//===--- StringUTF16.swift ------------------------------------------------===//
//
// This source file is part of the Swift.org open source project
//
// Copyright (c) 2014 - 2017 Apple Inc. and the Swift project authors
// Licensed under Apache License v2.0 with Runtime Library Exception
//
// See https://swift.org/LICENSE.txt for license information
// See https://swift.org/CONTRIBUTORS.txt for the list of Swift project authors
//
//===----------------------------------------------------------------------===//

// FIXME(ABI)#71 : The UTF-16 string view should have a custom iterator type to
// allow performance optimizations of linear traversals.

extension String {
  /// A view of a string's contents as a collection of UTF-16 code units.
  ///
  /// You can access a string's view of UTF-16 code units by using its `utf16`
  /// property. A string's UTF-16 view encodes the string's Unicode scalar
  /// values as 16-bit integers.
  ///
  ///     let flowers = "Flowers 💐"
  ///     for v in flowers.utf16 {
  ///         print(v)
  ///     }
  ///     // 70
  ///     // 108
  ///     // 111
  ///     // 119
  ///     // 101
  ///     // 114
  ///     // 115
  ///     // 32
  ///     // 55357
  ///     // 56464
  ///
  /// Unicode scalar values that make up a string's contents can be up to 21
  /// bits long. The longer scalar values may need two `UInt16` values for
  /// storage. Those "pairs" of code units are called *surrogate pairs*.
  ///
  ///     let flowermoji = "💐"
  ///     for v in flowermoji.unicodeScalars {
  ///         print(v, v.value)
  ///     }
  ///     // 💐 128144
  ///
  ///     for v in flowermoji.utf16 {
  ///         print(v)
  ///     }
  ///     // 55357
  ///     // 56464
  ///
  /// To convert a `String.UTF16View` instance back into a string, use the
  /// `String` type's `init(_:)` initializer.
  ///
  ///     let favemoji = "My favorite emoji is 🎉"
  ///     if let i = favemoji.utf16.firstIndex(where: { $0 >= 128 }) {
  ///         let asciiPrefix = String(favemoji.utf16[..<i])!
  ///         print(asciiPrefix)
  ///     }
  ///     // Prints "My favorite emoji is "
  ///
  /// UTF16View Elements Match NSString Characters
  /// ============================================
  ///
  /// The UTF-16 code units of a string's `utf16` view match the elements
  /// accessed through indexed `NSString` APIs.
  ///
  ///     print(flowers.utf16.count)
  ///     // Prints "10"
  ///
  ///     let nsflowers = flowers as NSString
  ///     print(nsflowers.length)
  ///     // Prints "10"
  ///
  /// Unlike `NSString`, however, `String.UTF16View` does not use integer
  /// indices. If you need to access a specific position in a UTF-16 view, use
  /// Swift's index manipulation methods. The following example accesses the
  /// fourth code unit in both the `flowers` and `nsflowers` strings:
  ///
  ///     print(nsflowers.character(at: 3))
  ///     // Prints "119"
  ///
  ///     let i = flowers.utf16.index(flowers.utf16.startIndex, offsetBy: 3)
  ///     print(flowers.utf16[i])
  ///     // Prints "119"
  ///
  /// Although the Swift overlay updates many Objective-C methods to return
  /// native Swift indices and index ranges, some still return instances of
  /// `NSRange`. To convert an `NSRange` instance to a range of
  /// `String.Index`, use the `Range(_:in:)` initializer, which takes an
  /// `NSRange` and a string as arguments.
  ///
  ///     let snowy = "❄️ Let it snow! ☃️"
  ///     let nsrange = NSRange(location: 3, length: 12)
  ///     if let range = Range(nsrange, in: snowy) {
  ///         print(snowy[range])
  ///     }
  ///     // Prints "Let it snow!"
  @frozen
  public struct UTF16View: Sendable {
    @usableFromInline
    internal var _guts: _StringGuts

    @inlinable
    internal init(_ guts: _StringGuts) {
      self._guts = guts
      _invariantCheck()
    }
  }
}

extension String.UTF16View {
  #if !INTERNAL_CHECKS_ENABLED
  @inlinable @inline(__always) internal func _invariantCheck() {}
  #else
  @usableFromInline @inline(never) @_effects(releasenone)
  internal func _invariantCheck() {
    _internalInvariant(
      startIndex.transcodedOffset == 0 && endIndex.transcodedOffset == 0)
  }
  #endif // INTERNAL_CHECKS_ENABLED
}

extension String.UTF16View: BidirectionalCollection {
  public typealias Index = String.Index

  /// The position of the first code unit if the `String` is
  /// nonempty; identical to `endIndex` otherwise.
  @inlinable @inline(__always)
  public var startIndex: Index { return _guts.startIndex }

  /// The "past the end" position---that is, the position one greater than
  /// the last valid subscript argument.
  ///
  /// In an empty UTF-16 view, `endIndex` is equal to `startIndex`.
  @inlinable @inline(__always)
  public var endIndex: Index { return _guts.endIndex }
  
  @inlinable @inline(__always)
  public func index(after idx: Index) -> Index {
    if _slowPath(_guts.isForeign) { return _foreignIndex(after: idx) }
    if _guts.isASCII { return idx.nextEncoded }

    // For a BMP scalar (1-3 UTF-8 code units), advance past it. For a non-BMP
    // scalar, use a transcoded offset first.

    // TODO: If transcoded is 1, can we just skip ahead 4?

    let idx = _utf16AlignNativeIndex(idx)

    let len = _guts.fastUTF8ScalarLength(startingAt: idx._encodedOffset)
    if len == 4 && idx.transcodedOffset == 0 {
      return idx.nextTranscoded
    }
    return idx.strippingTranscoding.encoded(offsetBy: len)._scalarAligned
  }

  @inlinable @inline(__always)
  public func index(before idx: Index) -> Index {
    _precondition(!idx.isZeroPosition)
    if _slowPath(_guts.isForeign) { return _foreignIndex(before: idx) }
    if _guts.isASCII { return idx.priorEncoded }

    if idx.transcodedOffset != 0 {
      _internalInvariant(idx.transcodedOffset == 1)
      return idx.strippingTranscoding
    }

    let idx = _utf16AlignNativeIndex(idx)
    let len = _guts.fastUTF8ScalarLength(endingAt: idx._encodedOffset)
    if len == 4 {
      // 2 UTF-16 code units comprise this scalar; advance to the beginning and
      // start mid-scalar transcoding
      return idx.encoded(offsetBy: -len).nextTranscoded
    }

    // Single UTF-16 code unit
    _internalInvariant((1...3) ~= len)
    return idx.encoded(offsetBy: -len)._scalarAligned
  }

  public func index(_ i: Index, offsetBy n: Int) -> Index {
    if _slowPath(_guts.isForeign) {
      return _foreignIndex(i, offsetBy: n)
    }

    let lowerOffset = _nativeGetOffset(for: i)
    let result = _nativeGetIndex(for: lowerOffset + n)
    return result
  }

  public func index(
    _ i: Index, offsetBy n: Int, limitedBy limit: Index
  ) -> Index? {
    if _slowPath(_guts.isForeign) {
      return _foreignIndex(i, offsetBy: n, limitedBy: limit)
    }

    let iOffset = _nativeGetOffset(for: i)
    let limitOffset = _nativeGetOffset(for: limit)

    // If distance < 0, limit has no effect if it is greater than i.
    if _slowPath(n < 0 && limit <= i && limitOffset > iOffset + n) {
      return nil
    }
    // If distance > 0, limit has no effect if it is less than i.
    if _slowPath(n >= 0 && limit >= i && limitOffset < iOffset + n) {
      return nil
    }

    let result = _nativeGetIndex(for: iOffset + n)
    return result
  }

  public func distance(from start: Index, to end: Index) -> Int {
    if _slowPath(_guts.isForeign) {
      return _foreignDistance(from: start, to: end)
    }

    let lower = _nativeGetOffset(for: start)
    let upper = _nativeGetOffset(for: end)
    return upper &- lower
  }

  @inlinable
  public var count: Int {
    if _slowPath(_guts.isForeign) {
      return _foreignCount()
    }
    return _nativeGetOffset(for: endIndex)
  }

  /// Accesses the code unit at the given position.
  ///
  /// The following example uses the subscript to print the value of a
  /// string's first UTF-16 code unit.
  ///
  ///     let greeting = "Hello, friend!"
  ///     let i = greeting.utf16.startIndex
  ///     print("First character's UTF-16 code unit: \(greeting.utf16[i])")
  ///     // Prints "First character's UTF-16 code unit: 72"
  ///
  /// - Parameter position: A valid index of the view. `position` must be
  ///   less than the view's end index.
  @inlinable @inline(__always)
  public subscript(idx: Index) -> UTF16.CodeUnit {
    String(_guts)._boundsCheck(idx)

    if _fastPath(_guts.isFastUTF8) {
      let scalar = _guts.fastUTF8Scalar(
        startingAt: _guts.scalarAlign(idx)._encodedOffset)
      return scalar.utf16[idx.transcodedOffset]
    }

    return _foreignSubscript(position: idx)
  }
}

extension String.UTF16View {
  @frozen
  public struct Iterator: IteratorProtocol, Sendable {
    @usableFromInline
    internal var _guts: _StringGuts

    @usableFromInline
    internal var _position: Int = 0

    @usableFromInline
    internal var _end: Int

    // If non-nil, return this value for `next()` (and set it to nil).
    //
    // This is set when visiting a non-BMP scalar: the leading surrogate is
    // returned, this field is set with the value of the trailing surrogate, and
    // `_position` is advanced to the start of the next scalar.
    @usableFromInline
    internal var _nextIsTrailingSurrogate: UInt16? = nil

    @inlinable
    internal init(_ guts: _StringGuts) {
      self._end = guts.count
      self._guts = guts
    }

    @inlinable
    public mutating func next() -> UInt16? {
      if _slowPath(_nextIsTrailingSurrogate != nil) {
        let trailing = self._nextIsTrailingSurrogate._unsafelyUnwrappedUnchecked
        self._nextIsTrailingSurrogate = nil
        return trailing
      }
      guard _fastPath(_position < _end) else { return nil }

      let (scalar, len) = _guts.errorCorrectedScalar(startingAt: _position)
      _position &+= len

      if _slowPath(scalar.value > UInt16.max) {
        self._nextIsTrailingSurrogate = scalar.utf16[1]
        return scalar.utf16[0]
      }
      return UInt16(truncatingIfNeeded: scalar.value)
    }
  }
  @inlinable
  public __consuming func makeIterator() -> Iterator {
    return Iterator(_guts)
  }
}


extension String.UTF16View: CustomStringConvertible {
  @inlinable @inline(__always)
  public var description: String { return String(_guts) }
}

extension String.UTF16View: CustomDebugStringConvertible {
  public var debugDescription: String {
    return "StringUTF16(\(self.description.debugDescription))"
  }
}

extension String {
  /// A UTF-16 encoding of `self`.
  @inlinable
  public var utf16: UTF16View {
    @inline(__always) get { return UTF16View(_guts) }
    @inline(__always) set { self = String(newValue._guts) }
  }

  /// Creates a string corresponding to the given sequence of UTF-16 code units.
  @inlinable @inline(__always)
  @available(swift, introduced: 4.0)
  public init(_ utf16: UTF16View) {
    self.init(utf16._guts)
  }
}

// Index conversions
extension String.UTF16View.Index {
  /// Creates an index in the given UTF-16 view that corresponds exactly to the
  /// specified string position.
  ///
  /// If the index passed as `sourcePosition` represents either the start of a
  /// Unicode scalar value or the position of a UTF-16 trailing surrogate,
  /// then the initializer succeeds. If `sourcePosition` does not have an
  /// exact corresponding position in `target`, then the result is `nil`. For
  /// example, an attempt to convert the position of a UTF-8 continuation byte
  /// results in `nil`.
  ///
  /// The following example finds the position of a space in a string and then
  /// converts that position to an index in the string's `utf16` view.
  ///
  ///     let cafe = "Café 🍵"
  ///
  ///     let stringIndex = cafe.firstIndex(of: "é")!
  ///     let utf16Index = String.Index(stringIndex, within: cafe.utf16)!
  ///
  ///     print(String(cafe.utf16[...utf16Index])!)
  ///     // Prints "Café"
  ///
  /// - Parameters:
  ///   - sourcePosition: A position in at least one of the views of the string
  ///     shared by `target`.
  ///   - target: The `UTF16View` in which to find the new position.
  public init?(
    _ idx: String.Index, within target: String.UTF16View
  ) {
    if _slowPath(target._guts.isForeign) {
      guard idx._foreignIsWithin(target) else { return nil }
    } else {
      guard target._guts.isOnUnicodeScalarBoundary(idx) else { return nil }
    }

    self = idx
  }

  /// Returns the position in the given view of Unicode scalars that
  /// corresponds exactly to this index.
  ///
  /// This index must be a valid index of `String(unicodeScalars).utf16`.
  ///
  /// This example first finds the position of a space (UTF-16 code point `32`)
  /// in a string's `utf16` view and then uses this method to find the same
  /// position in the string's `unicodeScalars` view.
  ///
  ///     let cafe = "Café 🍵"
  ///     let i = cafe.utf16.firstIndex(of: 32)!
  ///     let j = i.samePosition(in: cafe.unicodeScalars)!
  ///     print(String(cafe.unicodeScalars[..<j]))
  ///     // Prints "Café"
  ///
  /// - Parameter unicodeScalars: The view to use for the index conversion.
  ///   This index must be a valid index of at least one view of the string
  ///   shared by `unicodeScalars`.
  /// - Returns: The position in `unicodeScalars` that corresponds exactly to
  ///   this index. If this index does not have an exact corresponding
  ///   position in `unicodeScalars`, this method returns `nil`. For example,
  ///   an attempt to convert the position of a UTF-16 trailing surrogate
  ///   returns `nil`.
  public func samePosition(
    in unicodeScalars: String.UnicodeScalarView
  ) -> String.UnicodeScalarIndex? {
    return String.UnicodeScalarIndex(self, within: unicodeScalars)
  }
}

#if SWIFT_ENABLE_REFLECTION
// Reflection
extension String.UTF16View: CustomReflectable {
  /// Returns a mirror that reflects the UTF-16 view of a string.
  public var customMirror: Mirror {
    return Mirror(self, unlabeledChildren: self)
  }
}
#endif

// Slicing
extension String.UTF16View {
  public typealias SubSequence = Substring.UTF16View

  public subscript(r: Range<Index>) -> Substring.UTF16View {
    return Substring.UTF16View(self, _bounds: r)
  }
}

// Foreign string support
extension String.UTF16View {
  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignIndex(after i: Index) -> Index {
    _internalInvariant(_guts.isForeign)
    return i.strippingTranscoding.nextEncoded
  }

  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignIndex(before i: Index) -> Index {
    _internalInvariant(_guts.isForeign)
    return i.strippingTranscoding.priorEncoded
  }

  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignSubscript(position i: Index) -> UTF16.CodeUnit {
    _internalInvariant(_guts.isForeign)
    return _guts.foreignErrorCorrectedUTF16CodeUnit(at: i.strippingTranscoding)
  }

  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignDistance(from start: Index, to end: Index) -> Int {
    _internalInvariant(_guts.isForeign)

    // Ignore transcoded offsets, i.e. scalar align if-and-only-if from a
    // transcoded view
    return end._encodedOffset - start._encodedOffset
  }

  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignIndex(
    _ i: Index, offsetBy n: Int, limitedBy limit: Index
  ) -> Index? {
    _internalInvariant(_guts.isForeign)
    let l = limit._encodedOffset - i._encodedOffset
    if n > 0 ? l >= 0 && l < n : l <= 0 && n < l {
      return nil
    }
    return i.strippingTranscoding.encoded(offsetBy: n)
  }

  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignIndex(_ i: Index, offsetBy n: Int) -> Index {
    _internalInvariant(_guts.isForeign)
    return i.strippingTranscoding.encoded(offsetBy: n)
  }

  @usableFromInline @inline(never)
  @_effects(releasenone)
  internal func _foreignCount() -> Int {
    _internalInvariant(_guts.isForeign)
    return endIndex._encodedOffset - startIndex._encodedOffset
  }

  // Align a native UTF-8 index to a valid UTF-16 position. If there is a
  // transcoded offset already, this is already a valid UTF-16 position
  // (referring to the second surrogate) and returns `idx`. Otherwise, this will
  // scalar-align the index. This is needed because we may be passed a
  // non-scalar-aligned index from the UTF8View.
  @_alwaysEmitIntoClient // Swift 5.1
  @inline(__always)
  internal func _utf16AlignNativeIndex(_ idx: String.Index) -> String.Index {
    _internalInvariant(!_guts.isForeign)
    guard idx.transcodedOffset == 0 else { return idx }
    return _guts.scalarAlign(idx)
  }
}

extension String.Index {
  @usableFromInline @inline(never) // opaque slow-path
  @_effects(releasenone)
  internal func _foreignIsWithin(_ target: String.UTF16View) -> Bool {
    _internalInvariant(target._guts.isForeign)

    // If we're transcoding, we're a UTF-8 view index, not UTF-16.
    return self.transcodedOffset == 0
  }
}

// Breadcrumb-aware acceleration
extension _StringGuts {
  @inline(__always)
  fileprivate func _useBreadcrumbs(forEncodedOffset offset: Int) -> Bool {
    return hasBreadcrumbs && offset >= _StringBreadcrumbs.breadcrumbStride
  }
}

extension String.UTF16View {
  
#if SWIFT_STDLIB_ENABLE_VECTOR_TYPES
  @inline(__always)
  internal func _utf16Length<U: SIMD, S: SIMD>(
    readPtr: inout UnsafeRawPointer,
    endPtr: UnsafeRawPointer,
    unsignedSIMDType: U.Type,
    signedSIMDType: S.Type
  ) -> Int where U.Scalar == UInt8, S.Scalar == Int8 {
    var utf16Count = 0
    
    while readPtr + MemoryLayout<U>.stride < endPtr {
      //Find the number of continuations (0b10xxxxxx)
      let sValue = Builtin.loadRaw(readPtr._rawValue) as S
      let continuations = S.zero.replacing(with: S.one, where: sValue .< -65 + 1)
            
      //Find the number of 4 byte code points (0b11110xxx)
      let uValue = Builtin.loadRaw(readPtr._rawValue) as U
      let fourBytes = S.zero.replacing(
        with: S.one,
        where: unsafeBitCast(
          uValue .>= 0b11110000,
          to: SIMDMask<S.MaskStorage>.self
        )
      )
      
      utf16Count &+= U.scalarCount + Int((fourBytes &- continuations).wrappedSum())
            
      readPtr += MemoryLayout<U>.stride
    }
    
    return utf16Count
  }
#endif
  
  @inline(__always)
  internal func _utf16Distance(from start: Index, to end: Index) -> Int {
    _internalInvariant(end.transcodedOffset == 0 || end.transcodedOffset == 1)
        
    return (end.transcodedOffset - start.transcodedOffset) + _guts.withFastUTF8(
      range: start._encodedOffset ..< end._encodedOffset
    ) { utf8 in
      let rawBuffer = UnsafeRawBufferPointer(utf8)
      guard rawBuffer.count > 0 else { return 0 }
      
      var utf16Count = 0
      var readPtr = rawBuffer.baseAddress.unsafelyUnwrapped
      let initialReadPtr = readPtr
      let endPtr = readPtr + rawBuffer.count
      
      //eat leading continuations
      while readPtr < endPtr {
        let byte = readPtr.load(as: UInt8.self)
        if !UTF8.isContinuation(byte) {
          break
        }
        readPtr += 1
      }

#if SWIFT_STDLIB_ENABLE_VECTOR_TYPES
      // TODO: Currently, using SIMD sizes above SIMD8 is slower
      // Once that's fixed we should go up to SIMD64 here
      
      utf16Count &+= _utf16Length(
        readPtr: &readPtr,
        endPtr: endPtr,
        unsignedSIMDType: SIMD8<UInt8>.self,
        signedSIMDType: SIMD8<Int8>.self
      )
   
      //TO CONSIDER: SIMD widths <8 here
      
      //back up to the start of the current scalar if we may have a trailing
      //incomplete scalar
      if utf16Count > 0 && UTF8.isContinuation(readPtr.load(as: UInt8.self)) {
        while readPtr > initialReadPtr && UTF8.isContinuation(readPtr.load(as: UInt8.self)) {
          readPtr -= 1
        }
        
        //The trailing scalar may be incomplete, subtract it out and check below
        let byte = readPtr.load(as: UInt8.self)
        let len = _utf8ScalarLength(byte)
        utf16Count &-= len == 4 ? 2 : 1
        if readPtr == initialReadPtr {
          //if we backed up all the way and didn't hit a non-continuation, then
          //we don't have any complete scalars, and we should bail.
          return 0
        }
      }
#endif

      //trailing bytes
      while readPtr < endPtr {
        let byte = readPtr.load(as: UInt8.self)
        let len = _utf8ScalarLength(byte)
        // if we don't have enough bytes left, we don't have a complete scalar,
        // so don't add it to the count.
        if readPtr + len <= endPtr {
          utf16Count &+= len == 4 ? 2 : 1
        }
        readPtr += len
      }

      return utf16Count
    }
  }
  
  @usableFromInline
  @_effects(releasenone)
  internal func _nativeGetOffset(for idx: Index) -> Int {
    // Trivial and common: start
    if idx == startIndex { return 0 }

    if _guts.isASCII {
      _internalInvariant(idx.transcodedOffset == 0)
      return idx._encodedOffset
    }

    let idx = _utf16AlignNativeIndex(idx)

    guard _guts._useBreadcrumbs(forEncodedOffset: idx._encodedOffset) else {
      return _utf16Distance(from: startIndex, to: idx)
    }

    // Simple and common: endIndex aka `length`.
    let breadcrumbsPtr = _guts.getBreadcrumbsPtr()
    if idx == endIndex { return breadcrumbsPtr.pointee.utf16Length }

    // Otherwise, find the nearest lower-bound breadcrumb and count from there
    let (crumb, crumbOffset) = breadcrumbsPtr.pointee.getBreadcrumb(
      forIndex: idx)
    
    return crumbOffset + _utf16Distance(from: crumb, to: idx)
  }

  @usableFromInline
  @_effects(releasenone)
  internal func _nativeGetIndex(for offset: Int) -> Index {
    // Trivial and common: start
    if offset == 0 { return startIndex }

    if _guts.isASCII { return Index(_encodedOffset: offset) }

    guard _guts._useBreadcrumbs(forEncodedOffset: offset) else {
      return _index(startIndex, offsetBy: offset)
    }

    // Simple and common: endIndex aka `length`.
    let breadcrumbsPtr = _guts.getBreadcrumbsPtr()
    if offset == breadcrumbsPtr.pointee.utf16Length { return endIndex }

    // Otherwise, find the nearest lower-bound breadcrumb and advance that
    let (crumb, remaining) = breadcrumbsPtr.pointee.getBreadcrumb(
      forOffset: offset)
    if remaining == 0 { return crumb }

    return _guts.withFastUTF8 { utf8 in
      var readIdx = crumb._encodedOffset
      let readEnd = utf8.count
      _internalInvariant(readIdx < readEnd)

      var utf16I = 0
      let utf16End: Int = remaining

      // Adjust for sub-scalar initial transcoding: If we're starting the scan
      // at a trailing surrogate, then we set our starting count to be -1 so as
      // offset counting the leading surrogate.
      if crumb.transcodedOffset != 0 {
        utf16I = -1
      }

      while true {
        let len = _utf8ScalarLength(utf8[_unchecked: readIdx])
        let utf16Len = len == 4 ? 2 : 1
        utf16I &+= utf16Len

        if utf16I >= utf16End {
          // Uncommon: final sub-scalar transcoded offset
          if _slowPath(utf16I > utf16End) {
            _internalInvariant(utf16Len == 2)
            return Index(encodedOffset: readIdx, transcodedOffset: 1)
          }
          return Index(_encodedOffset: readIdx &+ len)._scalarAligned
        }

        readIdx &+= len
      }
    }
  }

  // Copy (i.e. transcode to UTF-16) our contents into a buffer. `alignedRange`
  // means that the indices are part of the UTF16View.indices -- they are either
  // scalar-aligned or transcoded (e.g. derived from the UTF-16 view). They do
  // not need to go through an alignment check.
  internal func _nativeCopy(
    into buffer: UnsafeMutableBufferPointer<UInt16>,
    alignedRange range: Range<String.Index>
  ) {
    _internalInvariant(_guts.isFastUTF8)
    _internalInvariant(
      range.lowerBound == _utf16AlignNativeIndex(range.lowerBound))
    _internalInvariant(
      range.upperBound == _utf16AlignNativeIndex(range.upperBound))

    if _slowPath(range.isEmpty) { return }

    let isASCII = _guts.isASCII
    return _guts.withFastUTF8 { utf8 in
      var writeIdx = 0
      let writeEnd = buffer.count
      var readIdx = range.lowerBound._encodedOffset
      let readEnd = range.upperBound._encodedOffset

      if isASCII {
        _internalInvariant(range.lowerBound.transcodedOffset == 0)
        _internalInvariant(range.upperBound.transcodedOffset == 0)
        while readIdx < readEnd {
          _internalInvariant(utf8[readIdx] < 0x80)
          buffer[_unchecked: writeIdx] = UInt16(
            truncatingIfNeeded: utf8[_unchecked: readIdx])
          readIdx &+= 1
          writeIdx &+= 1
        }
        return
      }

      // Handle mid-transcoded-scalar initial index
      if _slowPath(range.lowerBound.transcodedOffset != 0) {
        _internalInvariant(range.lowerBound.transcodedOffset == 1)
        let (scalar, len) = _decodeScalar(utf8, startingAt: readIdx)
        buffer[writeIdx] = scalar.utf16[1]
        readIdx &+= len
        writeIdx &+= 1
      }

      // Transcode middle
      while readIdx < readEnd {
        let (scalar, len) = _decodeScalar(utf8, startingAt: readIdx)
        buffer[writeIdx] = scalar.utf16[0]
        readIdx &+= len
        writeIdx &+= 1
        if _slowPath(scalar.utf16.count == 2) {
          buffer[writeIdx] = scalar.utf16[1]
          writeIdx &+= 1
        }
      }

      // Handle mid-transcoded-scalar final index
      if _slowPath(range.upperBound.transcodedOffset == 1) {
        _internalInvariant(writeIdx < writeEnd)
        let (scalar, _) = _decodeScalar(utf8, startingAt: readIdx)
        _internalInvariant(scalar.utf16.count == 2)

        buffer[writeIdx] = scalar.utf16[0]
        writeIdx &+= 1
      }
      _internalInvariant(writeIdx <= writeEnd)
    }
  }
}