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 # Abstract class for manipulation of sequences of characters
17 intrude import collection
::array
25 # High-level abstraction for all text representations
29 redef type OTHER: Text
31 # Type of self (used for factorization of several methods, ex : substring_from, empty...)
34 # Gets a view on the chars of the Text object
36 # assert "hello".chars.to_a == ['h', 'e', 'l', 'l', 'o']
37 fun chars
: SequenceRead[Char] is abstract
39 # Gets a view on the bytes of the Text object
41 # assert "hello".bytes.to_a == [104u8, 101u8, 108u8, 108u8, 111u8]
42 fun bytes
: SequenceRead[Byte] is abstract
44 # Number of characters contained in self.
46 # assert "12345".length == 5
47 # assert "".length == 0
48 # assert "あいうえお".length == 5
49 fun length
: Int is abstract
51 # Number of bytes in `self`
53 # assert "12345".bytelen == 5
54 # assert "あいうえお".bytelen == 15
55 fun bytelen
: Int is abstract
59 # assert "abcd".substring(1, 2) == "bc"
60 # assert "abcd".substring(-1, 2) == "a"
61 # assert "abcd".substring(1, 0) == ""
62 # assert "abcd".substring(2, 5) == "cd"
63 # assert "あいうえお".substring(1,3) == "いうえ"
65 # A `from` index < 0 will be replaced by 0.
66 # Unless a `count` value is > 0 at the same time.
67 # In this case, `from += count` and `count -= from`.
68 fun substring
(from
: Int, count
: Int): SELFTYPE is abstract
70 # Iterates on the substrings of self if any
71 private fun substrings
: Iterator[FlatText] is abstract
73 # Is the current Text empty (== "")
76 # assert not "foo".is_empty
77 fun is_empty
: Bool do return self.length
== 0
79 # Returns an empty Text of the right type
81 # This method is used internally to get the right
82 # implementation of an empty string.
83 protected fun empty
: SELFTYPE is abstract
85 # Gets the first char of the Text
87 # DEPRECATED : Use self.chars.first instead
88 fun first
: Char do return self.chars
[0]
90 # Access a character at `index` in the string.
92 # assert "abcd"[2] == 'c'
94 # DEPRECATED : Use self.chars.[] instead
95 fun [](index
: Int): Char do return self.chars
[index
]
97 # Gets the index of the first occurence of 'c'
99 # Returns -1 if not found
101 # DEPRECATED : Use self.chars.index_of instead
102 fun index_of
(c
: Char): Int
104 return index_of_from
(c
, 0)
107 # Gets the last char of self
109 # DEPRECATED : Use self.chars.last instead
110 fun last
: Char do return self.chars
[length-1
]
112 # Gets the index of the first occurence of ´c´ starting from ´pos´
114 # Returns -1 if not found
116 # DEPRECATED : Use self.chars.index_of_from instead
117 fun index_of_from
(c
: Char, pos
: Int): Int
119 var iter
= self.chars
.iterator_from
(pos
)
121 if iter
.item
== c
then return iter
.index
127 # Gets the last index of char ´c´
129 # Returns -1 if not found
131 # DEPRECATED : Use self.chars.last_index_of instead
132 fun last_index_of
(c
: Char): Int
134 return last_index_of_from
(c
, length
- 1)
137 # Return a null terminated char *
138 fun to_cstring
: NativeString is abstract
140 # The index of the last occurrence of an element starting from pos (in reverse order).
142 # var s = "/etc/bin/test/test.nit"
143 # assert s.last_index_of_from('/', s.length-1) == 13
144 # assert s.last_index_of_from('/', 12) == 8
146 # Returns -1 if not found
148 # DEPRECATED : Use self.chars.last_index_of_from instead
149 fun last_index_of_from
(item
: Char, pos
: Int): Int do return chars
.last_index_of_from
(item
, pos
)
151 # Gets an iterator on the chars of self
153 # DEPRECATED : Use self.chars.iterator instead
154 fun iterator
: Iterator[Char]
156 return self.chars
.iterator
160 # Gets an Array containing the chars of self
162 # DEPRECATED : Use self.chars.to_a instead
163 fun to_a
: Array[Char] do return chars
.to_a
165 # Create a substring from `self` beginning at the `from` position
167 # assert "abcd".substring_from(1) == "bcd"
168 # assert "abcd".substring_from(-1) == "abcd"
169 # assert "abcd".substring_from(2) == "cd"
171 # As with substring, a `from` index < 0 will be replaced by 0
172 fun substring_from
(from
: Int): SELFTYPE
174 if from
>= self.length
then return empty
175 if from
< 0 then from
= 0
176 return substring
(from
, length
- from
)
179 # Does self have a substring `str` starting from position `pos`?
181 # assert "abcd".has_substring("bc",1) == true
182 # assert "abcd".has_substring("bc",2) == false
184 # Returns true iff all characters of `str` are presents
185 # at the expected index in `self.`
186 # The first character of `str` being at `pos`, the second
187 # character being at `pos+1` and so on...
189 # This means that all characters of `str` need to be inside `self`.
191 # assert "abcd".has_substring("xab", -1) == false
192 # assert "abcd".has_substring("cdx", 2) == false
194 # And that the empty string is always a valid substring.
196 # assert "abcd".has_substring("", 2) == true
197 # assert "abcd".has_substring("", 200) == true
198 fun has_substring
(str
: String, pos
: Int): Bool
200 if str
.is_empty
then return true
201 if pos
< 0 or pos
+ str
.length
> length
then return false
202 var myiter
= self.chars
.iterator_from
(pos
)
203 var itsiter
= str
.chars
.iterator
204 while myiter
.is_ok
and itsiter
.is_ok
do
205 if myiter
.item
!= itsiter
.item
then return false
209 if itsiter
.is_ok
then return false
213 # Is this string prefixed by `prefix`?
215 # assert "abcd".has_prefix("ab") == true
216 # assert "abcbc".has_prefix("bc") == false
217 # assert "ab".has_prefix("abcd") == false
218 fun has_prefix
(prefix
: String): Bool do return has_substring
(prefix
,0)
220 # Is this string suffixed by `suffix`?
222 # assert "abcd".has_suffix("abc") == false
223 # assert "abcd".has_suffix("bcd") == true
224 fun has_suffix
(suffix
: String): Bool do return has_substring
(suffix
, length
- suffix
.length
)
226 # Returns `self` as the corresponding integer
228 # assert "123".to_i == 123
229 # assert "-1".to_i == -1
230 # assert "0x64".to_i == 100
231 # assert "0b1100_0011".to_i== 195
232 # assert "--12".to_i == 12
234 # REQUIRE: `self`.`is_int`
235 fun to_i
: Int is abstract
237 # If `self` contains a float, return the corresponding float
239 # assert "123".to_f == 123.0
240 # assert "-1".to_f == -1.0
241 # assert "-1.2e-3".to_f == -0.0012
245 return to_s
.to_cstring
.atof
248 # If `self` contains only digits and alpha <= 'f', return the corresponding integer.
250 # assert "ff".to_hex == 255
251 fun to_hex
(pos
, ln
: nullable Int): Int do
253 if pos
== null then pos
= 0
254 if ln
== null then ln
= length
- pos
256 for i
in [pos
.. max
[ do
258 res
+= self[i
].from_hex
263 # If `self` contains only digits <= '7', return the corresponding integer.
265 # assert "714".to_oct == 460
266 fun to_oct
: Int do return a_to
(8)
268 # If `self` contains only '0' et '1', return the corresponding integer.
270 # assert "101101".to_bin == 45
271 fun to_bin
: Int do return a_to
(2)
273 # If `self` contains only digits '0' .. '9', return the corresponding integer.
275 # assert "108".to_dec == 108
276 fun to_dec
: Int do return a_to
(10)
278 # If `self` contains only digits and letters, return the corresponding integer in a given base
280 # assert "120".a_to(3) == 15
281 fun a_to
(base
: Int) : Int
286 for j
in [0..length
[ do
308 # Is this string in a valid numeric format compatible with `to_f`?
310 # assert "123".is_numeric == true
311 # assert "1.2".is_numeric == true
312 # assert "-1.2".is_numeric == true
313 # assert "-1.23e-2".is_numeric == true
314 # assert "1..2".is_numeric == false
315 # assert "".is_numeric == false
318 var has_point
= false
320 for i
in [0..length
[ do
322 if not c
.is_numeric
then
323 if c
== '.' and not has_point
then
325 else if c
== 'e' and e_index
== -1 and i
> 0 and i
< length
- 1 and chars
[i-1
] != '-' then
327 else if c
== '-' and i
== e_index
+ 1 and i
< length
- 1 then
336 # Returns `true` if the string contains only Hex chars
338 # assert "048bf".is_hex == true
339 # assert "ABCDEF".is_hex == true
340 # assert "0G".is_hex == false
343 for i
in [0..length
[ do
345 if not (c
>= 'a' and c
<= 'f') and
346 not (c
>= 'A' and c
<= 'F') and
347 not (c
>= '0' and c
<= '9') then return false
352 # Returns `true` if the string contains only Binary digits
354 # assert "1101100".is_bin == true
355 # assert "1101020".is_bin == false
357 for i
in chars
do if i
!= '0' and i
!= '1' then return false
361 # Returns `true` if the string contains only Octal digits
363 # assert "213453".is_oct == true
364 # assert "781".is_oct == false
366 for i
in chars
do if i
< '0' or i
> '7' then return false
370 # Returns `true` if the string contains only Decimal digits
372 # assert "10839".is_dec == true
373 # assert "164F".is_dec == false
375 for i
in chars
do if i
< '0' or i
> '9' then return false
379 # Are all letters in `self` upper-case ?
381 # assert "HELLO WORLD".is_upper == true
382 # assert "%$&%!".is_upper == true
383 # assert "hello world".is_upper == false
384 # assert "Hello World".is_upper == false
387 for i
in [0..length
[ do
389 if char
.is_lower
then return false
394 # Are all letters in `self` lower-case ?
396 # assert "hello world".is_lower == true
397 # assert "%$&%!".is_lower == true
398 # assert "Hello World".is_lower == false
401 for i
in [0..length
[ do
403 if char
.is_upper
then return false
408 # Removes the whitespaces at the beginning of self
410 # assert " \n\thello \n\t".l_trim == "hello \n\t"
412 # `Char::is_whitespace` determines what is a whitespace.
415 var iter
= self.chars
.iterator
417 if not iter
.item
.is_whitespace
then break
420 if iter
.index
== length
then return self.empty
421 return self.substring_from
(iter
.index
)
424 # Removes the whitespaces at the end of self
426 # assert " \n\thello \n\t".r_trim == " \n\thello"
428 # `Char::is_whitespace` determines what is a whitespace.
431 var iter
= self.chars
.reverse_iterator
433 if not iter
.item
.is_whitespace
then break
436 if iter
.index
< 0 then return self.empty
437 return self.substring
(0, iter
.index
+ 1)
440 # Trims trailing and preceding white spaces
442 # assert " Hello World ! ".trim == "Hello World !"
443 # assert "\na\nb\tc\t".trim == "a\nb\tc"
445 # `Char::is_whitespace` determines what is a whitespace.
446 fun trim
: SELFTYPE do return (self.l_trim
).r_trim
448 # Is the string non-empty but only made of whitespaces?
450 # assert " \n\t ".is_whitespace == true
451 # assert " hello ".is_whitespace == false
452 # assert "".is_whitespace == false
454 # `Char::is_whitespace` determines what is a whitespace.
455 fun is_whitespace
: Bool
457 if is_empty
then return false
458 for c
in self.chars
do
459 if not c
.is_whitespace
then return false
464 # Returns `self` removed from its last line terminator (if any).
466 # assert "Hello\n".chomp == "Hello"
467 # assert "Hello".chomp == "Hello"
469 # assert "\n".chomp == ""
470 # assert "".chomp == ""
472 # Line terminators are `"\n"`, `"\r\n"` and `"\r"`.
473 # A single line terminator, the last one, is removed.
475 # assert "\r\n".chomp == ""
476 # assert "\r\n\n".chomp == "\r\n"
477 # assert "\r\n\r\n".chomp == "\r\n"
478 # assert "\r\n\r".chomp == "\r\n"
480 # Note: unlike with most IO methods like `Reader::read_line`,
481 # a single `\r` is considered here to be a line terminator and will be removed.
485 if len
== 0 then return self
486 var l
= self.chars
.last
488 return substring
(0, len-1
)
489 else if l
!= '\n' then
491 else if len
> 1 and self.chars
[len-2
] == '\r' then
492 return substring
(0, len-2
)
494 return substring
(0, len-1
)
498 # Justify `self` in a space of `length`
500 # `left` is the space ratio on the left side.
501 # * 0.0 for left-justified (no space at the left)
502 # * 1.0 for right-justified (all spaces at the left)
503 # * 0.5 for centered (half the spaces at the left)
505 # `char`, or `' '` by default, is repeated to pad the empty space.
509 # assert "hello".justify(10, 0.0) == "hello "
510 # assert "hello".justify(10, 1.0) == " hello"
511 # assert "hello".justify(10, 0.5) == " hello "
512 # assert "hello".justify(10, 0.5, '.') == "..hello..."
514 # If `length` is not enough, `self` is returned as is.
516 # assert "hello".justify(2, 0.0) == "hello"
518 # REQUIRE: `left >= 0.0 and left <= 1.0`
519 # ENSURE: `self.length <= length implies result.length == length`
520 # ENSURE: `self.length >= length implies result == self`
521 fun justify
(length
: Int, left
: Float, char
: nullable Char): String
523 var pad
= (char
or else ' ').to_s
524 var diff
= length
- self.length
525 if diff
<= 0 then return to_s
526 assert left
>= 0.0 and left
<= 1.0
527 var before
= (diff
.to_f
* left
).to_i
528 return pad
* before
+ self + pad
* (diff-before
)
531 # Mangle a string to be a unique string only made of alphanumeric characters and underscores.
533 # This method is injective (two different inputs never produce the same
534 # output) and the returned string always respect the following rules:
536 # * Contains only US-ASCII letters, digits and underscores.
537 # * Never starts with a digit.
538 # * Never ends with an underscore.
539 # * Never contains two contiguous underscores.
541 # assert "42_is/The answer!".to_cmangle == "_52d2_is_47dThe_32danswer_33d"
542 # assert "__".to_cmangle == "_95d_95d"
543 # assert "__d".to_cmangle == "_95d_d"
544 # assert "_d_".to_cmangle == "_d_95d"
545 # assert "_42".to_cmangle == "_95d42"
546 # assert "foo".to_cmangle == "foo"
547 # assert "".to_cmangle == ""
548 fun to_cmangle
: String
550 if is_empty
then return ""
552 var underscore
= false
556 if c
>= '0' and c
<= '9' then
558 res
.append
(c
.code_point
.to_s
)
562 for i
in [start
..length
[ do
564 if (c
>= 'a' and c
<= 'z') or (c
>='A' and c
<= 'Z') then
570 res
.append
('_'.code_point
.to_s
)
573 if c
>= '0' and c
<= '9' then
576 else if c
== '_' then
581 res
.append
(c
.code_point
.to_s
)
587 res
.append
('_'.code_point
.to_s
)
593 # Escape " \ ' and non printable characters using the rules of literal C strings and characters
595 # assert "abAB12<>&".escape_to_c == "abAB12<>&"
596 # assert "\n\"'\\".escape_to_c == "\\n\\\"\\'\\\\"
598 # Most non-printable characters (bellow ASCII 32) are escaped to an octal form `\nnn`.
599 # Three digits are always used to avoid following digits to be interpreted as an element
600 # of the octal sequence.
602 # assert "{0.code_point}{1.code_point}{8.code_point}{31.code_point}{32.code_point}".escape_to_c == "\\000\\001\\010\\037 "
604 # The exceptions are the common `\t` and `\n`.
605 fun escape_to_c
: String
608 for i
in [0..length
[ do
612 else if c
== '\t' then
614 else if c
== '"' then
616 else if c == '\'' then
618 else if c == '\\
' then
620 else if c.code_point < 32 then
622 var oct = c.code_point.to_base(8)
623 # Force 3 octal digits since it is the
624 # maximum allowed in the C specification
625 if oct.length == 1 then
628 else if oct.length == 2 then
639 # Escape additionnal characters
640 # The result might no be legal in C but be used in other languages
642 # assert "ab|\{\}".escape_more_to_c("|\{\}") == "ab\\|\\\{\\\}"
643 fun escape_more_to_c(chars: String): String
646 for c in escape_to_c.chars do
647 if chars.chars.has(c) then
655 # Escape to C plus braces
657 # assert "\n\"'\\\
{\}".escape_to_nit == "\\n\\\
"\\'\\\\\\\{\\\}"
658 fun escape_to_nit
: String do return escape_more_to_c
("\{\}")
660 # Escape to POSIX Shell (sh).
662 # Abort if the text contains a null byte.
664 # assert "\n\"'\\\{\}0".escape_to_sh == "'\n\"'\\''\\\{\}0'"
665 fun escape_to_sh
: String do
668 for i in [0..length[ do
673 assert without_null_byte
: c
!= '\0'
681 # Escape to include in a Makefile
683 # Unfortunately, some characters are not escapable in Makefile.
684 # These characters are `;`, `|`, `\`, and the non-printable ones.
685 # They will be rendered as `"?{hex}"`.
686 fun escape_to_mk: String do
688 for i in [0..length[ do
692 else if c == ':' or c == ' ' or c == '#' then
695 else if c
.code_point
< 32 or c
== ';' or c
== '|' or c
== '\\' or c
== '=' then
696 b
.append
("?{c.code_point.to_base(16)}")
704 # Return a string where Nit escape sequences are transformed.
707 # assert s.length == 2
708 # var u = s.unescape_nit
709 # assert u.length == 1
710 # assert u.chars[0].code_point == 10 # (the ASCII value of the "new line" character)
711 fun unescape_nit
: String
713 var res
= new Buffer.with_cap
(self.length
)
714 var was_slash
= false
715 for i
in [0..length
[ do
717 if not was_slash
then
728 else if c
== 'r' then
730 else if c
== 't' then
732 else if c
== '0' then
741 # Returns `self` with all characters escaped with their UTF-16 representation
743 # assert "Aèあ𐏓".escape_to_utf16 == "\\u0041\\u00e8\\u3042\\ud800\\udfd3"
744 fun escape_to_utf16
: String do
746 for i
in chars
do buf
.append i
.escape_to_utf16
750 # Returns the Unicode char escaped by `self`
752 # assert "\\u0041".from_utf16_escape == 'A'
753 # assert "\\ud800\\udfd3".from_utf16_escape == '𐏓'
754 # assert "\\u00e8".from_utf16_escape == 'è'
755 # assert "\\u3042".from_utf16_escape == 'あ'
756 fun from_utf16_escape
(pos
, ln
: nullable Int): Char do
757 if pos
== null then pos
= 0
758 if ln
== null then ln
= length
- pos
759 if ln
< 6 then return 0xFFFD.code_point
760 var cp
= from_utf16_digit
(pos
+ 2)
761 if cp
< 0xD800 then return cp
.code_point
762 if cp
> 0xDFFF then return cp
.code_point
763 if cp
> 0xDBFF then return 0xFFFD.code_point
764 if ln
== 6 then return 0xFFFD.code_point
765 if ln
< 12 then return 0xFFFD.code_point
767 cp
+= from_utf16_digit
(pos
+ 8)
768 var cplo
= cp
& 0xFFFF
769 if cplo
< 0xDC00 then return 0xFFFD.code_point
770 if cplo
> 0xDFFF then return 0xFFFD.code_point
771 return cp
.from_utf16_surr
.code_point
774 # Returns a UTF-16 escape value
776 # var s = "\\ud800\\udfd3"
777 # assert s.from_utf16_digit(2) == 0xD800
778 # assert s.from_utf16_digit(8) == 0xDFD3
779 fun from_utf16_digit
(pos
: nullable Int): Int do
780 if pos
== null then pos
= 0
781 return to_hex
(pos
, 4)
784 # Encode `self` to percent (or URL) encoding
786 # assert "aBc09-._~".to_percent_encoding == "aBc09-._~"
787 # assert "%()< >".to_percent_encoding == "%25%28%29%3c%20%3e"
788 # assert ".com/post?e=asdf&f=123".to_percent_encoding == ".com%2fpost%3fe%3dasdf%26f%3d123"
789 # assert "éあいう".to_percent_encoding == "%c3%a9%e3%81%82%e3%81%84%e3%81%86"
790 fun to_percent_encoding
: String
794 for i
in [0..length
[ do
796 if (c
>= '0' and c
<= '9') or
797 (c
>= 'a' and c
<= 'z') or
798 (c
>= 'A' and c
<= 'Z') or
799 c
== '-' or c
== '.' or
804 var bytes
= c
.to_s
.bytes
805 for b
in bytes
do buf
.append
"%{b.to_i.to_hex}"
812 # Decode `self` from percent (or URL) encoding to a clear string
814 # Replace invalid use of '%' with '?'.
816 # assert "aBc09-._~".from_percent_encoding == "aBc09-._~"
817 # assert "%25%28%29%3c%20%3e".from_percent_encoding == "%()< >"
818 # assert ".com%2fpost%3fe%3dasdf%26f%3d123".from_percent_encoding == ".com/post?e=asdf&f=123"
819 # assert "%25%28%29%3C%20%3E".from_percent_encoding == "%()< >"
820 # assert "incomplete %".from_percent_encoding == "incomplete ?"
821 # assert "invalid % usage".from_percent_encoding == "invalid ? usage"
822 # assert "%c3%a9%e3%81%82%e3%81%84%e3%81%86".from_percent_encoding == "éあいう"
823 fun from_percent_encoding
: String
826 var has_percent
= false
834 # If no transformation is needed, return self as a string
835 if not has_percent
then return to_s
837 var buf
= new NativeString(len
)
843 if i
+ 2 >= length
then
844 # What follows % has been cut off
848 var hex_s
= substring
(i
, 2)
850 var hex_i
= hex_s
.to_hex
854 # What follows a % is not Hex
859 else buf
[l
] = c
.ascii
865 return buf
.to_s_unsafe
(l
)
868 # Escape the characters `<`, `>`, `&`, `"`, `'` and `/` as HTML/XML entity references.
870 # assert "a&b-<>\"x\"/'".html_escape == "a&b-<>"x"/'"
872 # SEE: <https://www.owasp.org/index.php/XSS_%28Cross_Site_Scripting%29_Prevention_Cheat_Sheet#RULE_.231_-_HTML_Escape_Before_Inserting_Untrusted_Data_into_HTML_Element_Content>
873 fun html_escape
: String
877 for i
in [0..length
[ do
881 else if c
== '<' then
883 else if c
== '>' then
885 else if c
== '"' then
887 else if c
== '\'' then
889 else if c == '/' then
898 # Two pieces of text are equals if thez have the same characters in the same order.
900 # assert "hello" == "hello"
901 # assert "hello" != "HELLO"
902 # assert "hello" == "hel"+"lo"
904 # Things that are not Text are not equal.
907 # assert "9" != ['9']
910 # assert "9".chars.first == '9' # equality of Char
911 # assert "9".chars == ['9'] # equality of Sequence
912 # assert "9".to_i == 9 # equality of Int
915 if o == null then return false
916 if not o isa Text then return false
917 if self.is_same_instance(o) then return true
918 if self.length != o.length then return false
919 return self.chars == o.chars
922 # Lexicographical comparaison
924 # assert "abc" < "xy"
925 # assert "ABC" < "abc"
928 var self_chars = self.chars.iterator
929 var other_chars = other.chars.iterator
931 while self_chars.is_ok and other_chars.is_ok do
932 if self_chars.item < other_chars.item then return true
933 if self_chars.item > other_chars.item then return false
938 if self_chars.is_ok then
945 # Escape string used in labels for graphviz
947 # assert ">><<".escape_to_dot == "\\>\\>\\<\\<"
948 fun escape_to_dot: String
950 return escape_more_to_c("|\{\}<>")
953 private var hash_cache: nullable Int = null
957 if hash_cache == null then
958 # djb2 hash algorithm
961 for i in [0..length[ do
963 h = (h << 5) + h + char.code_point
968 return hash_cache.as(not null)
971 # Format `self` by replacing each `%n` with the `n`th item of `args`
973 # The character `%` followed by something other than a number are left as is.
974 # To represent a `%` followed by a number, double the `%`, as in `%%7`.
976 # assert "This %0 is a %1.".format("String", "formatted String") == "This String is a formatted String."
977 # assert "Do not escape % nor %%1".format("unused") == "Do not escape % nor %1"
978 fun format(args: Object...): String do
979 var s = new Array[Text]
983 if self[i] == '%' then
987 while i < length and self[i].is_numeric do
991 var ciph_len = i - ciph_st
992 if ciph_len == 0 then
993 # What follows '%' is not a number.
994 s.push substring(curr_st, i - curr_st)
995 if i < length and self[i] == '%' then
1003 var arg_index = substring(ciph_st, ciph_len).to_i
1004 if arg_index >= args.length then continue
1006 s.push substring(curr_st, fmt_st - curr_st)
1007 s.push args[arg_index].to_s
1014 s.push substring(curr_st, length - curr_st)
1018 # Return the Levenshtein distance between two strings
1021 # assert "abcd".levenshtein_distance("abcd") == 0
1022 # assert "".levenshtein_distance("abcd") == 4
1023 # assert "abcd".levenshtein_distance("") == 4
1024 # assert "abcd".levenshtein_distance("xyz") == 4
1025 # assert "abcd".levenshtein_distance("xbdy") == 3
1027 fun levenshtein_distance(other: String): Int
1029 var slen = self.length
1030 var olen = other.length
1033 if slen == 0 then return olen
1034 if olen == 0 then return slen
1035 if self == other then return 0
1037 # previous row of distances
1038 var v0 = new Array[Int].with_capacity(olen+1)
1040 # current row of distances
1041 var v1 = new Array[Int].with_capacity(olen+1)
1043 for j in [0..olen] do
1044 # prefix insert cost
1048 for i in [0..slen[ do
1050 # prefix delete cost
1053 for j in [0..olen[ do
1055 var cost1 = v1[j] + 1
1057 var cost2 = v0[j + 1] + 1
1058 # same char cost (+0)
1061 if self[i] != other[j] then cost3 += 1
1063 v1[j+1] = cost1.min(cost2).min(cost3)
1067 # * v1 become v0 in the next iteration
1068 # * old v0 is reused as the new v1
1077 # Copies `n` bytes from `self` at `src_offset` into `dest` starting at `dest_offset`
1079 # Basically a high-level synonym of NativeString::copy_to
1081 # REQUIRE: `n` must be large enough to contain `len` bytes
1083 # var ns = new NativeString(8)
1084 # "Text is String".copy_to_native(ns, 8, 2, 0)
1085 # assert ns.to_s_unsafe(8) == "xt is St"
1087 fun copy_to_native(dest: NativeString, n, src_offset, dest_offset: Int) do
1088 var mypos = src_offset
1089 var itspos = dest_offset
1091 dest[itspos] = self.bytes[mypos]
1100 # All kinds of array-based text representations.
1101 abstract class FlatText
1104 # Underlying C-String (`char*`)
1106 # Warning : Might be void in some subclasses, be sure to check
1107 # if set before using it.
1108 var items: NativeString is noinit
1110 # Returns a char* starting at position `first_byte`
1112 # WARNING: If you choose to use this service, be careful of the following.
1114 # Strings and NativeString are *ideally* always allocated through a Garbage Collector.
1115 # Since the GC tracks the use of the pointer for the beginning of the char*, it may be
1116 # deallocated at any moment, rendering the pointer returned by this function invalid.
1117 # Any access to freed memory may very likely cause undefined behaviour or a crash.
1118 # (Failure to do so will most certainly result in long and painful debugging hours)
1120 # The only safe use of this pointer is if it is ephemeral (e.g. read in a C function
1121 # then immediately return).
1123 # As always, do not modify the content of the String in C code, if this is what you want
1124 # copy locally the char* as Nit Strings are immutable.
1125 fun fast_cstring: NativeString is abstract
1127 redef var length = 0
1129 redef var bytelen = 0
1140 redef fun copy_to_native(dest, n, src_offset, dest_offset) do
1141 items.copy_to(dest, n, src_offset, dest_offset)
1145 # Abstract class for the SequenceRead compatible
1146 # views on the chars of any Text
1147 private abstract class StringCharView
1148 super SequenceRead[Char]
1152 var target: SELFTYPE
1154 redef fun is_empty do return target.is_empty
1156 redef fun length do return target.length
1158 redef fun iterator: IndexedIterator[Char] do return self.iterator_from(0)
1160 redef fun reverse_iterator do return self.reverse_iterator_from(self.length - 1)
1163 # Abstract class for the SequenceRead compatible
1164 # views on the bytes of any Text
1165 private abstract class StringByteView
1166 super SequenceRead[Byte]
1170 var target: SELFTYPE
1172 redef fun is_empty do return target.is_empty
1174 redef fun length do return target.bytelen
1176 redef fun iterator do return self.iterator_from(0)
1178 redef fun reverse_iterator do return self.reverse_iterator_from(target.bytelen - 1)
1181 # Immutable sequence of characters.
1183 # String objects may be created using literals.
1185 # assert "Hello World!" isa String
1186 abstract class String
1189 redef type SELFTYPE: String is fixed
1191 redef fun to_s do return self
1193 # Concatenates `o` to `self`
1195 # assert "hello" + "world" == "helloworld"
1196 # assert "" + "hello" + "" == "hello"
1197 fun +(o: Text): SELFTYPE is abstract
1199 # Concatenates self `i` times
1201 # assert "abc" * 4 == "abcabcabcabc"
1202 # assert "abc" * 1 == "abc"
1203 # assert "abc" * 0 == ""
1204 fun *(i: Int): SELFTYPE is abstract
1206 # Insert `s` at `pos`.
1208 # assert "helloworld".insert_at(" ", 5) == "hello world"
1209 fun insert_at(s: String, pos: Int): SELFTYPE is abstract
1211 redef fun substrings is abstract
1213 # Returns a reversed version of self
1215 # assert "hello".reversed == "olleh"
1216 # assert "bob".reversed == "bob"
1217 # assert "".reversed == ""
1218 fun reversed: SELFTYPE is abstract
1220 # A upper case version of `self`
1222 # assert "Hello World!".to_upper == "HELLO WORLD!"
1223 fun to_upper: SELFTYPE is abstract
1225 # A lower case version of `self`
1227 # assert "Hello World!".to_lower == "hello world!"
1228 fun to_lower : SELFTYPE is abstract
1230 # Takes a camel case `self` and converts it to snake case
1232 # assert "randomMethodId".to_snake_case == "random_method_id"
1234 # The rules are the following:
1236 # An uppercase is always converted to a lowercase
1238 # assert "HELLO_WORLD".to_snake_case == "hello_world"
1240 # An uppercase that follows a lowercase is prefixed with an underscore
1242 # assert "HelloTheWORLD".to_snake_case == "hello_the_world"
1244 # An uppercase that follows an uppercase and is followed by a lowercase, is prefixed with an underscore
1246 # assert "HelloTHEWorld".to_snake_case == "hello_the_world"
1248 # All other characters are kept as is; `self` does not need to be a proper CamelCased string.
1250 # assert "=-_H3ll0Th3W0rld_-=".to_snake_case == "=-_h3ll0th3w0rld_-="
1251 fun to_snake_case: SELFTYPE
1253 if self.is_lower then return self
1255 var new_str = new Buffer.with_cap(self.length)
1256 var prev_is_lower = false
1257 var prev_is_upper = false
1259 for i in [0..length[ do
1261 if char.is_lower then
1263 prev_is_lower = true
1264 prev_is_upper = false
1265 else if char.is_upper then
1266 if prev_is_lower then
1268 else if prev_is_upper and i+1 < length and chars[i+1].is_lower then
1271 new_str.add(char.to_lower)
1272 prev_is_lower = false
1273 prev_is_upper = true
1276 prev_is_lower = false
1277 prev_is_upper = false
1284 # Takes a snake case `self` and converts it to camel case
1286 # assert "random_method_id".to_camel_case == "randomMethodId"
1288 # If the identifier is prefixed by an underscore, the underscore is ignored
1290 # assert "_private_field".to_camel_case == "_privateField"
1292 # If `self` is upper, it is returned unchanged
1294 # assert "RANDOM_ID".to_camel_case == "RANDOM_ID"
1296 # If there are several consecutive underscores, they are considered as a single one
1298 # assert "random__method_id".to_camel_case == "randomMethodId"
1299 fun to_camel_case: SELFTYPE
1301 if self.is_upper then return self
1303 var new_str = new Buffer
1304 var is_first_char = true
1305 var follows_us = false
1307 for i in [0..length[ do
1309 if is_first_char then
1311 is_first_char = false
1312 else if char == '_
' then
1314 else if follows_us then
1315 new_str.add(char.to_upper)
1325 # Returns a capitalized `self`
1327 # Letters that follow a letter are lowercased
1328 # Letters that follow a non-letter are upcased.
1330 # SEE : `Char::is_letter` for the definition of letter.
1332 # assert "jAVASCRIPT".capitalized == "Javascript"
1333 # assert "i am root".capitalized == "I Am Root"
1334 # assert "ab_c -ab0c ab\nc".capitalized == "Ab_C -Ab0C Ab\nC"
1335 fun capitalized: SELFTYPE do
1336 if length == 0 then return self
1338 var buf = new Buffer.with_cap(length)
1340 var curr = chars[0].to_upper
1344 for i in [1 .. length[ do
1347 if prev.is_letter then
1348 buf[i] = curr.to_lower
1350 buf[i] = curr.to_upper
1358 # A mutable sequence of characters.
1359 abstract class Buffer
1362 # Returns an arbitrary subclass of `Buffer` with default parameters
1365 # Returns an instance of a subclass of `Buffer` with `i` base capacity
1366 new with_cap(i: Int) is abstract
1368 redef type SELFTYPE: Buffer is fixed
1370 # Specific implementations MUST set this to `true` in order to invalidate caches
1371 protected var is_dirty = true
1373 # Copy-On-Write flag
1375 # If the `Buffer` was to_s'd
, the next in-place altering
1376 # operation will cause the current `Buffer` to be re-allocated.
1378 # The flag will then be set at `false`.
1379 protected var written
= false
1381 # Modifies the char contained at pos `index`
1383 # DEPRECATED : Use self.chars.[]= instead
1384 fun []=(index
: Int, item
: Char) is abstract
1386 # Adds a char `c` at the end of self
1388 # DEPRECATED : Use self.chars.add instead
1389 fun add
(c
: Char) is abstract
1393 # var b = new Buffer
1395 # assert not b.is_empty
1398 fun clear
is abstract
1400 # Enlarges the subsequent array containing the chars of self
1401 fun enlarge
(cap
: Int) is abstract
1403 # Adds the content of text `s` at the end of self
1405 # var b = new Buffer
1408 # assert b == "helloworld"
1409 fun append
(s
: Text) is abstract
1411 # `self` is appended in such a way that `self` is repeated `r` times
1413 # var b = new Buffer
1416 # assert b == "hellohellohello"
1417 fun times
(r
: Int) is abstract
1419 # Reverses itself in-place
1421 # var b = new Buffer
1424 # assert b == "olleh"
1425 fun reverse
is abstract
1427 # Changes each lower-case char in `self` by its upper-case variant
1429 # var b = new Buffer
1430 # b.append("Hello World!")
1432 # assert b == "HELLO WORLD!"
1433 fun upper
is abstract
1435 # Changes each upper-case char in `self` by its lower-case variant
1437 # var b = new Buffer
1438 # b.append("Hello World!")
1440 # assert b == "hello world!"
1441 fun lower
is abstract
1443 # Capitalizes each word in `self`
1445 # Letters that follow a letter are lowercased
1446 # Letters that follow a non-letter are upcased.
1448 # SEE: `Char::is_letter` for the definition of a letter.
1450 # var b = new FlatBuffer.from("jAVAsCriPt")
1452 # assert b == "Javascript"
1453 # b = new FlatBuffer.from("i am root")
1455 # assert b == "I Am Root"
1456 # b = new FlatBuffer.from("ab_c -ab0c ab\nc")
1458 # assert b == "Ab_C -Ab0C Ab\nC"
1460 if length
== 0 then return
1461 var c
= self[0].to_upper
1464 for i
in [1 .. length
[ do
1467 if prev
.is_letter
then
1468 self[i
] = c
.to_lower
1470 self[i
] = c
.to_upper
1477 if is_dirty
then hash_cache
= null
1481 # In Buffers, the internal sequence of character is mutable
1482 # Thus, `chars` can be used to modify the buffer.
1483 redef fun chars
: Sequence[Char] is abstract
1486 # View for chars on Buffer objects, extends Sequence
1487 # for mutation operations
1488 private abstract class BufferCharView
1489 super StringCharView
1490 super Sequence[Char]
1492 redef type SELFTYPE: Buffer
1496 # View for bytes on Buffer objects, extends Sequence
1497 # for mutation operations
1498 private abstract class BufferByteView
1499 super StringByteView
1501 redef type SELFTYPE: Buffer
1505 # User readable representation of `self`.
1506 fun to_s
: String do return inspect
1508 # The class name of the object in NativeString format.
1509 private fun native_class_name
: NativeString is intern
1511 # The class name of the object.
1513 # assert 5.class_name == "Int"
1514 fun class_name
: String do return native_class_name
.to_s
1516 # Developer readable representation of `self`.
1517 # Usually, it uses the form "<CLASSNAME:#OBJECTID bla bla bla>"
1520 return "<{inspect_head}>"
1523 # Return "CLASSNAME:#OBJECTID".
1524 # This function is mainly used with the redefinition of the inspect method
1525 protected fun inspect_head
: String
1527 return "{class_name}:#{object_id.to_hex}"
1532 # assert true.to_s == "true"
1533 # assert false.to_s == "false"
1545 # C function to calculate the length of the `NativeString` to receive `self`
1546 private fun byte_to_s_len
: Int `{
1547 return snprintf(NULL, 0, "0x%02x", self);
1550 # C function to convert an nit Int to a NativeString (char*)
1551 private fun native_byte_to_s
(nstr
: NativeString, strlen
: Int) `{
1552 snprintf(nstr, strlen, "0x%02x", self);
1555 # Displayable byte in its hexadecimal form (0x..)
1557 # assert 1.to_b.to_s == "0x01"
1558 # assert (-123).to_b.to_s == "0x85"
1560 var nslen
= byte_to_s_len
1561 var ns
= new NativeString(nslen
+ 1)
1563 native_byte_to_s
(ns
, nslen
+ 1)
1564 return ns
.to_s_unsafe
(nslen
)
1570 # Wrapper of strerror C function
1571 private fun strerror_ext
: NativeString `{ return strerror((int)self); `}
1573 # Returns a string describing error number
1574 fun strerror: String do return strerror_ext.to_s
1576 # Fill `s
` with the digits in base `base
` of `self` (and with the '-' sign if negative).
1577 # assume < to_c max const of char
1578 private fun fill_buffer(s: Buffer, base: Int)
1585 else if self == 0 then
1592 var pos = digit_count(base) - 1
1593 while pos >= 0 and n > 0 do
1594 s.chars[pos] = (n % base).to_c
1600 # C function to calculate the length of the `NativeString` to receive `self`
1601 private fun int_to_s_len: Int `{
1602 return snprintf
(NULL, 0, "%ld", self);
1605 # C function to convert an nit Int to a NativeString (char*)
1606 private fun native_int_to_s(nstr: NativeString, strlen: Int) `{
1607 snprintf
(nstr
, strlen
, "%ld", self);
1610 # String representation of `self` in the given `base
`
1613 # assert 15.to_base(10) == "15"
1614 # assert 15.to_base(16) == "f"
1615 # assert 15.to_base(2) == "1111"
1616 # assert (-10).to_base(3) == "-101"
1618 fun to_base(base: Int): String
1620 var l = digit_count(base)
1623 for x in [0..l[ do s.add(' ')
1624 fill_buffer(s, base)
1629 # return displayable int in hexadecimal
1631 # assert 1.to_hex == "1"
1632 # assert (-255).to_hex == "-ff"
1633 fun to_hex: String do return to_base(16)
1637 # Pretty representation of `self`, with decimals as needed from 1 to a maximum of 3
1639 # assert 12.34.to_s == "12.34"
1640 # assert (-0120.030).to_s == "-120.03"
1642 # see `to_precision
` for a custom precision.
1644 var str = to_precision( 3 )
1645 if is_inf != 0 or is_nan then return str
1646 var len = str.length
1647 for i in [0..len-1] do
1649 var c = str.chars[j]
1652 else if c == '.' then
1653 return str.substring( 0, j+2 )
1655 return str.substring( 0, j+1 )
1661 # `String` representation of `self` with the given number of `decimals
`
1663 # assert 12.345.to_precision(0) == "12"
1664 # assert 12.345.to_precision(3) == "12.345"
1665 # assert (-12.345).to_precision(3) == "-12.345"
1666 # assert (-0.123).to_precision(3) == "-0.123"
1667 # assert 0.999.to_precision(2) == "1.00"
1668 # assert 0.999.to_precision(4) == "0.9990"
1669 fun to_precision(decimals: Int): String
1671 if is_nan then return "nan"
1673 var isinf = self.is_inf
1676 else if isinf == -1 then
1680 if decimals == 0 then return self.to_i.to_s
1682 for i in [0..decimals[ do f = f * 10.0
1689 if i == 0 then return "0." + "0"*decimals
1691 # Prepare both parts of the float, before and after the "."
1696 if sl > decimals then
1697 # Has something before the "."
1698 p1 = s.substring(0, sl-decimals)
1699 p2 = s.substring(sl-decimals, decimals)
1702 p2 = "0"*(decimals-sl) + s
1705 if i < 0 then p1 = "-" + p1
1707 return p1 + "." + p2
1713 # Returns a sequence with the UTF-8 bytes of `self`
1715 # assert 'a'.bytes == [0x61u8]
1716 # assert 'ま'.bytes == [0xE3u8, 0x81u8, 0xBEu8]
1717 fun bytes: SequenceRead[Byte] do return to_s.bytes
1719 # Is `self` an UTF-16 surrogate pair ?
1720 fun is_surrogate: Bool do
1722 return cp >= 0xD800 and cp <= 0xDFFF
1725 # Length of `self` in a UTF-8 String
1726 private fun u8char_len: Int do
1727 var c = self.code_point
1728 if c < 0x80 then return 1
1729 if c <= 0x7FF then return 2
1730 if c <= 0xFFFF then return 3
1731 if c <= 0x10FFFF then return 4
1732 # Bad character format
1736 # assert 'x'.to_s == "x"
1739 var ns = new NativeString(ln + 1)
1741 return ns.to_s_unsafe(ln)
1744 # Returns `self` escaped to UTF-16
1746 # i.e. Represents `self`.`code_point
` using UTF-16 codets escaped
1749 # assert 'A'.escape_to_utf16 == "\\u0041"
1750 # assert 'è'.escape_to_utf16 == "\\u00e8"
1751 # assert 'あ'.escape_to_utf16 == "\\u3042"
1752 # assert '𐏓'.escape_to_utf16 == "\\ud800\\udfd3"
1753 fun escape_to_utf16: String do
1756 if cp < 0xD800 or (cp >= 0xE000 and cp <= 0xFFFF) then
1757 buf = new Buffer.with_cap(6)
1758 buf.append("\\u0000")
1761 for i in hx.chars.reverse_iterator do
1766 buf = new Buffer.with_cap(12)
1767 buf.append("\\u0000\\u0000")
1768 var lo = (((cp - 0x10000) & 0x3FF) + 0xDC00).to_hex
1769 var hi = ((((cp - 0x10000) & 0xFFC00) >> 10) + 0xD800).to_hex
1784 private fun u8char_tos(r: NativeString, len: Int) `{
1791 r
[0] = 0xC0 | ((self & 0x7C0) >> 6);
1792 r
[1] = 0x80 | (self & 0x3F);
1795 r
[0] = 0xE0 | ((self & 0xF000) >> 12);
1796 r
[1] = 0x80 | ((self & 0xFC0) >> 6);
1797 r
[2] = 0x80 | (self & 0x3F);
1800 r
[0] = 0xF0 | ((self & 0x1C0000) >> 18);
1801 r
[1] = 0x80 | ((self & 0x3F000) >> 12);
1802 r
[2] = 0x80 | ((self & 0xFC0) >> 6);
1803 r
[3] = 0x80 | (self & 0x3F);
1808 # Returns true if the char is a numerical digit
1810 # assert '0'.is_numeric
1811 # assert '9'.is_numeric
1812 # assert not 'a'.is_numeric
1813 # assert not '?'.is_numeric
1815 # FIXME: Works on ASCII-range only
1816 fun is_numeric: Bool
1818 return self >= '0' and self <= '9'
1821 # Returns true if the char is an alpha digit
1823 # assert 'a'.is_alpha
1824 # assert 'Z'.is_alpha
1825 # assert not '0'.is_alpha
1826 # assert not '?'.is_alpha
1828 # FIXME: Works on ASCII-range only
1831 return (self >= 'a' and self <= 'z') or (self >= 'A' and self <= 'Z')
1834 # Is `self` an hexadecimal digit ?
1836 # assert 'A'.is_hexdigit
1837 # assert not 'G'.is_hexdigit
1838 # assert 'a'.is_hexdigit
1839 # assert not 'g'.is_hexdigit
1840 # assert '5'.is_hexdigit
1841 fun is_hexdigit: Bool do return (self >= '0' and self <= '9') or (self >= 'A' and self <= 'F') or
1842 (self >= 'a' and self <= 'f')
1844 # Returns true if the char is an alpha or a numeric digit
1846 # assert 'a'.is_alphanumeric
1847 # assert 'Z'.is_alphanumeric
1848 # assert '0'.is_alphanumeric
1849 # assert '9'.is_alphanumeric
1850 # assert not '?'.is_alphanumeric
1852 # FIXME: Works on ASCII-range only
1853 fun is_alphanumeric: Bool
1855 return self.is_numeric or self.is_alpha
1858 # Returns `self` to its int value
1860 # REQUIRE: `is_hexdigit
`
1861 fun from_hex: Int do
1862 if self >= '0' and self <= '9' then return code_point - 0x30
1863 if self >= 'A' and self <= 'F' then return code_point - 0x37
1864 if self >= 'a' and self <= 'f' then return code_point - 0x57
1865 # Happens if self is not a hexdigit
1866 assert self.is_hexdigit
1867 # To make flow analysis happy
1872 redef class Collection[E]
1873 # String representation of the content of the collection.
1875 # The standard representation is the list of elements separated with commas.
1878 # assert [1,2,3].to_s == "[1,2,3]"
1879 # assert [1..3].to_s == "[1,2,3]"
1880 # assert (new Array[Int]).to_s == "[]" # empty collection
1883 # Subclasses may return a more specific string representation.
1886 return "[" + join(",") + "]"
1889 # Concatenate elements without separators
1892 # assert [1,2,3].plain_to_s == "123"
1893 # assert [11..13].plain_to_s == "111213"
1894 # assert (new Array[Int]).plain_to_s == "" # empty collection
1896 fun plain_to_s: String
1899 for e in self do if e != null then s.append(e.to_s)
1903 # Concatenate and separate each elements with `separator
`.
1905 # Only concatenate if `separator
== null`.
1907 # assert [1, 2, 3].join(":") == "1:2:3"
1908 # assert [1..3].join(":") == "1:2:3"
1909 # assert [1..3].join == "123"
1911 # if `last_separator
` is given, then it is used to separate the last element.
1913 # assert [1, 2, 3, 4].join(", ", " and ") == "1, 2, 3 and 4"
1914 fun join(separator: nullable Text, last_separator: nullable Text): String
1916 if is_empty then return ""
1918 var s = new Buffer # Result
1923 if e != null then s.append(e.to_s)
1925 if last_separator == null then last_separator = separator
1927 # Concat other items
1933 if separator != null then s.append(separator)
1935 if last_separator != null then s.append(last_separator)
1937 if e != null then s.append(e.to_s)
1943 redef class Map[K,V]
1944 # Concatenate couples of key value.
1945 # Key and value are separated by `couple_sep
`.
1946 # Couples are separated by `sep
`.
1948 # var m = new HashMap[Int, String]
1951 # assert m.join("; ", "=") == "1=one; 10=ten"
1952 fun join(sep, couple_sep: String): String is abstract
1956 private var args_cache: nullable Sequence[String] = null
1958 # The arguments of the program as given by the OS
1959 fun program_args: Sequence[String]
1961 if _args_cache == null then init_args
1962 return _args_cache.as(not null)
1965 # The name of the program as given by the OS
1966 fun program_name: String
1968 return native_argv(0).to_s
1971 # Initialize `program_args
` with the contents of `native_argc
` and `native_argv
`.
1972 private fun init_args
1974 var argc = native_argc
1975 var args = new Array[String].with_capacity(0)
1978 args[i-1] = native_argv(i).to_s
1984 # First argument of the main C function.
1985 private fun native_argc: Int is intern
1987 # Second argument of the main C function.
1988 private fun native_argv(i: Int): NativeString is intern
1991 # Comparator that efficienlty use `to_s
` to compare things
1993 # The comparaison call `to_s
` on object and use the result to order things.
1995 # var a = [1, 2, 3, 10, 20]
1996 # (new CachedAlphaComparator).sort(a)
1997 # assert a == [1, 10, 2, 20, 3]
1999 # Internally the result of `to_s
` is cached in a HashMap to counter
2000 # uneficient implementation of `to_s
`.
2002 # Note: it caching is not usefull, see `alpha_comparator
`
2003 class CachedAlphaComparator
2005 redef type COMPARED: Object
2007 private var cache = new HashMap[Object, String]
2009 private fun do_to_s(a: Object): String do
2010 if cache.has_key(a) then return cache[a]
2016 redef fun compare(a, b) do
2017 return do_to_s(a) <=> do_to_s(b)
2021 # see `alpha_comparator
`
2022 private class AlphaComparator
2024 redef fun compare(a, b) do return a.to_s <=> b.to_s
2027 # Stateless comparator that naively use `to_s
` to compare things.
2029 # Note: the result of `to_s
` is not cached, thus can be invoked a lot
2030 # on a single instace. See `CachedAlphaComparator` as an alternative.
2032 # var a = [1, 2, 3, 10, 20]
2033 # alpha_comparator.sort(a)
2034 # assert a == [1, 10, 2, 20, 3]
2035 fun alpha_comparator: Comparator do return once new AlphaComparator
2037 # The arguments of the program as given by the OS
2038 fun args: Sequence[String]
2040 return sys.program_args
2043 redef class NativeString
2044 # Get a `String` from the data at `self` copied into Nit memory
2046 # Require: `self` is a null-terminated string.
2047 fun to_s_with_copy: String is abstract
2049 # Get a `String` from `length
` bytes at `self`
2051 # The result may point to the data at `self` or
2052 # it may make a copy in Nit controlled memory.
2053 # This method should only be used when `self` was allocated by the Nit GC,
2054 # or when manually controlling the deallocation of `self`.
2055 fun to_s_with_length(length: Int): String is abstract
2057 # Get a `String` from the raw `length
` bytes at `self`
2059 # The default value of `length
` is the number of bytes before
2060 # the first null character.
2062 # The created `String` points to the data at `self`.
2063 # This method should be used when `self` was allocated by the Nit GC,
2064 # or when manually controlling the deallocation of `self`.
2066 # /!\: This service does not clean the items for compliance with UTF-8,
2067 # use only when the data has already been verified as valid UTF-8.
2068 fun to_s_unsafe(length: nullable Int): String is abstract
2070 # Get a `String` from the raw `bytelen
` bytes at `self` with `unilen
` Unicode characters
2072 # The created `String` points to the data at `self`.
2073 # This method should be used when `self` was allocated by the Nit GC,
2074 # or when manually controlling the deallocation of `self`.
2076 # /!\: This service does not clean the items for compliance with UTF-8,
2077 # use only when the data has already been verified as valid UTF-8.
2079 # SEE: `abstract_text
::Text` for more info on the difference
2080 # between `Text::bytelen
` and `Text::length
`.
2081 fun to_s_full(bytelen, unilen: Int): String is abstract
2084 redef class NativeArray[E]
2085 # Join all the elements using `to_s
`
2087 # REQUIRE: `self isa NativeArray[String]`
2088 # REQUIRE: all elements are initialized
2089 fun native_to_s: String is abstract