Nothing earth shattering today, but I thought it was neat. Every now and then I come across a document I would like to add to my test suites for a project, but there’s a problem. The document might contain email addresses, phone numbers, who knows what, and I don’t really want that to be sitting around in my codebase, but the structure of the document matters. The solution? Pipe it through this:
tr 'A-Za-z0-9' 'a-ff-aa-fa-ff-aa-fa-ff-aa-f001122334455'
Now all of your punctuation and such is preserved for your parsing pleasure, but the data is somewhat obfuscated. Of course use it at your discretion.
For bonus points, I actually made it into a TextMate command. Now I can have my test data, and feel reasonably safe about it.