v/vlib/builtin/utf8.v

209 lines
5.7 KiB
V

// Copyright (c) 2019-2024 Alexander Medvednikov. All rights reserved.
// Use of this source code is governed by an MIT license
// that can be found in the LICENSE file.
module builtin
// utf8_char_len returns the length in bytes of a UTF-8 encoded codepoint that starts with the byte `b`.
pub fn utf8_char_len(b u8) int {
return ((0xe5000000 >> ((b >> 3) & 0x1e)) & 3) + 1
}
// Convert utf32 to utf8
// utf32 == Codepoint
pub fn utf32_to_str(code u32) string {
unsafe {
mut buffer := malloc_noscan(5)
res := utf32_to_str_no_malloc(code, mut buffer)
if res.len == 0 {
// the buffer was not used at all
free(buffer)
}
return res
}
}
@[manualfree; unsafe]
pub fn utf32_to_str_no_malloc(code u32, mut buf &u8) string {
unsafe {
len := utf32_decode_to_buffer(code, mut buf)
if len == 0 {
return ''
}
buf[len] = 0
return tos(buf, len)
}
}
@[manualfree; unsafe]
pub fn utf32_decode_to_buffer(code u32, mut buf &u8) int {
unsafe {
icode := int(code) // Prevents doing casts everywhere
mut buffer := &u8(buf)
if icode <= 127 {
// 0x7F
buffer[0] = u8(icode)
return 1
} else if icode <= 2047 {
// 0x7FF
buffer[0] = 192 | u8(icode >> 6) // 0xC0 - 110xxxxx
buffer[1] = 128 | u8(icode & 63) // 0x80 - 0x3F - 10xxxxxx
return 2
} else if icode <= 65535 {
// 0xFFFF
buffer[0] = 224 | u8(icode >> 12) // 0xE0 - 1110xxxx
buffer[1] = 128 | (u8(icode >> 6) & 63) // 0x80 - 0x3F - 10xxxxxx
buffer[2] = 128 | u8(icode & 63) // 0x80 - 0x3F - 10xxxxxx
return 3
}
// 0x10FFFF
else if icode <= 1114111 {
buffer[0] = 240 | u8(icode >> 18) // 0xF0 - 11110xxx
buffer[1] = 128 | (u8(icode >> 12) & 63) // 0x80 - 0x3F - 10xxxxxx
buffer[2] = 128 | (u8(icode >> 6) & 63) // 0x80 - 0x3F - 10xxxxxx
buffer[3] = 128 | u8(icode & 63) // 0x80 - 0x3F - 10xxxxxx
return 4
}
}
return 0
}
// Convert utf8 to utf32
// the original implementation did not check for
// valid utf8 in the string, and could result in
// values greater than the utf32 spec
// it has been replaced by `utf8_to_utf32` which
// has an option return type.
//
// this function is left for backward compatibility
// it is used in vlib/builtin/string.v,
// and also in vlib/v/gen/c/cgen.v
pub fn (_rune string) utf32_code() int {
if _rune.len > 4 {
return 0
}
return int(impl_utf8_to_utf32(&u8(_rune.str), _rune.len))
}
// convert array of utf8 bytes to single utf32 value
// will error if more than 4 bytes are submitted
pub fn (_bytes []u8) utf8_to_utf32() !rune {
if _bytes.len > 4 {
return error('attempted to decode too many bytes, utf-8 is limited to four bytes maximum')
}
return impl_utf8_to_utf32(&u8(_bytes.data), _bytes.len)
}
@[direct_array_access]
fn impl_utf8_to_utf32(_bytes &u8, _bytes_len int) rune {
if _bytes_len == 0 {
return 0
}
// return ASCII unchanged
if _bytes_len == 1 {
return unsafe { rune(_bytes[0]) }
}
mut b := u8(int(unsafe { _bytes[0] }))
b = b << _bytes_len
mut res := rune(b)
mut shift := 6 - _bytes_len
for i := 1; i < _bytes_len; i++ {
c := rune(unsafe { _bytes[i] })
res = rune(res) << shift
res |= c & 63 // 0x3f
shift = 6
}
return res
}
// Calculate string length for formatting, i.e. number of "characters"
// This is simplified implementation. if you need specification compliant width,
// use utf8.east_asian.display_width.
pub fn utf8_str_visible_length(s string) int {
mut l := 0
mut ul := 1
for i := 0; i < s.len; i += ul {
c := unsafe { s.str[i] }
ul = ((0xe5000000 >> ((unsafe { s.str[i] } >> 3) & 0x1e)) & 3) + 1
if i + ul > s.len { // incomplete UTF-8 sequence
return l
}
l++
// avoid the match if not needed
if ul == 1 {
continue
}
// recognize combining characters and wide characters
match ul {
2 {
r := u64((u16(c) << 8) | unsafe { s.str[i + 1] })
if r >= 0xcc80 && r < 0xcdb0 {
// diacritical marks
l--
}
}
3 {
r := u64((u32(c) << 16) | unsafe { (u32(s.str[i + 1]) << 8) | s.str[i + 2] })
// diacritical marks extended
// diacritical marks supplement
// diacritical marks for symbols
// TODO: remove this workaround for v2's parser
// vfmt off
if (r >= 0xe1aab0 && r <= 0xe1ac7f) ||
(r >= 0xe1b780 && r <= 0xe1b87f) ||
(r >= 0xe28390 && r <= 0xe2847f) ||
(r >= 0xefb8a0 && r <= 0xefb8af) {
// diacritical marks
l--
}
// Hangru
// CJK Unified Ideographics
// Hangru
// CJK
else if (r >= 0xe18480 && r <= 0xe1859f) ||
(r >= 0xe2ba80 && r <= 0xe2bf95) ||
(r >= 0xe38080 && r <= 0xe4b77f) ||
(r >= 0xe4b880 && r <= 0xea807f) ||
(r >= 0xeaa5a0 && r <= 0xeaa79f) ||
(r >= 0xeab080 && r <= 0xed9eaf) ||
(r >= 0xefa480 && r <= 0xefac7f) ||
(r >= 0xefb8b8 && r <= 0xefb9af) {
// half marks
l++
}
// vfmt on
}
4 {
r := u64((u32(c) << 24) | unsafe {
(u32(s.str[i + 1]) << 16) | (u32(s.str[i + 2]) << 8) | s.str[i + 3]
})
// Enclosed Ideographic Supplement
// Emoji
// CJK Unified Ideographs Extension B-G
// TODO: remove this workaround for v2's parser
// vfmt off
if (r >= 0x0f9f8880 && r <= 0xf09f8a8f) ||
(r >= 0xf09f8c80 && r <= 0xf09f9c90) ||
(r >= 0xf09fa490 && r <= 0xf09fa7af) ||
(r >= 0xf0a08080 && r <= 0xf180807f) {
l++
}
// vfmt on
}
else {}
}
}
return l
}
// string_to_ansi_not_null_terminated returns an ANSI version of the string `_str`.
// NOTE: This is most useful for converting a vstring to an ANSI string under Windows.
// NOTE: The ANSI string return is not null-terminated, then you can use `os.write_file_array` write an ANSI file.
pub fn string_to_ansi_not_null_terminated(_str string) []u8 {
wstr := _str.to_wide()
mut ansi := wide_to_ansi(wstr)
if ansi.len > 0 {
unsafe { ansi.len-- } // remove tailing zero
}
return ansi
}