ChangeLog.md

1.0.1 / 2024-01-23

• Switched to using require_relative to improve load-times.
• Correctly alias :long/:ulong to :int32/:uint32.
• Disable ANSI styling in {Hexdump#hexdump} if the output is not a TTY.
• Removed the old hexdump/extensions file in favor of hexdump/core_ext.

1.0.0 / 2021-08-31

• Added the ability to hexdump typed data. Available type:s are :
  • :char - signed 8bit ASCII character
  • :uchar - unsigned 8bit ASCII character
  • :int8 - signed 8bit integer
  • :uin8 / :byte - unsigned 8bit integer (default)
  • :int16 / :short - signed 16bit integer
  • :int16_le / :short_le - signed 16bit integer (little endian)
  • :int16_be / :short_be - signed 16bit integer (big endian)
  • :int16_ne / :short_ne - signed 16bit integer (network endian)
  • :uint16 / :ushort - unsigned 16bit integer
  • :uint16_le / :ushort_le - unsigned 16bit integer (little endian)
  • :uint16_be / :ushort_be - unsigned 16bit integer (big endian)
  • :uint16_ne / :ushort_ne - unsigned 16bit integer (network endian)
  • :int32 / :int / :long - signed 32bit integer
  • :int32_le / :int_le / :long_le - signed 32bit integer (little endian)
  • :int32_be / :int_be / :long_be - signed 32bit integer (big endian)
  • :int32_ne / :int_ne / :long_ne - signed 32bit integer (network endian)
  • :uint32 / :uint / :ulong - unsigned 32bit integer
  • :uint32_le / :uint_le / :ulong_le - unsigned 32bit integer (little endian)
  • :uint32_be / :uint_be / :ulong_be - unsigned 32bit integer (big endian)
  • :uint32_ne / :uint_ne / :ulong_ne - unsigned 32bit integer (network endian)
  • :int64 / :long_long - signed 64bit integer
  • :int64_le / :long_long_le - signed 64bit integer (little endian)
  • :int64_be / :long_long_be - signed 64bit integer (big endian)
  • :int64_ne / :long_long_ne - signed 64bit integer (network endian)
  • :uint64 / :ulong_long - unsigned 64bit integer
  • :uint64_le / :ulong_long_le - unsigned 64bit integer (little endian)
  • :uint64_be / :ulong_long_be - unsigned 64bit integer (big endian)
  • :uint64_ne / ulong_long_ne - unsigned 64bit integer (network endian)
  • :float - single precision 32bit floating-point number
  • :float_le - single precision 32bit floating-point number (little endian)
  • :float_be - single precision 32bit floating-point number (big endian)
  • :float_ne - single precision 32bit floating-point number (network endian)
  • :double - double precision 64bit floating-point number
  • :double_le - double precision 64bit floating-point number (little endian)
  • :double_be - double precision 64bit floating-point number (big endian)
  • :double_ne - double precision 64bit floating-point number (network endian)
• Added support for optionally skipping N bytes or reading at most N bytes of data (ex: offset: 10 and length: 100).
• Added support for optional zero-padding of the data so it aligns with the type size (ex: zero_pad: true).
• Added support for displaying the index in a separate numeric base than the numeric columns (ex: index_base: 10, base: 2). Defaults to base 16.
• Added support for omitting repeating rows with a * (ex: repeating: true or repeating: false to disable).
• Added support for grouping columns together like GNU hexdump -C (ex: group_columns: 4).
• Added support for grouping characters together (ex: group_chars: 4).
• Added support for grouping characters together to align with the type's size (ex: group_chars: :type).
• Added support for displaying UTF-8 characters or other character encodings.
• Added support for disabling the characters column entirely (ex: chars_column: false).
• Added support for ANSI styling and highlighting (ex: style: {...} and highlights: {...}).
• Added {Hexdump::ModuleMethods} to provide a top-level {Hexdump::ModuleMethods#hexdump hexdump} method.
  • Include {Hexdump::ModuleMethods} into {Kernel} by default.
• Added preliminary truffleruby support.
  • Only issue was oracle/truffleruby#2426
• hexdump methods now accept a block for configuring the hexdump instance.
• Renamed Hexdump::Dumper to {Hexdump::Hexdump}.
• Renamed the :width option to :columns.

0.3.0 / 2021-04-10

• Require Ruby >= 2.0.0.
• Added support for Ruby 3.0.
• {Hexdump.dump}, {Hexdump#dump}, and Hexdump::Dumper#initialize now accept keyword arguments.
• Hexdump::Dumper#each now returns the total number of bytes read.
• Hexdump::Dumper#dump now prints the final number of bytes read on the last line.
• Micro-optimizations to improve performance on JRuby and TruffleRuby.

0.2.4 / 2020-09-26

• Default the :output option of {Hexdump.dump} to $stdout instead of STDOUT, to support overriding the default stdout stream (@kylekyle).

0.2.3 / 2012-05-28

• Fixed a typo in the gemspec, which incorrectly set required_rubygems_version to the same value as required_ruby_version.

0.2.2 / 2012-05-27

• Added {Hexdump::VERSION}.
• Rescue RangeErrors in Hexdump::Dumper#format_printable.
• Fixed a typo in benchmarks/hexdump.rb (thanks ColMcp).
• Fixed a typo in Source URL, in the README (thanks Lawrence Woodman).
• Documented the :endian option for {Hexdump.dump}.
• Replaced ore-tasks with rubygems-tasks.

0.2.1 / 2011-06-11

• Fixed a major bug in Hexdump::Dumper#dump, where the line buffers were not being cleared.

0.2.0 / 2011-06-11

• Added Hexdump::Dumper.
• Added support for hexdumping 1, 2, 4, 8 byte words.
• Added support for hexdumping Little and Big Endian words.
• Optimized the hexdump algorithm to not use Arrays, use a lookup table of printable characters and cache formatted numbers.
• Opted into test.rubygems.org.

0.1.0 / 2011-03-05

• Initial release:
  • Can hexdump any Object supporting the each_byte method.
  • Can send the hexdump output to any Object supporting the << method.
  • Can yield each line of hexdump, instead of printing the output.
  • Supports printing hexadecimal, decimal, octal and binary bytes.
  • Makes {String}, {StringIO}, {IO}, {File} objects hexdumpable.
  • Fast-ish.