|
1
|
|
|
/* |
|
2
|
|
|
* Copyright 2014-2020, Armenak Grigoryan, and individual contributors as indicated |
|
3
|
|
|
* by the @authors tag. See the copyright.txt in the distribution for a |
|
4
|
|
|
* full listing of individual contributors. |
|
5
|
|
|
* |
|
6
|
|
|
* This is free software; you can redistribute it and/or modify it |
|
7
|
|
|
* under the terms of the GNU Lesser General Public License as |
|
8
|
|
|
* published by the Free Software Foundation; either version 2.1 of |
|
9
|
|
|
* the License, or (at your option) any later version. |
|
10
|
|
|
* |
|
11
|
|
|
* This software 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 GNU |
|
14
|
|
|
* Lesser General Public License for more details. |
|
15
|
|
|
*/ |
|
16
|
|
|
package com.strider.datadefender.specialcase; |
|
17
|
|
|
|
|
18
|
|
|
import com.strider.datadefender.discoverer.Discoverer.ColumnMatch; |
|
19
|
|
|
import java.io.BufferedReader; |
|
20
|
|
|
import java.io.FileReader; |
|
21
|
|
|
import java.io.IOException; |
|
22
|
|
|
|
|
23
|
|
|
import java.util.ArrayList; |
|
24
|
|
|
import java.util.List; |
|
25
|
|
|
import java.util.Locale; |
|
26
|
|
|
|
|
27
|
|
|
import com.strider.datadefender.discoverer.Probability; |
|
28
|
|
|
import com.strider.datadefender.database.metadata.TableMetaData.ColumnMetaData; |
|
29
|
|
|
import java.util.Objects; |
|
30
|
|
|
|
|
31
|
|
|
import lombok.extern.log4j.Log4j2; |
|
32
|
|
|
import org.apache.commons.lang3.StringUtils; |
|
33
|
|
|
|
|
34
|
|
|
/** |
|
35
|
|
|
* @author Armenak Grigoryan |
|
36
|
|
|
*/ |
|
37
|
|
|
@Log4j2 |
|
38
|
|
|
public class PhiDetector implements SpecialCase { |
|
39
|
|
|
|
|
40
|
|
|
private static final String PHI_FILE = "phi.txt"; |
|
41
|
|
|
private static List phiList = new ArrayList(); |
|
42
|
|
|
|
|
43
|
|
|
static { |
|
44
|
|
|
try { |
|
45
|
|
|
log.info("*** reading from " + PHI_FILE); |
|
46
|
|
|
|
|
47
|
|
|
try (BufferedReader br = new BufferedReader(new FileReader(PHI_FILE))) { |
|
48
|
|
|
for (String line; (line = br.readLine()) != null; ) { |
|
49
|
|
|
phiList.add(line); |
|
50
|
|
|
} |
|
51
|
|
|
} |
|
52
|
|
|
} catch (IOException ioe) { |
|
53
|
|
|
log.error(ioe.toString()); |
|
54
|
|
|
} |
|
55
|
|
|
} |
|
56
|
|
|
|
|
57
|
|
|
/** |
|
58
|
|
|
* Generates random 9-digit student number |
|
59
|
|
|
* @param data |
|
60
|
|
|
* @param text |
|
61
|
|
|
* @return String |
|
62
|
|
|
*/ |
|
63
|
|
|
public static ColumnMatch isPhiTerm(final ColumnMetaData data, final String text) { |
|
64
|
|
|
if (StringUtils.isNotBlank(text) |
|
65
|
|
|
&& Objects.equals(String.class, data.getColumnType()) |
|
66
|
|
|
&& phiList.contains(text.trim().toLowerCase(Locale.ENGLISH))) { |
|
67
|
|
|
|
|
68
|
|
|
log.debug("PHI detected: " + text); |
|
69
|
|
|
return new ColumnMatch( |
|
70
|
|
|
data, |
|
71
|
|
|
100.0, |
|
72
|
|
|
"phi", |
|
73
|
|
|
List.of(new Probability(text, 1.00)) |
|
74
|
|
|
); |
|
75
|
|
|
} |
|
76
|
|
|
return null; |
|
77
|
|
|
} |
|
78
|
|
|
} |
armenak /
DataDefender
