dymokomi/luce-regex
Source of release 0.6.1
0.6.1 / src / unicode / classify.lucb
#==============================================================================================
#
# classify - Code point categorizing: the Latin-1 class bits and the non-ASCII spaces
#
# DESCRIPTION:
# Port of QuickJS libunicode.c lines 1825-1917: lre_ctype_bits, the table behind the
# inline byte predicates of libunicode.h, and lre_is_space_non_ascii.
#
#==============================================================================================
# mark: Latin-1 classes ========================================================================
# lre codepoint categorizing functions
# The C file's one-letter macros S, D, X, U, L, _ and d.
let ct_s: u8 = unicode_c_space
let ct_d: u8 = unicode_c_digit
let ct_x: u8 = unicode_c_xdigit
let ct_u: u8 = unicode_c_upper
let ct_l: u8 = unicode_c_lower
let ct_under: u8 = unicode_c_under
let ct_dollar: u8 = unicode_c_dollar
## The `unicode_c_*` class bits of each Latin-1 byte.
pub let lre_ctype_bits: u8[256] = [
0, 0, 0, 0, 0, 0, 0, 0,
0, ct_s, ct_s, ct_s, ct_s, ct_s, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
ct_s, 0, 0, 0, ct_dollar, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
ct_x | ct_d, ct_x | ct_d, ct_x | ct_d, ct_x | ct_d, ct_x | ct_d, ct_x | ct_d, ct_x | ct_d, ct_x | ct_d,
ct_x | ct_d, ct_x | ct_d, 0, 0, 0, 0, 0, 0,
0, ct_x | ct_u, ct_x | ct_u, ct_x | ct_u, ct_x | ct_u, ct_x | ct_u, ct_x | ct_u, ct_u,
ct_u, ct_u, ct_u, ct_u, ct_u, ct_u, ct_u, ct_u,
ct_u, ct_u, ct_u, ct_u, ct_u, ct_u, ct_u, ct_u,
ct_u, ct_u, ct_u, 0, 0, 0, 0, ct_under,
0, ct_x | ct_l, ct_x | ct_l, ct_x | ct_l, ct_x | ct_l, ct_x | ct_l, ct_x | ct_l, ct_l,
ct_l, ct_l, ct_l, ct_l, ct_l, ct_l, ct_l, ct_l,
ct_l, ct_l, ct_l, ct_l, ct_l, ct_l, ct_l, ct_l,
ct_l, ct_l, ct_l, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
ct_s, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
]
# mark: Spaces =================================================================================
# code point ranges for Zs,Zl or Zp property
let char_range_s: u16[21] = [
10,
0x0009, 0x000D + 1,
0x0020, 0x0020 + 1,
0x00A0, 0x00A0 + 1,
0x1680, 0x1680 + 1,
0x2000, 0x200A + 1,
# 2028;LINE SEPARATOR;Zl;0;WS;;;;;N;;;;;
# 2029;PARAGRAPH SEPARATOR;Zp;0;B;;;;;N;;;;;
0x2028, 0x2029 + 1,
0x202F, 0x202F + 1,
0x205F, 0x205F + 1,
0x3000, 0x3000 + 1,
# FEFF;ZERO WIDTH NO-BREAK SPACE;Cf;0;BN;;;;;N;BYTE ORDER MARK;;;;
0xFEFF, 0xFEFF + 1,
]
## True if the code point `c` (at least 256) is a JavaScript white space or line terminator.
pub func lre_is_space_non_ascii(c: u32) -> bool:
let n = char_range_s.length
var i: usize = 5
while i < n:
let low: u32 = char_range_s[i]
let high: u32 = char_range_s[i + 1]
if c < low:
return false
if c < high:
return true
i += 2
return false