Used to redact a hash by removing, digesting or encrypting keys.
Makes use of attr_encrypted
to perform encryption.
Useful if you have large JSON objects or similar that are coming in through API calls or elsewhere and you want to scrub the private data before storing it to the database.
Add this line to your application's Gemfile:
gem 'hash_redactor'
And then execute:
$ bundle
Or install it yourself as:
$ gem install hash_redactor
Initialize the HashRedactor with a set of fields that you want to redact in hashes that get passed to it.
You can choose 3 ways to redact each field:
:keep
- The field is untouched (for use in:whitelist
mode):remove
- The field is simply deleted:digest
- The field is passed through a one way hash function (SHA256):encrypt
- The field is encrypted
The modes can be specified as strings also (to facilitate easy loading from a JSONB field)
If you plan on using digest or encrypt, then you will need to specify :digest_salt
or :encryption_key
respectively in the options.
fields_to_redact = {
:ssn => :remove,
:email => :digest,
:medical_history => :encrypt
}
redactor = HashRedactor::HashRedactor.new({
redact: fields_to_redact,
encryption_key: 'a really secure key no one will ever guess it',
digest_salt: 'a secret salt'
})
Once initialized, you can simply call #redact
on hashes of data.
If a key you have specified to redact is missing, HashRedactor will silently ignore it's absence.
Any key's not specified in your redact hash will be left unchanged.
raw_data = {
ssn: 'a very personal number',
email: 'personal@email.com',
medical_history: 'Some very private information'
plaintext: 'This will never be changed'
}
safe_data = redactor.redact(raw_data)
safe_data[:ssn] # nil
safe_data[:email_digest] # One way hash of the email
safe_data[:encrypted_medical_history] # Encrypted medical history
safe_data[:plaintext] # This will never be changed
To retrieve your encrypted data, pass it through decrypt.
loaded_data = redactor.decrypt(safe_data)
loaded_data[:medical_history] # 'Some very private information'
Digest may be used in cases where you don't want to keep the data due to privacy concerns, but want to compare it. For example, maybe you have a large database of people whom you never need to contact by email, but you want to use the email address to check for duplicates.
To enhance protection of the data you should provide a salt in the configuration options and you should not store that salt in your database.
Once the field has been digested, it's key will have _digest appended.
Encryption is for fields that you do need to recover, but don't want to store in plaintext in the database.
attr_encrypted
is used to perform encryption.
Once the field has been encrypted, the original field is removed from the hash, and replaced with :encrypted_FIELDNAME
and :encrypted_FIELDNAME_iv
- both these fields must be present in the hash for decrypt to succeed.
For example, if the field you want to encrypt has the key :email
, then the redacred hash will have the keys :encrypted_email
and :encrypted_email_iv
.
From the example above, after running #redact, the safe_data hash would look something like this:
{
:email_digest=>"VXJsQeG81HYWOb30XfpBdbqEFH1f4VaFLTqHCdxCmj8=", :encrypted_medical_history=>"2JIN3Yhxvm/m7qlE+n4pMT9yckXuPa+2IlMBFQMcbP1pcwyrG7wy0TP4scgx\n",
:encrypted_medical_history_iv=>"tqMQa1aYTdJuMhrD\n",
:plaintext => 'This will never be changed'
}
Default options are:
digest_salt: "",
encryption_key: nil,
encode: true,
encode_iv: true,
default_encoding: 'm',
filter_mode: :blacklist,
digest_empty: true
Salt to use when applying digest method
Key to use for encryption and decryption of values
Determines how (if at all) to encode the encrypted data and it's iv
The default encoding is m (base64). You can change this by setting encode: 'some encoding'
. See Arrary#pack for more encoding options.
Filter mode determines how keys that are not included in the redact option are handled.
It can be either :whitelist
or :blacklist
In :blacklist
mode the default, it will leave unspecified keys untouched.
In :whitelist
mode any keys not specified will be removes.
redactor = HashRedactor::HashRedactor.new({
redact: { :id => :keep, :ssn => :remove, :history => :encrypt, :email => :digest" },
encryption_key: 'a really secure key no one will ever guess it',
digest_salt: 'a secret salt'
})
data = {
id: 42,
ssn: 'my ssn number',
history: 'Intriguing"
email: 'personal@email.com',
age: "that's personal"
}
result = reactor.redact data
result[:id] # 42
result[:ssn] # nil
result[:encrypted_history] # encrypted value
result[:email_digest] # digest of email
result[:age] # "that's personal"
result = reactor.redact data, :filter_mode => :whitelist
result[:id] # 42 (because of :id => :keep)
result[:ssn] # nil
result[:encrypted_history] # encrypted value
result[:email_digest] # digest of email
result[:age] # nil (because it wasn't explicitly whitelisted)
Note: To prevent accidental deletion of digest information during repeated loading and unloading data, the digest of all values is implicitly assumed to be :keep.
eg If your redact hash includes :email => :digest
, it is assumed to also contain :email_digest => :keep
Determines if the empty string or nil should be digested, defaults to true for backwards compatibility.
data = { empty: '', not_a_thing: nil }
redactor = HashRedactor::HashRedactor.new({
:empty => :digest, :not_a_thing => :digest
})
result = redactor.redact(data)
result[:empty_digest] # some digest
result[:not_a_thing_digest] # some digest
redactor.redact(data, digest_empty: false)
result[:empty_digest] # ''
result[:not_a_thing_digest] # nil
After checking out the repo, run bin/setup
to install dependencies. Then, run bin/console
for an interactive prompt that will allow you to experiment.
To install this gem onto your local machine, run bundle exec rake install
. To release a new version, update the version number in version.rb
, and then run bundle exec rake release
to create a git tag for the version, push git commits and tags, and push the .gem
file to rubygems.org.
- Fork it ( https://github.com/chrisjensen/hash_redactor/fork )
- Create your feature branch (
git checkout -b my-new-feature
) - Commit your changes (
git commit -am 'Add some feature'
) - Push to the branch (
git push origin my-new-feature
) - Create a new Pull Request