组合类

What Is the Canonical Combining Class?

The Canonical Combining Class (CCC) is an integer property (range 0–240) assigned to every Unicode character. It specifies how combining marks—characters that attach to a preceding base character—are reordered relative to one another during Unicode Normalization. Most base characters and non-combining characters have CCC = 0 (Non-combining). Combining diacritical marks carry non-zero values that determine their stacking order.

The core rule is the Canonical Ordering Algorithm: when two adjacent combining marks both have non-zero CCC values, the one with the lower value is placed closer to the base character in the normalized form. Two marks with equal non-zero CCC values are considered blocked and their relative order is preserved.

CCC in Practice

Consider the letter a with two diacritics: a cedilla (CCC=202) and an ogonek (CCC=202 as well). Because they share the same CCC, their order is kept stable. But an above-combining mark like combining breve (CCC=228) and a below-combining mark like combining macron below (CCC=220) would sort by their values during normalization, placing the CCC=220 mark before the CCC=228 mark in NFD.

import unicodedata

marks = [
    ("\u0300", "COMBINING GRAVE ACCENT"),        # CCC=230
    ("\u0327", "COMBINING CEDILLA"),              # CCC=202
    ("\u0328", "COMBINING OGONEK"),               # CCC=202
    ("\u0331", "COMBINING MACRON BELOW"),         # CCC=220
    ("\u0952", "DEVANAGARI STRESS SIGN ANUDATTA"),# CCC=220
]

for char, name in marks:
    ccc = unicodedata.combining(char)
    print(f"  CCC={ccc:3}  {name}")

# CCC=230  COMBINING GRAVE ACCENT
# CCC=202  COMBINING CEDILLA
# CCC=202  COMBINING OGONEK
# CCC=220  COMBINING MACRON BELOW
# CCC=220  DEVANAGARI STRESS SIGN ANUDATTA

# Normalization puts the sequence into canonical order:
text = "a\u0328\u0300"   # a + ogonek (CCC=202) + grave (CCC=230)
nfd = unicodedata.normalize("NFD", text)
# NFD preserves order here because 202 < 230, ogonek stays first
print([f"U+{ord(c):04X}" for c in nfd])
# ['U+0061', 'U+0328', 'U+0300']

Named CCC Values

A few CCC values have names defined in the standard: 0 (Not_Reordered), 1 (Overlay), 6 (Han_Reading), 7 (Nukta), 8 (Kana_Voicing), 9 (Virama), and 10 (CCC10) through 199 (CCC199) for specific positioning classes. Values 200–240 are used for particular combining categories such as Below (CCC=220), Above (CCC=230), and Double_Below (CCC=233).

Quick Facts