ferret 0.1.3 → 0.1.4

data/test/unit/index/tc_index_reader.rb

@@ -597,7 +597,7 @@ class IndexReaderTest < Test::Unit::TestCase
   def test_ir_read_while_optimizing_on_disk()
     dpath = File.join(File.dirname(__FILE__),
                        '../../temp/fsdir')
-    fs_dir = Ferret::Store::FSDirectory.get_directory(dpath, true)
+    fs_dir = Ferret::Store::FSDirectory.new(dpath, true)
 
     iw = IndexWriter.new(fs_dir, :analyzer => WhiteSpaceAnalyzer.new(), :create => true)
     docs = IndexTestHelper.prepare_ir_test_docs()

data/test/unit/index/tc_segment_infos.rb

@@ -63,7 +63,7 @@ class SegmentInfoTest < Test::Unit::TestCase
     assert_equal(si.name, "seg1")
     @dir.close()
     @dpath = File.dirname(__FILE__) + '/../../temp/fsdir'
-    @dir = Ferret::Store::FSDirectory.get_directory(@dpath, true)
+    @dir = Ferret::Store::FSDirectory.new(@dpath, true)
     si.name = "seg2"
     si.doc_count += 2
     si.directory = @dir

data/test/unit/index/th_doc.rb

@@ -234,7 +234,7 @@ module IndexTestHelper
       doc.boost = i+1
 
       fields.each_pair do |field, text|
-        doc << Field.new(field, text, Field::Store::NO, Field::Index::TOKENIZED, Field::TermVector::NO, i+1)
+        doc << Field.new(field, text, Field::Store::YES, Field::Index::TOKENIZED, Field::TermVector::NO, false)
       end
       docs << doc
     end

data/test/unit/search/tc_index_searcher.rb

@@ -46,6 +46,12 @@ class IndexSearcherTest < Test::Unit::TestCase
     end
   end
 
+  def test_get_doc()
+    assert_equal(18, @is.max_doc)
+    assert_equal("20050930", @is.doc(0).values(:date))
+    assert_equal("cat1/sub2/subsub2", @is.doc(4)[:cat])
+  end
+
   def test_term_query
     tq = TermQuery.new(Term.new("field", "word2"));
     tq.boost = 100

data/test/unit/store/tc_fs_store.rb

@@ -26,7 +26,7 @@ class FSStoreTest < Test::Unit::TestCase
   def setup
     @dpath = File.join(File.dirname(__FILE__),
                        '../../temp/fsdir')
-    @dir = FSDirectory.get_directory(@dpath, true)
+    @dir = FSDirectory.new(@dpath, true)
   end
 
   def teardown
@@ -39,12 +39,12 @@ class FSStoreTest < Test::Unit::TestCase
                          '/../../temp/cachetest')
     assert(! FSDirectory.directory_cache[dir_path],
            "this directory should not be cached yet") 
-    @dir1 = FSDirectory.get_directory(dir_path, true)
+    @dir1 = FSDirectory.new(dir_path, true)
     assert(FSDirectory.directory_cache[dir_path],
            "this directory should now be cached") 
     assert_equal(@dir1.ref_count, 1,
                  "There is one reference so the refcount should now be 1")
-    @dir2 = FSDirectory.get_directory(dir_path, true)
+    @dir2 = FSDirectory.new(dir_path, true)
     assert(@dir1 === @dir2,
            "The directory should be cached so the same directory object should have been returned")
     assert_equal(@dir1.ref_count, 2,

data/test/unit/utils/tc_bit_vector.rb

@@ -4,6 +4,14 @@ require File.dirname(__FILE__) + "/../../test_helper"
 class BitVectorTest < Test::Unit::TestCase
   include Ferret::Utils
 
+  def test_bignum_conversion()
+    j = 256
+    10.times do
+      j *= j
+      assert_equal(j, BitVector.string_to_bignum(BitVector.bignum_to_string(j)))
+    end
+  end
+
   def test_bv()
     bv = BitVector.new
     assert_equal(0, bv.count)

data/test/unit/utils/tc_thread.rb

@@ -0,0 +1,61 @@
+require File.dirname(__FILE__) + "/../../test_helper"
+require 'thread'
+
+
+class ThreadTest < Test::Unit::TestCase
+  include Ferret::Utils
+
+  NUM_THREADS = 100
+
+  def test_basic_get_and_set()
+    Thread.current.clear_local
+    b = "hello"
+    Thread.current.set_local(b, "dave")
+    assert_equal("dave", Thread.current.get_local(b))
+  end
+
+  def test_objects_die
+    Thread.current.clear_local
+    a = []
+    10.times {|i| a[i] = "#{i}"; Thread.current.set_local(a[i], i) }
+    10.times {|i| assert_equal(i, Thread.current.get_local(a[i])) }
+    assert_equal(10, Thread.current.local_size)
+    GC.start
+    assert_equal(10, Thread.current.local_size)
+    10.times {|i| a[i] = nil; }
+    #puts w
+    
+    # this is a hack to get the GC to collect the last ref created above
+    x = WeakKeyHash.new()
+    10.times {|i| a[i] = "#{i}"; x[a[i]] = i }
+
+    assert_equal(10, Thread.current.local_size)
+    GC.start
+    assert(0, Thread.current.local_size)
+  end
+
+  class ThreadTester
+    def initialize(val)
+      Thread.current.set_local(self, val)
+    end
+    def inc
+      val = Thread.current.get_local(self) + 1
+      Thread.current.set_local(self, val)
+      return val
+    end
+  end
+
+  def single_thread
+    tt = ThreadTester.new(start = rand(10000000))
+    ((start+1)..start+11).each {|i| assert_equal(i, tt.inc) }
+  end
+
+  def test_threads_dont_share
+    threads = []
+    NUM_THREADS.times do
+      threads << Thread.new { single_thread }
+    end
+
+    threads.each {|t| t.join}
+  end
+end

data/test/unit/utils/tc_weak_key_hash.rb

@@ -4,11 +4,12 @@ require File.dirname(__FILE__) + "/../../test_helper"
 class WeakKeyHashTest < Test::Unit::TestCase
   include Ferret::Utils
 
-  def test_marshalling()
+  def test_objects_are_destroyed()
     w = WeakKeyHash.new()
     a = []
     10.times {|i| a[i] = "#{i}"; w[a[i]] = i }
     10.times {|i| assert_equal(i, w[a[i]]) }
+    assert_equal(10, w.size)
     10.times {|i| a[i] = nil; }
     #puts w
     
@@ -16,7 +17,6 @@ class WeakKeyHashTest < Test::Unit::TestCase
     x = WeakKeyHash.new()
     10.times {|i| a[i] = "#{i}"; x[a[i]] = i }
 
-    assert_equal(10, w.size)
     GC.start
     #puts w.size
     #puts w

data/test/utils/number_to_spoken.rb

@@ -0,0 +1,132 @@
+# Author: Matthew D Moss
+#
+# Writtern for ruby quiz #25
+#
+class JapaneseTranslator
+    # My knowledge of counting Japanese is limited, so this may not
+    # be entirely correct; in particular, I don't know what rules
+    # to follow after 'hyaku man' (1,000,000).
+    # I also combine a digit with its group, such as 'gohyaku' rather
+    # than 'go hyaku'; I just like reading it better that way.
+
+    DIGITS = %w(zero ichi ni san yon go roku nana hachi kyu)
+    GROUPS = %w(nothingtoseeheremovealong ju hyaku sen)
+    MAN = 10000
+
+    def to_spoken(val)
+        case val <=> 0
+        when -1
+            '- ' + to_spoken(-val)
+        when 0
+            DIGITS[0]
+        else
+            group(val, 0)
+        end
+    end
+
+    private
+
+    def group(val, level)
+        if val >= MAN
+            group(val / MAN, 0) + 'man ' + group(val % MAN, 0)
+        else
+            case val
+            when 0
+                ''
+            when 1
+                level == 0 ? DIGITS[val] : GROUPS[level]
+            when 2...10
+                DIGITS[val] + (GROUPS[level] if level > 0).to_s
+            else
+                group(val / 10, level+1) + ' ' + group(val % 10, level)
+            end
+        end
+    end
+end
+
+
+class USEnglishTranslator
+    # Formal, US English. Optional 'and'. Will not produce things
+    # such as 'twelve hundred' but rather 'one thousand two hundred'.
+    # The use of 'and' is incomplete; it is sometimes missed.
+
+    DIGITS = %w(zero one two three four five six seven eight nine)
+    TEENS  = %w(ten eleven twelve thirteen fourteen fifteen sixteen
+                seventeen eighteen nineteen)
+    TENS   = %w(hello world twenty thirty forty fifty sixty seventy
+                eighty ninety)
+    GROUPS = %w(thousand million billion trillion quadrillion
+                quintillion sextillion septillion octillion nonillion
+                decillion)
+    K = 1000
+
+    def initialize(conjunction = true)
+        @conjunction = conjunction
+    end
+
+    def to_spoken(val)
+        case val <=> 0
+        when -1
+            'negative ' + to_spoken(-val)
+        when 0
+            DIGITS[0]
+        else
+            group(val, 0).flatten.join(' ')
+        end
+    end
+
+    private
+
+    def group(val, level)
+        x = group(val / K, level + 1) << GROUPS[level] if val >= K
+        x.to_a << under_1000(val % K, level)
+    end
+
+    def under_1000(val, level)
+        x = [DIGITS[val / 100]] << 'hundred' if val >= 100
+        x.to_a << under_100(val % 100, (level == 0 and not x.nil?))
+    end
+
+    def under_100(val, junction)
+        x = [('and' if @conjunction and junction)]    # wyf?
+        case val
+        when 0
+            []
+        when 1...10
+            x << DIGITS[val]
+        when 10...20
+            x << TEENS[val - 10]
+        else
+            d = val % 10
+            x << (TENS[val / 10] + ('-' + DIGITS[d] if d != 0).to_s)
+        end
+    end
+end
+
+
+class Integer
+    def to_spoken(translator = USEnglishTranslator.new)
+        translator.to_spoken(self).squeeze(' ').strip
+    end
+end
+
+if $0 == __FILE__
+    SAMPLES = [ 0, 1, 2, 5, 10, 11, 14, 18, 20, 21, 29, 33, 42, 50, 87, 99,
+                100, 101, 110, 167, 199, 200, 201, 276, 300, 314, 500, 610,
+                1000, 1039, 1347, 2309, 3098, 23501, 32767, 70000, 5480283,
+                2435489238, 234100090000, -42, -2001 ]
+
+    TRANSLATORS = { 'US English' => USEnglishTranslator.new,
+                    'Japanese'   => JapaneseTranslator.new }
+
+
+    # main
+    TRANSLATORS.each do |lang, translator|
+        puts
+        puts lang
+        puts '-' * lang.length
+        SAMPLES.each do |val|
+            puts "%12d => %s" % [val, val.to_spoken(translator)]
+        end
+    end
+end

metadata

@@ -3,8 +3,8 @@ rubygems_version: 0.8.11
 specification_version: 1
 name: ferret
 version: !ruby/object:Gem::Version 
-  version: 0.1.3
-date: 2005-10-25 00:00:00 +09:00
+  version: 0.1.4
+date: 2005-11-01 00:00:00 +09:00
 summary: Ruby indexing library.
 require_paths: 
   - lib
@@ -164,6 +164,7 @@ files:
   - lib/ferret/utils/string_helper.rb
   - lib/ferret/utils/number_tools.rb
   - lib/ferret/utils/date_tools.rb
+  - lib/ferret/utils/thread_local.rb
   - test/test_helper.rb
   - test/test_all.rb
   - test/unit/ts_document.rb
@@ -180,6 +181,7 @@ files:
   - test/unit/utils/tc_date_tools.rb
   - test/unit/utils/tc_parameter.rb
   - test/unit/utils/tc_weak_key_hash.rb
+  - test/unit/utils/tc_thread.rb
   - test/unit/analysis/tc_lower_case_tokenizer.rb
   - test/unit/analysis/tc_lower_case_filter.rb
   - test/unit/analysis/tc_porter_stem_filter.rb
@@ -230,6 +232,9 @@ files:
   - test/longrunning/tm_store.rb
   - test/benchmark/tb_rw_vint.rb
   - test/benchmark/tb_ram_store.rb
+  - test/functional/thread_safety_index_test.rb
+  - test/functional/thread_safety_test.rb
+  - test/utils/number_to_spoken.rb
   - test/unit/analysis/data/wordfile
   - rake_utils/code_statistics.rb
 test_files: []