码空间

What is the Unicode Code Space?

The code space is the complete range of integer values available for Unicode code points: U+0000 through U+10FFFF, totaling exactly 1,114,112 positions. Think of it as the full address space of the Unicode standard — every character, symbol, and abstract entity that could ever be assigned a Unicode code point must fit within this range.

The code space is not fully occupied. As of Unicode 16.0, approximately 154,998 of these positions are assigned to characters. The remaining ~959,000 positions are either unassigned (available for future characters), reserved, or permanently designated as noncharacters or private use.

The Numbers

U+0000      →  U+10FFFF
0           →  1,114,111

Total code points:  1,114,112
= 17 planes × 65,536 points per plane
= 17 × 0x10000
= 0x110000 (hex)
= 2^21 - 2^16 = ? (not a round power of 2 — see below)

The value 1,114,112 is not a round power of two. It equals 17 × 65,536, which results from the deliberate choice to have 17 planes of 65,536 each. The upper limit of U+10FFFF was set to match the maximum value expressible by UTF-16 surrogate pairs, making UTF-16 the natural encoding boundary.

Why U+10FFFF as the Upper Limit?

UTF-16 uses surrogate pairs to encode supplementary characters. Each surrogate half occupies a 10-bit value, so a pair provides 20 bits of additional addressing: 2^20 = 1,048,576 supplementary code points. Adding the 65,536 BMP positions yields exactly 1,114,112 — the size of the code space. The U+10FFFF upper limit was thus engineered to keep UTF-16 and the code space in perfect alignment.

Code Space Breakdown

Code Space vs Character Repertoire

The code space defines positions; the character repertoire is the subset of those positions that are currently assigned. Unicode's stability policies ensure that once a character is assigned to a position, that assignment is permanent — no code point is ever recycled or reassigned to a different character.

# Python: check if a code point is within the Unicode code space
def is_valid_code_point(cp: int) -> bool:
    return 0x0000 <= cp <= 0x10FFFF

# Check for surrogate range (not real characters)
def is_surrogate(cp: int) -> bool:
    return 0xD800 <= cp <= 0xDFFF

# Check for noncharacter
def is_noncharacter(cp: int) -> bool:
    last_two = cp & 0xFFFF
    return last_two in (0xFFFE, 0xFFFF) or 0xFDD0 <= cp <= 0xFDEF

Historical Context

The original Unicode proposal (1988) envisioned a 16-bit code space of 65,536 characters. Engineers believed this would be sufficient for all world languages. By Unicode 2.0 (1996) it was clear the CJK ideograph extensions alone would exceed this limit. The standard was extended to 21 bits (the current code space), but the legacy 16-bit assumption is why surrogate pairs exist in UTF-16 and why JavaScript's String.length counts UTF-16 code units rather than Unicode code points.

Quick Facts