UrlEncode - ObjectVision/GeoDMS GitHub Wiki

String functions UrlEncode

The UrlEncode function encodes strings for safe use in URLs by escaping special characters.

UrlEncode(strings: E->String) -> E->String

Converts strings to URL-safe format: a space becomes +, and every other byte outside the unreserved set becomes a percent-encoded %XX sequence with UPPERCASE hex digits (& becomes %26, < becomes %3C).

UrlEncode is the exact inverse of UrlDecode: UrlDecode(UrlEncode(x)) returns x for every string, non-ASCII content included.

Characters that remain unencoded - the unreserved set of RFC2396:

• Letters (A-Z, a-z) and digits (0-9)
• The mark characters -, _, ., !, ~, *, ', (, )

Everything else is encoded, in particular:

• A space becomes + (not %20), which is what UrlDecode reads back as a space
• A literal + becomes %2B, for that same reason
• Control characters and the reserved URL characters #, $, &, ,, /, :, ;, =, ?, @, [, ]
• Every non-ASCII byte, as one %XX per byte; UTF-8 text is therefore encoded as its UTF-8 byte sequence

Time complexity: O(n × L) where n is the number of strings and L is the average string length.

Encoded strings are typically longer than input strings (each special character becomes 3 characters).

unit<uint32> QueryParams: nrofrows = 3;
attribute<String> values (QueryParams) := union_data(QueryParams,
    'hello world',
    'price=100&tax=20',
    'name=Müller'
);

attribute<String> encoded (QueryParams) := UrlEncode(values);
// encoded = {'hello+world', 'price%3D100%26tax%3D20', 'name%3DM%C3%BCller'}

// Building URL query string
parameter<String> base_url := 'https://api.example.com/search?q=';
attribute<String> full_url (QueryParams) := base_url + encoded;

• UrlDecode - reverse operation
• HtmlEncode - for HTML content
• String functions

20.13.0