Implement more of the API

README.rdoc

@@ -32,6 +32,11 @@ I've tested this on these platforms:
 - Debian Linux
 - Arch Linux
 
+and these rubies:
+
+- MRI 1.9.1
+- MRI 1.8.7
+
 YMMV.
 
 == Note on Patches/Pull Requests

Rakefile

@@ -4,14 +4,14 @@ require 'rake'
 begin
   require 'jeweler'
   Jeweler::Tasks.new do |gem|
-    gem.name = "icu-chardet-ffi"
-    gem.summary = %Q{Tiny FFI wrapper for ICU's UCharsetDetector.}
+    gem.name        = "icu-chardet-ffi"
+    gem.summary     = %Q{Tiny FFI wrapper for ICU's UCharsetDetector.}
     gem.description = %Q{Tiny FFI wrapper for ICU's UCharsetDetector.}
-    gem.email = "jari.bakken@gmail.com"
-    gem.homepage = "http://github.com/jarib/icu-chardet-ffi"
-    gem.authors = ["Jari Bakken"]
+    gem.email       = "jari.bakken@gmail.com"
+    gem.homepage    = "http://github.com/jarib/icu-chardet-ffi"
+    gem.authors     = ["Jari Bakken"]
 
-    gem.add_dependency "ffi", "0.6.3"
+    gem.add_dependency "ffi", ">= 0.6.3"
     gem.add_development_dependency "rspec", ">= 1.3.0"
     # gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
   end

lib/icu-chardet-ffi.rb

@@ -17,13 +17,22 @@ module ICU
       suffix = '_44'
     end
 
+    #
+    # http://icu-project.org/apiref/icu4c/ucsdet_8h.html
+    #
+
     attach_function "ucsdet_open#{suffix}", [:pointer], :pointer
     attach_function "ucsdet_close#{suffix}", [:pointer], :void
     attach_function "ucsdet_setText#{suffix}", [:pointer, :string, :int, :pointer], :void
+    attach_function "ucsdet_setDeclaredEncoding#{suffix}", [:pointer, :string, :int, :pointer], :void
     attach_function "ucsdet_detect#{suffix}", [:pointer, :pointer], :pointer
+    attach_function "ucsdet_detectAll#{suffix}", [:pointer, :pointer, :pointer], :pointer
     attach_function "ucsdet_getName#{suffix}", [:pointer, :pointer], :string
     attach_function "ucsdet_getConfidence#{suffix}", [:pointer, :pointer], :int
+    attach_function "ucsdet_getLanguage#{suffix}", [:pointer, :pointer], :string
     attach_function "ucsdet_getAllDetectableCharsets#{suffix}", [:pointer, :pointer], :pointer
+    attach_function "ucsdet_isInputFilterEnabled#{suffix}", [:pointer], :bool
+    attach_function "ucsdet_enableInputFilter#{suffix}", [:pointer, :bool], :bool
     attach_function "u_errorName#{suffix}", [:int], :string
     attach_function "uenum_count#{suffix}", [:pointer, :pointer], :int
     attach_function "uenum_close#{suffix}", [:pointer], :void
@@ -38,7 +47,14 @@ module ICU
         alias_method :ucsdet_detect, "ucsdet_detect#{suffix}"
         alias_method :ucsdet_getName, "ucsdet_getName#{suffix}"
         alias_method :ucsdet_getConfidence, "ucsdet_getConfidence#{suffix}"
+        alias_method :ucsdet_getLanguage, "ucsdet_getLanguage#{suffix}"
+        alias_method :ucsdet_isInputFilterEnabled, "ucsdet_isInputFilterEnabled#{suffix}"
+        alias_method :ucsdet_enableInputFilter, "ucsdet_enableInputFilter#{suffix}"
+        alias_method :ucsdet_setDeclaredEncoding, "ucsdet_setDeclaredEncoding#{suffix}"
         alias_method :u_errorName, "u_errorName#{suffix}"
+        alias_method :uenum_count, "uenum_count#{suffix}"
+        alias_method :uenum_count, "uenum_close#{suffix}"
+        alias_method :uenum_count, "uenum_next#{suffix}"
       end
     end
 
@@ -51,7 +67,7 @@ module ICU
     end
 
     class Detector
-      Match = Struct.new(:name, :confidence)
+      Match = Struct.new(:name, :confidence, :language)
 
       def initialize
         check_status do |ptr|
@@ -59,31 +75,50 @@ module ICU
         end
       end
 
+      def input_filter_enabled?
+        CharDet.ucsdet_isInputFilterEnabled @detector
+      end
+
+      def input_filter_enabled=(bool)
+        CharDet.ucsdet_enableInputFilter(@detector, !!bool)
+      end
+
+      def declared_encoding=(str)
+        check_status do |ptr|
+          CharDet.ucsdet_setDeclaredEncoding(@detector, str, str.bytesize, ptr)
+        end
+      end
+
       def close
         CharDet.ucsdet_close @detector
       end
 
       def detect(str)
-        check_status do |ptr|
-          CharDet.ucsdet_setText(@detector, str, str.bytesize, ptr)
-        end
+        set_text(str)
 
         match_ptr = nil
-
         check_status do |ptr|
           match_ptr = CharDet.ucsdet_detect(@detector, ptr)
         end
 
-        result = Match.new
-        check_status do |ptr|
-          result.name = CharDet.ucsdet_getName(match_ptr, ptr)
+        match_ptr_to_ruby(match_ptr) unless match_ptr.null?
+      end
+
+      def detect_all(str)
+        set_text(str)
+
+        matches_found_ptr = FFI::MemoryPointer.new :int
+        array_ptr = nil
+
+        check_status do |status|
+          array_ptr = CharDet.ucsdet_detectAll(@detector, matches_found_ptr, status)
         end
 
-        check_status do |ptr|
-          result.confidence = CharDet.ucsdet_getConfidence(match_ptr, ptr)
-        end
+        length = matches_found_ptr.read_int
 
-        result
+        array_ptr.read_array_of_pointer(length).map do |match|
+          match_ptr_to_ruby(match)
+        end
       end
 
       def detectable_charsets
@@ -93,7 +128,7 @@ module ICU
           enum_ptr = CharDet.ucsdet_getAllDetectableCharsets(@detector, ptr)
         end
 
-        result = enumeration_to_array(enum_ptr)
+        result = enum_ptr_to_array(enum_ptr)
         CharDet.uenum_close(enum_ptr)
 
         result
@@ -112,20 +147,44 @@ module ICU
         end
       end
 
-      def enumeration_to_array(ptr)
+      def enum_ptr_to_array(enum_ptr)
         length = 0
         check_status do |status|
-          length = CharDet.uenum_count(ptr, status)
+          length = CharDet.uenum_count(enum_ptr, status)
         end
 
         result = []
         0.upto(length - 1) do |idx|
-          check_status { |st| result << CharDet.uenum_next(ptr, nil, st) }
+          check_status { |st| result << CharDet.uenum_next(enum_ptr, nil, st) }
         end
 
         result
       end
 
+      def match_ptr_to_ruby(match_ptr)
+        result = Match.new
+
+        check_status do |ptr|
+          result.name = CharDet.ucsdet_getName(match_ptr, ptr)
+        end
+
+        check_status do |ptr|
+          result.confidence = CharDet.ucsdet_getConfidence(match_ptr, ptr)
+        end
+
+        check_status do |ptr|
+          result.language = CharDet.ucsdet_getLanguage(match_ptr, ptr)
+        end
+
+        result
+      end
+
+      def set_text(text)
+        check_status do |status|
+          CharDet.ucsdet_setText(@detector, text, text.bytesize, status)
+        end
+      end
+
     end # Detector
   end # CharDet
 end # ICU

spec/chardet_spec.rb

@@ -3,20 +3,37 @@
 require 'spec_helper'
 
 describe ICU::CharDet::Detector do
+
   before { @d = ICU::CharDet::Detector.new }
   after { @d.close }
 
-
   it "should recognize UTF-8" do
-    @d.detect("æåø").name.should == "UTF-8"
+    m = @d.detect("æåø")
+    m.name.should == "UTF-8"
+    m.language.should be_kind_of(String)
   end
 
   it "has a list of detectable charsets" do
     cs = @d.detectable_charsets
     cs.should be_kind_of(Array)
     cs.should_not be_empty
-    
+
     cs.first.should be_kind_of(String)
   end
 
+  it "should disable / enable the input filter" do
+    @d.input_filter_enabled?.should be_false
+    @d.input_filter_enabled = true
+    @d.input_filter_enabled?.should be_true
+  end
+
+  it "should should set declared encoding" do
+    @d.declared_encoding = "UTF-8"
+  end
+
+  it "should detect several matching encodings" do
+    r = @d.detect_all("foo bar")
+    r.should be_instance_of(Array)
+  end
+
 end

spec/spec_helper.rb

@@ -1,5 +1,7 @@
 $LOAD_PATH.unshift(File.dirname(__FILE__))
 $LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+
+require "rubygems"
 require 'icu-chardet-ffi'
 require 'spec'
 require 'spec/autorun'