Variable-length integer used in Apple `'dcmp' (0)` and `'dcmp' (1)` compressed resource formats: Nim parsing library

A variable-length integer, in the format used by the 0xfe chunks in the 'dcmp' (0) and 'dcmp' (1) resource compression formats. See the dcmp_0 and dcmp_1 specs for more information about these compression formats.

This variable-length integer format can store an integer x in any of the following ways:

  • In a single byte, if 0 <= x <= 0x7f (7-bit unsigned integer)
  • In 2 bytes, if -0x4000 <= x <= 0x3eff (15-bit signed integer with the highest 0x100 values unavailable)
  • In 5 bytes, if -0x80000000 <= x <= 0x7fffffff (32-bit signed integer)

In practice, values are always stored in the smallest possible format, but technically any of the larger formats could be used as well.

Application

Mac OS

KS implementation details

License: MIT
Minimal Kaitai Struct required: 0.8

This page hosts a formal specification of Variable-length integer used in Apple `'dcmp' (0)` and `'dcmp' (1)` compressed resource formats using Kaitai Struct. This specification can be automatically translated into a variety of programming languages to get a parsing library.

Nim source code to parse Variable-length integer used in Apple `'dcmp' (0)` and `'dcmp' (1)` compressed resource formats

dcmp_variable_length_integer.nim

import kaitai_struct_nim_runtime
import options

type
  DcmpVariableLengthInteger* = ref object of KaitaiStruct
    `first`*: uint8
    `more`*: int32
    `parent`*: KaitaiStruct
    `valueInst`: int
    `valueInstFlag`: bool

proc read*(_: typedesc[DcmpVariableLengthInteger], io: KaitaiStream, root: KaitaiStruct, parent: KaitaiStruct): DcmpVariableLengthInteger

proc value*(this: DcmpVariableLengthInteger): int


##[
A variable-length integer,
in the format used by the 0xfe chunks in the `'dcmp' (0)` and `'dcmp' (1)` resource compression formats.
See the dcmp_0 and dcmp_1 specs for more information about these compression formats.

This variable-length integer format can store an integer `x` in any of the following ways:

* In a single byte,
  if `0 <= x <= 0x7f`
  (7-bit unsigned integer)
* In 2 bytes,
  if `-0x4000 <= x <= 0x3eff`
  (15-bit signed integer with the highest `0x100` values unavailable)
* In 5 bytes, if `-0x80000000 <= x <= 0x7fffffff`
  (32-bit signed integer)

In practice,
values are always stored in the smallest possible format,
but technically any of the larger formats could be used as well.

@see <a href="https://github.com/dgelessus/python-rsrcfork/blob/f891a6e/src/rsrcfork/compress/common.py">Source</a>
]##
proc read*(_: typedesc[DcmpVariableLengthInteger], io: KaitaiStream, root: KaitaiStruct, parent: KaitaiStruct): DcmpVariableLengthInteger =
  template this: untyped = result
  this = new(DcmpVariableLengthInteger)
  let root = if root == nil: cast[DcmpVariableLengthInteger](this) else: cast[DcmpVariableLengthInteger](root)
  this.io = io
  this.root = root
  this.parent = parent


  ##[
  The first byte of the variable-length integer.
This determines which storage format is used.

* For the 1-byte format,
  this encodes the entire value of the value.
* For the 2-byte format,
  this encodes the high 7 bits of the value,
  minus `0xc0`.
  The highest bit of the value,
  i. e. the second-highest bit of this field,
  is the sign bit.
* For the 5-byte format,
  this is always `0xff`.

  ]##
  let firstExpr = this.io.readU1()
  this.first = firstExpr

  ##[
  The remaining bytes of the variable-length integer.

* For the 1-byte format,
  this is not present.
* For the 2-byte format,
  this encodes the low 8 bits of the value.
* For the 5-byte format,
  this encodes the entire value.

  ]##
  if this.first >= 128:
    block:
      let on = this.first
      if on == 255:
        let moreExpr = this.io.readS4be()
        this.more = moreExpr
      else:
        let moreExpr = int32(this.io.readU1())
        this.more = moreExpr

proc value(this: DcmpVariableLengthInteger): int = 

  ##[
  The decoded value of the variable-length integer.

  ]##
  if this.valueInstFlag:
    return this.valueInst
  let valueInstExpr = int((if this.first == 255: this.more else: (if this.first >= 128: (((this.first shl 8) or this.more) - 49152) else: this.first)))
  this.valueInst = valueInstExpr
  this.valueInstFlag = true
  return this.valueInst

proc fromFile*(_: typedesc[DcmpVariableLengthInteger], filename: string): DcmpVariableLengthInteger =
  DcmpVariableLengthInteger.read(newKaitaiFileStream(filename), nil, nil)