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 # All the array-based text representations
14 intrude import abstract_text
22 private class FlatSubstringsIter
23 super Iterator[FlatText]
25 var tgt
: nullable FlatText
29 return tgt
.as(not null)
32 redef fun is_ok
do return tgt
!= null
34 redef fun next
do tgt
= null
39 # First byte of the NativeString
40 protected fun first_byte
: Int do return 0
42 # Last byte of the NativeString
43 protected fun last_byte
: Int do return first_byte
+ _bytelen
- 1
45 # Cache of the latest position (char) explored in the string
48 # Cached position (bytes) in the NativeString underlying the String
51 # Index of the character `index` in `_items`
52 fun char_to_byte_index
(index
: Int): Int do
53 var dpos
= index
- _position
58 if its
[b
] & 0x80u
8 == 0x00u
8 then
61 b
+= its
.length_of_char_at
(b
)
68 b
= its
.find_beginning_of_char_at
(b
- 1)
73 if dpos
== 0 then return b
77 # Find best insertion point
78 var delta_begin
= index
79 var delta_end
= (ln
- 1) - index
80 var delta_cache
= (pos
- index
).abs
83 if delta_cache
< min
then min
= delta_cache
84 if delta_end
< min
then min
= delta_end
89 if min
== delta_cache
then
92 else if min
== delta_begin
then
96 ns_i
= its
.find_beginning_of_char_at
(last_byte
)
100 ns_i
= its
.char_to_byte_index_cached
(index
, my_i
, ns_i
)
108 # By escaping `self` to HTML, how many more bytes will be needed ?
109 fun chars_to_html_escape
: Int do
118 else if c
== 0x3Eu
8 then
120 else if c
== 0x26u
8 then
122 else if c
== 0x22u
8 then
124 else if c
== 0x27u
8 then
126 else if c
== 0x2Fu
8 then
134 redef fun html_escape
136 var extra
= chars_to_html_escape
137 if extra
== 0 then return to_s
141 var nlen
= extra
+ _bytelen
142 var nits
= new NativeString(nlen
)
147 # Some HTML characters are used as meta-data, they need
148 # to be replaced by an HTML-Escaped equivalent
152 # * 0x26 (&) => &
153 # * 0x22 (") => "
154 # * 0x27 (') => '
155 # * 0x2F (/) => /
157 nits
[outpos
] = 0x26u
8
158 nits
[outpos
+ 1] = 0x6Cu
8
159 nits
[outpos
+ 2] = 0x74u
8
160 nits
[outpos
+ 3] = 0x3Bu
8
162 else if c
== 0x3Eu
8 then
163 nits
[outpos
] = 0x26u
8
164 nits
[outpos
+ 1] = 0x67u
8
165 nits
[outpos
+ 2] = 0x74u
8
166 nits
[outpos
+ 3] = 0x3Bu
8
168 else if c
== 0x26u
8 then
169 nits
[outpos
] = 0x26u
8
170 nits
[outpos
+ 1] = 0x61u
8
171 nits
[outpos
+ 2] = 0x6Du
8
172 nits
[outpos
+ 3] = 0x70u
8
173 nits
[outpos
+ 4] = 0x3Bu
8
175 else if c
== 0x22u
8 then
176 nits
[outpos
] = 0x26u
8
177 nits
[outpos
+ 1] = 0x23u
8
178 nits
[outpos
+ 2] = 0x33u
8
179 nits
[outpos
+ 3] = 0x34u
8
180 nits
[outpos
+ 4] = 0x3Bu
8
182 else if c
== 0x27u
8 then
183 nits
[outpos
] = 0x26u
8
184 nits
[outpos
+ 1] = 0x23u
8
185 nits
[outpos
+ 2] = 0x33u
8
186 nits
[outpos
+ 3] = 0x39u
8
187 nits
[outpos
+ 4] = 0x3Bu
8
189 else if c
== 0x2Fu
8 then
190 nits
[outpos
] = 0x26u
8
191 nits
[outpos
+ 1] = 0x23u
8
192 nits
[outpos
+ 2] = 0x34u
8
193 nits
[outpos
+ 3] = 0x37u
8
194 nits
[outpos
+ 4] = 0x3Bu
8
202 var s
= new FlatString.with_infos
(nits
, nlen
, 0)
206 # By escaping `self` to C, how many more bytes will be needed ?
208 # This enables a double-optimization in `escape_to_c` since if this
209 # method returns 0, then `self` does not need escaping and can be
211 fun chars_to_escape_to_c
: Int do
220 else if c
== 0x09u
8 then
222 else if c
== 0x22u
8 then
224 else if c
== 0x27u
8 then
226 else if c
== 0x5Cu
8 then
228 else if c
< 32u8
then
236 redef fun escape_to_c
do
237 var ln_extra
= chars_to_escape_to_c
238 if ln_extra
== 0 then return self.to_s
241 var nlen
= _bytelen
+ ln_extra
242 var nns
= new NativeString(nlen
)
249 # Any byte with value < 32 is a control character
250 # All their uses will be replaced by their octal
253 # There are two exceptions however:
258 # Aside from the code points above, the following are:
265 nns
[opos
+ 1] = 0x74u
8
267 else if c
== 0x0Au
8 then
269 nns
[opos
+ 1] = 0x6Eu
8
271 else if c
== 0x22u
8 then
273 nns
[opos
+ 1] = 0x22u
8
275 else if c
== 0x27u
8 then
277 nns
[opos
+ 1] = 0x27u
8
279 else if c
== 0x5Cu
8 then
281 nns
[opos
+ 1] = 0x5Cu
8
283 else if c
< 32u8
then
285 nns
[opos
+ 1] = 0x30u
8
286 nns
[opos
+ 2] = ((c
& 0x38u
8) >> 3) + 0x30u
8
287 nns
[opos
+ 3] = (c
& 0x07u
8) + 0x30u
8
295 return nns
.to_s_unsafe
(nlen
)
298 redef fun [](index
) do
302 # * ~70% want the next char
303 # * ~23% want the previous
304 # * ~7% want the same char
306 # So it makes sense to shortcut early. And early is here.
307 var dpos
= index
- _position
309 if dpos
== 1 and index
< len
- 1 then
312 if c
& 0x80u
8 == 0x00u
8 then
313 # We want the next, and current is easy.
314 # So next is easy to find!
318 # The rest will be done by `dpos==0` bellow.
321 else if dpos
== -1 and index
> 1 then
324 if c
& 0x80u
8 == 0x00u
8 then
325 # We want the previous, and it is easy.
334 # We know what we want (+0 or +1) just get it now!
337 if c
& 0x80u
8 == 0x00u
8 then return c
.ascii
338 return items
.char_at
(b
)
341 assert index
>= 0 and index
< len
342 return fetch_char_at
(index
)
345 # Gets a `Char` at `index` in `self`
347 # WARNING: Use at your own risks as no bound-checking is done
348 fun fetch_char_at
(index
: Int): Char do
349 var i
= char_to_byte_index
(index
)
352 if b
& 0x80u
8 == 0x00u
8 then return b
.ascii
353 return items
.char_at
(i
)
356 # If `self` contains only digits and alpha <= 'f', return the corresponding integer.
358 # assert "ff".to_hex == 255
359 redef fun to_hex
(pos
, ln
) do
361 if pos
== null then pos
= 0
362 if ln
== null then ln
= length
- pos
363 pos
= char_to_byte_index
(pos
)
366 for i
in [pos
.. max
[ do
368 res
+= its
[i
].ascii
.from_hex
373 redef fun copy_to_native
(dst
, n
, src_off
, dst_off
) do
374 _items
.copy_to
(dst
, n
, first_byte
+ src_off
, dst_off
)
378 # Immutable strings of characters.
379 abstract class FlatString
383 # Index at which `self` begins in `_items`, inclusively
384 redef var first_byte
is noinit
386 redef var chars
= new FlatStringCharView(self) is lazy
388 redef var bytes
= new FlatStringByteView(self) is lazy
390 redef var to_cstring
is lazy
do
392 var new_items
= new NativeString(blen
+ 1)
393 _items
.copy_to
(new_items
, blen
, _first_byte
, 0)
394 new_items
[blen
] = 0u8
398 redef fun reversed
do
399 var b
= new FlatBuffer.with_capacity
(_bytelen
+ 1)
402 b
.add
self.fetch_char_at
(i
)
405 var s
= b
.to_s
.as(FlatString)
406 s
._length
= self._length
410 redef fun fast_cstring
do return _items
.fast_cstring
(_first_byte
)
412 redef fun substring
(from
, count
)
414 if count
<= 0 then return ""
418 if count
<= 0 then return ""
423 if (count
+ from
) > ln
then count
= ln
- from
424 if count
<= 0 then return ""
425 var end_index
= from
+ count
- 1
426 return substring_impl
(from
, count
, end_index
)
429 private fun substring_impl
(from
, count
, end_index
: Int): String do
430 var cache
= _position
431 var dfrom
= (cache
- from
).abs
432 var dend
= (end_index
- from
).abs
437 bytefrom
= char_to_byte_index
(from
)
438 byteto
= char_to_byte_index
(end_index
)
440 byteto
= char_to_byte_index
(end_index
)
441 bytefrom
= char_to_byte_index
(from
)
445 byteto
+= its
.length_of_char_at
(byteto
) - 1
447 var s
= new FlatString.full
(its
, byteto
- bytefrom
+ 1, bytefrom
, count
)
451 redef fun empty
do return "".as(FlatString)
455 var outstr
= new FlatBuffer.with_capacity
(self._bytelen
+ 1)
461 outstr
.add
(chars
[pos
].to_upper
)
470 var outstr
= new FlatBuffer.with_capacity
(self._bytelen
+ 1)
476 outstr
.add
(chars
[pos
].to_lower
)
485 for i
in chars
do i
.output
488 ##################################################
489 # String Specific Methods #
490 ##################################################
492 # Low-level creation of a new string with minimal data.
494 # `_items` will be used as is, without copy, to retrieve the characters of the string.
495 # Aliasing issues is the responsibility of the caller.
496 private new with_infos
(items
: NativeString, bytelen
, from
: Int)
498 var len
= items
.utf8_length
(from
, bytelen
)
499 if bytelen
== len
then return new ASCIIFlatString.full_data
(items
, bytelen
, from
, len
)
500 return new UnicodeFlatString.full_data
(items
, bytelen
, from
, len
)
503 # Low-level creation of a new string with all the data.
505 # `_items` will be used as is, without copy, to retrieve the characters of the string.
506 # Aliasing issues is the responsibility of the caller.
507 private new full
(items
: NativeString, bytelen
, from
, length
: Int)
509 if bytelen
== length
then return new ASCIIFlatString.full_data
(items
, bytelen
, from
, length
)
510 return new UnicodeFlatString.full_data
(items
, bytelen
, from
, length
)
515 if not other
isa FlatText then return super
517 if self.object_id
== other
.object_id
then return true
519 var my_length
= _bytelen
521 if other
._bytelen
!= my_length
then return false
523 var my_index
= _first_byte
524 var its_index
= other
.first_byte
526 var last_iteration
= my_index
+ my_length
528 var its_items
= other
._items
529 var my_items
= self._items
531 while my_index
< last_iteration
do
532 if my_items
[my_index
] != its_items
[its_index
] then return false
542 if not other
isa FlatText then return super
544 if self.object_id
== other
.object_id
then return false
547 var itsits
= other
._items
550 var obt
= other
.bytelen
552 var minln
= if mbt
< obt
then mbt
else obt
553 var mst
= _first_byte
554 var ost
= other
.first_byte
556 for i
in [0 .. minln
[ do
557 var my_curr_char
= myits
[mst
]
558 var its_curr_char
= itsits
[ost
]
560 if my_curr_char
> its_curr_char
then return false
561 if my_curr_char
< its_curr_char
then return true
574 var nlen
= mlen
+ slen
576 var mifrom
= _first_byte
577 if s
isa FlatText then
579 var sifrom
= s
.first_byte
580 var ns
= new NativeString(nlen
+ 1)
581 mits
.copy_to
(ns
, mlen
, mifrom
, 0)
582 sits
.copy_to
(ns
, slen
, sifrom
, mlen
)
583 return new FlatString.full
(ns
, nlen
, 0, _length
+ o
.length
)
590 var mybtlen
= _bytelen
591 var new_bytelen
= mybtlen
* i
593 var newlen
= mylen
* i
596 var ns
= new NativeString(new_bytelen
+ 1)
597 ns
[new_bytelen
] = 0u8
600 its
.copy_to
(ns
, mybtlen
, fb
, offset
)
604 return new FlatString.full
(ns
, new_bytelen
, 0, newlen
)
609 if hash_cache
== null then
610 # djb2 hash algorithm
614 var my_items
= _items
618 h
= (h
<< 5) + h
+ my_items
[i
].to_i
625 return hash_cache
.as(not null)
628 redef fun substrings
do return new FlatSubstringsIter(self)
631 # Regular Nit UTF-8 strings
632 private class UnicodeFlatString
635 init full_data
(items
: NativeString, bytelen
, from
, length
: Int) do
637 self._length
= length
638 self._bytelen
= bytelen
643 redef fun substring_from
(from
) do
644 if from
>= self._length
then return empty
645 if from
<= 0 then return self
646 var c
= char_to_byte_index
(from
)
647 var st
= c
- _first_byte
648 var fln
= bytelen
- st
649 return new FlatString.full
(items
, fln
, c
, _length
- from
)
653 # Special cases of String where all the characters are ASCII-based
655 # Optimizes access operations to O(1) complexity.
656 private class ASCIIFlatString
659 init full_data
(items
: NativeString, bytelen
, from
, length
: Int) do
661 self._length
= length
662 self._bytelen
= bytelen
668 assert idx
< _bytelen
and idx
>= 0
669 return _items
[idx
+ _first_byte
].ascii
672 redef fun substring
(from
, count
) do
674 if count
<= 0 then return ""
675 if (count
+ from
) > ln
then count
= ln
- from
676 if count
<= 0 then return ""
679 if count
<= 0 then return ""
682 return new ASCIIFlatString.full_data
(_items
, count
, from
+ _first_byte
, count
)
685 redef fun reversed
do
686 var b
= new FlatBuffer.with_capacity
(_bytelen
+ 1)
692 var s
= b
.to_s
.as(FlatString)
696 redef fun char_to_byte_index
(index
) do return index
+ _first_byte
698 redef fun substring_impl
(from
, count
, end_index
) do
699 return new ASCIIFlatString.full_data
(_items
, count
, from
+ _first_byte
, count
)
702 redef fun fetch_char_at
(i
) do return _items
[i
+ _first_byte
].ascii
705 private class FlatStringCharReverseIterator
706 super IndexedIterator[Char]
708 var target
: FlatString
712 redef fun is_ok
do return curr_pos
>= 0
714 redef fun item
do return target
[curr_pos
]
716 redef fun next
do curr_pos
-= 1
718 redef fun index
do return curr_pos
722 private class FlatStringCharIterator
723 super IndexedIterator[Char]
725 var target
: FlatString
727 var max
: Int is noautoinit
731 init do max
= target
._length
- 1
733 redef fun is_ok
do return curr_pos
<= max
735 redef fun item
do return target
[curr_pos
]
737 redef fun next
do curr_pos
+= 1
739 redef fun index
do return curr_pos
743 private class FlatStringCharView
746 redef type SELFTYPE: FlatString
748 redef fun [](index
) do return target
[index
]
750 redef fun iterator_from
(start
) do return new FlatStringCharIterator(target
, start
)
752 redef fun reverse_iterator_from
(start
) do return new FlatStringCharReverseIterator(target
, start
)
756 private class FlatStringByteReverseIterator
757 super IndexedIterator[Byte]
759 var target
: FlatString
761 var target_items
: NativeString is noautoinit
768 target_items
= tgt
._items
769 curr_pos
+= tgt
._first_byte
772 redef fun is_ok
do return curr_pos
>= target
._first_byte
774 redef fun item
do return target_items
[curr_pos
]
776 redef fun next
do curr_pos
-= 1
778 redef fun index
do return curr_pos
- target
._first_byte
782 private class FlatStringByteIterator
783 super IndexedIterator[Byte]
785 var target
: FlatString
787 var target_items
: NativeString is noautoinit
794 target_items
= tgt
._items
795 curr_pos
+= tgt
._first_byte
798 redef fun is_ok
do return curr_pos
<= target
.last_byte
800 redef fun item
do return target_items
[curr_pos
]
802 redef fun next
do curr_pos
+= 1
804 redef fun index
do return curr_pos
- target
._first_byte
808 private class FlatStringByteView
811 redef type SELFTYPE: FlatString
815 # Check that the index (+ _first_byte) is not larger than last_byte
816 # In other terms, if the index is valid
818 assert index
>= 0 and index
< target
._bytelen
819 var ind
= index
+ target
._first_byte
820 return target
._items
[ind
]
823 redef fun iterator_from
(start
) do return new FlatStringByteIterator(target
, start
)
825 redef fun reverse_iterator_from
(start
) do return new FlatStringByteReverseIterator(target
, start
)
830 redef new do return new FlatBuffer
832 redef new with_cap
(i
) do return new FlatBuffer.with_capacity
(i
)
835 # Mutable strings of characters.
840 redef var chars
: Sequence[Char] = new FlatBufferCharView(self) is lazy
842 redef var bytes
= new FlatBufferByteView(self) is lazy
844 private var char_cache
: Int = -1
846 private var byte_cache
: Int = -1
848 private var capacity
= 0
850 # Real items, used as cache for when to_cstring is called
851 private var real_items
: NativeString is noinit
853 redef fun fast_cstring
do return _items
.fast_cstring
(0)
855 redef fun substrings
do return new FlatSubstringsIter(self)
857 # Re-copies the `NativeString` into a new one and sets it as the new `Buffer`
859 # This happens when an operation modifies the current `Buffer` and
860 # the Copy-On-Write flag `written` is set at true.
862 var nns
= new NativeString(capacity
)
863 if _bytelen
!= 0 then _items
.copy_to
(nns
, _bytelen
, 0, 0)
868 # Shifts the content of the buffer by `len` bytes to the right, starting at byte `from`
870 # Internal only, does not modify _bytelen or length, this is the caller's responsability
871 private fun rshift_bytes
(from
: Int, len
: Int) do
875 if bt
+ len
> capacity
then
876 capacity
= capacity
* 2 + 2
877 nit
= new NativeString(capacity
)
878 oit
.copy_to
(nit
, 0, 0, from
)
880 oit
.copy_to
(nit
, bt
- from
, from
, from
+ len
)
883 # Shifts the content of the buffer by `len` bytes to the left, starting at `from`
885 # Internal only, does not modify _bytelen or length, this is the caller's responsability
886 private fun lshift_bytes
(from
: Int, len
: Int) do
888 it
.copy_to
(it
, _bytelen
- from
, from
, from
- len
)
891 redef fun []=(index
, item
)
893 assert index
>= 0 and index
<= _length
894 if written
then reset
896 if index
== _length
then
901 var ip
= it
.char_to_byte_index
(index
)
902 var c
= it
.char_at
(ip
)
903 var clen
= c
.u8char_len
904 var itemlen
= item
.u8char_len
905 var size_diff
= itemlen
- clen
906 if size_diff
> 0 then
907 rshift_bytes
(ip
+ clen
, size_diff
)
908 else if size_diff
< 0 then
909 lshift_bytes
(ip
+ clen
, -size_diff
)
911 _bytelen
+= size_diff
912 it
.set_char_at
(ip
, item
)
917 if written
then reset
919 var clen
= c
.u8char_len
922 _items
.set_char_at
(bt
, c
)
937 redef fun empty
do return new Buffer
939 redef fun enlarge
(cap
)
942 if cap
<= c
then return
943 if c
<= 16 then c
= 16
944 while c
<= cap
do c
= c
* 2
945 # The COW flag can be set at false here, since
946 # it does a copy of the current `Buffer`
949 var a
= new NativeString(c
)
952 if bln
> 0 then it
.copy_to
(a
, bln
, 0, 0)
962 if bln
== 0 then _items
= new NativeString(1)
963 return new FlatString.full
(_items
, bln
, 0, _length
)
970 var new_native
= new NativeString(bln
+ 1)
971 new_native
[bln
] = 0u8
972 if _length
> 0 then _items
.copy_to
(new_native
, bln
, 0, 0)
973 real_items
= new_native
979 # Create a new empty string.
982 # Low-level creation a new buffer with given data.
984 # `_items` will be used as is, without copy, to store the characters of the buffer.
985 # Aliasing issues is the responsibility of the caller.
987 # If `_items` is shared, `written` should be set to true after the creation
988 # so that a modification will do a copy-on-write.
989 private init with_infos
(items
: NativeString, capacity
, bytelen
, length
: Int)
992 self.capacity
= capacity
993 self._bytelen
= bytelen
994 self._length
= length
997 # Create a new string copied from `s`.
1000 _items
= new NativeString(s
.bytelen
)
1001 for i
in s
.substrings
do i
._items
.copy_to
(_items
, i
._bytelen
, first_byte
, 0)
1002 _bytelen
= s
.bytelen
1004 _capacity
= _bytelen
1007 # Create a new empty string with a given capacity.
1008 init with_capacity
(cap
: Int)
1011 _items
= new NativeString(cap
)
1018 if s
.is_empty
then return
1021 var nln
= _bytelen
+ sl
1023 if s
isa FlatText then
1024 s
._items
.copy_to
(_items
, sl
, s
.first_byte
, _bytelen
)
1026 for i
in s
.substrings
do append i
1033 # Copies the content of self in `dest`
1034 fun copy
(start
: Int, len
: Int, dest
: Buffer, new_start
: Int)
1036 var self_chars
= self.chars
1037 var dest_chars
= dest
.chars
1038 for i
in [0..len-1
] do
1039 dest_chars
[new_start
+i
] = self_chars
[start
+i
]
1043 redef fun substring
(from
, count
)
1046 if from
< 0 then from
= 0
1047 if (from
+ count
) > _length
then count
= _length
- from
1048 if count
<= 0 then return new Buffer
1050 var bytefrom
= its
.char_to_byte_index
(from
)
1051 var byteto
= its
.char_to_byte_index
(count
+ from
- 1)
1052 byteto
+= its
.char_at
(byteto
).u8char_len
- 1
1053 var byte_length
= byteto
- bytefrom
+ 1
1054 var r_items
= new NativeString(byte_length
)
1055 its
.copy_to
(r_items
, byte_length
, bytefrom
, 0)
1056 return new FlatBuffer.with_infos
(r_items
, byte_length
, byte_length
, count
)
1059 redef fun append_substring_impl
(s
, from
, length
) do
1060 if length
<= 0 then return
1061 if not s
isa FlatText then
1065 var bytest
= s
.char_to_byte_index
(from
)
1066 var bytend
= s
.char_to_byte_index
(from
+ length
- 1)
1067 var btln
= bytend
- bytest
+ 1
1068 enlarge
(btln
+ _bytelen
)
1069 s
._items
.copy_to
(_items
, btln
, bytest
, _bytelen
)
1077 var ns
= new FlatBuffer.with_capacity
(capacity
)
1078 for i
in chars
.reverse_iterator
do ns
.add i
1082 redef fun times
(repeats
)
1085 var x
= new FlatString.full
(_items
, bln
, 0, _length
)
1086 for i
in [1 .. repeats
[ do
1093 if written
then reset
1094 for i
in [0 .. _length
[ do self[i
] = self[i
].to_upper
1099 if written
then reset
1100 for i
in [0 .. _length
[ do self[i
] = self[i
].to_lower
1104 private class FlatBufferByteReverseIterator
1105 super IndexedIterator[Byte]
1107 var target
: FlatBuffer
1109 var target_items
: NativeString is noautoinit
1113 init do target_items
= target
._items
1115 redef fun index
do return curr_pos
1117 redef fun is_ok
do return curr_pos
>= 0
1119 redef fun item
do return target_items
[curr_pos
]
1121 redef fun next
do curr_pos
-= 1
1125 private class FlatBufferByteView
1126 super BufferByteView
1128 redef type SELFTYPE: FlatBuffer
1130 redef fun [](index
) do return target
._items
[index
]
1132 redef fun iterator_from
(pos
) do return new FlatBufferByteIterator(target
, pos
)
1134 redef fun reverse_iterator_from
(pos
) do return new FlatBufferByteReverseIterator(target
, pos
)
1138 private class FlatBufferByteIterator
1139 super IndexedIterator[Byte]
1141 var target
: FlatBuffer
1143 var target_items
: NativeString is noautoinit
1147 init do target_items
= target
._items
1149 redef fun index
do return curr_pos
1151 redef fun is_ok
do return curr_pos
< target
._bytelen
1153 redef fun item
do return target_items
[curr_pos
]
1155 redef fun next
do curr_pos
+= 1
1159 private class FlatBufferCharReverseIterator
1160 super IndexedIterator[Char]
1162 var target
: FlatBuffer
1166 redef fun index
do return curr_pos
1168 redef fun is_ok
do return curr_pos
>= 0
1170 redef fun item
do return target
[curr_pos
]
1172 redef fun next
do curr_pos
-= 1
1176 private class FlatBufferCharView
1177 super BufferCharView
1179 redef type SELFTYPE: FlatBuffer
1181 redef fun [](index
) do return target
[index
]
1183 redef fun []=(index
, item
)
1185 assert index
>= 0 and index
<= length
1186 if index
== length
then
1190 target
[index
] = item
1203 fun enlarge
(cap
: Int)
1210 var s_length
= s
.length
1211 if target
.capacity
< s
.length
then enlarge
(s_length
+ target
._length
)
1212 for i
in s
do target
.add i
1215 redef fun iterator_from
(pos
) do return new FlatBufferCharIterator(target
, pos
)
1217 redef fun reverse_iterator_from
(pos
) do return new FlatBufferCharReverseIterator(target
, pos
)
1221 private class FlatBufferCharIterator
1222 super IndexedIterator[Char]
1224 var target
: FlatBuffer
1226 var max
: Int is noautoinit
1230 init do max
= target
._length
- 1
1232 redef fun index
do return curr_pos
1234 redef fun is_ok
do return curr_pos
<= max
1236 redef fun item
do return target
[curr_pos
]
1238 redef fun next
do curr_pos
+= 1
1242 redef class NativeString
1245 return to_s_with_length
(cstring_length
)
1248 redef fun to_s_with_length
(length
)
1251 return clean_utf8
(length
)
1254 redef fun to_s_full
(bytelen
, unilen
) do
1255 return new FlatString.full
(self, bytelen
, 0, unilen
)
1258 redef fun to_s_unsafe
(len
) do
1259 if len
== null then len
= cstring_length
1260 return new FlatString.with_infos
(self, len
, 0)
1263 redef fun to_s_with_copy
do return to_s_with_copy_and_length
(cstring_length
)
1265 # Get a `String` from `length` bytes at `self` copied into Nit memory
1266 fun to_s_with_copy_and_length
(length
: Int): String
1268 var r
= clean_utf8
(length
)
1269 if r
.items
!= self then return r
1270 var new_self
= new NativeString(length
+ 1)
1271 copy_to
(new_self
, length
, 0, 0)
1272 var str
= new FlatString.with_infos
(new_self
, length
, 0)
1273 new_self
[length
] = 0u8
1274 str
.to_cstring
= new_self
1278 # Cleans a NativeString if necessary
1279 fun clean_utf8
(len
: Int): FlatString do
1280 var replacements
: nullable Array[Int] = null
1281 var end_length
= len
1287 var i
= fetch_4_chars
(pos
)
1288 if i
& 0x80808080 != 0 then break
1293 if rem
== 0 then break
1295 if b
& 0x80u
8 == 0x00u
8 then
1301 var nxst
= length_of_char_at
(pos
)
1304 ok_st
= b
& 0x80u
8 == 0u8
1305 else if nxst
== 2 then
1306 ok_st
= b
& 0xE0u
8 == 0xC0u
8
1307 else if nxst
== 3 then
1308 ok_st
= b
& 0xF0u
8 == 0xE0u
8
1310 ok_st
= b
& 0xF8u
8 == 0xF0u
8
1313 if replacements
== null then replacements
= new Array[Int]
1314 replacements
.add pos
1322 var c
= char_at
(pos
)
1323 var cp
= c
.code_point
1325 ok_c
= cp
>= 0 and cp
<= 0x7F
1326 else if nxst
== 2 then
1327 ok_c
= cp
>= 0x80 and cp
<= 0x7FF
1328 else if nxst
== 3 then
1329 ok_c
= cp
>= 0x800 and cp
<= 0xFFFF
1330 ok_c
= ok_c
and not (cp
>= 0xD800 and cp
<= 0xDFFF) and cp
!= 0xFFFE and cp
!= 0xFFFF
1332 ok_c
= cp
>= 0x10000 and cp
<= 0x10FFFF
1335 if replacements
== null then replacements
= new Array[Int]
1336 replacements
.add pos
1343 var clen
= c
.u8char_len
1349 if end_length
!= len
then
1350 ret
= new NativeString(end_length
)
1353 var repls
= replacements
.as(not null)
1354 var r
= repls
.items
.as(not null)
1355 var imax
= repls
.length
1356 for i
in [0 .. imax
[ do
1358 var chkln
= repl_pos
- old_repl
1359 copy_to
(ret
, chkln
, old_repl
, off
)
1362 ret
[off
+ 1] = 0xBFu
8
1363 ret
[off
+ 2] = 0xBDu
8
1364 old_repl
= repl_pos
+ 1
1367 copy_to
(ret
, len
- old_repl
, old_repl
, off
)
1369 return new FlatString.full
(ret
, end_length
, 0, chr_ln
)
1372 # Sets the next bytes at position `pos` to the value of `c`, encoded in UTF-8
1374 # Very unsafe, make sure to have room for this char prior to calling this function.
1375 private fun set_char_at
(pos
: Int, c
: Char) do
1376 var cp
= c
.code_point
1381 var ln
= c
.u8char_len
1383 self[pos
] = (0xC0 | ((cp
& 0x7C0) >> 6)).to_b
1384 self[pos
+ 1] = (0x80 | (cp
& 0x3F)).to_b
1385 else if ln
== 3 then
1386 self[pos
] = (0xE0 | ((cp
& 0xF000) >> 12)).to_b
1387 self[pos
+ 1] = (0x80 | ((cp
& 0xFC0) >> 6)).to_b
1388 self[pos
+ 2] = (0x80 | (cp
& 0x3F)).to_b
1389 else if ln
== 4 then
1390 self[pos
] = (0xF0 | ((cp
& 0x1C0000) >> 18)).to_b
1391 self[pos
+ 1] = (0x80 | ((cp
& 0x3F000) >> 12)).to_b
1392 self[pos
+ 2] = (0x80 | ((cp
& 0xFC0) >> 6)).to_b
1393 self[pos
+ 3] = (0x80 | (cp
& 0x3F)).to_b
1399 # return displayable int in base 10 and signed
1401 # assert 1.to_s == "1"
1402 # assert (-123).to_s == "-123"
1404 # Fast case for common numbers
1405 if self == 0 then return "0"
1406 if self == 1 then return "1"
1408 var nslen
= int_to_s_len
1409 var ns
= new NativeString(nslen
+ 1)
1411 native_int_to_s
(ns
, nslen
+ 1)
1412 return new FlatString.full
(ns
, nslen
, 0, nslen
)
1416 redef class Array[E
]
1418 # Fast implementation
1419 redef fun plain_to_s
1422 if l
== 0 then return ""
1423 var its
= _items
.as(not null)
1425 if l
== 1 then if first
== null then return "" else return first
.to_s
1426 var na
= new NativeArray[String](l
)
1432 if itsi
== null then
1442 var ns
= new NativeString(sl
+ 1)
1448 if tmp
isa FlatString then
1449 var tpl
= tmp
._bytelen
1450 tmp
._items
.copy_to
(ns
, tpl
, tmp
._first_byte
, off
)
1453 for j
in tmp
.substrings
do
1454 var s
= j
.as(FlatString)
1455 var slen
= s
._bytelen
1456 s
._items
.copy_to
(ns
, slen
, s
._first_byte
, off
)
1462 return new FlatString.with_infos
(ns
, sl
, 0)
1466 redef class NativeArray[E
]
1467 redef fun native_to_s
do
1468 assert self isa NativeArray[String]
1479 var ns
= new NativeString(sl
+ 1)
1485 if tmp
isa FlatString then
1486 var tpl
= tmp
._bytelen
1487 tmp
._items
.copy_to
(ns
, tpl
, tmp
._first_byte
, off
)
1490 for j
in tmp
.substrings
do
1491 var s
= j
.as(FlatString)
1492 var slen
= s
._bytelen
1493 s
._items
.copy_to
(ns
, slen
, s
._first_byte
, off
)
1499 return new FlatString.with_infos
(ns
, sl
, 0)
1503 redef class Map[K
,V
]
1504 redef fun join
(sep
, couple_sep
)
1506 if is_empty
then return ""
1508 var s
= new Buffer # Result
1514 s
.append
("{k or else "<null>"}{couple_sep}{e or else "<null>"}")
1516 # Concat other _items
1522 s
.append
("{k or else "<null>"}{couple_sep}{e or else "<null>"}")