Update for a few new paths

TechAndCheck · Sep 12, 2024 · 578ed82 · 578ed82
1 parent 9ebc1e3
commit 578ed82
Show file tree

Hide file tree

Showing 5 changed files with 48 additions and 19 deletions.
diff --git a/Gemfile.lock b/Gemfile.lock
@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    forki (0.2.5)
+    forki (0.2.8)
       apparition
       capybara
       oj
@@ -35,14 +35,15 @@ GEM
       capybara (~> 3.13, < 4)
       websocket-driver (>= 0.6.5)
     ast (2.4.2)
-    bigdecimal (3.1.5)
+    base64 (0.2.0)
+    bigdecimal (3.1.8)
     builder (3.2.4)
     byebug (11.1.3)
-    capybara (3.39.2)
+    capybara (3.40.0)
       addressable
       matrix
       mini_mime (>= 0.1.3)
-      nokogiri (~> 1.8)
+      nokogiri (~> 1.11)
       rack (>= 1.6.0)
       rack-test (>= 0.6.3)
       regexp_parser (>= 1.5, < 3.0)
@@ -54,7 +55,7 @@ GEM
     erubi (1.12.0)
     ethon (0.16.0)
       ffi (>= 1.15.0)
-    ffi (1.16.3)
+    ffi (1.17.0-arm64-darwin)
     i18n (1.13.0)
       concurrent-ruby (~> 1.0)
     json (2.6.3)
@@ -67,12 +68,12 @@ GEM
     minitest (5.18.0)
     nokogiri (1.15.1-arm64-darwin)
       racc (~> 1.4)
-    oj (3.16.3)
+    oj (3.16.4)
       bigdecimal (>= 3.0)
     parallel (1.23.0)
     parser (3.2.2.1)
       ast (~> 2.4.1)
-    public_suffix (5.0.4)
+    public_suffix (5.1.1)
     racc (1.6.2)
     rack (2.2.4)
     rack-test (2.1.0)
@@ -129,7 +130,8 @@ GEM
       rubocop-rails (~> 2.0)
     ruby-progressbar (1.13.0)
     rubyzip (2.3.2)
-    selenium-webdriver (4.16.0)
+    selenium-webdriver (4.21.1)
+      base64 (~> 0.2)
       rexml (~> 3.2, >= 3.2.5)
       rubyzip (>= 1.2.2, < 3.0)
       websocket (~> 1.0)

diff --git a/lib/forki/scrapers/post_scraper.rb b/lib/forki/scrapers/post_scraper.rb
@@ -204,7 +204,11 @@ def extract_video_post_data(graphql_strings)
       begin
         feedback_object = story_node_object["comet_sections"]["feedback"]["story"]["feedback_context"]["feedback_target_with_context"]["ufi_renderer"]["feedback"]
       rescue NoMethodError
-        feedback_object = story_node_object["comet_sections"]["feedback"]["story"]["comet_feed_ufi_container"]["story"]["story_ufi_container"]["story"]["feedback_context"]["feedback_target_with_context"]
+        begin
+          feedback_object = story_node_object["comet_sections"]["feedback"]["story"]["comet_feed_ufi_container"]["story"]["story_ufi_container"]["story"]["feedback_context"]["feedback_target_with_context"]
+        rescue NoMethodError
+          feedback_object = story_node_object["comet_sections"]["feedback"]["story"]["story_ufi_container"]["story"]["feedback_context"]["feedback_target_with_context"]
+        end
       end
 
       if feedback_object["comet_ufi_summary_and_actions_renderer"]["feedback"].key?("cannot_see_top_custom_reactions")
@@ -315,8 +319,14 @@ def extract_image_post_data(graphql_object_array)
         # TODO: These two branches are *super* similar, probably a lot of overlap
         attachments = graphql_object["node"]["comet_sections"]["content"]["story"]["attachments"]
 
-        if graphql_object["node"]["comet_sections"]["feedback"]["story"].key?("feedback_context")
-          feedback_object = graphql_object["node"]["comet_sections"]["feedback"]["story"]["feedback_context"]["feedback_target_with_context"]["ufi_renderer"]["feedback"]["comet_ufi_summary_and_actions_renderer"]["feedback"]
+        if graphql_object["node"]["comet_sections"]["feedback"]["story"].has_key?("story_ufi_container")
+          feedback_object = graphql_object["node"]["comet_sections"]["feedback"]["story"]["story_ufi_container"]["story"]["feedback_context"]["feedback_target_with_context"]["comet_ufi_summary_and_actions_renderer"]["feedback"]
+        elsif graphql_object["node"]["comet_sections"]["feedback"]["story"].dig("feedback_context")
+          begin
+            feedback_object = graphql_object["node"]["comet_sections"]["feedback"]["story"]["feedback_context"]["feedback_target_with_context"]["ufi_renderer"]["feedback"]["comet_ufi_summary_and_actions_renderer"]["feedback"]
+          rescue NoMethodError
+            debugger
+          end
         elsif graphql_object["node"]["comet_sections"]["feedback"]["story"].has_key?("comet_feed_ufi_container")
           feedback_object = graphql_object["node"]["comet_sections"]["feedback"]["story"]["comet_feed_ufi_container"]["story"]["story_ufi_container"]["story"]["feedback_context"]["feedback_target_with_context"]["comet_ufi_summary_and_actions_renderer"]["feedback"]
         else

diff --git a/lib/forki/scrapers/sieves/video_sieves/video_sieve_watch_tab.rb b/lib/forki/scrapers/sieves/video_sieves/video_sieve_watch_tab.rb
@@ -41,8 +41,12 @@ def self.check(graphql_objects)
   def self.sieve(graphql_objects)
     video_object = self.extractor(graphql_objects)
 
-    video_url = video_object["attachments"].first["media"]["browser_native_sd_url"]
-    video_preview_image_url = video_object["attachments"].first["media"]["preferred_thumbnail"]["image"]["uri"]
+    # video_url = video_object["attachments"].first["media"]["browser_native_sd_url"]
+    video_url = video_object["short_form_video_context"]["playback_video"]["browser_native_hd_url"]
+    video_url = video_object["short_form_video_context"]["playback_video"]["browser_native_sd_url"] if video_url.nil?
+
+    # video_preview_image_url = video_object["attachments"].first["media"]["preferred_thumbnail"]["image"]["uri"]
+    video_preview_image_url = video_object["short_form_video_context"]["video"]["first_frame_thumbnail"]
 
     if !video_object["feedback_context"].nil?
       feedback_object = video_object["feedback_context"]["feedback_target_with_context"]
@@ -51,16 +55,25 @@ def self.sieve(graphql_objects)
       feedback_object = feedback_object["feedback"] if feedback_object.has_key?("feedback")
     end
 
-    profile_link = video_object["attachments"].first["media"]["owner"]["url"]
+    begin
+      profile_link = video_object["attachments"].first["media"]["owner"]["url"]
+    rescue StandardError => e
+      profile_link = video_object["short_form_video_context"]["video_owner"]["url"]
+    end
+
     if profile_link.nil?
       filtered_json = graphql_objects.find { |go| go.has_key? "attachments" }
       profile_link = filtered_json["attachments"].first["media"]["creation_story"]["comet_sections"]["title"]["story"]["actors"].first["url"]
     end
 
-    if feedback_object.key?("cannot_see_top_custom_reactions")
-      reactions = feedback_object["cannot_see_top_custom_reactions"]["top_reactions"]["edges"]
-    else
-      reactions = feedback_object["top_reactions"]["edges"]
+    begin
+      if feedback_object.key?("cannot_see_top_custom_reactions")
+        reactions = feedback_object["cannot_see_top_custom_reactions"]["top_reactions"]["edges"]
+      else
+        reactions = feedback_object["top_reactions"]["edges"]
+      end
+    rescue StandardError => e
+      reactions = feedback_object["unified_reactors"]["count"]
     end
 
     post_details = {

diff --git a/lib/forki/version.rb b/lib/forki/version.rb
@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Forki
-  VERSION = "0.2.8"
+  VERSION = "0.2.9"
 end
diff --git a/test/post_test.rb b/test/post_test.rb
@@ -254,4 +254,8 @@ def test_a_pure_text_link
     assert !post.first.text.empty?
     assert_predicate post.first.text.length, :positive?
   end
+
+  def test_an_alternative_reel
+    Forki::Post.lookup("https://www.facebook.com/reel/2841887882618164")
+  end
 end