ByteVector.scala

/*
 * Copyright (c) 2013, Scodec
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 * 1. Redistributions of source code must retain the above copyright notice, this
 *    list of conditions and the following disclaimer.
 *
 * 2. Redistributions in binary form must reproduce the above copyright notice,
 *    this list of conditions and the following disclaimer in the documentation
 *    and/or other materials provided with the distribution.
 *
 * 3. Neither the name of the copyright holder nor the names of its contributors
 *    may be used to endorse or promote products derived from this software without
 *    specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

package scodec.bits

import java.io.{InputStream, OutputStream}
import java.nio.{ByteBuffer, CharBuffer}
import java.nio.charset.{CharacterCodingException, Charset, CodingErrorAction}
import java.util.UUID
import java.util.concurrent.atomic.{AtomicInteger, AtomicLong}

import scala.annotation.tailrec

/** An immutable vector of bytes, backed by a balanced binary tree of chunks. Most operations are
  * logarithmic in the depth of this tree, including `++`, `:+`, `+:`, `update`, and `insert`. Where
  * possible, operations return lazy views rather than copying any underlying bytes. Use `copy` to
  * copy all underlying bytes to a fresh, array-backed `ByteVector`.
  *
  * Unless otherwise noted, operations follow the same naming as the scala standard library
  * collections, though this class does not extend any of the standard scala collections. Use
  * `toIndexedSeq`, `toSeq`, or `toIterable` to obtain a regular `scala.collection` type.
  *
  * @groupname collection
  *   Collection Like Methods
  * @groupprio collection
  *   0
  *
  * @groupname bitwise
  *   Bitwise Operations
  * @groupprio bitwise
  *   1
  *
  * @groupname conversions
  *   Conversions
  * @groupprio conversions
  *   2
  *
  * @groupname buffer
  *   Buffering
  * @groupprio buffer
  *   3
  *
  * @groupname crypto
  *   Cryptography
  * @groupprio crypto
  *   4
  *
  * @define bitwiseOperationsReprDescription
  *   bit vector
  * @define returnsView
  *   This method returns a view and hence, is O(1). Call [[compact]] to generate a new strict
  *   vector.
  */
sealed abstract class ByteVector
    extends ByteVectorCrossPlatform
    with BitwiseOperations[ByteVector, Long]
    with Ordered[ByteVector]
    with Serializable {

  import ByteVector._

  /** Returns the number of bytes in this vector.
    * @group collection
    */
  def size: Long

  /** Returns the number of bytes in this vector, or `None` if the size does not fit into an `Int`.
    *
    * @group collection
    */
  final def intSize: Option[Int] = if (size <= Int.MaxValue) Some(size.toInt) else None

  /** Alias for [[size]].
    * @group collection
    */
  final def length: Long = size

  /** Returns true if this vector has no bytes.
    * @group collection
    */
  final def isEmpty: Boolean = size == 0

  /** Returns true if this vector has a non-zero number of bytes.
    * @group collection
    */
  final def nonEmpty: Boolean = !isEmpty

  /** Gets the byte at the specified index.
    * @throws IndexOutOfBoundsException
    *   if the specified index is not in `[0, size)`
    * @group collection
    */
  def get(index: Long): Byte = {
    checkIndex(index)
    getImpl(index)
  }

  protected def getImpl(index: Long): Byte

  /** Alias for [[get]].
    * @throws IndexOutOfBoundsException
    *   if the specified index is not in `[0, size)`
    * @group collection
    */
  final def apply(index: Long): Byte = get(index)

  /** Returns the byte at the specified index, or `None` if the index is out of range.
    * @group collection
    */
  final def lift(index: Long): Option[Byte] =
    if (index >= 0 && index < size) Some(apply(index))
    else None

  /** Returns a vector with the byte at the specified index replaced with the specified byte.
    * @group collection
    */
  final def update(idx: Long, b: Byte): ByteVector = {
    checkIndex(idx)
    (take(idx) :+ b) ++ drop(idx + 1)
  }

  /** Returns a vector with the specified byte inserted at the specified index.
    * @group collection
    */
  final def insert(idx: Long, b: Byte): ByteVector =
    (take(idx) :+ b) ++ drop(idx)

  /** Returns a vector with the specified byte vector inserted at the specified index.
    * @group collection
    */
  final def splice(idx: Long, b: ByteVector): ByteVector =
    take(idx) ++ b ++ drop(idx)

  /** Returns a vector with the specified byte vector replacing bytes `[idx, idx + b.size]`.
    * @group collection
    */
  final def patch(idx: Long, b: ByteVector): ByteVector =
    take(idx) ++ b ++ drop(idx + b.size)

  /** Returns a new byte vector representing this vector's contents followed by the specified
    * vector's contents.
    * @group collection
    */
  def ++(other: ByteVector): ByteVector =
    if (this.isEmpty) other
    else if (other.isEmpty) this
    else Chunks(Append(this, other)).bufferBy(64)

  /** Returns a new vector with the specified byte prepended.
    * @group collection
    */
  final def +:(byte: Byte): ByteVector = ByteVector(byte) ++ this

  /** Returns a new vector with the specified byte appended.
    * @group collection
    */
  def :+(byte: Byte): ByteVector =
    this ++ ByteVector(byte)

  /** Returns a vector of all bytes in this vector except the first `n` bytes.
    *
    * The resulting vector's size is `0 max (size - n)`.
    *
    * @group collection
    */
  def drop(n: Long): ByteVector = {
    val n1 = n.min(size).max(0)
    if (n1 == size) ByteVector.empty
    else if (n1 == 0) this
    else {
      @annotation.tailrec
      def go(cur: ByteVector, n1: Long, accR: List[ByteVector]): ByteVector =
        cur match {
          case Chunk(bs) => accR.foldLeft(Chunk(bs.drop(n1)): ByteVector)(_ ++ _).unbuffer
          case Append(l, r) =>
            if (n1 > l.size) go(r, n1 - l.size, accR)
            else go(l, n1, r :: accR)
          case b: Buffer =>
            if (n1 > b.hd.size) go(b.lastBytes, n1 - b.hd.size, accR)
            else go(b.hd, n1, b.lastBytes :: accR)
          case c: Chunks => go(c.chunks, n1, accR)
        }
      go(this, n1, Nil)
    }
  }

  /** Returns a vector of all bytes in this vector except the last `n` bytes.
    *
    * The resulting vector's size is `0 max (size - n)`.
    *
    * @group collection
    */
  final def dropRight(n: Long): ByteVector =
    take(size - n.max(0))

  /** Drops the longest prefix of this vector such that every byte of the prefix satisfies the
    * specific predicate.
    *
    * @group collection
    */
  final def dropWhile(f: Byte => Boolean): ByteVector = {
    var toDrop = 0L
    foreachSPartial(new F1BB {
      def apply(b: Byte) = {
        val cont = f(b)
        if (cont) toDrop += 1
        cont
      }
    })
    drop(toDrop)
  }

  /** Returns a vector of the first `n` bytes of this vector.
    *
    * The resulting vector's size is `n min size`.
    *
    * Note: if an `n`-byte vector is required, use the `acquire` method instead.
    *
    * @see
    *   acquire
    * @group collection
    */
  def take(n: Long): ByteVector = {
    val n1 = n.min(size).max(0)
    if (n1 == size) this
    else if (n1 == 0) ByteVector.empty
    else {
      @annotation.tailrec
      def go(accL: ByteVector, cur: ByteVector, n1: Long): ByteVector =
        cur match {
          case Chunk(bs) => accL ++ Chunk(bs.take(n1))
          case Append(l, r) =>
            if (n1 > l.size) go(accL ++ l, r, n1 - l.size)
            else go(accL, l, n1)
          case c: Chunks => go(accL, c.chunks, n1)
          case b: Buffer => go(accL, b.unbuffer, n1)
        }
      go(ByteVector.empty, this, n1)
    }
  }

  /** Returns a vector of the last `n` bytes of this vector.
    *
    * The resulting vector's size is `n min size`.
    *
    * @group collection
    */
  final def takeRight(n: Long): ByteVector =
    drop(size - n)

  /** Returns the longest prefix of this vector such that every byte satisfies the specific
    * predicate.
    *
    * @group collection
    */
  final def takeWhile(f: Byte => Boolean): ByteVector = {
    var toTake = 0L
    foreachSPartial(new F1BB {
      def apply(b: Byte) = {
        val cont = f(b)
        if (cont) toTake += 1
        cont
      }
    })
    take(toTake)
  }

  /** Returns a pair of vectors that is equal to `(take(n), drop(n))`.
    * @group collection
    */
  final def splitAt(n: Long): (ByteVector, ByteVector) = (take(n), drop(n))

  /** Returns a vector made up of the bytes starting at index `from` up to index `until`.
    * @group collection
    */
  final def slice(from: Long, until: Long): ByteVector =
    drop(from).take(until - (from.max(0)))

  /** Returns a vector whose contents are the results of taking the first `n` bytes of this vector.
    *
    * If this vector does not contain at least `n` bytes, an error message is returned.
    *
    * @see
    *   take
    * @group collection
    */
  def acquire(n: Long): Either[String, ByteVector] =
    if (n <= size) Right(take(n))
    else Left(s"cannot acquire $n bytes from a vector that contains $size bytes")

  /** Consumes the first `n` bytes of this vector and decodes them with the specified function,
    * resulting in a vector of the remaining bytes and the decoded value. If this vector does not
    * have `n` bytes or an error occurs while decoding, an error is returned instead.
    *
    * @group collection
    */
  final def consume[A](
      n: Long
  )(decode: ByteVector => Either[String, A]): Either[String, (ByteVector, A)] =
    for {
      toDecode <- acquire(n)
      decoded <- decode(toDecode)
    } yield (drop(n), decoded)

  /** Applies a binary operator to a start value and all elements of this vector, going left to
    * right.
    * @param z
    *   starting value
    * @param f
    *   operator to apply
    * @group collection
    */
  final def foldLeft[A](z: A)(f: (A, Byte) => A): A = {
    var acc = z
    foreachS(new F1BU { def apply(b: Byte) = acc = f(acc, b) })
    acc
  }

  /** Applies a binary operator to a start value and all elements of this vector, going right to
    * left.
    * @param z
    *   starting value
    * @param f
    *   operator to apply
    * @group collection
    */
  final def foldRight[A](z: A)(f: (Byte, A) => A): A =
    reverse.foldLeft(z)((tl, h) => f(h, tl))

  /** Applies a binary operator to a start value and all segments(views) of this ByteVector
    * expressed as read-only ByteBuffer, going left to right.
    * @param z
    *   Starting value
    * @param f
    *   operator to apply
    * @group collection
    */
  final def foldLeftBB[A](z: A)(f: (A, ByteBuffer) => A): A = {
    @annotation.tailrec
    def go(rem: List[ByteVector], a: A): A =
      rem match {
        case Chunk(bs) :: rem    => go(rem, f(a, bs.at.asByteBuffer(bs.offset, bs.size.toInt)))
        case Append(l, r) :: rem => go(l :: r :: rem, a)
        case Chunks(Append(l, r)) :: rem => go(l :: r :: rem, a)
        case (b: Buffer) :: rem          => go(b.unbuffer :: rem, a)
        case Nil                         => a
      }
    go(this :: Nil, z)
  }

  /** Applies a binary operator to a start value and all segments(views) of this ByteVector
    * expressed as read-only ByteBuffer, going right ot left.
    * @param z
    *   Starting value
    * @param f
    *   operator to apply
    * @group collection
    */
  final def foldRightBB[A](z: A)(f: (ByteBuffer, A) => A): A =
    reverse.foldLeftBB(z)((tl, h) => f(h, tl))

  /** Applies the specified function to each element of this vector.
    * @group collection
    */
  final def foreach(f: Byte => Unit): Unit = foreachS(new F1BU { def apply(b: Byte) = f(b) })

  private[scodec] final def foreachS(f: F1BU): Unit = foreachV(_.foreach(f))

  private[scodec] final def foreachSPartial(f: F1BB): Boolean = foreachVPartial(_.foreachPartial(f))

  private[scodec] final def foreachV(f: View => Unit): Unit = {
    @annotation.tailrec
    def go(rem: List[ByteVector]): Unit =
      rem match {
        case Chunk(bs) :: rem            => f(bs); go(rem)
        case Append(l, r) :: rem         => go(l :: r :: rem)
        case Chunks(Append(l, r)) :: rem => go(l :: r :: rem)
        case (b: Buffer) :: rem          => go(b.unbuffer :: rem)
        case Nil                         => ()
      }
    go(this :: Nil)
  }

  private[scodec] final def foreachVPartial(f: View => Boolean): Boolean = {
    @annotation.tailrec
    def go(rem: List[ByteVector]): Boolean =
      rem match {
        case Chunk(bs) :: rem            => if (f(bs)) go(rem) else false
        case Append(l, r) :: rem         => go(l :: r :: rem)
        case Chunks(Append(l, r)) :: rem => go(l :: r :: rem)
        case (b: Buffer) :: rem          => go(b.unbuffer :: rem)
        case Nil                         => true
      }
    go(this :: Nil)
  }

  /** Returns true if this byte vector starts with the specified vector.
    * @group collection
    */
  final def startsWith(b: ByteVector): Boolean =
    take(b.size) === b

  /** Returns true if this byte vector ends with the specified vector.
    * @group collection
    */
  final def endsWith(b: ByteVector): Boolean =
    takeRight(b.size) === b

  /** Finds the first index of the specified byte pattern in this vector.
    * @return
    *   index of slice or -1 if not found
    * @group collection
    */
  final def indexOfSlice(slice: ByteVector): Long = indexOfSlice(slice, 0)

  /** Finds the first index after `from` of the specified byte pattern in this vector.
    * @return
    *   index of slice or -1 if not found
    * @group collection
    */
  final def indexOfSlice(slice: ByteVector, from: Long): Long = {
    @annotation.tailrec
    def go(b: ByteVector, idx: Long): Long =
      if (b.startsWith(slice)) idx
      else if (b.isEmpty) -1
      else go(b.tail, idx + 1)
    go(drop(from), from)
  }

  /** Determines if the specified slice is in this vector.
    * @group collection
    */
  final def containsSlice(slice: ByteVector): Boolean = indexOfSlice(slice) >= 0

  // This was public before version 1.1.8 so it must stay here for bincompat
  // The public grouped method is adding via an extension method defined in the companion
  private[bits] final def grouped(chunkSize: Long): Stream[ByteVector] =
    groupedIterator(chunkSize).toStream

  private final def groupedIterator(chunkSize: Long): Iterator[ByteVector] =
    if (isEmpty) Iterator.empty
    else if (size <= chunkSize) Iterator(this)
    else Iterator(take(chunkSize)) ++ drop(chunkSize).groupedIterator(chunkSize)

  /** Returns the first byte of this vector or throws if vector is emtpy.
    * @group collection
    */
  final def head: Byte = apply(0)

  /** Returns the first byte of this vector or `None` if vector is emtpy.
    * @group collection
    */
  final def headOption: Option[Byte] = lift(0)

  /** Returns a vector of all bytes in this vector except the first byte.
    * @group collection
    */
  final def tail: ByteVector = drop(1)

  /** Returns a vector of all bytes in this vector except the last byte.
    * @group collection
    */
  final def init: ByteVector = dropRight(1)

  /** Returns the last byte in this vector or throws if vector is empty.
    * @group collection
    */
  final def last: Byte = apply(size - 1)

  /** Returns the last byte in this vector or returns `None` if vector is empty.
    * @group collection
    */
  final def lastOption: Option[Byte] = lift(size - 1)

  /** Alias for `padRight`.
    *
    * @throws IllegalArgumentException
    *   if `n < size`
    * @group collection
    */
  final def padTo(n: Long): ByteVector = padRight(n)

  /** Returns an `n`-byte vector whose contents are this vector's contents followed by 0 or more
    * zero bytes.
    *
    * @throws IllegalArgumentException
    *   if `n < size`
    * @group collection
    */
  final def padRight(n: Long): ByteVector =
    if (n < size) throw new IllegalArgumentException(s"ByteVector.padRight($n)")
    else this ++ ByteVector.fill(n - size)(0)

  /** Returns an `n`-bytes vector whose contents are 0 or more zero bytes followed by this vector's
    * contents.
    *
    * @throws IllegalArgumentException
    *   if `n < size`
    * @group collection
    */
  final def padLeft(n: Long): ByteVector =
    if (n < size) throw new IllegalArgumentException(s"ByteVector.padLeft($n)")
    else ByteVector.fill(n - size)(0) ++ this

  /** Returns a vector where each byte is the result of applying the specified function to the
    * corresponding byte in this vector. $returnsView
    * @group collection
    */
  final def map(f: Byte => Byte): ByteVector =
    ByteVector.viewAt((i: Long) => f(apply(i)), size)

  /** Returns a vector where each byte is the result of applying the specified function to the
    * corresponding byte in this vector. Only the least significant byte is used (the three most
    * significant bytes are ignored). $returnsView
    * @group collection
    */
  final def mapI(f: Byte => Int): ByteVector =
    map(f.andThen(_.toByte))

  private[scodec] final def mapS(f: F1B): ByteVector =
    ByteVector.view(new At { def apply(i: Long) = f(ByteVector.this(i)) }, size)

  /** Returns a vector with the bytes of this vector in reverse order. $returnsView
    * @group collection
    */
  final def reverse: ByteVector =
    ByteVector.viewAt((l: Long) => apply(size - l - 1), size)

  final def shiftLeft(n: Long): ByteVector =
    BitVector(this).shiftLeft(n).toByteVector

  final def shiftRight(n: Long, signExtension: Boolean): ByteVector =
    BitVector(this).shiftRight(n, signExtension).toByteVector

  final def rotateLeft(n: Long): ByteVector =
    BitVector(this).rotateLeft(n).toByteVector

  final def rotateRight(n: Long): ByteVector =
    BitVector(this).rotateRight(n).toByteVector

  /** Returns a vector with the same contents but represented as a single tree node internally.
    *
    * This may involve copying data, but has the advantage that lookups index directly into a single
    * node rather than traversing a logarithmic number of nodes in this tree.
    *
    * Calling this method on an already compacted vector is a no-op.
    *
    * @group collection
    */
  final def compact: ByteVector =
    this match {
      case Chunk(_) => this
      case _        => this.copy
    }

  /** Invokes `compact` on any subtrees whose size is `<= chunkSize`.
    * @group collection
    */
  final def partialCompact(chunkSize: Long): ByteVector =
    this match {
      case small if small.size <= chunkSize => small.compact
      case Append(l, r) => Append(l.partialCompact(chunkSize), r.partialCompact(chunkSize))
      case _            => this
    }

  /** Returns a vector with the same contents as this vector but with a single compacted node made
    * up by evaluating all internal nodes and concatenating their values.
    * @group collection
    */
  final def copy: ByteVector = {
    val sz = size
    if (sz <= Int.MaxValue) {
      val arr = this.toArray
      Chunk(View(new AtArray(arr), 0, sz))
    } else
      take(Int.MaxValue).copy ++ drop(Int.MaxValue).copy
  }

  /** Converts the contents of this vector to a byte array.
    *
    * @group conversions
    */
  final def toArray: Array[Byte] = {
    val buf = new Array[Byte](toIntSize(size))
    copyToArray(buf, 0)
    buf
  }

  /** Zero-copy version of [[toArray]]. In the case where this vector is a simple wrapper
    * around an underlying array, the array is returned directly instead of copying it.
    * In all other cases, a copy is returned.
    *
    * @group conversions
    */
  final def toArrayUnsafe: Array[Byte] = this match {
    case Chunk(v) => v.toArrayUnsafe
    case _        => toArray
  }

  /** Copies the contents of this vector to array `xs`, beginning at index `start`.
    *
    * @group conversions
    */
  final def copyToArray(xs: Array[Byte], start: Int): Unit = {
    var i = start
    foreachV { v =>
      v.copyToArray(xs, i); i += toIntSize(v.size)
    }
  }

  /** Copies `size` bytes of this vector, starting at index `offset`, to array `xs`, beginning at
    * index `start`.
    *
    * @group conversions
    */
  final def copyToArray(xs: Array[Byte], start: Int, offset: Long, size: Int): Unit = {
    var i = 0L
    var voffset = 0L
    foreachV { v =>
      if (i < size) {
        val reloff = (offset - voffset).max(0)
        if (voffset >= offset || reloff < v.size) {
          val sz = (size - i).min(v.size - reloff)
          v.copyToArray(xs, toIntSize(start + i), reloff, toIntSize(sz))
          i += sz
        }
        voffset += v.size
      }
    }
  }

  /** Copies as many bytes as possible to the given [[ByteBuffer]], starting from its current
    * position. This method will not overflow the buffer.
    *
    * @param buffer
    *   a ByteBuffer to copy bytes to
    * @return
    *   the number of bytes actually copied
    * @group conversions
    */
  final def copyToBuffer(buffer: ByteBuffer): Int = {
    var copied = 0
    foreachVPartial { v =>
      val copiedFromView = v.copyToBuffer(buffer)
      copied += copiedFromView
      copiedFromView == v.size
    }
    copied
  }

  /** Copies the contents of this vector to OutputStream `s`.
    *
    * @group conversions
    */
  final def copyToStream(s: OutputStream): Unit =
    foreachV(_.copyToStream(s))

  /** Creates new `InputStream` reading data from this `ByteVector`.
    */
  final def toInputStream: InputStream =
    new ByteVectorInputStream(this)

  /** Converts the contents of this vector to an `IndexedSeq`.
    *
    * @group conversions
    */
  final def toIndexedSeq: IndexedSeq[Byte] =
    new IndexedSeq[Byte] {
      val length = toIntSize(ByteVector.this.size)
      def apply(i: Int) = ByteVector.this.apply(i.toLong)
      override def foldRight[B](z: B)(op: (Byte, B) => B): B = ByteVector.this.foldRight(z)(op)
    }

  /** Converts the contents of this vector to a `Seq`.
    *
    * @group conversions
    */
  final def toSeq: Seq[Byte] = toIndexedSeq

  /** Converts the contents of this vector to an `Iterable`.
    *
    * @group conversions
    */
  final def toIterable: Iterable[Byte] = toIndexedSeq

  /** Converts the contents of this vector to a bit vector of `size * 8` bits.
    * @group conversions
    */
  final def toBitVector: BitVector = BitVector(this)

  /** Alias for [[toBitVector]].
    * @group conversions
    */
  final def bits: BitVector = toBitVector

  /** Allocate (unobservable) mutable scratch space at the end of this `ByteVector`, which will be
    * used to support fast `:+` and `++` of small vectors. A default chunk size is used.
    *
    * Note that `:+`, `++`, and `drop` on the result of a call to `buffer` are guaranteed to return
    * another buffered `ByteVector`.
    *
    * @group buffer
    */
  final def buffer: ByteVector = bufferBy(1024)

  /** Allocate (unobservable) mutable scratch space at the end of this `ByteVector`, with chunks of
    * the given size, which will be used to support fast `:+` and `++` of small vectors.
    *
    * Note that `:+`, `++`, and `drop` on the result of a call to `buffer` are guaranteed to return
    * another buffered `ByteVector`, with the same size scratch space.
    *
    * @group buffer
    */
  final def bufferBy(chunkSize: Int): ByteVector =
    this match {
      case b: Buffer =>
        if (b.lastChunk.length >= chunkSize) b
        else b.rebuffer(chunkSize)
      case _ => Buffer(new AtomicLong(0), 0, this, new Array[Byte](chunkSize), 0)
    }

  /** Collapse any buffered chunks at the end of this `ByteVector`, resulting in an unbuffered
    * `ByteVector`.
    *
    * @group buffer
    */
  def unbuffer: ByteVector = this

  /** Represents the contents of this vector as a read-only `java.nio.ByteBuffer`.
    *
    * The returned buffer is read-only with limit set to the minimum number of bytes needed to
    * represent the contents of this vector, position set to zero, and remaining set to the limit.
    *
    * @group conversions
    */
  final def toByteBuffer: ByteBuffer =
    this match {
      case Chunk(v) => v.asByteBuffer
      case _        => ByteBuffer.wrap(toArray).asReadOnlyBuffer()
    }

  /** Converts the contents of this byte vector to a binary string of `size * 8` digits.
    *
    * @group conversions
    */
  final def toBin: String = toBin(Bases.Alphabets.Binary)

  /** Converts the contents of this byte vector to a binary string of `size * 8` digits.
    *
    * @group conversions
    */
  final def toBin(alphabet: Bases.BinaryAlphabet): String = {
    val bldr = new StringBuilder
    foreachS {
      new F1BU {
        def apply(b: Byte) = {
          var n = 7
          while (n >= 0) {
            val idx = 1 & (b >> n)
            bldr.append(alphabet.toChar(idx))
            n -= 1
          }
        }
      }
    }
    bldr.toString
  }

  /** Converts the contents of this byte vector to a hexadecimal string of `size * 2` nibbles.
    *
    * @group conversions
    */
  final def toHex: String = toHex(Bases.Alphabets.HexLowercase)

  /** Converts the contents of this byte vector to a hexadecimal string of `size * 2` nibbles.
    *
    * @group conversions
    */
  final def toHex(alphabet: Bases.HexAlphabet): String = {
    val bldr = new StringBuilder
    foreachS {
      new F1BU {
        def apply(b: Byte) = {
          bldr
            .append(alphabet.toChar((b >> 4 & 0x0f).toByte.toInt))
            .append(alphabet.toChar((b & 0x0f).toByte.toInt))
          ()
        }
      }
    }
    bldr.toString
  }

  /** Generates a hex dump of this vector using the default format (with no color / ANSI escapes).
    * To customize the output, use the `HexDumpFormat` class instead.
    * For example, `HexDumpFormat.NoAscii.render(bytes)` or
    * `HexDumpFormat.Default.withIncludeAddressColumn(false).render(bytes)`.
    *
    * @group conversions
    */
  final def toHexDump: String = HexDumpFormat.NoAnsi.render(this)

  /** Colorized version of [[toHexDump]] that uses ANSI escape codes.
    *
    * @group conversions
    */
  final def toHexDumpColorized: String = HexDumpFormat.Default.render(this)

  /** Prints this vector as a colorized hex dump to standard out.
    */
  final def printHexDump(): Unit = HexDumpFormat.Default.print(this)

  /** Helper alias for [[toHex:String*]]
    *
    * @group conversions
    */
  final def toBase16: String = toHex

  /** Helper alias for [[toHex(alphabet:scodec\.bits\.Bases\.HexAlphabet):String*]]
    *
    * @group conversions
    */
  final def toBase16(alphabet: Bases.HexAlphabet): String = toHex(alphabet)

  /** Converts the contents of this vector to a base 32 string.
    *
    * @group conversions
    */
  final def toBase32: String = toBase32(Bases.Alphabets.Base32)

  /** Selects at most 8 bits from a byte array as a right aligned byte */
  private final def bitsAtOffset(bytes: Array[Byte], bitIndex: Long, length: Int): Int = {
    val i = (bitIndex / 8).toInt
    if (i >= bytes.length) 0
    else {
      val off = (bitIndex - (i * 8)).toInt
      val mask = ((1 << length) - 1) << (8 - length)
      val half = (bytes(i) << off) & mask
      val full =
        if (off + length <= 8 || i + 1 >= bytes.length) half
        else half | ((bytes(i + 1) & ((mask << (8 - off)) & 0xff)) >>> (8 - off))
      full >>> (8 - length)
    }
  }

  /** Converts the contents of this vector to a base 32 string using the specified alphabet.
    *
    * @group conversions
    */
  final def toBase32(alphabet: Bases.Base32Alphabet): String = {
    val bitsPerChar = 5
    val bytesPerGroup = 5
    val charsPerGroup = bytesPerGroup * 8 / bitsPerChar

    val bytes = toArray
    val bldr =
      CharBuffer.allocate((bytes.length + bytesPerGroup - 1) / bytesPerGroup * charsPerGroup)

    {
      var bidx: Long = 0
      while ((bidx / 8) < bytes.length) {
        val char = alphabet.toChar(bitsAtOffset(bytes, bidx, bitsPerChar))
        bldr.append(char)
        bidx += bitsPerChar
      }
    }

    if (alphabet.pad != 0.toChar) {
      val padLen =
        (((bytes.length + bitsPerChar - 1) / bitsPerChar * bitsPerChar) - bytes.length) * 8 / bitsPerChar
      var i = 0
      while (i < padLen) {
        bldr.append(alphabet.pad)
        i += 1
      }
    }

    bldr.flip.toString
  }

  /** Converts the contents of this vector to a base 58 string.
    *
    * @group conversions
    */
  final def toBase58: String = toBase58(Bases.Alphabets.Base58)

  /** Converts the contents of this vector to a base 58 string using the specified alphabet.
    *
    * @group conversions
    */
  final def toBase58(alphabet: Bases.Alphabet): String =
    if (isEmpty)
      ""
    else {
      val ZERO = BigInt(0)
      val RADIX = BigInt(58L)
      val ones = List.fill(takeWhile(_ == 0).length.toInt)('1')

      @tailrec
      def go(value: BigInt, chars: List[Char]): String =
        value match {
          case ZERO => (ones ++ chars).mkString
          case _ =>
            val (div, rem) = value /% RADIX
            go(div, alphabet.toChar(rem.toInt) +: chars)
        }
      go(BigInt(1, toArray), List.empty)
    }

  /** Converts the contents of this vector to a base 64 string.
    *
    * @group conversions
    */
  final def toBase64: String = toBase64(Bases.Alphabets.Base64)

  /** Converts the contents of this vector to a base 64 string using the specified alphabet.
    *
    * @group conversions
    */
  final def toBase64(alphabet: Bases.Base64Alphabet): String = {
    val bytes = toArray
    val bldr = CharBuffer.allocate(((bytes.length + 2) / 3) * 4)
    var idx = 0
    val mod = bytes.length % 3
    while (idx < bytes.length - mod) {
      var buffer =
        ((bytes(idx) & 0x0ff) << 16) | ((bytes(idx + 1) & 0x0ff) << 8) | (bytes(idx + 2) & 0x0ff)
      val fourth = buffer & 0x3f
      buffer = buffer >> 6
      val third = buffer & 0x3f
      buffer = buffer >> 6
      val second = buffer & 0x3f
      buffer = buffer >> 6
      val first = buffer
      bldr
        .append(alphabet.toChar(first))
        .append(alphabet.toChar(second))
        .append(alphabet.toChar(third))
        .append(alphabet.toChar(fourth))
      idx = idx + 3
    }
    if (mod == 1) {
      var buffer = (bytes(idx) & 0x0ff) << 4
      val second = buffer & 0x3f
      buffer = buffer >> 6
      val first = buffer
      bldr
        .append(alphabet.toChar(first))
        .append(alphabet.toChar(second))

      if (alphabet.pad != 0.toChar)
        bldr
          .append(alphabet.pad)
          .append(alphabet.pad)
    } else if (mod == 2) {
      var buffer = ((bytes(idx) & 0x0ff) << 10) | ((bytes(idx + 1) & 0x0ff) << 2)
      val third = buffer & 0x3f
      buffer = buffer >> 6
      val second = buffer & 0x3f
      buffer = buffer >> 6
      val first = buffer

      bldr
        .append(alphabet.toChar(first))
        .append(alphabet.toChar(second))
        .append(alphabet.toChar(third))

      if (alphabet.pad != 0.toChar) bldr.append(alphabet.pad)
    }
    bldr.flip.toString
  }

  /** Converts the contents of this vector to a base 64 string without padding.
    *
    * @group conversions
    */
  final def toBase64NoPad: String = toBase64(Bases.Alphabets.Base64NoPad)

  /** Converts the contents of this vector to a base 64 url string with padding.
    *
    * @group conversions
    */
  final def toBase64Url: String = toBase64(Bases.Alphabets.Base64Url)

  /** Converts the contents of this vector to a base 64 url string without padding.
    *
    * @group conversions
    */
  final def toBase64UrlNoPad: String = toBase64(Bases.Alphabets.Base64UrlNoPad)

  /** Converts the contents of this vector to a byte.
    *
    * @param signed
    *   whether sign extension should be performed
    * @throws IllegalArgumentException
    *   if size is greater than 8
    * @group conversions
    */
  final def toByte(signed: Boolean = true): Byte =
    bits.toByte(signed)

  /** Converts the contents of this vector to a short.
    *
    * @param signed
    *   whether sign extension should be performed
    * @param ordering
    *   order bytes should be processed in
    * @throws IllegalArgumentException
    *   if size is greater than 16
    * @group conversions
    */
  final def toShort(
      signed: Boolean = true,
      ordering: ByteOrdering = ByteOrdering.BigEndian
  ): Short =
    bits.toShort(signed, ordering)

  /** Converts the contents of this vector to an int.
    *
    * @param signed
    *   whether sign extension should be performed
    * @param ordering
    *   order bytes should be processed in
    * @throws IllegalArgumentException
    *   if size is greater than 32
    * @group conversions
    */
  final def toInt(signed: Boolean = true, ordering: ByteOrdering = ByteOrdering.BigEndian): Int =
    bits.toInt(signed, ordering)

  /** Converts the contents of this vector to an int.
    *
    * @param signed
    *   whether sign extension should be performed
    * @param ordering
    *   order bytes should be processed in
    * @throws IllegalArgumentException
    *   if size is greater than 64
    * @group conversions
    */
  final def toLong(signed: Boolean = true, ordering: ByteOrdering = ByteOrdering.BigEndian): Long =
    bits.toLong(signed, ordering)

  /** Converts the contents of this byte vector to a UUID.
    *
    * @throws IllegalArgumentException
    *   if size is not exactly 16.
    * @group conversions
    */
  final def toUUID: UUID = {
    // Sanity check
    if (size != 16)
      throw new IllegalArgumentException(
        s"Cannot convert ByteVector of size $size to UUID; must be 16 bytes"
      )
    // Convert
    val byteBuffer = toByteBuffer
    val mostSignificant = byteBuffer.getLong
    val leastSignificant = byteBuffer.getLong
    new UUID(mostSignificant, leastSignificant)
  }

  /** Decodes this vector as a string using the implicitly available charset.
    * @group conversions
    */
  final def decodeString(implicit charset: Charset): Either[CharacterCodingException, String] = {
    val decoder = charset.newDecoder
    try Right(decoder.decode(toByteBuffer).toString)
    catch {
      case e: CharacterCodingException => Left(e)
    }
  }

  /** Like [[decodeString]] but does not fail on bad input.
    * @group conversions
    */
  final def decodeStringLenient(
      replaceMalformedInput: Boolean = true,
      replaceUnmappableChars: Boolean = true,
      replacement: String = "�"
  )(implicit charset: Charset): String = {
    val decoder = charset.newDecoder
      .replaceWith(replacement)
      .onMalformedInput(
        if (replaceMalformedInput) CodingErrorAction.REPLACE else CodingErrorAction.IGNORE
      )
      .onUnmappableCharacter(
        if (replaceUnmappableChars) CodingErrorAction.REPLACE else CodingErrorAction.IGNORE
      )
    decoder.decode(toByteBuffer).toString
  }

  /** Decodes this vector as a string using the UTF-8 charset.
    * @group conversions
    */
  final def decodeUtf8: Either[CharacterCodingException, String] =
    decodeString(Charset.forName("UTF-8"))

  /** Like [[decodeUtf8]] but does not fail on bad input. */
  final def decodeUtf8Lenient: String =
    decodeStringLenient()(Charset.forName("UTF-8"))

  /** Decodes this vector as a string using the US-ASCII charset.
    * @group conversions
    */
  final def decodeAscii: Either[CharacterCodingException, String] =
    decodeString(Charset.forName("US-ASCII"))

  /** Like [[decodeAscii]] but does not fail on bad input. */
  final def decodeAsciiLenient: String =
    decodeStringLenient()(Charset.forName("US-ASCII"))

  final def not: ByteVector = mapS(new F1B { def apply(b: Byte) = (~b).toByte })

  final def or(other: ByteVector): ByteVector =
    zipWithS(other)(new F2B { def apply(b: Byte, b2: Byte) = (b | b2).toByte })

  final def and(other: ByteVector): ByteVector =
    zipWithS(other)(new F2B { def apply(b: Byte, b2: Byte) = (b & b2).toByte })

  final def xor(other: ByteVector): ByteVector =
    zipWithS(other)(new F2B { def apply(b: Byte, b2: Byte) = (b ^ b2).toByte })

  /** Returns a new vector where each byte is the result of evaluating the specified function
    * against the bytes of this vector and the specified vector at the corresponding index. The
    * resulting vector has size `this.size min other.size`. $returnsView
    * @group collection
    */
  final def zipWith(other: ByteVector)(f: (Byte, Byte) => Byte): ByteVector =
    zipWithS(other)(new F2B { def apply(b: Byte, b2: Byte) = f(b, b2) })

  /** See [[zipWith]] $returnsView
    * @group collection
    */
  final def zipWith2(other: ByteVector, other2: ByteVector)(
      f: (Byte, Byte, Byte) => Byte
  ): ByteVector =
    zipWithS(other, other2)(new F3B { def apply(b: Byte, b2: Byte, b3: Byte) = f(b, b2, b3) })

  /** See [[zipWith]] $returnsView
    * @group collection
    */
  final def zipWith3(other: ByteVector, other2: ByteVector, other3: ByteVector)(
      f: (Byte, Byte, Byte, Byte) => Byte
  ): ByteVector =
    zipWithS(other, other2, other3)(new F4B {
      def apply(b: Byte, b2: Byte, b3: Byte, b4: Byte) = f(b, b2, b3, b4)
    })

  private[scodec] final def zipWithS(other: ByteVector)(f: F2B): ByteVector = {
    val at = new At { def apply(i: Long) = f(ByteVector.this(i), other(i)) }
    Chunk(View(at, 0, size.min(other.size)))
  }

  private[scodec] final def zipWithS(other: ByteVector, other2: ByteVector)(f: F3B): ByteVector = {
    val at = new At { def apply(i: Long) = f(ByteVector.this(i), other(i), other2(i)) }
    Chunk(View(at, 0, size.min(other.size).min(other2.size)))
  }

  private[scodec] final def zipWithS(other: ByteVector, other2: ByteVector, other3: ByteVector)(
      f: F4B
  ): ByteVector = {
    val at = new At { def apply(i: Long) = f(ByteVector.this(i), other(i), other2(i), other3(i)) }
    Chunk(View(at, 0, size.min(other.size).min(other2.size).min(other3.size)))
  }

  /** Returns a new vector where each byte is the result of evaluating the specified function
    * against the bytes of this vector and the specified vector at the corresponding index. The
    * resulting vector has size `this.size min other.size`. Only the least significant byte is used
    * (the three most significant bytes are ignored). $returnsView
    * @group collection
    */
  final def zipWithI(other: ByteVector)(op: (Byte, Byte) => Int): ByteVector =
    zipWith(other) { case (l, r) => op(l, r).toByte }

  /** See [[zipWithI]] $returnsView
    * @group collection
    */
  final def zipWithI2(other: ByteVector, other2: ByteVector)(
      op: (Byte, Byte, Byte) => Int
  ): ByteVector =
    zipWith2(other, other2) { case (l, r1, r2) => op(l, r1, r2).toByte }

  /** See [[zipWithI]] $returnsView
    * @group collection
    */
  final def zipWithI3(other: ByteVector, other2: ByteVector, other3: ByteVector)(
      op: (Byte, Byte, Byte, Byte) => Int
  ): ByteVector =
    zipWith3(other, other2, other3) { case (l, r1, r2, r3) => op(l, r1, r2, r3).toByte }

  // implementation details, Object methods

  /** Calculates the hash code of this vector. The result is cached.
    * @group collection
    */
  override lazy val hashCode = {
    // todo: this could be recomputed more efficiently using the tree structure
    // given an associative hash function
    import util.hashing.MurmurHash3._
    val chunkSize = 1024 * 64L
    @annotation.tailrec
    def go(bytes: ByteVector, h: Int, iter: Int): Int =
      if (bytes.isEmpty) finalizeHash(h, iter)
      else go(bytes.drop(chunkSize), mix(h, bytesHash(bytes.take(chunkSize).toArray)), iter + 1)
    go(this, stringHash("ByteVector"), 1)
  }

  /** Returns true if the specified `ByteVector` has the same contents as this vector.
    * @group collection
    */
  final def ===(other: ByteVector): Boolean =
    if (this.eq(other)) true
    else {
      val s = this.size
      if (s != other.size)
        false
      else {
        @annotation.tailrec
        def go(i: Long): Boolean =
          if (i < s)
            if (this(i) == other(i)) go(i + 1)
            else false
          else
            true
        go(0)
      }
    }

  /** Returns true if the specified value is a `ByteVector` with the same contents as this vector.
    * @see
    *   [[ByteVector.===]]
    * @group collection
    */
  override def equals(other: Any) =
    other match {
      case that: ByteVector => this === that
      case _                => false
    }

  /** Display the size and bytes of this `ByteVector`. For bit vectors beyond a certain size, only a
    * hash of the contents are shown.
    * @group collection
    */
  override def toString: String =
    if (isEmpty) "ByteVector(empty)"
    else if (size < 512) s"ByteVector($size bytes, 0x$toHex)"
    else s"ByteVector($size bytes, #$hashCode)"

  private[bits] def pretty(prefix: String): String =
    this match {
      case Append(l, r) =>
        prefix + "bytes:append\n" +
          l.pretty(prefix + "  ") + "\n" +
          r.pretty(prefix + "  ")
      case Chunks(c) =>
        prefix + "bytes:chunks " + size + "\n" +
          c.left.pretty(prefix + "  ") + "\n" +
          c.right.pretty(prefix + "  ")
      case b: Buffer =>
        prefix + "bytes:buffer " + size + "\n" +
          b.hd.pretty(prefix + "  ") + "\n" +
          b.lastBytes.pretty(prefix + "  ")
      case Chunk(_) => prefix + (if (size < 16) "0x" + toHex else "#" + hashCode)
    }

  private def checkIndex(n: Long): Unit =
    if (n < 0 || n >= size)
      throw new IndexOutOfBoundsException(s"invalid index: $n for size $size")

  protected final def writeReplace(): AnyRef = new SerializationProxy(toArray)

  override def compare(that: ByteVector): Int =
    if (this.eq(that))
      0
    else {
      val thisLength = this.length
      val thatLength = that.length
      val commonLength = thisLength.min(thatLength)
      var i = 0
      while (i < commonLength) {
        val cmp = (this(i.toLong) & 0xff).compare(that(i.toLong) & 0xff)
        if (cmp != 0)
          return cmp
        i = i + 1
      }
      if (thisLength < thatLength)
        -1
      else if (thisLength > thatLength)
        1
      else
        0
    }
}

/** Companion for [[ByteVector]].
  *
  * @groupname constants
  *   Constants
  * @groupprio constants
  *   0
  *
  * @groupname constructors
  *   Constructors
  * @groupprio constructors
  *   1
  *
  * @groupname numeric
  *   Numeric Conversions
  * @groupprio numeric
  *   2
  *
  * @groupname base
  *   Base Conversions
  * @groupprio base
  *   3
  */
object ByteVector extends ByteVectorCompanionCrossPlatform {

  // various specialized function types

  private[scodec] abstract class F1B { def apply(b: Byte): Byte }
  private[scodec] abstract class F1BU { def apply(b: Byte): Unit }
  private[scodec] abstract class F1BB { def apply(b: Byte): Boolean }

  private[scodec] abstract class F2B { def apply(b: Byte, b2: Byte): Byte }
  private[scodec] abstract class F3B { def apply(b: Byte, b2: Byte, b3: Byte): Byte }
  private[scodec] abstract class F4B { def apply(b: Byte, b2: Byte, b3: Byte, b4: Byte): Byte }
  private[scodec] abstract class F5B {
    def apply(b: Byte, b2: Byte, b3: Byte, b4: Byte, b5: Byte): Byte
  }

  private[scodec] sealed abstract class At {
    def apply(i: Long): Byte
    def asByteBuffer(offset: Long, size: Int): ByteBuffer = {
      val arr = new Array[Byte](size)
      copyToArray(arr, 0, offset, size)
      ByteBuffer.wrap(arr).asReadOnlyBuffer()
    }
    def copyToArray(xs: Array[Byte], start: Int, offset: Long, size: Int): Unit = {
      var i = 0
      while (i < size) {
        xs(start + i) = apply(offset + i)
        i += 1
      }
    }
    def copyToBuffer(buffer: ByteBuffer, offset: Long, size: Int): Int = {
      var i = 0
      while (i < size && buffer.remaining > 0) {
        buffer.put(apply(offset + i))
        i += 1
      }
      i
    }
    def copyToStream(s: OutputStream, offset: Long, size: Long): Unit = {
      var i = 0
      while (i < size) {
        s.write(apply(offset + i).toInt)
        i += 1
      }
    }
  }

  private object AtEmpty extends At {
    def apply(i: Long) = throw new IllegalArgumentException("empty view")
    override def asByteBuffer(start: Long, size: Int): ByteBuffer =
      ByteBuffer.allocate(0).asReadOnlyBuffer()
    override def copyToBuffer(buffer: ByteBuffer, offset: Long, size: Int): Int = 0
  }

  private class AtArray(val arr: Array[Byte]) extends At {
    def apply(i: Long) = arr(i.toInt)

    override def asByteBuffer(start: Long, size: Int): ByteBuffer = {
      val b = ByteBuffer.wrap(arr, start.toInt, size).asReadOnlyBuffer()
      if (start == 0 && size == arr.length) b
      else b.slice()
    }

    override def copyToArray(xs: Array[Byte], start: Int, offset: Long, size: Int): Unit =
      System.arraycopy(arr, offset.toInt, xs, start, size)

    override def copyToStream(s: OutputStream, offset: Long, size: Long): Unit =
      s.write(arr, offset.toInt, toIntSize(size))

    override def copyToBuffer(buffer: ByteBuffer, offset: Long, size: Int): Int = {
      val toCopy = buffer.remaining.min(size)
      buffer.put(arr, offset.toInt, toCopy)
      toCopy
    }
  }

  private class AtByteBuffer(val buf: ByteBuffer) extends At {
    def apply(i: Long) = buf.get(i.toInt)

    override def copyToArray(xs: Array[Byte], start: Int, offset: Long, size: Int): Unit = {
      val n = buf.duplicate()
      n.position(offset.toInt)
      n.get(xs, start, size)
      ()
    }

    override def asByteBuffer(offset: Long, size: Int): ByteBuffer = {
      val b = buf.asReadOnlyBuffer()
      if (offset == 0 && b.position() == 0 && size == b.remaining()) b
      else {
        b.position(offset.toInt)
        b.limit(offset.toInt + size)
        b.slice()
      }
    }

    override def copyToBuffer(buffer: ByteBuffer, offset: Long, size: Int): Int = {
      val toCopy = buffer.remaining.min(size)
      buffer.put(asByteBuffer(offset, toCopy))
      toCopy
    }
  }

  private[bits] case class View(at: At, offset: Long, size: Long) {
    def apply(n: Long) = at(offset + n)
    def foreach(f: F1BU): Unit = {
      var i = 0L
      while (i < size) {
        f(at(offset + i))
        i += 1
      }
    }
    def foreachPartial(f: F1BB): Boolean = {
      var i = 0L
      var cont = true
      while (i < size && cont) {
        cont = f(at(offset + i))
        i += 1
      }
      cont
    }
    def asByteBuffer: ByteBuffer = at.asByteBuffer(offset, toIntSize(size))
    def copyToStream(s: OutputStream): Unit =
      at.copyToStream(s, offset, size)
    def copyToArray(xs: Array[Byte], start: Int): Unit =
      at.copyToArray(xs, start, offset, toIntSize(size))
    def copyToArray(xs: Array[Byte], start: Int, offset: Long, size: Int): Unit =
      at.copyToArray(xs, start, this.offset + offset, this.size.min(size.toLong).toInt)
    def toArray: Array[Byte] = {
      val arr = new Array[Byte](toIntSize(size))
      copyToArray(arr, 0)
      arr
    }
    def toArrayUnsafe: Array[Byte] =
      at match {
        case atarr: AtArray if offset == 0 && size == atarr.arr.size => atarr.arr
        case _                                                       => toArray
      }
    def copyToBuffer(buffer: ByteBuffer): Int =
      at.copyToBuffer(buffer, offset, toIntSize(size))
    def take(n: Long): View =
      if (n <= 0) View.empty
      else if (n >= size) this
      else View(at, offset, n)
    def drop(n: Long): View =
      if (n <= 0) this
      else if (n >= size) View.empty
      else View(at, offset + n, size - n)
  }

  private[bits] object View {
    val empty = View(AtEmpty, 0, 0)
  }

  private[bits] case class Chunk(bytes: View) extends ByteVector {
    def size = bytes.size
    def getImpl(i: Long) = bytes(i)
  }

  private[bits] case class Append(left: ByteVector, right: ByteVector) extends ByteVector {
    val size = left.size + right.size
    def getImpl(n: Long) =
      if (n < left.size) left.getImpl(n)
      else right.getImpl(n - left.size)
  }

  /** Empty byte vector.
    * @group constants
    */
  val empty: ByteVector = Chunk(View(AtEmpty, 0, 0))

  /** Constructs a `ByteVector` from a list of literal bytes. Only the least significant byte is
    * used of each integral value.
    * @group constructors
    */
  def apply[A: Integral](bytes: A*): ByteVector = {
    val integral = implicitly[Integral[A]]
    val buf = new Array[Byte](bytes.size)
    var i = 0
    bytes.foreach { b =>
      buf(i) = integral.toInt(b).toByte
      i += 1
    }
    view(buf)
  }

  /** Constructs a `ByteVector` from a collection of bytes.
    * @group constructors
    */
  def apply(bytes: Vector[Byte]): ByteVector = viewAt(idx => bytes(idx.toInt), bytes.size.toLong)

  /** Constructs a `ByteVector` from an `Array[Byte]`. The given `Array[Byte]` is copied to ensure
    * the resulting `ByteVector` is immutable. If this is not desired, use `ByteVector.view`.
    * @group constructors
    */
  def apply(bytes: Array[Byte]): ByteVector = {
    val copy: Array[Byte] = bytes.clone()
    view(copy)
  }

  /** Constructs a `ByteVector` from an `Array[Byte]`, an offset, and a length. The given
    * `Array[Byte]` is copied to ensure the resulting `ByteVector` is immutable. If this is not
    * desired, use `ByteVector.view`.
    * @group constructors
    */
  def apply(bytes: Array[Byte], offset: Int, length: Int): ByteVector = {
    val fresh: Array[Byte] = new Array[Byte](length)
    System.arraycopy(bytes, offset, fresh, 0, length)
    view(fresh)
  }

  /** Constructs a `ByteVector` from a `ByteBuffer`. The given `ByteBuffer` is is copied to ensure
    * the resulting `ByteVector` is immutable. If this is not desired, use `ByteVector.view`.
    *
    * The returned vector is a copy of a subsequence of the buffer, with bounds determined by the
    * buffer's position and limit at the time this method is called.
    *
    * @group constructors
    */
  def apply(buffer: ByteBuffer): ByteVector = {
    val c = buffer.duplicate()
    val arr = new Array[Byte](c.remaining)
    c.get(arr)
    view(arr)
  }

  /** Constructs a `ByteVector` from a `scala.collection` source of bytes.
    * @group constructors
    */
  def apply(bs: IterableOnce[Byte]): ByteVector =
    view(bs.iterator.toArray[Byte])

  /** Constructs a `ByteVector` from an `Array[Byte]`. Unlike `apply`, this does not make a copy of
    * the input array, so callers should take care not to modify the contents of the array passed to
    * this function.
    * @group constructors
    */
  def view(bytes: Array[Byte]): ByteVector =
    Chunk(View(new AtArray(bytes), 0, bytes.length.toLong))

  /** Constructs a `ByteVector` from a slice of an `Array[Byte]`. Unlike `apply`, this does not make
    * a copy of the input array, so callers should take care not to modify the contents of the array
    * passed to this function.
    * @group constructors
    */
  def view(bytes: Array[Byte], offset: Int, size: Int): ByteVector =
    Chunk(View(new AtArray(bytes), offset.toLong, size.toLong))

  /** Constructs a `ByteVector` from a `ByteBuffer`. Unlike `apply`, this does not make a copy of
    * the input buffer, so callers should take care not to modify the contents of the buffer passed
    * to this function.
    *
    * The returned vector is a view of a subsequence of the buffer, with bounds determined by the
    * buffer's position and limit at the time this method is called.
    *
    * @group constructors
    */
  def view(bytes: ByteBuffer): ByteVector = {
    val slice = bytes.slice()
    Chunk(View(new AtByteBuffer(slice), 0, slice.remaining.toLong))
  }

  /** Constructs a `ByteVector` from a function from `Long => Byte` and a size.
    * @group constructors
    */
  def viewAt(at: Long => Byte, size: Long): ByteVector =
    Chunk(View(new At { def apply(i: Long) = at(i) }, 0, size))

  /** Constructs a `ByteVector` from a function from `At` and a size.
    */
  private[scodec] def view(at: At, size: Long): ByteVector =
    Chunk(View(at, 0, size))

  /** Constructs a `ByteVector` from a function from `Long => Int` and a size, where the `Int`
    * returned by `at` must fit in a `Byte`.
    * @group constructors
    */
  def viewI(at: Long => Int, size: Long): ByteVector =
    Chunk(View(new At { def apply(i: Long) = at(i).toByte }, 0, size))

  /** Constructs a `ByteVector` of the given size, where all bytes have the value `b`.
    * @group constructors
    */
  def fill[A: Integral](size: Long)(b: A): ByteVector = {
    val integral = implicitly[Integral[A]]
    val value = integral.toInt(b).toByte
    Chunk(View(new At { def apply(i: Long) = value }, 0, size))
  }

  /** Constructs a `ByteVector` of the given size, where all bytes have the value `0`.
    * @group constructors
    */
  def low(size: Long): ByteVector = fill(size)(0)

  /** Constructs a `ByteVector` of the given size, where all bytes have the value `0xff`.
    * @group constructors
    */
  def high(size: Long): ByteVector = fill(size)(0xff)

  /** Constructs a `ByteVector` vector with the 2's complement encoding of the specified byte.
    * @param b
    *   value to encode
    * @group numeric
    */
  def fromByte(b: Byte): ByteVector =
    BitVector.fromByte(b, 8).bytes

  /** Constructs a `ByteVector` vector with the 2's complement encoding of the specified value.
    * @param s
    *   value to encode
    * @param size
    *   size of vector (<= 2)
    * @param ordering
    *   byte ordering of vector
    * @group numeric
    */
  def fromShort(
      s: Short,
      size: Int = 2,
      ordering: ByteOrdering = ByteOrdering.BigEndian
  ): ByteVector =
    BitVector.fromShort(s, size * 8, ordering).bytes

  /** Constructs a `ByteVector` with the 2's complement encoding of the specified value.
    * @param i
    *   value to encode
    * @param size
    *   size of vector (<= 4)
    * @param ordering
    *   byte ordering of vector
    * @group numeric
    */
  def fromInt(i: Int, size: Int = 4, ordering: ByteOrdering = ByteOrdering.BigEndian): ByteVector =
    BitVector.fromInt(i, size * 8, ordering).bytes

  /** Constructs a `ByteVector` with the 2's complement encoding of the specified value.
    * @param l
    *   value to encode
    * @param size
    *   size of vector (<= 8)
    * @param ordering
    *   byte ordering of vector
    * @group numeric
    */
  def fromLong(
      l: Long,
      size: Int = 8,
      ordering: ByteOrdering = ByteOrdering.BigEndian
  ): ByteVector =
    BitVector.fromLong(l, size * 8, ordering).bytes

  /** Constructs a `ByteVector` containing the binary representation of the specified UUID. The
    * bytes are in MSB-to-LSB order.
    *
    * @param u
    *   value to encode
    * @group conversions
    */
  final def fromUUID(u: UUID): ByteVector = {
    val buf = ByteBuffer.allocate(16)
    buf.putLong(u.getMostSignificantBits)
    buf.putLong(u.getLeastSignificantBits)
    // Go via Array[Byte] to avoid hanging on to intermediate ByteBuffer via AtByteBuffer.
    view(buf.array())
  }

  /** Constructs a `ByteVector` from a hexadecimal string or returns an error message if the string
    * is not valid hexadecimal.
    *
    * The string may start with a `0x` and it may contain whitespace or underscore characters.
    * @group base
    */
  def fromHexDescriptive(
      str: String,
      alphabet: Bases.HexAlphabet = Bases.Alphabets.HexLowercase
  ): Either[String, ByteVector] =
    fromHexInternal(str, alphabet).map { case (res, _) => res }

  private[bits] def fromHexInternal(
      str: String,
      alphabet: Bases.HexAlphabet
  ): Either[String, (ByteVector, Int)] = {
    val prefixed = (str.startsWith("0x")) || (str.startsWith("0X"))
    val withoutPrefix = if (prefixed) str.substring(2) else str
    var idx, hi, count = 0
    var midByte = false
    var err: String = null
    val bldr = ByteBuffer.allocate((str.size + 1) / 2)
    while (idx < withoutPrefix.length && (err eq null)) {
      val c = withoutPrefix(idx)
      if (!alphabet.ignore(c))
        try {
          val nibble = alphabet.toIndex(c)
          if (midByte) {
            bldr.put((hi | nibble).toByte)
            midByte = false
          } else {
            hi = (nibble << 4).toByte.toInt
            midByte = true
          }
          count += 1
        } catch {
          case _: IllegalArgumentException =>
            err = s"Invalid hexadecimal character '$c' at index ${idx + (if (prefixed) 2 else 0)}"
        }
      idx += 1
    }
    if (err eq null)
      Right(
        (
          if (midByte) {
            bldr.put(hi.toByte)
            bldr.flip()
            ByteVector(bldr).shiftRight(4, false)
          } else {
            bldr.flip()
            ByteVector(bldr)
          },
          count
        )
      )
    else Left(err)
  }

  /** Constructs a `ByteVector` from a hexadecimal string or returns `None` if the string is not
    * valid hexadecimal.
    *
    * The string may start with a `0x` and it may contain whitespace or underscore characters.
    * @group base
    */
  def fromHex(
      str: String,
      alphabet: Bases.HexAlphabet = Bases.Alphabets.HexLowercase
  ): Option[ByteVector] = fromHexDescriptive(str, alphabet).toOption

  /** Constructs a `ByteVector` from a hexadecimal string or throws an IllegalArgumentException if
    * the string is not valid hexadecimal.
    *
    * The string may start with a `0x` and it may contain whitespace or underscore characters.
    *
    * @throws IllegalArgumentException
    *   if the string is not valid hexadecimal
    * @group base
    */
  def fromValidHex(
      str: String,
      alphabet: Bases.HexAlphabet = Bases.Alphabets.HexLowercase
  ): ByteVector =
    fromHexDescriptive(str, alphabet).fold(msg => throw new IllegalArgumentException(msg), identity)

  /** Constructs a `ByteVector` from a binary string or returns an error message if the string is
    * not valid binary.
    *
    * The string may start with a `0b` and it may contain whitespace or underscore characters.
    * @group base
    */
  def fromBinDescriptive(
      str: String,
      alphabet: Bases.BinaryAlphabet = Bases.Alphabets.Binary
  ): Either[String, ByteVector] = fromBinInternal(str, alphabet).map { case (res, _) => res }

  private[bits] def fromBinInternal(
      str: String,
      alphabet: Bases.BinaryAlphabet = Bases.Alphabets.Binary
  ): Either[String, (ByteVector, Int)] = {
    val prefixed = (str.startsWith("0b")) || (str.startsWith("0B"))
    val withoutPrefix = if (prefixed) str.substring(2) else str
    var idx, byte, bits, count = 0
    var err: String = null
    val bldr = ByteBuffer.allocate((str.size + 7) / 8)
    while (idx < withoutPrefix.length && (err eq null)) {
      val c = withoutPrefix(idx)
      if (!alphabet.ignore(c))
        try {
          byte = (byte << 1) | (1 & alphabet.toIndex(c))
          bits += 1
          count += 1
        } catch {
          case _: IllegalArgumentException =>
            err = s"Invalid binary character '$c' at index ${idx + (if (prefixed) 2 else 0)}"
        }
      if (bits == 8) {
        bldr.put(byte.toByte)
        byte = 0
        bits = 0
      }
      idx += 1
    }
    if (err eq null)
      Right(
        (
          if (bits > 0) {
            bldr.put((byte << (8 - bits)).toByte)
            bldr.flip()
            ByteVector(bldr).shiftRight((8 - bits).toLong, false)
          } else {
            bldr.flip()
            ByteVector(bldr)
          },
          count
        )
      )
    else Left(err)
  }

  /** Constructs a `ByteVector` from a binary string or returns `None` if the string is not valid
    * binary.
    *
    * The string may start with a `0b` and it may contain whitespace or underscore characters.
    * @group base
    */
  def fromBin(
      str: String,
      alphabet: Bases.BinaryAlphabet = Bases.Alphabets.Binary
  ): Option[ByteVector] = fromBinDescriptive(str, alphabet).toOption

  /** Constructs a `ByteVector` from a binary string or throws an IllegalArgumentException if the
    * string is not valid binary.
    *
    * The string may start with a `0b` and it may contain whitespace or underscore characters.
    *
    * @throws IllegalArgumentException
    *   if the string is not valid binary
    * @group base
    */
  def fromValidBin(
      str: String,
      alphabet: Bases.BinaryAlphabet = Bases.Alphabets.Binary
  ): ByteVector =
    fromBinDescriptive(str, alphabet).fold(msg => throw new IllegalArgumentException(msg), identity)

  /** Constructs a `ByteVector` from a base 32 string or returns an error message if the string is
    * not valid base 32. An empty input string results in an empty ByteVector. The string may
    * contain whitespace characters and hyphens which are ignored.
    * @group base
    */
  def fromBase32Descriptive(
      str: String,
      alphabet: Bases.Base32Alphabet = Bases.Alphabets.Base32
  ): Either[String, ByteVector] = {
    val bitsPerChar = 5
    val bytesPerGroup = 5
    val charsPerGroup = bytesPerGroup * 8 / bitsPerChar

    val Pad = alphabet.pad
    var idx, bidx, buffer, padding = 0
    val acc = ByteBuffer.allocate((str.length + charsPerGroup - 1) / charsPerGroup * bytesPerGroup)
    while (idx < str.length) {
      val c = str(idx)

      if (Pad != 0.toChar && c == Pad)
        padding += 1
      else if (!alphabet.ignore(c)) {
        if (padding > 0)
          return Left(
            s"Unexpected character '$c' at index $idx after padding character; only '=' and whitespace characters allowed after first padding character"
          )

        val index =
          try alphabet.toIndex(c)
          catch {
            case _: IllegalArgumentException =>
              return Left(s"Invalid base 32 character '$c' at index $idx")
          }

        buffer |= (index << (8 - bitsPerChar) >>> bidx) & 0xff
        bidx += bitsPerChar

        if (bidx >= 8) {
          bidx -= 8
          acc.put(buffer.toByte)
          buffer = (index << (8 - bidx)) & 0xff
        }
      }

      idx += 1
    }

    if (bidx >= bitsPerChar)
      acc.put(buffer.toByte)

    acc.flip()
    val bytes = ByteVector.view(acc)

    val expectedPadding =
      (((bytes.length + bitsPerChar - 1) / bitsPerChar * bitsPerChar) - bytes.length) * 8 / bitsPerChar
    if (padding != 0 && padding != expectedPadding)
      return Left(
        s"Malformed padding - optionally expected $expectedPadding padding characters such that the quantum is completed"
      )

    Right(bytes)
  }

  /** Constructs a `ByteVector` from a base 64 string or returns `None` if the string is not valid
    * base 32. Details pertaining to base 32 decoding can be found in the comment for
    * fromBase32Descriptive. The string may contain whitespace characters which are ignored.
    * @group base
    */
  def fromBase32(
      str: String,
      alphabet: Bases.Base32Alphabet = Bases.Alphabets.Base32
  ): Option[ByteVector] = fromBase32Descriptive(str, alphabet).toOption

  /** Constructs a `ByteVector` from a base 32 string or throws an IllegalArgumentException if the
    * string is not valid base 32. Details pertaining to base 32 decoding can be found in the
    * comment for fromBase32Descriptive. The string may contain whitespace characters which are
    * ignored.
    *
    * @throws IllegalArgumentException
    *   if the string is not valid base 32
    * @group base
    */
  def fromValidBase32(
      str: String,
      alphabet: Bases.Base32Alphabet = Bases.Alphabets.Base32
  ): ByteVector =
    fromBase32Descriptive(str, alphabet)
      .fold(msg => throw new IllegalArgumentException(msg), identity)

  /** Constructs a `ByteVector` from a base 58 string or returns an error message if the string is
    * not valid base 58. It is similar to Base64 but has been modified to avoid both
    * non-alphanumeric characters and letters which might look ambiguous when printed. It is
    * therefore designed for human users who manually enter the data, copying from some visual
    * source Compared to Base64, the following similar-looking letters are omitted: 0 (zero), O
    * (capital o), I (capital i) and l (lower case L) as well as the non-alphanumeric characters +
    * (plus) and / (slash). The actual order of letters in the alphabet depends on the application,
    * the default order is the same used in Bitcoin An empty input string results in an empty
    * ByteVector. The string may contain whitespace characters which are ignored.
    * @group base
    */
  def fromBase58Descriptive(
      str: String,
      alphabet: Bases.Alphabet = Bases.Alphabets.Base58
  ): Either[String, ByteVector] = {
    val zeroLength = str.takeWhile(_ == '1').length
    val zeroes = ByteVector.fill(zeroLength.toLong)(0)
    val trim = str.splitAt(zeroLength)._2.toList
    val RADIX = BigInt(58L)
    try {
      val decoded = trim.foldLeft(BigInt(0)) { (a, c) =>
        try a * RADIX + BigInt(alphabet.toIndex(c))
        catch {
          case _: IllegalArgumentException =>
            val idx = trim.takeWhile(_ != c).length
            throw new IllegalArgumentException(s"Invalid base 58 character '$c' at index $idx")
        }
      }
      if (trim.isEmpty) Right(zeroes)
      else
        Right(
          zeroes ++ ByteVector(decoded.toByteArray.dropWhile(_ == 0))
        ) // drop because toByteArray sometimes prepends a zero
    } catch {
      case e: IllegalArgumentException => Left(e.getMessage)
    }
  }

  /** Constructs a `ByteVector` from a base 58 string or returns `None` if the string is not valid
    * base 58. Details pertaining to base 58 decoding can be found in the comment for
    * fromBase58Descriptive. The string may contain whitespace characters which are ignored.
    * @group base
    */
  def fromBase58(
      str: String,
      alphabet: Bases.Alphabet = Bases.Alphabets.Base58
  ): Option[ByteVector] = fromBase58Descriptive(str, alphabet).toOption

  /** Constructs a `ByteVector` from a base 58 string or throws an IllegalArgumentException if the
    * string is not valid base 58. Details pertaining to base 58 decoding can be found in the
    * comment for fromBase58Descriptive. The string may contain whitespace characters which are
    * ignored.
    *
    * @throws IllegalArgumentException
    *   if the string is not valid base 58
    * @group base
    */
  def fromValidBase58(str: String, alphabet: Bases.Alphabet = Bases.Alphabets.Base58): ByteVector =
    fromBase58Descriptive(str, alphabet)
      .fold(msg => throw new IllegalArgumentException(msg), identity)

  private val Base64PaddingError = Left(
    "Malformed padding - final quantum may optionally be padded with one or two padding characters such that the quantum is completed"
  )

  /** Constructs a `ByteVector` from a base 64 string or returns an error message if the string is
    * not valid base 64. If the final encoding quantum does not contain 4 characters, i.e. the total
    * number of characters is not evenly divisible by 4, padding is inferred if the final quantum
    * contains 2 or 3 characters. This is to say that padding is optional as long as the inferred
    * padding would yield a valid base 64 string. The input is considered invalid if the final
    * quantum only contains a single character. If padding characters are present, they must be used
    * in accordance with the base 64 specification and no padding characters will be inferred. An
    * empty input string results in an empty ByteVector. The string may contain whitespace
    * characters which are ignored.
    * @group base
    */
  def fromBase64Descriptive(
      str: String,
      alphabet: Bases.Base64Alphabet = Bases.Alphabets.Base64
  ): Either[String, ByteVector] = {
    val Pad = alphabet.pad
    var idx, bidx, buffer, mod, padding = 0
    val acc = new Array[Byte]((str.size + 3) / 4 * 3)
    while (idx < str.length) {
      str(idx) match {
        case c if alphabet.ignore(c) => // ignore
        case c =>
          val cidx =
            if (padding == 0)
              if (c == Pad)
                if (mod == 2 || mod == 3) {
                  padding += 1
                  0
                } else
                  return Base64PaddingError
              else
                try alphabet.toIndex(c)
                catch {
                  case _: IllegalArgumentException =>
                    return Left(s"Invalid base 64 character '$c' at index $idx")
                }
            else if (c == Pad)
              if (padding == 1 && mod == 3) {
                padding += 1
                0
              } else
                return Base64PaddingError
            else
              return Left(
                s"Unexpected character '$c' at index $idx after padding character; only '=' and whitespace characters allowed after first padding character"
              )
          mod match {
            case 0 =>
              buffer = cidx & 0x3f
              mod += 1
            case 1 =>
              buffer = (buffer << 6) | (cidx & 0x3f)
              mod += 1
            case 2 =>
              buffer = (buffer << 6) | (cidx & 0x3f)
              mod += 1
            case 3 =>
              buffer = (buffer << 6) | (cidx & 0x3f)
              mod = 0
              val c = buffer & 0x0ff
              val b = (buffer >> 8) & 0x0ff
              val a = (buffer >> 16) & 0x0ff
              acc(bidx) = a.toByte
              acc(bidx + 1) = b.toByte
              acc(bidx + 2) = c.toByte
              bidx += 3
          }
      }
      idx += 1
    }
    if (padding != 0 && mod != 0) Base64PaddingError
    else
      mod match {
        case 0 => Right(ByteVector(acc).take((bidx - padding).toLong))
        case 1 => Left("Final base 64 quantum had only 1 digit - must have at least 2 digits")
        case 2 =>
          acc(bidx) = ((buffer >> 4) & 0x0ff).toByte
          bidx += 1
          Right(ByteVector(acc).take(bidx.toLong))
        case 3 =>
          acc(bidx) = ((buffer >> 10) & 0x0ff).toByte
          acc(bidx + 1) = ((buffer >> 2) & 0x0ff).toByte
          bidx += 2
          Right(ByteVector(acc).take(bidx.toLong))
      }
  }

  /** Constructs a `ByteVector` from a base 64 string or returns `None` if the string is not valid
    * base 64. Details pertaining to base 64 decoding can be found in the comment for
    * fromBase64Descriptive. The string may contain whitespace characters which are ignored.
    * @group base
    */
  def fromBase64(
      str: String,
      alphabet: Bases.Base64Alphabet = Bases.Alphabets.Base64
  ): Option[ByteVector] = fromBase64Descriptive(str, alphabet).toOption

  /** Constructs a `ByteVector` from a base 64 string or throws an IllegalArgumentException if the
    * string is not valid base 64. Details pertaining to base 64 decoding can be found in the
    * comment for fromBase64Descriptive. The string may contain whitespace characters which are
    * ignored.
    *
    * @throws IllegalArgumentException
    *   if the string is not valid base 64
    * @group base
    */
  def fromValidBase64(
      str: String,
      alphabet: Bases.Base64Alphabet = Bases.Alphabets.Base64
  ): ByteVector =
    fromBase64Descriptive(str, alphabet)
      .fold(msg => throw new IllegalArgumentException(msg), identity)

  /** Encodes the specified string to a `ByteVector` using the implicitly available `Charset`.
    *
    * @group constructors
    */
  def encodeString(
      str: String
  )(implicit charset: Charset): Either[CharacterCodingException, ByteVector] = {
    val encoder = charset.newEncoder
    val buffer = CharBuffer.wrap(str)
    try Right(ByteVector(encoder.encode(buffer)))
    catch {
      case e: CharacterCodingException => Left(e)
    }
  }

  /** Encodes the specified string to a `ByteVector` using the UTF-8 charset.
    *
    * @group constructors
    */
  def encodeUtf8(str: String): Either[CharacterCodingException, ByteVector] =
    encodeString(str)(Charset.forName("UTF-8"))

  /** Encodes the specified string to a `ByteVector` using the US-ASCII charset.
    *
    * @group constructors
    */
  def encodeAscii(str: String): Either[CharacterCodingException, ByteVector] =
    encodeString(str)(Charset.forName("US-ASCII"))

  /** Concatenates all the given `ByteVector`s into a single instance.
    *
    * @group constructors
    */
  def concat(bvs: IterableOnce[ByteVector]): ByteVector =
    bvs.iterator.foldLeft(ByteVector.empty)(_ ++ _).unbuffer

  @SerialVersionUID(1L)
  private class SerializationProxy(private val bytes: Array[Byte]) extends Serializable {
    def readResolve: AnyRef = ByteVector.view(bytes)
  }

  private[bits] case class Chunks(chunks: Append) extends ByteVector {
    def size = chunks.size
    def getImpl(n: Long) = chunks.getImpl(n)

    override def ++(b: ByteVector): ByteVector =
      if (b.isEmpty) this
      else if (this.isEmpty) b
      else {
        @annotation.tailrec
        def go(chunks: Append, last: ByteVector): ByteVector = {
          val lastN = last.size
          if (lastN >= chunks.size || lastN * 2 <= chunks.right.size)
            Chunks(Append(chunks, last))
          else
            chunks.left match {
              case left: Append => go(left, Append(chunks.right, last))
              case _            => Chunks(Append(chunks, last))
            }
        }
        go(chunks, b.unbuffer) // unbuffering `b` avoids proliferation of unused scratch space
        // if `b` happens to have been built up using a buffer
      }
  }

  /*
   * `Buffer` is a `ByteVector` supporting fast `:+` and `++` operations, using
   * an (unobservable) mutable scratch `Array[Byte]` at the end of the vector.
   * The scratch space is written in an append-only fashion, which lets us
   * safely share the same `Array` between multiple `ByteVector` instances,
   * so long as we mutably append to this scratch space _at most once_ from a
   * given `ByteVector`.
   *
   * That is, given {{{
   *    val b = ByteVector.empty.buffer
   *    val b2 = b ++ foo
   *    val b3 = b ++ bar
   *    val b22 = b2 ++ baz
   *  }}}
   *
   * we obviously cannot have both `b2` and `b3` mutably append to the same scratch
   * space, as `b3` would then overwrite what was written by `b2`. The solution
   * adopted here (see [1]) is to let `b2` write mutable to the buffer, but `b3`,
   * evaluated later, must make a copy.
   *
   * [1]: http://www.serpentine.com/blog/2014/05/31/attoparsec/
   */
  private class Buffer(
      val id: AtomicLong,
      val stamp: Long,
      val hd: ByteVector,
      val lastChunk: Array[Byte],
      val lastSize: Int
  ) extends ByteVector {

    def size = hd.size + lastSize

    override def take(n: Long): ByteVector =
      if (n <= hd.size) hd.take(n)
      else hd ++ lastBytes.take(n - hd.size)

    override def drop(n: Long): ByteVector =
      if (n <= hd.size) Buffer(id, stamp, hd.drop(n), lastChunk, lastSize)
      else unbuffer.drop(n).bufferBy(lastChunk.length)

    def getImpl(n: Long): Byte =
      if (n < hd.size) hd.getImpl(n)
      else lastChunk((n - hd.size).toInt)

    override def :+(b: Byte): ByteVector =
      // threads race to increment id, winner gets to update tl mutably
      if (id.compareAndSet(stamp, stamp + 1) && lastSize < lastChunk.length) {
        lastChunk(lastSize) = b
        Buffer(id, stamp + 1, hd, lastChunk, lastSize + 1)
      } else // loser has to make a copy of the scratch space
        Buffer(new AtomicLong(0), 0, unbuffer, new Array[Byte](lastChunk.length), 0) :+ b

    // NB: if `bs` fits in scratch space and is itself a buffer, it will be
    // unbuffered before being added. This avoids proliferation of scratch space
    // for tiny ByteVectors
    override def ++(bs: ByteVector): ByteVector =
      if (bs.isEmpty) this
      else if (isEmpty) bs
      else
      // threads race to increment id, winner gets to update tl mutably
      if (id.compareAndSet(stamp, stamp + 1) && (lastChunk.length - lastSize > bs.size)) {
        bs.copyToArray(lastChunk, lastSize)
        Buffer(id, stamp + 1, hd, lastChunk, lastSize + bs.size.toInt)
      } else if (lastSize == 0) // just append directly to `hd`
        Buffer(id, stamp, (hd ++ bs).unbuffer, lastChunk, lastSize)
      else // loser has to make a copy of the scratch space
        Buffer(new AtomicLong(0), 0, unbuffer, new Array[Byte](lastChunk.length), 0) ++ bs

    def lastBytes = ByteVector.view(lastChunk).take(lastSize.toLong)

    override def unbuffer: ByteVector =
      // if last chunk more than half unused, copy to a fresh
      // bytevector, to avoid proliferation of scratch space
      hd ++ (if (lastSize * 2 < lastChunk.length) lastBytes.copy else lastBytes)

    def rebuffer(chunkSize: Int): ByteVector = {
      require(chunkSize > lastChunk.length)
      val lastChunk2 = new Array[Byte](chunkSize)
      lastChunk.copyToArray(lastChunk2, 0, lastChunk2.length)
      Buffer(new AtomicLong(0), 0, hd, lastChunk, lastSize)
    }
  }

  private object Buffer {
    def apply(
        id: AtomicLong,
        stamp: Long,
        hd: ByteVector,
        lastChunk: Array[Byte],
        lastSize: Int
    ): Buffer =
      new Buffer(id, stamp, hd.unbuffer, lastChunk, lastSize)
  }

  private[bits] def toIntSize(size: Long): Int =
    if (size <= Int.MaxValue) size.toInt
    else throw new IllegalArgumentException(s"size must be <= Int.MaxValue but is $size")

  /** Extractor used in support of pattern matching on the bytes of a vector.
    *
    * @group constructors
    */
  def unapplySeq(b: ByteVector): Some[Seq[Byte]] = Some(b.toIndexedSeq)

  implicit class GroupedOp(val self: ByteVector) extends AnyVal {

    /** Converts this vector in to a sequence of `chunkSize`-byte vectors.
      * @group collection
      */
    final def grouped(chunkSize: Long): Iterator[ByteVector] = self.groupedIterator(chunkSize)
  }

  private class ByteVectorInputStream(bv: ByteVector) extends InputStream {
    private val bvlen = bv.size.toInt
    private val pos = new CustomAtomicInteger(0)

    override def read(): Int = {
      val p = pos.getAndIncrement()
      if (p >= bvlen) -1 else bv.get(p.toLong) & 0xff
    }

    override def read(b: Array[Byte], off: Int, len: Int): Int = {
      var l: Int = -1

      val cpos: Int = pos.getAndUpdate_ { cpos =>
        l = Math.min(len, bvlen - cpos)
        cpos + l
      }

      if (cpos >= bvlen) return -1

      bv.copyToArray(b, off, cpos.toLong, l)

      l
    }

    override def available(): Int = Math.max(bvlen - pos.get(), 0)

    /* This is quite a hack. The problem is that the code has to be Scala-agnostic but Scala.js implementation of `AtomicInteger` differs
     * from the JVM standard one and it doesn't contain `getAndUpdate` method. So I took the method and put here the code with a different name
     * which adds the method on all platforms and everybody is happy. */
    private class CustomAtomicInteger(v: Int) extends AtomicInteger(v) {
      def getAndUpdate_(updateFunction: Int => Int): Int = {
        var prev = get
        var next = updateFunction(prev)
        while (!compareAndSet(prev, next)) {
          prev = get
          next = updateFunction(prev)
        }
        prev
      }
    }
  }
}