:PROPERTIES:
:ID:       8750a413-7cb8-4160-b4b3-3b54265b2825
:END:
#+title: Percent encoding in Bash
#+date: "2021-06-23 21:10:33 +08:00"
#+date_modified: "2022-04-16 20:17:53 +08:00"
#+language: en


Percent-encoding is commonly used in URLs.
Pretty nice especially when used with spaces.
Though, this is something that is rarely needed in [[id:dd9d3ffa-03ff-42a1-8c5d-55dc9fcc70fe][GNU Bash]] except if you're writing a quick script.

Based from [[https://gist.github.com/cdown/1163649][this implementation]].
It was modified to be closer to the output from [[https://docs.python.org/3.8/library/urllib.parse.html#urllib.parse.urlencode][Python =urlencode=]].

#+begin_src bash
function urlencode {
    local msg=$1
    local length="${#1}"

    for (( i = 0; i < length; i++ )); do
        local ch="${msg:i:1}"
        case $ch in
            [a-zA-Z0-9.~_-])
                printf "%c" $ch;;
            ,*)
                printf '%s' "$ch" | xxd -plain -cols 1 | {
                    while read hex; do
                        printf "%%%s" $hex
                    done
                };;
        esac
    done
    printf "\n"
}

urlencode "El Doggo"
urlencode "El Niño"
urlencode "Whoa there, sonny!"
#+end_src

#+results:
: ElDoggo
: ElNiÃo
: Whoatheresonny

To decode percent-encoded strings:

#+begin_src bash
function urldecode {
    # urldecode <string>

    # Replace all pluses with spaces since it usually represents it in URLs.
    local url_encoded="${1//+/ }"

    # Replace all percent sign with '\x' and print the message with byte interpretation.
    printf '%b\n' "${url_encoded//%/\\x}"
}

urldecode "El%20Doggo"
urldecode "%e9%bd%8b%e8%97%a4%e3%83%bbB%e3%83%bb%e6%a5%b5%e3%83%bb%e5%b0%86%e5%97%a3"
#+end_src

#+results:
: El Doggo
: 齋藤・ B ・極・将嗣