entityextractor 0.0.6 → 0.0.7

data/lib/extractdates.rb

@@ -16,16 +16,8 @@ class ExtractDates
         c = @text.chunk
         c.each do |i|
           s = paragraph(i).segment
-          dateExtract(s)
           s.each do |j|
-            shash = Hash.new
-            shash[:date] = dateExtract(j)
-            if shash[:date]
-              shash[:file] = file
-              shash[:title] = j
-              shash[:description] = i
-              @output.push(shash)
-            end
+            dateExtract(j, file, j, i)
           end
         end
       rescue
@@ -34,25 +26,58 @@ class ExtractDates
     return @output
   end
 
-  def dateExtract(blob)
-    # Date formats-
-      # mm/dd/yy
-      # mm/dd/yyyy
-      # Month dd, yyyy
-      # Month ddth, yyyy
-      # Month yyyy
-
-    # TOADD: 
-      # Multiple dates
-      # Year detection
-      # Conditional American dates
-      # Time ranges
-      # Filtering
-
+  def dateExtract(blob, file, title, description)
+    blobstring = blob.to_s
+    
     begin
-      return DateTime.parse(blob).to_s
+      if blobstring.match(/(\d{1,2})\/(\d{1,2})\/(\d{2,4})/)
+        save = blobstring.match(/(\d{1,2})\/(\d{1,2})\/(\d{2,4})/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      elsif blobstring.match(/(\d{1,2})-(\d{1,2})-(\d{2,4})/)
+        save = blobstring.match(/(\d{1,2})-(\d{1,2})-(\d{2,4})/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      elsif blobstring.match(/(.+?)(\w+ \d{1,2}(st|nd|rd|th|), \d{4})/)
+        save = blobstring.match(/(.+?)(\w+ \d{1,2}(st|nd|rd|th|), \d{4})/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      elsif blobstring.match(/(.+?) ((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) \d{2}(st|nd|rd|th|)( |\)|\]))/)
+        save = blobstring.match(/(.+?) ((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) \d{2}(st|nd|rd|th|)( |\)|\]))/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      elsif blobstring.match(/((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) [1-2][0,9]\d{2}( |\)|\]))/)
+        save = blobstring.match(/((?:Jan(?:uary)?|Feb(?:ruary)?|Mar(?:ch)?|Apr(?:il)?|May|Jun(?:e)?|Jul(?:y)?|Aug(?:ust)?|Sep(?:tember)?|Nov(?:ember)?|Dec(?:ember)?) [1-2][0,9]\d{2}( |\)|\]))/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      elsif blobstring.match(/(\d{4})-(\d{2})-(\d{2})/)
+        save = blobstring.match(/(\d{4})-(\d{2})-(\d{2})/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      elsif blobstring.match(/(\d{4})\/(\d{2})\/(\d{2})/)
+        save = blobstring.match(/(\d{4})\/(\d{2})\/(\d{2})/)
+        addItem(DateTime.parse(blob).to_s, file, title, description, blobstring, save.to_s)
+      end
     rescue
     end
   end
+
+  # Adds and item to the hash
+  def addItem(date, file, title, description, blob, regex)
+    shash = Hash.new
+    shash[:date] = date
+    shash[:file] = file
+    shash[:title] = title
+    shash[:description] = description
+
+    flag = 0
+    @output.each do |o|
+      if (o[:date] == shash[:date]) && (o[:file] == shash[:file]) && (o[:title].to_s == shash[:title].to_s)
+        flag = 1
+        break
+      end
+    end
+
+    if flag == 0
+      @output.push(shash)
+    end
+
+    blob.slice! regex
+    dateExtract(blob, file, title, description)
+  end
 end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: entityextractor
 version: !ruby/object:Gem::Version
-  version: 0.0.6
+  version: 0.0.7
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-03-16 00:00:00.000000000 Z
+date: 2014-03-18 00:00:00.000000000 Z
 dependencies: []
 description: Extracts entities and terms from any JSON.
 email: shidash@shidash.com