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
374 # Immutable strings of characters.
379 # Index at which `self` begins in `_items`, inclusively
380 redef var first_byte
is noinit
382 redef var chars
= new FlatStringCharView(self) is lazy
384 redef var bytes
= new FlatStringByteView(self) is lazy
386 redef var to_cstring
is lazy
do
388 var new_items
= new NativeString(blen
+ 1)
389 _items
.copy_to
(new_items
, blen
, _first_byte
, 0)
390 new_items
[blen
] = 0u8
394 redef fun reversed
do
395 var b
= new FlatBuffer.with_capacity
(_bytelen
+ 1)
398 b
.add
self.fetch_char_at
(i
)
401 var s
= b
.to_s
.as(FlatString)
402 s
._length
= self._length
406 redef fun fast_cstring
do return _items
.fast_cstring
(_first_byte
)
408 redef fun substring_from
(from
) do
409 if from
>= self._length
then return empty
410 if from
<= 0 then return self
411 var c
= char_to_byte_index
(from
)
412 var st
= c
- _first_byte
413 var fln
= bytelen
- st
414 return new FlatString.full
(items
, fln
, c
, _length
- from
)
417 redef fun substring
(from
, count
)
419 if count
<= 0 then return ""
423 if count
< 0 then return ""
428 if (count
+ from
) > ln
then count
= ln
- from
429 if count
<= 0 then return ""
430 var end_index
= from
+ count
- 1
431 return substring_impl
(from
, count
, end_index
)
434 private fun substring_impl
(from
, count
, end_index
: Int): String do
435 var cache
= _position
436 var dfrom
= (cache
- from
).abs
437 var dend
= (end_index
- from
).abs
442 bytefrom
= char_to_byte_index
(from
)
443 byteto
= char_to_byte_index
(end_index
)
445 byteto
= char_to_byte_index
(end_index
)
446 bytefrom
= char_to_byte_index
(from
)
450 byteto
+= its
.length_of_char_at
(byteto
) - 1
452 var s
= new FlatString.full
(its
, byteto
- bytefrom
+ 1, bytefrom
, count
)
456 redef fun empty
do return "".as(FlatString)
460 var outstr
= new FlatBuffer.with_capacity
(self._bytelen
+ 1)
466 outstr
.add
(chars
[pos
].to_upper
)
475 var outstr
= new FlatBuffer.with_capacity
(self._bytelen
+ 1)
481 outstr
.add
(chars
[pos
].to_lower
)
490 for i
in chars
do i
.output
493 ##################################################
494 # String Specific Methods #
495 ##################################################
497 # Low-level creation of a new string with minimal data.
499 # `_items` will be used as is, without copy, to retrieve the characters of the string.
500 # Aliasing issues is the responsibility of the caller.
501 private init with_infos
(items
: NativeString, bytelen
, from
: Int)
504 self._bytelen
= bytelen
507 _length
= _items
.utf8_length
(_first_byte
, bytelen
)
510 # Low-level creation of a new string with all the data.
512 # `_items` will be used as is, without copy, to retrieve the characters of the string.
513 # Aliasing issues is the responsibility of the caller.
514 private init full
(items
: NativeString, bytelen
, from
, length
: Int)
517 self._length
= length
518 self._bytelen
= bytelen
525 if not other
isa FlatText then return super
527 if self.object_id
== other
.object_id
then return true
529 var my_length
= _bytelen
531 if other
._bytelen
!= my_length
then return false
533 var my_index
= _first_byte
534 var its_index
= other
.first_byte
536 var last_iteration
= my_index
+ my_length
538 var its_items
= other
._items
539 var my_items
= self._items
541 while my_index
< last_iteration
do
542 if my_items
[my_index
] != its_items
[its_index
] then return false
552 if not other
isa FlatText then return super
554 if self.object_id
== other
.object_id
then return false
557 var itsits
= other
._items
560 var obt
= other
.bytelen
562 var minln
= if mbt
< obt
then mbt
else obt
563 var mst
= _first_byte
564 var ost
= other
.first_byte
566 for i
in [0 .. minln
[ do
567 var my_curr_char
= myits
[mst
]
568 var its_curr_char
= itsits
[ost
]
570 if my_curr_char
> its_curr_char
then return false
571 if my_curr_char
< its_curr_char
then return true
584 var nlen
= mlen
+ slen
586 var mifrom
= _first_byte
587 if s
isa FlatText then
589 var sifrom
= s
.first_byte
590 var ns
= new NativeString(nlen
+ 1)
591 mits
.copy_to
(ns
, mlen
, mifrom
, 0)
592 sits
.copy_to
(ns
, slen
, sifrom
, mlen
)
593 return new FlatString.full
(ns
, nlen
, 0, _length
+ o
.length
)
600 var mybtlen
= _bytelen
601 var new_bytelen
= mybtlen
* i
603 var newlen
= mylen
* i
606 var ns
= new NativeString(new_bytelen
+ 1)
607 ns
[new_bytelen
] = 0u8
610 its
.copy_to
(ns
, mybtlen
, fb
, offset
)
614 return new FlatString.full
(ns
, new_bytelen
, 0, newlen
)
620 if hash_cache
== null then
621 # djb2 hash algorithm
625 var my_items
= _items
629 h
= (h
<< 5) + h
+ my_items
[i
].to_i
636 return hash_cache
.as(not null)
639 redef fun substrings
do return new FlatSubstringsIter(self)
642 private class FlatStringCharReverseIterator
643 super IndexedIterator[Char]
645 var target
: FlatString
649 redef fun is_ok
do return curr_pos
>= 0
651 redef fun item
do return target
[curr_pos
]
653 redef fun next
do curr_pos
-= 1
655 redef fun index
do return curr_pos
659 private class FlatStringCharIterator
660 super IndexedIterator[Char]
662 var target
: FlatString
664 var max
: Int is noautoinit
668 init do max
= target
._length
- 1
670 redef fun is_ok
do return curr_pos
<= max
672 redef fun item
do return target
[curr_pos
]
674 redef fun next
do curr_pos
+= 1
676 redef fun index
do return curr_pos
680 private class FlatStringCharView
683 redef type SELFTYPE: FlatString
685 redef fun [](index
) do return target
[index
]
687 redef fun iterator_from
(start
) do return new FlatStringCharIterator(target
, start
)
689 redef fun reverse_iterator_from
(start
) do return new FlatStringCharReverseIterator(target
, start
)
693 private class FlatStringByteReverseIterator
694 super IndexedIterator[Byte]
696 var target
: FlatString
698 var target_items
: NativeString is noautoinit
705 target_items
= tgt
._items
706 curr_pos
+= tgt
._first_byte
709 redef fun is_ok
do return curr_pos
>= target
._first_byte
711 redef fun item
do return target_items
[curr_pos
]
713 redef fun next
do curr_pos
-= 1
715 redef fun index
do return curr_pos
- target
._first_byte
719 private class FlatStringByteIterator
720 super IndexedIterator[Byte]
722 var target
: FlatString
724 var target_items
: NativeString is noautoinit
731 target_items
= tgt
._items
732 curr_pos
+= tgt
._first_byte
735 redef fun is_ok
do return curr_pos
<= target
.last_byte
737 redef fun item
do return target_items
[curr_pos
]
739 redef fun next
do curr_pos
+= 1
741 redef fun index
do return curr_pos
- target
._first_byte
745 private class FlatStringByteView
748 redef type SELFTYPE: FlatString
752 # Check that the index (+ _first_byte) is not larger than last_byte
753 # In other terms, if the index is valid
755 assert index
>= 0 and index
< target
._bytelen
756 var ind
= index
+ target
._first_byte
757 return target
._items
[ind
]
760 redef fun iterator_from
(start
) do return new FlatStringByteIterator(target
, start
)
762 redef fun reverse_iterator_from
(start
) do return new FlatStringByteReverseIterator(target
, start
)
767 redef new do return new FlatBuffer
769 redef new with_cap
(i
) do return new FlatBuffer.with_capacity
(i
)
772 # Mutable strings of characters.
777 redef var chars
: Sequence[Char] = new FlatBufferCharView(self) is lazy
779 redef var bytes
= new FlatBufferByteView(self) is lazy
781 private var char_cache
: Int = -1
783 private var byte_cache
: Int = -1
785 private var capacity
= 0
787 # Real items, used as cache for when to_cstring is called
788 private var real_items
: NativeString is noinit
790 redef fun fast_cstring
do return _items
.fast_cstring
(0)
792 redef fun substrings
do return new FlatSubstringsIter(self)
794 # Re-copies the `NativeString` into a new one and sets it as the new `Buffer`
796 # This happens when an operation modifies the current `Buffer` and
797 # the Copy-On-Write flag `written` is set at true.
799 var nns
= new NativeString(capacity
)
800 if _bytelen
!= 0 then _items
.copy_to
(nns
, _bytelen
, 0, 0)
805 # Shifts the content of the buffer by `len` bytes to the right, starting at byte `from`
807 # Internal only, does not modify _bytelen or length, this is the caller's responsability
808 private fun rshift_bytes
(from
: Int, len
: Int) do
812 if bt
+ len
> capacity
then
813 capacity
= capacity
* 2 + 2
814 nit
= new NativeString(capacity
)
815 oit
.copy_to
(nit
, 0, 0, from
)
817 oit
.copy_to
(nit
, bt
- from
, from
, from
+ len
)
820 # Shifts the content of the buffer by `len` bytes to the left, starting at `from`
822 # Internal only, does not modify _bytelen or length, this is the caller's responsability
823 private fun lshift_bytes
(from
: Int, len
: Int) do
825 it
.copy_to
(it
, _bytelen
- from
, from
, from
- len
)
828 redef fun []=(index
, item
)
830 assert index
>= 0 and index
<= _length
831 if written
then reset
833 if index
== _length
then
838 var ip
= it
.char_to_byte_index
(index
)
839 var c
= it
.char_at
(ip
)
840 var clen
= c
.u8char_len
841 var itemlen
= item
.u8char_len
842 var size_diff
= itemlen
- clen
843 if size_diff
> 0 then
844 rshift_bytes
(ip
+ clen
, size_diff
)
845 else if size_diff
< 0 then
846 lshift_bytes
(ip
+ clen
, -size_diff
)
848 _bytelen
+= size_diff
849 it
.set_char_at
(ip
, item
)
854 if written
then reset
856 var clen
= c
.u8char_len
859 _items
.set_char_at
(bt
, c
)
866 if written
then reset
871 redef fun empty
do return new Buffer
873 redef fun enlarge
(cap
)
876 if cap
<= c
then return
877 while c
<= cap
do c
= c
* 2 + 2
878 # The COW flag can be set at false here, since
879 # it does a copy of the current `Buffer`
882 var a
= new NativeString(c
+1)
885 if bln
> 0 then it
.copy_to
(a
, bln
, 0, 0)
895 if bln
== 0 then _items
= new NativeString(1)
896 return new FlatString.full
(_items
, bln
, 0, _length
)
903 var new_native
= new NativeString(bln
+ 1)
904 new_native
[bln
] = 0u8
905 if _length
> 0 then _items
.copy_to
(new_native
, bln
, 0, 0)
906 real_items
= new_native
912 # Create a new empty string.
915 # Low-level creation a new buffer with given data.
917 # `_items` will be used as is, without copy, to store the characters of the buffer.
918 # Aliasing issues is the responsibility of the caller.
920 # If `_items` is shared, `written` should be set to true after the creation
921 # so that a modification will do a copy-on-write.
922 private init with_infos
(items
: NativeString, capacity
, bytelen
, length
: Int)
925 self.capacity
= capacity
926 self._bytelen
= bytelen
927 self._length
= length
930 # Create a new string copied from `s`.
933 _items
= new NativeString(s
.bytelen
)
934 if s
isa FlatText then
937 for i
in substrings
do i
.as(FlatString)._items
.copy_to
(_items
, i
._bytelen
, 0, 0)
945 # Create a new empty string with a given capacity.
946 init with_capacity
(cap
: Int)
949 _items
= new NativeString(cap
+ 1)
956 if s
.is_empty
then return
959 var nln
= _bytelen
+ sl
961 if s
isa FlatText then
962 s
._items
.copy_to
(_items
, sl
, s
.first_byte
, _bytelen
)
964 for i
in s
.substrings
do append i
971 # Copies the content of self in `dest`
972 fun copy
(start
: Int, len
: Int, dest
: Buffer, new_start
: Int)
974 var self_chars
= self.chars
975 var dest_chars
= dest
.chars
976 for i
in [0..len-1
] do
977 dest_chars
[new_start
+i
] = self_chars
[start
+i
]
981 redef fun substring
(from
, count
)
984 if from
< 0 then from
= 0
985 if (from
+ count
) > _length
then count
= _length
- from
986 if count
<= 0 then return new Buffer
988 var bytefrom
= its
.char_to_byte_index
(from
)
989 var byteto
= its
.char_to_byte_index
(count
+ from
- 1)
990 byteto
+= its
.char_at
(byteto
).u8char_len
- 1
991 var byte_length
= byteto
- bytefrom
+ 1
992 var r_items
= new NativeString(byte_length
)
993 its
.copy_to
(r_items
, byte_length
, bytefrom
, 0)
994 return new FlatBuffer.with_infos
(r_items
, byte_length
, byte_length
, count
)
1000 var ns
= new FlatBuffer.with_capacity
(capacity
)
1001 for i
in chars
.reverse_iterator
do ns
.add i
1005 redef fun times
(repeats
)
1008 var x
= new FlatString.full
(_items
, bln
, 0, _length
)
1009 for i
in [1 .. repeats
[ do
1016 if written
then reset
1017 for i
in [0 .. _length
[ do self[i
] = self[i
].to_upper
1022 if written
then reset
1023 for i
in [0 .. _length
[ do self[i
] = self[i
].to_lower
1027 private class FlatBufferByteReverseIterator
1028 super IndexedIterator[Byte]
1030 var target
: FlatBuffer
1032 var target_items
: NativeString is noautoinit
1036 init do target_items
= target
._items
1038 redef fun index
do return curr_pos
1040 redef fun is_ok
do return curr_pos
>= 0
1042 redef fun item
do return target_items
[curr_pos
]
1044 redef fun next
do curr_pos
-= 1
1048 private class FlatBufferByteView
1049 super BufferByteView
1051 redef type SELFTYPE: FlatBuffer
1053 redef fun [](index
) do return target
._items
[index
]
1055 redef fun iterator_from
(pos
) do return new FlatBufferByteIterator(target
, pos
)
1057 redef fun reverse_iterator_from
(pos
) do return new FlatBufferByteReverseIterator(target
, pos
)
1061 private class FlatBufferByteIterator
1062 super IndexedIterator[Byte]
1064 var target
: FlatBuffer
1066 var target_items
: NativeString is noautoinit
1070 init do target_items
= target
._items
1072 redef fun index
do return curr_pos
1074 redef fun is_ok
do return curr_pos
< target
._bytelen
1076 redef fun item
do return target_items
[curr_pos
]
1078 redef fun next
do curr_pos
+= 1
1082 private class FlatBufferCharReverseIterator
1083 super IndexedIterator[Char]
1085 var target
: FlatBuffer
1089 redef fun index
do return curr_pos
1091 redef fun is_ok
do return curr_pos
>= 0
1093 redef fun item
do return target
[curr_pos
]
1095 redef fun next
do curr_pos
-= 1
1099 private class FlatBufferCharView
1100 super BufferCharView
1102 redef type SELFTYPE: FlatBuffer
1104 redef fun [](index
) do return target
[index
]
1106 redef fun []=(index
, item
)
1108 assert index
>= 0 and index
<= length
1109 if index
== length
then
1113 target
[index
] = item
1126 fun enlarge
(cap
: Int)
1133 var s_length
= s
.length
1134 if target
.capacity
< s
.length
then enlarge
(s_length
+ target
._length
)
1135 for i
in s
do target
.add i
1138 redef fun iterator_from
(pos
) do return new FlatBufferCharIterator(target
, pos
)
1140 redef fun reverse_iterator_from
(pos
) do return new FlatBufferCharReverseIterator(target
, pos
)
1144 private class FlatBufferCharIterator
1145 super IndexedIterator[Char]
1147 var target
: FlatBuffer
1149 var max
: Int is noautoinit
1153 init do max
= target
._length
- 1
1155 redef fun index
do return curr_pos
1157 redef fun is_ok
do return curr_pos
<= max
1159 redef fun item
do return target
[curr_pos
]
1161 redef fun next
do curr_pos
+= 1
1165 redef class NativeString
1168 return to_s_with_length
(cstring_length
)
1171 # Returns `self` as a String of `length`.
1172 redef fun to_s_with_length
(length
): FlatString
1175 return clean_utf8
(length
)
1178 redef fun to_s_full
(bytelen
, unilen
) do
1179 return new FlatString.full
(self, bytelen
, 0, unilen
)
1182 redef fun to_s_unsafe
(len
) do
1183 if len
== null then len
= cstring_length
1184 return new FlatString.with_infos
(self, len
, 0)
1187 # Returns `self` as a new String.
1188 redef fun to_s_with_copy
: FlatString
1190 var length
= cstring_length
1191 var r
= clean_utf8
(length
)
1192 if r
.items
!= self then return r
1193 var new_self
= new NativeString(length
+ 1)
1194 copy_to
(new_self
, length
, 0, 0)
1195 var str
= new FlatString.with_infos
(new_self
, length
, 0)
1196 new_self
[length
] = 0u8
1197 str
.to_cstring
= new_self
1201 # Cleans a NativeString if necessary
1202 fun clean_utf8
(len
: Int): FlatString do
1203 var replacements
: nullable Array[Int] = null
1204 var end_length
= len
1210 var i
= fetch_4_chars
(pos
)
1211 if i
& 0x80808080 != 0 then break
1216 if rem
== 0 then break
1218 if b
& 0x80u
8 == 0x00u
8 then
1224 var nxst
= length_of_char_at
(pos
)
1227 ok_st
= b
& 0x80u
8 == 0u8
1228 else if nxst
== 2 then
1229 ok_st
= b
& 0xE0u
8 == 0xC0u
8
1230 else if nxst
== 3 then
1231 ok_st
= b
& 0xF0u
8 == 0xE0u
8
1233 ok_st
= b
& 0xF8u
8 == 0xF0u
8
1236 if replacements
== null then replacements
= new Array[Int]
1237 replacements
.add pos
1245 var c
= char_at
(pos
)
1246 var cp
= c
.code_point
1248 ok_c
= cp
>= 0 and cp
<= 0x7F
1249 else if nxst
== 2 then
1250 ok_c
= cp
>= 0x80 and cp
<= 0x7FF
1251 else if nxst
== 3 then
1252 ok_c
= cp
>= 0x800 and cp
<= 0xFFFF
1253 ok_c
= ok_c
and not (cp
>= 0xD800 and cp
<= 0xDFFF) and cp
!= 0xFFFE and cp
!= 0xFFFF
1255 ok_c
= cp
>= 0x10000 and cp
<= 0x10FFFF
1258 if replacements
== null then replacements
= new Array[Int]
1259 replacements
.add pos
1266 var clen
= c
.u8char_len
1272 if end_length
!= len
then
1273 ret
= new NativeString(end_length
)
1276 var repls
= replacements
.as(not null)
1277 var r
= repls
.items
.as(not null)
1278 var imax
= repls
.length
1279 for i
in [0 .. imax
[ do
1281 var chkln
= repl_pos
- old_repl
1282 copy_to
(ret
, chkln
, old_repl
, off
)
1285 ret
[off
+ 1] = 0xBFu
8
1286 ret
[off
+ 2] = 0xBDu
8
1287 old_repl
= repl_pos
+ 1
1290 copy_to
(ret
, len
- old_repl
, old_repl
, off
)
1292 return new FlatString.full
(ret
, end_length
, 0, chr_ln
)
1295 # Sets the next bytes at position `pos` to the value of `c`, encoded in UTF-8
1297 # Very unsafe, make sure to have room for this char prior to calling this function.
1298 private fun set_char_at
(pos
: Int, c
: Char) do
1299 if c
.code_point
< 128 then
1300 self[pos
] = c
.code_point
.to_b
1303 var ln
= c
.u8char_len
1304 native_set_char
(pos
, c
, ln
)
1307 private fun native_set_char
(pos
: Int, c
: Char, ln
: Int) `{
1308 char* dst = self + pos;
1314 dst[0] = 0xC0 | ((c & 0x7C0) >> 6);
1315 dst[1] = 0x80 | (c & 0x3F);
1318 dst[0] = 0xE0 | ((c & 0xF000) >> 12);
1319 dst[1] = 0x80 | ((c & 0xFC0) >> 6);
1320 dst[2] = 0x80 | (c & 0x3F);
1323 dst[0] = 0xF0 | ((c & 0x1C0000) >> 18);
1324 dst[1] = 0x80 | ((c & 0x3F000) >> 12);
1325 dst[2] = 0x80 | ((c & 0xFC0) >> 6);
1326 dst[3] = 0x80 | (c & 0x3F);
1333 # return displayable int in base 10 and signed
1335 # assert 1.to_s == "1"
1336 # assert (-123).to_s == "-123"
1338 # Fast case for common numbers
1339 if self == 0 then return "0"
1340 if self == 1 then return "1"
1342 var nslen
= int_to_s_len
1343 var ns
= new NativeString(nslen
+ 1)
1345 native_int_to_s
(ns
, nslen
+ 1)
1346 return new FlatString.full
(ns
, nslen
, 0, nslen
)
1350 redef class Array[E
]
1352 # Fast implementation
1353 redef fun plain_to_s
1356 if l
== 0 then return ""
1357 var its
= _items
.as(not null)
1359 if l
== 1 then if first
== null then return "" else return first
.to_s
1360 var na
= new NativeArray[String](l
)
1366 if itsi
== null then
1376 var ns
= new NativeString(sl
+ 1)
1382 if tmp
isa FlatString then
1383 var tpl
= tmp
._bytelen
1384 tmp
._items
.copy_to
(ns
, tpl
, tmp
._first_byte
, off
)
1387 for j
in tmp
.substrings
do
1388 var s
= j
.as(FlatString)
1389 var slen
= s
._bytelen
1390 s
._items
.copy_to
(ns
, slen
, s
._first_byte
, off
)
1396 return new FlatString.with_infos
(ns
, sl
, 0)
1400 redef class NativeArray[E
]
1401 redef fun native_to_s
do
1402 assert self isa NativeArray[String]
1413 var ns
= new NativeString(sl
+ 1)
1419 if tmp
isa FlatString then
1420 var tpl
= tmp
._bytelen
1421 tmp
._items
.copy_to
(ns
, tpl
, tmp
._first_byte
, off
)
1424 for j
in tmp
.substrings
do
1425 var s
= j
.as(FlatString)
1426 var slen
= s
._bytelen
1427 s
._items
.copy_to
(ns
, slen
, s
._first_byte
, off
)
1433 return new FlatString.with_infos
(ns
, sl
, 0)
1437 redef class Map[K
,V
]
1438 redef fun join
(sep
, couple_sep
)
1440 if is_empty
then return ""
1442 var s
= new Buffer # Result
1448 s
.append
("{k or else "<null>"}{couple_sep}{e or else "<null>"}")
1450 # Concat other _items
1456 s
.append
("{k or else "<null>"}{couple_sep}{e or else "<null>"}")