picky 1.5.2 → 1.5.3

data/lib/picky/indexed/indexes.rb

@@ -1,45 +1,61 @@
 module Indexed
-  
+
   # Registers the indexes held at runtime, for queries.
   #
   class Indexes
-    
+
     attr_reader :indexes, :index_mapping
-    
+
     each_delegate :load_from_cache,
                   :to => :indexes
-    
+
     def initialize
       clear
     end
-    
+
+    def to_s
+      indexes.indented_to_s
+    end
+
     # Clears the indexes and the mapping.
     #
     def clear
       @indexes       = []
       @index_mapping = {}
     end
-    
+
     # Reloads all indexes, one after another,
     # in the order they were added.
     #
     def reload
       load_from_cache
     end
-    
+
     # Registers an index with the indexes.
     #
     def register index
       self.indexes << index
       self.index_mapping[index.name] = index
     end
-    
+
+    # Load each index, and analyze it.
+    #
+    # Returns a hash with the findings.
+    #
+    def analyze
+      result = {}
+      self.indexes.each do |index|
+        index.analyze result
+      end
+      result
+    end
+
     # Extracts an index, given its identifier.
     #
     def [] identifier
       index_mapping[identifier.to_sym]
     end
-    
+
   end
-  
+
 end

data/lib/picky/indexing/indexes.rb

@@ -1,11 +1,11 @@
 module Indexing
-  
+
   # Registers the indexes held at index time, for indexing.
   #
   class Indexes
-    
+
     attr_reader :indexes
-    
+
     each_delegate :take_snapshot,
                   :generate_caches,
                   :backup_caches,
@@ -14,57 +14,59 @@ module Indexing
                   :clear_caches,
                   :create_directory_structure,
                   :to => :indexes
-    
+
     def initialize
       clear
     end
-    
+
+    def to_s
+      indexes.indented_to_s
+    end
+
     # Clears the array of indexes.
     #
     def clear
       @indexes = []
     end
-    
+
     # Registers an index with the indexes.
     #
     def register index
       self.indexes << index
     end
-    
+
     # Runs the indexers in parallel (index + cache).
     #
     # TODO Spec.
     #
     def index randomly = true
       take_snapshot
-      
+
       # Run in parallel.
       #
-      timed_exclaim "INDEXING USING #{Cores.max_processors} PROCESSORS, IN #{randomly ? 'RANDOM' : 'GIVEN'} ORDER."
-      
-      # TODO Think about having serial work units.
+      timed_exclaim "Indexing using #{Cores.max_processors} processors, in #{randomly ? 'random' : 'given'} order."
+
+      # Run indexing/caching forked.
       #
       Cores.forked self.indexes, { randomly: randomly } do |an_index|
         an_index.index
-      # TODO
-      # end
-      # Cores.forked self.indexes, { randomly: randomly } do |an_index|
         an_index.cache
       end
-      timed_exclaim "INDEXING FINISHED."
+
+      timed_exclaim "Indexing finished."
     end
-    
+
     # For integration testing – indexes for the tests without forking and shouting ;)
     #
     def index_for_tests
       take_snapshot
-      
+
       self.indexes.each do |an_index|
         an_index.index
         an_index.cache
       end
     end
-    
+
     # Generate only the index for the given index:category pair.
     #
     def generate_index_only index_name, category_name = nil
@@ -77,23 +79,23 @@ module Indexing
       found = find index_name, category_name
       found.generate_caches if found
     end
-    
+
     # Find a given index:category pair.
     #
     def find index_name, category_name
       index_name = index_name.to_sym
-      
+
       indexes.each do |index|
         next unless index.name == index_name
-        
+
         return index unless category_name
-        
+
         found = index.categories.find category_name
         return found if found
       end
-      
+
       raise %Q{Index "#{index_name}" not found. Possible indexes: "#{indexes.map(&:name).join('", "')}".}
     end
-    
+
   end
 end

data/lib/picky/interfaces/live_parameters.rb

@@ -9,12 +9,12 @@ module Interfaces
   # Important Note: This will only work in Master/Child configurations.
   #
   class LiveParameters
-    
+
     def initialize
       @child, @parent = IO.pipe
       start_master_process_thread
     end
-    
+
     # This runs a thread that listens to child processes.
     #
     def start_master_process_thread
@@ -30,18 +30,19 @@ module Interfaces
           exclaim "Trying to update MASTER configuration."
           try_updating_configuration_with configuration_hash
           kill_each_worker_except pid
-        # TODO rescue on error.
-          
+
+          # Fails hard on an error.
+          #
         end
       end
     end
-    
+
     # TODO This needs to be webserver agnostic.
     #
     def worker_pids
       Unicorn::HttpServer::WORKERS.keys
     end
-    
+
     # Taken from Unicorn.
     #
     def kill_each_worker_except pid
@@ -61,14 +62,14 @@ module Interfaces
     def remove_worker wpid
       worker = Unicorn::HttpServer::WORKERS.delete(wpid) and worker.tmp.close rescue nil
     end
-    
+
     # Updates any parameters with the ones given and
     # returns the updated params.
     #
     # The params are a strictly defined hash of:
     #   * querying_removes_characters: Regexp
     #   * querying_stopwords:          Regexp
-    #   TODO etc.
+    #   * querying_splits_text_on:     Regexp
     #
     # This first tries to update in the child process,
     # and if successful, in the parent process
@@ -103,7 +104,7 @@ module Interfaces
     def close_child
       @child.close unless @child.closed?
     end
-    
+
     class CouldNotUpdateConfigurationError < StandardError
       attr_reader :config_key
       def initialize config_key, message
@@ -111,9 +112,9 @@ module Interfaces
         @config_key = config_key
       end
     end
-    
+
     # Tries updating the configuration in the child process or parent process.
-    # 
+    #
     def try_updating_configuration_with configuration_hash
       current_key = nil
       begin
@@ -128,7 +129,7 @@ module Interfaces
         raise CouldNotUpdateConfigurationError.new current_key, e.message
       end
     end
-    
+
     def extract_configuration
       {
         querying_removes_characters: querying_removes_characters,
@@ -136,7 +137,7 @@ module Interfaces
         querying_splits_text_on:     querying_splits_text_on
       }
     end
-    
+
     # TODO Move to Interface object.
     #
     def querying_removes_characters
@@ -157,11 +158,17 @@ module Interfaces
     def querying_splits_text_on= new_value
       Tokenizers::Query.default.instance_variable_set(:@splits_text_on_regexp, %r{#{new_value}})
     end
-    
+
+    #
+    #
+    def to_s
+      "Suckerfish Live Interface (Use the picky-live gem to introspect)"
+    end
+
   end
-  
+
   # Aka.
   #
   ::LiveParameters = LiveParameters
-  
+
 end

data/lib/picky/internals/extensions/object.rb

@@ -1,22 +1,29 @@
 class Object # :nodoc:all
-  
+
   # Puts a text in the form:
   #   12:34:56: text here
   #
   def timed_exclaim text
     exclaim "#{Time.now.strftime("%H:%M:%S")}: #{text}"
   end
-  
+
   # Just puts the given text.
   #
   def exclaim text
     puts text
   end
-  
+
   # Puts a text that informs the user of a missing gem.
   #
-  def puts_gem_missing gem_name, message
-    puts "#{gem_name} gem missing!\nTo use #{message}, you need to:\n  1. Add the following line to Gemfile:\n     gem '#{gem_name}'\n  2. Then, run:\n     bundle update\n"
+  def warn_gem_missing gem_name, message
+    warn "#{gem_name} gem missing!\nTo use #{message}, you need to:\n  1. Add the following line to Gemfile:\n     gem '#{gem_name}'\n  2. Then, run:\n     bundle update\n"
   end
-  
+
+  # Indents each line by <tt>amount=2</tt> spaces.
+  #
+  def indented_to_s amount = 2
+    ary = self.respond_to?(:join) ? self : self.to_s.split("\n")
+    ary.map { |s| "#{" "*amount}#{s}"}.join("\n")
+  end
+
 end

data/lib/picky/internals/frontend_adapters/rack.rb

@@ -7,18 +7,8 @@ module Internals
     # TODO Rename to Routing again. Push everything back into appropriate Adapters.
     #
     class Rack # :nodoc:all
-  
-      @@defaults = {
-        query_key:    'query'.freeze,
-        offset_key:   'offset'.freeze,
-        content_type: 'application/octet-stream'.freeze # TODO Wrong.
-      }
-  
-      def initialize
-        @defaults = @@defaults.dup
-      end
-  
-      # 
+
+      #
       #
       def reset_routes
         @routes = ::Rack::Mount::RouteSet.new
@@ -29,13 +19,13 @@ module Internals
       def finalize
         routes.freeze
       end
-  
+
       # Routing simply delegates to the route set to handle a request.
       #
       def call env
         routes.call env
       end
-  
+
       # API method.
       #
       def route options = {}
@@ -61,7 +51,7 @@ module Internals
       end
       def route_one url, query, route_options = {}
         raise RouteTargetNilError.new(url) unless query
-        routes.add_route Internals::Adapters::Rack.app_for(query, route_options), default_options(url, route_options)
+        routes.add_route Internals::Adapters::Rack.app_for(query, route_options), default_options(url, route_options), {}, query.to_s
       end
       class RouteTargetNilError < StandardError
         def initialize url
@@ -81,9 +71,9 @@ module Internals
       def default status
         answer nil, STATUSES[status]
       end
-  
-  
-  
+
+
+
       # TODO Can Rack handle this for me?
       #
       # Note: Rack-mount already handles the 404.
@@ -92,21 +82,21 @@ module Internals
         200 => lambda { |_| [200, { 'Content-Type' => 'text/html', 'Content-Length' => '0' }, ['']] },
         404 => lambda { |_| [404, { 'Content-Type' => 'text/html', 'Content-Length' => '0' }, ['']] }
       }
-  
+
       #
       #
       def default_options url, route_options = {}
         url = normalized url
-    
+
         options = { request_method: 'GET' }.merge route_options
-    
+
         options[:path_info] = url if url
-    
+
         options.delete :content_type
-      
+
         query_params = options.delete :query
         options[:query_string] = %r{#{generate_query_string(query_params)}} if query_params
-    
+
         options
       end
       #
@@ -117,38 +107,44 @@ module Internals
         k, v = query_params.first
         "#{k}=#{v}"
       end
-    
+
       # Setup a route that answers using the given app.
       #
       def answer url = nil, app = nil
         routes.add_route (app || STATUSES[200]), default_options(url)
       end
-  
+
       # Returns a regular expression for the url if it is given a String-like object.
       #
       def normalized url
         url.respond_to?(:to_str) ? %r{#{url}} : url
       end
-  
+
       # Returns true if there are no routes defined.
       #
       def empty?
         routes.length.zero?
       end
-  
-      # TODO Beautify.
+
+      # TODO Beautify. Rewrite!
       #
       def to_s
+        max_length = routes.instance_variable_get(:@routes).reduce(0) do |current_max, route|
+          route_length = route.conditions[:path_info].source.to_s.size
+          route_length > current_max ? route_length : current_max
+        end
+        "Note: Anchored (\u2713) regexps are faster, e.g. /\\A.*\\Z/ or /^.*$/.\n\n" +
         routes.instance_variable_get(:@routes).map do |route|
           path_info = route.conditions[:path_info]
           anchored = ::Rack::Mount::Utils.regexp_anchored?(path_info)
           anchored_ok = anchored ? "\u2713" : " "
-          "#{anchored_ok}  #{path_info.source}"
-        end.join "\n"
+          source = path_info.source
+          "#{anchored_ok}  #{source.ljust(max_length)} => #{route.name}"
+        end.join("\n")
       end
-  
+
     end
-  
+
   end
-  
+
 end

data/lib/picky/internals/index/backend.rb

@@ -95,8 +95,7 @@ module Internals
         similarity.restore
         configuration.restore
       end
-    
-    
+      
       # Delete all index files.
       #
       def delete

data/lib/picky/internals/index/file/basic.rb

@@ -1,11 +1,11 @@
 module Internals
 
   module Index
-  
+
     # Handles all aspects of index files, such as dumping/loading.
     #
     module File
-    
+
       # Base class for all index files.
       #
       # Provides necessary helper methods for its
@@ -14,22 +14,26 @@ module Internals
       # dump/load methods.
       #
       class Basic
-      
+
         attr_reader :cache_path
-      
+
         # An index cache takes a path, without file extension,
         # which will be provided by the subclasses.
         #
         def initialize cache_path
           @cache_path = "#{cache_path}.#{extension}"
         end
-      
+
+        def to_s
+          cache_path
+        end
+
         # The default extension for index files is "index".
         #
         def extension
           :index
         end
-      
+
         # Will copy the index file to a location that
         # is in a directory named "backup" right under
         # the directory the index file is in.
@@ -49,7 +53,7 @@ module Internals
         def prepare_backup target
           FileUtils.mkdir target unless Dir.exists?(target)
         end
-      
+
         # Copies the file from its backup location back
         # to the original location.
         #
@@ -62,16 +66,16 @@ module Internals
           dir, name = ::File.split path
           ::File.join dir, 'backup', name
         end
-      
+
         # Deletes the file.
         #
         def delete
           `rm -Rf #{cache_path}`
         end
-      
+
         # Checks.
         #
-      
+
         # Is this cache file suspiciously small?
         # (less than 8 Bytes of size)
         #
@@ -91,11 +95,11 @@ module Internals
         def size_of path
           `ls -l #{path} | awk '{print $5}'`.to_i
         end
-      
+
       end
-    
+
     end
-  
+
   end
-  
+
 end

data/lib/picky/internals/index/files.rb

@@ -1,12 +1,12 @@
 module Internals
 
   module Index
-  
+
     class Files < Backend
-    
+
       def initialize bundle_name, config
         super bundle_name, config
-      
+
         # Note: We marshal the similarity, as the
         #       Yajl json lib cannot load symbolized
         #       values, just keys.
@@ -16,9 +16,19 @@ module Internals
         @similarity    = File::Marshal.new config.index_path(bundle_name, :similarity)
         @configuration = File::JSON.new    config.index_path(bundle_name, :configuration)
       end
-    
+
+      def to_s
+        <<-FILES
+Files:
+#{"Index:      #{@index}".indented_to_s}
+#{"Weights:    #{@weights}".indented_to_s}
+#{"Similarity: #{@similarity}".indented_to_s}
+#{"Config:     #{@configuration}".indented_to_s}
+FILES
+      end
+
     end
-  
+
   end
-  
+
 end

data/lib/picky/internals/index/redis/basic.rb

@@ -49,7 +49,10 @@ module Internals
         # Deletes the Redis index namespace.
         #
         def delete
-          # TODO @backend.
+          # Not implemented here.
+          # Note: backend.flushdb might be the way to go,
+          #       but since we cannot delete by key pattern,
+          #       we don't do anything.
         end
       
         # Checks.
@@ -58,19 +61,23 @@ module Internals
         # Is this cache suspiciously small?
         #
         def cache_small?
-          false # TODO
+          size < 1
         end
         # Is the cache ok?
         #
         # A small cache is still ok.
         #
         def cache_ok?
-          false # TODO
+          size > 0
         end
         # Extracts the size of the file in Bytes.
         #
-        def size_of path
-          # TODO
+        # Note: This is a very forgiving implementation.
+        #       But as long as Redis does not implement
+        #       DBSIZE KEYPATTERN, we are stuck with this.
+        #
+        def size
+          backend.dbsize
         end
       
       end

data/lib/picky/internals/index/redis.rb

@@ -8,8 +8,8 @@ module Internals
     
       def initialize bundle_name, config
         super bundle_name, config
-      
-        # TODO
+
+        # Refine a few Redis "types".
         #
         @index         = Redis::ListHash.new "#{config.identifier}:#{bundle_name}:index"
         @weights       = Redis::StringHash.new "#{config.identifier}:#{bundle_name}:weights"