textquery 0.1.1 → 0.1.2

data/README.rdoc

@@ -9,6 +9,9 @@ Textquery is a simple PEG grammar with support for:
 - OR
 - NOT (- is an alias)
 - 'quoted strings'
+- fuzzy matching
+- case (in)sensitive
+- custom delimeters
 
 == Example
 
@@ -25,6 +28,14 @@ Textquery is a simple PEG grammar with support for:
   q.match?("b")                                             # => false
   q.match?("a b cdefg")                                     # => true
 
+  TextQuery.new("a~").match?("adf")                         # => true
+  TextQuery.new("~a").match?("dfa")                         # => true
+  TextQuery.new("~a~").match?("daf")                        # => true
+  TextQuery.new("2~a~1").match?("edaf")                     # => true
+  TextQuery.new("2~a~2").match?("edaf")                     # => false
+
+  TextQuery.new("a", :ignorecase => true).match?("A b cD")  # => true
+
 == License
 
 (The MIT License)

data/VERSION

@@ -1 +1 @@
-0.1.1
+0.1.2

data/lib/textquery/textquery.rb

@@ -8,7 +8,25 @@ end
 
 class WordMatch < Treetop::Runtime::SyntaxNode
   def eval(text, opt)
-    not text.match("^#{query}#{opt[:delim]}|#{opt[:delim]}#{query}#{opt[:delim]}|#{opt[:delim]}#{query}$|^#{query}$").nil?
+    fuzzy = query.match(/(\d)*(~)?([^~]+)(~)?(\d)*$/)
+
+    q = []
+    q.push "."                                    if fuzzy[2]
+    q.push fuzzy[1].nil? ? "*" : "{#{fuzzy[1]}}"  if fuzzy[2]
+    q.push fuzzy[3]
+    q.push "."                                    if fuzzy[4]
+    q.push fuzzy[5].nil? ? "*" : "{#{fuzzy[5]}}"  if fuzzy[4]
+    q = q.join
+
+    regex = "^#{q}#{opt[:delim]}|#{opt[:delim]}#{q}#{opt[:delim]}|#{opt[:delim]}#{q}$|^#{q}$"
+
+    if opt[:ignorecase]
+      regex = Regexp.new(regex, Regexp::IGNORECASE, 'utf8')
+    else
+      regex = Regexp.new(regex, nil, 'utf8')
+    end
+
+    not text.match(regex).nil?
   end
 
   def query

data/spec/textquery_spec.rb

@@ -160,14 +160,43 @@ describe TextQuery do
     q.eval("a b cdefg").should be_true
   end
 
+  it "should support fuzzy matching" do
+    parse("a~").eval("adf").should be_true
+    parse("~a").eval("dfa").should be_true
+    parse("~a~").eval("daf").should be_true
+
+    parse("1~a~1").eval("daf").should be_true
+    parse("2~a~1").eval("daf").should be_false
+    parse("1~a~2").eval("daf").should be_false
+
+    parse("~a~3").eval("daffy").should be_true
+    parse("a~1").eval("adf").should be_false
+    
+    parse("a~1 AND b").eval("adf b").should be_false
+    parse("a~2 AND b").eval("adf b").should be_true
+    parse("a~3 AND b").eval("adf b").should be_false
+  end
+
   it "should work on CJK text" do
-    JP = "に入れるわけにはいかないので、プラグインの出力が同一であることでもって同一性を判定する"
+    JP = "仕様変更は出し尽くしてしまいß"
 
-    q = TextQuery.new("に入".mb_chars, :delim => '')
+    q = TextQuery.new("変更", :delim => '')
     q.eval(JP).should be_true
-    q.eval("けにはい").should be_false
+    q.eval("変ま").should be_false
+    q.parse("は出").eval(JP).should be_true
+
+    q = TextQuery.new
+    q.parse("~出~").eval(JP).should be_true
+    q.parse("~出~ AND NOT ~尽~").eval(JP).should be_false
+    q.parse("~更は出~ OR ~尽く~").eval(JP).should be_true
+  end
+
+  it "should be case insensitive" do
+    TextQuery.new("a", :ignorecase => true).match?("A b cD").should be_true
+    TextQuery.new("a AND CD", :ignorecase => true).match?("A b cD").should be_true
 
-    q.parse("れるわ AND が同".mb_chars).eval(JP).should be_true
-    q.parse("れるわ AND NOT す".mb_chars).eval(JP).should be_false
+    TextQuery.new("a", :ignorecase => false).match?("A b cD").should be_false
+    TextQuery.new("a AND CD", :ignorecase => false).match?("A b cD").should be_false
   end
+
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: textquery
 version: !ruby/object:Gem::Version 
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors: 
 - Ilya Grigorik
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-12-29 00:00:00 -05:00
+date: 2009-12-30 00:00:00 -05:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency