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 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
151 var iter
= self.chars
.reverse_iterator_from
(pos
)
153 if iter
.item
== item
then return iter
.index
159 # Gets an iterator on the chars of self
161 # DEPRECATED : Use self.chars.iterator instead
162 fun iterator
: Iterator[Char]
164 return self.chars
.iterator
168 # Gets an Array containing the chars of self
170 # DEPRECATED : Use self.chars.to_a instead
171 fun to_a
: Array[Char] do return chars
.to_a
173 # Create a substring from `self` beginning at the `from` position
175 # assert "abcd".substring_from(1) == "bcd"
176 # assert "abcd".substring_from(-1) == "abcd"
177 # assert "abcd".substring_from(2) == "cd"
179 # As with substring, a `from` index < 0 will be replaced by 0
180 fun substring_from
(from
: Int): SELFTYPE
182 if from
>= self.length
then return empty
183 if from
< 0 then from
= 0
184 return substring
(from
, length
- from
)
187 # Does self have a substring `str` starting from position `pos`?
189 # assert "abcd".has_substring("bc",1) == true
190 # assert "abcd".has_substring("bc",2) == false
192 # Returns true iff all characters of `str` are presents
193 # at the expected index in `self.`
194 # The first character of `str` being at `pos`, the second
195 # character being at `pos+1` and so on...
197 # This means that all characters of `str` need to be inside `self`.
199 # assert "abcd".has_substring("xab", -1) == false
200 # assert "abcd".has_substring("cdx", 2) == false
202 # And that the empty string is always a valid substring.
204 # assert "abcd".has_substring("", 2) == true
205 # assert "abcd".has_substring("", 200) == true
206 fun has_substring
(str
: String, pos
: Int): Bool
208 if str
.is_empty
then return true
209 if pos
< 0 or pos
+ str
.length
> length
then return false
210 var myiter
= self.chars
.iterator_from
(pos
)
211 var itsiter
= str
.chars
.iterator
212 while myiter
.is_ok
and itsiter
.is_ok
do
213 if myiter
.item
!= itsiter
.item
then return false
217 if itsiter
.is_ok
then return false
221 # Is this string prefixed by `prefix`?
223 # assert "abcd".has_prefix("ab") == true
224 # assert "abcbc".has_prefix("bc") == false
225 # assert "ab".has_prefix("abcd") == false
226 fun has_prefix
(prefix
: String): Bool do return has_substring
(prefix
,0)
228 # Is this string suffixed by `suffix`?
230 # assert "abcd".has_suffix("abc") == false
231 # assert "abcd".has_suffix("bcd") == true
232 fun has_suffix
(suffix
: String): Bool do return has_substring
(suffix
, length
- suffix
.length
)
234 # Returns `self` as the corresponding integer
236 # assert "123".to_i == 123
237 # assert "-1".to_i == -1
238 # assert "0x64".to_i == 100
239 # assert "0b1100_0011".to_i== 195
240 # assert "--12".to_i == 12
242 # REQUIRE: `self`.`is_int`
243 fun to_i
: Int is abstract
245 # If `self` contains a float, return the corresponding float
247 # assert "123".to_f == 123.0
248 # assert "-1".to_f == -1.0
249 # assert "-1.2e-3".to_f == -0.0012
253 return to_s
.to_cstring
.atof
256 # If `self` contains only digits and alpha <= 'f', return the corresponding integer.
258 # assert "ff".to_hex == 255
259 fun to_hex
: Int do return a_to
(16)
261 # If `self` contains only digits <= '7', return the corresponding integer.
263 # assert "714".to_oct == 460
264 fun to_oct
: Int do return a_to
(8)
266 # If `self` contains only '0' et '1', return the corresponding integer.
268 # assert "101101".to_bin == 45
269 fun to_bin
: Int do return a_to
(2)
271 # If `self` contains only digits '0' .. '9', return the corresponding integer.
273 # assert "108".to_dec == 108
274 fun to_dec
: Int do return a_to
(10)
276 # If `self` contains only digits and letters, return the corresponding integer in a given base
278 # assert "120".a_to(3) == 15
279 fun a_to
(base
: Int) : Int
284 for j
in [0..length
[ do
306 # Returns `true` if the string contains only Numeric values (and one "," or one "." character)
308 # assert "123".is_numeric == true
309 # assert "1.2".is_numeric == true
310 # assert "1,2".is_numeric == true
311 # assert "1..2".is_numeric == false
314 var has_point_or_comma
= false
315 for i
in [0..length
[ do
317 if not c
.is_numeric
then
318 if (c
== '.' or c
== ',') and not has_point_or_comma
then
319 has_point_or_comma
= true
328 # Returns `true` if the string contains only Hex chars
330 # assert "048bf".is_hex == true
331 # assert "ABCDEF".is_hex == true
332 # assert "0G".is_hex == false
335 for i
in [0..length
[ do
337 if not (c
>= 'a' and c
<= 'f') and
338 not (c
>= 'A' and c
<= 'F') and
339 not (c
>= '0' and c
<= '9') then return false
344 # Returns `true` if the string contains only Binary digits
346 # assert "1101100".is_bin == true
347 # assert "1101020".is_bin == false
349 for i
in chars
do if i
!= '0' and i
!= '1' then return false
353 # Returns `true` if the string contains only Octal digits
355 # assert "213453".is_oct == true
356 # assert "781".is_oct == false
358 for i
in chars
do if i
< '0' or i
> '7' then return false
362 # Returns `true` if the string contains only Decimal digits
364 # assert "10839".is_dec == true
365 # assert "164F".is_dec == false
367 for i
in chars
do if i
< '0' or i
> '9' then return false
371 # Are all letters in `self` upper-case ?
373 # assert "HELLO WORLD".is_upper == true
374 # assert "%$&%!".is_upper == true
375 # assert "hello world".is_upper == false
376 # assert "Hello World".is_upper == false
379 for i
in [0..length
[ do
381 if char
.is_lower
then return false
386 # Are all letters in `self` lower-case ?
388 # assert "hello world".is_lower == true
389 # assert "%$&%!".is_lower == true
390 # assert "Hello World".is_lower == false
393 for i
in [0..length
[ do
395 if char
.is_upper
then return false
400 # Removes the whitespaces at the beginning of self
402 # assert " \n\thello \n\t".l_trim == "hello \n\t"
404 # `Char::is_whitespace` determines what is a whitespace.
407 var iter
= self.chars
.iterator
409 if not iter
.item
.is_whitespace
then break
412 if iter
.index
== length
then return self.empty
413 return self.substring_from
(iter
.index
)
416 # Removes the whitespaces at the end of self
418 # assert " \n\thello \n\t".r_trim == " \n\thello"
420 # `Char::is_whitespace` determines what is a whitespace.
423 var iter
= self.chars
.reverse_iterator
425 if not iter
.item
.is_whitespace
then break
428 if iter
.index
< 0 then return self.empty
429 return self.substring
(0, iter
.index
+ 1)
432 # Trims trailing and preceding white spaces
434 # assert " Hello World ! ".trim == "Hello World !"
435 # assert "\na\nb\tc\t".trim == "a\nb\tc"
437 # `Char::is_whitespace` determines what is a whitespace.
438 fun trim
: SELFTYPE do return (self.l_trim
).r_trim
440 # Is the string non-empty but only made of whitespaces?
442 # assert " \n\t ".is_whitespace == true
443 # assert " hello ".is_whitespace == false
444 # assert "".is_whitespace == false
446 # `Char::is_whitespace` determines what is a whitespace.
447 fun is_whitespace
: Bool
449 if is_empty
then return false
450 for c
in self.chars
do
451 if not c
.is_whitespace
then return false
456 # Returns `self` removed from its last line terminator (if any).
458 # assert "Hello\n".chomp == "Hello"
459 # assert "Hello".chomp == "Hello"
461 # assert "\n".chomp == ""
462 # assert "".chomp == ""
464 # Line terminators are `"\n"`, `"\r\n"` and `"\r"`.
465 # A single line terminator, the last one, is removed.
467 # assert "\r\n".chomp == ""
468 # assert "\r\n\n".chomp == "\r\n"
469 # assert "\r\n\r\n".chomp == "\r\n"
470 # assert "\r\n\r".chomp == "\r\n"
472 # Note: unlike with most IO methods like `Reader::read_line`,
473 # a single `\r` is considered here to be a line terminator and will be removed.
477 if len
== 0 then return self
478 var l
= self.chars
.last
480 return substring
(0, len-1
)
481 else if l
!= '\n' then
483 else if len
> 1 and self.chars
[len-2
] == '\r' then
484 return substring
(0, len-2
)
486 return substring
(0, len-1
)
490 # Justify a self in a space of `length`
492 # `left` is the space ratio on the left side.
493 # * 0.0 for left-justified (no space at the left)
494 # * 1.0 for right-justified (all spaces at the left)
495 # * 0.5 for centered (half the spaces at the left)
499 # assert "hello".justify(10, 0.0) == "hello "
500 # assert "hello".justify(10, 1.0) == " hello"
501 # assert "hello".justify(10, 0.5) == " hello "
503 # If `length` is not enough, `self` is returned as is.
505 # assert "hello".justify(2, 0.0) == "hello"
507 # REQUIRE: `left >= 0.0 and left <= 1.0`
508 # ENSURE: `self.length <= length implies result.length == length`
509 # ENSURE: `self.length >= length implies result == self`
510 fun justify
(length
: Int, left
: Float): String
512 var diff
= length
- self.length
513 if diff
<= 0 then return to_s
514 assert left
>= 0.0 and left
<= 1.0
515 var before
= (diff
.to_f
* left
).to_i
516 return " " * before
+ self + " " * (diff-before
)
519 # Mangle a string to be a unique string only made of alphanumeric characters and underscores.
521 # This method is injective (two different inputs never produce the same
522 # output) and the returned string always respect the following rules:
524 # * Contains only US-ASCII letters, digits and underscores.
525 # * Never starts with a digit.
526 # * Never ends with an underscore.
527 # * Never contains two contiguous underscores.
529 # assert "42_is/The answer!".to_cmangle == "_52d2_is_47dThe_32danswer_33d"
530 # assert "__".to_cmangle == "_95d_95d"
531 # assert "__d".to_cmangle == "_95d_d"
532 # assert "_d_".to_cmangle == "_d_95d"
533 # assert "_42".to_cmangle == "_95d42"
534 # assert "foo".to_cmangle == "foo"
535 # assert "".to_cmangle == ""
536 fun to_cmangle
: String
538 if is_empty
then return ""
540 var underscore
= false
544 if c
>= '0' and c
<= '9' then
546 res
.append
(c
.ascii
.to_s
)
550 for i
in [start
..length
[ do
552 if (c
>= 'a' and c
<= 'z') or (c
>='A' and c
<= 'Z') then
558 res
.append
('_'.ascii
.to_s
)
561 if c
>= '0' and c
<= '9' then
564 else if c
== '_' then
569 res
.append
(c
.ascii
.to_s
)
575 res
.append
('_'.ascii
.to_s
)
581 # Escape " \ ' and non printable characters using the rules of literal C strings and characters
583 # assert "abAB12<>&".escape_to_c == "abAB12<>&"
584 # assert "\n\"'\\".escape_to_c == "\\n\\\"\\'\\\\"
586 # Most non-printable characters (bellow ASCII 32) are escaped to an octal form `\nnn`.
587 # Three digits are always used to avoid following digits to be interpreted as an element
588 # of the octal sequence.
590 # assert "{0.ascii}{1.ascii}{8.ascii}{31.ascii}{32.ascii}".escape_to_c == "\\000\\001\\010\\037 "
592 # The exceptions are the common `\t` and `\n`.
593 fun escape_to_c
: String
596 for i
in [0..length
[ do
600 else if c
== '\t' then
602 else if c
== '\0' then
604 else if c
== '"' then
606 else if c == '\'' then
608 else if c == '\\
' then
610 else if c.ascii < 32 then
612 var oct = c.ascii.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
.ascii
< 32 or c
== ';' or c
== '|' or c
== '\\' or c
== '=' then
686 b
.append
("?{c.ascii.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].ascii == 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 # Encode `self` to percent (or URL) encoding
733 # assert "aBc09-._~".to_percent_encoding == "aBc09-._~"
734 # assert "%()< >".to_percent_encoding == "%25%28%29%3c%20%3e"
735 # assert ".com/post?e=asdf&f=123".to_percent_encoding == ".com%2fpost%3fe%3dasdf%26f%3d123"
736 fun to_percent_encoding
: String
740 for i
in [0..length
[ do
742 if (c
>= '0' and c
<= '9') or
743 (c
>= 'a' and c
<= 'z') or
744 (c
>= 'A' and c
<= 'Z') or
745 c
== '-' or c
== '.' or
749 else buf
.append
"%{c.ascii.to_hex}"
755 # Decode `self` from percent (or URL) encoding to a clear string
757 # Replace invalid use of '%' with '?'.
759 # assert "aBc09-._~".from_percent_encoding == "aBc09-._~"
760 # assert "%25%28%29%3c%20%3e".from_percent_encoding == "%()< >"
761 # assert ".com%2fpost%3fe%3dasdf%26f%3d123".from_percent_encoding == ".com/post?e=asdf&f=123"
762 # assert "%25%28%29%3C%20%3E".from_percent_encoding == "%()< >"
763 # assert "incomplete %".from_percent_encoding == "incomplete ?"
764 # assert "invalid % usage".from_percent_encoding == "invalid ? usage"
765 fun from_percent_encoding
: String
773 if i
+ 2 >= length
then
774 # What follows % has been cut off
778 var hex_s
= substring
(i
, 2)
780 var hex_i
= hex_s
.to_hex
784 # What follows a % is not Hex
797 # Escape the characters `<`, `>`, `&`, `"`, `'` and `/` as HTML/XML entity references.
799 # assert "a&b-<>\"x\"/'".html_escape == "a&b-<>"x"/'"
801 # 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>
802 fun html_escape
: String
806 for i
in [0..length
[ do
810 else if c
== '<' then
812 else if c
== '>' then
814 else if c
== '"' then
816 else if c
== '\'' then
818 else if c == '/' then
827 # Two pieces of text are equals if thez have the same characters in the same order.
829 # assert "hello" == "hello"
830 # assert "hello" != "HELLO"
831 # assert "hello" == "hel"+"lo"
833 # Things that are not Text are not equal.
836 # assert "9" != ['9']
839 # assert "9".chars.first == '9' # equality of Char
840 # assert "9".chars == ['9'] # equality of Sequence
841 # assert "9".to_i == 9 # equality of Int
844 if o == null then return false
845 if not o isa Text then return false
846 if self.is_same_instance(o) then return true
847 if self.length != o.length then return false
848 return self.chars == o.chars
851 # Lexicographical comparaison
853 # assert "abc" < "xy"
854 # assert "ABC" < "abc"
857 var self_chars = self.chars.iterator
858 var other_chars = other.chars.iterator
860 while self_chars.is_ok and other_chars.is_ok do
861 if self_chars.item < other_chars.item then return true
862 if self_chars.item > other_chars.item then return false
867 if self_chars.is_ok then
874 # Escape string used in labels for graphviz
876 # assert ">><<".escape_to_dot == "\\>\\>\\<\\<"
877 fun escape_to_dot: String
879 return escape_more_to_c("|\{\}<>")
882 private var hash_cache: nullable Int = null
886 if hash_cache == null then
887 # djb2 hash algorithm
890 for i in [0..length[ do
892 h = (h << 5) + h + char.ascii
897 return hash_cache.as(not null)
900 # Gives the formatted string back as a Nit string with `args` in place
902 # assert "This %1 is a %2.".format("String", "formatted String") == "This String is a formatted String."
903 # assert "\\%1 This string".format("String") == "\\%1 This string"
904 fun format(args: Object...): String do
905 var s = new Array[Text]
909 # Skip escaped characters
910 if self[i] == '\\
' then
913 else if self[i] == '%' then
917 while i < length and self[i].is_numeric do
922 var ciph_len = fmt_end - ciph_st + 1
924 var arg_index = substring(ciph_st, ciph_len).to_i - 1
925 if arg_index >= args.length then continue
927 s.push substring(curr_st, fmt_st - curr_st)
928 s.push args[arg_index].to_s
933 s.push substring(curr_st, length - curr_st)
937 # Copies `n` bytes from `self` at `src_offset` into `dest` starting at `dest_offset`
939 # Basically a high-level synonym of NativeString::copy_to
941 # REQUIRE: `n` must be large enough to contain `len` bytes
943 # var ns = new NativeString(8)
944 # "Text is String".copy_to_native(ns, 8, 2, 0)
945 # assert ns.to_s_with_length(8) == "xt is St"
947 fun copy_to_native(dest: NativeString, n, src_offset, dest_offset: Int) do
948 var mypos = src_offset
949 var itspos = dest_offset
951 dest[itspos] = self.bytes[mypos]
960 # All kinds of array-based text representations.
961 abstract class FlatText
964 # Underlying C-String (`char*`)
966 # Warning : Might be void in some subclasses, be sure to check
967 # if set before using it.
968 private var items: NativeString is noinit
970 # Real items, used as cache for to_cstring is called
971 private var real_items: nullable NativeString = null
973 # Returns a char* starting at position `first_byte`
975 # WARNING: If you choose to use this service, be careful of the following.
977 # Strings and NativeString are *ideally* always allocated through a Garbage Collector.
978 # Since the GC tracks the use of the pointer for the beginning of the char*, it may be
979 # deallocated at any moment, rendering the pointer returned by this function invalid.
980 # Any access to freed memory may very likely cause undefined behaviour or a crash.
981 # (Failure to do so will most certainly result in long and painful debugging hours)
983 # The only safe use of this pointer is if it is ephemeral (e.g. read in a C function
984 # then immediately return).
986 # As always, do not modify the content of the String in C code, if this is what you want
987 # copy locally the char* as Nit Strings are immutable.
988 private fun fast_cstring: NativeString is abstract
992 redef var bytelen = 0
1003 redef fun copy_to_native(dest, n, src_offset, dest_offset) do
1004 items.copy_to(dest, n, src_offset, dest_offset)
1008 # Abstract class for the SequenceRead compatible
1009 # views on the chars of any Text
1010 private abstract class StringCharView
1011 super SequenceRead[Char]
1015 var target: SELFTYPE
1017 redef fun is_empty do return target.is_empty
1019 redef fun length do return target.length
1021 redef fun iterator: IndexedIterator[Char] do return self.iterator_from(0)
1023 redef fun reverse_iterator do return self.reverse_iterator_from(self.length - 1)
1026 # Abstract class for the SequenceRead compatible
1027 # views on the bytes of any Text
1028 private abstract class StringByteView
1029 super SequenceRead[Byte]
1033 var target: SELFTYPE
1035 redef fun is_empty do return target.is_empty
1037 redef fun length do return target.length
1039 redef fun iterator do return self.iterator_from(0)
1041 redef fun reverse_iterator do return self.reverse_iterator_from(target.bytelen - 1)
1044 # Immutable sequence of characters.
1046 # String objects may be created using literals.
1048 # assert "Hello World!" isa String
1049 abstract class String
1052 redef type SELFTYPE: String is fixed
1054 redef fun to_s do return self
1056 # Concatenates `o` to `self`
1058 # assert "hello" + "world" == "helloworld"
1059 # assert "" + "hello" + "" == "hello"
1060 fun +(o: Text): SELFTYPE is abstract
1062 # Concatenates self `i` times
1064 # assert "abc" * 4 == "abcabcabcabc"
1065 # assert "abc" * 1 == "abc"
1066 # assert "abc" * 0 == ""
1067 fun *(i: Int): SELFTYPE is abstract
1069 # Insert `s` at `pos`.
1071 # assert "helloworld".insert_at(" ", 5) == "hello world"
1072 fun insert_at(s: String, pos: Int): SELFTYPE is abstract
1074 redef fun substrings is abstract
1076 # Returns a reversed version of self
1078 # assert "hello".reversed == "olleh"
1079 # assert "bob".reversed == "bob"
1080 # assert "".reversed == ""
1081 fun reversed: SELFTYPE is abstract
1083 # A upper case version of `self`
1085 # assert "Hello World!".to_upper == "HELLO WORLD!"
1086 fun to_upper: SELFTYPE is abstract
1088 # A lower case version of `self`
1090 # assert "Hello World!".to_lower == "hello world!"
1091 fun to_lower : SELFTYPE is abstract
1093 # Takes a camel case `self` and converts it to snake case
1095 # assert "randomMethodId".to_snake_case == "random_method_id"
1097 # The rules are the following:
1099 # An uppercase is always converted to a lowercase
1101 # assert "HELLO_WORLD".to_snake_case == "hello_world"
1103 # An uppercase that follows a lowercase is prefixed with an underscore
1105 # assert "HelloTheWORLD".to_snake_case == "hello_the_world"
1107 # An uppercase that follows an uppercase and is followed by a lowercase, is prefixed with an underscore
1109 # assert "HelloTHEWorld".to_snake_case == "hello_the_world"
1111 # All other characters are kept as is; `self` does not need to be a proper CamelCased string.
1113 # assert "=-_H3ll0Th3W0rld_-=".to_snake_case == "=-_h3ll0th3w0rld_-="
1114 fun to_snake_case: SELFTYPE
1116 if self.is_lower then return self
1118 var new_str = new Buffer.with_cap(self.length)
1119 var prev_is_lower = false
1120 var prev_is_upper = false
1122 for i in [0..length[ do
1124 if char.is_lower then
1126 prev_is_lower = true
1127 prev_is_upper = false
1128 else if char.is_upper then
1129 if prev_is_lower then
1131 else if prev_is_upper and i+1 < length and chars[i+1].is_lower then
1134 new_str.add(char.to_lower)
1135 prev_is_lower = false
1136 prev_is_upper = true
1139 prev_is_lower = false
1140 prev_is_upper = false
1147 # Takes a snake case `self` and converts it to camel case
1149 # assert "random_method_id".to_camel_case == "randomMethodId"
1151 # If the identifier is prefixed by an underscore, the underscore is ignored
1153 # assert "_private_field".to_camel_case == "_privateField"
1155 # If `self` is upper, it is returned unchanged
1157 # assert "RANDOM_ID".to_camel_case == "RANDOM_ID"
1159 # If there are several consecutive underscores, they are considered as a single one
1161 # assert "random__method_id".to_camel_case == "randomMethodId"
1162 fun to_camel_case: SELFTYPE
1164 if self.is_upper then return self
1166 var new_str = new Buffer
1167 var is_first_char = true
1168 var follows_us = false
1170 for i in [0..length[ do
1172 if is_first_char then
1174 is_first_char = false
1175 else if char == '_
' then
1177 else if follows_us then
1178 new_str.add(char.to_upper)
1188 # Returns a capitalized `self`
1190 # Letters that follow a letter are lowercased
1191 # Letters that follow a non-letter are upcased.
1193 # SEE : `Char::is_letter` for the definition of letter.
1195 # assert "jAVASCRIPT".capitalized == "Javascript"
1196 # assert "i am root".capitalized == "I Am Root"
1197 # assert "ab_c -ab0c ab\nc".capitalized == "Ab_C -Ab0C Ab\nC"
1198 fun capitalized: SELFTYPE do
1199 if length == 0 then return self
1201 var buf = new Buffer.with_cap(length)
1203 var curr = chars[0].to_upper
1207 for i in [1 .. length[ do
1210 if prev.is_letter then
1211 buf[i] = curr.to_lower
1213 buf[i] = curr.to_upper
1221 # A mutable sequence of characters.
1222 abstract class Buffer
1225 # Returns an arbitrary subclass of `Buffer` with default parameters
1228 # Returns an instance of a subclass of `Buffer` with `i` base capacity
1229 new with_cap(i: Int) is abstract
1231 redef type SELFTYPE: Buffer is fixed
1233 # Specific implementations MUST set this to `true` in order to invalidate caches
1234 protected var is_dirty = true
1236 # Copy-On-Write flag
1238 # If the `Buffer` was to_s'd
, the next in-place altering
1239 # operation will cause the current `Buffer` to be re-allocated.
1241 # The flag will then be set at `false`.
1242 protected var written
= false
1244 # Modifies the char contained at pos `index`
1246 # DEPRECATED : Use self.chars.[]= instead
1247 fun []=(index
: Int, item
: Char) is abstract
1249 # Adds a char `c` at the end of self
1251 # DEPRECATED : Use self.chars.add instead
1252 fun add
(c
: Char) is abstract
1256 # var b = new Buffer
1258 # assert not b.is_empty
1261 fun clear
is abstract
1263 # Enlarges the subsequent array containing the chars of self
1264 fun enlarge
(cap
: Int) is abstract
1266 # Adds the content of text `s` at the end of self
1268 # var b = new Buffer
1271 # assert b == "helloworld"
1272 fun append
(s
: Text) is abstract
1274 # `self` is appended in such a way that `self` is repeated `r` times
1276 # var b = new Buffer
1279 # assert b == "hellohellohello"
1280 fun times
(r
: Int) is abstract
1282 # Reverses itself in-place
1284 # var b = new Buffer
1287 # assert b == "olleh"
1288 fun reverse
is abstract
1290 # Changes each lower-case char in `self` by its upper-case variant
1292 # var b = new Buffer
1293 # b.append("Hello World!")
1295 # assert b == "HELLO WORLD!"
1296 fun upper
is abstract
1298 # Changes each upper-case char in `self` by its lower-case variant
1300 # var b = new Buffer
1301 # b.append("Hello World!")
1303 # assert b == "hello world!"
1304 fun lower
is abstract
1306 # Capitalizes each word in `self`
1308 # Letters that follow a letter are lowercased
1309 # Letters that follow a non-letter are upcased.
1311 # SEE: `Char::is_letter` for the definition of a letter.
1313 # var b = new FlatBuffer.from("jAVAsCriPt")
1315 # assert b == "Javascript"
1316 # b = new FlatBuffer.from("i am root")
1318 # assert b == "I Am Root"
1319 # b = new FlatBuffer.from("ab_c -ab0c ab\nc")
1321 # assert b == "Ab_C -Ab0C Ab\nC"
1323 if length
== 0 then return
1324 var c
= self[0].to_upper
1327 for i
in [1 .. length
[ do
1330 if prev
.is_letter
then
1331 self[i
] = c
.to_lower
1333 self[i
] = c
.to_upper
1340 if is_dirty
then hash_cache
= null
1344 # In Buffers, the internal sequence of character is mutable
1345 # Thus, `chars` can be used to modify the buffer.
1346 redef fun chars
: Sequence[Char] is abstract
1349 # View for chars on Buffer objects, extends Sequence
1350 # for mutation operations
1351 private abstract class BufferCharView
1352 super StringCharView
1353 super Sequence[Char]
1355 redef type SELFTYPE: Buffer
1359 # View for bytes on Buffer objects, extends Sequence
1360 # for mutation operations
1361 private abstract class BufferByteView
1362 super StringByteView
1364 redef type SELFTYPE: Buffer
1368 # User readable representation of `self`.
1369 fun to_s
: String do return inspect
1371 # The class name of the object in NativeString format.
1372 private fun native_class_name
: NativeString is intern
1374 # The class name of the object.
1376 # assert 5.class_name == "Int"
1377 fun class_name
: String do return native_class_name
.to_s
1379 # Developer readable representation of `self`.
1380 # Usually, it uses the form "<CLASSNAME:#OBJECTID bla bla bla>"
1383 return "<{inspect_head}>"
1386 # Return "CLASSNAME:#OBJECTID".
1387 # This function is mainly used with the redefinition of the inspect method
1388 protected fun inspect_head
: String
1390 return "{class_name}:#{object_id.to_hex}"
1395 # assert true.to_s == "true"
1396 # assert false.to_s == "false"
1408 # C function to calculate the length of the `NativeString` to receive `self`
1409 private fun byte_to_s_len
: Int `{
1410 return snprintf(NULL, 0, "0x%02x", self);
1413 # C function to convert an nit Int to a NativeString (char*)
1414 private fun native_byte_to_s
(nstr
: NativeString, strlen
: Int) `{
1415 snprintf(nstr, strlen, "0x%02x", self);
1418 # Displayable byte in its hexadecimal form (0x..)
1420 # assert 1.to_b.to_s == "0x01"
1421 # assert (-123).to_b.to_s == "0x85"
1423 var nslen
= byte_to_s_len
1424 var ns
= new NativeString(nslen
+ 1)
1426 native_byte_to_s
(ns
, nslen
+ 1)
1427 return ns
.to_s_with_length
(nslen
)
1433 # Wrapper of strerror C function
1434 private fun strerror_ext
: NativeString `{ return strerror(self); `}
1436 # Returns a string describing error number
1437 fun strerror: String do return strerror_ext.to_s
1439 # Fill `s
` with the digits in base `base
` of `self` (and with the '-' sign if 'signed' and negative).
1440 # assume < to_c max const of char
1441 private fun fill_buffer(s: Buffer, base: Int, signed: Bool)
1448 else if self == 0 then
1455 var pos = digit_count(base) - 1
1456 while pos >= 0 and n > 0 do
1457 s.chars[pos] = (n % base).to_c
1463 # C function to calculate the length of the `NativeString` to receive `self`
1464 private fun int_to_s_len: Int `{
1465 return snprintf
(NULL, 0, "%ld", self);
1468 # C function to convert an nit Int to a NativeString (char*)
1469 private fun native_int_to_s(nstr: NativeString, strlen: Int) `{
1470 snprintf
(nstr
, strlen
, "%ld", self);
1473 # return displayable int in base base and signed
1474 fun to_base(base: Int, signed: Bool): String is abstract
1476 # return displayable int in hexadecimal
1478 # assert 1.to_hex == "1"
1479 # assert (-255).to_hex == "-ff"
1480 fun to_hex: String do return to_base(16,false)
1484 # Pretty representation of `self`, with decimals as needed from 1 to a maximum of 3
1486 # assert 12.34.to_s == "12.34"
1487 # assert (-0120.030).to_s == "-120.03"
1489 # see `to_precision
` for a custom precision.
1491 var str = to_precision( 3 )
1492 if is_inf != 0 or is_nan then return str
1493 var len = str.length
1494 for i in [0..len-1] do
1496 var c = str.chars[j]
1499 else if c == '.' then
1500 return str.substring( 0, j+2 )
1502 return str.substring( 0, j+1 )
1508 # `String` representation of `self` with the given number of `decimals
`
1510 # assert 12.345.to_precision(0) == "12"
1511 # assert 12.345.to_precision(3) == "12.345"
1512 # assert (-12.345).to_precision(3) == "-12.345"
1513 # assert (-0.123).to_precision(3) == "-0.123"
1514 # assert 0.999.to_precision(2) == "1.00"
1515 # assert 0.999.to_precision(4) == "0.9990"
1516 fun to_precision(decimals: Int): String
1518 if is_nan then return "nan"
1520 var isinf = self.is_inf
1523 else if isinf == -1 then
1527 if decimals == 0 then return self.to_i.to_s
1529 for i in [0..decimals[ do f = f * 10.0
1536 if i == 0 then return "0." + "0"*decimals
1538 # Prepare both parts of the float, before and after the "."
1543 if sl > decimals then
1544 # Has something before the "."
1545 p1 = s.substring(0, sl-decimals)
1546 p2 = s.substring(sl-decimals, decimals)
1549 p2 = "0"*(decimals-sl) + s
1552 if i < 0 then p1 = "-" + p1
1554 return p1 + "." + p2
1560 # Length of `self` in a UTF-8 String
1561 private fun u8char_len: Int do
1563 if c < 0x80 then return 1
1564 if c <= 0x7FF then return 2
1565 if c <= 0xFFFF then return 3
1566 if c <= 0x10FFFF then return 4
1567 # Bad character format
1571 # assert 'x'.to_s == "x"
1574 var ns = new NativeString(ln + 1)
1576 return ns.to_s_with_length(ln)
1579 private fun u8char_tos(r: NativeString, len: Int) `{
1586 r
[0] = 0xC0 | ((self & 0x7C0) >> 6);
1587 r
[1] = 0x80 | (self & 0x3F);
1590 r
[0] = 0xE0 | ((self & 0xF000) >> 12);
1591 r
[1] = 0x80 | ((self & 0xFC0) >> 6);
1592 r
[2] = 0x80 | (self & 0x3F);
1595 r
[0] = 0xF0 | ((self & 0x1C0000) >> 18);
1596 r
[1] = 0x80 | ((self & 0x3F000) >> 12);
1597 r
[2] = 0x80 | ((self & 0xFC0) >> 6);
1598 r
[3] = 0x80 | (self & 0x3F);
1603 # Returns true if the char is a numerical digit
1605 # assert '0'.is_numeric
1606 # assert '9'.is_numeric
1607 # assert not 'a'.is_numeric
1608 # assert not '?'.is_numeric
1610 # FIXME: Works on ASCII-range only
1611 fun is_numeric: Bool
1613 return self >= '0' and self <= '9'
1616 # Returns true if the char is an alpha digit
1618 # assert 'a'.is_alpha
1619 # assert 'Z'.is_alpha
1620 # assert not '0'.is_alpha
1621 # assert not '?'.is_alpha
1623 # FIXME: Works on ASCII-range only
1626 return (self >= 'a' and self <= 'z') or (self >= 'A' and self <= 'Z')
1629 # Returns true if the char is an alpha or a numeric digit
1631 # assert 'a'.is_alphanumeric
1632 # assert 'Z'.is_alphanumeric
1633 # assert '0'.is_alphanumeric
1634 # assert '9'.is_alphanumeric
1635 # assert not '?'.is_alphanumeric
1637 # FIXME: Works on ASCII-range only
1638 fun is_alphanumeric: Bool
1640 return self.is_numeric or self.is_alpha
1644 redef class Collection[E]
1645 # String representation of the content of the collection.
1647 # The standard representation is the list of elements separated with commas.
1650 # assert [1,2,3].to_s == "[1,2,3]"
1651 # assert [1..3].to_s == "[1,2,3]"
1652 # assert (new Array[Int]).to_s == "[]" # empty collection
1655 # Subclasses may return a more specific string representation.
1658 return "[" + join(",") + "]"
1661 # Concatenate elements without separators
1664 # assert [1,2,3].plain_to_s == "123"
1665 # assert [11..13].plain_to_s == "111213"
1666 # assert (new Array[Int]).plain_to_s == "" # empty collection
1668 fun plain_to_s: String
1671 for e in self do if e != null then s.append(e.to_s)
1675 # Concatenate and separate each elements with `separator
`.
1677 # Only concatenate if `separator
== null`.
1679 # assert [1, 2, 3].join(":") == "1:2:3"
1680 # assert [1..3].join(":") == "1:2:3"
1681 # assert [1..3].join == "123"
1682 fun join(separator: nullable Text): String
1684 if is_empty then return ""
1686 var s = new Buffer # Result
1691 if e != null then s.append(e.to_s)
1693 # Concat other items
1696 if separator != null then s.append(separator)
1698 if e != null then s.append(e.to_s)
1705 redef class Map[K,V]
1706 # Concatenate couples of key value.
1707 # Key and value are separated by `couple_sep
`.
1708 # Couples are separated by `sep
`.
1710 # var m = new HashMap[Int, String]
1713 # assert m.join("; ", "=") == "1=one; 10=ten"
1714 fun join(sep, couple_sep: String): String is abstract
1718 private var args_cache: nullable Sequence[String] = null
1720 # The arguments of the program as given by the OS
1721 fun program_args: Sequence[String]
1723 if _args_cache == null then init_args
1724 return _args_cache.as(not null)
1727 # The name of the program as given by the OS
1728 fun program_name: String
1730 return native_argv(0).to_s
1733 # Initialize `program_args
` with the contents of `native_argc
` and `native_argv
`.
1734 private fun init_args
1736 var argc = native_argc
1737 var args = new Array[String].with_capacity(0)
1740 args[i-1] = native_argv(i).to_s
1746 # First argument of the main C function.
1747 private fun native_argc: Int is intern
1749 # Second argument of the main C function.
1750 private fun native_argv(i: Int): NativeString is intern
1753 # Comparator that efficienlty use `to_s
` to compare things
1755 # The comparaison call `to_s
` on object and use the result to order things.
1757 # var a = [1, 2, 3, 10, 20]
1758 # (new CachedAlphaComparator).sort(a)
1759 # assert a == [1, 10, 2, 20, 3]
1761 # Internally the result of `to_s
` is cached in a HashMap to counter
1762 # uneficient implementation of `to_s
`.
1764 # Note: it caching is not usefull, see `alpha_comparator
`
1765 class CachedAlphaComparator
1767 redef type COMPARED: Object
1769 private var cache = new HashMap[Object, String]
1771 private fun do_to_s(a: Object): String do
1772 if cache.has_key(a) then return cache[a]
1778 redef fun compare(a, b) do
1779 return do_to_s(a) <=> do_to_s(b)
1783 # see `alpha_comparator
`
1784 private class AlphaComparator
1786 redef fun compare(a, b) do return a.to_s <=> b.to_s
1789 # Stateless comparator that naively use `to_s
` to compare things.
1791 # Note: the result of `to_s
` is not cached, thus can be invoked a lot
1792 # on a single instace. See `CachedAlphaComparator` as an alternative.
1794 # var a = [1, 2, 3, 10, 20]
1795 # alpha_comparator.sort(a)
1796 # assert a == [1, 10, 2, 20, 3]
1797 fun alpha_comparator: Comparator do return once new AlphaComparator
1799 # The arguments of the program as given by the OS
1800 fun args: Sequence[String]
1802 return sys.program_args
1805 redef class NativeString
1806 # Returns `self` as a new String.
1807 fun to_s_with_copy: String is abstract
1809 # Returns `self` as a String of `length
`.
1810 fun to_s_with_length(length: Int): String is abstract
1813 redef class NativeArray[E]
1814 # Join all the elements using `to_s
`
1816 # REQUIRE: `self isa NativeArray[String]`
1817 # REQUIRE: all elements are initialized
1818 fun native_to_s: String is abstract