atp_scraper 0.3.0 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: ee413b367b7003e9748268c3e6ebac13a5758db0
-  data.tar.gz: dc366cb110b4063407919d9eb782f9f59af4ba7c
+  metadata.gz: 3cef6acd9e40a30750c787843d0726e19d4f9ef9
+  data.tar.gz: 41574d18eceb43ea314c589ba581e0aaef5fc30e
 SHA512:
-  metadata.gz: ee1d39cc20915134812ce6d6f1781c3cf0a6eb11d878e3a4218df8b47ab502389c9c3acd23dfd144dc145836e4a434ad996757be4a067e7ed78aefa318573bd3
-  data.tar.gz: 89e81c1b0251913fcd810671425b6773fdb2173f7c14a81293f1715b883d05529e2794a4c5c51b49383a8ac6ba393a517394475b01583adb5738913a778b0d19
+  metadata.gz: 5aee0fc52a7da12114bd6d33b52d16623a48efdd25486f3de6a79e507674012d18b8d5af7afe949fe249d473383a3f6d3e5009b69d48ce151e09f7a68643462c
+  data.tar.gz: f42d99878a62911900e7c769fc5329cf25160aa3fe2d30aa8e473615290e2f1746d3f5970d5f85469f3e94bf19c377b814d3f25cd6367e88b81204d4046655e5

data/lib/atp_scraper/activity.rb

@@ -14,12 +14,17 @@
 module AtpScraper
   # Scrape activity data
   class Activity
-    def pickup_activity_data(activity_doc)
+    def initialize(html, html_charset = 'utf-8')
+      @activity_doc = AtpScraper::Html.parse(html, html_charset)
+      @player_name = pickup_player_name(@activity_doc)
+    end
+
+    def pickup_activity_data
       result = []
       player = {}
-      player[:name] = pickup_player_name(activity_doc)
+      player[:name] = @player_name
 
-      search_tournaments_doc(activity_doc).each do |tournament_doc|
+      search_tournaments_doc(@activity_doc).each do |tournament_doc|
         tournament = pickup_tournament_info(tournament_doc)
         player[:rank] = pickup_player_rank(tournament[:caption])
         search_records_doc(tournament_doc).each do |record_doc|
@@ -55,7 +60,8 @@ module AtpScraper
         tournament_location: tournament[:location],
         tournament_start_date: tournament[:date][:start],
         tournament_end_date: tournament[:date][:end],
-        tournament_surface: tournament[:surface]
+        tournament_surface: tournament[:surface],
+        tournament_surface_inout: tournament[:surface_inout]
       }
     end
 
@@ -87,13 +93,15 @@ module AtpScraper
 
     def pickup_tournament_info(tournament_doc)
       tournament_date = pickup_text(tournament_doc, ".tourney-dates")
+      surface = pickup_surface(tournament_doc)
       {
         name: pickup_text(tournament_doc, ".tourney-title"),
         location: pickup_text(tournament_doc, ".tourney-location"),
         date: divide_tournament_date(tournament_date),
         year: tournament_date[0, 4],
         caption: pickup_text(tournament_doc, ".activity-tournament-caption"),
-        surface: pickup_surface(tournament_doc)
+        surface: surface[:surface],
+        surface_inout: surface[:inout]
       }
     end
 
@@ -114,10 +122,19 @@ module AtpScraper
     end
 
     def pickup_surface(tournament_doc)
-      tournament_doc
+      surface = tournament_doc
         .css(".tourney-details")[1]
         .css(".item-details")
         .first.content.gsub(/\t|\s/, "")
+      divide_surface(surface)
+    end
+
+    def divide_surface(surface)
+      if (surface.match(/^Outdoor/))
+        return { surface: surface.gsub(/Outdoor/, ''), inout: "Outdoor" }
+      else
+        return { surface: surface.gsub(/Indoor/, ''), inout: "Indoor" }
+      end
     end
   end
 end

data/lib/atp_scraper/get.rb

@@ -3,16 +3,16 @@ module AtpScraper
   class Get
     def self.singles_ranking(rank_range = nil)
       request_uri = "/en/rankings/singles?rankRange=#{rank_range}"
-      ranking_doc = AtpScraper::Html.get_and_parse(request_uri)
-      ranking = AtpScraper::Ranking.new
-      ranking.pickup_ranking_data(ranking_doc)
+      ranking_html = AtpScraper::Html.get(request_uri)
+      ranking = AtpScraper::Ranking.new(ranking_html[:html], ranking_html[:charset])
+      ranking.pickup_ranking_data
     end
 
     def self.player_activity(player_id, year)
       request_uri = "/players/anything/#{player_id}/player-activity?year=#{year}"
-      activity_doc = AtpScraper::Html.get_and_parse(request_uri)
-      activity = AtpScraper::Activity.new
-      activity.pickup_activity_data(activity_doc)
+      activity_html = AtpScraper::Html.get(request_uri)
+      activity = AtpScraper::Activity.new(activity_html[:html], activity_html[:charset])
+      activity.pickup_activity_data
     end
   end
 end

data/lib/atp_scraper/html.rb

@@ -4,11 +4,6 @@ module AtpScraper
   # Get and parse html from atpworldtour.com
   class Html
     BASE = "http://www.atpworldtour.com"
-    def self.get_and_parse(uri)
-      html = get(uri)
-      parse(html[:html], html[:charset])
-    end
-
     def self.get(uri)
       charset = nil
       html = open(BASE + uri) do |f|

data/lib/atp_scraper/ranking.rb

@@ -9,9 +9,13 @@
 module AtpScraper
   # Scrape ranking data
   class Ranking
-    def pickup_ranking_data(ranking_doc)
+    def initialize(html, html_charset = 'utf-8')
+      @ranking_doc = AtpScraper::Html.parse(html, html_charset)
+    end
+
+    def pickup_ranking_data
       result = []
-      search_player_doc(ranking_doc).each do |player_doc|
+      search_player_doc(@ranking_doc).each do |player_doc|
         result.push(pickup_player_data(player_doc))
       end
       result

data/lib/atp_scraper/version.rb

@@ -1,3 +1,3 @@
 module AtpScraper
-  VERSION = "0.3.0"
+  VERSION = "0.4.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: atp_scraper
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: ruby
 authors:
 - mosuke5
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2016-02-19 00:00:00.000000000 Z
+date: 2016-02-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: nokogiri