24
|
1 |
/*
|
|
2 |
* Copyright (c) 2010 Nokia Corporation and/or its subsidiary(-ies).
|
|
3 |
* All rights reserved.
|
|
4 |
* This component and the accompanying materials are made available
|
|
5 |
* under the terms of "Eclipse Public License v1.0"
|
|
6 |
* which accompanies this distribution, and is available
|
|
7 |
* at the URL "http://www.eclipse.org/legal/epl-v10.html".
|
|
8 |
*
|
|
9 |
* Initial Contributors:
|
|
10 |
* Nokia Corporation - initial contribution.
|
|
11 |
*
|
|
12 |
* Contributors:
|
|
13 |
*
|
|
14 |
* Description:
|
|
15 |
*
|
|
16 |
*/
|
|
17 |
|
|
18 |
#include "itk.h"
|
|
19 |
|
|
20 |
#include "thaianalysis.h"
|
|
21 |
|
|
22 |
#include "CLucene.h"
|
|
23 |
|
|
24 |
#include <iostream>
|
|
25 |
|
|
26 |
#include "testutils.h"
|
|
27 |
|
|
28 |
#include "evaluationtool.h"
|
|
29 |
#include "analysisunittest.h"
|
|
30 |
#include "thaistatemachine.h"
|
|
31 |
|
|
32 |
|
|
33 |
using namespace analysis;
|
|
34 |
using namespace evaluationtool;
|
|
35 |
using namespace lucene::util;
|
|
36 |
|
|
37 |
using namespace lucene::analysis;
|
|
38 |
|
|
39 |
#define THAI_TEXTCORPUS "c:\\data\\analysisunittestcorpus\\thai\\corpus.txt"
|
|
40 |
|
|
41 |
#define BUFFER_SIZE 512
|
|
42 |
|
|
43 |
void printBreaks(BreakIterator& breaks, const wchar_t* text)
|
|
44 |
{
|
|
45 |
breaks.setText( text );
|
|
46 |
|
|
47 |
while (breaks.hasNext())
|
|
48 |
{
|
|
49 |
int begin = breaks.current();
|
|
50 |
int end = breaks.next();
|
|
51 |
|
|
52 |
wchar_t buf[BUFFER_SIZE];
|
|
53 |
memcpy(buf, text+begin, sizeof(wchar_t)*(end-begin));
|
|
54 |
buf[end-begin] = '\0';
|
|
55 |
printf(" '%S'", buf);
|
|
56 |
}
|
|
57 |
printf("\n");
|
|
58 |
}
|
|
59 |
|
|
60 |
void thaiBreakIteratorTest(Itk::TestMgr* testMgr)
|
|
61 |
{
|
|
62 |
std::auto_ptr<BreakIterator> breaks( ThaiAnalysisInfra::theInstance()->createBreakIterator() );
|
|
63 |
|
|
64 |
FileReader reader(THAI_TEXTCORPUS, "UTF-8");
|
|
65 |
|
|
66 |
Corpus corpus(THAI_TEXTCORPUS);
|
|
67 |
|
|
68 |
for (int i = 0; i < corpus.size(); i++) {
|
|
69 |
printBreaks(*breaks, corpus[i]);
|
|
70 |
}
|
|
71 |
}
|
|
72 |
|
|
73 |
void thaiAnalyzerTest(Itk::TestMgr* testMgr)
|
|
74 |
{
|
|
75 |
ThaiAnalyzer analyzer;
|
|
76 |
|
|
77 |
Corpus corpus(THAI_TEXTCORPUS);
|
|
78 |
|
|
79 |
for (int i = 0; i < corpus.size(); i++) {
|
|
80 |
printTokens(analyzer, corpus[i]);
|
|
81 |
}
|
|
82 |
}
|
|
83 |
|
|
84 |
void thaiAnalyzerSpeed(Itk::TestMgr* testMgr)
|
|
85 |
{
|
|
86 |
ThaiAnalyzer analyzer;
|
|
87 |
FileReader reader(THAI_TEXTCORPUS, "UTF-8");
|
|
88 |
int filesize = Cpt::filesize(THAI_TEXTCORPUS);
|
|
89 |
|
|
90 |
Itk::Timestamp begin;
|
|
91 |
Itk::getTimestamp(&begin);
|
|
92 |
|
|
93 |
auto_ptr<TokenStream> stream( analyzer.tokenStream( NULL, &reader ) );
|
|
94 |
lucene::analysis::Token token;
|
|
95 |
while (stream->next(&token)); // go throught all tokens
|
|
96 |
|
|
97 |
Itk::Timestamp end;
|
|
98 |
Itk::getTimestamp(&end);
|
|
99 |
|
|
100 |
long time = Itk::getElapsedMs(&end, &begin);
|
|
101 |
ITK_REPORT( testMgr, "Thai analysis time", "%d ms / %d KB", time, (filesize/1000));
|
|
102 |
ITK_REPORT( testMgr, "Thai analysis speed", "%d KB/s", (filesize / time));
|
|
103 |
}
|
|
104 |
|
|
105 |
|
|
106 |
void thaiControlSpeed(Itk::TestMgr* testMgr)
|
|
107 |
{
|
|
108 |
lucene::analysis::standard::StandardAnalyzer analyzer;
|
|
109 |
FileReader reader(THAI_TEXTCORPUS, "UTF-8");
|
|
110 |
int filesize = Cpt::filesize(THAI_TEXTCORPUS);
|
|
111 |
|
|
112 |
Itk::Timestamp begin;
|
|
113 |
Itk::getTimestamp(&begin);
|
|
114 |
|
|
115 |
auto_ptr<TokenStream> stream( analyzer.tokenStream( NULL, &reader ) );
|
|
116 |
lucene::analysis::Token token;
|
|
117 |
while (stream->next(&token)); // go throught all tokens
|
|
118 |
|
|
119 |
Itk::Timestamp end;
|
|
120 |
Itk::getTimestamp(&end);
|
|
121 |
|
|
122 |
long time = Itk::getElapsedMs(&end, &begin);
|
|
123 |
ITK_REPORT( testMgr, "Thai control time", "%d ms / %d KB", time, (filesize/1000));
|
|
124 |
ITK_REPORT( testMgr, "Thai control speed", "%d KB/s", (filesize / time));
|
|
125 |
}
|
|
126 |
|
|
127 |
|
|
128 |
|
|
129 |
Itk::TesterBase * CreateThaiAnalysisUnitTest()
|
|
130 |
{
|
|
131 |
using namespace Itk;
|
|
132 |
|
|
133 |
SuiteTester
|
|
134 |
* testSuite =
|
|
135 |
new SuiteTester( "thai" );
|
|
136 |
|
|
137 |
testSuite->add( "breaks", thaiBreakIteratorTest, "breaks" );
|
|
138 |
testSuite->add( "analyzer", thaiAnalyzerTest, "analyzer" );
|
|
139 |
testSuite->add( "analyzerSpeed", thaiAnalyzerSpeed );
|
|
140 |
testSuite->add( "controlSpeed", thaiControlSpeed );
|
|
141 |
|
|
142 |
return testSuite;
|
|
143 |
}
|
|
144 |
|