lib: Perfized `to_hex` and have it work anywhere in a `Text`
[nit.git] / lib / core / text / abstract_text.nit
1 # This file is part of NIT ( http://www.nitlanguage.org ).
2 #
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
9 # another product.
10
11 # Abstract class for manipulation of sequences of characters
12 module abstract_text
13
14 import native
15 import math
16 import collection
17 intrude import collection::array
18
19 in "C" `{
20 #include <stdio.h>
21 #include <stdlib.h>
22 #include <string.h>
23 `}
24
25 # High-level abstraction for all text representations
26 abstract class Text
27 super Comparable
28
29 redef type OTHER: Text
30
31 # Type of self (used for factorization of several methods, ex : substring_from, empty...)
32 type SELFTYPE: Text
33
34 # Gets a view on the chars of the Text object
35 #
36 # assert "hello".chars.to_a == ['h', 'e', 'l', 'l', 'o']
37 fun chars: SequenceRead[Char] is abstract
38
39 # Gets a view on the bytes of the Text object
40 #
41 # assert "hello".bytes.to_a == [104u8, 101u8, 108u8, 108u8, 111u8]
42 fun bytes: SequenceRead[Byte] is abstract
43
44 # Number of characters contained in self.
45 #
46 # assert "12345".length == 5
47 # assert "".length == 0
48 # assert "あいうえお".length == 5
49 fun length: Int is abstract
50
51 # Number of bytes in `self`
52 #
53 # assert "12345".bytelen == 5
54 # assert "あいうえお".bytelen == 15
55 fun bytelen: Int is abstract
56
57 # Create a substring.
58 #
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"
64 #
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
69
70 # Iterates on the substrings of self if any
71 private fun substrings: Iterator[FlatText] is abstract
72
73 # Is the current Text empty (== "")
74 #
75 # assert "".is_empty
76 # assert not "foo".is_empty
77 fun is_empty: Bool do return self.length == 0
78
79 # Returns an empty Text of the right type
80 #
81 # This method is used internally to get the right
82 # implementation of an empty string.
83 protected fun empty: SELFTYPE is abstract
84
85 # Gets the first char of the Text
86 #
87 # DEPRECATED : Use self.chars.first instead
88 fun first: Char do return self.chars[0]
89
90 # Access a character at `index` in the string.
91 #
92 # assert "abcd"[2] == 'c'
93 #
94 # DEPRECATED : Use self.chars.[] instead
95 fun [](index: Int): Char do return self.chars[index]
96
97 # Gets the index of the first occurence of 'c'
98 #
99 # Returns -1 if not found
100 #
101 # DEPRECATED : Use self.chars.index_of instead
102 fun index_of(c: Char): Int
103 do
104 return index_of_from(c, 0)
105 end
106
107 # Gets the last char of self
108 #
109 # DEPRECATED : Use self.chars.last instead
110 fun last: Char do return self.chars[length-1]
111
112 # Gets the index of the first occurence of ´c´ starting from ´pos´
113 #
114 # Returns -1 if not found
115 #
116 # DEPRECATED : Use self.chars.index_of_from instead
117 fun index_of_from(c: Char, pos: Int): Int
118 do
119 var iter = self.chars.iterator_from(pos)
120 while iter.is_ok do
121 if iter.item == c then return iter.index
122 iter.next
123 end
124 return -1
125 end
126
127 # Gets the last index of char ´c´
128 #
129 # Returns -1 if not found
130 #
131 # DEPRECATED : Use self.chars.last_index_of instead
132 fun last_index_of(c: Char): Int
133 do
134 return last_index_of_from(c, length - 1)
135 end
136
137 # Return a null terminated char *
138 fun to_cstring: NativeString is abstract
139
140 # The index of the last occurrence of an element starting from pos (in reverse order).
141 #
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
145 #
146 # Returns -1 if not found
147 #
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)
150
151 # Gets an iterator on the chars of self
152 #
153 # DEPRECATED : Use self.chars.iterator instead
154 fun iterator: Iterator[Char]
155 do
156 return self.chars.iterator
157 end
158
159
160 # Gets an Array containing the chars of self
161 #
162 # DEPRECATED : Use self.chars.to_a instead
163 fun to_a: Array[Char] do return chars.to_a
164
165 # Create a substring from `self` beginning at the `from` position
166 #
167 # assert "abcd".substring_from(1) == "bcd"
168 # assert "abcd".substring_from(-1) == "abcd"
169 # assert "abcd".substring_from(2) == "cd"
170 #
171 # As with substring, a `from` index < 0 will be replaced by 0
172 fun substring_from(from: Int): SELFTYPE
173 do
174 if from >= self.length then return empty
175 if from < 0 then from = 0
176 return substring(from, length - from)
177 end
178
179 # Does self have a substring `str` starting from position `pos`?
180 #
181 # assert "abcd".has_substring("bc",1) == true
182 # assert "abcd".has_substring("bc",2) == false
183 #
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...
188 #
189 # This means that all characters of `str` need to be inside `self`.
190 #
191 # assert "abcd".has_substring("xab", -1) == false
192 # assert "abcd".has_substring("cdx", 2) == false
193 #
194 # And that the empty string is always a valid substring.
195 #
196 # assert "abcd".has_substring("", 2) == true
197 # assert "abcd".has_substring("", 200) == true
198 fun has_substring(str: String, pos: Int): Bool
199 do
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
206 myiter.next
207 itsiter.next
208 end
209 if itsiter.is_ok then return false
210 return true
211 end
212
213 # Is this string prefixed by `prefix`?
214 #
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)
219
220 # Is this string suffixed by `suffix`?
221 #
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)
225
226 # Returns `self` as the corresponding integer
227 #
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
233 #
234 # REQUIRE: `self`.`is_int`
235 fun to_i: Int is abstract
236
237 # If `self` contains a float, return the corresponding float
238 #
239 # assert "123".to_f == 123.0
240 # assert "-1".to_f == -1.0
241 # assert "-1.2e-3".to_f == -0.0012
242 fun to_f: Float
243 do
244 # Shortcut
245 return to_s.to_cstring.atof
246 end
247
248 # If `self` contains only digits and alpha <= 'f', return the corresponding integer.
249 #
250 # assert "ff".to_hex == 255
251 fun to_hex(pos, ln: nullable Int): Int do
252 var res = 0
253 if pos == null then pos = 0
254 if ln == null then ln = length - pos
255 var max = pos + ln
256 for i in [pos .. max[ do
257 res <<= 4
258 res += self[i].from_hex
259 end
260 return res
261 end
262
263 # If `self` contains only digits <= '7', return the corresponding integer.
264 #
265 # assert "714".to_oct == 460
266 fun to_oct: Int do return a_to(8)
267
268 # If `self` contains only '0' et '1', return the corresponding integer.
269 #
270 # assert "101101".to_bin == 45
271 fun to_bin: Int do return a_to(2)
272
273 # If `self` contains only digits '0' .. '9', return the corresponding integer.
274 #
275 # assert "108".to_dec == 108
276 fun to_dec: Int do return a_to(10)
277
278 # If `self` contains only digits and letters, return the corresponding integer in a given base
279 #
280 # assert "120".a_to(3) == 15
281 fun a_to(base: Int) : Int
282 do
283 var i = 0
284 var neg = false
285
286 for j in [0..length[ do
287 var c = chars[j]
288 var v = c.to_i
289 if v > base then
290 if neg then
291 return -i
292 else
293 return i
294 end
295 else if v < 0 then
296 neg = true
297 else
298 i = i * base + v
299 end
300 end
301 if neg then
302 return -i
303 else
304 return i
305 end
306 end
307
308 # Returns `true` if the string contains only Numeric values (and one "," or one "." character)
309 #
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
314 fun is_numeric: Bool
315 do
316 var has_point_or_comma = false
317 for i in [0..length[ do
318 var c = chars[i]
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
322 else
323 return false
324 end
325 end
326 end
327 return true
328 end
329
330 # Returns `true` if the string contains only Hex chars
331 #
332 # assert "048bf".is_hex == true
333 # assert "ABCDEF".is_hex == true
334 # assert "0G".is_hex == false
335 fun is_hex: Bool
336 do
337 for i in [0..length[ do
338 var c = chars[i]
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
342 end
343 return true
344 end
345
346 # Returns `true` if the string contains only Binary digits
347 #
348 # assert "1101100".is_bin == true
349 # assert "1101020".is_bin == false
350 fun is_bin: Bool do
351 for i in chars do if i != '0' and i != '1' then return false
352 return true
353 end
354
355 # Returns `true` if the string contains only Octal digits
356 #
357 # assert "213453".is_oct == true
358 # assert "781".is_oct == false
359 fun is_oct: Bool do
360 for i in chars do if i < '0' or i > '7' then return false
361 return true
362 end
363
364 # Returns `true` if the string contains only Decimal digits
365 #
366 # assert "10839".is_dec == true
367 # assert "164F".is_dec == false
368 fun is_dec: Bool do
369 for i in chars do if i < '0' or i > '9' then return false
370 return true
371 end
372
373 # Are all letters in `self` upper-case ?
374 #
375 # assert "HELLO WORLD".is_upper == true
376 # assert "%$&%!".is_upper == true
377 # assert "hello world".is_upper == false
378 # assert "Hello World".is_upper == false
379 fun is_upper: Bool
380 do
381 for i in [0..length[ do
382 var char = chars[i]
383 if char.is_lower then return false
384 end
385 return true
386 end
387
388 # Are all letters in `self` lower-case ?
389 #
390 # assert "hello world".is_lower == true
391 # assert "%$&%!".is_lower == true
392 # assert "Hello World".is_lower == false
393 fun is_lower: Bool
394 do
395 for i in [0..length[ do
396 var char = chars[i]
397 if char.is_upper then return false
398 end
399 return true
400 end
401
402 # Removes the whitespaces at the beginning of self
403 #
404 # assert " \n\thello \n\t".l_trim == "hello \n\t"
405 #
406 # `Char::is_whitespace` determines what is a whitespace.
407 fun l_trim: SELFTYPE
408 do
409 var iter = self.chars.iterator
410 while iter.is_ok do
411 if not iter.item.is_whitespace then break
412 iter.next
413 end
414 if iter.index == length then return self.empty
415 return self.substring_from(iter.index)
416 end
417
418 # Removes the whitespaces at the end of self
419 #
420 # assert " \n\thello \n\t".r_trim == " \n\thello"
421 #
422 # `Char::is_whitespace` determines what is a whitespace.
423 fun r_trim: SELFTYPE
424 do
425 var iter = self.chars.reverse_iterator
426 while iter.is_ok do
427 if not iter.item.is_whitespace then break
428 iter.next
429 end
430 if iter.index < 0 then return self.empty
431 return self.substring(0, iter.index + 1)
432 end
433
434 # Trims trailing and preceding white spaces
435 #
436 # assert " Hello World ! ".trim == "Hello World !"
437 # assert "\na\nb\tc\t".trim == "a\nb\tc"
438 #
439 # `Char::is_whitespace` determines what is a whitespace.
440 fun trim: SELFTYPE do return (self.l_trim).r_trim
441
442 # Is the string non-empty but only made of whitespaces?
443 #
444 # assert " \n\t ".is_whitespace == true
445 # assert " hello ".is_whitespace == false
446 # assert "".is_whitespace == false
447 #
448 # `Char::is_whitespace` determines what is a whitespace.
449 fun is_whitespace: Bool
450 do
451 if is_empty then return false
452 for c in self.chars do
453 if not c.is_whitespace then return false
454 end
455 return true
456 end
457
458 # Returns `self` removed from its last line terminator (if any).
459 #
460 # assert "Hello\n".chomp == "Hello"
461 # assert "Hello".chomp == "Hello"
462 #
463 # assert "\n".chomp == ""
464 # assert "".chomp == ""
465 #
466 # Line terminators are `"\n"`, `"\r\n"` and `"\r"`.
467 # A single line terminator, the last one, is removed.
468 #
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"
473 #
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.
476 fun chomp: SELFTYPE
477 do
478 var len = length
479 if len == 0 then return self
480 var l = self.chars.last
481 if l == '\r' then
482 return substring(0, len-1)
483 else if l != '\n' then
484 return self
485 else if len > 1 and self.chars[len-2] == '\r' then
486 return substring(0, len-2)
487 else
488 return substring(0, len-1)
489 end
490 end
491
492 # Justify a self in a space of `length`
493 #
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)
498 #
499 # Examples
500 #
501 # assert "hello".justify(10, 0.0) == "hello "
502 # assert "hello".justify(10, 1.0) == " hello"
503 # assert "hello".justify(10, 0.5) == " hello "
504 #
505 # If `length` is not enough, `self` is returned as is.
506 #
507 # assert "hello".justify(2, 0.0) == "hello"
508 #
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
513 do
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)
519 end
520
521 # Mangle a string to be a unique string only made of alphanumeric characters and underscores.
522 #
523 # This method is injective (two different inputs never produce the same
524 # output) and the returned string always respect the following rules:
525 #
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.
530 #
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
539 do
540 if is_empty then return ""
541 var res = new Buffer
542 var underscore = false
543 var start = 0
544 var c = chars[0]
545
546 if c >= '0' and c <= '9' then
547 res.add('_')
548 res.append(c.code_point.to_s)
549 res.add('d')
550 start = 1
551 end
552 for i in [start..length[ do
553 c = chars[i]
554 if (c >= 'a' and c <= 'z') or (c >='A' and c <= 'Z') then
555 res.add(c)
556 underscore = false
557 continue
558 end
559 if underscore then
560 res.append('_'.code_point.to_s)
561 res.add('d')
562 end
563 if c >= '0' and c <= '9' then
564 res.add(c)
565 underscore = false
566 else if c == '_' then
567 res.add(c)
568 underscore = true
569 else
570 res.add('_')
571 res.append(c.code_point.to_s)
572 res.add('d')
573 underscore = false
574 end
575 end
576 if underscore then
577 res.append('_'.code_point.to_s)
578 res.add('d')
579 end
580 return res.to_s
581 end
582
583 # Escape " \ ' and non printable characters using the rules of literal C strings and characters
584 #
585 # assert "abAB12<>&".escape_to_c == "abAB12<>&"
586 # assert "\n\"'\\".escape_to_c == "\\n\\\"\\'\\\\"
587 #
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.
591 #
592 # assert "{0.code_point}{1.code_point}{8.code_point}{31.code_point}{32.code_point}".escape_to_c == "\\000\\001\\010\\037 "
593 #
594 # The exceptions are the common `\t` and `\n`.
595 fun escape_to_c: String
596 do
597 var b = new Buffer
598 for i in [0..length[ do
599 var c = chars[i]
600 if c == '\n' then
601 b.append("\\n")
602 else if c == '\t' then
603 b.append("\\t")
604 else if c == '"' then
605 b.append("\\\"")
606 else if c == '\'' then
607 b.append("\\\'")
608 else if c == '\\' then
609 b.append("\\\\")
610 else if c.code_point < 32 then
611 b.add('\\')
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
616 b.add('0')
617 b.add('0')
618 else if oct.length == 2 then
619 b.add('0')
620 end
621 b.append(oct)
622 else
623 b.add(c)
624 end
625 end
626 return b.to_s
627 end
628
629 # Escape additionnal characters
630 # The result might no be legal in C but be used in other languages
631 #
632 # assert "ab|\{\}".escape_more_to_c("|\{\}") == "ab\\|\\\{\\\}"
633 fun escape_more_to_c(chars: String): String
634 do
635 var b = new Buffer
636 for c in escape_to_c.chars do
637 if chars.chars.has(c) then
638 b.add('\\')
639 end
640 b.add(c)
641 end
642 return b.to_s
643 end
644
645 # Escape to C plus braces
646 #
647 # assert "\n\"'\\\{\}".escape_to_nit == "\\n\\\"\\'\\\\\\\{\\\}"
648 fun escape_to_nit: String do return escape_more_to_c("\{\}")
649
650 # Escape to POSIX Shell (sh).
651 #
652 # Abort if the text contains a null byte.
653 #
654 # assert "\n\"'\\\{\}0".escape_to_sh == "'\n\"'\\''\\\{\}0'"
655 fun escape_to_sh: String do
656 var b = new Buffer
657 b.chars.add '\''
658 for i in [0..length[ do
659 var c = chars[i]
660 if c == '\'' then
661 b.append("'\\''")
662 else
663 assert without_null_byte: c != '\0'
664 b.add(c)
665 end
666 end
667 b.chars.add '\''
668 return b.to_s
669 end
670
671 # Escape to include in a Makefile
672 #
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
677 var b = new Buffer
678 for i in [0..length[ do
679 var c = chars[i]
680 if c == '$' then
681 b.append("$$")
682 else if c == ':' or c == ' ' or c == '#' then
683 b.add('\\')
684 b.add(c)
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)}")
687 else
688 b.add(c)
689 end
690 end
691 return b.to_s
692 end
693
694 # Return a string where Nit escape sequences are transformed.
695 #
696 # var s = "\\n"
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
702 do
703 var res = new Buffer.with_cap(self.length)
704 var was_slash = false
705 for i in [0..length[ do
706 var c = chars[i]
707 if not was_slash then
708 if c == '\\' then
709 was_slash = true
710 else
711 res.add(c)
712 end
713 continue
714 end
715 was_slash = false
716 if c == 'n' then
717 res.add('\n')
718 else if c == 'r' then
719 res.add('\r')
720 else if c == 't' then
721 res.add('\t')
722 else if c == '0' then
723 res.add('\0')
724 else
725 res.add(c)
726 end
727 end
728 return res.to_s
729 end
730
731 # Returns `self` with all characters escaped with their UTF-16 representation
732 #
733 # assert "Aèあ𐏓".escape_to_utf16 == "\\u0041\\u00e8\\u3042\\ud800\\udfd3"
734 fun escape_to_utf16: String do
735 var buf = new Buffer
736 for i in chars do buf.append i.escape_to_utf16
737 return buf.to_s
738 end
739
740 # Returns the Unicode char escaped by `self`
741 #
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
747 var ln = length
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
753 var cp = 0
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
758 cp += cplo - 0xDC00
759 cp += 0x10000
760 return cp.code_point
761 end
762
763 # Encode `self` to percent (or URL) encoding
764 #
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
770 do
771 var buf = new Buffer
772
773 for i in [0..length[ do
774 var c = chars[i]
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
779 c == '_' or c == '~'
780 then
781 buf.add c
782 else
783 var bytes = c.to_s.bytes
784 for b in bytes do buf.append "%{b.to_i.to_hex}"
785 end
786 end
787
788 return buf.to_s
789 end
790
791 # Decode `self` from percent (or URL) encoding to a clear string
792 #
793 # Replace invalid use of '%' with '?'.
794 #
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
803 do
804 var len = bytelen
805 var has_percent = false
806 for c in chars do
807 if c == '%' then
808 len -= 2
809 has_percent = true
810 end
811 end
812
813 # If no transformation is needed, return self as a string
814 if not has_percent then return to_s
815
816 var buf = new NativeString(len)
817 var i = 0
818 var l = 0
819 while i < length do
820 var c = chars[i]
821 if c == '%' then
822 if i + 2 >= length then
823 # What follows % has been cut off
824 buf[l] = '?'.ascii
825 else
826 i += 1
827 var hex_s = substring(i, 2)
828 if hex_s.is_hex then
829 var hex_i = hex_s.to_hex
830 buf[l] = hex_i.to_b
831 i += 1
832 else
833 # What follows a % is not Hex
834 buf[l] = '?'.ascii
835 i -= 1
836 end
837 end
838 else buf[l] = c.ascii
839
840 i += 1
841 l += 1
842 end
843
844 return buf.to_s_with_length(l)
845 end
846
847 # Escape the characters `<`, `>`, `&`, `"`, `'` and `/` as HTML/XML entity references.
848 #
849 # assert "a&b-<>\"x\"/'".html_escape == "a&amp;b-&lt;&gt;&#34;x&#34;&#47;&#39;"
850 #
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
853 do
854 var buf = new Buffer
855
856 for i in [0..length[ do
857 var c = chars[i]
858 if c == '&' then
859 buf.append "&amp;"
860 else if c == '<' then
861 buf.append "&lt;"
862 else if c == '>' then
863 buf.append "&gt;"
864 else if c == '"' then
865 buf.append "&#34;"
866 else if c == '\'' then
867 buf.append "&#39;"
868 else if c == '/' then
869 buf.append "&#47;"
870 else buf.add c
871 end
872
873 return buf.to_s
874 end
875
876 # Equality of text
877 # Two pieces of text are equals if thez have the same characters in the same order.
878 #
879 # assert "hello" == "hello"
880 # assert "hello" != "HELLO"
881 # assert "hello" == "hel"+"lo"
882 #
883 # Things that are not Text are not equal.
884 #
885 # assert "9" != '9'
886 # assert "9" != ['9']
887 # assert "9" != 9
888 #
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
892 redef fun ==(o)
893 do
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
899 end
900
901 # Lexicographical comparaison
902 #
903 # assert "abc" < "xy"
904 # assert "ABC" < "abc"
905 redef fun <(other)
906 do
907 var self_chars = self.chars.iterator
908 var other_chars = other.chars.iterator
909
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
913 self_chars.next
914 other_chars.next
915 end
916
917 if self_chars.is_ok then
918 return false
919 else
920 return true
921 end
922 end
923
924 # Escape string used in labels for graphviz
925 #
926 # assert ">><<".escape_to_dot == "\\>\\>\\<\\<"
927 fun escape_to_dot: String
928 do
929 return escape_more_to_c("|\{\}<>")
930 end
931
932 private var hash_cache: nullable Int = null
933
934 redef fun hash
935 do
936 if hash_cache == null then
937 # djb2 hash algorithm
938 var h = 5381
939
940 for i in [0..length[ do
941 var char = chars[i]
942 h = (h << 5) + h + char.code_point
943 end
944
945 hash_cache = h
946 end
947 return hash_cache.as(not null)
948 end
949
950 # Gives the formatted string back as a Nit string with `args` in place
951 #
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]
956 var curr_st = 0
957 var i = 0
958 while i < length do
959 # Skip escaped characters
960 if self[i] == '\\' then
961 i += 1
962 # In case of format
963 else if self[i] == '%' then
964 var fmt_st = i
965 i += 1
966 var ciph_st = i
967 while i < length and self[i].is_numeric do
968 i += 1
969 end
970 i -= 1
971 var fmt_end = i
972 var ciph_len = fmt_end - ciph_st + 1
973
974 var arg_index = substring(ciph_st, ciph_len).to_i - 1
975 if arg_index >= args.length then continue
976
977 s.push substring(curr_st, fmt_st - curr_st)
978 s.push args[arg_index].to_s
979 curr_st = i + 1
980 end
981 i += 1
982 end
983 s.push substring(curr_st, length - curr_st)
984 return s.plain_to_s
985 end
986
987 # Copies `n` bytes from `self` at `src_offset` into `dest` starting at `dest_offset`
988 #
989 # Basically a high-level synonym of NativeString::copy_to
990 #
991 # REQUIRE: `n` must be large enough to contain `len` bytes
992 #
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"
996 #
997 fun copy_to_native(dest: NativeString, n, src_offset, dest_offset: Int) do
998 var mypos = src_offset
999 var itspos = dest_offset
1000 while n > 0 do
1001 dest[itspos] = self.bytes[mypos]
1002 itspos += 1
1003 mypos += 1
1004 n -= 1
1005 end
1006 end
1007
1008 end
1009
1010 # All kinds of array-based text representations.
1011 abstract class FlatText
1012 super Text
1013
1014 # Underlying C-String (`char*`)
1015 #
1016 # Warning : Might be void in some subclasses, be sure to check
1017 # if set before using it.
1018 var items: NativeString is noinit
1019
1020 # Returns a char* starting at position `first_byte`
1021 #
1022 # WARNING: If you choose to use this service, be careful of the following.
1023 #
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)
1029 #
1030 # The only safe use of this pointer is if it is ephemeral (e.g. read in a C function
1031 # then immediately return).
1032 #
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
1036
1037 redef var length = 0
1038
1039 redef var bytelen = 0
1040
1041 redef fun output
1042 do
1043 var i = 0
1044 while i < length do
1045 items[i].output
1046 i += 1
1047 end
1048 end
1049
1050 redef fun copy_to_native(dest, n, src_offset, dest_offset) do
1051 items.copy_to(dest, n, src_offset, dest_offset)
1052 end
1053 end
1054
1055 # Abstract class for the SequenceRead compatible
1056 # views on the chars of any Text
1057 private abstract class StringCharView
1058 super SequenceRead[Char]
1059
1060 type SELFTYPE: Text
1061
1062 var target: SELFTYPE
1063
1064 redef fun is_empty do return target.is_empty
1065
1066 redef fun length do return target.length
1067
1068 redef fun iterator: IndexedIterator[Char] do return self.iterator_from(0)
1069
1070 redef fun reverse_iterator do return self.reverse_iterator_from(self.length - 1)
1071 end
1072
1073 # Abstract class for the SequenceRead compatible
1074 # views on the bytes of any Text
1075 private abstract class StringByteView
1076 super SequenceRead[Byte]
1077
1078 type SELFTYPE: Text
1079
1080 var target: SELFTYPE
1081
1082 redef fun is_empty do return target.is_empty
1083
1084 redef fun length do return target.bytelen
1085
1086 redef fun iterator do return self.iterator_from(0)
1087
1088 redef fun reverse_iterator do return self.reverse_iterator_from(target.bytelen - 1)
1089 end
1090
1091 # Immutable sequence of characters.
1092 #
1093 # String objects may be created using literals.
1094 #
1095 # assert "Hello World!" isa String
1096 abstract class String
1097 super Text
1098
1099 redef type SELFTYPE: String is fixed
1100
1101 redef fun to_s do return self
1102
1103 # Concatenates `o` to `self`
1104 #
1105 # assert "hello" + "world" == "helloworld"
1106 # assert "" + "hello" + "" == "hello"
1107 fun +(o: Text): SELFTYPE is abstract
1108
1109 # Concatenates self `i` times
1110 #
1111 # assert "abc" * 4 == "abcabcabcabc"
1112 # assert "abc" * 1 == "abc"
1113 # assert "abc" * 0 == ""
1114 fun *(i: Int): SELFTYPE is abstract
1115
1116 # Insert `s` at `pos`.
1117 #
1118 # assert "helloworld".insert_at(" ", 5) == "hello world"
1119 fun insert_at(s: String, pos: Int): SELFTYPE is abstract
1120
1121 redef fun substrings is abstract
1122
1123 # Returns a reversed version of self
1124 #
1125 # assert "hello".reversed == "olleh"
1126 # assert "bob".reversed == "bob"
1127 # assert "".reversed == ""
1128 fun reversed: SELFTYPE is abstract
1129
1130 # A upper case version of `self`
1131 #
1132 # assert "Hello World!".to_upper == "HELLO WORLD!"
1133 fun to_upper: SELFTYPE is abstract
1134
1135 # A lower case version of `self`
1136 #
1137 # assert "Hello World!".to_lower == "hello world!"
1138 fun to_lower : SELFTYPE is abstract
1139
1140 # Takes a camel case `self` and converts it to snake case
1141 #
1142 # assert "randomMethodId".to_snake_case == "random_method_id"
1143 #
1144 # The rules are the following:
1145 #
1146 # An uppercase is always converted to a lowercase
1147 #
1148 # assert "HELLO_WORLD".to_snake_case == "hello_world"
1149 #
1150 # An uppercase that follows a lowercase is prefixed with an underscore
1151 #
1152 # assert "HelloTheWORLD".to_snake_case == "hello_the_world"
1153 #
1154 # An uppercase that follows an uppercase and is followed by a lowercase, is prefixed with an underscore
1155 #
1156 # assert "HelloTHEWorld".to_snake_case == "hello_the_world"
1157 #
1158 # All other characters are kept as is; `self` does not need to be a proper CamelCased string.
1159 #
1160 # assert "=-_H3ll0Th3W0rld_-=".to_snake_case == "=-_h3ll0th3w0rld_-="
1161 fun to_snake_case: SELFTYPE
1162 do
1163 if self.is_lower then return self
1164
1165 var new_str = new Buffer.with_cap(self.length)
1166 var prev_is_lower = false
1167 var prev_is_upper = false
1168
1169 for i in [0..length[ do
1170 var char = chars[i]
1171 if char.is_lower then
1172 new_str.add(char)
1173 prev_is_lower = true
1174 prev_is_upper = false
1175 else if char.is_upper then
1176 if prev_is_lower then
1177 new_str.add('_')
1178 else if prev_is_upper and i+1 < length and chars[i+1].is_lower then
1179 new_str.add('_')
1180 end
1181 new_str.add(char.to_lower)
1182 prev_is_lower = false
1183 prev_is_upper = true
1184 else
1185 new_str.add(char)
1186 prev_is_lower = false
1187 prev_is_upper = false
1188 end
1189 end
1190
1191 return new_str.to_s
1192 end
1193
1194 # Takes a snake case `self` and converts it to camel case
1195 #
1196 # assert "random_method_id".to_camel_case == "randomMethodId"
1197 #
1198 # If the identifier is prefixed by an underscore, the underscore is ignored
1199 #
1200 # assert "_private_field".to_camel_case == "_privateField"
1201 #
1202 # If `self` is upper, it is returned unchanged
1203 #
1204 # assert "RANDOM_ID".to_camel_case == "RANDOM_ID"
1205 #
1206 # If there are several consecutive underscores, they are considered as a single one
1207 #
1208 # assert "random__method_id".to_camel_case == "randomMethodId"
1209 fun to_camel_case: SELFTYPE
1210 do
1211 if self.is_upper then return self
1212
1213 var new_str = new Buffer
1214 var is_first_char = true
1215 var follows_us = false
1216
1217 for i in [0..length[ do
1218 var char = chars[i]
1219 if is_first_char then
1220 new_str.add(char)
1221 is_first_char = false
1222 else if char == '_' then
1223 follows_us = true
1224 else if follows_us then
1225 new_str.add(char.to_upper)
1226 follows_us = false
1227 else
1228 new_str.add(char)
1229 end
1230 end
1231
1232 return new_str.to_s
1233 end
1234
1235 # Returns a capitalized `self`
1236 #
1237 # Letters that follow a letter are lowercased
1238 # Letters that follow a non-letter are upcased.
1239 #
1240 # SEE : `Char::is_letter` for the definition of letter.
1241 #
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
1247
1248 var buf = new Buffer.with_cap(length)
1249
1250 var curr = chars[0].to_upper
1251 var prev = curr
1252 buf[0] = curr
1253
1254 for i in [1 .. length[ do
1255 prev = curr
1256 curr = self[i]
1257 if prev.is_letter then
1258 buf[i] = curr.to_lower
1259 else
1260 buf[i] = curr.to_upper
1261 end
1262 end
1263
1264 return buf.to_s
1265 end
1266 end
1267
1268 # A mutable sequence of characters.
1269 abstract class Buffer
1270 super Text
1271
1272 # Returns an arbitrary subclass of `Buffer` with default parameters
1273 new is abstract
1274
1275 # Returns an instance of a subclass of `Buffer` with `i` base capacity
1276 new with_cap(i: Int) is abstract
1277
1278 redef type SELFTYPE: Buffer is fixed
1279
1280 # Specific implementations MUST set this to `true` in order to invalidate caches
1281 protected var is_dirty = true
1282
1283 # Copy-On-Write flag
1284 #
1285 # If the `Buffer` was to_s'd, the next in-place altering
1286 # operation will cause the current `Buffer` to be re-allocated.
1287 #
1288 # The flag will then be set at `false`.
1289 protected var written = false
1290
1291 # Modifies the char contained at pos `index`
1292 #
1293 # DEPRECATED : Use self.chars.[]= instead
1294 fun []=(index: Int, item: Char) is abstract
1295
1296 # Adds a char `c` at the end of self
1297 #
1298 # DEPRECATED : Use self.chars.add instead
1299 fun add(c: Char) is abstract
1300
1301 # Clears the buffer
1302 #
1303 # var b = new Buffer
1304 # b.append "hello"
1305 # assert not b.is_empty
1306 # b.clear
1307 # assert b.is_empty
1308 fun clear is abstract
1309
1310 # Enlarges the subsequent array containing the chars of self
1311 fun enlarge(cap: Int) is abstract
1312
1313 # Adds the content of text `s` at the end of self
1314 #
1315 # var b = new Buffer
1316 # b.append "hello"
1317 # b.append "world"
1318 # assert b == "helloworld"
1319 fun append(s: Text) is abstract
1320
1321 # `self` is appended in such a way that `self` is repeated `r` times
1322 #
1323 # var b = new Buffer
1324 # b.append "hello"
1325 # b.times 3
1326 # assert b == "hellohellohello"
1327 fun times(r: Int) is abstract
1328
1329 # Reverses itself in-place
1330 #
1331 # var b = new Buffer
1332 # b.append("hello")
1333 # b.reverse
1334 # assert b == "olleh"
1335 fun reverse is abstract
1336
1337 # Changes each lower-case char in `self` by its upper-case variant
1338 #
1339 # var b = new Buffer
1340 # b.append("Hello World!")
1341 # b.upper
1342 # assert b == "HELLO WORLD!"
1343 fun upper is abstract
1344
1345 # Changes each upper-case char in `self` by its lower-case variant
1346 #
1347 # var b = new Buffer
1348 # b.append("Hello World!")
1349 # b.lower
1350 # assert b == "hello world!"
1351 fun lower is abstract
1352
1353 # Capitalizes each word in `self`
1354 #
1355 # Letters that follow a letter are lowercased
1356 # Letters that follow a non-letter are upcased.
1357 #
1358 # SEE: `Char::is_letter` for the definition of a letter.
1359 #
1360 # var b = new FlatBuffer.from("jAVAsCriPt")
1361 # b.capitalize
1362 # assert b == "Javascript"
1363 # b = new FlatBuffer.from("i am root")
1364 # b.capitalize
1365 # assert b == "I Am Root"
1366 # b = new FlatBuffer.from("ab_c -ab0c ab\nc")
1367 # b.capitalize
1368 # assert b == "Ab_C -Ab0C Ab\nC"
1369 fun capitalize do
1370 if length == 0 then return
1371 var c = self[0].to_upper
1372 self[0] = c
1373 var prev = c
1374 for i in [1 .. length[ do
1375 prev = c
1376 c = self[i]
1377 if prev.is_letter then
1378 self[i] = c.to_lower
1379 else
1380 self[i] = c.to_upper
1381 end
1382 end
1383 end
1384
1385 redef fun hash
1386 do
1387 if is_dirty then hash_cache = null
1388 return super
1389 end
1390
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
1394 end
1395
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]
1401
1402 redef type SELFTYPE: Buffer
1403
1404 end
1405
1406 # View for bytes on Buffer objects, extends Sequence
1407 # for mutation operations
1408 private abstract class BufferByteView
1409 super StringByteView
1410
1411 redef type SELFTYPE: Buffer
1412 end
1413
1414 redef class Object
1415 # User readable representation of `self`.
1416 fun to_s: String do return inspect
1417
1418 # The class name of the object in NativeString format.
1419 private fun native_class_name: NativeString is intern
1420
1421 # The class name of the object.
1422 #
1423 # assert 5.class_name == "Int"
1424 fun class_name: String do return native_class_name.to_s
1425
1426 # Developer readable representation of `self`.
1427 # Usually, it uses the form "<CLASSNAME:#OBJECTID bla bla bla>"
1428 fun inspect: String
1429 do
1430 return "<{inspect_head}>"
1431 end
1432
1433 # Return "CLASSNAME:#OBJECTID".
1434 # This function is mainly used with the redefinition of the inspect method
1435 protected fun inspect_head: String
1436 do
1437 return "{class_name}:#{object_id.to_hex}"
1438 end
1439 end
1440
1441 redef class Bool
1442 # assert true.to_s == "true"
1443 # assert false.to_s == "false"
1444 redef fun to_s
1445 do
1446 if self then
1447 return once "true"
1448 else
1449 return once "false"
1450 end
1451 end
1452 end
1453
1454 redef class Byte
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);
1458 `}
1459
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);
1463 `}
1464
1465 # Displayable byte in its hexadecimal form (0x..)
1466 #
1467 # assert 1.to_b.to_s == "0x01"
1468 # assert (-123).to_b.to_s == "0x85"
1469 redef fun to_s do
1470 var nslen = byte_to_s_len
1471 var ns = new NativeString(nslen + 1)
1472 ns[nslen] = 0u8
1473 native_byte_to_s(ns, nslen + 1)
1474 return ns.to_s_with_length(nslen)
1475 end
1476 end
1477
1478 redef class Int
1479
1480 # Wrapper of strerror C function
1481 private fun strerror_ext: NativeString `{ return strerror(self); `}
1482
1483 # Returns a string describing error number
1484 fun strerror: String do return strerror_ext.to_s
1485
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)
1489 do
1490 var n: Int
1491 # Sign
1492 if self < 0 then
1493 n = - self
1494 s.chars[0] = '-'
1495 else if self == 0 then
1496 s.chars[0] = '0'
1497 return
1498 else
1499 n = self
1500 end
1501 # Fill digits
1502 var pos = digit_count(base) - 1
1503 while pos >= 0 and n > 0 do
1504 s.chars[pos] = (n % base).to_c
1505 n = n / base # /
1506 pos -= 1
1507 end
1508 end
1509
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);
1513 `}
1514
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);
1518 `}
1519
1520 # return displayable int in base base and signed
1521 fun to_base(base: Int, signed: Bool): String is abstract
1522
1523 # return displayable int in hexadecimal
1524 #
1525 # assert 1.to_hex == "1"
1526 # assert (-255).to_hex == "-ff"
1527 fun to_hex: String do return to_base(16,false)
1528 end
1529
1530 redef class Float
1531 # Pretty representation of `self`, with decimals as needed from 1 to a maximum of 3
1532 #
1533 # assert 12.34.to_s == "12.34"
1534 # assert (-0120.030).to_s == "-120.03"
1535 #
1536 # see `to_precision` for a custom precision.
1537 redef fun to_s do
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
1542 var j = len-1-i
1543 var c = str.chars[j]
1544 if c == '0' then
1545 continue
1546 else if c == '.' then
1547 return str.substring( 0, j+2 )
1548 else
1549 return str.substring( 0, j+1 )
1550 end
1551 end
1552 return str
1553 end
1554
1555 # `String` representation of `self` with the given number of `decimals`
1556 #
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
1564 do
1565 if is_nan then return "nan"
1566
1567 var isinf = self.is_inf
1568 if isinf == 1 then
1569 return "inf"
1570 else if isinf == -1 then
1571 return "-inf"
1572 end
1573
1574 if decimals == 0 then return self.to_i.to_s
1575 var f = self
1576 for i in [0..decimals[ do f = f * 10.0
1577 if self > 0.0 then
1578 f = f + 0.5
1579 else
1580 f = f - 0.5
1581 end
1582 var i = f.to_i
1583 if i == 0 then return "0." + "0"*decimals
1584
1585 # Prepare both parts of the float, before and after the "."
1586 var s = i.abs.to_s
1587 var sl = s.length
1588 var p1
1589 var p2
1590 if sl > decimals then
1591 # Has something before the "."
1592 p1 = s.substring(0, sl-decimals)
1593 p2 = s.substring(sl-decimals, decimals)
1594 else
1595 p1 = "0"
1596 p2 = "0"*(decimals-sl) + s
1597 end
1598
1599 if i < 0 then p1 = "-" + p1
1600
1601 return p1 + "." + p2
1602 end
1603 end
1604
1605 redef class Char
1606
1607 # Returns a sequence with the UTF-8 bytes of `self`
1608 #
1609 # assert 'a'.bytes == [0x61u8]
1610 # assert 'ま'.bytes == [0xE3u8, 0x81u8, 0xBEu8]
1611 fun bytes: SequenceRead[Byte] do return to_s.bytes
1612
1613 # Is `self` an UTF-16 surrogate pair ?
1614 fun is_surrogate: Bool do
1615 var cp = code_point
1616 return cp >= 0xD800 and cp <= 0xDFFF
1617 end
1618
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
1627 return 1
1628 end
1629
1630 # assert 'x'.to_s == "x"
1631 redef fun to_s do
1632 var ln = u8char_len
1633 var ns = new NativeString(ln + 1)
1634 u8char_tos(ns, ln)
1635 return ns.to_s_with_length(ln)
1636 end
1637
1638 # Returns `self` escaped to UTF-16
1639 #
1640 # i.e. Represents `self`.`code_point` using UTF-16 codets escaped
1641 # with a `\u`
1642 #
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
1648 var cp = code_point
1649 var buf: Buffer
1650 if cp < 0xD800 or (cp >= 0xE000 and cp <= 0xFFFF) then
1651 buf = new Buffer.with_cap(6)
1652 buf.append("\\u0000")
1653 var hx = cp.to_hex
1654 var outid = 5
1655 for i in hx.chars.reverse_iterator do
1656 buf[outid] = i
1657 outid -= 1
1658 end
1659 else
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
1664 var out = 2
1665 for i in hi do
1666 buf[out] = i
1667 out += 1
1668 end
1669 out = 8
1670 for i in lo do
1671 buf[out] = i
1672 out += 1
1673 end
1674 end
1675 return buf.to_s
1676 end
1677
1678 private fun u8char_tos(r: NativeString, len: Int) `{
1679 r[len] = '\0';
1680 switch(len){
1681 case 1:
1682 r[0] = self;
1683 break;
1684 case 2:
1685 r[0] = 0xC0 | ((self & 0x7C0) >> 6);
1686 r[1] = 0x80 | (self & 0x3F);
1687 break;
1688 case 3:
1689 r[0] = 0xE0 | ((self & 0xF000) >> 12);
1690 r[1] = 0x80 | ((self & 0xFC0) >> 6);
1691 r[2] = 0x80 | (self & 0x3F);
1692 break;
1693 case 4:
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);
1698 break;
1699 }
1700 `}
1701
1702 # Returns true if the char is a numerical digit
1703 #
1704 # assert '0'.is_numeric
1705 # assert '9'.is_numeric
1706 # assert not 'a'.is_numeric
1707 # assert not '?'.is_numeric
1708 #
1709 # FIXME: Works on ASCII-range only
1710 fun is_numeric: Bool
1711 do
1712 return self >= '0' and self <= '9'
1713 end
1714
1715 # Returns true if the char is an alpha digit
1716 #
1717 # assert 'a'.is_alpha
1718 # assert 'Z'.is_alpha
1719 # assert not '0'.is_alpha
1720 # assert not '?'.is_alpha
1721 #
1722 # FIXME: Works on ASCII-range only
1723 fun is_alpha: Bool
1724 do
1725 return (self >= 'a' and self <= 'z') or (self >= 'A' and self <= 'Z')
1726 end
1727
1728 # Is `self` an hexadecimal digit ?
1729 #
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')
1737
1738 # Returns true if the char is an alpha or a numeric digit
1739 #
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
1745 #
1746 # FIXME: Works on ASCII-range only
1747 fun is_alphanumeric: Bool
1748 do
1749 return self.is_numeric or self.is_alpha
1750 end
1751
1752 # Returns `self` to its int value
1753 #
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
1762 abort
1763 end
1764 end
1765
1766 redef class Collection[E]
1767 # String representation of the content of the collection.
1768 #
1769 # The standard representation is the list of elements separated with commas.
1770 #
1771 # ~~~
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
1775 # ~~~
1776 #
1777 # Subclasses may return a more specific string representation.
1778 redef fun to_s
1779 do
1780 return "[" + join(",") + "]"
1781 end
1782
1783 # Concatenate elements without separators
1784 #
1785 # ~~~
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
1789 # ~~~
1790 fun plain_to_s: String
1791 do
1792 var s = new Buffer
1793 for e in self do if e != null then s.append(e.to_s)
1794 return s.to_s
1795 end
1796
1797 # Concatenate and separate each elements with `separator`.
1798 #
1799 # Only concatenate if `separator == null`.
1800 #
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
1805 do
1806 if is_empty then return ""
1807
1808 var s = new Buffer # Result
1809
1810 # Concat first item
1811 var i = iterator
1812 var e = i.item
1813 if e != null then s.append(e.to_s)
1814
1815 # Concat other items
1816 i.next
1817 while i.is_ok do
1818 if separator != null then s.append(separator)
1819 e = i.item
1820 if e != null then s.append(e.to_s)
1821 i.next
1822 end
1823 return s.to_s
1824 end
1825 end
1826
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`.
1831 #
1832 # var m = new HashMap[Int, String]
1833 # m[1] = "one"
1834 # m[10] = "ten"
1835 # assert m.join("; ", "=") == "1=one; 10=ten"
1836 fun join(sep, couple_sep: String): String is abstract
1837 end
1838
1839 redef class Sys
1840 private var args_cache: nullable Sequence[String] = null
1841
1842 # The arguments of the program as given by the OS
1843 fun program_args: Sequence[String]
1844 do
1845 if _args_cache == null then init_args
1846 return _args_cache.as(not null)
1847 end
1848
1849 # The name of the program as given by the OS
1850 fun program_name: String
1851 do
1852 return native_argv(0).to_s
1853 end
1854
1855 # Initialize `program_args` with the contents of `native_argc` and `native_argv`.
1856 private fun init_args
1857 do
1858 var argc = native_argc
1859 var args = new Array[String].with_capacity(0)
1860 var i = 1
1861 while i < argc do
1862 args[i-1] = native_argv(i).to_s
1863 i += 1
1864 end
1865 _args_cache = args
1866 end
1867
1868 # First argument of the main C function.
1869 private fun native_argc: Int is intern
1870
1871 # Second argument of the main C function.
1872 private fun native_argv(i: Int): NativeString is intern
1873 end
1874
1875 # Comparator that efficienlty use `to_s` to compare things
1876 #
1877 # The comparaison call `to_s` on object and use the result to order things.
1878 #
1879 # var a = [1, 2, 3, 10, 20]
1880 # (new CachedAlphaComparator).sort(a)
1881 # assert a == [1, 10, 2, 20, 3]
1882 #
1883 # Internally the result of `to_s` is cached in a HashMap to counter
1884 # uneficient implementation of `to_s`.
1885 #
1886 # Note: it caching is not usefull, see `alpha_comparator`
1887 class CachedAlphaComparator
1888 super Comparator
1889 redef type COMPARED: Object
1890
1891 private var cache = new HashMap[Object, String]
1892
1893 private fun do_to_s(a: Object): String do
1894 if cache.has_key(a) then return cache[a]
1895 var res = a.to_s
1896 cache[a] = res
1897 return res
1898 end
1899
1900 redef fun compare(a, b) do
1901 return do_to_s(a) <=> do_to_s(b)
1902 end
1903 end
1904
1905 # see `alpha_comparator`
1906 private class AlphaComparator
1907 super Comparator
1908 redef fun compare(a, b) do return a.to_s <=> b.to_s
1909 end
1910
1911 # Stateless comparator that naively use `to_s` to compare things.
1912 #
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.
1915 #
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
1920
1921 # The arguments of the program as given by the OS
1922 fun args: Sequence[String]
1923 do
1924 return sys.program_args
1925 end
1926
1927 redef class NativeString
1928 # Returns `self` as a new String.
1929 fun to_s_with_copy: String is abstract
1930
1931 # Returns `self` as a String of `length`.
1932 fun to_s_with_length(length: Int): String is abstract
1933
1934 # Returns `self` as a String with `bytelen` and `length` set
1935 #
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
1939 end
1940
1941 redef class NativeArray[E]
1942 # Join all the elements using `to_s`
1943 #
1944 # REQUIRE: `self isa NativeArray[String]`
1945 # REQUIRE: all elements are initialized
1946 fun native_to_s: String is abstract
1947 end