tidus 1.0.9 → 1.1.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/tidus/strategies/ean_anonymizer.rb +6 -0
- data/lib/tidus/strategies/postgresql/ean_anonymizer.rb +147 -0
- data/lib/tidus/version.rb +1 -1
- metadata +4 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 2abe2b87088b2b0c6b190b31dd6deaa8db0fcbe5
|
4
|
+
data.tar.gz: 1816fedc56bf0c901745eb49db0ae6d9df71c87e
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 506f5b3c3f53586e3c9bdc909fc4cddb01b3b3c22ab85110a744e0478b38a504dc57ae5112fbb282b59d23b9decbc85518917b74fd4025a52336e11eb6fe907d
|
7
|
+
data.tar.gz: 3ed54eb937f1be9ee00d96764b5e6636386a75ad4c6e861077ef1d43ecc4f1158e37e9cc7a98acb27d2ddd560bd5100fbc21eec3a78d9555636a28857232e59e
|
@@ -0,0 +1,147 @@
|
|
1
|
+
module Tidus
|
2
|
+
module Postgresql
|
3
|
+
class EanAnonymizer
|
4
|
+
BASE_MAPPING = '0123456789'.freeze
|
5
|
+
DEFAULT_MAPPING_START = 1.freeze
|
6
|
+
|
7
|
+
def self.anonymize(table_name, column_name, options = {})
|
8
|
+
mapping_snippet = build_mapped_digit_snippet(column_name, options)
|
9
|
+
|
10
|
+
query = <<-SQL
|
11
|
+
(
|
12
|
+
SELECT
|
13
|
+
string_agg(new_digits.digit::TEXT, ''::TEXT)
|
14
|
+
FROM (
|
15
|
+
(
|
16
|
+
SELECT
|
17
|
+
pos,
|
18
|
+
digit
|
19
|
+
FROM (
|
20
|
+
#{mapping_snippet}
|
21
|
+
) AS where_sub
|
22
|
+
ORDER BY pos ASC
|
23
|
+
)
|
24
|
+
UNION ALL
|
25
|
+
(
|
26
|
+
SELECT
|
27
|
+
LENGTH(#{column_name}::TEXT) AS pos,
|
28
|
+
(
|
29
|
+
10
|
30
|
+
-
|
31
|
+
(
|
32
|
+
(
|
33
|
+
SELECT
|
34
|
+
SUM(digit)
|
35
|
+
FROM (
|
36
|
+
#{mapping_snippet}
|
37
|
+
) AS where_sub
|
38
|
+
WHERE pos % 2 = LENGTH(#{column_name}::TEXT) % 2
|
39
|
+
)
|
40
|
+
+
|
41
|
+
(
|
42
|
+
SELECT
|
43
|
+
SUM(digit)
|
44
|
+
FROM (
|
45
|
+
#{mapping_snippet}
|
46
|
+
) AS where_sub
|
47
|
+
WHERE pos % 2 = (LENGTH(#{column_name}::TEXT) - 1) % 2
|
48
|
+
) * 3
|
49
|
+
) % 10
|
50
|
+
) % 10 AS digit
|
51
|
+
)
|
52
|
+
) new_digits
|
53
|
+
)
|
54
|
+
SQL
|
55
|
+
|
56
|
+
return query.gsub!(/\n/, '').gsub!(/\ +/, ' ')
|
57
|
+
end
|
58
|
+
|
59
|
+
# Generates a new mapping if no cache_key is given
|
60
|
+
# Generates a new mapping if the cache_key is unknown
|
61
|
+
# Reuses the found mapping if the cache_key is known
|
62
|
+
def self.retrieve_mapping(cache_key)
|
63
|
+
@cached_mappings ||= {}
|
64
|
+
mapping = @cached_mappings[cache_key]
|
65
|
+
|
66
|
+
if mapping == nil
|
67
|
+
mapping = {
|
68
|
+
base: BASE_MAPPING,
|
69
|
+
replacement: BASE_MAPPING.split('').shuffle.join('')
|
70
|
+
}
|
71
|
+
end
|
72
|
+
|
73
|
+
if cache_key != nil
|
74
|
+
@cached_mappings[cache_key] = mapping
|
75
|
+
end
|
76
|
+
|
77
|
+
return mapping
|
78
|
+
end
|
79
|
+
|
80
|
+
def self.build_mapped_digit_snippet(column_name, options)
|
81
|
+
substrings = build_digit_mapping(column_name, options)
|
82
|
+
|
83
|
+
return <<-SQL
|
84
|
+
SELECT
|
85
|
+
*
|
86
|
+
FROM (
|
87
|
+
SELECT
|
88
|
+
ROW_NUMBER() over () AS pos,
|
89
|
+
digit::INT
|
90
|
+
FROM (
|
91
|
+
SELECT
|
92
|
+
REGEXP_SPLIT_TO_TABLE(
|
93
|
+
#{substrings.join(' || ')},
|
94
|
+
''
|
95
|
+
) AS digit
|
96
|
+
) AS sub
|
97
|
+
) AS sub
|
98
|
+
WHERE pos < LENGTH(#{column_name}::TEXT)
|
99
|
+
SQL
|
100
|
+
end
|
101
|
+
|
102
|
+
def self.build_digit_mapping(column_name, options)
|
103
|
+
mapping = retrieve_mapping(options[:cache_key])
|
104
|
+
|
105
|
+
options[:start] ||= DEFAULT_MAPPING_START
|
106
|
+
|
107
|
+
# reduce length by 1 to take away the check digit
|
108
|
+
total_length = "LENGTH(#{column_name}::TEXT) - 1"
|
109
|
+
length = options[:length] || total_length
|
110
|
+
|
111
|
+
substrings = []
|
112
|
+
|
113
|
+
if options[:start] != DEFAULT_MAPPING_START
|
114
|
+
substrings << build_substring_snippet(
|
115
|
+
column_name,
|
116
|
+
DEFAULT_MAPPING_START,
|
117
|
+
options[:start] - DEFAULT_MAPPING_START
|
118
|
+
)
|
119
|
+
end
|
120
|
+
|
121
|
+
translate_part = <<-SNIPPET
|
122
|
+
TRANSLATE(
|
123
|
+
#{build_substring_snippet(column_name, options[:start], length)},
|
124
|
+
'#{mapping[:base]}',
|
125
|
+
'#{mapping[:replacement]}'
|
126
|
+
)
|
127
|
+
SNIPPET
|
128
|
+
|
129
|
+
substrings << translate_part.gsub(/\n/, '').gsub(/\ +/, ' ')
|
130
|
+
|
131
|
+
if options[:length] != nil
|
132
|
+
substrings << build_substring_snippet(
|
133
|
+
column_name,
|
134
|
+
"#{options[:start]} + #{options[:length]}",
|
135
|
+
"#{total_length} - #{options[:start]}"
|
136
|
+
)
|
137
|
+
end
|
138
|
+
|
139
|
+
return substrings
|
140
|
+
end
|
141
|
+
|
142
|
+
def self.build_substring_snippet(column_name, start, length)
|
143
|
+
return "SUBSTRING(#{column_name}::TEXT, #{start}, #{length})"
|
144
|
+
end
|
145
|
+
end
|
146
|
+
end
|
147
|
+
end
|
data/lib/tidus/version.rb
CHANGED
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: tidus
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.0
|
4
|
+
version: 1.1.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Tobias Schoknecht
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2019-01-14 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: rake
|
@@ -79,10 +79,12 @@ files:
|
|
79
79
|
- lib/tidus/query.rb
|
80
80
|
- lib/tidus/strategies/base_selector.rb
|
81
81
|
- lib/tidus/strategies/cond_anonymizer.rb
|
82
|
+
- lib/tidus/strategies/ean_anonymizer.rb
|
82
83
|
- lib/tidus/strategies/email_anonymizer.rb
|
83
84
|
- lib/tidus/strategies/null_anonymizer.rb
|
84
85
|
- lib/tidus/strategies/overlay_anonymizer.rb
|
85
86
|
- lib/tidus/strategies/postgresql/cond_anonymizer.rb
|
87
|
+
- lib/tidus/strategies/postgresql/ean_anonymizer.rb
|
86
88
|
- lib/tidus/strategies/postgresql/email_anonymizer.rb
|
87
89
|
- lib/tidus/strategies/postgresql/null_anonymizer.rb
|
88
90
|
- lib/tidus/strategies/postgresql/overlay_anonymizer.rb
|