Move the core::char module to its own directory

3 files changed, 1713 insertions, 0 deletions

diff --git a/src/libcore/char/mod.rs b/src/libcore/char/mod.rs
new file mode 100644
index 00000000000..c6b620e5238
--- /dev/null
+++ b/src/libcore/char/mod.rs
@@ -0,0 +1,928 @@
+// Copyright 2012-2014 The Rust Project Developers. See the COPYRIGHT
+// file at the top-level directory of this distribution and at
+// http://rust-lang.org/COPYRIGHT.
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+//! Character manipulation.
+//!
+//! For more details, see ::core::unicode::char (a.k.a. std::char)
+
+#![allow(non_snake_case)]
+#![stable(feature = "core_char", since = "1.2.0")]
+
+mod printable;
+
+use self::printable::is_printable;
+use convert::TryFrom;
+use fmt::{self, Write};
+use slice;
+use str::{from_utf8_unchecked_mut, FromStr};
+use iter::FusedIterator;
+use mem::transmute;
+
+// UTF-8 ranges and tags for encoding characters
+const TAG_CONT: u8    = 0b1000_0000;
+const TAG_TWO_B: u8   = 0b1100_0000;
+const TAG_THREE_B: u8 = 0b1110_0000;
+const TAG_FOUR_B: u8  = 0b1111_0000;
+const MAX_ONE_B: u32   =     0x80;
+const MAX_TWO_B: u32   =    0x800;
+const MAX_THREE_B: u32 =  0x10000;
+
+/*
+    Lu  Uppercase_Letter        an uppercase letter
+    Ll  Lowercase_Letter        a lowercase letter
+    Lt  Titlecase_Letter        a digraphic character, with first part uppercase
+    Lm  Modifier_Letter         a modifier letter
+    Lo  Other_Letter            other letters, including syllables and ideographs
+    Mn  Nonspacing_Mark         a nonspacing combining mark (zero advance width)
+    Mc  Spacing_Mark            a spacing combining mark (positive advance width)
+    Me  Enclosing_Mark          an enclosing combining mark
+    Nd  Decimal_Number          a decimal digit
+    Nl  Letter_Number           a letterlike numeric character
+    No  Other_Number            a numeric character of other type
+    Pc  Connector_Punctuation   a connecting punctuation mark, like a tie
+    Pd  Dash_Punctuation        a dash or hyphen punctuation mark
+    Ps  Open_Punctuation        an opening punctuation mark (of a pair)
+    Pe  Close_Punctuation       a closing punctuation mark (of a pair)
+    Pi  Initial_Punctuation     an initial quotation mark
+    Pf  Final_Punctuation       a final quotation mark
+    Po  Other_Punctuation       a punctuation mark of other type
+    Sm  Math_Symbol             a symbol of primarily mathematical use
+    Sc  Currency_Symbol         a currency sign
+    Sk  Modifier_Symbol         a non-letterlike modifier symbol
+    So  Other_Symbol            a symbol of other type
+    Zs  Space_Separator         a space character (of various non-zero widths)
+    Zl  Line_Separator          U+2028 LINE SEPARATOR only
+    Zp  Paragraph_Separator     U+2029 PARAGRAPH SEPARATOR only
+    Cc  Control                 a C0 or C1 control code
+    Cf  Format                  a format control character
+    Cs  Surrogate               a surrogate code point
+    Co  Private_Use             a private-use character
+    Cn  Unassigned              a reserved unassigned code point or a noncharacter
+*/
+
+/// The highest valid code point a `char` can have.
+///
+/// A [`char`] is a [Unicode Scalar Value], which means that it is a [Code
+/// Point], but only ones within a certain range. `MAX` is the highest valid
+/// code point that's a valid [Unicode Scalar Value].
+///
+/// [`char`]: ../../std/primitive.char.html
+/// [Unicode Scalar Value]: http://www.unicode.org/glossary/#unicode_scalar_value
+/// [Code Point]: http://www.unicode.org/glossary/#code_point
+#[stable(feature = "rust1", since = "1.0.0")]
+pub const MAX: char = '\u{10ffff}';
+
+/// `U+FFFD REPLACEMENT CHARACTER` (�) is used in Unicode to represent a
+/// decoding error.
+///
+/// It can occur, for example, when giving ill-formed UTF-8 bytes to
+/// [`String::from_utf8_lossy`](../../std/string/struct.String.html#method.from_utf8_lossy).
+#[stable(feature = "decode_utf16", since = "1.9.0")]
+pub const REPLACEMENT_CHARACTER: char = '\u{FFFD}';
+
+/// Converts a `u32` to a `char`.
+///
+/// Note that all [`char`]s are valid [`u32`]s, and can be cast to one with
+/// [`as`]:
+///
+/// ```
+/// let c = '💯';
+/// let i = c as u32;
+///
+/// assert_eq!(128175, i);
+/// ```
+///
+/// However, the reverse is not true: not all valid [`u32`]s are valid
+/// [`char`]s. `from_u32()` will return `None` if the input is not a valid value
+/// for a [`char`].
+///
+/// [`char`]: ../../std/primitive.char.html
+/// [`u32`]: ../../std/primitive.u32.html
+/// [`as`]: ../../book/first-edition/casting-between-types.html#as
+///
+/// For an unsafe version of this function which ignores these checks, see
+/// [`from_u32_unchecked`].
+///
+/// [`from_u32_unchecked`]: fn.from_u32_unchecked.html
+///
+/// # Examples
+///
+/// Basic usage:
+///
+/// ```
+/// use std::char;
+///
+/// let c = char::from_u32(0x2764);
+///
+/// assert_eq!(Some('❤'), c);
+/// ```
+///
+/// Returning `None` when the input is not a valid [`char`]:
+///
+/// ```
+/// use std::char;
+///
+/// let c = char::from_u32(0x110000);
+///
+/// assert_eq!(None, c);
+/// ```
+#[inline]
+#[stable(feature = "rust1", since = "1.0.0")]
+pub fn from_u32(i: u32) -> Option<char> {
+    char::try_from(i).ok()
+}
+
+/// Converts a `u32` to a `char`, ignoring validity.
+///
+/// Note that all [`char`]s are valid [`u32`]s, and can be cast to one with
+/// [`as`]:
+///
+/// ```
+/// let c = '💯';
+/// let i = c as u32;
+///
+/// assert_eq!(128175, i);
+/// ```
+///
+/// However, the reverse is not true: not all valid [`u32`]s are valid
+/// [`char`]s. `from_u32_unchecked()` will ignore this, and blindly cast to
+/// [`char`], possibly creating an invalid one.
+///
+/// [`char`]: ../../std/primitive.char.html
+/// [`u32`]: ../../std/primitive.u32.html
+/// [`as`]: ../../book/first-edition/casting-between-types.html#as
+///
+/// # Safety
+///
+/// This function is unsafe, as it may construct invalid `char` values.
+///
+/// For a safe version of this function, see the [`from_u32`] function.
+///
+/// [`from_u32`]: fn.from_u32.html
+///
+/// # Examples
+///
+/// Basic usage:
+///
+/// ```
+/// use std::char;
+///
+/// let c = unsafe { char::from_u32_unchecked(0x2764) };
+///
+/// assert_eq!('❤', c);
+/// ```
+#[inline]
+#[stable(feature = "char_from_unchecked", since = "1.5.0")]
+pub unsafe fn from_u32_unchecked(i: u32) -> char {
+    transmute(i)
+}
+
+#[stable(feature = "char_convert", since = "1.13.0")]
+impl From<char> for u32 {
+    #[inline]
+    fn from(c: char) -> Self {
+        c as u32
+    }
+}
+
+/// Maps a byte in 0x00...0xFF to a `char` whose code point has the same value, in U+0000 to U+00FF.
+///
+/// Unicode is designed such that this effectively decodes bytes
+/// with the character encoding that IANA calls ISO-8859-1.
+/// This encoding is compatible with ASCII.
+///
+/// Note that this is different from ISO/IEC 8859-1 a.k.a. ISO 8859-1 (with one less hyphen),
+/// which leaves some "blanks", byte values that are not assigned to any character.
+/// ISO-8859-1 (the IANA one) assigns them to the C0 and C1 control codes.
+///
+/// Note that this is *also* different from Windows-1252 a.k.a. code page 1252,
+/// which is a superset ISO/IEC 8859-1 that assigns some (not all!) blanks
+/// to punctuation and various Latin characters.
+///
+/// To confuse things further, [on the Web](https://encoding.spec.whatwg.org/)
+/// `ascii`, `iso-8859-1`, and `windows-1252` are all aliases
+/// for a superset of Windows-1252 that fills the remaining blanks with corresponding
+/// C0 and C1 control codes.
+#[stable(feature = "char_convert", since = "1.13.0")]
+impl From<u8> for char {
+    #[inline]
+    fn from(i: u8) -> Self {
+        i as char
+    }
+}
+
+
+/// An error which can be returned when parsing a char.
+#[stable(feature = "char_from_str", since = "1.20.0")]
+#[derive(Clone, Debug, PartialEq, Eq)]
+pub struct ParseCharError {
+    kind: CharErrorKind,
+}
+
+impl ParseCharError {
+    #[unstable(feature = "char_error_internals",
+               reason = "this method should not be available publicly",
+               issue = "0")]
+    #[doc(hidden)]
+    pub fn __description(&self) -> &str {
+        match self.kind {
+            CharErrorKind::EmptyString => {
+                "cannot parse char from empty string"
+            },
+            CharErrorKind::TooManyChars => "too many characters in string"
+        }
+    }
+}
+
+#[derive(Copy, Clone, Debug, PartialEq, Eq)]
+enum CharErrorKind {
+    EmptyString,
+    TooManyChars,
+}
+
+#[stable(feature = "char_from_str", since = "1.20.0")]
+impl fmt::Display for ParseCharError {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        self.__description().fmt(f)
+    }
+}
+
+
+#[stable(feature = "char_from_str", since = "1.20.0")]
+impl FromStr for char {
+    type Err = ParseCharError;
+
+    #[inline]
+    fn from_str(s: &str) -> Result<Self, Self::Err> {
+        let mut chars = s.chars();
+        match (chars.next(), chars.next()) {
+            (None, _) => {
+                Err(ParseCharError { kind: CharErrorKind::EmptyString })
+            },
+            (Some(c), None) => Ok(c),
+            _ => {
+                Err(ParseCharError { kind: CharErrorKind::TooManyChars })
+            }
+        }
+    }
+}
+
+
+#[stable(feature = "try_from", since = "1.26.0")]
+impl TryFrom<u32> for char {
+    type Error = CharTryFromError;
+
+    #[inline]
+    fn try_from(i: u32) -> Result<Self, Self::Error> {
+        if (i > MAX as u32) || (i >= 0xD800 && i <= 0xDFFF) {
+            Err(CharTryFromError(()))
+        } else {
+            Ok(unsafe { from_u32_unchecked(i) })
+        }
+    }
+}
+
+/// The error type returned when a conversion from u32 to char fails.
+#[stable(feature = "try_from", since = "1.26.0")]
+#[derive(Copy, Clone, Debug, PartialEq, Eq)]
+pub struct CharTryFromError(());
+
+#[stable(feature = "try_from", since = "1.26.0")]
+impl fmt::Display for CharTryFromError {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        "converted integer out of range for `char`".fmt(f)
+    }
+}
+
+/// Converts a digit in the given radix to a `char`.
+///
+/// A 'radix' here is sometimes also called a 'base'. A radix of two
+/// indicates a binary number, a radix of ten, decimal, and a radix of
+/// sixteen, hexadecimal, to give some common values. Arbitrary
+/// radices are supported.
+///
+/// `from_digit()` will return `None` if the input is not a digit in
+/// the given radix.
+///
+/// # Panics
+///
+/// Panics if given a radix larger than 36.
+///
+/// # Examples
+///
+/// Basic usage:
+///
+/// ```
+/// use std::char;
+///
+/// let c = char::from_digit(4, 10);
+///
+/// assert_eq!(Some('4'), c);
+///
+/// // Decimal 11 is a single digit in base 16
+/// let c = char::from_digit(11, 16);
+///
+/// assert_eq!(Some('b'), c);
+/// ```
+///
+/// Returning `None` when the input is not a digit:
+///
+/// ```
+/// use std::char;
+///
+/// let c = char::from_digit(20, 10);
+///
+/// assert_eq!(None, c);
+/// ```
+///
+/// Passing a large radix, causing a panic:
+///
+/// ```
+/// use std::thread;
+/// use std::char;
+///
+/// let result = thread::spawn(|| {
+///     // this panics
+///     let c = char::from_digit(1, 37);
+/// }).join();
+///
+/// assert!(result.is_err());
+/// ```
+#[inline]
+#[stable(feature = "rust1", since = "1.0.0")]
+pub fn from_digit(num: u32, radix: u32) -> Option<char> {
+    if radix > 36 {
+        panic!("from_digit: radix is too high (maximum 36)");
+    }
+    if num < radix {
+        let num = num as u8;
+        if num < 10 {
+            Some((b'0' + num) as char)
+        } else {
+            Some((b'a' + num - 10) as char)
+        }
+    } else {
+        None
+    }
+}
+
+// NB: the stabilization and documentation for this trait is in
+// unicode/char.rs, not here
+#[allow(missing_docs)] // docs in libunicode/u_char.rs
+#[doc(hidden)]
+#[unstable(feature = "core_char_ext",
+           reason = "the stable interface is `impl char` in later crate",
+           issue = "32110")]
+pub trait CharExt {
+    #[stable(feature = "core", since = "1.6.0")]
+    fn is_digit(self, radix: u32) -> bool;
+    #[stable(feature = "core", since = "1.6.0")]
+    fn to_digit(self, radix: u32) -> Option<u32>;
+    #[stable(feature = "core", since = "1.6.0")]
+    fn escape_unicode(self) -> EscapeUnicode;
+    #[stable(feature = "core", since = "1.6.0")]
+    fn escape_default(self) -> EscapeDefault;
+    #[stable(feature = "char_escape_debug", since = "1.20.0")]
+    fn escape_debug(self) -> EscapeDebug;
+    #[stable(feature = "core", since = "1.6.0")]
+    fn len_utf8(self) -> usize;
+    #[stable(feature = "core", since = "1.6.0")]
+    fn len_utf16(self) -> usize;
+    #[stable(feature = "unicode_encode_char", since = "1.15.0")]
+    fn encode_utf8(self, dst: &mut [u8]) -> &mut str;
+    #[stable(feature = "unicode_encode_char", since = "1.15.0")]
+    fn encode_utf16(self, dst: &mut [u16]) -> &mut [u16];
+}
+
+#[stable(feature = "core", since = "1.6.0")]
+impl CharExt for char {
+    #[inline]
+    fn is_digit(self, radix: u32) -> bool {
+        self.to_digit(radix).is_some()
+    }
+
+    #[inline]
+    fn to_digit(self, radix: u32) -> Option<u32> {
+        if radix > 36 {
+            panic!("to_digit: radix is too high (maximum 36)");
+        }
+        let val = match self {
+          '0' ... '9' => self as u32 - '0' as u32,
+          'a' ... 'z' => self as u32 - 'a' as u32 + 10,
+          'A' ... 'Z' => self as u32 - 'A' as u32 + 10,
+          _ => return None,
+        };
+        if val < radix { Some(val) }
+        else { None }
+    }
+
+    #[inline]
+    fn escape_unicode(self) -> EscapeUnicode {
+        let c = self as u32;
+
+        // or-ing 1 ensures that for c==0 the code computes that one
+        // digit should be printed and (which is the same) avoids the
+        // (31 - 32) underflow
+        let msb = 31 - (c | 1).leading_zeros();
+
+        // the index of the most significant hex digit
+        let ms_hex_digit = msb / 4;
+        EscapeUnicode {
+            c: self,
+            state: EscapeUnicodeState::Backslash,
+            hex_digit_idx: ms_hex_digit as usize,
+        }
+    }
+
+    #[inline]
+    fn escape_default(self) -> EscapeDefault {
+        let init_state = match self {
+            '\t' => EscapeDefaultState::Backslash('t'),
+            '\r' => EscapeDefaultState::Backslash('r'),
+            '\n' => EscapeDefaultState::Backslash('n'),
+            '\\' | '\'' | '"' => EscapeDefaultState::Backslash(self),
+            '\x20' ... '\x7e' => EscapeDefaultState::Char(self),
+            _ => EscapeDefaultState::Unicode(self.escape_unicode())
+        };
+        EscapeDefault { state: init_state }
+    }
+
+    #[inline]
+    fn escape_debug(self) -> EscapeDebug {
+        let init_state = match self {
+            '\t' => EscapeDefaultState::Backslash('t'),
+            '\r' => EscapeDefaultState::Backslash('r'),
+            '\n' => EscapeDefaultState::Backslash('n'),
+            '\\' | '\'' | '"' => EscapeDefaultState::Backslash(self),
+            c if is_printable(c) => EscapeDefaultState::Char(c),
+            c => EscapeDefaultState::Unicode(c.escape_unicode()),
+        };
+        EscapeDebug(EscapeDefault { state: init_state })
+    }
+
+    #[inline]
+    fn len_utf8(self) -> usize {
+        let code = self as u32;
+        if code < MAX_ONE_B {
+            1
+        } else if code < MAX_TWO_B {
+            2
+        } else if code < MAX_THREE_B {
+            3
+        } else {
+            4
+        }
+    }
+
+    #[inline]
+    fn len_utf16(self) -> usize {
+        let ch = self as u32;
+        if (ch & 0xFFFF) == ch { 1 } else { 2 }
+    }
+
+    #[inline]
+    fn encode_utf8(self, dst: &mut [u8]) -> &mut str {
+        let code = self as u32;
+        unsafe {
+            let len =
+            if code < MAX_ONE_B && !dst.is_empty() {
+                *dst.get_unchecked_mut(0) = code as u8;
+                1
+            } else if code < MAX_TWO_B && dst.len() >= 2 {
+                *dst.get_unchecked_mut(0) = (code >> 6 & 0x1F) as u8 | TAG_TWO_B;
+                *dst.get_unchecked_mut(1) = (code & 0x3F) as u8 | TAG_CONT;
+                2
+            } else if code < MAX_THREE_B && dst.len() >= 3  {
+                *dst.get_unchecked_mut(0) = (code >> 12 & 0x0F) as u8 | TAG_THREE_B;
+                *dst.get_unchecked_mut(1) = (code >>  6 & 0x3F) as u8 | TAG_CONT;
+                *dst.get_unchecked_mut(2) = (code & 0x3F) as u8 | TAG_CONT;
+                3
+            } else if dst.len() >= 4 {
+                *dst.get_unchecked_mut(0) = (code >> 18 & 0x07) as u8 | TAG_FOUR_B;
+                *dst.get_unchecked_mut(1) = (code >> 12 & 0x3F) as u8 | TAG_CONT;
+                *dst.get_unchecked_mut(2) = (code >>  6 & 0x3F) as u8 | TAG_CONT;
+                *dst.get_unchecked_mut(3) = (code & 0x3F) as u8 | TAG_CONT;
+                4
+            } else {
+                panic!("encode_utf8: need {} bytes to encode U+{:X}, but the buffer has {}",
+                    from_u32_unchecked(code).len_utf8(),
+                    code,
+                    dst.len())
+            };
+            from_utf8_unchecked_mut(dst.get_unchecked_mut(..len))
+        }
+    }
+
+    #[inline]
+    fn encode_utf16(self, dst: &mut [u16]) -> &mut [u16] {
+        let mut code = self as u32;
+        unsafe {
+            if (code & 0xFFFF) == code && !dst.is_empty() {
+                // The BMP falls through (assuming non-surrogate, as it should)
+                *dst.get_unchecked_mut(0) = code as u16;
+                slice::from_raw_parts_mut(dst.as_mut_ptr(), 1)
+            } else if dst.len() >= 2 {
+                // Supplementary planes break into surrogates.
+                code -= 0x1_0000;
+                *dst.get_unchecked_mut(0) = 0xD800 | ((code >> 10) as u16);
+                *dst.get_unchecked_mut(1) = 0xDC00 | ((code as u16) & 0x3FF);
+                slice::from_raw_parts_mut(dst.as_mut_ptr(), 2)
+            } else {
+                panic!("encode_utf16: need {} units to encode U+{:X}, but the buffer has {}",
+                    from_u32_unchecked(code).len_utf16(),
+                    code,
+                    dst.len())
+            }
+        }
+    }
+}
+
+/// Returns an iterator that yields the hexadecimal Unicode escape of a
+/// character, as `char`s.
+///
+/// This `struct` is created by the [`escape_unicode`] method on [`char`]. See
+/// its documentation for more.
+///
+/// [`escape_unicode`]: ../../std/primitive.char.html#method.escape_unicode
+/// [`char`]: ../../std/primitive.char.html
+#[derive(Clone, Debug)]
+#[stable(feature = "rust1", since = "1.0.0")]
+pub struct EscapeUnicode {
+    c: char,
+    state: EscapeUnicodeState,
+
+    // The index of the next hex digit to be printed (0 if none),
+    // i.e. the number of remaining hex digits to be printed;
+    // increasing from the least significant digit: 0x543210
+    hex_digit_idx: usize,
+}
+
+// The enum values are ordered so that their representation is the
+// same as the remaining length (besides the hexadecimal digits). This
+// likely makes `len()` a single load from memory) and inline-worth.
+#[derive(Clone, Debug)]
+enum EscapeUnicodeState {
+    Done,
+    RightBrace,
+    Value,
+    LeftBrace,
+    Type,
+    Backslash,
+}
+
+#[stable(feature = "rust1", since = "1.0.0")]
+impl Iterator for EscapeUnicode {
+    type Item = char;
+
+    fn next(&mut self) -> Option<char> {
+        match self.state {
+            EscapeUnicodeState::Backslash => {
+                self.state = EscapeUnicodeState::Type;
+                Some('\\')
+            }
+            EscapeUnicodeState::Type => {
+                self.state = EscapeUnicodeState::LeftBrace;
+                Some('u')
+            }
+            EscapeUnicodeState::LeftBrace => {
+                self.state = EscapeUnicodeState::Value;
+                Some('{')
+            }
+            EscapeUnicodeState::Value => {
+                let hex_digit = ((self.c as u32) >> (self.hex_digit_idx * 4)) & 0xf;
+                let c = from_digit(hex_digit, 16).unwrap();
+                if self.hex_digit_idx == 0 {
+                    self.state = EscapeUnicodeState::RightBrace;
+                } else {
+                    self.hex_digit_idx -= 1;
+                }
+                Some(c)
+            }
+            EscapeUnicodeState::RightBrace => {
+                self.state = EscapeUnicodeState::Done;
+                Some('}')
+            }
+            EscapeUnicodeState::Done => None,
+        }
+    }
+
+    #[inline]
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        let n = self.len();
+        (n, Some(n))
+    }
+
+    #[inline]
+    fn count(self) -> usize {
+        self.len()
+    }
+
+    fn last(self) -> Option<char> {
+        match self.state {
+            EscapeUnicodeState::Done => None,
+
+            EscapeUnicodeState::RightBrace |
+            EscapeUnicodeState::Value |
+            EscapeUnicodeState::LeftBrace |
+            EscapeUnicodeState::Type |
+            EscapeUnicodeState::Backslash => Some('}'),
+        }
+    }
+}
+
+#[stable(feature = "exact_size_escape", since = "1.11.0")]
+impl ExactSizeIterator for EscapeUnicode {
+    #[inline]
+    fn len(&self) -> usize {
+        // The match is a single memory access with no branching
+        self.hex_digit_idx + match self.state {
+            EscapeUnicodeState::Done => 0,
+            EscapeUnicodeState::RightBrace => 1,
+            EscapeUnicodeState::Value => 2,
+            EscapeUnicodeState::LeftBrace => 3,
+            EscapeUnicodeState::Type => 4,
+            EscapeUnicodeState::Backslash => 5,
+        }
+    }
+}
+
+#[stable(feature = "fused", since = "1.26.0")]
+impl FusedIterator for EscapeUnicode {}
+
+#[stable(feature = "char_struct_display", since = "1.16.0")]
+impl fmt::Display for EscapeUnicode {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        for c in self.clone() {
+            f.write_char(c)?;
+        }
+        Ok(())
+    }
+}
+
+/// An iterator that yields the literal escape code of a `char`.
+///
+/// This `struct` is created by the [`escape_default`] method on [`char`]. See
+/// its documentation for more.
+///
+/// [`escape_default`]: ../../std/primitive.char.html#method.escape_default
+/// [`char`]: ../../std/primitive.char.html
+#[derive(Clone, Debug)]
+#[stable(feature = "rust1", since = "1.0.0")]
+pub struct EscapeDefault {
+    state: EscapeDefaultState
+}
+
+#[derive(Clone, Debug)]
+enum EscapeDefaultState {
+    Done,
+    Char(char),
+    Backslash(char),
+    Unicode(EscapeUnicode),
+}
+
+#[stable(feature = "rust1", since = "1.0.0")]
+impl Iterator for EscapeDefault {
+    type Item = char;
+
+    fn next(&mut self) -> Option<char> {
+        match self.state {
+            EscapeDefaultState::Backslash(c) => {
+                self.state = EscapeDefaultState::Char(c);
+                Some('\\')
+            }
+            EscapeDefaultState::Char(c) => {
+                self.state = EscapeDefaultState::Done;
+                Some(c)
+            }
+            EscapeDefaultState::Done => None,
+            EscapeDefaultState::Unicode(ref mut iter) => iter.next(),
+        }
+    }
+
+    #[inline]
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        let n = self.len();
+        (n, Some(n))
+    }
+
+    #[inline]
+    fn count(self) -> usize {
+        self.len()
+    }
+
+    fn nth(&mut self, n: usize) -> Option<char> {
+        match self.state {
+            EscapeDefaultState::Backslash(c) if n == 0 => {
+                self.state = EscapeDefaultState::Char(c);
+                Some('\\')
+            },
+            EscapeDefaultState::Backslash(c) if n == 1 => {
+                self.state = EscapeDefaultState::Done;
+                Some(c)
+            },
+            EscapeDefaultState::Backslash(_) => {
+                self.state = EscapeDefaultState::Done;
+                None
+            },
+            EscapeDefaultState::Char(c) => {
+                self.state = EscapeDefaultState::Done;
+
+                if n == 0 {
+                    Some(c)
+                } else {
+                    None
+                }
+            },
+            EscapeDefaultState::Done => return None,
+            EscapeDefaultState::Unicode(ref mut i) => return i.nth(n),
+        }
+    }
+
+    fn last(self) -> Option<char> {
+        match self.state {
+            EscapeDefaultState::Unicode(iter) => iter.last(),
+            EscapeDefaultState::Done => None,
+            EscapeDefaultState::Backslash(c) | EscapeDefaultState::Char(c) => Some(c),
+        }
+    }
+}
+
+#[stable(feature = "exact_size_escape", since = "1.11.0")]
+impl ExactSizeIterator for EscapeDefault {
+    fn len(&self) -> usize {
+        match self.state {
+            EscapeDefaultState::Done => 0,
+            EscapeDefaultState::Char(_) => 1,
+            EscapeDefaultState::Backslash(_) => 2,
+            EscapeDefaultState::Unicode(ref iter) => iter.len(),
+        }
+    }
+}
+
+#[stable(feature = "fused", since = "1.26.0")]
+impl FusedIterator for EscapeDefault {}
+
+#[stable(feature = "char_struct_display", since = "1.16.0")]
+impl fmt::Display for EscapeDefault {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        for c in self.clone() {
+            f.write_char(c)?;
+        }
+        Ok(())
+    }
+}
+
+/// An iterator that yields the literal escape code of a `char`.
+///
+/// This `struct` is created by the [`escape_debug`] method on [`char`]. See its
+/// documentation for more.
+///
+/// [`escape_debug`]: ../../std/primitive.char.html#method.escape_debug
+/// [`char`]: ../../std/primitive.char.html
+#[stable(feature = "char_escape_debug", since = "1.20.0")]
+#[derive(Clone, Debug)]
+pub struct EscapeDebug(EscapeDefault);
+
+#[stable(feature = "char_escape_debug", since = "1.20.0")]
+impl Iterator for EscapeDebug {
+    type Item = char;
+    fn next(&mut self) -> Option<char> { self.0.next() }
+    fn size_hint(&self) -> (usize, Option<usize>) { self.0.size_hint() }
+}
+
+#[stable(feature = "char_escape_debug", since = "1.20.0")]
+impl ExactSizeIterator for EscapeDebug { }
+
+#[stable(feature = "fused", since = "1.26.0")]
+impl FusedIterator for EscapeDebug {}
+
+#[stable(feature = "char_escape_debug", since = "1.20.0")]
+impl fmt::Display for EscapeDebug {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        fmt::Display::fmt(&self.0, f)
+    }
+}
+
+
+
+/// An iterator over an iterator of bytes of the characters the bytes represent
+/// as UTF-8
+#[unstable(feature = "decode_utf8", issue = "33906")]
+#[derive(Clone, Debug)]
+pub struct DecodeUtf8<I: Iterator<Item = u8>>(::iter::Peekable<I>);
+
+/// Decodes an `Iterator` of bytes as UTF-8.
+#[unstable(feature = "decode_utf8", issue = "33906")]
+#[inline]
+pub fn decode_utf8<I: IntoIterator<Item = u8>>(i: I) -> DecodeUtf8<I::IntoIter> {
+    DecodeUtf8(i.into_iter().peekable())
+}
+
+/// `<DecodeUtf8 as Iterator>::next` returns this for an invalid input sequence.
+#[unstable(feature = "decode_utf8", issue = "33906")]
+#[derive(PartialEq, Eq, Debug)]
+pub struct InvalidSequence(());
+
+#[unstable(feature = "decode_utf8", issue = "33906")]
+impl<I: Iterator<Item = u8>> Iterator for DecodeUtf8<I> {
+    type Item = Result<char, InvalidSequence>;
+    #[inline]
+
+    fn next(&mut self) -> Option<Result<char, InvalidSequence>> {
+        self.0.next().map(|first_byte| {
+            // Emit InvalidSequence according to
+            // Unicode §5.22 Best Practice for U+FFFD Substitution
+            // http://www.unicode.org/versions/Unicode9.0.0/ch05.pdf#G40630
+
+            // Roughly: consume at least one byte,
+            // then validate one byte at a time and stop before the first unexpected byte
+            // (which might be the valid start of the next byte sequence).
+
+            let mut code_point;
+            macro_rules! first_byte {
+                ($mask: expr) => {
+                    code_point = u32::from(first_byte & $mask)
+                }
+            }
+            macro_rules! continuation_byte {
+                () => { continuation_byte!(0x80...0xBF) };
+                ($range: pat) => {
+                    match self.0.peek() {
+                        Some(&byte @ $range) => {
+                            code_point = (code_point << 6) | u32::from(byte & 0b0011_1111);
+                            self.0.next();
+                        }
+                        _ => return Err(InvalidSequence(()))
+                    }
+                }
+            }
+
+            match first_byte {
+                0x00...0x7F => {
+                    first_byte!(0b1111_1111);
+                }
+                0xC2...0xDF => {
+                    first_byte!(0b0001_1111);
+                    continuation_byte!();
+                }
+                0xE0 => {
+                    first_byte!(0b0000_1111);
+                    continuation_byte!(0xA0...0xBF);  // 0x80...0x9F here are overlong
+                    continuation_byte!();
+                }
+                0xE1...0xEC | 0xEE...0xEF => {
+                    first_byte!(0b0000_1111);
+                    continuation_byte!();
+                    continuation_byte!();
+                }
+                0xED => {
+                    first_byte!(0b0000_1111);
+                    continuation_byte!(0x80...0x9F);  // 0xA0..0xBF here are surrogates
+                    continuation_byte!();
+                }
+                0xF0 => {
+                    first_byte!(0b0000_0111);
+                    continuation_byte!(0x90...0xBF);  // 0x80..0x8F here are overlong
+                    continuation_byte!();
+                    continuation_byte!();
+                }
+                0xF1...0xF3 => {
+                    first_byte!(0b0000_0111);
+                    continuation_byte!();
+                    continuation_byte!();
+                    continuation_byte!();
+                }
+                0xF4 => {
+                    first_byte!(0b0000_0111);
+                    continuation_byte!(0x80...0x8F);  // 0x90..0xBF here are beyond char::MAX
+                    continuation_byte!();
+                    continuation_byte!();
+                }
+                _ => return Err(InvalidSequence(()))  // Illegal first byte, overlong, or beyond MAX
+            }
+            unsafe {
+                Ok(from_u32_unchecked(code_point))
+            }
+        })
+    }
+
+    #[inline]
+    fn size_hint(&self) -> (usize, Option<usize>) {
+        let (lower, upper) = self.0.size_hint();
+
+        // A code point is at most 4 bytes long.
+        let min_code_points = lower / 4;
+
+        (min_code_points, upper)
+    }
+}
+
+#[unstable(feature = "decode_utf8", issue = "33906")]
+impl<I: FusedIterator<Item = u8>> FusedIterator for DecodeUtf8<I> {}
diff --git a/src/libcore/char/printable.py b/src/libcore/char/printable.py
new file mode 100644
index 00000000000..484822e10aa
--- /dev/null
+++ b/src/libcore/char/printable.py
@@ -0,0 +1,254 @@
+#!/usr/bin/env python
+#
+# Copyright 2011-2016 The Rust Project Developers. See the COPYRIGHT
+# file at the top-level directory of this distribution and at
+# http://rust-lang.org/COPYRIGHT.
+#
+# Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+# http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+# <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+# option. This file may not be copied, modified, or distributed
+# except according to those terms.
+
+# This script uses the following Unicode tables:
+# - UnicodeData.txt
+
+
+from collections import namedtuple
+import csv
+import os
+import subprocess
+
+NUM_CODEPOINTS=0x110000
+
+def to_ranges(iter):
+    current = None
+    for i in iter:
+        if current is None or i != current[1] or i in (0x10000, 0x20000):
+            if current is not None:
+                yield tuple(current)
+            current = [i, i + 1]
+        else:
+            current[1] += 1
+    if current is not None:
+        yield tuple(current)
+
+def get_escaped(codepoints):
+    for c in codepoints:
+        if (c.class_ or "Cn") in "Cc Cf Cs Co Cn Zl Zp Zs".split() and c.value != ord(' '):
+            yield c.value
+
+def get_file(f):
+    try:
+        return open(os.path.basename(f))
+    except FileNotFoundError:
+        subprocess.run(["curl", "-O", f], check=True)
+        return open(os.path.basename(f))
+
+Codepoint = namedtuple('Codepoint', 'value class_')
+
+def get_codepoints(f):
+    r = csv.reader(f, delimiter=";")
+    prev_codepoint = 0
+    class_first = None
+    for row in r:
+        codepoint = int(row[0], 16)
+        name = row[1]
+        class_ = row[2]
+
+        if class_first is not None:
+            if not name.endswith("Last>"):
+                raise ValueError("Missing Last after First")
+
+        for c in range(prev_codepoint + 1, codepoint):
+            yield Codepoint(c, class_first)
+
+        class_first = None
+        if name.endswith("First>"):
+            class_first = class_
+
+        yield Codepoint(codepoint, class_)
+        prev_codepoint = codepoint
+
+    if class_first != None:
+        raise ValueError("Missing Last after First")
+
+    for c in range(prev_codepoint + 1, NUM_CODEPOINTS):
+        yield Codepoint(c, None)
+
+def compress_singletons(singletons):
+    uppers = [] # (upper, # items in lowers)
+    lowers = []
+
+    for i in singletons:
+        upper = i >> 8
+        lower = i & 0xff
+        if len(uppers) == 0 or uppers[-1][0] != upper:
+            uppers.append((upper, 1))
+        else:
+            upper, count = uppers[-1]
+            uppers[-1] = upper, count + 1
+        lowers.append(lower)
+
+    return uppers, lowers
+
+def compress_normal(normal):
+    # lengths 0x00..0x7f are encoded as 00, 01, ..., 7e, 7f
+    # lengths 0x80..0x7fff are encoded as 80 80, 80 81, ..., ff fe, ff ff
+    compressed = [] # [truelen, (truelenaux), falselen, (falselenaux)]
+
+    prev_start = 0
+    for start, count in normal:
+        truelen = start - prev_start
+        falselen = count
+        prev_start = start + count
+
+        assert truelen < 0x8000 and falselen < 0x8000
+        entry = []
+        if truelen > 0x7f:
+            entry.append(0x80 | (truelen >> 8))
+            entry.append(truelen & 0xff)
+        else:
+            entry.append(truelen & 0x7f)
+        if falselen > 0x7f:
+            entry.append(0x80 | (falselen >> 8))
+            entry.append(falselen & 0xff)
+        else:
+            entry.append(falselen & 0x7f)
+
+        compressed.append(entry)
+
+    return compressed
+
+def print_singletons(uppers, lowers, uppersname, lowersname):
+    print("const {}: &'static [(u8, u8)] = &[".format(uppersname))
+    for u, c in uppers:
+        print("    ({:#04x}, {}),".format(u, c))
+    print("];")
+    print("const {}: &'static [u8] = &[".format(lowersname))
+    for i in range(0, len(lowers), 8):
+        print("    {}".format(" ".join("{:#04x},".format(l) for l in lowers[i:i+8])))
+    print("];")
+
+def print_normal(normal, normalname):
+    print("const {}: &'static [u8] = &[".format(normalname))
+    for v in normal:
+        print("    {}".format(" ".join("{:#04x},".format(i) for i in v)))
+    print("];")
+
+def main():
+    file = get_file("http://www.unicode.org/Public/UNIDATA/UnicodeData.txt")
+
+    codepoints = get_codepoints(file)
+
+    CUTOFF=0x10000
+    singletons0 = []
+    singletons1 = []
+    normal0 = []
+    normal1 = []
+    extra = []
+
+    for a, b in to_ranges(get_escaped(codepoints)):
+        if a > 2 * CUTOFF:
+            extra.append((a, b - a))
+        elif a == b - 1:
+            if a & CUTOFF:
+                singletons1.append(a & ~CUTOFF)
+            else:
+                singletons0.append(a)
+        elif a == b - 2:
+            if a & CUTOFF:
+                singletons1.append(a & ~CUTOFF)
+                singletons1.append((a + 1) & ~CUTOFF)
+            else:
+                singletons0.append(a)
+                singletons0.append(a + 1)
+        else:
+            if a >= 2 * CUTOFF:
+                extra.append((a, b - a))
+            elif a & CUTOFF:
+                normal1.append((a & ~CUTOFF, b - a))
+            else:
+                normal0.append((a, b - a))
+
+    singletons0u, singletons0l = compress_singletons(singletons0)
+    singletons1u, singletons1l = compress_singletons(singletons1)
+    normal0 = compress_normal(normal0)
+    normal1 = compress_normal(normal1)
+
+    print("""\
+// Copyright 2012-2017 The Rust Project Developers. See the COPYRIGHT
+// file at the top-level directory of this distribution and at
+// http://rust-lang.org/COPYRIGHT.
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+// NOTE: The following code was generated by "src/libcore/char/printable.py",
+//       do not edit directly!
+
+fn check(x: u16, singletonuppers: &[(u8, u8)], singletonlowers: &[u8],
+         normal: &[u8]) -> bool {
+    let xupper = (x >> 8) as u8;
+    let mut lowerstart = 0;
+    for &(upper, lowercount) in singletonuppers {
+        let lowerend = lowerstart + lowercount as usize;
+        if xupper == upper {
+            for &lower in &singletonlowers[lowerstart..lowerend] {
+                if lower == x as u8 {
+                    return false;
+                }
+            }
+        } else if xupper < upper {
+            break;
+        }
+        lowerstart = lowerend;
+    }
+
+    let mut x = x as i32;
+    let mut normal = normal.iter().cloned();
+    let mut current = true;
+    while let Some(v) = normal.next() {
+        let len = if v & 0x80 != 0 {
+            ((v & 0x7f) as i32) << 8 | normal.next().unwrap() as i32
+        } else {
+            v as i32
+        };
+        x -= len;
+        if x < 0 {
+            break;
+        }
+        current = !current;
+    }
+    current
+}
+
+pub(crate) fn is_printable(x: char) -> bool {
+    let x = x as u32;
+    let lower = x as u16;
+    if x < 0x10000 {
+        check(lower, SINGLETONS0U, SINGLETONS0L, NORMAL0)
+    } else if x < 0x20000 {
+        check(lower, SINGLETONS1U, SINGLETONS1L, NORMAL1)
+    } else {\
+""")
+    for a, b in extra:
+        print("        if 0x{:x} <= x && x < 0x{:x} {{".format(a, a + b))
+        print("            return false;")
+        print("        }")
+    print("""\
+        true
+    }
+}\
+""")
+    print()
+    print_singletons(singletons0u, singletons0l, 'SINGLETONS0U', 'SINGLETONS0L')
+    print_singletons(singletons1u, singletons1l, 'SINGLETONS1U', 'SINGLETONS1L')
+    print_normal(normal0, 'NORMAL0')
+    print_normal(normal1, 'NORMAL1')
+
+if __name__ == '__main__':
+    main()
diff --git a/src/libcore/char/printable.rs b/src/libcore/char/printable.rs
new file mode 100644
index 00000000000..ce011fab187
--- /dev/null
+++ b/src/libcore/char/printable.rs
@@ -0,0 +1,531 @@
+// Copyright 2012-2017 The Rust Project Developers. See the COPYRIGHT
+// file at the top-level directory of this distribution and at
+// http://rust-lang.org/COPYRIGHT.
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+// NOTE: The following code was generated by "src/libcore/char/printable.py",
+//       do not edit directly!
+
+fn check(x: u16, singletonuppers: &[(u8, u8)], singletonlowers: &[u8],
+         normal: &[u8]) -> bool {
+    let xupper = (x >> 8) as u8;
+    let mut lowerstart = 0;
+    for &(upper, lowercount) in singletonuppers {
+        let lowerend = lowerstart + lowercount as usize;
+        if xupper == upper {
+            for &lower in &singletonlowers[lowerstart..lowerend] {
+                if lower == x as u8 {
+                    return false;
+                }
+            }
+        } else if xupper < upper {
+            break;
+        }
+        lowerstart = lowerend;
+    }
+
+    let mut x = x as i32;
+    let mut normal = normal.iter().cloned();
+    let mut current = true;
+    while let Some(v) = normal.next() {
+        let len = if v & 0x80 != 0 {
+            ((v & 0x7f) as i32) << 8 | normal.next().unwrap() as i32
+        } else {
+            v as i32
+        };
+        x -= len;
+        if x < 0 {
+            break;
+        }
+        current = !current;
+    }
+    current
+}
+
+pub(crate) fn is_printable(x: char) -> bool {
+    let x = x as u32;
+    let lower = x as u16;
+    if x < 0x10000 {
+        check(lower, SINGLETONS0U, SINGLETONS0L, NORMAL0)
+    } else if x < 0x20000 {
+        check(lower, SINGLETONS1U, SINGLETONS1L, NORMAL1)
+    } else {
+        if 0x2a6d7 <= x && x < 0x2a700 {
+            return false;
+        }
+        if 0x2b735 <= x && x < 0x2b740 {
+            return false;
+        }
+        if 0x2b81e <= x && x < 0x2b820 {
+            return false;
+        }
+        if 0x2cea2 <= x && x < 0x2ceb0 {
+            return false;
+        }
+        if 0x2ebe1 <= x && x < 0x2f800 {
+            return false;
+        }
+        if 0x2fa1e <= x && x < 0xe0100 {
+            return false;
+        }
+        if 0xe01f0 <= x && x < 0x110000 {
+            return false;
+        }
+        true
+    }
+}
+
+const SINGLETONS0U: &'static [(u8, u8)] = &[
+    (0x00, 1),
+    (0x03, 5),
+    (0x05, 8),
+    (0x06, 3),
+    (0x07, 4),
+    (0x08, 8),
+    (0x09, 16),
+    (0x0a, 27),
+    (0x0b, 25),
+    (0x0c, 22),
+    (0x0d, 18),
+    (0x0e, 22),
+    (0x0f, 4),
+    (0x10, 3),
+    (0x12, 18),
+    (0x13, 9),
+    (0x16, 1),
+    (0x17, 5),
+    (0x18, 2),
+    (0x19, 3),
+    (0x1a, 7),
+    (0x1d, 1),
+    (0x1f, 22),
+    (0x20, 3),
+    (0x2b, 5),
+    (0x2c, 2),
+    (0x2d, 11),
+    (0x2e, 1),
+    (0x30, 3),
+    (0x31, 3),
+    (0x32, 2),
+    (0xa7, 1),
+    (0xa8, 2),
+    (0xa9, 2),
+    (0xaa, 4),
+    (0xab, 8),
+    (0xfa, 2),
+    (0xfb, 5),
+    (0xfd, 4),
+    (0xfe, 3),
+    (0xff, 9),
+];
+const SINGLETONS0L: &'static [u8] = &[
+    0xad, 0x78, 0x79, 0x8b, 0x8d, 0xa2, 0x30, 0x57,
+    0x58, 0x60, 0x88, 0x8b, 0x8c, 0x90, 0x1c, 0x1d,
+    0xdd, 0x0e, 0x0f, 0x4b, 0x4c, 0x2e, 0x2f, 0x3f,
+    0x5c, 0x5d, 0x5f, 0xb5, 0xe2, 0x84, 0x8d, 0x8e,
+    0x91, 0x92, 0xa9, 0xb1, 0xba, 0xbb, 0xc5, 0xc6,
+    0xc9, 0xca, 0xde, 0xe4, 0xe5, 0x04, 0x11, 0x12,
+    0x29, 0x31, 0x34, 0x37, 0x3a, 0x3b, 0x3d, 0x49,
+    0x4a, 0x5d, 0x84, 0x8e, 0x92, 0xa9, 0xb1, 0xb4,
+    0xba, 0xbb, 0xc6, 0xca, 0xce, 0xcf, 0xe4, 0xe5,
+    0x00, 0x04, 0x0d, 0x0e, 0x11, 0x12, 0x29, 0x31,
+    0x34, 0x3a, 0x3b, 0x45, 0x46, 0x49, 0x4a, 0x5e,
+    0x64, 0x65, 0x84, 0x91, 0x9b, 0x9d, 0xc9, 0xce,
+    0xcf, 0x04, 0x0d, 0x11, 0x29, 0x45, 0x49, 0x57,
+    0x64, 0x65, 0x84, 0x8d, 0x91, 0xa9, 0xb4, 0xba,
+    0xbb, 0xc5, 0xc9, 0xdf, 0xe4, 0xe5, 0xf0, 0x04,
+    0x0d, 0x11, 0x45, 0x49, 0x64, 0x65, 0x80, 0x81,
+    0x84, 0xb2, 0xbc, 0xbe, 0xbf, 0xd5, 0xd7, 0xf0,
+    0xf1, 0x83, 0x85, 0x86, 0x89, 0x8b, 0x8c, 0x98,
+    0xa0, 0xa4, 0xa6, 0xa8, 0xa9, 0xac, 0xba, 0xbe,
+    0xbf, 0xc5, 0xc7, 0xce, 0xcf, 0xda, 0xdb, 0x48,
+    0x98, 0xbd, 0xcd, 0xc6, 0xce, 0xcf, 0x49, 0x4e,
+    0x4f, 0x57, 0x59, 0x5e, 0x5f, 0x89, 0x8e, 0x8f,
+    0xb1, 0xb6, 0xb7, 0xbf, 0xc1, 0xc6, 0xc7, 0xd7,
+    0x11, 0x16, 0x17, 0x5b, 0x5c, 0xf6, 0xf7, 0xfe,
+    0xff, 0x80, 0x0d, 0x6d, 0x71, 0xde, 0xdf, 0x0e,
+    0x0f, 0x1f, 0x6e, 0x6f, 0x1c, 0x1d, 0x5f, 0x7d,
+    0x7e, 0xae, 0xaf, 0xfa, 0x16, 0x17, 0x1e, 0x1f,
+    0x46, 0x47, 0x4e, 0x4f, 0x58, 0x5a, 0x5c, 0x5e,
+    0x7e, 0x7f, 0xb5, 0xc5, 0xd4, 0xd5, 0xdc, 0xf0,
+    0xf1, 0xf5, 0x72, 0x73, 0x8f, 0x74, 0x75, 0x96,
+    0x97, 0xc9, 0x2f, 0x5f, 0x26, 0x2e, 0x2f, 0xa7,
+    0xaf, 0xb7, 0xbf, 0xc7, 0xcf, 0xd7, 0xdf, 0x9a,
+    0x40, 0x97, 0x98, 0x2f, 0x30, 0x8f, 0x1f, 0xff,
+    0xaf, 0xfe, 0xff, 0xce, 0xff, 0x4e, 0x4f, 0x5a,
+    0x5b, 0x07, 0x08, 0x0f, 0x10, 0x27, 0x2f, 0xee,
+    0xef, 0x6e, 0x6f, 0x37, 0x3d, 0x3f, 0x42, 0x45,
+    0x90, 0x91, 0xfe, 0xff, 0x53, 0x67, 0x75, 0xc8,
+    0xc9, 0xd0, 0xd1, 0xd8, 0xd9, 0xe7, 0xfe, 0xff,
+];
+const SINGLETONS1U: &'static [(u8, u8)] = &[
+    (0x00, 6),
+    (0x01, 1),
+    (0x03, 1),
+    (0x04, 2),
+    (0x08, 8),
+    (0x09, 2),
+    (0x0a, 3),
+    (0x0b, 2),
+    (0x10, 1),
+    (0x11, 4),
+    (0x12, 5),
+    (0x13, 18),
+    (0x14, 2),
+    (0x15, 2),
+    (0x1a, 3),
+    (0x1c, 5),
+    (0x1d, 4),
+    (0x24, 1),
+    (0x6a, 3),
+    (0x6b, 2),
+    (0xbc, 2),
+    (0xd1, 2),
+    (0xd4, 12),
+    (0xd5, 9),
+    (0xd6, 2),
+    (0xd7, 2),
+    (0xda, 1),
+    (0xe0, 5),
+    (0xe8, 2),
+    (0xee, 32),
+    (0xf0, 4),
+    (0xf1, 1),
+    (0xf9, 1),
+];
+const SINGLETONS1L: &'static [u8] = &[
+    0x0c, 0x27, 0x3b, 0x3e, 0x4e, 0x4f, 0x8f, 0x9e,
+    0x9e, 0x9f, 0x06, 0x07, 0x09, 0x36, 0x3d, 0x3e,
+    0x56, 0xf3, 0xd0, 0xd1, 0x04, 0x14, 0x18, 0x56,
+    0x57, 0xbd, 0x35, 0xce, 0xcf, 0xe0, 0x12, 0x87,
+    0x89, 0x8e, 0x9e, 0x04, 0x0d, 0x0e, 0x11, 0x12,
+    0x29, 0x31, 0x34, 0x3a, 0x3b, 0x45, 0x46, 0x49,
+    0x4a, 0x4e, 0x4f, 0x64, 0x65, 0x5a, 0x5c, 0xb6,
+    0xb7, 0x84, 0x85, 0x9d, 0x09, 0x37, 0x90, 0x91,
+    0xa8, 0x07, 0x0a, 0x3b, 0x3e, 0x6f, 0x5f, 0xee,
+    0xef, 0x5a, 0x62, 0x9a, 0x9b, 0x27, 0x28, 0x55,
+    0x9d, 0xa0, 0xa1, 0xa3, 0xa4, 0xa7, 0xa8, 0xad,
+    0xba, 0xbc, 0xc4, 0x06, 0x0b, 0x0c, 0x15, 0x1d,
+    0x3a, 0x3f, 0x45, 0x51, 0xa6, 0xa7, 0xcc, 0xcd,
+    0xa0, 0x07, 0x19, 0x1a, 0x22, 0x25, 0xc5, 0xc6,
+    0x04, 0x20, 0x23, 0x25, 0x26, 0x28, 0x33, 0x38,
+    0x3a, 0x48, 0x4a, 0x4c, 0x50, 0x53, 0x55, 0x56,
+    0x58, 0x5a, 0x5c, 0x5e, 0x60, 0x63, 0x65, 0x66,
+    0x6b, 0x73, 0x78, 0x7d, 0x7f, 0x8a, 0xa4, 0xaa,
+    0xaf, 0xb0, 0xc0, 0xd0, 0x2f, 0x3f,
+];
+const NORMAL0: &'static [u8] = &[
+    0x00, 0x20,
+    0x5f, 0x22,
+    0x82, 0xdf, 0x04,
+    0x82, 0x44, 0x08,
+    0x1b, 0x05,
+    0x05, 0x11,
+    0x81, 0xac, 0x0e,
+    0x3b, 0x05,
+    0x6b, 0x35,
+    0x1e, 0x16,
+    0x80, 0xdf, 0x03,
+    0x19, 0x08,
+    0x01, 0x04,
+    0x22, 0x03,
+    0x0a, 0x04,
+    0x34, 0x04,
+    0x07, 0x03,
+    0x01, 0x07,
+    0x06, 0x07,
+    0x10, 0x0b,
+    0x50, 0x0f,
+    0x12, 0x07,
+    0x55, 0x08,
+    0x02, 0x04,
+    0x1c, 0x0a,
+    0x09, 0x03,
+    0x08, 0x03,
+    0x07, 0x03,
+    0x02, 0x03,
+    0x03, 0x03,
+    0x0c, 0x04,
+    0x05, 0x03,
+    0x0b, 0x06,
+    0x01, 0x0e,
+    0x15, 0x05,
+    0x3a, 0x03,
+    0x11, 0x07,
+    0x06, 0x05,
+    0x10, 0x08,
+    0x56, 0x07,
+    0x02, 0x07,
+    0x15, 0x0d,
+    0x50, 0x04,
+    0x43, 0x03,
+    0x2d, 0x03,
+    0x01, 0x04,
+    0x11, 0x06,
+    0x0f, 0x0c,
+    0x3a, 0x04,
+    0x1d, 0x25,
+    0x0d, 0x06,
+    0x4c, 0x20,
+    0x6d, 0x04,
+    0x6a, 0x25,
+    0x80, 0xc8, 0x05,
+    0x82, 0xb0, 0x03,
+    0x1a, 0x06,
+    0x82, 0xfd, 0x03,
+    0x59, 0x07,
+    0x15, 0x0b,
+    0x17, 0x09,
+    0x14, 0x0c,
+    0x14, 0x0c,
+    0x6a, 0x06,
+    0x0a, 0x06,
+    0x1a, 0x06,
+    0x58, 0x08,
+    0x2b, 0x05,
+    0x46, 0x0a,
+    0x2c, 0x04,
+    0x0c, 0x04,
+    0x01, 0x03,
+    0x31, 0x0b,
+    0x2c, 0x04,
+    0x1a, 0x06,
+    0x0b, 0x03,
+    0x80, 0xac, 0x06,
+    0x0a, 0x06,
+    0x1f, 0x41,
+    0x4c, 0x04,
+    0x2d, 0x03,
+    0x74, 0x08,
+    0x3c, 0x03,
+    0x0f, 0x03,
+    0x3c, 0x37,
+    0x08, 0x08,
+    0x2a, 0x06,
+    0x82, 0xff, 0x11,
+    0x18, 0x08,
+    0x2f, 0x11,
+    0x2d, 0x03,
+    0x20, 0x10,
+    0x21, 0x0f,
+    0x80, 0x8c, 0x04,
+    0x82, 0x97, 0x19,
+    0x0b, 0x15,
+    0x87, 0x5a, 0x03,
+    0x16, 0x19,
+    0x04, 0x10,
+    0x80, 0xf4, 0x05,
+    0x2f, 0x05,
+    0x3b, 0x07,
+    0x02, 0x0e,
+    0x18, 0x09,
+    0x80, 0xaa, 0x36,
+    0x74, 0x0c,
+    0x80, 0xd6, 0x1a,
+    0x0c, 0x05,
+    0x80, 0xff, 0x05,
+    0x80, 0xb6, 0x05,
+    0x24, 0x0c,
+    0x9b, 0xc6, 0x0a,
+    0xd2, 0x2b, 0x15,
+    0x84, 0x8d, 0x03,
+    0x37, 0x09,
+    0x81, 0x5c, 0x14,
+    0x80, 0xb8, 0x08,
+    0x80, 0xb8, 0x3f,
+    0x35, 0x04,
+    0x0a, 0x06,
+    0x38, 0x08,
+    0x46, 0x08,
+    0x0c, 0x06,
+    0x74, 0x0b,
+    0x1e, 0x03,
+    0x5a, 0x04,
+    0x59, 0x09,
+    0x80, 0x83, 0x18,
+    0x1c, 0x0a,
+    0x16, 0x09,
+    0x46, 0x0a,
+    0x80, 0x8a, 0x06,
+    0xab, 0xa4, 0x0c,
+    0x17, 0x04,
+    0x31, 0xa1, 0x04,
+    0x81, 0xda, 0x26,
+    0x07, 0x0c,
+    0x05, 0x05,
+    0x80, 0xa5, 0x11,
+    0x81, 0x6d, 0x10,
+    0x78, 0x28,
+    0x2a, 0x06,
+    0x4c, 0x04,
+    0x80, 0x8d, 0x04,
+    0x80, 0xbe, 0x03,
+    0x1b, 0x03,
+    0x0f, 0x0d,
+];
+const NORMAL1: &'static [u8] = &[
+    0x5e, 0x22,
+    0x7b, 0x05,
+    0x03, 0x04,
+    0x2d, 0x03,
+    0x65, 0x04,
+    0x01, 0x2f,
+    0x2e, 0x80, 0x82,
+    0x1d, 0x03,
+    0x31, 0x0f,
+    0x1c, 0x04,
+    0x24, 0x09,
+    0x1e, 0x05,
+    0x2b, 0x05,
+    0x44, 0x04,
+    0x0e, 0x2a,
+    0x80, 0xaa, 0x06,
+    0x24, 0x04,
+    0x24, 0x04,
+    0x28, 0x08,
+    0x34, 0x0b,
+    0x01, 0x80, 0x90,
+    0x81, 0x37, 0x09,
+    0x16, 0x0a,
+    0x08, 0x80, 0x98,
+    0x39, 0x03,
+    0x63, 0x08,
+    0x09, 0x30,
+    0x16, 0x05,
+    0x21, 0x03,
+    0x1b, 0x05,
+    0x01, 0x40,
+    0x38, 0x04,
+    0x4b, 0x05,
+    0x28, 0x04,
+    0x03, 0x04,
+    0x09, 0x08,
+    0x09, 0x07,
+    0x40, 0x20,
+    0x27, 0x04,
+    0x0c, 0x09,
+    0x36, 0x03,
+    0x3a, 0x05,
+    0x1a, 0x07,
+    0x04, 0x0c,
+    0x07, 0x50,
+    0x49, 0x37,
+    0x33, 0x0d,
+    0x33, 0x07,
+    0x06, 0x81, 0x60,
+    0x1f, 0x81, 0x81,
+    0x4e, 0x04,
+    0x1e, 0x0f,
+    0x43, 0x0e,
+    0x19, 0x07,
+    0x0a, 0x06,
+    0x44, 0x0c,
+    0x27, 0x09,
+    0x75, 0x0b,
+    0x3f, 0x41,
+    0x2a, 0x06,
+    0x3b, 0x05,
+    0x0a, 0x06,
+    0x51, 0x06,
+    0x01, 0x05,
+    0x10, 0x03,
+    0x05, 0x80, 0x8b,
+    0x5e, 0x22,
+    0x48, 0x08,
+    0x0a, 0x80, 0xa6,
+    0x5e, 0x22,
+    0x45, 0x0b,
+    0x0a, 0x06,
+    0x0d, 0x13,
+    0x38, 0x08,
+    0x0a, 0x36,
+    0x1a, 0x03,
+    0x0f, 0x04,
+    0x10, 0x81, 0x60,
+    0x53, 0x0c,
+    0x01, 0x81, 0x00,
+    0x48, 0x08,
+    0x53, 0x1d,
+    0x39, 0x81, 0x07,
+    0x46, 0x0a,
+    0x1d, 0x03,
+    0x47, 0x49,
+    0x37, 0x03,
+    0x0e, 0x08,
+    0x0a, 0x82, 0xa6,
+    0x83, 0x9a, 0x66,
+    0x75, 0x0b,
+    0x80, 0xc4, 0x8a, 0xbc,
+    0x84, 0x2f, 0x8f, 0xd1,
+    0x82, 0x47, 0xa1, 0xb9,
+    0x82, 0x39, 0x07,
+    0x2a, 0x04,
+    0x02, 0x60,
+    0x26, 0x0a,
+    0x46, 0x0a,
+    0x28, 0x05,
+    0x13, 0x83, 0x70,
+    0x45, 0x0b,
+    0x2f, 0x10,
+    0x11, 0x40,
+    0x02, 0x1e,
+    0x97, 0xed, 0x13,
+    0x82, 0xf3, 0xa5, 0x0d,
+    0x81, 0x1f, 0x51,
+    0x81, 0x8c, 0x89, 0x04,
+    0x6b, 0x05,
+    0x0d, 0x03,
+    0x09, 0x07,
+    0x10, 0x93, 0x60,
+    0x80, 0xf6, 0x0a,
+    0x73, 0x08,
+    0x6e, 0x17,
+    0x46, 0x80, 0xba,
+    0x57, 0x09,
+    0x12, 0x80, 0x8e,
+    0x81, 0x47, 0x03,
+    0x85, 0x42, 0x0f,
+    0x15, 0x85, 0x50,
+    0x2b, 0x87, 0xd5,
+    0x80, 0xd7, 0x29,
+    0x4b, 0x05,
+    0x0a, 0x04,
+    0x02, 0x84, 0xa0,
+    0x3c, 0x06,
+    0x01, 0x04,
+    0x55, 0x05,
+    0x1b, 0x34,
+    0x02, 0x81, 0x0e,
+    0x2c, 0x04,
+    0x64, 0x0c,
+    0x56, 0x0a,
+    0x0d, 0x03,
+    0x5c, 0x04,
+    0x3d, 0x39,
+    0x1d, 0x0d,
+    0x2c, 0x04,
+    0x09, 0x07,
+    0x02, 0x0e,
+    0x06, 0x80, 0x9a,
+    0x83, 0xd5, 0x0b,
+    0x0d, 0x03,
+    0x09, 0x07,
+    0x74, 0x0c,
+    0x55, 0x2b,
+    0x0c, 0x04,
+    0x38, 0x08,
+    0x0a, 0x06,
+    0x28, 0x08,
+    0x1e, 0x52,
+    0x0c, 0x04,
+    0x3d, 0x03,
+    0x1c, 0x14,
+    0x18, 0x28,
+    0x01, 0x0f,
+    0x17, 0x86, 0x19,
+];