cmess 0.1.0.281 → 0.1.1.283

data/lib/cmess.rb

@@ -36,6 +36,8 @@
 #                   convert them back -- this is where cinderella comes in,
 #                   sorting the good ones into the pot and the (potentially)
 #                   bad ones into the crop... (see Cinderella)
+# bconv::           Convert between bibliographic (and other) encodings.
+#                   (see BConv)
 # decode_entities:: Decode HTML entities in a string. (see DecodeEntities)
 
 module CMess

data/lib/cmess/bconv.rb

@@ -0,0 +1,98 @@
+#--
+###############################################################################
+#                                                                             #
+# A component of cmess, the encoding tool-box.                                #
+#                                                                             #
+# Copyright (C) 2008 University of Cologne,                                   #
+#                    Albertus-Magnus-Platz,                                   #
+#                    50932 Cologne, Germany                                   #
+#                                                                             #
+# Authors:                                                                    #
+#     Jens Wille <jens.wille@uni-koeln.de>                                    #
+#                                                                             #
+# cmess is free software; you can redistribute it and/or modify it under the  #
+# terms of the GNU General Public License as published by the Free Software   #
+# Foundation; either version 3 of the License, or (at your option) any later  #
+# version.                                                                    #
+#                                                                             #
+# cmess is distributed in the hope that it will be useful, but WITHOUT ANY    #
+# WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS   #
+# FOR A PARTICULAR PURPOSE. See the GNU General Public License for more       #
+# details.                                                                    #
+#                                                                             #
+# You should have received a copy of the GNU General Public License along     #
+# with cmess. If not, see <http://www.gnu.org/licenses/>.                     #
+#                                                                             #
+###############################################################################
+#++
+
+require 'iconv'
+require 'cmess'
+
+# Convert between bibliographic (and other) encodings.
+
+module CMess::BConv
+
+  extend self
+
+  # our version ;-)
+  VERSION = '0.0.1'
+
+  INTERMEDIATE_ENCODING = 'utf-8'
+
+  def encodings(chartab)
+    chartab[chartab.keys.first].keys.map { |encoding|
+      encoding.upcase unless encoding =~ /\A__/
+    }.compact.sort
+  end
+
+  def convert(input, output, source_encoding, target_encoding, chartab)
+    source_encoding.upcase!
+    target_encoding.upcase!
+
+    encodings = self.encodings(chartab)
+
+    if encodings.include?(source_encoding)
+      if encodings.include?(target_encoding)
+        charmap = chartab.inject({}) { |hash, (code, map)|
+          hash.update(map[source_encoding] => map[target_encoding].pack('U*'))
+        }
+
+        input.each_byte { |byte|
+          output.print charmap[[byte]] || charmap[[byte, input.getc]]
+        }
+      else
+        iconv = Iconv.new(target_encoding, INTERMEDIATE_ENCODING)
+
+        charmap = chartab.inject({}) { |hash, (code, map)|
+          hash.update(map[source_encoding] => [code.to_i(16)].pack('U*'))
+        }
+
+        input.each_byte { |byte|
+          output.print iconv.iconv(charmap[[byte]] || charmap[[byte, input.getc]])
+        }
+      end
+    else
+      if encodings.include?(target_encoding)
+        iconv = Iconv.new(INTERMEDIATE_ENCODING, source_encoding)
+
+        charmap = chartab.inject({}) { |hash, (code, map)|
+          hash.update(code.to_i(16) => map[target_encoding].pack('U*'))
+        }
+
+        input.each { |line|
+          iconv.iconv(line).unpack('U*').each { |byte|
+            output.print charmap[byte]
+          }
+        }
+      else
+        iconv = Iconv.new(target_encoding, source_encoding)
+
+        input.each { |line|
+          output.puts iconv.iconv(line)
+        }
+      end
+    end
+  end
+
+end

data/lib/cmess/cli.rb

@@ -28,52 +28,54 @@
 
 module CMess::CLI
 
-    def ensure_readable(file)
-      abort "Can't find input file: #{file}" unless File.readable?(file)
-    end
+  DATA_DIR = File.expand_path(File.join(File.dirname(__FILE__), '..', '..', 'data'))
 
-    def ensure_directory(dir)
-      abort "Directory not found: #{dir}" unless File.directory?(dir)
-    end
+  def ensure_readable(file)
+    abort "Can't find input file: #{file}" unless File.readable?(file)
+  end
 
-    def open_file_in_place(file)
-      ensure_readable(file)
-      [File.readlines(file), File.open(file, 'w')]
-    end
+  def ensure_directory(dir)
+    abort "Directory not found: #{dir}" unless File.directory?(dir)
+  end
+
+  def open_file_in_place(file)
+    ensure_readable(file)
+    [File.readlines(file), File.open(file, 'w')]
+  end
 
-    def open_file_or_std(file, mode = 'r')
-      if file == '-'
-        case mode
-          when 'r': STDIN
-          when 'w': STDOUT
-          when 'a': STDERR
-          else      raise ArgumentError, "don't know how to handle mode '#{mode}'"
-        end
-      else
-        ensure_readable(file) unless mode == 'w'
-        File.open(file, mode)
+  def open_file_or_std(file, mode = 'r')
+    if file == '-'
+      case mode
+        when 'r': STDIN
+        when 'w': STDOUT
+        when 'a': STDERR
+        else      raise ArgumentError, "don't know how to handle mode '#{mode}'"
       end
+    else
+      ensure_readable(file) unless mode == 'w'
+      File.open(file, mode)
     end
+  end
 
-    def determine_system_encoding
-      ENV['SYSTEM_ENCODING']   ||
-      ENV['LANG'][/\.(.*)/, 1] ||
-      system_encoding_not_found
-    end
+  def determine_system_encoding
+    ENV['SYSTEM_ENCODING']   ||
+    ENV['LANG'][/\.(.*)/, 1] ||
+    system_encoding_not_found
+  end
 
-    def system_encoding_not_found
-      not_found = lambda {
-        abort <<-EOT
+  def system_encoding_not_found
+    not_found = lambda {
+      abort <<-EOT
 Your system's encoding couldn't be determined automatically -- please specify it
 explicitly via the SYSTEM_ENCODING environment variable or via the '-t' option.
-        EOT
-      }
+      EOT
+    }
 
-      def not_found.to_s
-        'NOT FOUND'
-      end
-
-      not_found
+    def not_found.to_s
+      'NOT FOUND'
     end
 
+    not_found
   end
+
+end

data/lib/cmess/version.rb

@@ -30,7 +30,7 @@ module CMess::Version
 
   MAJOR = 0
   MINOR = 1
-  TINY  = 0
+  TINY  = 1
 
   class << self
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: cmess
 version: !ruby/object:Gem::Version 
-  version: 0.1.0.281
+  version: 0.1.1.283
 platform: ruby
 authors: 
 - Jens Wille
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2008-09-15 00:00:00 +02:00
+date: 2008-09-16 00:00:00 +02:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -32,10 +32,11 @@ dependencies:
       - !ruby/object:Gem::Version 
         version: "0"
     version: 
-description: "Assist with handling messed up encodings (Currently includes the following tools: cinderella, decode_entities, guess_encoding)"
+description: "Assist with handling messed up encodings (Currently includes the following tools: bconv, cinderella, decode_entities, guess_encoding)"
 email: jens.wille@uni-koeln.de
 executables: 
 - cinderella
+- bconv
 - decode_entities
 - guess_encoding
 extensions: []
@@ -46,6 +47,7 @@ extra_rdoc_files:
 - README
 files: 
 - lib/cmess.rb
+- lib/cmess/bconv.rb
 - lib/cmess/version.rb
 - lib/cmess/guess_encoding.rb
 - lib/cmess/cli.rb
@@ -55,6 +57,7 @@ files:
 - lib/cmess/guess_encoding/encoding.rb
 - lib/cmess/guess_encoding/automatic.rb
 - bin/cinderella
+- bin/bconv
 - bin/decode_entities
 - bin/guess_encoding
 - COPYING
@@ -94,6 +97,7 @@ files:
 - data/csets/unicode/basic_latin.yaml
 - data/csets/unicode/cyrillic.yaml
 - data/test_chars.yaml
+- data/chartab.yaml
 has_rdoc: true
 homepage: http://prometheus.rubyforge.org/cmess
 post_install_message: 
@@ -127,6 +131,6 @@ rubyforge_project: prometheus
 rubygems_version: 1.2.0
 signing_key: 
 specification_version: 2
-summary: "Assist with handling messed up encodings (Currently includes the following tools: cinderella, decode_entities, guess_encoding)"
+summary: "Assist with handling messed up encodings (Currently includes the following tools: bconv, cinderella, decode_entities, guess_encoding)"
 test_files: []