math_metadata_lookup 0.1.1 → 0.1.2

data/README.md

@@ -5,6 +5,12 @@ This utility search mathematical reviews sites and fetches metadata about articl
 It returns results as one of text, xml, html, yaml or ruby formats.
 
 
+Installation
+------------
+
+    gem install math_metadata_lookup
+
+
 Command line usage example
 --------------------------
 
@@ -56,14 +62,48 @@ Hash arguments are:
 * article id is unknown 
   * **:title** String
   * **:authors** Array of strings
+  * **:year**
 
 Returns instance of class Result.
 
 
 #Lookup#author( hash )
 
+Search for authors "name forms".
+
 Hash arguments are:
 
 * **:name** String. Author name.
 
 Returns instance of class Result.
+
+
+#Lookup#heuristic( hash )
+
+Returns only one best match from each site where similarity is higher then threshold.
+It runs article searh with first two words from title and only surnames from author names.
+The result of search is sorted by similarity and articles with similarity less then threshold are deleted.
+Similarity is count as weighted average from title, authors and year using Levenshtein distance method.
+The Levenshtein distance function is run on full given title and full given names.
+
+Hash arguments are:
+
+* **:title**   String
+* **:author**  Array of strings
+* **:year**    String
+* **:threshold** Float. Range: 0.0...1.0. Default: 0.6
+
+Returns instance of class Result.
+
+
+#Lookup#reference( hash )
+
+Parse reference string and run heuristic.
+
+Hash arguments are:
+
+* **:reference**   String
+* **:threshold**   Float. Range: 0.0...1.0. Default: 0.6
+
+Returns instance of class Result.
+

data/TODO

@@ -2,9 +2,7 @@ Tasks
 =====
 
 o cache
-- heuristic
-  x joining results
-  - guess relevant article if possible
+o tests
 
 
 Completed
@@ -13,3 +11,7 @@ Completed
 * year argument
 * support for references
 * xml output
+* heuristic
+  x joining results
+  * guess relevant article if possible
+

data/bin/math_metadata_lookup

@@ -9,6 +9,7 @@ $:.unshift File.expand_path(File.join(File.dirname(__FILE__), "../lib"))
 def print_help
   puts "
   Search mathematical reviews sites and fetches metadata about articles.
+  https://github.com/pejuko/math_metadata_lookup
 
   #{$0} <command> <options>
 
@@ -30,7 +31,7 @@ def print_help
 
 
   heuristic -- returns only one best match from each site where
-               similarity is higher the threshold
+               similarity is higher then threshold
 
     --title, -t <title>
     --author, -a <author>       -- repeatable option
@@ -60,7 +61,6 @@ def print_help
 end
 
 require 'pp'
-require 'yaml'
 require 'rubygems'
 require 'math_metadata_lookup'
 

data/lib/math_metadata_lookup/result.rb

@@ -1,6 +1,8 @@
 # -*-: coding: utf-8 -*-
 # vi: fenc=utf-8:expandtab:ts=2:sw=2:sts=2
 
+require 'ya2yaml'
+
 module MathMetadata
 
   class Result
@@ -33,10 +35,11 @@ module MathMetadata
     def to_html
       result = ""
       @metadata.each do |site|
+        next unless site[:result]
         result << %~
 <div class="site">
     <h3>Site: #{site[:name]}</h3>~
-        site[:result].each do |entity|
+        site[:result].to_a.each do |entity|
           result << entity.to_html
         end
         result << %~</div>~
@@ -51,6 +54,7 @@ module MathMetadata
       result << %~<?xml version="1.0" encoding="utf-8"?>
 <mml>~
       @metadata.each do |site|
+        next unless site[:result]
         result << %~
     <site name="#{site[:name]}">~
         site[:result].each do |entity|
@@ -67,7 +71,7 @@ module MathMetadata
 
 
     def to_yaml
-      @metadata.to_yaml
+      @metadata.ya2yaml
     end
 
 

data/lib/math_metadata_lookup/site.rb

@@ -3,7 +3,6 @@
 
 require 'htmlentities'
 require 'open-uri'
-require 'i18n'
 require 'cgi'
 
 

data/lib/math_metadata_lookup/tools.rb

@@ -2,6 +2,7 @@
 # vi: fenc=utf-8:expandtab:ts=2:sw=2:sts=2
 
 require 'unicode'
+require 'unidecoder'
 
 module MathMetadata
 
@@ -50,7 +51,7 @@ module MathMetadata
     def normalize_name( name )
       # only latin chars
       trans = latex_to_utf8(name.to_s)
-      trans = I18n.transliterate(trans)
+      trans = trans.to_ascii
 
       # remove Jr. 
       trans.sub! %r{\bjr\.(\b|$)}i, ' '
@@ -74,8 +75,8 @@ module MathMetadata
 
 
     def normalize_text( s )
-      str = latex_to_utf8(s)
-      str = I18n.transliterate(str).downcase
+      str = latex_to_utf8(s.to_s)
+      str = str.to_ascii.downcase
       str = remove_punctuation(str)
       str.gsub!(%r{\W+}, ' ')
       str.gsub!(%r{(?:the|a|of|)\s+}i, ' ')

data/math_metadata_lookup.gemspec

@@ -13,13 +13,14 @@ spec = Gem::Specification.new do |s|
   s.email = "pejuko@gmail.com"
   s.authors = ["Petr Kovar"]
   s.name = 'math_metadata_lookup'
-  s.version = '0.1.1'
+  s.version = '0.1.2'
   s.date = Time.now.strftime("%Y-%m-%d")
-  s.add_dependency('i18n', '>= 0.5.0')
   s.add_dependency('unicode')
+  s.add_dependency('unidecoder')
+  s.add_dependency('ya2yaml')
   s.require_path = 'lib'
   s.files = ["bin/math_metadata_lookup", "README.md", "math_metadata_lookup.gemspec", "TODO", "Rakefile"]
-  s.files += Dir["lib/**/*.rb"]
+  s.files += Dir["lib/**/*.rb", "resources/*"]
   s.executables = ["math_metadata_lookup"]
   s.description = <<EOF
 This utility/library search mathematical reviews sites and fetches metadata about articles.

data/resources/math_metadata_lookup.js

@@ -0,0 +1,4 @@
+function toggle_references( id ) {
+  var ref = document.getElementById( id );
+  ref.style.display = (obj.style.display == 'none') ? 'block' : 'none';
+}

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 1
-  - 1
-  version: 0.1.1
+  - 2
+  version: 0.1.2
 platform: ruby
 authors: 
 - Petr Kovar
@@ -14,11 +14,11 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-01-25 00:00:00 +01:00
+date: 2011-01-27 00:00:00 +01:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
-  name: i18n
+  name: unicode
   prerelease: false
   requirement: &id001 !ruby/object:Gem::Requirement 
     none: false
@@ -27,13 +27,11 @@ dependencies:
       - !ruby/object:Gem::Version 
         segments: 
         - 0
-        - 5
-        - 0
-        version: 0.5.0
+        version: "0"
   type: :runtime
   version_requirements: *id001
 - !ruby/object:Gem::Dependency 
-  name: unicode
+  name: unidecoder
   prerelease: false
   requirement: &id002 !ruby/object:Gem::Requirement 
     none: false
@@ -45,6 +43,19 @@ dependencies:
         version: "0"
   type: :runtime
   version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  name: ya2yaml
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        segments: 
+        - 0
+        version: "0"
+  type: :runtime
+  version_requirements: *id003
 description: |
   This utility/library search mathematical reviews sites and fetches metadata about articles.
   It can return results as one of text, xml, html, yaml or ruby formats.
@@ -73,6 +84,7 @@ files:
 - lib/math_metadata_lookup/reference.rb
 - lib/math_metadata_lookup/entity.rb
 - lib/math_metadata_lookup/author.rb
+- resources/math_metadata_lookup.js
 has_rdoc: true
 homepage: http://github.com/pejuko/math_metadata_lookup
 licenses: []