Text

String Inspector

Every way of counting a string at once — UTF-8 bytes, UTF-16 code units, code points and grapheme clusters — plus the code point behind each character and a flag on anything invisible, combining or deceptive.

Text

Try

UTF-8 bytes

What goes over the wire

UTF-16 units

JS .length

Code points

[...str].length

Graphemes

What a reader counts

The four counts disagree, which is the usual source of truncated emoji and off-by-one column limits: a database VARCHAR(n) counts one of them, your validator probably counts another.

Findings

Astral plane characters

Code points above U+FFFF take two UTF-16 units, so slicing by JS string index can cut one in half.

Characters6 code points

0063

0061

0066

00E9

␣

0020

👋

1F44B

Code point tableBytes and units each character actually occupies

#	Char	Code point	Name	UTF-8	UTF-16	Escape
0	c	U+0063	ASCII	63	0063	\u0063
1	a	U+0061	ASCII	61	0061	\u0061
2	f	U+0066	ASCII	66	0066	\u0066
3	é	U+00E9	Latin-1 supplement	C3 A9	00E9	\u00E9
4	␣	U+0020	Whitespace	20	0020	\u0020
5	👋	U+1F44B	Supplementary plane	F0 9F 91 8B	D83D DC4B	\u{1F44B}

NormalisationSame text under each Unicode normalisation form

Form	Code points	UTF-8 bytes	Changes input	Result
NFC	6	10	no	café 👋
NFD	7	11	yes	café 👋
NFKC	6	10	no	café 👋
NFKD	7	11	yes	café 👋

UTF-8 hexContains multi-byte sequences

63 61 66 c3 a9 20 f0 9f 91 8b