045"; |
227 |
|
228 |
File f8 = new File( |
229 |
"test-data/normalization/patSplitChain8.xml");
|
230 |
assertTrue(f8.exists()); |
231 |
assertNotNull(f8); |
232 |
|
233 |
List<String> r8 = parser(f8, c8); |
234 |
assertNotNull(r8); |
235 |
|
236 |
assertEquals(5.548d, Double.parseDouble(((String) r8.get(0)).trim()), 0); |
237 |
assertEquals(5478.254, Double.parseDouble(((String) r8.get(1)).trim()), |
238 |
0);
|
239 |
assertEquals(0.24578457, Double |
240 |
.parseDouble(((String) r8.get(2)).trim()), 0); |
241 |
assertEquals(256.21450045, Double.parseDouble(((String) r8.get(3)) |
242 |
.trim()), 0);
|
243 |
|
244 |
log.debug("TestNormAlgorithm. Test finished");
|
245 |
} |
246 |
|
247 |
/**
|
248 |
*
|
249 |
*/
|
250 |
public void testSplitFixed() { |
251 |
|
252 |
log.debug("TestNormAlgorithm. Test tokens fixed");
|
253 |
|
254 |
String chain = "esto;/es;;; una_prueba; de un/ split de una cadena_de texto"; |
255 |
|
256 |
int parts = 4; |
257 |
boolean join = true; |
258 |
String[] separators = { " " }; |
259 |
List<String> result = DefaultNormalizationAlgorithm |
260 |
.splitChainBySeparators(chain, parts, separators, join); |
261 |
log.debug("Cadena inicial: " + chain);
|
262 |
for (int i = 0; i < result.size(); i++) { |
263 |
log.debug("Subcadena" + i + ": " + (String) result.get(i)); |
264 |
} |
265 |
assertEquals("esto;/es;;;", result.get(0)); |
266 |
assertEquals("una_prueba;", result.get(1)); |
267 |
assertEquals("de", result.get(2)); |
268 |
assertEquals("un/ split de una cadena_de texto", result.get(3)); |
269 |
|
270 |
log.debug("TestNormAlgorithm. Test tokens fixed finished");
|
271 |
|
272 |
} |
273 |
|
274 |
/**
|
275 |
*
|
276 |
*/
|
277 |
public void testSplitSeparators() { |
278 |
|
279 |
log.debug("TestNormAlgorithm. Test tokens with separators");
|
280 |
|
281 |
String chain = "esto;/es;;; una_prueba; de un/ split de una cadena_de texto"; |
282 |
|
283 |
int parts = 4; |
284 |
boolean join = true; |
285 |
String[] separators = { " " }; |
286 |
List<String> result = DefaultNormalizationAlgorithm |
287 |
.splitChainBySeparators(chain, parts, separators, join); |
288 |
System.out.println("Cadena inicial: " + chain); |
289 |
for (int i = 0; i < result.size(); i++) { |
290 |
System.out.println("Subcadena" + i + ": " + (String) result.get(i)); |
291 |
} |
292 |
assertEquals("esto;/es;;;", result.get(0)); |
293 |
assertEquals("una_prueba;", result.get(1)); |
294 |
assertEquals("de", result.get(2)); |
295 |
assertEquals("un/ split de una cadena_de texto", result.get(3)); |
296 |
|
297 |
log.debug("TestNormAlgorithm. Test tokens with separators finished");
|
298 |
|
299 |
} |
300 |
|
301 |
/**
|
302 |
*
|
303 |
* @param f
|
304 |
* @param chain
|
305 |
* @return
|
306 |
* @throws PersistenceException
|
307 |
* @throws IOException
|
308 |
*/
|
309 |
private List<String> parser(File f, String chain) |
310 |
throws PersistenceException, IOException { |
311 |
|
312 |
PersistenceManager manager = ToolsLocator.getPersistenceManager(); |
313 |
|
314 |
Patternnormalization pat = new DefaultPatternnormalization();
|
315 |
FileReader reader = new FileReader(f); |
316 |
PersistentState state = manager.loadState(reader); |
317 |
pat.loadFromState(state); |
318 |
reader.close(); |
319 |
|
320 |
NormalizationAlgorithm na = new DefaultNormalizationAlgorithm(pat);
|
321 |
|
322 |
List<String> result = na.splitChain(chain); |
323 |
return result;
|
324 |
|
325 |
} |
326 |
|
327 |
} |