Tweet Compression
tedunangst.com
tedunangst.com
And to some degree of surprise, there are indeed several strange bits in the implementation.
One example:
document.getElementById("wrapped").innerHTML = w.replace(/&/g, "&").replace(/</g, "<")
Are we really so used to unsafe APIs that we prefer escaping even if the safe API is actually shorter? document.getElementById("wrapped").textContent = w
Also, this is declaring a global variable "i" for no reason - probably forgot a "var i"? function burrito() {
var c = ...
var w = ...
var u = ...
for (i = 0; i < fillings.length; i++) {
...
Finally, the preference for unsafe APIs continues, treating plain text as regex, this time without proper escaping: var fillings = [
[ "Th", "Þ" ],
[ "th", "þ" ],
...
w = w.replace(new RegExp(fillings[i][0], "g"), fillings[i][1])
While this is non-critical from a security perspective (because the replacement list is hard coded into the program), this begs for bugs. Good luck if you want to put something into your replacement list than contains a ".", "+" or similar. At least a small warning at the top of the list would be a nice service to the future self.