Package it.unimi.dsi.mg4j.document

Examples of it.unimi.dsi.mg4j.document.TRECDocumentCollection$TRECDocumentDescriptor

27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
    OutputStream outputStream = new FileOutputStream(temp);
    IOUtils.copy(this.getClass().getResourceAsStream("testChar255.data"),
        outputStream);
    outputStream.close();

    TRECDocumentCollection collection = new TRECDocumentCollection(
        new String[] { temp.toString() },
        CompositeDocumentFactory
            .getFactory(new DocumentFactory[] {
                new TRECHeaderDocumentFactory(),
                new HtmlDocumentFactory( new String[] { "encoding=ISO-8859-1" } ) } ),
        4, // Very small, to induce fragmentation
        false);

    try {
      DocumentIterator iter = collection.iterator();
      Document d;
      while ((d = iter.nextDocument()) != null)
        d.title();
    } catch (IllegalStateException e) {
      assertTrue(false);
View Full Code Here
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
    IOUtils.copy( this.getClass().getResourceAsStream( "testContents.data" ), outputStream );
    outputStream.close();
    IOUtils.copy( this.getClass().getResourceAsStream( "testContentsAgain.data" ), outputStreamAgain );
    outputStreamAgain.close();

    TRECDocumentCollection collection = new TRECDocumentCollection(
        new String[] { temp.toString(), tempAgain.toString() },
        CompositeDocumentFactory
            .getFactory(new DocumentFactory[] {
                new TRECHeaderDocumentFactory(),
                new HtmlDocumentFactory( new String[] { "encoding=ISO-8859-1" } ) } ),
        4, // Very small, to induce fragmentation
        false);

    DocumentIterator iter = collection.iterator();
    Document d = null;

    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0001/", d.uri());
    assertEquals("GX001", d.title());

    final int textIndex = collection.factory().fieldIndex( "text" );
   
    assertEquals( "Line 1\n     The line 2!\n  Mamma\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
   
    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0002/", d.uri());
    assertEquals("GX002", d.title());

    assertEquals( "Contents of this file reside on one line only\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );

    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0003/", d.uri());
    assertEquals("GX003", d.title());

    assertEquals( "Line 1\nLine 2\nLine 3\nLine 4\nLine 5\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );

    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0004/", d.uri());
    assertEquals("GX004", d.title());

    assertEquals( "New content 0\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
   
    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0005/", d.uri());
    assertEquals("GX005", d.title());

    assertEquals( "New content 1\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );

    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0006/", d.uri());
    assertEquals("GX006", d.title());

    assertEquals( "New content 2\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );

    d = iter.nextDocument();
    assertNotNull(d);
    assertEquals("http://gx0007/", d.uri());
    assertEquals("GX007", d.title());

    assertEquals( "", IOUtils.toString( (Reader)d.content( textIndex ) ) );

    d = iter.nextDocument();
    assertNull(d);
    iter.close();
   
    d = collection.document( 0 );
    assertNotNull(d);
    assertEquals("http://gx0001/", d.uri());
    assertEquals("GX001", d.title());

    assertEquals( "Line 1\n     The line 2!\n  Mamma\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
    d.close();
   
    d = collection.document( 1 );
    assertNotNull(d);
    assertEquals("http://gx0002/", d.uri());
    assertEquals("GX002", d.title());

    assertEquals( "Contents of this file reside on one line only\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
    d.close();

    d = collection.document( 2 );
    assertNotNull(d);
    assertEquals("http://gx0003/", d.uri());
    assertEquals("GX003", d.title());

    assertEquals( "Line 1\nLine 2\nLine 3\nLine 4\nLine 5\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
    d.close();

    d = collection.document( 3 );
    assertNotNull(d);
    assertEquals("http://gx0004/", d.uri());
    assertEquals("GX004", d.title());

    assertEquals( "New content 0\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
    d.close();
   
    d = collection.document( 4 );
    assertNotNull(d);
    assertEquals("http://gx0005/", d.uri());
    assertEquals("GX005", d.title());

    assertEquals( "New content 1\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
    d.close();

    d = collection.document( 5 );
    assertNotNull(d);
    assertEquals("http://gx0006/", d.uri());
    assertEquals("GX006", d.title());

    assertEquals( "New content 2\n", IOUtils.toString( (Reader)d.content( textIndex ) ) );
    d.close();

    d = collection.document( 6 );
    assertNotNull(d);
    assertEquals("http://gx0007/", d.uri());
    assertEquals("GX007", d.title());

    assertEquals( "", IOUtils.toString( (Reader)d.content( textIndex ) ) );
View Full Code Here
TOP

Related Classes of it.unimi.dsi.mg4j.document.TRECDocumentCollection$TRECDocumentDescriptor

  • com.martiansoftware.jsap.FlaggedOption
  • com.martiansoftware.jsap.JSAPResult
  • com.martiansoftware.jsap.SimpleJSAP
  • com.martiansoftware.jsap.Switch
  • com.martiansoftware.jsap.UnflaggedOption
  • it.unimi.dsi.fastutil.io.FastBufferedInputStream
  • it.unimi.dsi.io.SegmentedInputStream
  • it.unimi.dsi.logging.ProgressLogger
  • it.unimi.dsi.mg4j.document.TRECDocumentCollectionTest
  • java.io.InputStream
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.