google-video 0.5.0 → 0.6.0

data/CHANGELOG

@@ -1,3 +1,24 @@
+* 2006/11/16
+
+- [shaper] renamed Video#yesterday_view_count to 
+  Video#view_count_yesterday for consistency, and aliased to maintain
+  backwards-compatibility.
+
+- [shaper] Updated to reflect video detail page edits:
+  + parse new metrics on email and embed count/rank for today and yesterday
+  + correct parsing to accurately grab from new layout of rank and view 
+    count for today and yesterday
+  + turned Video#rank_delta into a method that returns the delta between
+    yesterday and today's rank.
+
+- [shaper] Updated hpricot usage for brevity: use Element#[] for referencing
+  attributes and Element#% for slicing out elements where we know there
+  can be only one.
+
+- [shaper] Fixed typo in rdoc for VideoSearchResponse.
+
+- [shaper] Optimized string building for URLs and the like.
+
 * 2006/11/07
 
 - [shaper] Initial version.

data/Rakefile

@@ -6,7 +6,7 @@ require 'rake/gempackagetask'
 
 spec = Gem::Specification.new do |s|
   s.name = 'google-video'
-  s.version = '0.5.0'
+  s.version = '0.6.0'
   s.author = 'Walter Korman'
   s.email = 'shaper@wgks.org'
   s.platform = Gem::Platform::RUBY

data/lib/google-video.rb

@@ -64,11 +64,18 @@ module GoogleVideo
     # only available via a details request: the current rank of this video on
     # the site.
     attr_reader :rank
-    
-    # only available via a details request: the change in rank this video has
-    # seen since last ranking: if positive, a move up, if negative, a move
-    # down.
-    attr_reader :rank_delta
+
+    # only available via a details request: the rank of this video on the site
+    # yesterday.
+    attr_reader :rank_yesterday
+
+    # only available via a details request: the current rank of this video on
+    # the site based on email metrics.
+    attr_reader :rank_email
+
+    # only available via a details request: the current rank of this video on
+    # the site based on embedded html metrics.
+    attr_reader :rank_embed
 
     # the number of ratings this video has received.
     attr_reader :rating_count
@@ -115,7 +122,18 @@ module GoogleVideo
     
     # only available via a details request: the number of views this video
     # received yesterday.
-    attr_reader :yesterday_view_count
+    attr_reader :view_count_yesterday
+
+    # same as Video#view_count_yesterday providing backwards-compatibility.
+    alias_method :yesterday_view_count, :view_count_yesterday
+
+    # only available via a details request: the number of views this video
+    # has received based on email metrics.
+    attr_reader :view_count_email
+
+    # only available via a details request: the number of views this video
+    # has received based on embedded html metrics.
+    attr_reader :view_count_embed
 
     # the default width in pixels for the video embed html.
     @@DEFAULT_WIDTH = 400
@@ -144,6 +162,13 @@ module GoogleVideo
  src="http://video.google.com/googleplayer.swf?docId=#{@doc_id}&hl=en" flashvars=""> </embed>
 edoc
     end
+        
+    # only available via a details request: the change in rank this video has
+    # seen since yesterday's rank: if positive, a move up, if negative, a move
+    # down.
+    def rank_delta
+      (@rank_yesterday - @rank)
+    end
   end
 
   # Provides a miniature snapshot of a point in time within a full video.  On
@@ -331,7 +356,7 @@ edoc
     end
   end
 
-  # Describes a response to a VideoSearchQuery.
+  # Describes a response to a VideoSearchRequest.
   class VideoSearchResponse
     # the url with which the request was made of the Google Video service.
     attr_reader :request_url
@@ -439,7 +464,7 @@ edoc
 
       # parse the overall search query stats
       regexp_stats = Regexp.new(/([0-9,]+) \- ([0-9,]+)<\/b> of about <b>([0-9,]+)<\/b> \(<b>([0-9.]+)/)
-      row = (doc/"#resultsheadertable/tr/td/font").first
+      row = (doc%"#resultsheadertable/tr/td/font")
       if !regexp_stats.match(row.inner_html)
         raise GoogleVideoException.new("failed to parse search query stats")
       end
@@ -450,18 +475,18 @@ edoc
       rows = doc/"table[@class='searchresult']/tr"
       rows.each do |row|
         # parse the thumbnail image
-        thumbnail_image_url = _decode_html((row/"img[@class='searchresultimg']").first.attributes['src'])
+        thumbnail_image_url = _decode_html((row%"img[@class='searchresultimg']")['src'])
 
         # parse the title and page url
-        a_title = (row/"div[@class='resulttitle']/a").first
-        page_url = 'http://' + @host + '/' + _decode_html(a_title.attributes['href'])
+        a_title = (row%"div[@class='resulttitle']/a")
+        page_url = "http://#{@host}/#{_decode_html(a_title['href'])}"
         title = _decode_html(a_title.inner_html.strip)
 
         # parse the description text
-        description = _decode_html((row/"div[@class='snippet']").first.inner_html.strip)
+        description = _decode_html((row%"div[@class='snippet']").inner_html.strip)
 
         # parse the upload username
-        span_channel = (row/"span[@class='channel']").first
+        span_channel = (row%"span[@class='channel']")
         channel_html = (span_channel) ? span_channel.inner_html : ''
         channel_html =~ /([^\-]+)/
         upload_user = _clean_string($1)
@@ -470,11 +495,11 @@ edoc
         star_count = _parse_star_elements(row/"img[@class='star']")
 
         # rating count
-        span_raters = (row/"span[@id='numOfRaters']").first
+        span_raters = (row%"span[@id='numOfRaters']")
         rating_count = (span_raters) ? span_raters.inner_html.to_i : 0
 
         # duration
-        span_date = (row/"span[@class='date']").first
+        span_date = (row%"span[@class='date']")
         date_html = span_date.inner_html
         date_html =~ /([^\-]+) \- (.*)$/
         duration = _clean_string($1)
@@ -519,20 +544,20 @@ edoc
       title = (doc/"div[@id='pvprogtitle']").inner_html.strip
 
       # parse description
-      font_description = (doc/"div[@id='description']/font").first
+      font_description = (doc%"div[@id='description']/font")
       description = (font_description) ? font_description.all_text.strip : ''
-      span_wholedescr = (doc/"span[@id='wholedescr']").first
+      span_wholedescr = (doc%"span[@id='wholedescr']")
       if (span_wholedescr)
         description += ' ' + span_wholedescr.all_text.strip
       end
       description = _decode_html(description)
       
       # parse star count
-      span_rating = (doc/"span[@id='communityRating']").first
+      span_rating = (doc%"span[@id='communityRating']")
       star_count = _parse_star_elements(span_rating/"img[@class='star']")
 
       # parse rating count
-      span_raters = (doc/"span[@id='numOfRaters']").first
+      span_raters = (doc%"span[@id='numOfRaters']")
       rating_count = (span_raters) ? span_raters.inner_html.to_i : 0
 
       # parse upload user, duration, upload date, upload user domain, upload
@@ -542,7 +567,7 @@ edoc
       # rows of text, with only the middle row (in the three-row scenario)
       # containing duration and upload date, omnipresent.  still, we buckle
       # down and have at it with fervor and tenacity.
-      duration_etc_html = (doc/"div[@id='durationetc']").first.inner_html
+      duration_etc_html = (doc%"div[@id='durationetc']").inner_html
       duration_parts = duration_etc_html.split(/<br[^>]+>/)
       # see if the first line looks like it has a date formatted ala 'Nov 9, 2006'
       if (duration_parts[0] =~ /\-  [A-Za-z]{3} \d+, \d{4}/)
@@ -563,7 +588,7 @@ edoc
 
       # parse the upload user url and domain if present
       if (upload_user_domain =~ /<a.*?href="([^"]+)"[^>]+>([^<]+)<\/a>/)
-        upload_user_url = 'http://' + @host + _decode_html(_clean_string($1))
+        upload_user_url = "http://#{@host}#{_decode_html(_clean_string($1))}"
         upload_user_domain = _clean_string($2)
       else
         upload_user_url = ''
@@ -571,28 +596,21 @@ edoc
       end
       
       # pull out view count and rank info table row elements
-      ( tr_total_views, tr_views_yesterday, tr_rank ) = (doc/"table[@id='statsall']/tr")
-
-      # parse view count
-      tr_total_views.inner_html =~ /<b>([^<]+)<\/b>/
-      view_count = _human_number_to_int($1)
-
-      # parse yesterday's view count
-      tr_views_yesterday.inner_html =~ /\s+([0-9,]+) yesterday/
-      yesterday_view_count = _human_number_to_int($1)
+      tr_statsall = (doc/"div[@id='statsall']/table/tr")
 
-      # parse rank
-      tr_rank.inner_html =~ /rank ([0-9,]+)/
-      rank = _human_number_to_int($1)
+      # remove the first row which just contains header info
+      tr_statsall.shift
 
-      # parse rank delta
-      (tr_rank/"span").first.inner_html =~ /\(([0-9\+\-,]+)\)/
-      rank_delta = _human_number_to_int($1)
+      # parse each of the view count and rank rows
+      (view_count, rank) = _parse_statsall_row(tr_statsall.shift)
+      (view_count_yesterday, rank_yesterday) = _parse_statsall_row(tr_statsall.shift)
+      (view_count_email, rank_email) = _parse_statsall_row(tr_statsall.shift)
+      (view_count_embed, rank_embed) = _parse_statsall_row(tr_statsall.shift)
 
       # pull out the url to the video .gvp file if prsent
-      img_download = (doc/"img[@src='/static/btn_download.gif']").first
+      img_download = (doc%"img[@src='/static/btn_download.gif']")
       if (img_download)
-        onclick_html = img_download.attributes['onclick']
+        onclick_html = img_download['onclick']
         onclick_script = _decode_html(onclick_html)
         onclick_script =~ /onDownloadClick\(([^\)]+)\)/
         video_file_url = onclick_script.split(",")[1].gsub(/"/, '')
@@ -608,7 +626,7 @@ edoc
       
       # pull out the playlist entries
       playlist_entries = []
-      table_upnext = (doc/"table[@id='upnexttable']").first
+      table_upnext = (doc%"table[@id='upnexttable']")
       (table_upnext/"tr").each do |tr_playlist|
         playlist_entries << _parse_playlist_entry(tr_playlist)
       end
@@ -619,7 +637,9 @@ edoc
                         :page_url => url,
                         :playlist_entries => playlist_entries,
                         :rank => rank,
-                        :rank_delta => rank_delta,
+                        :rank_yesterday => rank_yesterday,
+                        :rank_email => rank_email,
+                        :rank_embed => rank_embed,
                         :rating_count => rating_count,
                         :star_count => star_count,
                         :title => title,
@@ -630,7 +650,9 @@ edoc
                         :video_file_url => video_file_url,
                         :video_frame_thumbnails => video_frame_thumbnails,
                         :view_count => view_count,
-                        :yesterday_view_count => yesterday_view_count)
+                        :view_count_yesterday => view_count_yesterday,
+                        :view_count_email => view_count_email,
+                        :view_count_embed => view_count_embed)
 
       # build and return the response
       VideoDetailsResponse.new(:request_url => url, :video => video)
@@ -675,11 +697,11 @@ edoc
         rank_yesterday = td_rank_yesterday.inner_html.to_i
 
         # parse the video thumbnail image
-        thumbnail_image_url = _decode_html((td_thumbnail/"a/img").first.attributes['src'])
+        thumbnail_image_url = _decode_html((td_thumbnail%"a/img")['src'])
         
         # parse the detailed video info
-        a_video = (td_detail/"a").first
-        page_url = 'http://' + @host + a_video.attributes['href']
+        a_video = (td_detail%"a")
+        page_url = "http://#{@host}#{a_video['href']}"
 
         # title
         title = _decode_html(a_video.inner_html.strip)
@@ -688,11 +710,11 @@ edoc
         star_count = _parse_star_elements(td_detail/"div[@class='meta']/span/font/img[@class='star']")
 
         # rating count
-        span_raters = (td_detail/"div[@class='meta']/span/font/span[@id='numOfRaters']").first
+        span_raters = (td_detail%"div[@class='meta']/span/font/span[@id='numOfRaters']")
         rating_count = (span_raters) ? span_raters.inner_html.to_i : 0
 
         # duration
-        duration = (td_detail/"div[@class='meta']").first.all_text.gsub(/&nbsp;/, '').strip
+        duration = (td_detail%"div[@class='meta']").all_text.gsub(/&nbsp;/, '').strip
 
         # description
         description = _decode_html((td_detail).all_text.strip)
@@ -718,12 +740,31 @@ edoc
     end
 
     private
+
+    # Breakout method used by Client#video_details to make things more
+    # manageable, returning a tuple as (view count, rank) for the supplied
+    # data row within the "statsall" div element's statistics table.
+    def _parse_statsall_row (row)
+      td_alltime = row/"td"
+      view_count = _human_number_to_int(_parse_number(td_alltime[1].all_text))
+      rank = _human_number_to_int(_parse_number(td_alltime[2].all_text))
+      [ view_count, rank ]
+    end
+
+    # Given html containing some set of garbage and a human-readable number
+    # somewhere therein, e.g. "&nbsp; 554,200", returns an integer
+    # representing just the number.
+    def _parse_number (html)
+      html =~ /([0-9,\-]+)/
+      $1
+    end
+
     # Breakout method used by Client#video_details to make things more
     # manageable, returning a VideoFrameThumbnail constructed from the given
     # image element.
     def _parse_video_frame_thumbnail (frame_image)
-      title = frame_image.attributes['title']
-      thumbnail_image_url = _decode_html(frame_image.attributes['src'])
+      title = frame_image['title']
+      thumbnail_image_url = _decode_html(frame_image['src'])
       VideoFrameThumbnail.new(:title => title, :thumbnail_image_url => thumbnail_image_url)
     end
 
@@ -735,15 +776,15 @@ edoc
       ( td_thumbnail, td_playlist ) = (tr_playlist/"td")
 
       # parse thumbnail image
-      thumbnail_image_url = _decode_html((td_thumbnail/"img").first.attributes['src'])
+      thumbnail_image_url = _decode_html((td_thumbnail%"img")['src'])
 
       # parse the page url and title
-      a_page = (td_playlist/"a").first
-      page_url = 'http://' + @host + a_page.attributes['href']
-      title = _decode_html(a_page.attributes['title'])
+      a_page = (td_playlist%"a")
+      page_url = "http://#{@host}#{a_page['href']}"
+      title = _decode_html(a_page['title'])
 
       # parse the upload user and duration
-      meta_html = (td_playlist/"span[@class='meta']").first.inner_html
+      meta_html = (td_playlist%"span[@class='meta']").inner_html
       if (meta_html =~ /([^<]+)<br \/>(.*)/)
         upload_user = _clean_string($1)
         duration = _clean_string($2)
@@ -774,7 +815,7 @@ edoc
     def _parse_star_elements (elements)
       star_count = 0
       elements.each do |star|
-        star_src = star.attributes['src']
+        star_src = star['src']
         if (star_src =~ /starLittle\.gif$/)
           star_count += 1
         elsif (star_src =~ /starLittleHalf\.gif$/)
@@ -793,7 +834,7 @@ edoc
       url = "http://#{@host}/videoranking"
       if top_request.country
         country_code = TopVideosRequest::COUNTRIES[top_request.country]
-        url += "?cr=#{country_code}"
+        url.concat "?cr=#{country_code}"
       end
       url
     end
@@ -801,17 +842,17 @@ edoc
     def _search_url (search_request)
       query = search_request.query
       if (search_request.duration && search_request.duration != 'all')
-        query += '+duration:' + search_request.duration
+        query.concat "+duration:#{search_request.duration}"
       end
 
       url = "http://#{@host}/videosearch?q=#{URI.encode(query)}"
 
       if (search_request.page)
-        url += '&page=' + URI.encode(search_request.page.to_s)
+        url.concat "&page=#{URI.encode(search_request.page.to_s)}"
       end
 
       if (search_request.sort)
-        url += '&so=' + URI.encode(search_request.sort)
+        url.concat "&so=#{URI.encode(search_request.sort)}"
       end
 
       url

data/test/test_video_details.rb

@@ -87,10 +87,19 @@ class TestVideoDetails < Test::Unit::TestCase
 
     # poke and prod the video object
     assert_valid_video video
-    assert (video.view_count >= video.yesterday_view_count)
+    assert (video.view_count >= video.view_count_yesterday)
+    assert (video.view_count >= video.view_count_embed)
+    assert (video.view_count >= video.view_count_email)
     assert_match /http:\/\/video\.google\.com\/videogvp\/[^\.]+\.gvp\?docid=[0-9\-\+]+/, video.video_file_url
     assert_kind_of Numeric, video.rank
-    assert_kind_of Numeric, video.rank_delta
+    assert_kind_of Numeric, video.rank_yesterday
+    assert_kind_of Numeric, video.rank_email
+    assert_kind_of Numeric, video.rank_embed
+    assert_kind_of Numeric, video.view_count
+    assert_kind_of Numeric, video.view_count_yesterday
+    assert_kind_of Numeric, video.view_count_email
+    assert_kind_of Numeric, video.view_count_embed
+    assert_equal video.view_count_yesterday, video.yesterday_view_count
     assert_instance_of Time, video.upload_date
 
     # check the video frame thumbnails

metadata

@@ -3,8 +3,8 @@ rubygems_version: 0.9.0
 specification_version: 1
 name: google-video
 version: !ruby/object:Gem::Version 
-  version: 0.5.0
-date: 2006-11-09 00:00:00 -08:00
+  version: 0.6.0
+date: 2006-11-16 00:00:00 -08:00
 summary: A Ruby object-oriented interface to Google Video content.
 require_paths: 
 - lib