my_obfuscate 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/.document +5 -0
- data/.gitignore +6 -0
- data/LICENSE +19 -0
- data/README.rdoc +55 -0
- data/Rakefile +52 -0
- data/VERSION +1 -0
- data/lib/my_obfuscate.rb +172 -0
- data/my_obfuscate.gemspec +54 -0
- data/spec/my_obfuscate_spec.rb +196 -0
- data/spec/spec_helper.rb +10 -0
- metadata +78 -0
data/.document
ADDED
data/.gitignore
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,19 @@
|
|
1
|
+
Copyright (c) 2009 Honk
|
2
|
+
|
3
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
4
|
+
of this software and associated documentation files (the "Software"), to deal
|
5
|
+
in the Software without restriction, including without limitation the rights
|
6
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
7
|
+
copies of the Software, and to permit persons to whom the Software is
|
8
|
+
furnished to do so, subject to the following conditions:
|
9
|
+
|
10
|
+
The above copyright notice and this permission notice shall be included in
|
11
|
+
all copies or substantial portions of the Software.
|
12
|
+
|
13
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
14
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
15
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
16
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
17
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
18
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
19
|
+
THE SOFTWARE.
|
data/README.rdoc
ADDED
@@ -0,0 +1,55 @@
|
|
1
|
+
= my_obfuscate
|
2
|
+
|
3
|
+
Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.
|
4
|
+
|
5
|
+
= Example Usage
|
6
|
+
|
7
|
+
Make an obfuscator.rb script:
|
8
|
+
|
9
|
+
#!/usr/bin/env ruby
|
10
|
+
require "rubygems"
|
11
|
+
require "my_obfuscate"
|
12
|
+
|
13
|
+
obfuscator = MyObfuscate.new({
|
14
|
+
:people => {
|
15
|
+
:email => { :type => :email, :skip_regexes => [/^[\w\.\_]+@my_company\.com$/i] },
|
16
|
+
:ethnicity => { :type => :null },
|
17
|
+
:crypted_password => { :type => :fixed, :string => "SOME_FIXED_PASSWORD_FOR_EASE_OF_DEBUGGING" },
|
18
|
+
:salt => { :type => :fixed, :string => "SOME_THING" },
|
19
|
+
:remember_token => { :type => :null },
|
20
|
+
:remember_token_expires_at => { :type => :null },
|
21
|
+
:photo_file_name => { :type => :null },
|
22
|
+
:photo_content_type => { :type => :null },
|
23
|
+
:photo_file_size => { :type => :null },
|
24
|
+
:photo_updated_at => { :type => :null },
|
25
|
+
:postal_code => { :type => :fixed, :string => "94109" },
|
26
|
+
:name => { :type => :fixed, :string => "Production User" },
|
27
|
+
:relationship_status => { :type => :fixed, :one_of => ["Single", "Divorced", "Married", "Engaged", "In a Relationship"] },
|
28
|
+
:has_children => { :type => :integer, :between => 0..1 },
|
29
|
+
},
|
30
|
+
|
31
|
+
:invites => :truncate,
|
32
|
+
:invite_requests => :truncate,
|
33
|
+
|
34
|
+
:relationships => {
|
35
|
+
:account_id => { :type => :string, :length => 8, :chars => DatabaseDumpObfuscator::NUMBER_CHARS },
|
36
|
+
:code => { :type => :string, :length => 8, :chars => DatabaseDumpObfuscator::USERNAME_CHARS }
|
37
|
+
}
|
38
|
+
})
|
39
|
+
obfuscator.obfuscate(STDIN, STDOUT)
|
40
|
+
|
41
|
+
And to get an obfuscated dump:
|
42
|
+
mysqldump -c --add-drop-table -u user -p database | ruby obfuscator.rb > obfuscated_dump.sql
|
43
|
+
Note that the -c option on mysqldump is required to use my_obfuscator.
|
44
|
+
|
45
|
+
== Note on Patches/Pull Requests
|
46
|
+
|
47
|
+
* Fork the project.
|
48
|
+
* Make your feature addition or bug fix.
|
49
|
+
* Add tests for it. This is important so I don't break it in a future version unintentionally.
|
50
|
+
* Commit, do not mess with rakefile, version, or history. (If you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull)
|
51
|
+
* Send me a pull request. Bonus points for topic branches.
|
52
|
+
|
53
|
+
== Copyright
|
54
|
+
|
55
|
+
Copyright (c) 2009 Honk. See LICENSE for details.
|
data/Rakefile
ADDED
@@ -0,0 +1,52 @@
|
|
1
|
+
require 'rubygems'
|
2
|
+
require 'rake'
|
3
|
+
|
4
|
+
begin
|
5
|
+
require 'jeweler'
|
6
|
+
Jeweler::Tasks.new do |gem|
|
7
|
+
gem.name = "my_obfuscate"
|
8
|
+
gem.summary = %Q{Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.}
|
9
|
+
gem.description = %Q{Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.}
|
10
|
+
gem.email = "andrew@pivotallabs.com"
|
11
|
+
gem.homepage = "http://github.com/honkster/myobfuscate"
|
12
|
+
gem.authors = ["Andrew Cantino", "Dave Willett", "Mike Grafton", "Mason Glaves"]
|
13
|
+
gem.add_development_dependency "rspec"
|
14
|
+
gem.rubyforge_project = 'my-obfuscate'
|
15
|
+
end
|
16
|
+
|
17
|
+
Jeweler::RubyforgeTasks.new do |rubyforge|
|
18
|
+
rubyforge.doc_task = "rdoc"
|
19
|
+
end
|
20
|
+
rescue LoadError
|
21
|
+
puts "Jeweler (or a dependency) not available. Install it with: sudo gem install jeweler"
|
22
|
+
end
|
23
|
+
|
24
|
+
require 'spec/rake/spectask'
|
25
|
+
Spec::Rake::SpecTask.new(:spec) do |spec|
|
26
|
+
spec.libs << 'lib' << 'spec'
|
27
|
+
spec.spec_files = FileList['spec/**/*_spec.rb']
|
28
|
+
end
|
29
|
+
|
30
|
+
Spec::Rake::SpecTask.new(:rcov) do |spec|
|
31
|
+
spec.libs << 'lib' << 'spec'
|
32
|
+
spec.pattern = 'spec/**/*_spec.rb'
|
33
|
+
spec.rcov = true
|
34
|
+
end
|
35
|
+
|
36
|
+
task :spec => :check_dependencies
|
37
|
+
|
38
|
+
task :default => :spec
|
39
|
+
|
40
|
+
require 'rake/rdoctask'
|
41
|
+
Rake::RDocTask.new do |rdoc|
|
42
|
+
if File.exist?('VERSION')
|
43
|
+
version = File.read('VERSION')
|
44
|
+
else
|
45
|
+
version = ""
|
46
|
+
end
|
47
|
+
|
48
|
+
rdoc.rdoc_dir = 'rdoc'
|
49
|
+
rdoc.title = "my_obfuscate #{version}"
|
50
|
+
rdoc.rdoc_files.include('README*')
|
51
|
+
rdoc.rdoc_files.include('lib/**/*.rb')
|
52
|
+
end
|
data/VERSION
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
0.1.0
|
data/lib/my_obfuscate.rb
ADDED
@@ -0,0 +1,172 @@
|
|
1
|
+
# This can parse mysqldump outputs when the dumps have option -c, which includes column names in the insert statements.
|
2
|
+
require 'jcode'
|
3
|
+
|
4
|
+
class MyObfuscate
|
5
|
+
attr_accessor :config
|
6
|
+
|
7
|
+
INSERT_REGEX = /^\s*INSERT INTO `(.*?)` \((.*?)\) VALUES\s*/i
|
8
|
+
NUMBER_CHARS = "1234567890"
|
9
|
+
USERNAME_CHARS = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_" + NUMBER_CHARS
|
10
|
+
SENSIBLE_CHARS = USERNAME_CHARS + '+-=[{]}/?|!@#$%^&*()`~'
|
11
|
+
|
12
|
+
def initialize(configuration = {})
|
13
|
+
@config = configuration
|
14
|
+
end
|
15
|
+
|
16
|
+
# We assume that every INSERT INTO line occupies one line in the file, with no internal linebreaks.
|
17
|
+
def obfuscate(input_io, output_io)
|
18
|
+
input_io.each do |line|
|
19
|
+
if regex_result = INSERT_REGEX.match(line)
|
20
|
+
table_name = regex_result[1].to_sym
|
21
|
+
columns = regex_result[2].split(/`\s*,\s*`/).map { |col| col.gsub('`',"").to_sym }
|
22
|
+
if config[table_name]
|
23
|
+
output_io.puts obfuscate_line(line, table_name, columns)
|
24
|
+
else
|
25
|
+
output_io.write line
|
26
|
+
end
|
27
|
+
else
|
28
|
+
output_io.write line
|
29
|
+
end
|
30
|
+
end
|
31
|
+
end
|
32
|
+
|
33
|
+
def self.reasembling_each_insert(line, table_name, columns)
|
34
|
+
line = line.gsub(INSERT_REGEX, '').gsub(/\s*;\s*$/, '')
|
35
|
+
output = context_aware_mysql_string_split(line).map do |sub_insert|
|
36
|
+
result = yield(sub_insert)
|
37
|
+
result = result.map do |i|
|
38
|
+
if i.nil?
|
39
|
+
"NULL"
|
40
|
+
else
|
41
|
+
"'" + i + "'"
|
42
|
+
end
|
43
|
+
end
|
44
|
+
result = result.join(",")
|
45
|
+
"(" + result + ")"
|
46
|
+
end.join(",")
|
47
|
+
"INSERT INTO `#{table_name}` (`#{columns.join('`, `')}`) VALUES #{output};"
|
48
|
+
end
|
49
|
+
|
50
|
+
# Note: Strings must be quoted in single quotes!
|
51
|
+
def self.context_aware_mysql_string_split(string)
|
52
|
+
in_sub_insert = false
|
53
|
+
in_quoted_string = false
|
54
|
+
escaped = false
|
55
|
+
current_field = nil
|
56
|
+
length = string.length
|
57
|
+
index = 0
|
58
|
+
fields = []
|
59
|
+
output = []
|
60
|
+
string.each_char do |i|
|
61
|
+
if escaped
|
62
|
+
escaped = false
|
63
|
+
current_field ||= ""
|
64
|
+
current_field << i
|
65
|
+
else
|
66
|
+
if i == "\\"
|
67
|
+
escaped = true
|
68
|
+
current_field ||= ""
|
69
|
+
current_field << i
|
70
|
+
elsif i == "(" && !in_quoted_string && !in_sub_insert
|
71
|
+
in_sub_insert = true
|
72
|
+
elsif i == ")" && !in_quoted_string && in_sub_insert
|
73
|
+
fields << current_field unless current_field.nil?
|
74
|
+
output << fields unless fields.length == 0
|
75
|
+
in_sub_insert = false
|
76
|
+
fields = []
|
77
|
+
current_field = nil
|
78
|
+
elsif i == "'" && !in_quoted_string
|
79
|
+
fields << current_field unless current_field.nil?
|
80
|
+
current_field = ''
|
81
|
+
in_quoted_string = true
|
82
|
+
elsif i == "'" && in_quoted_string
|
83
|
+
fields << current_field unless current_field.nil?
|
84
|
+
current_field = nil
|
85
|
+
in_quoted_string = false
|
86
|
+
elsif i == "," && !in_quoted_string && in_sub_insert
|
87
|
+
fields << current_field unless current_field.nil?
|
88
|
+
current_field = nil
|
89
|
+
elsif i == "L" && !in_quoted_string && in_sub_insert && current_field == "NUL"
|
90
|
+
current_field = nil
|
91
|
+
fields << current_field
|
92
|
+
elsif (i == " " || i == "\t") && !in_quoted_string
|
93
|
+
# Don't add whitespace not in a string
|
94
|
+
elsif in_sub_insert
|
95
|
+
current_field ||= ""
|
96
|
+
current_field << i
|
97
|
+
end
|
98
|
+
end
|
99
|
+
index += 1
|
100
|
+
end
|
101
|
+
fields << current_field unless current_field.nil?
|
102
|
+
output << fields unless fields.length == 0
|
103
|
+
output
|
104
|
+
end
|
105
|
+
|
106
|
+
def self.apply_table_config(row, table_config, columns)
|
107
|
+
return row unless table_config.is_a?(Hash)
|
108
|
+
|
109
|
+
table_config.each do |column, definition|
|
110
|
+
index = columns.index(column)
|
111
|
+
|
112
|
+
if definition[:skip_regexes]
|
113
|
+
next if definition[:skip_regexes].any? {|regex| row[index] =~ regex}
|
114
|
+
end
|
115
|
+
|
116
|
+
row[index.to_i] = case definition[:type]
|
117
|
+
when :email
|
118
|
+
random_string(4..10, USERNAME_CHARS) + "@example.com"
|
119
|
+
when :string
|
120
|
+
random_string(definition[:length], definition[:chars] || SENSIBLE_CHARS)
|
121
|
+
when :integer
|
122
|
+
random_integer(definition[:between] || (0..1000)).to_s
|
123
|
+
when :fixed
|
124
|
+
if definition[:one_of]
|
125
|
+
definition[:one_of][(rand * definition[:one_of].length).to_i]
|
126
|
+
else
|
127
|
+
definition[:string]
|
128
|
+
end
|
129
|
+
when :null
|
130
|
+
nil
|
131
|
+
else
|
132
|
+
row[index]
|
133
|
+
end
|
134
|
+
end
|
135
|
+
row
|
136
|
+
end
|
137
|
+
|
138
|
+
def self.random_integer(between)
|
139
|
+
(between.min + (between.max - between.min) * rand).round
|
140
|
+
end
|
141
|
+
|
142
|
+
def self.random_string(length_or_range, chars)
|
143
|
+
length_or_range = (length_or_range..length_or_range) if length_or_range.is_a?(Fixnum)
|
144
|
+
times = random_integer(length_or_range)
|
145
|
+
out = ""
|
146
|
+
times.times { out << chars[rand * chars.length] }
|
147
|
+
out
|
148
|
+
end
|
149
|
+
|
150
|
+
def check_for_missing_columns(table_name, columns)
|
151
|
+
missing_columns = config[table_name].keys - columns
|
152
|
+
unless missing_columns.length == 0
|
153
|
+
error_message = missing_columns.map do |missing_column|
|
154
|
+
"Column '#{missing_column}' could not be found in table '#{table_name}', please fix your obfuscator config."
|
155
|
+
end.join("\n")
|
156
|
+
raise RuntimeError.new(error_message)
|
157
|
+
end
|
158
|
+
end
|
159
|
+
|
160
|
+
def obfuscate_line(line, table_name, columns)
|
161
|
+
table_config = config[table_name]
|
162
|
+
if table_config == :truncate
|
163
|
+
""
|
164
|
+
else
|
165
|
+
check_for_missing_columns(table_name, columns)
|
166
|
+
# Note: Remember to SQL escape strings in what you pass back.
|
167
|
+
MyObfuscate.reasembling_each_insert(line, table_name, columns) do |row|
|
168
|
+
MyObfuscate.apply_table_config(row, table_config, columns)
|
169
|
+
end
|
170
|
+
end
|
171
|
+
end
|
172
|
+
end
|
@@ -0,0 +1,54 @@
|
|
1
|
+
# Generated by jeweler
|
2
|
+
# DO NOT EDIT THIS FILE
|
3
|
+
# Instead, edit Jeweler::Tasks in Rakefile, and run `rake gemspec`
|
4
|
+
# -*- encoding: utf-8 -*-
|
5
|
+
|
6
|
+
Gem::Specification.new do |s|
|
7
|
+
s.name = %q{my_obfuscate}
|
8
|
+
s.version = "0.1.0"
|
9
|
+
|
10
|
+
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
11
|
+
s.authors = ["Andrew Cantino", "Dave Willett", "Mike Grafton", "Mason Glaves"]
|
12
|
+
s.date = %q{2009-10-01}
|
13
|
+
s.description = %q{Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.}
|
14
|
+
s.email = %q{andrew@pivotallabs.com}
|
15
|
+
s.extra_rdoc_files = [
|
16
|
+
"LICENSE",
|
17
|
+
"README.rdoc"
|
18
|
+
]
|
19
|
+
s.files = [
|
20
|
+
".document",
|
21
|
+
".gitignore",
|
22
|
+
"LICENSE",
|
23
|
+
"README.rdoc",
|
24
|
+
"Rakefile",
|
25
|
+
"VERSION",
|
26
|
+
"lib/my_obfuscate.rb",
|
27
|
+
"my_obfuscate.gemspec",
|
28
|
+
"spec/my_obfuscate_spec.rb",
|
29
|
+
"spec/spec_helper.rb"
|
30
|
+
]
|
31
|
+
s.homepage = %q{http://github.com/honkster/myobfuscate}
|
32
|
+
s.rdoc_options = ["--charset=UTF-8"]
|
33
|
+
s.require_paths = ["lib"]
|
34
|
+
s.rubyforge_project = %q{my-obfuscate}
|
35
|
+
s.rubygems_version = %q{1.3.5}
|
36
|
+
s.summary = %q{Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.}
|
37
|
+
s.test_files = [
|
38
|
+
"spec/my_obfuscate_spec.rb",
|
39
|
+
"spec/spec_helper.rb"
|
40
|
+
]
|
41
|
+
|
42
|
+
if s.respond_to? :specification_version then
|
43
|
+
current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
|
44
|
+
s.specification_version = 3
|
45
|
+
|
46
|
+
if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
|
47
|
+
s.add_development_dependency(%q<rspec>, [">= 0"])
|
48
|
+
else
|
49
|
+
s.add_dependency(%q<rspec>, [">= 0"])
|
50
|
+
end
|
51
|
+
else
|
52
|
+
s.add_dependency(%q<rspec>, [">= 0"])
|
53
|
+
end
|
54
|
+
end
|
@@ -0,0 +1,196 @@
|
|
1
|
+
require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
|
2
|
+
|
3
|
+
describe MyObfuscate do
|
4
|
+
describe "MyObfuscate.context_aware_mysql_string_split" do
|
5
|
+
it "should split a mysql string into fields" do
|
6
|
+
string = "('bob@bob.com','bob', 'somethingelse1', 25, '2', 10, 'hi')"
|
7
|
+
fields = [['bob@bob.com', 'bob', 'somethingelse1', '25', '2', '10', "hi"]]
|
8
|
+
MyObfuscate.context_aware_mysql_string_split(string).should == fields
|
9
|
+
end
|
10
|
+
|
11
|
+
it "should work ok with escaped characters" do
|
12
|
+
string = "('bob,@bob.c , om', 'bo\\', b', 'some\"thin\\gel\\\\\\'se1', 25, '2', 10, 'hi', 5)"
|
13
|
+
fields = [['bob,@bob.c , om', 'bo\\\', b', 'some"thin\\gel\\\\\\\'se1', '25', '2', '10', "hi", "5"]]
|
14
|
+
MyObfuscate.context_aware_mysql_string_split(string).should == fields
|
15
|
+
end
|
16
|
+
|
17
|
+
it "should work with multiple subinserts" do
|
18
|
+
string = "(1,2,3, '((m))(oo()s,e'), ('bob,@bob.c , om', 'bo\\', b', 'some\"thin\\gel\\\\\\'se1', 25, '2', 10, 'hi', 5)"
|
19
|
+
fields = [["1", "2", "3", "((m))(oo()s,e"], ['bob,@bob.c , om', 'bo\\\', b', 'some"thin\\gel\\\\\\\'se1', '25', '2', '10', "hi", "5"]]
|
20
|
+
MyObfuscate.context_aware_mysql_string_split(string).should == fields
|
21
|
+
end
|
22
|
+
|
23
|
+
it "should work ok with NULL values" do
|
24
|
+
string = "(NULL , 'bob@bob.com','bob', NULL, 25, '2', NULL, 'hi', NULL )"
|
25
|
+
fields = [[nil, 'bob@bob.com', 'bob', nil, '25', '2', nil, "hi", nil]]
|
26
|
+
MyObfuscate.context_aware_mysql_string_split(string).should == fields
|
27
|
+
end
|
28
|
+
|
29
|
+
it "should work with empty strings" do
|
30
|
+
string = "(NULL , '', '' , '', 25, '2','', 'hi','')"
|
31
|
+
fields = [[nil, '', '', '', '25', '2', '', "hi", '']]
|
32
|
+
MyObfuscate.context_aware_mysql_string_split(string).should == fields
|
33
|
+
end
|
34
|
+
end
|
35
|
+
|
36
|
+
describe "MyObfuscate.reasembling_each_insert" do
|
37
|
+
before do
|
38
|
+
@column_names = [:a, :b, :c, :d]
|
39
|
+
@test_insert = "INSERT INTO `some_table` (`a`, `b`, `c`, `d`) VALUES ('(\\'bob@bob.com','b()ob','some(thingelse1','25)('),('joe@joe.com','joe','somethingelse2','54');"
|
40
|
+
@test_insert_passes = [
|
41
|
+
["(\\'bob@bob.com","b()ob", "some(thingelse1", "25)("],
|
42
|
+
["joe@joe.com","joe", "somethingelse2", "54"]
|
43
|
+
]
|
44
|
+
end
|
45
|
+
|
46
|
+
it "should yield each subinsert and reassemble the result" do
|
47
|
+
count = 0
|
48
|
+
reassembled = MyObfuscate.reasembling_each_insert(@test_insert, "some_table", @column_names) do |sub_insert|
|
49
|
+
sub_insert.should == @test_insert_passes.shift
|
50
|
+
count += 1
|
51
|
+
sub_insert
|
52
|
+
end
|
53
|
+
count.should == 2
|
54
|
+
reassembled.should == @test_insert
|
55
|
+
end
|
56
|
+
end
|
57
|
+
|
58
|
+
describe "MyObfuscate.apply_table_config" do
|
59
|
+
it "should work on email addresses" do
|
60
|
+
new_row = MyObfuscate.apply_table_config(["blah", "something_else"], { :a => { :type => :email }}, [:a, :b])
|
61
|
+
new_row.length.should == 2
|
62
|
+
new_row.first.should =~ /^\w+\@\w+\.\w+$/
|
63
|
+
end
|
64
|
+
|
65
|
+
it "should work on strings" do
|
66
|
+
new_row = MyObfuscate.apply_table_config(["blah", "something_else", "something crazy"], { :b => { :type => :string, :length => 7 }}, [:a, :b, :c])
|
67
|
+
new_row.length.should == 3
|
68
|
+
new_row[1].length.should == 7
|
69
|
+
new_row[1].should_not == "something_else"
|
70
|
+
end
|
71
|
+
|
72
|
+
it "should be able to generate random integers in ranges" do
|
73
|
+
new_row = MyObfuscate.apply_table_config(["blah", "something_else", "5"], { :c => { :type => :integer, :between => 10..100 }}, [:a, :b, :c])
|
74
|
+
new_row.length.should == 3
|
75
|
+
new_row[2].to_i.to_s.should == new_row[2] # It should be an integer.
|
76
|
+
new_row[2].should_not == "5"
|
77
|
+
end
|
78
|
+
|
79
|
+
it "should be able to substitute fixed strings" do
|
80
|
+
new_row = MyObfuscate.apply_table_config(["blah", "something_else", "5"], { :b => { :type => :fixed, :string => "hello" }}, [:a, :b, :c])
|
81
|
+
new_row.length.should == 3
|
82
|
+
new_row[1].should == "hello"
|
83
|
+
end
|
84
|
+
|
85
|
+
it "should be able to substitute fixed strings from a random set" do
|
86
|
+
looking_for = ["hello", "world"]
|
87
|
+
original_looking_for = looking_for.dup
|
88
|
+
guard = 0
|
89
|
+
while !looking_for.empty? && guard < 1000
|
90
|
+
new_row = MyObfuscate.apply_table_config(["blah", "something_else", "5"], { :a => { :type => :fixed, :one_of => ["hello", "world"] }}, [:a, :b, :c])
|
91
|
+
new_row.length.should == 3
|
92
|
+
original_looking_for.should include(new_row[0])
|
93
|
+
looking_for.delete new_row[0]
|
94
|
+
guard += 1
|
95
|
+
end
|
96
|
+
looking_for.should be_empty
|
97
|
+
end
|
98
|
+
|
99
|
+
it "should be able to set things NULL" do
|
100
|
+
new_row = MyObfuscate.apply_table_config(["blah", "something_else", "5"], { :b => { :type => :null }}, [:a, :b, :c])
|
101
|
+
new_row.length.should == 3
|
102
|
+
new_row[1].should == nil
|
103
|
+
end
|
104
|
+
|
105
|
+
end
|
106
|
+
|
107
|
+
describe "#obfuscate" do
|
108
|
+
context "when there is nothing to obfuscate" do
|
109
|
+
it "should accept an IO object for input and output, and copy the input to the output" do
|
110
|
+
ddo = MyObfuscate.new
|
111
|
+
string = "hello, world\nsup?"
|
112
|
+
input = StringIO.new(string)
|
113
|
+
output = StringIO.new
|
114
|
+
ddo.obfuscate(input, output)
|
115
|
+
input.rewind
|
116
|
+
output.rewind
|
117
|
+
output.read.should == string
|
118
|
+
end
|
119
|
+
end
|
120
|
+
|
121
|
+
context "when the dump to obfuscate is missing columns" do
|
122
|
+
before do
|
123
|
+
@database_dump = StringIO.new(<<-SQL)
|
124
|
+
INSERT INTO `some_table` (`email`, `name`, `something`, `age`) VALUES ('bob@honk.com','bob', 'some\\'thin,ge())lse1', 25),('joe@joe.com','joe', 'somethingelse2', 54);
|
125
|
+
SQL
|
126
|
+
@ddo = MyObfuscate.new({
|
127
|
+
:some_table => {
|
128
|
+
:email => { :type => :email, :honk_email_skip => true },
|
129
|
+
:name => { :type => :string, :length => 8, :chars => MyObfuscate::USERNAME_CHARS },
|
130
|
+
:gender => { :type => :fixed, :string => "m" }
|
131
|
+
}})
|
132
|
+
@output = StringIO.new
|
133
|
+
end
|
134
|
+
|
135
|
+
it "should raise an error if a column name can't be found" do
|
136
|
+
lambda {
|
137
|
+
@ddo.obfuscate(@database_dump, @output)
|
138
|
+
}.should raise_error
|
139
|
+
end
|
140
|
+
end
|
141
|
+
|
142
|
+
context "when there is something to obfuscate" do
|
143
|
+
before do
|
144
|
+
@database_dump = StringIO.new(<<-SQL)
|
145
|
+
INSERT INTO `some_table` (`email`, `name`, `something`, `age`) VALUES ('bob@honk.com','bob', 'some\\'thin,ge())lse1', 25),('joe@joe.com','joe', 'somethingelse2', 54),('dontmurderme@direwolf.com','direwolf', 'somethingelse3', 44);
|
146
|
+
INSERT INTO `another_table` (`a`, `b`, `c`, `d`) VALUES (1,2,3,4), (5,6,7,8);
|
147
|
+
INSERT INTO `one_more_table` (`a`, `password`, `c`, `d,d`) VALUES ('hello','kjhjd^&dkjh', 'aawefjkafe'), ('hello1','kjhj!', 892938), ('hello2','moose!!', NULL);
|
148
|
+
INSERT INTO `an_ignored_table` (`col`, `col2`) VALUES ('hello','kjhjd^&dkjh'), ('hello1','kjhj!'), ('hello2','moose!!');
|
149
|
+
SQL
|
150
|
+
|
151
|
+
@ddo = MyObfuscate.new({
|
152
|
+
:some_table => {
|
153
|
+
:email => { :type => :email, :skip_regexes => [/^[\w\.\_]+@honk\.com$/i, /^dontmurderme@direwolf.com$/] },
|
154
|
+
:name => { :type => :string, :length => 8, :chars => MyObfuscate::USERNAME_CHARS },
|
155
|
+
:age => { :type => :integer, :between => 10...80 }
|
156
|
+
},
|
157
|
+
:another_table => :truncate,
|
158
|
+
:one_more_table => {
|
159
|
+
# Note: fixed strings must be pre-SQL escaped!
|
160
|
+
:password => { :type => :fixed, :string => "monkey" },
|
161
|
+
:c => { :type => :null }
|
162
|
+
}
|
163
|
+
})
|
164
|
+
@output = StringIO.new
|
165
|
+
@ddo.obfuscate(@database_dump, @output)
|
166
|
+
@output.rewind
|
167
|
+
@output_string = @output.read
|
168
|
+
end
|
169
|
+
|
170
|
+
it "should be able to truncate tables" do
|
171
|
+
@output_string.should_not include("INSERT INTO `another_table`")
|
172
|
+
@output_string.should include("INSERT INTO `one_more_table`")
|
173
|
+
end
|
174
|
+
|
175
|
+
it "should ignore tables that it doesn't know about" do
|
176
|
+
@output_string.should include("INSERT INTO `an_ignored_table` (`col`, `col2`) VALUES ('hello','kjhjd^&dkjh'), ('hello1','kjhj!'), ('hello2','moose!!');")
|
177
|
+
end
|
178
|
+
|
179
|
+
it "should obfuscate the tables" do
|
180
|
+
@output_string.should include("INSERT INTO `some_table` (`email`, `name`, `something`, `age`) VALUES (")
|
181
|
+
@output_string.should include("INSERT INTO `one_more_table` (`a`, `password`, `c`, `d,d`) VALUES (")
|
182
|
+
@output_string.should include("'some\\'thin,ge())lse1'")
|
183
|
+
@output_string.should include("INSERT INTO `one_more_table` (`a`, `password`, `c`, `d,d`) VALUES ('hello','monkey',NULL),('hello1','monkey',NULL),('hello2','monkey',NULL);")
|
184
|
+
@output_string.should_not include("INSERT INTO `one_more_table` (`a`, `password`, `c`, `d,d`) VALUES ('hello','kjhjd^&dkjh', 'aawefjkafe'), ('hello1','kjhj!', 892938), ('hello2','moose!!', NULL);")
|
185
|
+
@output_string.should_not include("INSERT INTO `one_more_table` (`a`, `password`, `c`, `d,d`) VALUES ('hello','kjhjd^&dkjh','aawefjkafe'),('hello1','kjhj!',892938),('hello2','moose!!',NULL);")
|
186
|
+
@output_string.should_not include("INSERT INTO `some_table` (`email`, `name`, `something`, `age`) VALUES ('bob@honk.com','bob', 'some\\'thin,ge())lse1', 25),('joe@joe.com','joe', 'somethingelse2', 54);")
|
187
|
+
end
|
188
|
+
|
189
|
+
it "honors a special case: on the people table, rows with anything@honk.com in a slot marked with :honk_email_skip do not change this slot" do
|
190
|
+
@output_string.should include("('bob@honk.com',")
|
191
|
+
@output_string.should include("('dontmurderme@direwolf.com',")
|
192
|
+
@output_string.should_not include("joe@joe.com")
|
193
|
+
end
|
194
|
+
end
|
195
|
+
end
|
196
|
+
end
|
data/spec/spec_helper.rb
ADDED
metadata
ADDED
@@ -0,0 +1,78 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: my_obfuscate
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.1.0
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Andrew Cantino
|
8
|
+
- Dave Willett
|
9
|
+
- Mike Grafton
|
10
|
+
- Mason Glaves
|
11
|
+
autorequire:
|
12
|
+
bindir: bin
|
13
|
+
cert_chain: []
|
14
|
+
|
15
|
+
date: 2009-10-01 00:00:00 -07:00
|
16
|
+
default_executable:
|
17
|
+
dependencies:
|
18
|
+
- !ruby/object:Gem::Dependency
|
19
|
+
name: rspec
|
20
|
+
type: :development
|
21
|
+
version_requirement:
|
22
|
+
version_requirements: !ruby/object:Gem::Requirement
|
23
|
+
requirements:
|
24
|
+
- - ">="
|
25
|
+
- !ruby/object:Gem::Version
|
26
|
+
version: "0"
|
27
|
+
version:
|
28
|
+
description: Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.
|
29
|
+
email: andrew@pivotallabs.com
|
30
|
+
executables: []
|
31
|
+
|
32
|
+
extensions: []
|
33
|
+
|
34
|
+
extra_rdoc_files:
|
35
|
+
- LICENSE
|
36
|
+
- README.rdoc
|
37
|
+
files:
|
38
|
+
- .document
|
39
|
+
- .gitignore
|
40
|
+
- LICENSE
|
41
|
+
- README.rdoc
|
42
|
+
- Rakefile
|
43
|
+
- VERSION
|
44
|
+
- lib/my_obfuscate.rb
|
45
|
+
- my_obfuscate.gemspec
|
46
|
+
- spec/my_obfuscate_spec.rb
|
47
|
+
- spec/spec_helper.rb
|
48
|
+
has_rdoc: true
|
49
|
+
homepage: http://github.com/honkster/myobfuscate
|
50
|
+
licenses: []
|
51
|
+
|
52
|
+
post_install_message:
|
53
|
+
rdoc_options:
|
54
|
+
- --charset=UTF-8
|
55
|
+
require_paths:
|
56
|
+
- lib
|
57
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
58
|
+
requirements:
|
59
|
+
- - ">="
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: "0"
|
62
|
+
version:
|
63
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
64
|
+
requirements:
|
65
|
+
- - ">="
|
66
|
+
- !ruby/object:Gem::Version
|
67
|
+
version: "0"
|
68
|
+
version:
|
69
|
+
requirements: []
|
70
|
+
|
71
|
+
rubyforge_project: my-obfuscate
|
72
|
+
rubygems_version: 1.3.5
|
73
|
+
signing_key:
|
74
|
+
specification_version: 3
|
75
|
+
summary: Standalone Ruby code for the selective rewriting of MySQL dumps in order to protect user privacy.
|
76
|
+
test_files:
|
77
|
+
- spec/my_obfuscate_spec.rb
|
78
|
+
- spec/spec_helper.rb
|