Shape Symbols

Filled and hollow are two different characters

In a Unicode name, BLACK means filled and WHITE means hollow, and the two are encoded separately: ■ U+25A0 BLACK SQUARE beside □ U+25A1 WHITE SQUARE, ● U+25CF BLACK CIRCLE beside ○ U+25CB WHITE CIRCLE, ◆ U+25C6 BLACK DIAMOND beside ◇ U+25C7 WHITE DIAMOND. No style setting turns one into the other. To change the fill, change the character.

The pairing runs out at the edge of the block. The media-control triangles from ⏩ to ⏺ U+23FA BLACK CIRCLE FOR RECORD are all BLACK, and this set holds no hollow counterpart for any of them.

Some of the plainest shapes are maths operators

▷ U+25B7 WHITE RIGHT-POINTING TRIANGLE and ◁ U+25C1 WHITE LEFT-POINTING TRIANGLE are general category Sm, mathematical symbol, while their filled partners ▶ and ◀ are So. The medium squares do the same thing: ◻ U+25FB WHITE MEDIUM SQUARE and ◼ U+25FC BLACK MEDIUM SQUARE are both Sm, though nothing in the name says so.

Twenty-nine characters here are Sm, including the whole run from ⊏ U+228F SQUARE IMAGE OF to ⊔ U+2294 SQUARE CUP. Those are relations built on a square outline rather than boxes to draw a border with, and a maths layout engine spaces every one of them as an operator.

The size word does not predict the width

▪ U+25AA BLACK SMALL SQUARE is its own codepoint next to U+25A0, and each pointing triangle has a SMALL twin. Through the main geometric run the pattern is tidy: the plain square, circle, diamond and pointing triangle are East Asian Width Ambiguous, one cell in a Western context and two in a CJK one, while every SMALL square and triangle beside them is Neutral.

Then it inverts. ◻ WHITE MEDIUM SQUARE is Neutral, but ◽ U+25FD WHITE MEDIUM SMALL SQUARE is Wide — the medium small square asks for two cells where the medium one asks for one. ⚪ U+26AA MEDIUM WHITE CIRCLE is Wide and ⚬ U+26AC MEDIUM SMALL WHITE CIRCLE is Neutral. Thirty of the 140 are Ambiguous and eight are Wide, and the eight are not the eight you would guess from the names.

Part-filled circles, and one alignment trap in them

The circles from ◐ to ◗ and from ◴ to ◷ are filled by halves and quadrants, which makes them usable as progress dots. The widths are not uniform across the family: ◐ U+25D0 CIRCLE WITH LEFT HALF BLACK and ◑ are Ambiguous, while ◒ U+25D2 CIRCLE WITH LOWER HALF BLACK and the quadrant forms are Neutral, so a row mixed from both halves of the family can change width partway along.

◌ U+25CC DOTTED CIRCLE is the odd one: a renderer may use it as the placeholder base on which to draw a combining mark that has no letter to sit on.

One of them is not a shape

ׯ U+05EF HEBREW YOD TRIANGLE is category Lo, a letter, and its bidirectional class is R. Pasting it into a line of Latin text opens a right-to-left run, and the bidirectional algorithm can reorder the punctuation around it, so a bracket or full stop beside it may appear at the other end of the phrase. It is in a shape set because of the word in its name.

⎡ to ⎦ are not shapes to paste on their own either: they are the corner and extension pieces for assembling one tall square bracket down the side of several lines, and singly they look like fragments, because that is what they are.

Every set is indexed on the symbol reference. For frames and rules built from connecting pieces rather than single shapes, see line and box drawing symbols. Nothing in this set is a star inside a circle; that is U+272A CIRCLED WHITE STAR, on star symbols.