src/libcore/char/methods.rs

   1 //! impl char {}
   2
   3 use crate::slice;
   4 use crate::str::from_utf8_unchecked_mut;
   5 use crate::unicode::printable::is_printable;
   6 use crate::unicode::{self, conversions};
   7
   8 use super::*;
   9
  10 #[lang = "char"]
  11 impl char {
  12     /// The highest valid code point a `char` can have.
  13     ///
  14     /// A `char` is a [Unicode Scalar Value], which means that it is a [Code
  15     /// Point], but only ones within a certain range. `MAX` is the highest valid
  16     /// code point that's a valid [Unicode Scalar Value].
  17     ///
  18     /// [Unicode Scalar Value]: http://www.unicode.org/glossary/#unicode_scalar_value
  19     /// [Code Point]: http://www.unicode.org/glossary/#code_point
  20     #[unstable(feature = "assoc_char_consts", reason = "recently added", issue = "71763")]
  21     pub const MAX: char = '\u{10ffff}';
  22
  23     /// `U+FFFD REPLACEMENT CHARACTER` (�) is used in Unicode to represent a
  24     /// decoding error.
  25     ///
  26     /// It can occur, for example, when giving ill-formed UTF-8 bytes to
  27     /// [`String::from_utf8_lossy`](string/struct.String.html#method.from_utf8_lossy).
  28     #[unstable(feature = "assoc_char_consts", reason = "recently added", issue = "71763")]
  29     pub const REPLACEMENT_CHARACTER: char = '\u{FFFD}';
  30
  31     /// The version of [Unicode](http://www.unicode.org/) that the Unicode parts of
  32     /// `char` and `str` methods are based on.
  33     ///
  34     /// New versions of Unicode are released regularly and subsequently all methods
  35     /// in the standard library depending on Unicode are updated. Therefore the
  36     /// behavior of some `char` and `str` methods and the value of this constant
  37     /// changes over time. This is *not* considered to be a breaking change.
  38     ///
  39     /// The version numbering scheme is explained in
  40     /// [Unicode 11.0 or later, Section 3.1 Versions of the Unicode Standard](https://www.unicode.org/versions/Unicode11.0.0/ch03.pdf#page=4).
  41     #[unstable(feature = "assoc_char_consts", reason = "recently added", issue = "71763")]
  42     pub const UNICODE_VERSION: (u8, u8, u8) = crate::unicode::UNICODE_VERSION;
  43
  44     /// Creates an iterator over the UTF-16 encoded code points in `iter`,
  45     /// returning unpaired surrogates as `Err`s.
  46     ///
  47     /// # Examples
  48     ///
  49     /// Basic usage:
  50     ///
  51     /// ```
  52     /// use std::char::decode_utf16;
  53     ///
  54     /// // 𝄞mus<invalid>ic<invalid>
  55     /// let v = [
  56     ///     0xD834, 0xDD1E, 0x006d, 0x0075, 0x0073, 0xDD1E, 0x0069, 0x0063, 0xD834,
  57     /// ];
  58     ///
  59     /// assert_eq!(
  60     ///     decode_utf16(v.iter().cloned())
  61     ///         .map(|r| r.map_err(|e| e.unpaired_surrogate()))
  62     ///         .collect::<Vec<_>>(),
  63     ///     vec![
  64     ///         Ok('𝄞'),
  65     ///         Ok('m'), Ok('u'), Ok('s'),
  66     ///         Err(0xDD1E),
  67     ///         Ok('i'), Ok('c'),
  68     ///         Err(0xD834)
  69     ///     ]
  70     /// );
  71     /// ```
  72     ///
  73     /// A lossy decoder can be obtained by replacing `Err` results with the replacement character:
  74     ///
  75     /// ```
  76     /// use std::char::{decode_utf16, REPLACEMENT_CHARACTER};
  77     ///
  78     /// // 𝄞mus<invalid>ic<invalid>
  79     /// let v = [
  80     ///     0xD834, 0xDD1E, 0x006d, 0x0075, 0x0073, 0xDD1E, 0x0069, 0x0063, 0xD834,
  81     /// ];
  82     ///
  83     /// assert_eq!(
  84     ///     decode_utf16(v.iter().cloned())
  85     ///        .map(|r| r.unwrap_or(REPLACEMENT_CHARACTER))
  86     ///        .collect::<String>(),
  87     ///     "𝄞mus�ic�"
  88     /// );
  89     /// ```
  90     #[unstable(feature = "assoc_char_funcs", reason = "recently added", issue = "71763")]
  91     #[inline]
  92     pub fn decode_utf16<I: IntoIterator<Item = u16>>(iter: I) -> DecodeUtf16<I::IntoIter> {
  93         super::decode::decode_utf16(iter)
  94     }
  95
  96     /// Converts a `u32` to a `char`.
  97     ///
  98     /// Note that all `char`s are valid [`u32`]s, and can be cast to one with
  99     /// `as`:
 100     ///
 101     /// ```
 102     /// let c = '💯';
 103     /// let i = c as u32;
 104     ///
 105     /// assert_eq!(128175, i);
 106     /// ```
 107     ///
 108     /// However, the reverse is not true: not all valid [`u32`]s are valid
 109     /// `char`s. `from_u32()` will return `None` if the input is not a valid value
 110     /// for a `char`.
 111     ///
 112     /// [`u32`]: primitive.u32.html
 113     ///
 114     /// For an unsafe version of this function which ignores these checks, see
 115     /// [`from_u32_unchecked`].
 116     ///
 117     /// [`from_u32_unchecked`]: #method.from_u32_unchecked
 118     ///
 119     /// # Examples
 120     ///
 121     /// Basic usage:
 122     ///
 123     /// ```
 124     /// use std::char;
 125     ///
 126     /// let c = char::from_u32(0x2764);
 127     ///
 128     /// assert_eq!(Some('❤'), c);
 129     /// ```
 130     ///
 131     /// Returning `None` when the input is not a valid `char`:
 132     ///
 133     /// ```
 134     /// use std::char;
 135     ///
 136     /// let c = char::from_u32(0x110000);
 137     ///
 138     /// assert_eq!(None, c);
 139     /// ```
 140     #[unstable(feature = "assoc_char_funcs", reason = "recently added", issue = "71763")]
 141     #[inline]
 142     pub fn from_u32(i: u32) -> Option<char> {
 143         super::convert::from_u32(i)
 144     }
 145
 146     /// Converts a `u32` to a `char`, ignoring validity.
 147     ///
 148     /// Note that all `char`s are valid [`u32`]s, and can be cast to one with
 149     /// `as`:
 150     ///
 151     /// ```
 152     /// let c = '💯';
 153     /// let i = c as u32;
 154     ///
 155     /// assert_eq!(128175, i);
 156     /// ```
 157     ///
 158     /// However, the reverse is not true: not all valid [`u32`]s are valid
 159     /// `char`s. `from_u32_unchecked()` will ignore this, and blindly cast to
 160     /// `char`, possibly creating an invalid one.
 161     ///
 162     /// [`u32`]: primitive.u32.html
 163     ///
 164     /// # Safety
 165     ///
 166     /// This function is unsafe, as it may construct invalid `char` values.
 167     ///
 168     /// For a safe version of this function, see the [`from_u32`] function.
 169     ///
 170     /// [`from_u32`]: #method.from_u32
 171     ///
 172     /// # Examples
 173     ///
 174     /// Basic usage:
 175     ///
 176     /// ```
 177     /// use std::char;
 178     ///
 179     /// let c = unsafe { char::from_u32_unchecked(0x2764) };
 180     ///
 181     /// assert_eq!('❤', c);
 182     /// ```
 183     #[unstable(feature = "assoc_char_funcs", reason = "recently added", issue = "71763")]
 184     #[inline]
 185     pub unsafe fn from_u32_unchecked(i: u32) -> char {
 186         // SAFETY: the safety contract must be upheld by the caller.
 187         unsafe { super::convert::from_u32_unchecked(i) }
 188     }
 189
 190     /// Converts a digit in the given radix to a `char`.
 191     ///
 192     /// A 'radix' here is sometimes also called a 'base'. A radix of two
 193     /// indicates a binary number, a radix of ten, decimal, and a radix of
 194     /// sixteen, hexadecimal, to give some common values. Arbitrary
 195     /// radices are supported.
 196     ///
 197     /// `from_digit()` will return `None` if the input is not a digit in
 198     /// the given radix.
 199     ///
 200     /// # Panics
 201     ///
 202     /// Panics if given a radix larger than 36.
 203     ///
 204     /// # Examples
 205     ///
 206     /// Basic usage:
 207     ///
 208     /// ```
 209     /// use std::char;
 210     ///
 211     /// let c = char::from_digit(4, 10);
 212     ///
 213     /// assert_eq!(Some('4'), c);
 214     ///
 215     /// // Decimal 11 is a single digit in base 16
 216     /// let c = char::from_digit(11, 16);
 217     ///
 218     /// assert_eq!(Some('b'), c);
 219     /// ```
 220     ///
 221     /// Returning `None` when the input is not a digit:
 222     ///
 223     /// ```
 224     /// use std::char;
 225     ///
 226     /// let c = char::from_digit(20, 10);
 227     ///
 228     /// assert_eq!(None, c);
 229     /// ```
 230     ///
 231     /// Passing a large radix, causing a panic:
 232     ///
 233     /// ```should_panic
 234     /// use std::char;
 235     ///
 236     /// // this panics
 237     /// char::from_digit(1, 37);
 238     /// ```
 239     #[unstable(feature = "assoc_char_funcs", reason = "recently added", issue = "71763")]
 240     #[inline]
 241     pub fn from_digit(num: u32, radix: u32) -> Option<char> {
 242         super::convert::from_digit(num, radix)
 243     }
 244
 245     /// Checks if a `char` is a digit in the given radix.
 246     ///
 247     /// A 'radix' here is sometimes also called a 'base'. A radix of two
 248     /// indicates a binary number, a radix of ten, decimal, and a radix of
 249     /// sixteen, hexadecimal, to give some common values. Arbitrary
 250     /// radices are supported.
 251     ///
 252     /// Compared to `is_numeric()`, this function only recognizes the characters
 253     /// `0-9`, `a-z` and `A-Z`.
 254     ///
 255     /// 'Digit' is defined to be only the following characters:
 256     ///
 257     /// * `0-9`
 258     /// * `a-z`
 259     /// * `A-Z`
 260     ///
 261     /// For a more comprehensive understanding of 'digit', see [`is_numeric`][is_numeric].
 262     ///
 263     /// [is_numeric]: #method.is_numeric
 264     ///
 265     /// # Panics
 266     ///
 267     /// Panics if given a radix larger than 36.
 268     ///
 269     /// # Examples
 270     ///
 271     /// Basic usage:
 272     ///
 273     /// ```
 274     /// assert!('1'.is_digit(10));
 275     /// assert!('f'.is_digit(16));
 276     /// assert!(!'f'.is_digit(10));
 277     /// ```
 278     ///
 279     /// Passing a large radix, causing a panic:
 280     ///
 281     /// ```should_panic
 282     /// // this panics
 283     /// '1'.is_digit(37);
 284     /// ```
 285     #[stable(feature = "rust1", since = "1.0.0")]
 286     #[inline]
 287     pub fn is_digit(self, radix: u32) -> bool {
 288         self.to_digit(radix).is_some()
 289     }
 290
 291     /// Converts a `char` to a digit in the given radix.
 292     ///
 293     /// A 'radix' here is sometimes also called a 'base'. A radix of two
 294     /// indicates a binary number, a radix of ten, decimal, and a radix of
 295     /// sixteen, hexadecimal, to give some common values. Arbitrary
 296     /// radices are supported.
 297     ///
 298     /// 'Digit' is defined to be only the following characters:
 299     ///
 300     /// * `0-9`
 301     /// * `a-z`
 302     /// * `A-Z`
 303     ///
 304     /// # Errors
 305     ///
 306     /// Returns `None` if the `char` does not refer to a digit in the given radix.
 307     ///
 308     /// # Panics
 309     ///
 310     /// Panics if given a radix larger than 36.
 311     ///
 312     /// # Examples
 313     ///
 314     /// Basic usage:
 315     ///
 316     /// ```
 317     /// assert_eq!('1'.to_digit(10), Some(1));
 318     /// assert_eq!('f'.to_digit(16), Some(15));
 319     /// ```
 320     ///
 321     /// Passing a non-digit results in failure:
 322     ///
 323     /// ```
 324     /// assert_eq!('f'.to_digit(10), None);
 325     /// assert_eq!('z'.to_digit(16), None);
 326     /// ```
 327     ///
 328     /// Passing a large radix, causing a panic:
 329     ///
 330     /// ```should_panic
 331     /// // this panics
 332     /// '1'.to_digit(37);
 333     /// ```
 334     #[stable(feature = "rust1", since = "1.0.0")]
 335     #[inline]
 336     pub fn to_digit(self, radix: u32) -> Option<u32> {
 337         assert!(radix <= 36, "to_digit: radix is too high (maximum 36)");
 338
 339         // the code is split up here to improve execution speed for cases where
 340         // the `radix` is constant and 10 or smaller
 341         let val = if radix <= 10 {
 342             match self {
 343                 '0'..='9' => self as u32 - '0' as u32,
 344                 _ => return None,
 345             }
 346         } else {
 347             match self {
 348                 '0'..='9' => self as u32 - '0' as u32,
 349                 'a'..='z' => self as u32 - 'a' as u32 + 10,
 350                 'A'..='Z' => self as u32 - 'A' as u32 + 10,
 351                 _ => return None,
 352             }
 353         };
 354
 355         if val < radix { Some(val) } else { None }
 356     }
 357
 358     /// Returns an iterator that yields the hexadecimal Unicode escape of a
 359     /// character as `char`s.
 360     ///
 361     /// This will escape characters with the Rust syntax of the form
 362     /// `\u{NNNNNN}` where `NNNNNN` is a hexadecimal representation.
 363     ///
 364     /// # Examples
 365     ///
 366     /// As an iterator:
 367     ///
 368     /// ```
 369     /// for c in '❤'.escape_unicode() {
 370     ///     print!("{}", c);
 371     /// }
 372     /// println!();
 373     /// ```
 374     ///
 375     /// Using `println!` directly:
 376     ///
 377     /// ```
 378     /// println!("{}", '❤'.escape_unicode());
 379     /// ```
 380     ///
 381     /// Both are equivalent to:
 382     ///
 383     /// ```
 384     /// println!("\\u{{2764}}");
 385     /// ```
 386     ///
 387     /// Using `to_string`:
 388     ///
 389     /// ```
 390     /// assert_eq!('❤'.escape_unicode().to_string(), "\\u{2764}");
 391     /// ```
 392     #[stable(feature = "rust1", since = "1.0.0")]
 393     #[inline]
 394     pub fn escape_unicode(self) -> EscapeUnicode {
 395         let c = self as u32;
 396
 397         // or-ing 1 ensures that for c==0 the code computes that one
 398         // digit should be printed and (which is the same) avoids the
 399         // (31 - 32) underflow
 400         let msb = 31 - (c | 1).leading_zeros();
 401
 402         // the index of the most significant hex digit
 403         let ms_hex_digit = msb / 4;
 404         EscapeUnicode {
 405             c: self,
 406             state: EscapeUnicodeState::Backslash,
 407             hex_digit_idx: ms_hex_digit as usize,
 408         }
 409     }
 410
 411     /// An extended version of `escape_debug` that optionally permits escaping
 412     /// Extended Grapheme codepoints. This allows us to format characters like
 413     /// nonspacing marks better when they're at the start of a string.
 414     #[inline]
 415     pub(crate) fn escape_debug_ext(self, escape_grapheme_extended: bool) -> EscapeDebug {
 416         let init_state = match self {
 417             '\t' => EscapeDefaultState::Backslash('t'),
 418             '\r' => EscapeDefaultState::Backslash('r'),
 419             '\n' => EscapeDefaultState::Backslash('n'),
 420             '\\' | '\'' | '"' => EscapeDefaultState::Backslash(self),
 421             _ if escape_grapheme_extended && self.is_grapheme_extended() => {
 422                 EscapeDefaultState::Unicode(self.escape_unicode())
 423             }
 424             _ if is_printable(self) => EscapeDefaultState::Char(self),
 425             _ => EscapeDefaultState::Unicode(self.escape_unicode()),
 426         };
 427         EscapeDebug(EscapeDefault { state: init_state })
 428     }
 429
 430     /// Returns an iterator that yields the literal escape code of a character
 431     /// as `char`s.
 432     ///
 433     /// This will escape the characters similar to the `Debug` implementations
 434     /// of `str` or `char`.
 435     ///
 436     /// # Examples
 437     ///
 438     /// As an iterator:
 439     ///
 440     /// ```
 441     /// for c in '\n'.escape_debug() {
 442     ///     print!("{}", c);
 443     /// }
 444     /// println!();
 445     /// ```
 446     ///
 447     /// Using `println!` directly:
 448     ///
 449     /// ```
 450     /// println!("{}", '\n'.escape_debug());
 451     /// ```
 452     ///
 453     /// Both are equivalent to:
 454     ///
 455     /// ```
 456     /// println!("\\n");
 457     /// ```
 458     ///
 459     /// Using `to_string`:
 460     ///
 461     /// ```
 462     /// assert_eq!('\n'.escape_debug().to_string(), "\\n");
 463     /// ```
 464     #[stable(feature = "char_escape_debug", since = "1.20.0")]
 465     #[inline]
 466     pub fn escape_debug(self) -> EscapeDebug {
 467         self.escape_debug_ext(true)
 468     }
 469
 470     /// Returns an iterator that yields the literal escape code of a character
 471     /// as `char`s.
 472     ///
 473     /// The default is chosen with a bias toward producing literals that are
 474     /// legal in a variety of languages, including C++11 and similar C-family
 475     /// languages. The exact rules are:
 476     ///
 477     /// * Tab is escaped as `\t`.
 478     /// * Carriage return is escaped as `\r`.
 479     /// * Line feed is escaped as `\n`.
 480     /// * Single quote is escaped as `\'`.
 481     /// * Double quote is escaped as `\"`.
 482     /// * Backslash is escaped as `\\`.
 483     /// * Any character in the 'printable ASCII' range `0x20` .. `0x7e`
 484     ///   inclusive is not escaped.
 485     /// * All other characters are given hexadecimal Unicode escapes; see
 486     ///   [`escape_unicode`][escape_unicode].
 487     ///
 488     /// [escape_unicode]: #method.escape_unicode
 489     ///
 490     /// # Examples
 491     ///
 492     /// As an iterator:
 493     ///
 494     /// ```
 495     /// for c in '"'.escape_default() {
 496     ///     print!("{}", c);
 497     /// }
 498     /// println!();
 499     /// ```
 500     ///
 501     /// Using `println!` directly:
 502     ///
 503     /// ```
 504     /// println!("{}", '"'.escape_default());
 505     /// ```
 506     ///
 507     ///
 508     /// Both are equivalent to:
 509     ///
 510     /// ```
 511     /// println!("\\\"");
 512     /// ```
 513     ///
 514     /// Using `to_string`:
 515     ///
 516     /// ```
 517     /// assert_eq!('"'.escape_default().to_string(), "\\\"");
 518     /// ```
 519     #[stable(feature = "rust1", since = "1.0.0")]
 520     #[inline]
 521     pub fn escape_default(self) -> EscapeDefault {
 522         let init_state = match self {
 523             '\t' => EscapeDefaultState::Backslash('t'),
 524             '\r' => EscapeDefaultState::Backslash('r'),
 525             '\n' => EscapeDefaultState::Backslash('n'),
 526             '\\' | '\'' | '"' => EscapeDefaultState::Backslash(self),
 527             '\x20'..='\x7e' => EscapeDefaultState::Char(self),
 528             _ => EscapeDefaultState::Unicode(self.escape_unicode()),
 529         };
 530         EscapeDefault { state: init_state }
 531     }
 532
 533     /// Returns the number of bytes this `char` would need if encoded in UTF-8.
 534     ///
 535     /// That number of bytes is always between 1 and 4, inclusive.
 536     ///
 537     /// # Examples
 538     ///
 539     /// Basic usage:
 540     ///
 541     /// ```
 542     /// let len = 'A'.len_utf8();
 543     /// assert_eq!(len, 1);
 544     ///
 545     /// let len = 'ß'.len_utf8();
 546     /// assert_eq!(len, 2);
 547     ///
 548     /// let len = 'ℝ'.len_utf8();
 549     /// assert_eq!(len, 3);
 550     ///
 551     /// let len = '💣'.len_utf8();
 552     /// assert_eq!(len, 4);
 553     /// ```
 554     ///
 555     /// The `&str` type guarantees that its contents are UTF-8, and so we can compare the length it
 556     /// would take if each code point was represented as a `char` vs in the `&str` itself:
 557     ///
 558     /// ```
 559     /// // as chars
 560     /// let eastern = '東';
 561     /// let capital = '京';
 562     ///
 563     /// // both can be represented as three bytes
 564     /// assert_eq!(3, eastern.len_utf8());
 565     /// assert_eq!(3, capital.len_utf8());
 566     ///
 567     /// // as a &str, these two are encoded in UTF-8
 568     /// let tokyo = "東京";
 569     ///
 570     /// let len = eastern.len_utf8() + capital.len_utf8();
 571     ///
 572     /// // we can see that they take six bytes total...
 573     /// assert_eq!(6, tokyo.len());
 574     ///
 575     /// // ... just like the &str
 576     /// assert_eq!(len, tokyo.len());
 577     /// ```
 578     #[stable(feature = "rust1", since = "1.0.0")]
 579     #[inline]
 580     pub fn len_utf8(self) -> usize {
 581         len_utf8(self as u32)
 582     }
 583
 584     /// Returns the number of 16-bit code units this `char` would need if
 585     /// encoded in UTF-16.
 586     ///
 587     /// See the documentation for [`len_utf8`] for more explanation of this
 588     /// concept. This function is a mirror, but for UTF-16 instead of UTF-8.
 589     ///
 590     /// [`len_utf8`]: #method.len_utf8
 591     ///
 592     /// # Examples
 593     ///
 594     /// Basic usage:
 595     ///
 596     /// ```
 597     /// let n = 'ß'.len_utf16();
 598     /// assert_eq!(n, 1);
 599     ///
 600     /// let len = '💣'.len_utf16();
 601     /// assert_eq!(len, 2);
 602     /// ```
 603     #[stable(feature = "rust1", since = "1.0.0")]
 604     #[inline]
 605     pub fn len_utf16(self) -> usize {
 606         let ch = self as u32;
 607         if (ch & 0xFFFF) == ch { 1 } else { 2 }
 608     }
 609
 610     /// Encodes this character as UTF-8 into the provided byte buffer,
 611     /// and then returns the subslice of the buffer that contains the encoded character.
 612     ///
 613     /// # Panics
 614     ///
 615     /// Panics if the buffer is not large enough.
 616     /// A buffer of length four is large enough to encode any `char`.
 617     ///
 618     /// # Examples
 619     ///
 620     /// In both of these examples, 'ß' takes two bytes to encode.
 621     ///
 622     /// ```
 623     /// let mut b = [0; 2];
 624     ///
 625     /// let result = 'ß'.encode_utf8(&mut b);
 626     ///
 627     /// assert_eq!(result, "ß");
 628     ///
 629     /// assert_eq!(result.len(), 2);
 630     /// ```
 631     ///
 632     /// A buffer that's too small:
 633     ///
 634     /// ```should_panic
 635     /// let mut b = [0; 1];
 636     ///
 637     /// // this panics
 638     /// 'ß'.encode_utf8(&mut b);
 639     /// ```
 640     #[stable(feature = "unicode_encode_char", since = "1.15.0")]
 641     #[inline]
 642     pub fn encode_utf8(self, dst: &mut [u8]) -> &mut str {
 643         // SAFETY: `char` is not a surrogate, so this is valid UTF-8.
 644         unsafe { from_utf8_unchecked_mut(encode_utf8_raw(self as u32, dst)) }
 645     }
 646
 647     /// Encodes this character as UTF-16 into the provided `u16` buffer,
 648     /// and then returns the subslice of the buffer that contains the encoded character.
 649     ///
 650     /// # Panics
 651     ///
 652     /// Panics if the buffer is not large enough.
 653     /// A buffer of length 2 is large enough to encode any `char`.
 654     ///
 655     /// # Examples
 656     ///
 657     /// In both of these examples, '𝕊' takes two `u16`s to encode.
 658     ///
 659     /// ```
 660     /// let mut b = [0; 2];
 661     ///
 662     /// let result = '𝕊'.encode_utf16(&mut b);
 663     ///
 664     /// assert_eq!(result.len(), 2);
 665     /// ```
 666     ///
 667     /// A buffer that's too small:
 668     ///
 669     /// ```should_panic
 670     /// let mut b = [0; 1];
 671     ///
 672     /// // this panics
 673     /// '𝕊'.encode_utf16(&mut b);
 674     /// ```
 675     #[stable(feature = "unicode_encode_char", since = "1.15.0")]
 676     #[inline]
 677     pub fn encode_utf16(self, dst: &mut [u16]) -> &mut [u16] {
 678         encode_utf16_raw(self as u32, dst)
 679     }
 680
 681     /// Returns `true` if this `char` has the `Alphabetic` property.
 682     ///
 683     /// `Alphabetic` is described in Chapter 4 (Character Properties) of the [Unicode Standard] and
 684     /// specified in the [Unicode Character Database][ucd] [`DerivedCoreProperties.txt`].
 685     ///
 686     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 687     /// [ucd]: https://www.unicode.org/reports/tr44/
 688     /// [`DerivedCoreProperties.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/DerivedCoreProperties.txt
 689     ///
 690     /// # Examples
 691     ///
 692     /// Basic usage:
 693     ///
 694     /// ```
 695     /// assert!('a'.is_alphabetic());
 696     /// assert!('京'.is_alphabetic());
 697     ///
 698     /// let c = '💝';
 699     /// // love is many things, but it is not alphabetic
 700     /// assert!(!c.is_alphabetic());
 701     /// ```
 702     #[stable(feature = "rust1", since = "1.0.0")]
 703     #[inline]
 704     pub fn is_alphabetic(self) -> bool {
 705         match self {
 706             'a'..='z' | 'A'..='Z' => true,
 707             c => c > '\x7f' && unicode::Alphabetic(c),
 708         }
 709     }
 710
 711     /// Returns `true` if this `char` has the `Lowercase` property.
 712     ///
 713     /// `Lowercase` is described in Chapter 4 (Character Properties) of the [Unicode Standard] and
 714     /// specified in the [Unicode Character Database][ucd] [`DerivedCoreProperties.txt`].
 715     ///
 716     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 717     /// [ucd]: https://www.unicode.org/reports/tr44/
 718     /// [`DerivedCoreProperties.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/DerivedCoreProperties.txt
 719     ///
 720     /// # Examples
 721     ///
 722     /// Basic usage:
 723     ///
 724     /// ```
 725     /// assert!('a'.is_lowercase());
 726     /// assert!('δ'.is_lowercase());
 727     /// assert!(!'A'.is_lowercase());
 728     /// assert!(!'Δ'.is_lowercase());
 729     ///
 730     /// // The various Chinese scripts and punctuation do not have case, and so:
 731     /// assert!(!'中'.is_lowercase());
 732     /// assert!(!' '.is_lowercase());
 733     /// ```
 734     #[stable(feature = "rust1", since = "1.0.0")]
 735     #[inline]
 736     pub fn is_lowercase(self) -> bool {
 737         match self {
 738             'a'..='z' => true,
 739             c => c > '\x7f' && unicode::Lowercase(c),
 740         }
 741     }
 742
 743     /// Returns `true` if this `char` has the `Uppercase` property.
 744     ///
 745     /// `Uppercase` is described in Chapter 4 (Character Properties) of the [Unicode Standard] and
 746     /// specified in the [Unicode Character Database][ucd] [`DerivedCoreProperties.txt`].
 747     ///
 748     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 749     /// [ucd]: https://www.unicode.org/reports/tr44/
 750     /// [`DerivedCoreProperties.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/DerivedCoreProperties.txt
 751     ///
 752     /// # Examples
 753     ///
 754     /// Basic usage:
 755     ///
 756     /// ```
 757     /// assert!(!'a'.is_uppercase());
 758     /// assert!(!'δ'.is_uppercase());
 759     /// assert!('A'.is_uppercase());
 760     /// assert!('Δ'.is_uppercase());
 761     ///
 762     /// // The various Chinese scripts and punctuation do not have case, and so:
 763     /// assert!(!'中'.is_uppercase());
 764     /// assert!(!' '.is_uppercase());
 765     /// ```
 766     #[stable(feature = "rust1", since = "1.0.0")]
 767     #[inline]
 768     pub fn is_uppercase(self) -> bool {
 769         match self {
 770             'A'..='Z' => true,
 771             c => c > '\x7f' && unicode::Uppercase(c),
 772         }
 773     }
 774
 775     /// Returns `true` if this `char` has the `White_Space` property.
 776     ///
 777     /// `White_Space` is specified in the [Unicode Character Database][ucd] [`PropList.txt`].
 778     ///
 779     /// [ucd]: https://www.unicode.org/reports/tr44/
 780     /// [`PropList.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/PropList.txt
 781     ///
 782     /// # Examples
 783     ///
 784     /// Basic usage:
 785     ///
 786     /// ```
 787     /// assert!(' '.is_whitespace());
 788     ///
 789     /// // a non-breaking space
 790     /// assert!('\u{A0}'.is_whitespace());
 791     ///
 792     /// assert!(!'越'.is_whitespace());
 793     /// ```
 794     #[stable(feature = "rust1", since = "1.0.0")]
 795     #[inline]
 796     pub fn is_whitespace(self) -> bool {
 797         match self {
 798             ' ' | '\x09'..='\x0d' => true,
 799             c => c > '\x7f' && unicode::White_Space(c),
 800         }
 801     }
 802
 803     /// Returns `true` if this `char` satisfies either [`is_alphabetic()`] or [`is_numeric()`].
 804     ///
 805     /// [`is_alphabetic()`]: #method.is_alphabetic
 806     /// [`is_numeric()`]: #method.is_numeric
 807     ///
 808     /// # Examples
 809     ///
 810     /// Basic usage:
 811     ///
 812     /// ```
 813     /// assert!('٣'.is_alphanumeric());
 814     /// assert!('7'.is_alphanumeric());
 815     /// assert!('৬'.is_alphanumeric());
 816     /// assert!('¾'.is_alphanumeric());
 817     /// assert!('①'.is_alphanumeric());
 818     /// assert!('K'.is_alphanumeric());
 819     /// assert!('و'.is_alphanumeric());
 820     /// assert!('藏'.is_alphanumeric());
 821     /// ```
 822     #[stable(feature = "rust1", since = "1.0.0")]
 823     #[inline]
 824     pub fn is_alphanumeric(self) -> bool {
 825         self.is_alphabetic() || self.is_numeric()
 826     }
 827
 828     /// Returns `true` if this `char` has the general category for control codes.
 829     ///
 830     /// Control codes (code points with the general category of `Cc`) are described in Chapter 4
 831     /// (Character Properties) of the [Unicode Standard] and specified in the [Unicode Character
 832     /// Database][ucd] [`UnicodeData.txt`].
 833     ///
 834     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 835     /// [ucd]: https://www.unicode.org/reports/tr44/
 836     /// [`UnicodeData.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
 837     ///
 838     /// # Examples
 839     ///
 840     /// Basic usage:
 841     ///
 842     /// ```
 843     /// // U+009C, STRING TERMINATOR
 844     /// assert!('\9c'.is_control());
 845     /// assert!(!'q'.is_control());
 846     /// ```
 847     #[stable(feature = "rust1", since = "1.0.0")]
 848     #[inline]
 849     pub fn is_control(self) -> bool {
 850         unicode::Cc(self)
 851     }
 852
 853     /// Returns `true` if this `char` has the `Grapheme_Extend` property.
 854     ///
 855     /// `Grapheme_Extend` is described in [Unicode Standard Annex #29 (Unicode Text
 856     /// Segmentation)][uax29] and specified in the [Unicode Character Database][ucd]
 857     /// [`DerivedCoreProperties.txt`].
 858     ///
 859     /// [uax29]: https://www.unicode.org/reports/tr29/
 860     /// [ucd]: https://www.unicode.org/reports/tr44/
 861     /// [`DerivedCoreProperties.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/DerivedCoreProperties.txt
 862     #[inline]
 863     pub(crate) fn is_grapheme_extended(self) -> bool {
 864         unicode::Grapheme_Extend(self)
 865     }
 866
 867     /// Returns `true` if this `char` has one of the general categories for numbers.
 868     ///
 869     /// The general categories for numbers (`Nd` for decimal digits, `Nl` for letter-like numeric
 870     /// characters, and `No` for other numeric characters) are specified in the [Unicode Character
 871     /// Database][ucd] [`UnicodeData.txt`].
 872     ///
 873     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 874     /// [ucd]: https://www.unicode.org/reports/tr44/
 875     /// [`UnicodeData.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
 876     ///
 877     /// # Examples
 878     ///
 879     /// Basic usage:
 880     ///
 881     /// ```
 882     /// assert!('٣'.is_numeric());
 883     /// assert!('7'.is_numeric());
 884     /// assert!('৬'.is_numeric());
 885     /// assert!('¾'.is_numeric());
 886     /// assert!('①'.is_numeric());
 887     /// assert!(!'K'.is_numeric());
 888     /// assert!(!'و'.is_numeric());
 889     /// assert!(!'藏'.is_numeric());
 890     /// ```
 891     #[stable(feature = "rust1", since = "1.0.0")]
 892     #[inline]
 893     pub fn is_numeric(self) -> bool {
 894         match self {
 895             '0'..='9' => true,
 896             c => c > '\x7f' && unicode::N(c),
 897         }
 898     }
 899
 900     /// Returns an iterator that yields the lowercase mapping of this `char` as one or more
 901     /// `char`s.
 902     ///
 903     /// If this `char` does not have a lowercase mapping, the iterator yields the same `char`.
 904     ///
 905     /// If this `char` has a one-to-one lowercase mapping given by the [Unicode Character
 906     /// Database][ucd] [`UnicodeData.txt`], the iterator yields that `char`.
 907     ///
 908     /// [ucd]: https://www.unicode.org/reports/tr44/
 909     /// [`UnicodeData.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
 910     ///
 911     /// If this `char` requires special considerations (e.g. multiple `char`s) the iterator yields
 912     /// the `char`(s) given by [`SpecialCasing.txt`].
 913     ///
 914     /// [`SpecialCasing.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/SpecialCasing.txt
 915     ///
 916     /// This operation performs an unconditional mapping without tailoring. That is, the conversion
 917     /// is independent of context and language.
 918     ///
 919     /// In the [Unicode Standard], Chapter 4 (Character Properties) discusses case mapping in
 920     /// general and Chapter 3 (Conformance) discusses the default algorithm for case conversion.
 921     ///
 922     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 923     ///
 924     /// # Examples
 925     ///
 926     /// As an iterator:
 927     ///
 928     /// ```
 929     /// for c in 'İ'.to_lowercase() {
 930     ///     print!("{}", c);
 931     /// }
 932     /// println!();
 933     /// ```
 934     ///
 935     /// Using `println!` directly:
 936     ///
 937     /// ```
 938     /// println!("{}", 'İ'.to_lowercase());
 939     /// ```
 940     ///
 941     /// Both are equivalent to:
 942     ///
 943     /// ```
 944     /// println!("i\u{307}");
 945     /// ```
 946     ///
 947     /// Using `to_string`:
 948     ///
 949     /// ```
 950     /// assert_eq!('C'.to_lowercase().to_string(), "c");
 951     ///
 952     /// // Sometimes the result is more than one character:
 953     /// assert_eq!('İ'.to_lowercase().to_string(), "i\u{307}");
 954     ///
 955     /// // Characters that do not have both uppercase and lowercase
 956     /// // convert into themselves.
 957     /// assert_eq!('山'.to_lowercase().to_string(), "山");
 958     /// ```
 959     #[stable(feature = "rust1", since = "1.0.0")]
 960     #[inline]
 961     pub fn to_lowercase(self) -> ToLowercase {
 962         ToLowercase(CaseMappingIter::new(conversions::to_lower(self)))
 963     }
 964
 965     /// Returns an iterator that yields the uppercase mapping of this `char` as one or more
 966     /// `char`s.
 967     ///
 968     /// If this `char` does not have a uppercase mapping, the iterator yields the same `char`.
 969     ///
 970     /// If this `char` has a one-to-one uppercase mapping given by the [Unicode Character
 971     /// Database][ucd] [`UnicodeData.txt`], the iterator yields that `char`.
 972     ///
 973     /// [ucd]: https://www.unicode.org/reports/tr44/
 974     /// [`UnicodeData.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
 975     ///
 976     /// If this `char` requires special considerations (e.g. multiple `char`s) the iterator yields
 977     /// the `char`(s) given by [`SpecialCasing.txt`].
 978     ///
 979     /// [`SpecialCasing.txt`]: https://www.unicode.org/Public/UCD/latest/ucd/SpecialCasing.txt
 980     ///
 981     /// This operation performs an unconditional mapping without tailoring. That is, the conversion
 982     /// is independent of context and language.
 983     ///
 984     /// In the [Unicode Standard], Chapter 4 (Character Properties) discusses case mapping in
 985     /// general and Chapter 3 (Conformance) discusses the default algorithm for case conversion.
 986     ///
 987     /// [Unicode Standard]: https://www.unicode.org/versions/latest/
 988     ///
 989     /// # Examples
 990     ///
 991     /// As an iterator:
 992     ///
 993     /// ```
 994     /// for c in 'ß'.to_uppercase() {
 995     ///     print!("{}", c);
 996     /// }
 997     /// println!();
 998     /// ```
 999     ///
1000     /// Using `println!` directly:
1001     ///
1002     /// ```
1003     /// println!("{}", 'ß'.to_uppercase());
1004     /// ```
1005     ///
1006     /// Both are equivalent to:
1007     ///
1008     /// ```
1009     /// println!("SS");
1010     /// ```
1011     ///
1012     /// Using `to_string`:
1013     ///
1014     /// ```
1015     /// assert_eq!('c'.to_uppercase().to_string(), "C");
1016     ///
1017     /// // Sometimes the result is more than one character:
1018     /// assert_eq!('ß'.to_uppercase().to_string(), "SS");
1019     ///
1020     /// // Characters that do not have both uppercase and lowercase
1021     /// // convert into themselves.
1022     /// assert_eq!('山'.to_uppercase().to_string(), "山");
1023     /// ```
1024     ///
1025     /// # Note on locale
1026     ///
1027     /// In Turkish, the equivalent of 'i' in Latin has five forms instead of two:
1028     ///
1029     /// * 'Dotless': I / ı, sometimes written ï
1030     /// * 'Dotted': İ / i
1031     ///
1032     /// Note that the lowercase dotted 'i' is the same as the Latin. Therefore:
1033     ///
1034     /// ```
1035     /// let upper_i = 'i'.to_uppercase().to_string();
1036     /// ```
1037     ///
1038     /// The value of `upper_i` here relies on the language of the text: if we're
1039     /// in `en-US`, it should be `"I"`, but if we're in `tr_TR`, it should
1040     /// be `"İ"`. `to_uppercase()` does not take this into account, and so:
1041     ///
1042     /// ```
1043     /// let upper_i = 'i'.to_uppercase().to_string();
1044     ///
1045     /// assert_eq!(upper_i, "I");
1046     /// ```
1047     ///
1048     /// holds across languages.
1049     #[stable(feature = "rust1", since = "1.0.0")]
1050     #[inline]
1051     pub fn to_uppercase(self) -> ToUppercase {
1052         ToUppercase(CaseMappingIter::new(conversions::to_upper(self)))
1053     }
1054
1055     /// Checks if the value is within the ASCII range.
1056     ///
1057     /// # Examples
1058     ///
1059     /// ```
1060     /// let ascii = 'a';
1061     /// let non_ascii = '❤';
1062     ///
1063     /// assert!(ascii.is_ascii());
1064     /// assert!(!non_ascii.is_ascii());
1065     /// ```
1066     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
1067     #[rustc_const_stable(feature = "const_ascii_methods_on_intrinsics", since = "1.32.0")]
1068     #[inline]
1069     pub const fn is_ascii(&self) -> bool {
1070         *self as u32 <= 0x7F
1071     }
1072
1073     /// Makes a copy of the value in its ASCII upper case equivalent.
1074     ///
1075     /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z',
1076     /// but non-ASCII letters are unchanged.
1077     ///
1078     /// To uppercase the value in-place, use [`make_ascii_uppercase`].
1079     ///
1080     /// To uppercase ASCII characters in addition to non-ASCII characters, use
1081     /// [`to_uppercase`].
1082     ///
1083     /// # Examples
1084     ///
1085     /// ```
1086     /// let ascii = 'a';
1087     /// let non_ascii = '❤';
1088     ///
1089     /// assert_eq!('A', ascii.to_ascii_uppercase());
1090     /// assert_eq!('❤', non_ascii.to_ascii_uppercase());
1091     /// ```
1092     ///
1093     /// [`make_ascii_uppercase`]: #method.make_ascii_uppercase
1094     /// [`to_uppercase`]: #method.to_uppercase
1095     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
1096     #[inline]
1097     pub fn to_ascii_uppercase(&self) -> char {
1098         if self.is_ascii() { (*self as u8).to_ascii_uppercase() as char } else { *self }
1099     }
1100
1101     /// Makes a copy of the value in its ASCII lower case equivalent.
1102     ///
1103     /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z',
1104     /// but non-ASCII letters are unchanged.
1105     ///
1106     /// To lowercase the value in-place, use [`make_ascii_lowercase`].
1107     ///
1108     /// To lowercase ASCII characters in addition to non-ASCII characters, use
1109     /// [`to_lowercase`].
1110     ///
1111     /// # Examples
1112     ///
1113     /// ```
1114     /// let ascii = 'A';
1115     /// let non_ascii = '❤';
1116     ///
1117     /// assert_eq!('a', ascii.to_ascii_lowercase());
1118     /// assert_eq!('❤', non_ascii.to_ascii_lowercase());
1119     /// ```
1120     ///
1121     /// [`make_ascii_lowercase`]: #method.make_ascii_lowercase
1122     /// [`to_lowercase`]: #method.to_lowercase
1123     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
1124     #[inline]
1125     pub fn to_ascii_lowercase(&self) -> char {
1126         if self.is_ascii() { (*self as u8).to_ascii_lowercase() as char } else { *self }
1127     }
1128
1129     /// Checks that two values are an ASCII case-insensitive match.
1130     ///
1131     /// Equivalent to `to_ascii_lowercase(a) == to_ascii_lowercase(b)`.
1132     ///
1133     /// # Examples
1134     ///
1135     /// ```
1136     /// let upper_a = 'A';
1137     /// let lower_a = 'a';
1138     /// let lower_z = 'z';
1139     ///
1140     /// assert!(upper_a.eq_ignore_ascii_case(&lower_a));
1141     /// assert!(upper_a.eq_ignore_ascii_case(&upper_a));
1142     /// assert!(!upper_a.eq_ignore_ascii_case(&lower_z));
1143     /// ```
1144     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
1145     #[inline]
1146     pub fn eq_ignore_ascii_case(&self, other: &char) -> bool {
1147         self.to_ascii_lowercase() == other.to_ascii_lowercase()
1148     }
1149
1150     /// Converts this type to its ASCII upper case equivalent in-place.
1151     ///
1152     /// ASCII letters 'a' to 'z' are mapped to 'A' to 'Z',
1153     /// but non-ASCII letters are unchanged.
1154     ///
1155     /// To return a new uppercased value without modifying the existing one, use
1156     /// [`to_ascii_uppercase`].
1157     ///
1158     /// # Examples
1159     ///
1160     /// ```
1161     /// let mut ascii = 'a';
1162     ///
1163     /// ascii.make_ascii_uppercase();
1164     ///
1165     /// assert_eq!('A', ascii);
1166     /// ```
1167     ///
1168     /// [`to_ascii_uppercase`]: #method.to_ascii_uppercase
1169     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
1170     #[inline]
1171     pub fn make_ascii_uppercase(&mut self) {
1172         *self = self.to_ascii_uppercase();
1173     }
1174
1175     /// Converts this type to its ASCII lower case equivalent in-place.
1176     ///
1177     /// ASCII letters 'A' to 'Z' are mapped to 'a' to 'z',
1178     /// but non-ASCII letters are unchanged.
1179     ///
1180     /// To return a new lowercased value without modifying the existing one, use
1181     /// [`to_ascii_lowercase`].
1182     ///
1183     /// # Examples
1184     ///
1185     /// ```
1186     /// let mut ascii = 'A';
1187     ///
1188     /// ascii.make_ascii_lowercase();
1189     ///
1190     /// assert_eq!('a', ascii);
1191     /// ```
1192     ///
1193     /// [`to_ascii_lowercase`]: #method.to_ascii_lowercase
1194     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
1195     #[inline]
1196     pub fn make_ascii_lowercase(&mut self) {
1197         *self = self.to_ascii_lowercase();
1198     }
1199
1200     /// Checks if the value is an ASCII alphabetic character:
1201     ///
1202     /// - U+0041 'A' ..= U+005A 'Z', or
1203     /// - U+0061 'a' ..= U+007A 'z'.
1204     ///
1205     /// # Examples
1206     ///
1207     /// ```
1208     /// let uppercase_a = 'A';
1209     /// let uppercase_g = 'G';
1210     /// let a = 'a';
1211     /// let g = 'g';
1212     /// let zero = '0';
1213     /// let percent = '%';
1214     /// let space = ' ';
1215     /// let lf = '\n';
1216     /// let esc: char = 0x1b_u8.into();
1217     ///
1218     /// assert!(uppercase_a.is_ascii_alphabetic());
1219     /// assert!(uppercase_g.is_ascii_alphabetic());
1220     /// assert!(a.is_ascii_alphabetic());
1221     /// assert!(g.is_ascii_alphabetic());
1222     /// assert!(!zero.is_ascii_alphabetic());
1223     /// assert!(!percent.is_ascii_alphabetic());
1224     /// assert!(!space.is_ascii_alphabetic());
1225     /// assert!(!lf.is_ascii_alphabetic());
1226     /// assert!(!esc.is_ascii_alphabetic());
1227     /// ```
1228     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1229     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1230     #[inline]
1231     pub const fn is_ascii_alphabetic(&self) -> bool {
1232         match *self {
1233             'A'..='Z' | 'a'..='z' => true,
1234             _ => false,
1235         }
1236     }
1237
1238     /// Checks if the value is an ASCII uppercase character:
1239     /// U+0041 'A' ..= U+005A 'Z'.
1240     ///
1241     /// # Examples
1242     ///
1243     /// ```
1244     /// let uppercase_a = 'A';
1245     /// let uppercase_g = 'G';
1246     /// let a = 'a';
1247     /// let g = 'g';
1248     /// let zero = '0';
1249     /// let percent = '%';
1250     /// let space = ' ';
1251     /// let lf = '\n';
1252     /// let esc: char = 0x1b_u8.into();
1253     ///
1254     /// assert!(uppercase_a.is_ascii_uppercase());
1255     /// assert!(uppercase_g.is_ascii_uppercase());
1256     /// assert!(!a.is_ascii_uppercase());
1257     /// assert!(!g.is_ascii_uppercase());
1258     /// assert!(!zero.is_ascii_uppercase());
1259     /// assert!(!percent.is_ascii_uppercase());
1260     /// assert!(!space.is_ascii_uppercase());
1261     /// assert!(!lf.is_ascii_uppercase());
1262     /// assert!(!esc.is_ascii_uppercase());
1263     /// ```
1264     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1265     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1266     #[inline]
1267     pub const fn is_ascii_uppercase(&self) -> bool {
1268         match *self {
1269             'A'..='Z' => true,
1270             _ => false,
1271         }
1272     }
1273
1274     /// Checks if the value is an ASCII lowercase character:
1275     /// U+0061 'a' ..= U+007A 'z'.
1276     ///
1277     /// # Examples
1278     ///
1279     /// ```
1280     /// let uppercase_a = 'A';
1281     /// let uppercase_g = 'G';
1282     /// let a = 'a';
1283     /// let g = 'g';
1284     /// let zero = '0';
1285     /// let percent = '%';
1286     /// let space = ' ';
1287     /// let lf = '\n';
1288     /// let esc: char = 0x1b_u8.into();
1289     ///
1290     /// assert!(!uppercase_a.is_ascii_lowercase());
1291     /// assert!(!uppercase_g.is_ascii_lowercase());
1292     /// assert!(a.is_ascii_lowercase());
1293     /// assert!(g.is_ascii_lowercase());
1294     /// assert!(!zero.is_ascii_lowercase());
1295     /// assert!(!percent.is_ascii_lowercase());
1296     /// assert!(!space.is_ascii_lowercase());
1297     /// assert!(!lf.is_ascii_lowercase());
1298     /// assert!(!esc.is_ascii_lowercase());
1299     /// ```
1300     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1301     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1302     #[inline]
1303     pub const fn is_ascii_lowercase(&self) -> bool {
1304         match *self {
1305             'a'..='z' => true,
1306             _ => false,
1307         }
1308     }
1309
1310     /// Checks if the value is an ASCII alphanumeric character:
1311     ///
1312     /// - U+0041 'A' ..= U+005A 'Z', or
1313     /// - U+0061 'a' ..= U+007A 'z', or
1314     /// - U+0030 '0' ..= U+0039 '9'.
1315     ///
1316     /// # Examples
1317     ///
1318     /// ```
1319     /// let uppercase_a = 'A';
1320     /// let uppercase_g = 'G';
1321     /// let a = 'a';
1322     /// let g = 'g';
1323     /// let zero = '0';
1324     /// let percent = '%';
1325     /// let space = ' ';
1326     /// let lf = '\n';
1327     /// let esc: char = 0x1b_u8.into();
1328     ///
1329     /// assert!(uppercase_a.is_ascii_alphanumeric());
1330     /// assert!(uppercase_g.is_ascii_alphanumeric());
1331     /// assert!(a.is_ascii_alphanumeric());
1332     /// assert!(g.is_ascii_alphanumeric());
1333     /// assert!(zero.is_ascii_alphanumeric());
1334     /// assert!(!percent.is_ascii_alphanumeric());
1335     /// assert!(!space.is_ascii_alphanumeric());
1336     /// assert!(!lf.is_ascii_alphanumeric());
1337     /// assert!(!esc.is_ascii_alphanumeric());
1338     /// ```
1339     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1340     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1341     #[inline]
1342     pub const fn is_ascii_alphanumeric(&self) -> bool {
1343         match *self {
1344             '0'..='9' | 'A'..='Z' | 'a'..='z' => true,
1345             _ => false,
1346         }
1347     }
1348
1349     /// Checks if the value is an ASCII decimal digit:
1350     /// U+0030 '0' ..= U+0039 '9'.
1351     ///
1352     /// # Examples
1353     ///
1354     /// ```
1355     /// let uppercase_a = 'A';
1356     /// let uppercase_g = 'G';
1357     /// let a = 'a';
1358     /// let g = 'g';
1359     /// let zero = '0';
1360     /// let percent = '%';
1361     /// let space = ' ';
1362     /// let lf = '\n';
1363     /// let esc: char = 0x1b_u8.into();
1364     ///
1365     /// assert!(!uppercase_a.is_ascii_digit());
1366     /// assert!(!uppercase_g.is_ascii_digit());
1367     /// assert!(!a.is_ascii_digit());
1368     /// assert!(!g.is_ascii_digit());
1369     /// assert!(zero.is_ascii_digit());
1370     /// assert!(!percent.is_ascii_digit());
1371     /// assert!(!space.is_ascii_digit());
1372     /// assert!(!lf.is_ascii_digit());
1373     /// assert!(!esc.is_ascii_digit());
1374     /// ```
1375     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1376     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1377     #[inline]
1378     pub const fn is_ascii_digit(&self) -> bool {
1379         match *self {
1380             '0'..='9' => true,
1381             _ => false,
1382         }
1383     }
1384
1385     /// Checks if the value is an ASCII hexadecimal digit:
1386     ///
1387     /// - U+0030 '0' ..= U+0039 '9', or
1388     /// - U+0041 'A' ..= U+0046 'F', or
1389     /// - U+0061 'a' ..= U+0066 'f'.
1390     ///
1391     /// # Examples
1392     ///
1393     /// ```
1394     /// let uppercase_a = 'A';
1395     /// let uppercase_g = 'G';
1396     /// let a = 'a';
1397     /// let g = 'g';
1398     /// let zero = '0';
1399     /// let percent = '%';
1400     /// let space = ' ';
1401     /// let lf = '\n';
1402     /// let esc: char = 0x1b_u8.into();
1403     ///
1404     /// assert!(uppercase_a.is_ascii_hexdigit());
1405     /// assert!(!uppercase_g.is_ascii_hexdigit());
1406     /// assert!(a.is_ascii_hexdigit());
1407     /// assert!(!g.is_ascii_hexdigit());
1408     /// assert!(zero.is_ascii_hexdigit());
1409     /// assert!(!percent.is_ascii_hexdigit());
1410     /// assert!(!space.is_ascii_hexdigit());
1411     /// assert!(!lf.is_ascii_hexdigit());
1412     /// assert!(!esc.is_ascii_hexdigit());
1413     /// ```
1414     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1415     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1416     #[inline]
1417     pub const fn is_ascii_hexdigit(&self) -> bool {
1418         match *self {
1419             '0'..='9' | 'A'..='F' | 'a'..='f' => true,
1420             _ => false,
1421         }
1422     }
1423
1424     /// Checks if the value is an ASCII punctuation character:
1425     ///
1426     /// - U+0021 ..= U+002F `! " # $ % & ' ( ) * + , - . /`, or
1427     /// - U+003A ..= U+0040 `: ; < = > ? @`, or
1428     /// - U+005B ..= U+0060 ``[ \ ] ^ _ ` ``, or
1429     /// - U+007B ..= U+007E `{ | } ~`
1430     ///
1431     /// # Examples
1432     ///
1433     /// ```
1434     /// let uppercase_a = 'A';
1435     /// let uppercase_g = 'G';
1436     /// let a = 'a';
1437     /// let g = 'g';
1438     /// let zero = '0';
1439     /// let percent = '%';
1440     /// let space = ' ';
1441     /// let lf = '\n';
1442     /// let esc: char = 0x1b_u8.into();
1443     ///
1444     /// assert!(!uppercase_a.is_ascii_punctuation());
1445     /// assert!(!uppercase_g.is_ascii_punctuation());
1446     /// assert!(!a.is_ascii_punctuation());
1447     /// assert!(!g.is_ascii_punctuation());
1448     /// assert!(!zero.is_ascii_punctuation());
1449     /// assert!(percent.is_ascii_punctuation());
1450     /// assert!(!space.is_ascii_punctuation());
1451     /// assert!(!lf.is_ascii_punctuation());
1452     /// assert!(!esc.is_ascii_punctuation());
1453     /// ```
1454     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1455     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1456     #[inline]
1457     pub const fn is_ascii_punctuation(&self) -> bool {
1458         match *self {
1459             '!'..='/' | ':'..='@' | '['..='`' | '{'..='~' => true,
1460             _ => false,
1461         }
1462     }
1463
1464     /// Checks if the value is an ASCII graphic character:
1465     /// U+0021 '!' ..= U+007E '~'.
1466     ///
1467     /// # Examples
1468     ///
1469     /// ```
1470     /// let uppercase_a = 'A';
1471     /// let uppercase_g = 'G';
1472     /// let a = 'a';
1473     /// let g = 'g';
1474     /// let zero = '0';
1475     /// let percent = '%';
1476     /// let space = ' ';
1477     /// let lf = '\n';
1478     /// let esc: char = 0x1b_u8.into();
1479     ///
1480     /// assert!(uppercase_a.is_ascii_graphic());
1481     /// assert!(uppercase_g.is_ascii_graphic());
1482     /// assert!(a.is_ascii_graphic());
1483     /// assert!(g.is_ascii_graphic());
1484     /// assert!(zero.is_ascii_graphic());
1485     /// assert!(percent.is_ascii_graphic());
1486     /// assert!(!space.is_ascii_graphic());
1487     /// assert!(!lf.is_ascii_graphic());
1488     /// assert!(!esc.is_ascii_graphic());
1489     /// ```
1490     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1491     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1492     #[inline]
1493     pub const fn is_ascii_graphic(&self) -> bool {
1494         match *self {
1495             '!'..='~' => true,
1496             _ => false,
1497         }
1498     }
1499
1500     /// Checks if the value is an ASCII whitespace character:
1501     /// U+0020 SPACE, U+0009 HORIZONTAL TAB, U+000A LINE FEED,
1502     /// U+000C FORM FEED, or U+000D CARRIAGE RETURN.
1503     ///
1504     /// Rust uses the WhatWG Infra Standard's [definition of ASCII
1505     /// whitespace][infra-aw]. There are several other definitions in
1506     /// wide use. For instance, [the POSIX locale][pct] includes
1507     /// U+000B VERTICAL TAB as well as all the above characters,
1508     /// but—from the very same specification—[the default rule for
1509     /// "field splitting" in the Bourne shell][bfs] considers *only*
1510     /// SPACE, HORIZONTAL TAB, and LINE FEED as whitespace.
1511     ///
1512     /// If you are writing a program that will process an existing
1513     /// file format, check what that format's definition of whitespace is
1514     /// before using this function.
1515     ///
1516     /// [infra-aw]: https://infra.spec.whatwg.org/#ascii-whitespace
1517     /// [pct]: http://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap07.html#tag_07_03_01
1518     /// [bfs]: http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_06_05
1519     ///
1520     /// # Examples
1521     ///
1522     /// ```
1523     /// let uppercase_a = 'A';
1524     /// let uppercase_g = 'G';
1525     /// let a = 'a';
1526     /// let g = 'g';
1527     /// let zero = '0';
1528     /// let percent = '%';
1529     /// let space = ' ';
1530     /// let lf = '\n';
1531     /// let esc: char = 0x1b_u8.into();
1532     ///
1533     /// assert!(!uppercase_a.is_ascii_whitespace());
1534     /// assert!(!uppercase_g.is_ascii_whitespace());
1535     /// assert!(!a.is_ascii_whitespace());
1536     /// assert!(!g.is_ascii_whitespace());
1537     /// assert!(!zero.is_ascii_whitespace());
1538     /// assert!(!percent.is_ascii_whitespace());
1539     /// assert!(space.is_ascii_whitespace());
1540     /// assert!(lf.is_ascii_whitespace());
1541     /// assert!(!esc.is_ascii_whitespace());
1542     /// ```
1543     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1544     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1545     #[inline]
1546     pub const fn is_ascii_whitespace(&self) -> bool {
1547         match *self {
1548             '\t' | '\n' | '\x0C' | '\r' | ' ' => true,
1549             _ => false,
1550         }
1551     }
1552
1553     /// Checks if the value is an ASCII control character:
1554     /// U+0000 NUL ..= U+001F UNIT SEPARATOR, or U+007F DELETE.
1555     /// Note that most ASCII whitespace characters are control
1556     /// characters, but SPACE is not.
1557     ///
1558     /// # Examples
1559     ///
1560     /// ```
1561     /// let uppercase_a = 'A';
1562     /// let uppercase_g = 'G';
1563     /// let a = 'a';
1564     /// let g = 'g';
1565     /// let zero = '0';
1566     /// let percent = '%';
1567     /// let space = ' ';
1568     /// let lf = '\n';
1569     /// let esc: char = 0x1b_u8.into();
1570     ///
1571     /// assert!(!uppercase_a.is_ascii_control());
1572     /// assert!(!uppercase_g.is_ascii_control());
1573     /// assert!(!a.is_ascii_control());
1574     /// assert!(!g.is_ascii_control());
1575     /// assert!(!zero.is_ascii_control());
1576     /// assert!(!percent.is_ascii_control());
1577     /// assert!(!space.is_ascii_control());
1578     /// assert!(lf.is_ascii_control());
1579     /// assert!(esc.is_ascii_control());
1580     /// ```
1581     #[stable(feature = "ascii_ctype_on_intrinsics", since = "1.24.0")]
1582     #[rustc_const_unstable(feature = "const_ascii_ctype_on_intrinsics", issue = "68983")]
1583     #[inline]
1584     pub const fn is_ascii_control(&self) -> bool {
1585         match *self {
1586             '\0'..='\x1F' | '\x7F' => true,
1587             _ => false,
1588         }
1589     }
1590 }
1591
1592 #[inline]
1593 fn len_utf8(code: u32) -> usize {
1594     if code < MAX_ONE_B {
1595         1
1596     } else if code < MAX_TWO_B {
1597         2
1598     } else if code < MAX_THREE_B {
1599         3
1600     } else {
1601         4
1602     }
1603 }
1604
1605 /// Encodes a raw u32 value as UTF-8 into the provided byte buffer,
1606 /// and then returns the subslice of the buffer that contains the encoded character.
1607 ///
1608 /// Unlike `char::encode_utf8`, this method also handles codepoints in the surrogate range.
1609 /// (Creating a `char` in the surrogate range is UB.)
1610 /// The result is valid [generalized UTF-8] but not valid UTF-8.
1611 ///
1612 /// [generalized UTF-8]: https://simonsapin.github.io/wtf-8/#generalized-utf8
1613 ///
1614 /// # Panics
1615 ///
1616 /// Panics if the buffer is not large enough.
1617 /// A buffer of length four is large enough to encode any `char`.
1618 #[unstable(feature = "char_internals", reason = "exposed only for libstd", issue = "none")]
1619 #[doc(hidden)]
1620 #[inline]
1621 pub fn encode_utf8_raw(code: u32, dst: &mut [u8]) -> &mut [u8] {
1622     let len = len_utf8(code);
1623     match (len, &mut dst[..]) {
1624         (1, [a, ..]) => {
1625             *a = code as u8;
1626         }
1627         (2, [a, b, ..]) => {
1628             *a = (code >> 6 & 0x1F) as u8 | TAG_TWO_B;
1629             *b = (code & 0x3F) as u8 | TAG_CONT;
1630         }
1631         (3, [a, b, c, ..]) => {
1632             *a = (code >> 12 & 0x0F) as u8 | TAG_THREE_B;
1633             *b = (code >> 6 & 0x3F) as u8 | TAG_CONT;
1634             *c = (code & 0x3F) as u8 | TAG_CONT;
1635         }
1636         (4, [a, b, c, d, ..]) => {
1637             *a = (code >> 18 & 0x07) as u8 | TAG_FOUR_B;
1638             *b = (code >> 12 & 0x3F) as u8 | TAG_CONT;
1639             *c = (code >> 6 & 0x3F) as u8 | TAG_CONT;
1640             *d = (code & 0x3F) as u8 | TAG_CONT;
1641         }
1642         _ => panic!(
1643             "encode_utf8: need {} bytes to encode U+{:X}, but the buffer has {}",
1644             len,
1645             code,
1646             dst.len(),
1647         ),
1648     };
1649     &mut dst[..len]
1650 }
1651
1652 /// Encodes a raw u32 value as UTF-16 into the provided `u16` buffer,
1653 /// and then returns the subslice of the buffer that contains the encoded character.
1654 ///
1655 /// Unlike `char::encode_utf16`, this method also handles codepoints in the surrogate range.
1656 /// (Creating a `char` in the surrogate range is UB.)
1657 ///
1658 /// # Panics
1659 ///
1660 /// Panics if the buffer is not large enough.
1661 /// A buffer of length 2 is large enough to encode any `char`.
1662 #[unstable(feature = "char_internals", reason = "exposed only for libstd", issue = "none")]
1663 #[doc(hidden)]
1664 #[inline]
1665 pub fn encode_utf16_raw(mut code: u32, dst: &mut [u16]) -> &mut [u16] {
1666     // SAFETY: each arm checks whether there are enough bits to write into
1667     unsafe {
1668         if (code & 0xFFFF) == code && !dst.is_empty() {
1669             // The BMP falls through
1670             *dst.get_unchecked_mut(0) = code as u16;
1671             slice::from_raw_parts_mut(dst.as_mut_ptr(), 1)
1672         } else if dst.len() >= 2 {
1673             // Supplementary planes break into surrogates.
1674             code -= 0x1_0000;
1675             *dst.get_unchecked_mut(0) = 0xD800 | ((code >> 10) as u16);
1676             *dst.get_unchecked_mut(1) = 0xDC00 | ((code as u16) & 0x3FF);
1677             slice::from_raw_parts_mut(dst.as_mut_ptr(), 2)
1678         } else {
1679             panic!(
1680                 "encode_utf16: need {} units to encode U+{:X}, but the buffer has {}",
1681                 from_u32_unchecked(code).len_utf16(),
1682                 code,
1683                 dst.len(),
1684             )
1685         }
1686     }
1687 }