保留码位

What is a Reserved Code Point?

A reserved code point is a position in the Unicode code space that has not yet been assigned to any character and is not permanently designated for a specific purpose (like noncharacters or private use). The Unicode Consortium holds these positions in reserve for potential future character assignments. As new scripts, symbols, and characters are added in future Unicode versions, they are taken from the pool of reserved code points.

Reserved code points are distinct from: - Unassigned code points: Often used interchangeably with "reserved," but technically "unassigned" means not yet having a character assignment, while "reserved" may imply more deliberate designation - Noncharacters: 66 code points permanently reserved and never to be assigned characters - Private Use Area: Permanently designated for user-defined characters

Current State

As of Unicode 16.0 (154,998 assigned characters), approximately 819,000 code points are unassigned — a vast majority of the 1,114,112 total code space. The Unicode Consortium has far more space than it currently needs:

Total code space:       1,114,112
Assigned characters:      154,998  (~13.9%)
Private Use Area:         137,468  (~12.4%)
Surrogates:                 2,048  ( ~0.2%)
Noncharacters:                 66  ( ~0.01%)
Available (unassigned):  ~819,000  (~73.5%)

Where Reserved Code Points Appear

Reserved code points are scattered throughout the code space, not concentrated in one region. Some patterns:

• Gaps within blocks: A block may have some code points assigned and others reserved (e.g., the Greek block has specific reserved positions where uncommon letters were not initially added)
• Entire sub-ranges: Planes 4–13 (U+40000–U+DFFFF) are entirely unassigned
• Within the BMP: Scattered positions within named blocks

Handling Reserved Code Points

Applications should treat reserved code points gracefully:

import unicodedata

def classify_code_point(cp: int) -> str:
    char = chr(cp)
    category = unicodedata.category(char)
    # Cn = Unassigned (reserved/not yet assigned)
    if category == "Cn":
        return "unassigned/reserved"
    elif category == "Co":
        return "private use"
    elif category == "Cs":
        return "surrogate"
    else:
        return f"assigned ({category})"

print(classify_code_point(0x0041))    # assigned (Lu)
print(classify_code_point(0xE001))    # private use
print(classify_code_point(0xD800))    # surrogate
print(classify_code_point(0x0378))    # unassigned/reserved

Stability Guarantee

A core Unicode stability policy states that reserved code points may become assigned in future versions, but: - An assigned code point is never unassigned - A code point is never reassigned to a different character - The properties of reserved code points may change when they are assigned

This means software written today that skips or rejects reserved code points may need updating when those points are assigned in a future Unicode version.

The U+0378 Example

U+0378 is an example of a reserved code point within the Greek block (U+0370–U+03FF). The Greek block contains letters and symbols, but U+0378 and U+0379 have no assigned characters. They were skipped in the original Greek assignments and remain reserved pending any future need.

Quick Facts