Waste less randomness when generating random sequences of the first 64 characters of your alphabet.
A lot of modules out there just modulo(random, alphabet.length),
which wastes a lot of randomness to duplicate choices.
Especially in cases where the to-be-used prefix of your alphabet is
at most 36 characters long, or (in this module's case) is a square number,
you can generate digits more easily and less wasteful using JavaScript
numbers' .toString() method, and keep unused digits for later.
On my development machine, one random number usually produces 8 to 9 random characters, so this module drains my randomness pool a low slower than a naive implementation.
For more details, see the test/ directory.
var rand64 = require('rand64-pmb'), urlSafe;
function demo(func, args) { console.log(repeat(4, func, args)); }
demo(rand64, [ /* no arguments */ ]);
// example output: [ 'HqWLD2En', 'qCqip+1T', 'em/kSMNZ', 'KkkPsb9U' ]
demo(rand64, [ 3 ]);
// ex. [ '+ip', '9qx', 'gOz', 'Fb0' ]
urlSafe = rand64.custom({ charset: rand64.charsets.url, len: 12 });
demo(urlSafe, []);
// ex. [ '9qTssF8_qUIY', 'LU-zHa_0y1rZ', 'UVCFa2d2ljPy', 'q9QF9DDNjW-7' ]
// If the alphabet is too short, the base64 alphabet is appended, so these
// parameters are effectively the same as in the previous exmaple:
urlSafe = rand64.custom({ charset: '-_' });
demo(urlSafe, [ 12 ]);
// ex. [ 'isKpZpAaNM-I', 'h-PqEoSa_uki', '0-lmKvH1WFt-', 'Qqkwley8y1KK' ]ISC