Downloads a gem and dumps its documents into a Sqlite3 file

gem-to-sqlite.rb

# gem-to-sqlite.rb
#
# Usage: ruby gem-to-sqlite.rb <gem_name> <gem_version>
# Example: ruby gem-to-sqlite.rb rake 13.0.6

require 'bundler/inline'

gemfile do
  source 'https://rubygems.org'
  gem 'yard', '~> 0.9.28'
  gem 'sqlite3', '~> 1.4'
end

require 'yard'
require 'sqlite3'
require 'net/http'
require 'tmpdir'
require 'rubygems/package'

class GemToSqlite
  def initialize(gem_name, gem_version)
    @gem_name = gem_name
    @gem_version = gem_version
  end

  def download_gem
    gem_url = "https://rubygems.org/downloads/#{@gem_name}-#{@gem_version}.gem"
    uri = URI(gem_url)
    file_path = File.join(Dir.tmpdir, "#{@gem_name}-#{@gem_version}.gem")
    
    File.write(file_path, Net::HTTP.get(uri))
    file_path
  end

  def extract_gem(file_path)
    spec = nil
    gem_dir = File.join(Dir.tmpdir, "#{@gem_name}-#{@gem_version}")

    # Extract the gem to a temporary directory
    Gem::Package.new(file_path).extract_files(gem_dir)
    spec = Gem::Package.new(file_path).spec

    # Parse the gem documentation using YARD
    YARD::Registry.clear
    Dir.glob(File.join(gem_dir, '**', '*.rb')).each do |file|
      YARD::Parser::SourceParser.parse(file)
    end

    spec
  end

  def create_database
    db = SQLite3::Database.new("#{@gem_name}-#{@gem_version}.db")

    db.execute <<-SQL
      CREATE TABLE IF NOT EXISTS gem_spec (
        id INTEGER PRIMARY KEY,
        name TEXT,
        version TEXT
      );
    SQL

    db.execute <<-SQL
      CREATE TABLE IF NOT EXISTS gem_docs (
        id INTEGER PRIMARY KEY,
        path TEXT,
        type TEXT,
        doc TEXT
      );
    SQL

    db
  end

  def store_gem_spec(db, spec)
    db.execute("INSERT INTO gem_spec (name, version) VALUES (?, ?)", [@gem_name, @gem_version])
  end

  def store_docs(db)
    YARD::Registry.all.each do |object|
      path = object.path
      type = object.type.to_s
      doc = object.docstring.to_raw

      db.execute("INSERT INTO gem_docs (path, type, doc) VALUES (?, ?, ?)", [path, type, doc])
    end
  end

  def run
    gem_path = download_gem
    spec = extract_gem(gem_path)

    db = create_database
    store_gem_spec(db, spec)
    store_docs(db)

    puts "Documentation has been stored in #{@gem_name}-#{@gem_version}.db"
  end
end

if ARGV.length != 2
  puts "Usage: ruby gem-to-sqlite.rb <gem_name> <gem_version>"
  exit 1
end

gem_name = ARGV[0]
gem_version = ARGV[1]

gem_to_sqlite = GemToSqlite.new(gem_name, gem_version)
gem_to_sqlite.run