A script to anonymize names in a text file.
For instance, suppose you are writing a case study and you don't want to shame the innocent or give publicity to the guilty. But you don't want to have to remember pseudonyms while you are writing the document. All you need to do is wrap names you want to anonymize in double curly brackets:
And then {{Jon}} made the most embarrassing mistake imaginable.
If you save that case study in a file named case_study.md, you can
anonymize it thusly:
$ anonymize.rb case_study.md
And then Filiberto made the most embarrassing mistake imaginable.
If you later write:
{{Jon}} was quite relieved that nobody blamed him for his mistake.
It will be anonymized as:
Filberto was quite relieved that nobody blamed him for his mistake.
That way you don't need to memorize Jon => Filberto when writing
since anonymize.rb keeps track for you. Of course, it might not be
Filberto next time you run it. Maybe you'll end up substituting Yung
for {{Jon}} next time.
The script can also work with first and last names if you use an underscore:
It's a good thing {{Jon_Ericson}} doesn't use his full name.
If {{Jon}}has already been defined, the script will make the same
substitution for the first name and create a new substitution for
{{Ericson}}. This may or may not be what you want. If you want to
differentiate two people who happen to have the same name, you can add
a number or other distinguishing feature:
{{Jon1}} hoped he wouldn't get dragged into this mess.
Getting a random substitution every time stops being fun after
awhile. Sometimes you want stability. It also can be handy to have a
list of substitutions if you need to decode the anonymized
document. The script can --save the substitutions as a Ruby hash:
$ anonymize.rb case_study.md -s substitutions.rb
Next time you run the script, --load and reuse the substitutions:
$ anonymize.rb case_study.md -l substitutions.rb
Since you probably want to both load and save substitutions at the
same time, you could use both options or the --cache option:
$ anonymize.rb case_study.md -l substitutions.rb -s substitutions.rb
$ anonymize.rb case_study.md -c substitutions.rb
Both commands are exactly the same.
The script uses the Namey gem
to pick a random name from the
US Census Bureau database. Some of those names are a little unusual. So you can specify just the --common names:
$ anonymize.rb case_study.md --common
Alternatively, you can select --rare names.
Probably the biggest oddity is that the new names won't have the same gender associations as the originals. This can make reading more difficult. I personally prefer to use "they" and "them" for all pronouns when the names are going to be anonymized. That way the pronouns work out whatever the name. But it might be preferable to assign gender to each name.
While Namey does provide a way to select names that are commonly male or commonly female in the data, this script doesn't support a way to assign gender. It's more a practical consideration than a political one; I don't want to invent a new syntax for it. There's also the issue that the data isn't exactly binary. For instance John is #2 on the male list (3.271%) and #819 on the female list (0.012%). So if someone is named John, you can reasonably guess they are a man, but you can't rule out they are a woman.
If sticking to one gender is important to you, I recommend using the
--cache option and manually editing the substitutions file to your
liking.
Namey tries to fix the capitalization of names. Since the census bureau encoded all the names in ALL CAPS, it munged names like McMasters and O'Brien.