1 # This file is part of NIT ( http://www.nitlanguage.org ).
3 # This file is free software, which comes along with NIT. This software is
4 # distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
5 # without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
6 # PARTICULAR PURPOSE. You can modify it is you want, provided this header
7 # is kept unaltered, and a notification of the changes is added.
8 # You are allowed to redistribute it and sell it, alone or is a part of
11 # Native structures for text and bytes
23 #include <libkern/OSByteOrder.h>
24 #define be32toh(x) OSSwapBigToHostInt32(x)
27 #define be32toh(val) _byteswap_ulong(val)
31 #define be32toh(val) betoh32(val)
39 # Gives the length of the UTF-8 char starting with `self`
41 if self & 0b1000_0000 == 0 then
43 else if self & 0b1110_0000 == 0b1100_0000 then
45 else if self & 0b1111_0000 == 0b1110_0000 then
47 else if self & 0b1111_1000 == 0b1111_0000 then
54 # Is `self` a valid UTF-8 sequence start ?
57 # assert 0.is_valid_utf8_start
58 # assert 0xC0.is_valid_utf8_start
59 # assert 0xE0.is_valid_utf8_start
60 # assert 0xF0.is_valid_utf8_start
62 fun is_valid_utf8_start
: Bool do
63 if self & 0x80 == 0 then return true
64 if self & 0b1110_0000 == 0b1100_0000 then return true
65 if self & 0b1111_0000 == 0b1110_0000 then return true
66 if self & 0b1111_1000 == 0b1111_0000 then return true
72 # Returns the code_point from a utf16 surrogate pair
74 # assert 0xD83DDE02u32.from_utf16_surr == 0x1F602u32
75 fun from_utf16_surr
: UInt32 do
76 var hi
= (self & 0xFFFF0000u
32) >> 16
77 var lo
= self & 0xFFFFu
32
79 cp
+= (hi
- 0xD800u
32) << 10
85 # The character which code point (unicode-wise) is `self`
87 # assert 65u32.code_point == 'A'
88 # assert 10u32.code_point == '\n'
89 # assert 0x220Bu32.code_point == '∋'
90 fun code_point
: Char `{ return self; `}
95 # Used as underlying implementation for `String` and some other `Text`.
96 extern class CString `{ char* `}
97 # Create a new `CString` with the capacity for `length` characters
98 new(length
: Int) is intern
100 # Get a char* starting at `index`.
102 # WARNING: Unsafe for extern code, use only for temporary
103 # pointer manipulation purposes (e.g. write to file or such)
104 fun fast_cstring
(index
: Int): CString is intern
106 # Get char at `index`.
107 fun [](index
: Int): Int is intern
109 # Set char `item` at index.
110 fun []=(index
: Int, item
: Int) is intern
112 # Copy `self` to `dest`.
113 fun copy_to
(dest
: CString, length
: Int, from
: Int, to
: Int) is intern
115 redef fun ==(o
) is intern do return is_same_instance
(o
)
117 redef fun !=(o
) is intern do return not is_same_instance
(o
)
119 # Position of the first nul character.
120 fun cstring_length
: Int
123 while self[l
] != 0 do l
+= 1
127 # Parse `self` as an Int.
128 fun atoi
: Int is intern
130 # Parse `self` as a Float.
131 fun atof
: Float `{ return atof(self); `}
133 # Gets the UTF-8 char at index `pos
`
135 # Index is expressed in Unicode chars
138 # assert "かきく".as(FlatString).items.char_at(0) == 'か'
141 # If the char at position pos is an invalid Unicode char,
142 # the Unicode replacement character � (0xFFFD) will be used.
145 # assert "かきく".as(FlatString).items.char_at(1) == '�'
147 fun char_at(pos: Int): Char do
149 if c & 0x80 == 0 then return c.code_point
150 var b = fetch_4_hchars(pos)
152 if b & 0xC00000u32 != 0x800000u32 then return 0xFFFD.code_point
153 if b & 0xE0000000u32 == 0xC0000000u32 then
154 ret |= (b & 0x1F000000u32) >> 18
155 ret |= (b & 0x3F0000u32) >> 16
156 return ret.code_point
158 if not b & 0xC000u32 == 0x8000u32 then return 0xFFFD.code_point
159 if b & 0xF0000000u32 == 0xE0000000u32 then
160 ret |= (b & 0xF000000u32) >> 12
161 ret |= (b & 0x3F0000u32) >> 10
162 ret |= (b & 0x3F00u32) >> 8
163 return ret.code_point
165 if not b & 0xC0u32 == 0x80u32 then return 0xFFFD.code_point
166 if b & 0xF8000000u32 == 0xF0000000u32 then
167 ret |= (b & 0x7000000u32) >> 6
168 ret |= (b & 0x3F0000u32) >> 4
169 ret |= (b & 0x3F00u32) >> 2
171 return ret.code_point
173 return 0xFFFD.code_point
176 # Gets the byte index of char at position `n
` in UTF-8 String
177 fun char_to_byte_index(n: Int): Int do return char_to_byte_index_cached(n, 0, 0)
179 # Gets the length of the character at position `pos
` (1 if invalid sequence)
180 fun length_of_char_at(pos: Int): Int do
182 if c & 0x80 == 0x00 then
184 else if c & 0xE0 == 0xC0 and self[pos + 1] & 0xC0 == 0x80 then
186 else if c & 0xF0 == 0xE0 and self[pos + 1] & 0xC0 == 0x80 and self[pos + 2] & 0xC0 == 0x80 then
188 else if c & 0xF8 == 0xF0 and self[pos + 1] & 0xC0 == 0x80 and self[pos + 2] & 0xC0 == 0x80 and self[pos + 3] & 0xC0 == 0x80 then
195 # Gets the byte index of char at position `n
` in UTF-8 String
197 # `char_from
` and `byte_from
` are cached values to seek from.
199 # NOTE: char_from and byte_from are not guaranteed to be valid cache values
200 # It it up to the client to ensure the validity of the information
201 fun char_to_byte_index_cached(n, char_from, byte_from: Int): Int do
209 var i = fetch_4_chars(ns_i)
210 if i & 0x80808080u32 != 0u32 then break
215 if dist == 0 then break
216 ns_i += length_of_char_at(ns_i)
223 var i = fetch_4_chars(ns_i - 4)
224 if i & 0x80808080u32 != 0u32 then break
229 if dist == 0 then break
230 ns_i = find_beginning_of_char_at(ns_i - 1)
238 # Gets the char index of byte at position `n
` in a UTF-8 String
240 # `char_from
` and `byte_from
` are cached values to seek from.
242 # NOTE: char_from and byte_from are not guaranteed to be valid cache values
243 # It it up to the client to ensure the validity of the information
244 fun byte_to_char_index_cached(n, char_from, byte_from: Int): Int do
249 ns_i += length_of_char_at(ns_i)
254 ns_i = find_beginning_of_char_at(ns_i - 1)
261 # Returns the beginning position of the char at position `pos
`
263 # If the char is invalid UTF-8, `pos
` is returned as-is
266 # assert "abc".items.find_beginning_of_char_at(2) == 2
267 # assert "か".items.find_beginning_of_char_at(1) == 0
268 # assert [0x41, 233].to_s.items.find_beginning_of_char_at(1) == 1
270 fun find_beginning_of_char_at(pos: Int): Int do
273 if c & 0x80 == 0x00 then return pos
274 while c & 0xC0 == 0x80 do
279 if length_of_char_at(stpos) >= (endpos - stpos + 1) then return pos
283 # Number of UTF-8 characters in `self` starting at `from
`, for a length of `byte_length
`
284 fun utf8_length(from, byte_length: Int): Int is intern do
287 while byte_length > 0 do
288 while byte_length >= 4 do
289 var i = fetch_4_chars(st)
290 if i & 0x80808080u32 != 0u32 then break
295 if byte_length == 0 then break
296 var cln = length_of_char_at(st)
304 # Fetch 4 chars in `self` at `pos
`
305 fun fetch_4_chars(pos: Int): UInt32 is intern `{ return *((uint32_t*)(self+pos)); `}
307 # Fetch 4 chars in `self` at `pos`
308 fun fetch_4_hchars
(pos
: Int): UInt32 is intern `{ return (uint32_t)be32toh(*((uint32_t*)(self+pos))); `}
310 # Right shifts `len
` bytes of `self` from `sh
` bytes starting at position `pos
`
311 fun rshift(sh, len, pos: Int) do
312 copy_to(self, len, pos, pos + sh)
315 # Left shifts `len
` bytes of `self` from `sh
` bytes starting at position `pos
`
316 fun lshift(sh, len, pos: Int) do
317 copy_to(self, len, pos, pos - sh)
320 # Sets the contents of `self` to `value
` for `len
` bytes
321 fun memset(value, len: Int) `{
323 memset
(self, value
, len
);