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 # Returns `true` if the string contains only Numeric values (and one "," or one "." character)
310 # assert "123".is_numeric == true
311 # assert "1.2".is_numeric == true
312 # assert "1,2".is_numeric == true
313 # assert "1..2".is_numeric == false
316 var has_point_or_comma
= false
317 for i
in [0..length
[ do
319 if not c
.is_numeric
then
320 if (c
== '.' or c
== ',') and not has_point_or_comma
then
321 has_point_or_comma
= true
330 # Returns `true` if the string contains only Hex chars
332 # assert "048bf".is_hex == true
333 # assert "ABCDEF".is_hex == true
334 # assert "0G".is_hex == false
337 for i
in [0..length
[ do
339 if not (c
>= 'a' and c
<= 'f') and
340 not (c
>= 'A' and c
<= 'F') and
341 not (c
>= '0' and c
<= '9') then return false
346 # Returns `true` if the string contains only Binary digits
348 # assert "1101100".is_bin == true
349 # assert "1101020".is_bin == false
351 for i
in chars
do if i
!= '0' and i
!= '1' then return false
355 # Returns `true` if the string contains only Octal digits
357 # assert "213453".is_oct == true
358 # assert "781".is_oct == false
360 for i
in chars
do if i
< '0' or i
> '7' then return false
364 # Returns `true` if the string contains only Decimal digits
366 # assert "10839".is_dec == true
367 # assert "164F".is_dec == false
369 for i
in chars
do if i
< '0' or i
> '9' then return false
373 # Are all letters in `self` upper-case ?
375 # assert "HELLO WORLD".is_upper == true
376 # assert "%$&%!".is_upper == true
377 # assert "hello world".is_upper == false
378 # assert "Hello World".is_upper == false
381 for i
in [0..length
[ do
383 if char
.is_lower
then return false
388 # Are all letters in `self` lower-case ?
390 # assert "hello world".is_lower == true
391 # assert "%$&%!".is_lower == true
392 # assert "Hello World".is_lower == false
395 for i
in [0..length
[ do
397 if char
.is_upper
then return false
402 # Removes the whitespaces at the beginning of self
404 # assert " \n\thello \n\t".l_trim == "hello \n\t"
406 # `Char::is_whitespace` determines what is a whitespace.
409 var iter
= self.chars
.iterator
411 if not iter
.item
.is_whitespace
then break
414 if iter
.index
== length
then return self.empty
415 return self.substring_from
(iter
.index
)
418 # Removes the whitespaces at the end of self
420 # assert " \n\thello \n\t".r_trim == " \n\thello"
422 # `Char::is_whitespace` determines what is a whitespace.
425 var iter
= self.chars
.reverse_iterator
427 if not iter
.item
.is_whitespace
then break
430 if iter
.index
< 0 then return self.empty
431 return self.substring
(0, iter
.index
+ 1)
434 # Trims trailing and preceding white spaces
436 # assert " Hello World ! ".trim == "Hello World !"
437 # assert "\na\nb\tc\t".trim == "a\nb\tc"
439 # `Char::is_whitespace` determines what is a whitespace.
440 fun trim
: SELFTYPE do return (self.l_trim
).r_trim
442 # Is the string non-empty but only made of whitespaces?
444 # assert " \n\t ".is_whitespace == true
445 # assert " hello ".is_whitespace == false
446 # assert "".is_whitespace == false
448 # `Char::is_whitespace` determines what is a whitespace.
449 fun is_whitespace
: Bool
451 if is_empty
then return false
452 for c
in self.chars
do
453 if not c
.is_whitespace
then return false
458 # Returns `self` removed from its last line terminator (if any).
460 # assert "Hello\n".chomp == "Hello"
461 # assert "Hello".chomp == "Hello"
463 # assert "\n".chomp == ""
464 # assert "".chomp == ""
466 # Line terminators are `"\n"`, `"\r\n"` and `"\r"`.
467 # A single line terminator, the last one, is removed.
469 # assert "\r\n".chomp == ""
470 # assert "\r\n\n".chomp == "\r\n"
471 # assert "\r\n\r\n".chomp == "\r\n"
472 # assert "\r\n\r".chomp == "\r\n"
474 # Note: unlike with most IO methods like `Reader::read_line`,
475 # a single `\r` is considered here to be a line terminator and will be removed.
479 if len
== 0 then return self
480 var l
= self.chars
.last
482 return substring
(0, len-1
)
483 else if l
!= '\n' then
485 else if len
> 1 and self.chars
[len-2
] == '\r' then
486 return substring
(0, len-2
)
488 return substring
(0, len-1
)
492 # Justify a self in a space of `length`
494 # `left` is the space ratio on the left side.
495 # * 0.0 for left-justified (no space at the left)
496 # * 1.0 for right-justified (all spaces at the left)
497 # * 0.5 for centered (half the spaces at the left)
501 # assert "hello".justify(10, 0.0) == "hello "
502 # assert "hello".justify(10, 1.0) == " hello"
503 # assert "hello".justify(10, 0.5) == " hello "
505 # If `length` is not enough, `self` is returned as is.
507 # assert "hello".justify(2, 0.0) == "hello"
509 # REQUIRE: `left >= 0.0 and left <= 1.0`
510 # ENSURE: `self.length <= length implies result.length == length`
511 # ENSURE: `self.length >= length implies result == self`
512 fun justify
(length
: Int, left
: Float): String
514 var diff
= length
- self.length
515 if diff
<= 0 then return to_s
516 assert left
>= 0.0 and left
<= 1.0
517 var before
= (diff
.to_f
* left
).to_i
518 return " " * before
+ self + " " * (diff-before
)
521 # Mangle a string to be a unique string only made of alphanumeric characters and underscores.
523 # This method is injective (two different inputs never produce the same
524 # output) and the returned string always respect the following rules:
526 # * Contains only US-ASCII letters, digits and underscores.
527 # * Never starts with a digit.
528 # * Never ends with an underscore.
529 # * Never contains two contiguous underscores.
531 # assert "42_is/The answer!".to_cmangle == "_52d2_is_47dThe_32danswer_33d"
532 # assert "__".to_cmangle == "_95d_95d"
533 # assert "__d".to_cmangle == "_95d_d"
534 # assert "_d_".to_cmangle == "_d_95d"
535 # assert "_42".to_cmangle == "_95d42"
536 # assert "foo".to_cmangle == "foo"
537 # assert "".to_cmangle == ""
538 fun to_cmangle
: String
540 if is_empty
then return ""
542 var underscore
= false
546 if c
>= '0' and c
<= '9' then
548 res
.append
(c
.code_point
.to_s
)
552 for i
in [start
..length
[ do
554 if (c
>= 'a' and c
<= 'z') or (c
>='A' and c
<= 'Z') then
560 res
.append
('_'.code_point
.to_s
)
563 if c
>= '0' and c
<= '9' then
566 else if c
== '_' then
571 res
.append
(c
.code_point
.to_s
)
577 res
.append
('_'.code_point
.to_s
)
583 # Escape " \ ' and non printable characters using the rules of literal C strings and characters
585 # assert "abAB12<>&".escape_to_c == "abAB12<>&"
586 # assert "\n\"'\\".escape_to_c == "\\n\\\"\\'\\\\"
588 # Most non-printable characters (bellow ASCII 32) are escaped to an octal form `\nnn`.
589 # Three digits are always used to avoid following digits to be interpreted as an element
590 # of the octal sequence.
592 # assert "{0.code_point}{1.code_point}{8.code_point}{31.code_point}{32.code_point}".escape_to_c == "\\000\\001\\010\\037 "
594 # The exceptions are the common `\t` and `\n`.
595 fun escape_to_c
: String
598 for i
in [0..length
[ do
602 else if c
== '\t' then
604 else if c
== '"' then
606 else if c == '\'' then
608 else if c == '\\
' then
610 else if c.code_point < 32 then
612 var oct = c.code_point.to_base(8, false)
613 # Force 3 octal digits since it is the
614 # maximum allowed in the C specification
615 if oct.length == 1 then
618 else if oct.length == 2 then
629 # Escape additionnal characters
630 # The result might no be legal in C but be used in other languages
632 # assert "ab|\{\}".escape_more_to_c("|\{\}") == "ab\\|\\\{\\\}"
633 fun escape_more_to_c(chars: String): String
636 for c in escape_to_c.chars do
637 if chars.chars.has(c) then
645 # Escape to C plus braces
647 # assert "\n\"'\\\
{\}".escape_to_nit == "\\n\\\
"\\'\\\\\\\{\\\}"
648 fun escape_to_nit
: String do return escape_more_to_c
("\{\}")
650 # Escape to POSIX Shell (sh).
652 # Abort if the text contains a null byte.
654 # assert "\n\"'\\\{\}0".escape_to_sh == "'\n\"'\\''\\\{\}0'"
655 fun escape_to_sh
: String do
658 for i in [0..length[ do
663 assert without_null_byte
: c
!= '\0'
671 # Escape to include in a Makefile
673 # Unfortunately, some characters are not escapable in Makefile.
674 # These characters are `;`, `|`, `\`, and the non-printable ones.
675 # They will be rendered as `"?{hex}"`.
676 fun escape_to_mk: String do
678 for i in [0..length[ do
682 else if c == ':' or c == ' ' or c == '#' then
685 else if c
.code_point
< 32 or c
== ';' or c
== '|' or c
== '\\' or c
== '=' then
686 b
.append
("?{c.code_point.to_base(16, false)}")
694 # Return a string where Nit escape sequences are transformed.
697 # assert s.length == 2
698 # var u = s.unescape_nit
699 # assert u.length == 1
700 # assert u.chars[0].code_point == 10 # (the ASCII value of the "new line" character)
701 fun unescape_nit
: String
703 var res
= new Buffer.with_cap
(self.length
)
704 var was_slash
= false
705 for i
in [0..length
[ do
707 if not was_slash
then
718 else if c
== 'r' then
720 else if c
== 't' then
722 else if c
== '0' then
731 # Returns `self` with all characters escaped with their UTF-16 representation
733 # assert "Aèあ𐏓".escape_to_utf16 == "\\u0041\\u00e8\\u3042\\ud800\\udfd3"
734 fun escape_to_utf16
: String do
736 for i
in chars
do buf
.append i
.escape_to_utf16
740 # Returns the Unicode char escaped by `self`
742 # assert "\\u0041".from_utf16_escape == 'A'
743 # assert "\\ud800\\udfd3".from_utf16_escape == '𐏓'
744 # assert "\\u00e8".from_utf16_escape == 'è'
745 # assert "\\u3042".from_utf16_escape == 'あ'
746 fun from_utf16_escape
: Char do
748 if ln
!= 6 and ln
!= 12 then return 0xFFFD.code_point
749 var cphi
= substring
(2, 4).to_hex
750 if cphi
< 0xD800 then return cphi
.code_point
751 if cphi
> 0xDFFF then return cphi
.code_point
752 if cphi
> 0xDBFF then return 0xFFFD.code_point
754 cp
+= (cphi
- 0xD800) << 10
755 var cplo
= substring
(8, 4).to_hex
756 if cplo
< 0xDC00 then return 0xFFFD.code_point
757 if cplo
> 0xDFFF then return 0xFFFD.code_point
763 # Encode `self` to percent (or URL) encoding
765 # assert "aBc09-._~".to_percent_encoding == "aBc09-._~"
766 # assert "%()< >".to_percent_encoding == "%25%28%29%3c%20%3e"
767 # assert ".com/post?e=asdf&f=123".to_percent_encoding == ".com%2fpost%3fe%3dasdf%26f%3d123"
768 # assert "éあいう".to_percent_encoding == "%c3%a9%e3%81%82%e3%81%84%e3%81%86"
769 fun to_percent_encoding
: String
773 for i
in [0..length
[ do
775 if (c
>= '0' and c
<= '9') or
776 (c
>= 'a' and c
<= 'z') or
777 (c
>= 'A' and c
<= 'Z') or
778 c
== '-' or c
== '.' or
783 var bytes
= c
.to_s
.bytes
784 for b
in bytes
do buf
.append
"%{b.to_i.to_hex}"
791 # Decode `self` from percent (or URL) encoding to a clear string
793 # Replace invalid use of '%' with '?'.
795 # assert "aBc09-._~".from_percent_encoding == "aBc09-._~"
796 # assert "%25%28%29%3c%20%3e".from_percent_encoding == "%()< >"
797 # assert ".com%2fpost%3fe%3dasdf%26f%3d123".from_percent_encoding == ".com/post?e=asdf&f=123"
798 # assert "%25%28%29%3C%20%3E".from_percent_encoding == "%()< >"
799 # assert "incomplete %".from_percent_encoding == "incomplete ?"
800 # assert "invalid % usage".from_percent_encoding == "invalid ? usage"
801 # assert "%c3%a9%e3%81%82%e3%81%84%e3%81%86".from_percent_encoding == "éあいう"
802 fun from_percent_encoding
: String
805 var has_percent
= false
813 # If no transformation is needed, return self as a string
814 if not has_percent
then return to_s
816 var buf
= new NativeString(len
)
822 if i
+ 2 >= length
then
823 # What follows % has been cut off
827 var hex_s
= substring
(i
, 2)
829 var hex_i
= hex_s
.to_hex
833 # What follows a % is not Hex
838 else buf
[l
] = c
.ascii
844 return buf
.to_s_with_length
(l
)
847 # Escape the characters `<`, `>`, `&`, `"`, `'` and `/` as HTML/XML entity references.
849 # assert "a&b-<>\"x\"/'".html_escape == "a&b-<>"x"/'"
851 # 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>
852 fun html_escape
: String
856 for i
in [0..length
[ do
860 else if c
== '<' then
862 else if c
== '>' then
864 else if c
== '"' then
866 else if c
== '\'' then
868 else if c == '/' then
877 # Two pieces of text are equals if thez have the same characters in the same order.
879 # assert "hello" == "hello"
880 # assert "hello" != "HELLO"
881 # assert "hello" == "hel"+"lo"
883 # Things that are not Text are not equal.
886 # assert "9" != ['9']
889 # assert "9".chars.first == '9' # equality of Char
890 # assert "9".chars == ['9'] # equality of Sequence
891 # assert "9".to_i == 9 # equality of Int
894 if o == null then return false
895 if not o isa Text then return false
896 if self.is_same_instance(o) then return true
897 if self.length != o.length then return false
898 return self.chars == o.chars
901 # Lexicographical comparaison
903 # assert "abc" < "xy"
904 # assert "ABC" < "abc"
907 var self_chars = self.chars.iterator
908 var other_chars = other.chars.iterator
910 while self_chars.is_ok and other_chars.is_ok do
911 if self_chars.item < other_chars.item then return true
912 if self_chars.item > other_chars.item then return false
917 if self_chars.is_ok then
924 # Escape string used in labels for graphviz
926 # assert ">><<".escape_to_dot == "\\>\\>\\<\\<"
927 fun escape_to_dot: String
929 return escape_more_to_c("|\{\}<>")
932 private var hash_cache: nullable Int = null
936 if hash_cache == null then
937 # djb2 hash algorithm
940 for i in [0..length[ do
942 h = (h << 5) + h + char.code_point
947 return hash_cache.as(not null)
950 # Gives the formatted string back as a Nit string with `args` in place
952 # assert "This %1 is a %2.".format("String", "formatted String") == "This String is a formatted String."
953 # assert "\\%1 This string".format("String") == "\\%1 This string"
954 fun format(args: Object...): String do
955 var s = new Array[Text]
959 # Skip escaped characters
960 if self[i] == '\\
' then
963 else if self[i] == '%' then
967 while i < length and self[i].is_numeric do
972 var ciph_len = fmt_end - ciph_st + 1
974 var arg_index = substring(ciph_st, ciph_len).to_i - 1
975 if arg_index >= args.length then continue
977 s.push substring(curr_st, fmt_st - curr_st)
978 s.push args[arg_index].to_s
983 s.push substring(curr_st, length - curr_st)
987 # Copies `n` bytes from `self` at `src_offset` into `dest` starting at `dest_offset`
989 # Basically a high-level synonym of NativeString::copy_to
991 # REQUIRE: `n` must be large enough to contain `len` bytes
993 # var ns = new NativeString(8)
994 # "Text is String".copy_to_native(ns, 8, 2, 0)
995 # assert ns.to_s_with_length(8) == "xt is St"
997 fun copy_to_native(dest: NativeString, n, src_offset, dest_offset: Int) do
998 var mypos = src_offset
999 var itspos = dest_offset
1001 dest[itspos] = self.bytes[mypos]
1010 # All kinds of array-based text representations.
1011 abstract class FlatText
1014 # Underlying C-String (`char*`)
1016 # Warning : Might be void in some subclasses, be sure to check
1017 # if set before using it.
1018 var items: NativeString is noinit
1020 # Returns a char* starting at position `first_byte`
1022 # WARNING: If you choose to use this service, be careful of the following.
1024 # Strings and NativeString are *ideally* always allocated through a Garbage Collector.
1025 # Since the GC tracks the use of the pointer for the beginning of the char*, it may be
1026 # deallocated at any moment, rendering the pointer returned by this function invalid.
1027 # Any access to freed memory may very likely cause undefined behaviour or a crash.
1028 # (Failure to do so will most certainly result in long and painful debugging hours)
1030 # The only safe use of this pointer is if it is ephemeral (e.g. read in a C function
1031 # then immediately return).
1033 # As always, do not modify the content of the String in C code, if this is what you want
1034 # copy locally the char* as Nit Strings are immutable.
1035 fun fast_cstring: NativeString is abstract
1037 redef var length = 0
1039 redef var bytelen = 0
1050 redef fun copy_to_native(dest, n, src_offset, dest_offset) do
1051 items.copy_to(dest, n, src_offset, dest_offset)
1055 # Abstract class for the SequenceRead compatible
1056 # views on the chars of any Text
1057 private abstract class StringCharView
1058 super SequenceRead[Char]
1062 var target: SELFTYPE
1064 redef fun is_empty do return target.is_empty
1066 redef fun length do return target.length
1068 redef fun iterator: IndexedIterator[Char] do return self.iterator_from(0)
1070 redef fun reverse_iterator do return self.reverse_iterator_from(self.length - 1)
1073 # Abstract class for the SequenceRead compatible
1074 # views on the bytes of any Text
1075 private abstract class StringByteView
1076 super SequenceRead[Byte]
1080 var target: SELFTYPE
1082 redef fun is_empty do return target.is_empty
1084 redef fun length do return target.bytelen
1086 redef fun iterator do return self.iterator_from(0)
1088 redef fun reverse_iterator do return self.reverse_iterator_from(target.bytelen - 1)
1091 # Immutable sequence of characters.
1093 # String objects may be created using literals.
1095 # assert "Hello World!" isa String
1096 abstract class String
1099 redef type SELFTYPE: String is fixed
1101 redef fun to_s do return self
1103 # Concatenates `o` to `self`
1105 # assert "hello" + "world" == "helloworld"
1106 # assert "" + "hello" + "" == "hello"
1107 fun +(o: Text): SELFTYPE is abstract
1109 # Concatenates self `i` times
1111 # assert "abc" * 4 == "abcabcabcabc"
1112 # assert "abc" * 1 == "abc"
1113 # assert "abc" * 0 == ""
1114 fun *(i: Int): SELFTYPE is abstract
1116 # Insert `s` at `pos`.
1118 # assert "helloworld".insert_at(" ", 5) == "hello world"
1119 fun insert_at(s: String, pos: Int): SELFTYPE is abstract
1121 redef fun substrings is abstract
1123 # Returns a reversed version of self
1125 # assert "hello".reversed == "olleh"
1126 # assert "bob".reversed == "bob"
1127 # assert "".reversed == ""
1128 fun reversed: SELFTYPE is abstract
1130 # A upper case version of `self`
1132 # assert "Hello World!".to_upper == "HELLO WORLD!"
1133 fun to_upper: SELFTYPE is abstract
1135 # A lower case version of `self`
1137 # assert "Hello World!".to_lower == "hello world!"
1138 fun to_lower : SELFTYPE is abstract
1140 # Takes a camel case `self` and converts it to snake case
1142 # assert "randomMethodId".to_snake_case == "random_method_id"
1144 # The rules are the following:
1146 # An uppercase is always converted to a lowercase
1148 # assert "HELLO_WORLD".to_snake_case == "hello_world"
1150 # An uppercase that follows a lowercase is prefixed with an underscore
1152 # assert "HelloTheWORLD".to_snake_case == "hello_the_world"
1154 # An uppercase that follows an uppercase and is followed by a lowercase, is prefixed with an underscore
1156 # assert "HelloTHEWorld".to_snake_case == "hello_the_world"
1158 # All other characters are kept as is; `self` does not need to be a proper CamelCased string.
1160 # assert "=-_H3ll0Th3W0rld_-=".to_snake_case == "=-_h3ll0th3w0rld_-="
1161 fun to_snake_case: SELFTYPE
1163 if self.is_lower then return self
1165 var new_str = new Buffer.with_cap(self.length)
1166 var prev_is_lower = false
1167 var prev_is_upper = false
1169 for i in [0..length[ do
1171 if char.is_lower then
1173 prev_is_lower = true
1174 prev_is_upper = false
1175 else if char.is_upper then
1176 if prev_is_lower then
1178 else if prev_is_upper and i+1 < length and chars[i+1].is_lower then
1181 new_str.add(char.to_lower)
1182 prev_is_lower = false
1183 prev_is_upper = true
1186 prev_is_lower = false
1187 prev_is_upper = false
1194 # Takes a snake case `self` and converts it to camel case
1196 # assert "random_method_id".to_camel_case == "randomMethodId"
1198 # If the identifier is prefixed by an underscore, the underscore is ignored
1200 # assert "_private_field".to_camel_case == "_privateField"
1202 # If `self` is upper, it is returned unchanged
1204 # assert "RANDOM_ID".to_camel_case == "RANDOM_ID"
1206 # If there are several consecutive underscores, they are considered as a single one
1208 # assert "random__method_id".to_camel_case == "randomMethodId"
1209 fun to_camel_case: SELFTYPE
1211 if self.is_upper then return self
1213 var new_str = new Buffer
1214 var is_first_char = true
1215 var follows_us = false
1217 for i in [0..length[ do
1219 if is_first_char then
1221 is_first_char = false
1222 else if char == '_
' then
1224 else if follows_us then
1225 new_str.add(char.to_upper)
1235 # Returns a capitalized `self`
1237 # Letters that follow a letter are lowercased
1238 # Letters that follow a non-letter are upcased.
1240 # SEE : `Char::is_letter` for the definition of letter.
1242 # assert "jAVASCRIPT".capitalized == "Javascript"
1243 # assert "i am root".capitalized == "I Am Root"
1244 # assert "ab_c -ab0c ab\nc".capitalized == "Ab_C -Ab0C Ab\nC"
1245 fun capitalized: SELFTYPE do
1246 if length == 0 then return self
1248 var buf = new Buffer.with_cap(length)
1250 var curr = chars[0].to_upper
1254 for i in [1 .. length[ do
1257 if prev.is_letter then
1258 buf[i] = curr.to_lower
1260 buf[i] = curr.to_upper
1268 # A mutable sequence of characters.
1269 abstract class Buffer
1272 # Returns an arbitrary subclass of `Buffer` with default parameters
1275 # Returns an instance of a subclass of `Buffer` with `i` base capacity
1276 new with_cap(i: Int) is abstract
1278 redef type SELFTYPE: Buffer is fixed
1280 # Specific implementations MUST set this to `true` in order to invalidate caches
1281 protected var is_dirty = true
1283 # Copy-On-Write flag
1285 # If the `Buffer` was to_s'd
, the next in-place altering
1286 # operation will cause the current `Buffer` to be re-allocated.
1288 # The flag will then be set at `false`.
1289 protected var written
= false
1291 # Modifies the char contained at pos `index`
1293 # DEPRECATED : Use self.chars.[]= instead
1294 fun []=(index
: Int, item
: Char) is abstract
1296 # Adds a char `c` at the end of self
1298 # DEPRECATED : Use self.chars.add instead
1299 fun add
(c
: Char) is abstract
1303 # var b = new Buffer
1305 # assert not b.is_empty
1308 fun clear
is abstract
1310 # Enlarges the subsequent array containing the chars of self
1311 fun enlarge
(cap
: Int) is abstract
1313 # Adds the content of text `s` at the end of self
1315 # var b = new Buffer
1318 # assert b == "helloworld"
1319 fun append
(s
: Text) is abstract
1321 # `self` is appended in such a way that `self` is repeated `r` times
1323 # var b = new Buffer
1326 # assert b == "hellohellohello"
1327 fun times
(r
: Int) is abstract
1329 # Reverses itself in-place
1331 # var b = new Buffer
1334 # assert b == "olleh"
1335 fun reverse
is abstract
1337 # Changes each lower-case char in `self` by its upper-case variant
1339 # var b = new Buffer
1340 # b.append("Hello World!")
1342 # assert b == "HELLO WORLD!"
1343 fun upper
is abstract
1345 # Changes each upper-case char in `self` by its lower-case variant
1347 # var b = new Buffer
1348 # b.append("Hello World!")
1350 # assert b == "hello world!"
1351 fun lower
is abstract
1353 # Capitalizes each word in `self`
1355 # Letters that follow a letter are lowercased
1356 # Letters that follow a non-letter are upcased.
1358 # SEE: `Char::is_letter` for the definition of a letter.
1360 # var b = new FlatBuffer.from("jAVAsCriPt")
1362 # assert b == "Javascript"
1363 # b = new FlatBuffer.from("i am root")
1365 # assert b == "I Am Root"
1366 # b = new FlatBuffer.from("ab_c -ab0c ab\nc")
1368 # assert b == "Ab_C -Ab0C Ab\nC"
1370 if length
== 0 then return
1371 var c
= self[0].to_upper
1374 for i
in [1 .. length
[ do
1377 if prev
.is_letter
then
1378 self[i
] = c
.to_lower
1380 self[i
] = c
.to_upper
1387 if is_dirty
then hash_cache
= null
1391 # In Buffers, the internal sequence of character is mutable
1392 # Thus, `chars` can be used to modify the buffer.
1393 redef fun chars
: Sequence[Char] is abstract
1396 # View for chars on Buffer objects, extends Sequence
1397 # for mutation operations
1398 private abstract class BufferCharView
1399 super StringCharView
1400 super Sequence[Char]
1402 redef type SELFTYPE: Buffer
1406 # View for bytes on Buffer objects, extends Sequence
1407 # for mutation operations
1408 private abstract class BufferByteView
1409 super StringByteView
1411 redef type SELFTYPE: Buffer
1415 # User readable representation of `self`.
1416 fun to_s
: String do return inspect
1418 # The class name of the object in NativeString format.
1419 private fun native_class_name
: NativeString is intern
1421 # The class name of the object.
1423 # assert 5.class_name == "Int"
1424 fun class_name
: String do return native_class_name
.to_s
1426 # Developer readable representation of `self`.
1427 # Usually, it uses the form "<CLASSNAME:#OBJECTID bla bla bla>"
1430 return "<{inspect_head}>"
1433 # Return "CLASSNAME:#OBJECTID".
1434 # This function is mainly used with the redefinition of the inspect method
1435 protected fun inspect_head
: String
1437 return "{class_name}:#{object_id.to_hex}"
1442 # assert true.to_s == "true"
1443 # assert false.to_s == "false"
1455 # C function to calculate the length of the `NativeString` to receive `self`
1456 private fun byte_to_s_len
: Int `{
1457 return snprintf(NULL, 0, "0x%02x", self);
1460 # C function to convert an nit Int to a NativeString (char*)
1461 private fun native_byte_to_s
(nstr
: NativeString, strlen
: Int) `{
1462 snprintf(nstr, strlen, "0x%02x", self);
1465 # Displayable byte in its hexadecimal form (0x..)
1467 # assert 1.to_b.to_s == "0x01"
1468 # assert (-123).to_b.to_s == "0x85"
1470 var nslen
= byte_to_s_len
1471 var ns
= new NativeString(nslen
+ 1)
1473 native_byte_to_s
(ns
, nslen
+ 1)
1474 return ns
.to_s_with_length
(nslen
)
1480 # Wrapper of strerror C function
1481 private fun strerror_ext
: NativeString `{ return strerror(self); `}
1483 # Returns a string describing error number
1484 fun strerror: String do return strerror_ext.to_s
1486 # Fill `s
` with the digits in base `base
` of `self` (and with the '-' sign if 'signed' and negative).
1487 # assume < to_c max const of char
1488 private fun fill_buffer(s: Buffer, base: Int, signed: Bool)
1495 else if self == 0 then
1502 var pos = digit_count(base) - 1
1503 while pos >= 0 and n > 0 do
1504 s.chars[pos] = (n % base).to_c
1510 # C function to calculate the length of the `NativeString` to receive `self`
1511 private fun int_to_s_len: Int `{
1512 return snprintf
(NULL, 0, "%ld", self);
1515 # C function to convert an nit Int to a NativeString (char*)
1516 private fun native_int_to_s(nstr: NativeString, strlen: Int) `{
1517 snprintf
(nstr
, strlen
, "%ld", self);
1520 # return displayable int in base base and signed
1521 fun to_base(base: Int, signed: Bool): String is abstract
1523 # return displayable int in hexadecimal
1525 # assert 1.to_hex == "1"
1526 # assert (-255).to_hex == "-ff"
1527 fun to_hex: String do return to_base(16,false)
1531 # Pretty representation of `self`, with decimals as needed from 1 to a maximum of 3
1533 # assert 12.34.to_s == "12.34"
1534 # assert (-0120.030).to_s == "-120.03"
1536 # see `to_precision
` for a custom precision.
1538 var str = to_precision( 3 )
1539 if is_inf != 0 or is_nan then return str
1540 var len = str.length
1541 for i in [0..len-1] do
1543 var c = str.chars[j]
1546 else if c == '.' then
1547 return str.substring( 0, j+2 )
1549 return str.substring( 0, j+1 )
1555 # `String` representation of `self` with the given number of `decimals
`
1557 # assert 12.345.to_precision(0) == "12"
1558 # assert 12.345.to_precision(3) == "12.345"
1559 # assert (-12.345).to_precision(3) == "-12.345"
1560 # assert (-0.123).to_precision(3) == "-0.123"
1561 # assert 0.999.to_precision(2) == "1.00"
1562 # assert 0.999.to_precision(4) == "0.9990"
1563 fun to_precision(decimals: Int): String
1565 if is_nan then return "nan"
1567 var isinf = self.is_inf
1570 else if isinf == -1 then
1574 if decimals == 0 then return self.to_i.to_s
1576 for i in [0..decimals[ do f = f * 10.0
1583 if i == 0 then return "0." + "0"*decimals
1585 # Prepare both parts of the float, before and after the "."
1590 if sl > decimals then
1591 # Has something before the "."
1592 p1 = s.substring(0, sl-decimals)
1593 p2 = s.substring(sl-decimals, decimals)
1596 p2 = "0"*(decimals-sl) + s
1599 if i < 0 then p1 = "-" + p1
1601 return p1 + "." + p2
1607 # Returns a sequence with the UTF-8 bytes of `self`
1609 # assert 'a'.bytes == [0x61u8]
1610 # assert 'ま'.bytes == [0xE3u8, 0x81u8, 0xBEu8]
1611 fun bytes: SequenceRead[Byte] do return to_s.bytes
1613 # Is `self` an UTF-16 surrogate pair ?
1614 fun is_surrogate: Bool do
1616 return cp >= 0xD800 and cp <= 0xDFFF
1619 # Length of `self` in a UTF-8 String
1620 private fun u8char_len: Int do
1621 var c = self.code_point
1622 if c < 0x80 then return 1
1623 if c <= 0x7FF then return 2
1624 if c <= 0xFFFF then return 3
1625 if c <= 0x10FFFF then return 4
1626 # Bad character format
1630 # assert 'x'.to_s == "x"
1633 var ns = new NativeString(ln + 1)
1635 return ns.to_s_with_length(ln)
1638 # Returns `self` escaped to UTF-16
1640 # i.e. Represents `self`.`code_point
` using UTF-16 codets escaped
1643 # assert 'A'.escape_to_utf16 == "\\u0041"
1644 # assert 'è'.escape_to_utf16 == "\\u00e8"
1645 # assert 'あ'.escape_to_utf16 == "\\u3042"
1646 # assert '𐏓'.escape_to_utf16 == "\\ud800\\udfd3"
1647 fun escape_to_utf16: String do
1650 if cp < 0xD800 or (cp >= 0xE000 and cp <= 0xFFFF) then
1651 buf = new Buffer.with_cap(6)
1652 buf.append("\\u0000")
1655 for i in hx.chars.reverse_iterator do
1660 buf = new Buffer.with_cap(12)
1661 buf.append("\\u0000\\u0000")
1662 var lo = (((cp - 0x10000) & 0x3FF) + 0xDC00).to_hex
1663 var hi = ((((cp - 0x10000) & 0xFFC00) >> 10) + 0xD800).to_hex
1678 private fun u8char_tos(r: NativeString, len: Int) `{
1685 r
[0] = 0xC0 | ((self & 0x7C0) >> 6);
1686 r
[1] = 0x80 | (self & 0x3F);
1689 r
[0] = 0xE0 | ((self & 0xF000) >> 12);
1690 r
[1] = 0x80 | ((self & 0xFC0) >> 6);
1691 r
[2] = 0x80 | (self & 0x3F);
1694 r
[0] = 0xF0 | ((self & 0x1C0000) >> 18);
1695 r
[1] = 0x80 | ((self & 0x3F000) >> 12);
1696 r
[2] = 0x80 | ((self & 0xFC0) >> 6);
1697 r
[3] = 0x80 | (self & 0x3F);
1702 # Returns true if the char is a numerical digit
1704 # assert '0'.is_numeric
1705 # assert '9'.is_numeric
1706 # assert not 'a'.is_numeric
1707 # assert not '?'.is_numeric
1709 # FIXME: Works on ASCII-range only
1710 fun is_numeric: Bool
1712 return self >= '0' and self <= '9'
1715 # Returns true if the char is an alpha digit
1717 # assert 'a'.is_alpha
1718 # assert 'Z'.is_alpha
1719 # assert not '0'.is_alpha
1720 # assert not '?'.is_alpha
1722 # FIXME: Works on ASCII-range only
1725 return (self >= 'a' and self <= 'z') or (self >= 'A' and self <= 'Z')
1728 # Is `self` an hexadecimal digit ?
1730 # assert 'A'.is_hexdigit
1731 # assert not 'G'.is_hexdigit
1732 # assert 'a'.is_hexdigit
1733 # assert not 'g'.is_hexdigit
1734 # assert '5'.is_hexdigit
1735 fun is_hexdigit: Bool do return (self >= '0' and self <= '9') or (self >= 'A' and self <= 'F') or
1736 (self >= 'a' and self <= 'f')
1738 # Returns true if the char is an alpha or a numeric digit
1740 # assert 'a'.is_alphanumeric
1741 # assert 'Z'.is_alphanumeric
1742 # assert '0'.is_alphanumeric
1743 # assert '9'.is_alphanumeric
1744 # assert not '?'.is_alphanumeric
1746 # FIXME: Works on ASCII-range only
1747 fun is_alphanumeric: Bool
1749 return self.is_numeric or self.is_alpha
1752 # Returns `self` to its int value
1754 # REQUIRE: `is_hexdigit
`
1755 fun from_hex: Int do
1756 if self >= '0' and self <= '9' then return code_point - 0x30
1757 if self >= 'A' and self <= 'F' then return code_point - 0x37
1758 if self >= 'a' and self <= 'f' then return code_point - 0x57
1759 # Happens if self is not a hexdigit
1760 assert self.is_hexdigit
1761 # To make flow analysis happy
1766 redef class Collection[E]
1767 # String representation of the content of the collection.
1769 # The standard representation is the list of elements separated with commas.
1772 # assert [1,2,3].to_s == "[1,2,3]"
1773 # assert [1..3].to_s == "[1,2,3]"
1774 # assert (new Array[Int]).to_s == "[]" # empty collection
1777 # Subclasses may return a more specific string representation.
1780 return "[" + join(",") + "]"
1783 # Concatenate elements without separators
1786 # assert [1,2,3].plain_to_s == "123"
1787 # assert [11..13].plain_to_s == "111213"
1788 # assert (new Array[Int]).plain_to_s == "" # empty collection
1790 fun plain_to_s: String
1793 for e in self do if e != null then s.append(e.to_s)
1797 # Concatenate and separate each elements with `separator
`.
1799 # Only concatenate if `separator
== null`.
1801 # assert [1, 2, 3].join(":") == "1:2:3"
1802 # assert [1..3].join(":") == "1:2:3"
1803 # assert [1..3].join == "123"
1804 fun join(separator: nullable Text): String
1806 if is_empty then return ""
1808 var s = new Buffer # Result
1813 if e != null then s.append(e.to_s)
1815 # Concat other items
1818 if separator != null then s.append(separator)
1820 if e != null then s.append(e.to_s)
1827 redef class Map[K,V]
1828 # Concatenate couples of key value.
1829 # Key and value are separated by `couple_sep
`.
1830 # Couples are separated by `sep
`.
1832 # var m = new HashMap[Int, String]
1835 # assert m.join("; ", "=") == "1=one; 10=ten"
1836 fun join(sep, couple_sep: String): String is abstract
1840 private var args_cache: nullable Sequence[String] = null
1842 # The arguments of the program as given by the OS
1843 fun program_args: Sequence[String]
1845 if _args_cache == null then init_args
1846 return _args_cache.as(not null)
1849 # The name of the program as given by the OS
1850 fun program_name: String
1852 return native_argv(0).to_s
1855 # Initialize `program_args
` with the contents of `native_argc
` and `native_argv
`.
1856 private fun init_args
1858 var argc = native_argc
1859 var args = new Array[String].with_capacity(0)
1862 args[i-1] = native_argv(i).to_s
1868 # First argument of the main C function.
1869 private fun native_argc: Int is intern
1871 # Second argument of the main C function.
1872 private fun native_argv(i: Int): NativeString is intern
1875 # Comparator that efficienlty use `to_s
` to compare things
1877 # The comparaison call `to_s
` on object and use the result to order things.
1879 # var a = [1, 2, 3, 10, 20]
1880 # (new CachedAlphaComparator).sort(a)
1881 # assert a == [1, 10, 2, 20, 3]
1883 # Internally the result of `to_s
` is cached in a HashMap to counter
1884 # uneficient implementation of `to_s
`.
1886 # Note: it caching is not usefull, see `alpha_comparator
`
1887 class CachedAlphaComparator
1889 redef type COMPARED: Object
1891 private var cache = new HashMap[Object, String]
1893 private fun do_to_s(a: Object): String do
1894 if cache.has_key(a) then return cache[a]
1900 redef fun compare(a, b) do
1901 return do_to_s(a) <=> do_to_s(b)
1905 # see `alpha_comparator
`
1906 private class AlphaComparator
1908 redef fun compare(a, b) do return a.to_s <=> b.to_s
1911 # Stateless comparator that naively use `to_s
` to compare things.
1913 # Note: the result of `to_s
` is not cached, thus can be invoked a lot
1914 # on a single instace. See `CachedAlphaComparator` as an alternative.
1916 # var a = [1, 2, 3, 10, 20]
1917 # alpha_comparator.sort(a)
1918 # assert a == [1, 10, 2, 20, 3]
1919 fun alpha_comparator: Comparator do return once new AlphaComparator
1921 # The arguments of the program as given by the OS
1922 fun args: Sequence[String]
1924 return sys.program_args
1927 redef class NativeString
1928 # Returns `self` as a new String.
1929 fun to_s_with_copy: String is abstract
1931 # Returns `self` as a String of `length
`.
1932 fun to_s_with_length(length: Int): String is abstract
1934 # Returns `self` as a String with `bytelen
` and `length
` set
1936 # SEE: `abstract_text
::Text` for more infos on the difference
1937 # between `Text::bytelen
` and `Text::length
`
1938 fun to_s_full(bytelen, unilen: Int): String is abstract
1941 redef class NativeArray[E]
1942 # Join all the elements using `to_s
`
1944 # REQUIRE: `self isa NativeArray[String]`
1945 # REQUIRE: all elements are initialized
1946 fun native_to_s: String is abstract