Blank space to copy and paste

Apps tidy what you type before they save it. Most cut spaces from the start and end of a field, and many squeeze a run of spaces down to one; Instagram, for example, collapses extra spaces and trims blank lines in bios and captions. All of that tidying looks for characters that Unicode classes as white space: the ordinary space, the no-break space, the em and en spaces and every other space in the second set.

The Braille blank and the Hangul fillers are not white space. U+2800 is the Braille cell with none of its eight dots raised, and the fillers are Korean letters that hold the place of a missing part of a syllable. Code that strips spaces leaves them alone, which is why they can keep an empty line or an empty-looking name in place.

What we checked: the Unicode 16.0 properties of every character on this page, and what JavaScript’s built-in trim(), a function many websites use to tidy form fields, does to each one. It removed every space in the width list plus the zero-width no-break space (U+FEFF), and kept the Braille blank, the Hangul fillers and the zero-width characters in the third set. We have not tested each character in every app. Apps run their own clean-up on top of that, and an update can change what survives.

Tip: To space out an Instagram caption, write the paragraphs first, then put a single Braille blank on its own line between each pair. Look at the preview before you post.

Widths are measured in ems. An em is the size of the font, so in 16-pixel text an em space is 16 pixels wide and an en space 8. Most of the others are fractions of that, and a few are tied to a character: the figure space matches a digit, which is how columns of numbers line up, and the punctuation space matches a full stop. The ideographic space is as wide as one Chinese or Japanese character.

Those are the widths Unicode intends. The font decides the real one, and an app may draw an unfamiliar space as an ordinary one. The quads are leftovers from metal type: Unicode treats the em quad as the same character as the em space, and the en quad as the en space, so tools that normalise text can quietly swap one for the other.

Each space in the second set is wrapped in two zero-width spaces (U+200B), one on each side. They take no room, but they keep a standard trim from removing the space when it lands at the start or end of a line. The codes shown under each row list all three characters.

• The zero-width space (U+200B) takes no room at all. It is the right pick for a hidden break point and the wrong one when you want a gap people can see.
• The zero-width joiner (U+200D) is left out on purpose. It glues emoji together: 👩, U+200D and 💻 in a row become a single emoji, 👩‍💻. Pasted between two emoji by mistake, it can merge them.
• Unicode marks the Hangul fillers as default-ignorable, which allows a phone to draw them with no width at all. A gap made of them on one phone can vanish on another; the Braille blank is steadier.
• Direction marks (U+200E and U+200F) are invisible, but in text that mixes right-to-left and left-to-right scripts they can move a full stop or a bracket to the other side.
• Blanks still count. Each one uses up a character of a 150-character Instagram bio, and every wrapped space in the second set counts as three.
• Screen readers treat blanks unevenly: some skip them and some announce them, so a column of twenty can mean twenty pauses for someone listening.