github_repo_statistics 2.3.6 → 2.3.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: bcd986a058fd4af9648d2900297d60aaf9a8f53d76f2a58c6a50adbfb40411d0
-  data.tar.gz: 4e3c39e106d2e4bd0776f4806c0329a49e7279798c48fce3561d342cd47306c8
+  metadata.gz: 722258cfcbda685136e97c5b844f38b4da2937dbf3693b5131c628aad5131672
+  data.tar.gz: e0377225ed1e088b3eef9ff283ebf254225b36546eae619378f28df4f64cd800
 SHA512:
-  metadata.gz: 7fe19b7e6e44588af15ce682f7b2a0a0e3a219d5c4c4d7ba041e33ff57869adc1093f6219681329e16279680cd37913d3bb1815b00eebb97b16309c960cdc975
-  data.tar.gz: 5636c1a35d16b9cddcd0084bd41423833a3b607d9ebcad0338caa02e62e7a23e17a301e0a1b62ef3283612b75affcd4f113efec6bba8de333be0b0879547d575
+  metadata.gz: 37cc51e5dd9c5c40ee4b816fcbfa39e0c288fe6fa17c298890e6f6dbc508923dd094f2bd75969bc906695c69a6e5bec6d36ccdb380b8c1d9a455247b86ce2af4
+  data.tar.gz: 355b86a47bb3f55ac5426256a840e3e4a77320a093524f77c93627d73b08c53df3762a5c9c8bab34992f182ba7a74575806442933855858a4f17d245d3986cf2

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    github_repo_statistics (2.3.5)
+    github_repo_statistics (2.3.7)
       date
       faraday-retry
       google-cloud-bigquery

data/lib/github_repo_statistics/github_repo_statistics.rb

@@ -227,9 +227,10 @@ class GithubRepoStatistics
         team.match(/#{TEAM_REGEX}/)[0].upcase
       end.reject { |e| EXCLUSIONS&.include?(e) }
 
+      teams = calculate_percentile(teams, 90)
+
       total_changes += commit_count
       all_teams << teams
-      teams = teams.uniq
 
       if teams.count > 1
         files_changed_by_many_teams += 1
@@ -245,6 +246,30 @@ class GithubRepoStatistics
      file_team_map]
   end
 
+  def calculate_percentile(arr, percentile)
+    # Count occurrences of each unique element
+    counts = arr.each_with_object(Hash.new(0)) { |item, hash| hash[item] += 1 }
+
+    # Sort elements by their counts in descending order
+    sorted_counts = counts.sort_by { |k, v| -v }.to_h
+
+    # Calculate the cut-off for the percentile
+    total_count = arr.size
+    cutoff = total_count * (percentile / 100.0)
+
+    # Select elements that meet the percentile criteria
+    selected_elements = []
+    cumulative_count = 0
+
+    sorted_counts.each do |item, count|
+      cumulative_count += count
+      selected_elements << item
+      break if cumulative_count >= cutoff
+    end
+
+    selected_elements
+  end
+
   def filter_files(file_team_map:, size: BIG_FILE_SIZE)
     file_team_map.select do |file_path|
       next unless File.exist?(file_path)

data/lib/github_repo_statistics/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 class GithubRepoStatistics
-  VERSION = '2.3.6'
+  VERSION = '2.3.7'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: github_repo_statistics
 version: !ruby/object:Gem::Version
-  version: 2.3.6
+  version: 2.3.7
 platform: ruby
 authors:
 - Serghei Moret
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-05-28 00:00:00.000000000 Z
+date: 2024-06-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: date