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)
36 # Gives the length of the UTF-8 char starting with `self`
38 if self & 0b1000_0000u
8 == 0u8
then
40 else if self & 0b1110_0000u
8 == 0b1100_0000u
8 then
42 else if self & 0b1111_0000u
8 == 0b1110_0000u
8 then
44 else if self & 0b1111_1000u
8 == 0b1111_0000u
8 then
51 # Is `self` a valid UTF-8 sequence start ?
54 # assert 0u8.is_valid_utf8_start
55 # assert 0xC0u8.is_valid_utf8_start
56 # assert 0xE0u8.is_valid_utf8_start
57 # assert 0xF0u8.is_valid_utf8_start
59 fun is_valid_utf8_start
: Bool do
60 if self & 0x80u
8 == 0u8
then return true
61 if self & 0b1110_0000u
8 == 0b1100_0000u
8 then return true
62 if self & 0b1111_0000u
8 == 0b1110_0000u
8 then return true
63 if self & 0b1111_1000u
8 == 0b1111_0000u
8 then return true
69 # Returns the code_point from a utf16 surrogate pair
71 # assert 0xD83DDE02u32.from_utf16_surr == 0x1F602u32
72 fun from_utf16_surr
: UInt32 do
73 var hi
= (self & 0xFFFF0000u
32) >> 16
74 var lo
= self & 0xFFFFu
32
76 cp
+= (hi
- 0xD800u
32) << 10
82 # The character which code point (unicode-wise) is `self`
84 # assert 65u32.code_point == 'A'
85 # assert 10u32.code_point == '\n'
86 # assert 0x220Bu32.code_point == '∋'
87 fun code_point
: Char `{ return self; `}
92 # Used as underlying implementation for `String` and some other `Text`.
93 extern class CString `{ char* `}
94 # Create a new `CString` with the capacity for `length` characters
95 new(length
: Int) is intern
97 # Get a char* starting at `index`.
99 # WARNING: Unsafe for extern code, use only for temporary
100 # pointer manipulation purposes (e.g. write to file or such)
101 fun fast_cstring
(index
: Int): CString is intern
103 # Get char at `index`.
104 fun [](index
: Int): Byte is intern
106 # Set char `item` at index.
107 fun []=(index
: Int, item
: Byte) is intern
109 # Copy `self` to `dest`.
110 fun copy_to
(dest
: CString, length
: Int, from
: Int, to
: Int) is intern
112 redef fun ==(o
) is intern do return is_same_instance
(o
)
114 redef fun !=(o
) is intern do return not is_same_instance
(o
)
116 # Position of the first nul character.
117 fun cstring_length
: Int
120 while self[l
] != 0u8
do l
+= 1
124 # Parse `self` as an Int.
125 fun atoi
: Int is intern
127 # Parse `self` as a Float.
128 fun atof
: Float `{ return atof(self); `}
130 # Gets the UTF-8 char at index `pos
`
132 # Index is expressed in Unicode chars
135 # assert "かきく".as(FlatString).items.char_at(0) == 'か'
138 # If the char at position pos is an invalid Unicode char,
139 # the Unicode replacement character � (0xFFFD) will be used.
142 # assert "かきく".as(FlatString).items.char_at(1) == '�'
144 fun char_at(pos: Int): Char do
146 if c & 0x80u8 == 0u8 then return c.ascii
147 var b = fetch_4_hchars(pos)
149 if b & 0xC00000u32 != 0x800000u32 then return 0xFFFD.code_point
150 if b & 0xE0000000u32 == 0xC0000000u32 then
151 ret |= (b & 0x1F000000u32) >> 18
152 ret |= (b & 0x3F0000u32) >> 16
153 return ret.code_point
155 if not b & 0xC000u32 == 0x8000u32 then return 0xFFFD.code_point
156 if b & 0xF0000000u32 == 0xE0000000u32 then
157 ret |= (b & 0xF000000u32) >> 12
158 ret |= (b & 0x3F0000u32) >> 10
159 ret |= (b & 0x3F00u32) >> 8
160 return ret.code_point
162 if not b & 0xC0u32 == 0x80u32 then return 0xFFFD.code_point
163 if b & 0xF8000000u32 == 0xF0000000u32 then
164 ret |= (b & 0x7000000u32) >> 6
165 ret |= (b & 0x3F0000u32) >> 4
166 ret |= (b & 0x3F00u32) >> 2
168 return ret.code_point
170 return 0xFFFD.code_point
173 # Gets the byte index of char at position `n
` in UTF-8 String
174 fun char_to_byte_index(n: Int): Int do return char_to_byte_index_cached(n, 0, 0)
176 # Gets the length of the character at position `pos
` (1 if invalid sequence)
177 fun length_of_char_at(pos: Int): Int do
179 if c & 0x80u8 == 0x00u8 then
181 else if c & 0xE0u8 == 0xC0u8 and self[pos + 1] & 0xC0u8 == 0x80u8 then
183 else if c & 0xF0u8 == 0xE0u8 and self[pos + 1] & 0xC0u8 == 0x80u8 and self[pos + 2] & 0xC0u8 == 0x80u8 then
185 else if c & 0xF8u8 == 0xF0u8 and self[pos + 1] & 0xC0u8 == 0x80u8 and self[pos + 2] & 0xC0u8 == 0x80u8 and self[pos + 3] & 0xC0u8 == 0x80u8 then
192 # Gets the byte index of char at position `n
` in UTF-8 String
194 # `char_from
` and `byte_from
` are cached values to seek from.
196 # NOTE: char_from and byte_from are not guaranteed to be valid cache values
197 # It it up to the client to ensure the validity of the information
198 fun char_to_byte_index_cached(n, char_from, byte_from: Int): Int do
206 var i = fetch_4_chars(ns_i)
207 if i & 0x80808080u32 != 0u32 then break
212 if dist == 0 then break
213 ns_i += length_of_char_at(ns_i)
220 var i = fetch_4_chars(ns_i - 4)
221 if i & 0x80808080u32 != 0u32 then break
226 if dist == 0 then break
227 ns_i = find_beginning_of_char_at(ns_i - 1)
235 # Gets the char index of byte at position `n
` in a UTF-8 String
237 # `char_from
` and `byte_from
` are cached values to seek from.
239 # NOTE: char_from and byte_from are not guaranteed to be valid cache values
240 # It it up to the client to ensure the validity of the information
241 fun byte_to_char_index_cached(n, char_from, byte_from: Int): Int do
246 ns_i += length_of_char_at(ns_i)
251 ns_i = find_beginning_of_char_at(ns_i - 1)
258 # Returns the beginning position of the char at position `pos
`
260 # If the char is invalid UTF-8, `pos
` is returned as-is
263 # assert "abc".items.find_beginning_of_char_at(2) == 2
264 # assert "か".items.find_beginning_of_char_at(1) == 0
265 # assert [0x41u8, 233u8].to_s.items.find_beginning_of_char_at(1) == 1
267 fun find_beginning_of_char_at(pos: Int): Int do
270 if c & 0x80u8 == 0x00u8 then return pos
271 while c & 0xC0u8 == 0x80u8 do
276 if length_of_char_at(stpos) >= (endpos - stpos + 1) then return pos
280 # Number of UTF-8 characters in `self` starting at `from
`, for a length of `byte_length
`
281 fun utf8_length(from, byte_length: Int): Int is intern do
284 while byte_length > 0 do
285 while byte_length >= 4 do
286 var i = fetch_4_chars(st)
287 if i & 0x80808080u32 != 0u32 then break
292 if byte_length == 0 then break
293 var cln = length_of_char_at(st)
301 # Fetch 4 chars in `self` at `pos
`
302 fun fetch_4_chars(pos: Int): UInt32 is intern `{ return *((uint32_t*)(self+pos)); `}
304 # Fetch 4 chars in `self` at `pos`
305 fun fetch_4_hchars
(pos
: Int): UInt32 is intern `{ return (uint32_t)be32toh(*((uint32_t*)(self+pos))); `}
307 # Right shifts `len
` bytes of `self` from `sh
` bytes starting at position `pos
`
308 fun rshift(sh, len, pos: Int) do
309 copy_to(self, len, pos, pos + sh)
312 # Left shifts `len
` bytes of `self` from `sh
` bytes starting at position `pos
`
313 fun lshift(sh, len, pos: Int) do
314 copy_to(self, len, pos, pos - sh)