Skip to content

Commit

Permalink
Remove duplicate repositories.
Browse files Browse the repository at this point in the history
refs #171.
Github assigns unique id(Repository#gh_id) to every repository.
1) Removed the duplicate repositories while maintaining the first as per creation date.
2) Migrated the associations to the original repository.
  • Loading branch information
prasadsurase committed Dec 12, 2016
1 parent c87ccbc commit 2afd194
Showing 1 changed file with 32 additions and 0 deletions.
32 changes: 32 additions & 0 deletions lib/tasks/utils.rake
Original file line number Diff line number Diff line change
Expand Up @@ -59,4 +59,36 @@ namespace :utils do
end
end

desc "Delete duplicate repositories"
task remove_dup_repos: :environment do
repos = Repository.pluck :gh_id
dup_repos = repos.select{|i| repos.count(i) > 1 }
dup_repos.each do |dup_repo|
dups = Repository.where(gh_id: dup_repo).asc(:created_at).to_a
original_repo = dups[0]
dups[1..-1].each do |dup|
# assign the users to the original repo
dup.users.each{|u| original_repo.users << u unless original_repo.users.include?(u) }

# assign the judges to the original repo
dup.judges.each{|a| original_repo.judges << a unless original_repo.judges.include?(a) }

# assign the commits to the original repo
dup.commits.each{|c| original_repo.commits << c unless original_repo.commits.include?(c) }

# assign the activities to the original repo
dup.activities.each{|a| original_repo.activities << a unless original_repo.activities.include?(a) }

# assign the code files to the original repo
dup.code_files.each{|a| original_repo.code_files << a unless original_repo.code_files.include?(a) }

# assign the repositories to the original repo
dup.repositories.each{|a| original_repo.repositories << a unless original_repo.repositories.include?(a) }

# soft delete the repository
dup.destroy
end
end
end

end

0 comments on commit 2afd194

Please sign in to comment.