ob64 0.5.0 → 0.7.0

data/ext/ob64/pack_uuencode.h

@@ -0,0 +1,18 @@
+#ifndef PACK_UUENCODE_H
+#define PACK_UUENCODE_H
+
+#include <stddef.h>
+
+size_t pack_uuencode_encode(const char *src, size_t srclen, char *out);
+/* Encodes like native pack("u") with line_len (a multiple of 3, 3..63) bytes per line; returns the length */
+size_t pack_uuencode_lines(const char *src, size_t srclen, char *out, size_t line_len);
+size_t pack_uuencode_decode(const char *src, size_t srclen, char *out);
+/* Upper bound of the decoded length of srclen chars: srclen * 3 / 4 without overflow */
+size_t pack_uudecode_cap(size_t srclen);
+/*
+ * Decodes like native unpack("u") and returns the decoded length. out needs
+ * pack_uudecode_cap(srclen) + 3 bytes; NULL computes the length only.
+ */
+size_t pack_uudecode_lines(const char *src, size_t srclen, char *out);
+
+#endif /* PACK_UUENCODE_H */

data/lib/ob64/core_ext.rb

@@ -1,29 +1,6 @@
 # frozen_string_literal: true
 
 require "ob64"
-require "base64"
 
-module Ob64
-  module CoreExt
-    def strict_encode64(bin)
-      Ob64.strictier_encode(bin)
-    end
-
-    def strict_decode64(str)
-      Ob64.strictier_decode(str)
-    end
-
-    def urlsafe_encode64(bin, padding: true)
-      Ob64.urlsafe_encode(bin, padding: padding)
-    end
-
-    def urlsafe_decode64(str)
-      Ob64.urlsafe_decode(str)
-    end
-  end
-
-  ::Base64.prepend(CoreExt)
-  ::Base64.singleton_class.prepend(CoreExt)
-
-  private_constant :CoreExt
-end
+::Array.prepend(Ob64::Ext::Pack)
+::String.prepend(Ob64::Ext::Unpack)

data/lib/ob64/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Ob64
-  VERSION = "0.5.0"
+  VERSION = "0.7.0"
 end

data/lib/ob64.rb

@@ -3,94 +3,242 @@
 require_relative "ob64/version"
 require_relative "ob64_ext"
 
-# Methods for base64-encoding and -decoding strings.
+# Fast base64 and uuencode encoding and decoding: see {Base64::RFC4648},
+# {Base64::RFC2045} and {UU}.
 module Ob64
-  # The glue between Ruby and libbase64. See +../ext/libbase64/ob64/ob64_ext.c+.
-  module LibBase64; end
+  # The glue between Ruby and the C codecs. See +../ext/ob64/ob64_ext.c+.
+  module Ext; end
 
-  include LibBase64
-  extend LibBase64
+  include Ext
+  extend Ext
 
   module_function
 
-  # Returns the Base64-encoded version of +bin+.
-  # This method complies with RFC 4648.
-  # No line feeds are added.
+  # Returns the codec used by {Base64}, {UU} and, with +ob64/core_ext+,
+  # +pack+ and +unpack+ with +m+ and +u+. Defaults to {best_codec}.
   #
-  # @param bin [String]
-  # @return [String]
-  def encode(bin)
-    __encode(bin)
+  # @return [Symbol] one of {codecs}
+  def codec
+    __codec
   end
 
-  # Returns the Base64-decoded version of +string+.
-  # This method complies with RFC 4648.
-  # ArgumentError is raised if +string+ is incorrectly padded or contains
-  # non-alphabet characters. Note that CR or LF are also rejected.
+  # Forces a codec for the whole process: every thread and Ractor.
   #
-  # @param string [String]
-  # @return [String]
-  # @raise [ArgumentError] if +string+ cannot be decoded
-  def decode(string)
-    __decode(string)
+  # @param codec [Symbol] one of {codecs}
+  # @return [Symbol] +codec+
+  # @raise [ArgumentError] if +codec+ is unknown or this CPU cannot run it
+  def codec=(codec)
+    __set_codec(codec)
   end
 
-  # Returns the Base64-encoded version of +bin+.
-  # This method complies with ``Base 64 Encoding with URL and Filename Safe
-  # Alphabet'' in RFC 4648.
-  # The alphabet uses '-' instead of '+' and '_' instead of '/'.
-  # Note that the result can still contain '='.
-  # You can remove the padding by setting +padding+ as false.
+  # Returns the codecs this CPU runs, best last. +:scalar+ is plain C.
   #
-  # @param bin [String]
-  # @param padding [Boolean] - if the output must be padded
-  # @return [String]
-  def urlsafe_encode(bin, padding: true)
-    string = __encode(bin)
-    string.chomp!("==") || string.chomp!("=") unless padding
-    string.tr!("+/", "-_")
-    string
+  # @return [Array<Symbol>] among +:scalar+, +:avx2+, +:avx512+, +:neon64+
+  def codecs
+    __codecs
   end
 
-  # Returns the Base64-decoded version of +string+.
-  # This method complies with ``Base 64 Encoding with URL and Filename Safe
-  # Alphabet'' in RFC 4648.
-  # The alphabet uses '-' instead of '+' and '_' instead of '/'.
+  # Returns the best codec this CPU runs, the default {codec}.
   #
-  # The padding character is optional.
-  # This method accepts both correctly-padded and unpadded input.
-  # Note that it still rejects incorrectly-padded input.
-  #
-  # @param string [String]
-  # @return [String]
-  # @raise [ArgumentError] if +string+ cannot be decoded
-  def urlsafe_decode(string)
-    if !string.end_with?("=") && string.length % 4 != 0
-      string = string.ljust((string.length + 3) & ~3, "=")
-      string.tr!("-_", "+/")
-    else
-      string = string.tr("-_", "+/")
-    end
-    __decode(string)
+  # @return [Symbol] one of {codecs}
+  def best_codec
+    codecs.last
   end
 
-  # Returns the length of the Base64-encoded version of +bin+.
-  #
-  # @param bin [String]
-  # @param padding [Boolean] - if the Base64-encoded version of +bin+ will be padded
-  # @return [Integer]
-  def encoded_length_of(bin, padding: true)
-    __encoded_length_of(bin, padding)
+  # Base64 encoding and decoding, one module per RFC: {RFC4648} and {RFC2045}.
+  module Base64
+    # Base64 as in RFC 4648: no line breaks, strict decoding.
+    module RFC4648
+      include Ext
+      extend Ext
+
+      module_function
+
+      # Returns the Base64-encoded version of +bin+, like
+      # <tt>[bin].pack("m0")</tt>. No line feeds are added.
+      #
+      # @param bin [String]
+      # @return [String]
+      def encode(bin)
+        __encode(bin, false, true)
+      end
+
+      # Returns the Base64-decoded version of +string+, like
+      # <tt>string.unpack1("m0")</tt>.
+      # ArgumentError is raised if +string+ is incorrectly padded or contains
+      # non-alphabet characters. Note that CR or LF are also rejected.
+      #
+      # With +loose+, the padding character is optional: both correctly-padded
+      # and unpadded input are accepted. Incorrectly-padded input is still
+      # rejected.
+      #
+      # @param string [String]
+      # @param loose [Boolean] - if unpadded input is accepted
+      # @return [String]
+      # @raise [ArgumentError] if +string+ cannot be decoded
+      def decode(string, loose: false)
+        __decode(string, false, loose)
+      end
+
+      # Returns the Base64-encoded version of +bin+ with the ``URL and Filename
+      # Safe'' alphabet: '-' instead of '+' and '_' instead of '/'.
+      # Note that the result can still contain '='.
+      # You can remove the padding by setting +padding+ as false.
+      #
+      # @param bin [String]
+      # @param padding [Boolean] - if the output must be padded
+      # @return [String]
+      def encode_urlsafe(bin, padding: true)
+        __encode(bin, true, padding)
+      end
+
+      # Returns the Base64-decoded version of +string+ with the ``URL and
+      # Filename Safe'' alphabet: '-' instead of '+' and '_' instead of '/'.
+      #
+      # The padding character is optional.
+      # This method accepts both correctly-padded and unpadded input.
+      # Note that it still rejects incorrectly-padded input.
+      #
+      # @param string [String]
+      # @return [String]
+      # @raise [ArgumentError] if +string+ cannot be decoded
+      def decode_urlsafe(string)
+        __decode(string, true, true)
+      end
+
+      # Returns the length of {encode} (or, with the same +padding+,
+      # {encode_urlsafe}) of +bin+.
+      #
+      # Fast: computed from +bin+'s length.
+      #
+      # @param bin [String]
+      # @param padding [Boolean] - if the Base64-encoded version of +bin+ will be padded
+      # @return [Integer]
+      def encoded_length_of(bin, padding: true)
+        __encoded_length_of(bin, padding)
+      end
+
+      # Returns the length of {decode} (or {decode_urlsafe}) of +string+,
+      # padded or not.
+      #
+      # Fast: computed from +string+'s length and its last two characters. It
+      # doesn't validate the rest, so decoding may still raise.
+      #
+      # @param string [String]
+      # @return [Integer]
+      # @raise [ArgumentError] if +string+ has an invalid length
+      def decoded_length_of(string)
+        __decoded_length_of(string)
+      end
+    end
+
+    # Base64 as in RFC 2045 (MIME): line breaks, lenient decoding.
+    module RFC2045
+      include Ext
+      extend Ext
+
+      module_function
+
+      # Returns the Base64-encoded version of +bin+ in lines of +line_length+
+      # characters, each ending in a line break. Like <tt>[bin].pack("m")</tt>
+      # by default; with +crlf+, lines end in "\r\n", as MIME requires.
+      # RFC 2045 allows lines of up to 76 characters; longer ones are allowed
+      # here, as in +pack+, but break the RFC.
+      #
+      # @param bin [String]
+      # @param line_length [Integer] - characters per line, a positive multiple of 4
+      # @param crlf [Boolean] - if lines end in "\r\n" instead of "\n"
+      # @return [String]
+      # @raise [ArgumentError] if +line_length+ is not a positive multiple of 4
+      def encode(bin, line_length = 60, crlf: false)
+        __encode_rfc_2045(bin, line_length, crlf)
+      end
+
+      # Returns the Base64-decoded version of +string+, like
+      # <tt>string.unpack1("m")</tt>: characters outside the alphabet, such
+      # as line breaks, are skipped. Decoding is lenient, as in Ruby, and
+      # never raises for malformed input.
+      #
+      # @param string [String]
+      # @return [String]
+      def decode(string)
+        __decode_rfc_2045(string)
+      end
+
+      # Returns the length of {encode} of +bin+.
+      #
+      # Fast: computed from +bin+'s length.
+      #
+      # @param bin [String]
+      # @param line_length [Integer] - characters per line, a positive multiple of 4
+      # @param crlf [Boolean] - if lines end in "\r\n" instead of "\n"
+      # @return [Integer]
+      # @raise [ArgumentError] if +line_length+ is not a positive multiple of 4
+      def encoded_length_of(bin, line_length = 60, crlf: false)
+        __encoded_length_of_rfc_2045(bin, line_length, crlf)
+      end
+
+      # Returns the length of {decode} of +string+.
+      #
+      # Costly: reads all of +string+, since skipped characters and +=+ decide
+      # the length. To size a buffer, <tt>string.bytesize * 3 / 4</tt> is
+      # always enough and fast.
+      #
+      # @param string [String]
+      # @return [Integer]
+      def decoded_length_of(string)
+        __decoded_length_of_rfc_2045(string)
+      end
+    end
   end
 
-  # Returns the length of the Base64-decoded version of +string+.
-  #
-  # ArgumentError is raised if +string+ has an invalid length.
-  #
-  # @param string [String]
-  # @return [Integer]
-  # @raise [ArgumentError] if +string+ has an invalid length
-  def decoded_length_of(string)
-    __decoded_length_of(string)
+  # Methods for uuencoding and -decoding strings.
+  module UU
+    include Ext
+    extend Ext
+
+    module_function
+
+    # Returns the uuencoded version of +bin+, like <tt>[bin].pack("u")</tt>:
+    # lines of up to 45 input bytes, each ending in a newline.
+    #
+    # @param bin [String]
+    # @return [String]
+    def encode(bin)
+      __uuencode(bin)
+    end
+
+    # Returns the uudecoded version of +string+, like <tt>string.unpack1("u")</tt>.
+    # Decoding is lenient, as in Ruby: it stops at the first invalid line and
+    # never raises for malformed input.
+    #
+    # @param string [String]
+    # @return [String]
+    def decode(string)
+      __uudecode(string)
+    end
+
+    # Returns the length of {encode} of +bin+.
+    #
+    # Fast: computed from +bin+'s length.
+    #
+    # @param bin [String]
+    # @return [Integer]
+    def encoded_length_of(bin)
+      __uuencoded_length_of(bin)
+    end
+
+    # Returns the length of {decode} of +string+.
+    # Like {decode}, never raises for malformed input.
+    #
+    # Costly: reads all of +string+, since each line's length character
+    # decides the length. To size a buffer, <tt>string.bytesize * 3 / 4</tt>
+    # is always enough and fast.
+    #
+    # @param string [String]
+    # @return [Integer]
+    def decoded_length_of(string)
+      __uudecoded_length_of(string)
+    end
   end
 end

data/ob64.gemspec

@@ -10,15 +10,13 @@ Gem::Specification.new do |spec|
   spec.email = ["joao.fernandes@ist.utl.pt"]
   spec.homepage = "https://github.com/jcmfernandes/ob64"
   spec.license = "MIT"
-  spec.summary = "A fast Base64 encoder and decoder."
-  spec.description = "A fast Base64 encoder and decoder that makes use of SIMD extensions."
+  spec.summary = "A fast Base64 and uuencode encoder and decoder."
+  spec.description = "A fast Base64 and uuencode encoder and decoder that makes use of SIMD extensions."
 
   # Specify which files should be added to the gem when it is released.
   # The `git ls-files -z` loads the files in the RubyGem that have been added into git.
   spec.files = Dir.chdir(File.expand_path(__dir__)) do
     `git ls-files -z`.split("\x0").reject { |f| f.match(%r{\A(?:test|spec|features)/}) }
-  end + Dir.chdir(File.join(File.expand_path(__dir__), "vendor", "libbase64")) do
-    `git ls-files -z`.split("\x0").reject { |f| f.match(%r{\A(?:test|spec|features)/}) }.map! { |f| "vendor/libbase64/#{f}" }
   end
   spec.require_paths = ["lib"]
   spec.extensions = ["ext/ob64/extconf.rb"]
@@ -30,5 +28,5 @@ Gem::Specification.new do |spec|
     "source_code_uri" => spec.homepage + "/tree/v#{spec.version}"
   }
 
-  spec.required_ruby_version = Gem::Requirement.new(">= 2.5.0")
+  spec.required_ruby_version = Gem::Requirement.new(">= 3.3.0")
 end

metadata

@@ -1,16 +1,16 @@
 --- !ruby/object:Gem::Specification
 name: ob64
 version: !ruby/object:Gem::Version
-  version: 0.5.0
+  version: 0.7.0
 platform: ruby
 authors:
 - João Fernandes
-autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-09-29 00:00:00.000000000 Z
+date: 1980-01-02 00:00:00.000000000 Z
 dependencies: []
-description: A fast Base64 encoder and decoder that makes use of SIMD extensions.
+description: A fast Base64 and uuencode encoder and decoder that makes use of SIMD
+  extensions.
 email:
 - joao.fernandes@ist.utl.pt
 executables: []
@@ -18,9 +18,9 @@ extensions:
 - ext/ob64/extconf.rb
 extra_rdoc_files: []
 files:
+- ".github/dependabot.yml"
 - ".github/workflows/main.yml"
 - ".gitignore"
-- ".gitmodules"
 - ".rspec"
 - ".standard.yml"
 - ".yardopts"
@@ -35,86 +35,32 @@ files:
 - bin/_guard-core
 - bin/console
 - bin/guard
+- bin/rake
 - bin/rspec
 - bin/setup
 - bin/standardrb
 - ext/ob64/extconf.rb
-- ext/ob64/libbase64.h
 - ext/ob64/ob64.h
 - ext/ob64/ob64_ext.c
+- ext/ob64/ob64_pack.c
+- ext/ob64/pack_base64.c
+- ext/ob64/pack_base64.h
+- ext/ob64/pack_generic.h
+- ext/ob64/pack_simd.h
+- ext/ob64/pack_uuencode.c
+- ext/ob64/pack_uuencode.h
 - lib/ob64.rb
 - lib/ob64/core_ext.rb
 - lib/ob64/version.rb
 - ob64.gemspec
-- vendor/libbase64/.gitignore
-- vendor/libbase64/.travis.yml
-- vendor/libbase64/CMakeLists.txt
-- vendor/libbase64/LICENSE
-- vendor/libbase64/Makefile
-- vendor/libbase64/README.md
-- vendor/libbase64/base64-benchmarks.png
-- vendor/libbase64/bin/base64.c
-- vendor/libbase64/cmake/Modules/TargetArch.cmake
-- vendor/libbase64/cmake/Modules/TargetSIMDInstructionSet.cmake
-- vendor/libbase64/cmake/base64-config.cmake.in
-- vendor/libbase64/cmake/config.h.in
-- vendor/libbase64/cmake/test-arch.c
-- vendor/libbase64/include/libbase64.h
-- vendor/libbase64/lib/arch/avx/codec.c
-- vendor/libbase64/lib/arch/avx2/codec.c
-- vendor/libbase64/lib/arch/avx2/dec_loop.c
-- vendor/libbase64/lib/arch/avx2/dec_reshuffle.c
-- vendor/libbase64/lib/arch/avx2/enc_loop.c
-- vendor/libbase64/lib/arch/avx2/enc_reshuffle.c
-- vendor/libbase64/lib/arch/avx2/enc_translate.c
-- vendor/libbase64/lib/arch/generic/32/dec_loop.c
-- vendor/libbase64/lib/arch/generic/32/enc_loop.c
-- vendor/libbase64/lib/arch/generic/64/enc_loop.c
-- vendor/libbase64/lib/arch/generic/codec.c
-- vendor/libbase64/lib/arch/generic/dec_head.c
-- vendor/libbase64/lib/arch/generic/dec_tail.c
-- vendor/libbase64/lib/arch/generic/enc_head.c
-- vendor/libbase64/lib/arch/generic/enc_tail.c
-- vendor/libbase64/lib/arch/neon32/codec.c
-- vendor/libbase64/lib/arch/neon32/dec_loop.c
-- vendor/libbase64/lib/arch/neon32/enc_loop.c
-- vendor/libbase64/lib/arch/neon32/enc_reshuffle.c
-- vendor/libbase64/lib/arch/neon32/enc_translate.c
-- vendor/libbase64/lib/arch/neon64/codec.c
-- vendor/libbase64/lib/arch/neon64/dec_loop.c
-- vendor/libbase64/lib/arch/neon64/enc_loop.c
-- vendor/libbase64/lib/arch/neon64/enc_reshuffle.c
-- vendor/libbase64/lib/arch/sse41/codec.c
-- vendor/libbase64/lib/arch/sse42/codec.c
-- vendor/libbase64/lib/arch/ssse3/codec.c
-- vendor/libbase64/lib/arch/ssse3/dec_loop.c
-- vendor/libbase64/lib/arch/ssse3/dec_reshuffle.c
-- vendor/libbase64/lib/arch/ssse3/enc_loop.c
-- vendor/libbase64/lib/arch/ssse3/enc_reshuffle.c
-- vendor/libbase64/lib/arch/ssse3/enc_translate.c
-- vendor/libbase64/lib/codec_choose.c
-- vendor/libbase64/lib/codecs.h
-- vendor/libbase64/lib/env.h
-- vendor/libbase64/lib/exports.txt
-- vendor/libbase64/lib/lib.c
-- vendor/libbase64/lib/lib_openmp.c
-- vendor/libbase64/lib/tables/.gitignore
-- vendor/libbase64/lib/tables/Makefile
-- vendor/libbase64/lib/tables/table_dec_32bit.h
-- vendor/libbase64/lib/tables/table_enc_12bit.h
-- vendor/libbase64/lib/tables/table_enc_12bit.py
-- vendor/libbase64/lib/tables/table_generator.c
-- vendor/libbase64/lib/tables/tables.c
-- vendor/libbase64/lib/tables/tables.h
 homepage: https://github.com/jcmfernandes/ob64
 licenses:
 - MIT
 metadata:
   homepage_uri: https://github.com/jcmfernandes/ob64
   changelog_uri: https://github.com/jcmfernandes/ob64/blob/master/CHANGELOG.md
-  documentation_uri: https://www.rubydoc.info/gems/ob64/0.5.0
-  source_code_uri: https://github.com/jcmfernandes/ob64/tree/v0.5.0
-post_install_message: 
+  documentation_uri: https://www.rubydoc.info/gems/ob64/0.7.0
+  source_code_uri: https://github.com/jcmfernandes/ob64/tree/v0.7.0
 rdoc_options: []
 require_paths:
 - lib
@@ -122,15 +68,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 2.5.0
+      version: 3.3.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.15
-signing_key: 
+rubygems_version: 4.0.23
 specification_version: 4
-summary: A fast Base64 encoder and decoder.
+summary: A fast Base64 and uuencode encoder and decoder.
 test_files: []

data/.gitmodules

@@ -1,3 +0,0 @@
-[submodule "vendor/libbase64"]
-	path = vendor/libbase64
-	url = https://github.com/jcmfernandes/base64-cmake

data/ext/ob64/libbase64.h

@@ -1 +0,0 @@
-#include "../../vendor/libbase64/include/libbase64.h"

data/vendor/libbase64/.gitignore

@@ -1,12 +0,0 @@
-*.o
-bin/base64
-lib/config.h
-test/benchmark
-test/test_base64
-
-# visual studio symbol db, etc.
-.vs/
-# build directory used by CMakePresets
-out/
-# private cmake presets
-CMakeUserPresets.json

data/vendor/libbase64/.travis.yml

@@ -1,71 +0,0 @@
-language: c
-services: docker
-
-jobs:
-  allow_failures:
-    - os: osx
-    - arch: ppc64le
-
-  include:
-    - name: linux-amd64-gcc
-      os: linux
-      arch: amd64
-      compiler: gcc
-      script: test/ci/amd64.sh
-
-    - name: linux-amd64-gcc +openmp
-      os: linux
-      arch: amd64
-      compiler: gcc
-      env: OPENMP=1 OMP_NUM_THREADS=4
-      script: test/ci/amd64.sh
-
-    - name: linux-amd64-clang
-      os: linux
-      arch: amd64
-      compiler: clang
-      script: test/ci/amd64.sh
-
-    - name: linux-arm32-gcc
-      compiler: gcc
-      script: test/ci/docker-arm32.sh
-
-    - name: linux-arm32-clang
-      compiler: clang
-      script: test/ci/docker-arm32.sh
-
-    - name: linux-powerpc-gcc
-      compiler: gcc
-      script: test/ci/docker-powerpc.sh
-
-    - name: linux-arm64-gcc
-      os: linux
-      arch: arm64
-      compiler: gcc
-      env: NEON64_CFLAGS="-march=armv8-a"
-      script: test/ci/generic.sh
-
-    - name: linux-arm64-clang
-      os: linux
-      arch: arm64
-      compiler: clang
-      env: NEON64_CFLAGS="-march=armv8-a"
-      script: test/ci/generic.sh
-
-    - name: linux-s390x-gcc
-      os: linux
-      arch: s390x
-      compiler: gcc
-      script: test/ci/generic.sh
-
-    - name: linux-ppc64le-gcc
-      os: linux
-      arch: ppc64le
-      compiler: gcc
-      script: test/ci/generic.sh
-
-    - name: osx-amd64-clang
-      os: osx
-      arch: amd64
-      compiler: clang
-      script: test/ci/amd64.sh