thomaspeklak-OfflineSearch 0.2.2

data/README

@@ -0,0 +1,100 @@
+==OfflineSearch
+OfflineSearch is a semantic offline search generator. It scans a directory of html files and generates a javascript search data file. This was primarly written for an offline html documentation and this is also the main target group. Of course it can be be useful on small websites, too.
+
+A frequency file of words can be outputted. This may help to find typos and gives statistical information, that can be used to tweak the stop word list. Maintaining the stop words is crucual to keep the index small. E.g. the index of 11MB of HTML files is about 1.2MB (including double metaphone data and if using a tweaked stop word list)
+
+The double metaphone algorithm can be used to find simlar terms if the search did not find any matches. This performs not to well, but an other (better) algorithm was not found that performed well on a large index.
+
+Default config files, stop word lists and search templates can be generated. Please see the section Usage for details.
+
+The search does not include any boolean logic (OR, NOT, XOR, ...). It always tries to find all specified terms.
+
+REMARK: Currently the search index is always written in UTF-8. Future releases will support other encodings.
+
+===Executable:
+ OfflineSearch
+
+===Usage:
+  OfflineSearch [options]
+    -c, --config=CONFIG_FILE         configuration file for the offline search
+  
+  Generators
+    -g, --generate-default-config    creates a default config file in the current directory
+    -w, --generate-default-stopwords creates a default stopword list in the current directory. Language flag is required.
+    -t, --generate-template          creates search template files in the current directory
+    -o, --generate-search-data       crawler the documents in the given docpath and generates the search data file
+  
+  Optional arguments
+  can also be specified in the config file
+  command line arguments will overwrite any given value in the config file
+    -d, --docpath=DOCPATH            path of the documents
+    -f=SEARCH_DATA_FILE              path and name of the search data file
+        --search-data-file
+    -s=STOPWORD_LIST                 stopword list, if none is specified the default stop word list is used
+        --stopword-list
+    -l, --language=LANGUAGE          required if you want to generate a default stopword list
+  
+    -h, --help                       Show this message
+
+===Config-File example:
+
+	language: english
+	storage: memory
+	crawler:
+	  docpath: ../docs
+	  docs: [html, htm]
+	  exceptions: 
+	  stopwords: 
+	  tags:
+      title: 150
+      h1: 50
+      h2: 25
+      h3: 18
+      h4: 13
+      h5: 11
+      h6: 9
+      strong: 7
+      b: 7
+      em: 5
+      i: 5
+      dt: 9
+      u: 4
+      a: 3
+	logger:
+	  file: STDOUT
+	  level: info
+	search_generator:
+	  search_data_file: search_data.js
+	  output_encoding: uft-8
+	  template: base
+	  relative_path_to_files: ../docs/
+	  output_frequency_to: frequency.txt
+    use_double_metaphone: true
+		
+===Templates===
+
+The shipped templates are very basic, as it is assumed that you integrate the search in your site. The only provide a guidance to how to implement the search on your site.
+
+
+___________________________________________________________________________
+The MIT License
+
+Copyright (c) 2008 Thomas Peklak
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in
+all copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+THE SOFTWARE.

data/bin/OfflineSearch

@@ -0,0 +1,3 @@
+#!/usr/local/bin/ruby -w
+
+require 'offline_search'

data/lib/action_controller.rb

@@ -0,0 +1,70 @@
+# ACTION CONTROLLER
+# checks which option is specified and executes the required scripts
+#
+# * $Author$
+# * $Rev$
+# * $LastChangedDate$
+#
+
+class ActionController
+  # checks the value of the global variable $action set by option parser
+  # valid actions are
+  # * genrating a default stopwords
+  # * genrating a default config file
+  # * genrating a template
+  # * genrating the search database
+  def initialize()
+    require "log_init"
+    unless defined?($action)
+      $logger.fatal("No action is defined. Please chose one of the following options:\n\t\t-o generate search index\n\t\t-t generate template files\n\t\t-o generate default stop words\n\t\t-g generate default config")
+      exit
+    end
+    case $action
+    when 'generate_default_stopwords'
+      generate_stopwords
+    when 'generate_default_config'
+      generate_config
+    when 'generate_template'
+			unless (['base','base+double_metaphone'].include?($config['template']))
+				$logger.error('Template not found')
+				$logger.info('Available templates: base,base+double_metaphone')
+				exit
+			end
+      generate_template
+    when 'generate_search'
+      verify_search_parameters
+      start_search
+    end
+  end
+  
+  private
+  def generate_stopwords
+		$logger.info("generating default stopwords")
+    require 'generate_default_stopwords'
+  end
+  def generate_config
+		$logger.info("generating default config")
+    require 'generate_default_config'
+  end
+  def generate_template
+		$logger.info("generating default template")
+    require 'generate_default_template'
+    TemplateGenerator.new($config['template'])
+  end
+  def verify_search_parameters
+    require 'option_validator'
+    OptionValidator.new
+  end
+  def start_search
+    require "stop_words"
+    require "crawler"
+    require "search_generator"
+
+    crawler = Crawler.new
+    crawler.find_files
+    crawler.parse_files
+
+    generator = SearchGenerator.new(crawler.get_stored_files, crawler.get_terms)
+    generator.generate
+  end
+end

data/lib/config.yaml

@@ -0,0 +1,31 @@
+language: german
+storage: memory
+crawler:
+  docpath: ../docs
+  docs: [html, htm]
+  exceptions: _dab
+  stopwords: 
+  tags:
+    title: 60
+    h1: 40
+    h2: 25
+    h3: 18
+    h4: 13
+    h5: 11
+    h6: 9
+    strong: 7
+    b: 7
+    em: 5
+    i: 5
+    dt: 9
+    u: 4
+    a: 3
+logger:
+  file: STDOUT
+  level: info
+search_generator:
+  search_data_file: search_data_js
+  output_encoding: uft-8
+  template: base
+  relative_path_to_files: D
+  output_frequency_to: frequency.txt

data/lib/config_default.yaml

@@ -0,0 +1,32 @@
+language: english
+storage: memory
+crawler:
+  docpath: ../docs
+  docs: [html, htm]
+  exceptions: 
+  stopwords: 
+  tags:
+    title: 60
+    h1: 40
+    h2: 25
+    h3: 18
+    h4: 13
+    h5: 11
+    h6: 9
+    strong: 7
+    b: 7
+    em: 5
+    i: 5
+    dt: 9
+    u: 4
+    a: 3
+logger:
+  file: STDOUT
+  level: info
+search_generator:
+  search_data_file: search_data.js
+  output_encoding: uft-8
+  template: base
+  relative_path_to_files: 
+  output_frequency_to: frequency.txt
+  use_double_metaphone: true

data/lib/crawler.rb

@@ -0,0 +1,237 @@
+# CRAWLER
+# searches dirctory for files
+# parses files for keywords, semantic keyword rank and pagerank
+#
+# * $Author$
+# * $Rev$
+# * $LastChangedDate$
+
+require 'rexml/document'
+require 'rubygems'
+require 'hpricot'
+require 'Kconv'
+require 'entity_converter'
+require 'filefinder'
+require 'progressbar'
+
+class Crawler
+  attr_writer :resource
+  # requires a docpath set in the config file and a temporary storage handler
+  def initialize
+    @resource = $config['crawler']['docpath']
+    require "temporary_storage"
+    @storage=Temporary_Storage.new($config['storage'])
+  end
+
+  # serach the given docpath for files with a valid extension and excludes files that should not be indexed
+  # returns an array of files
+  def find_files()
+    @files = FileFinder::find(@resource,:types=>$config['crawler']['docs'],:excludes=>$config['crawler']['exceptions'])
+    if (@files.empty?)
+      $logger.error('no files found in directory')
+      exit
+    end
+		@files_size = @files.length
+    @files
+  end
+
+  # takes an array of files an iterates through it. each file is read into a string and sent to a doccrawler for further processing
+  # if the file is a valid XHTML file, the file is processed with REXML otherwise Hpricot is used and a warning is written to the log
+  # no value is returned
+  def parse_files
+		i = 0
+		pbar = ProgressBar.new('indexing',@files_size)
+    @files.each do |file|
+      $logger.info("processing #{file}")
+      File.open(file,'r') do |f|
+				lines = f.read().gsub("\n",' ').gsub("\r",'')
+				#convert entities before a new Hpricot doc is created, otherwise the entities are not converted correctly
+				doc = HpricotCrawler.new(lines.decode_html_entities,file,@storage)
+        doc.crawler_and_store
+				i += 1
+				pbar.set(i)
+      end
+    end
+    @storage.calculate_pageranks_from_links
+  end
+
+  # returns a hash of the parsed documents
+  def get_stored_files
+    @storage.get_files
+  end
+
+  # returns a hash of the indexed terms with ranks and links to the documents
+  def get_terms
+    @storage.get_terms
+  end
+
+  ###### HELPER METHODS
+  private
+
+  # This abstract class parses a file and tries to extract semantic information
+  class DocCrawler
+    # tries to ignore external links an convert internal links
+    def resolve_link(link,dir)
+			link = File.basename(link)
+      case
+        when link =~ /^(http|ftp|mailto)/
+          return nil
+        when link =~ /^[\/a-zA-Z0-9_-]/
+          return (File.expand_path(dir+'/'+link)).gsub(@expanded_doc_path,'')
+        when link =~ /^\./
+          return (File.expand_path(dir+'/'+link)).gsub(@expanded_doc_path,'')
+        else
+          return nil
+      end
+    end
+
+    # method invokes other methods to get certain information about the document. These methods are implemented in the child classes
+    def crawler_and_store
+      @storage.store_file(resolve_link(@file,File.dirname(@file)),get_title)
+      @storage.store_link(get_hrefs)
+      split_and_store
+    end
+
+    private
+
+    # splits textblocks and stores terms in the storage. this method splits an all characters that are non aplhpa
+    def split_and_store()
+			numbers = '0'..'9'
+      get_texts.each do |text_block|
+        rank=text_block.semantic_value
+				unless (rank.nil?)
+					text_block.to_s.downcase.umlaut_to_downcase.decode_html_entities.split(/[^a-zäöüß0-9]+/).each do |term|
+						@storage.store_term(term,rank) unless ((term.size < 2 && !numbers.include?(term)) || $stop_words.has_key?(term))
+					end
+				end
+      end
+    end
+  end
+
+  # parses valid XHTML documents and extracts information
+  class XmlCrawler < DocCrawler
+    def initialize(lines,file,storage)
+      @file = file
+      @storage = storage
+      begin
+        @xml = REXML::Document.new(lines)
+      rescue REXML::ParseException
+        raise
+      end
+			@expanded_doc_path = File.expand_path($config['crawler']['docpath'])
+    end
+
+    private
+
+    # extracts and returns the title of the document
+    def get_title
+      @xml.elements.each('//head//title/text()')
+    end
+
+    # extracts all texts and returns an array of REXML::Texts if no block is given
+    # if a block is given then the texts are passed to it
+    def get_texts
+      texts=@xml.elements.each("//body//text()")
+      texts.delete_if { |t| t.to_s.lstrip.empty?}
+      if block_given?
+        yield texts
+      else
+        texts
+      end
+    end
+
+    # returns an array of internal links in the document
+    def get_hrefs
+      a=@xml.elements.to_a("//a")
+      href= Array.new
+      a.each do |anker|
+        link=resolve_link(anker.attributes.get_attribute('href').value,File.dirname(@file)) if anker.attributes.get_attribute('href')
+        href << link unless link.nil?
+      end
+      href
+    end
+
+    # extends REXML::Text with the functionaliy to extract semantic information
+    class REXML::Text
+      # stores an array of meaningful tags with their rank value
+      def self.store_semantics(tags)
+        @@semantic_tags=tags
+      end
+
+      # extracts the semantic value of a text block
+      def semantic_value
+        REXML::Text.store_semantics($config['crawler']['tags'].keys) unless defined?(@@semantic_tags)
+        rank = 1
+        node = parent
+				return nil if(node.name == 'script')
+        while @@semantic_tags.include?(node.name)
+          rank += $config['crawler']['tags'][node.name]
+          node = node.parent
+					return nil if(node.name == 'script')
+        end
+        rank
+      end
+    end
+  end
+
+  # parses non valid XHTML documents and extracts information
+  class HpricotCrawler < DocCrawler
+    def initialize(lines,file,storage)
+      @doc = Hpricot(lines)
+      @file = file
+      @storage = storage
+			@expanded_doc_path = File.expand_path($config['crawler']['docpath'])
+    end
+
+    private
+
+    # extracts and returns the title of the document
+    def get_title
+      @doc.at('//head/title').inner_text
+    end
+
+    # extracts all texts and returns an array of REXML::Texts if no block is given
+    # if a block is given then the texts are passed to it
+    def get_texts
+      texts = Array.new
+      @doc.traverse_text { |text|  texts << text unless text.to_s.strip.empty?}
+      if block_given?
+        yield texts
+      else
+        texts
+      end
+    end
+
+    # returns an array of internal links in the document
+    def get_hrefs
+      links = Array.new
+      (@doc/'a[@href]').each {|a|
+				href = a[:href]
+				links << resolve_link(a.get_attribute('href'),File.dirname(@file)) unless href[0,1] == '#' || href[0,10] == 'javascript'
+			}
+      links
+    end
+
+    # extends Hpricot::Text with the functionaliy to extract semantic information
+    class Hpricot::Text
+      # stores an array of meaningful tags with their rank value
+      def self.store_semantics(tags)
+        @@semantic_tags=tags
+      end
+
+      # extracts the semantic value of a text block
+      def semantic_value
+        Hpricot::Text.store_semantics($config['crawler']['tags'].keys) unless defined?(@@semantic_tags)
+        rank = 1
+        node = parent
+				return nil if(node.name == 'script')
+        while @@semantic_tags.include?(node.name)
+          rank += $config['crawler']['tags'][node.name]
+          node = node.parent
+					return nil if(node.name == 'script')
+        end
+        rank
+      end
+    end
+  end
+end