Text and Unicode

ransu.string(12); // 'k3Xq7bWm1zPd'
ransu.string(8, ransu.alphabets.base58);
ransu.hex(32);
ransu.shuffleString("ransu");

Uniform for any alphabet size, not just powers of two. Sizes that are a power of two take a bit-slicing fast path; the rest use rejection.

Alphabets

lower, upper, letters, digits, alphanumeric, hex, hexUpper, binary, octal, base32, base32hex, base58, base62, base64url, unambiguous, ascii.

unambiguous drops the characters people misread when transcribing a code by hand — 0/O, 1/l/I.

Emoji and astral characters

length counts characters, never UTF-16 units, and a surrogate pair is never split:

ransu.string(5, "🍎🍊🍇"); // '🍇🍎🍇🍊🍎'
ransu.string(5, ransu.graphemes("👍🏽🎉")); // skin tones stay intact

graphemes() splits on user-perceived characters using Intl.Segmenter, so emoji sequences, combining marks and flags survive.

Full Unicode

ransu.char({ blocks: "emoji" }); // '🚀'
ransu.chars(10, { blocks: "kana" });
ransu.codePoint({ blocks: ["greek", "cyrillic"] });
ransu.chars(20, { ranges: [[0x4e00, 0x9fff]] });

Option	Effect
ranges	Explicit inclusive [first, last] pairs
blocks	One or more names from unicodeRanges
bmpOnly	Keep every character to a single UTF-16 unit
allowControl / allowPrivateUse / allowNoncharacters	Opt back in
filter	Arbitrary per-code-point rejection

Blocks: ascii, latin, latin1, latinExtended, greek, cyrillic, hebrew, arabic, devanagari, thai, hiragana, katakana, kana, kanji, hangul, cjk, punctuation, currency, arrows, math, box, blockElements, geometric, braille, emoji, symbols, printable, bmp, all.

What can never come out

Surrogates are not scalar values, so they are excluded with no way to opt in — otherwise you could build a malformed string. Controls, private-use areas and noncharacters are excluded by default and can be re-enabled.

all includes a great many unassigned code points, which render as tofu. That is why printable is the default. Deciding which code points are assigned would need a large Unicode table, which ransu deliberately does not ship.

Reusing options

Building the range set costs a normalisation pass. When the same options are used repeatedly, build it once:

import { CodePointSet } from "ransu/unicode";

const kana = new CodePointSet({ blocks: "kana" });
ransu.char(kana);