How to type Roman Numeral One

Ⅰ lives in the Number Forms block (U+2150-U+218F), fractions and Roman numerals as single characters. Within Unicode it is classified as Nl, a letter-like number. Roman numerals are the familiar example. Its bidirectional class makes it left-to-right.

Its lowercase counterpart is ⅰ (U+2170). They are separate codepoints, so a byte comparison treats them as different characters even though a case-insensitive search does not.

What happens to Ⅰ in a search box

Ⅰ is a compatibility character: Unicode says it can be rewritten as I, one character. Anything that normalises text to NFKC before comparing it does exactly that, which is why Ⅰ pasted into a search field, a URL slug or a username can come back as I and match text that never contained Ⅰ at all. NFC, the form this site uses, leaves it alone. Unicode also records its numeric value as 1, so software that asks a character what number it is gets 1 rather than an error.

What Ⅰ gets confused with

Unicode's own confusables table, the one used to catch lookalike domain names, groups Ⅰ with 1 (U+0031, Digit One), I (U+0049, Latin Capital Letter I), l (U+006C, Latin Small Letter L), | (U+007C, Vertical Bar) and Ι (U+0399, Greek Capital Letter Iota). They are separate characters that render close enough to be mistaken for one another, so a search or a string comparison will separate them where your eye will not. If Roman Numeral One is not matching where you expect it to, one of these is the usual reason.

How Roman Numeral One behaves in software

Ⅰ is East Asian Ambiguous: one column in a Western locale and two in a CJK one. That is why a table that lines up on your machine can break on somebody else's, with no change to the text at all. It is a left-to-right character, so it holds that direction even inside a right-to-left paragraph.

Every way to write Roman Numeral One

If Ⅰ arrives as â… 

Its three UTF-8 bytes E2 85 A0 are being read one at a time as Windows-1252. The text is undamaged; the reader has to be told it is UTF-8.