Class: DataDuck::Commands

Inherits:
Object
  • Object
show all
Defined in:
lib/dataduck/commands.rb

Defined Under Namespace

Classes: Namespace

Class Method Summary collapse

Class Method Details

.acceptable_commands ⇒ Object 



34
35
36
 
# File 'lib/dataduck/commands.rb', line 34

def self.acceptable_commands
  ['console', 'dbconsole', 'etl', 'quickstart', 'show']
end

.console ⇒ Object 



52
53
54
55
56
 
# File 'lib/dataduck/commands.rb', line 52

def self.console
  require "irb"
  ARGV.clear
  IRB.start
end

.dbconsole(where = "destination") ⇒ Object 



58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
 
# File 'lib/dataduck/commands.rb', line 58

def self.dbconsole(where = "destination")
  which_database = nil
  if where == "destination"
    which_database = DataDuck::Destination.only_destination
  elsif where == "source"
    which_database = DataDuck::Source.only_source
  else
    found_source = DataDuck::Source.source(where, true)
    found_destination = DataDuck::Destination.destination(where, true)
    if found_source && found_destination
      raise ArgumentError.new("Ambiguous call to dbconsole for #{ where } since there is both a source and destination named #{ where }.")
    end

    which_database = found_source if found_source
    which_database = found_destination if found_destination
  end

  if which_database.nil?
    raise ArgumentError.new("Could not find database '#{ where }'")
  end

  puts "Connecting to #{ where }..."
  which_database.dbconsole
end

.etl(what = nil) ⇒ Object 



83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
 
# File 'lib/dataduck/commands.rb', line 83

def self.etl(what = nil)
  if what.nil?
    puts "You need to specify a table name or 'all'. Usage: dataduck etl all OR datduck etl my_table_name"
    return
  end

  only_destination = DataDuck::Destination.only_destination

  if what == "all"
    etl = ETL.new(destinations: [only_destination], autoload_tables: true)
    etl.process!
  else
    table_name_camelized = DataDuck::Util.underscore_to_camelcase(what)
    require DataDuck.project_root + "/src/tables/#{ what }.rb"
    table_class = Object.const_get(table_name_camelized)
    if !(table_class <= DataDuck::Table)
      raise Exception.new("Table class #{ table_name_camelized } must inherit from DataDuck::Table")
    end

    table = table_class.new
    etl = ETL.new(destinations: [only_destination], autoload_tables: false, tables: [table])
    etl.process_table!(table)
  end
end

.help ⇒ Object 



108
109
110
111
 
# File 'lib/dataduck/commands.rb', line 108

def self.help
  puts "Usage: dataduck commandname"
  puts "Commands: #{ acceptable_commands.sort.join(' ') }"
end

.prompt_choices(choices = []) ⇒ Object 



19
20
21
22
23
24
25
26
27
28
29
30
31
32
 
# File 'lib/dataduck/commands.rb', line 19

def self.prompt_choices(choices = [])
  while true
    print "Enter a number 0 - #{ choices.length - 1}\n"
    choices.each_with_index do |choice, idx|
      choice_name = choice.is_a?(String) ? choice : choice[1]
      print "#{ idx }: #{ choice_name }\n"
    end
    choice = STDIN.gets.strip.to_i
    if 0 <= choice && choice < choices.length
      selected = choices[choice]
      return selected.is_a?(String) ? selected : selected[0]
    end
  end
end

.quickstart ⇒ Object 



137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
 
# File 'lib/dataduck/commands.rb', line 137

def self.quickstart
  puts "Welcome to DataDuck!"
  puts "This quickstart wizard will help you set up DataDuck."

  puts "What kind of database would you like to source from?"
  db_type = prompt_choices([
      [:mysql, "MySQL"],
      [:postgresql, "PostgreSQL"],
      [:other, "other"],
  ])

  if db_type == :other
    puts "You've selected 'other'. Unfortunately, those are the only choices supported at the moment. Contact us at DataDuckETL.com to request support for your database."
    exit
  end

  puts "Enter the source hostname:"
  source_host = STDIN.gets.strip

  puts "Enter the name of the database when connecting to #{ source_host }:"
  source_database = STDIN.gets.strip

  puts "Enter the source's port:"
  source_port = STDIN.gets.strip.to_i

  puts "Enter the username:"
  source_username = STDIN.gets.strip

  puts "Enter the password:"
  source_password = STDIN.noecho(&:gets).chomp

  db_class = {
      mysql: DataDuck::MysqlSource,
      postgresql: DataDuck::PostgresqlSource,
  }[db_type]

  db_source = db_class.new({
      'db_type' => db_type.to_s,
      'host' => source_host,
      'database' => source_database,
      'port' => source_port,
      'username' => source_username,
      'password' => source_password,
  })

  puts "Connecting to source database..."
  table_names = db_source.table_names
  puts "Connection successful. Detected #{ table_names.length } tables."
  puts "Creating scaffolding..."
  table_names.each do |table_name|
    DataDuck::Commands.quickstart_create_table(table_name, db_source)
  end

  config_obj = {
    'sources' => {
      'my_database' => {
        'type' => db_type.to_s,
        'host' => source_host,
        'database' => source_database,
        'port' => source_port,
        'username' => source_username,
        'password' => source_password,
      }
    },
    'destinations' => {
      'my_destination' => {
        'type'  => 'redshift',
        'aws_key'  => 'YOUR_AWS_KEY',
        'aws_secret'  => 'YOUR_AWS_SECRET',
        's3_bucket'  => 'YOUR_BUCKET',
        's3_region'  => 'YOUR_BUCKET_REGION',
        'host'  => 'redshift.somekeygoeshere.us-west-2.redshift.amazonaws.com',
        'port'  => 5439,
        'database'  => 'main',
        'schema'  => 'public',
        'username'  => 'YOUR_UESRNAME',
        'password'  => 'YOUR_PASSWORD',
      }
    }
  }

  DataDuck::Commands.quickstart_save_file("#{ DataDuck.project_root }/config/secret/#{ DataDuck.environment }.yml", config_obj.to_yaml)
  DataDuck::Commands.quickstart_save_main
  DataDuck::Commands.quickstart_update_gitignore

  puts "Quickstart complete!"
  puts "You still need to edit your config/secret/*.yml file with your AWS and Redshift credentials."
  puts "Run your ETL with: ruby src/main.rb"
end

.quickstart_create_table(table_name, db) ⇒ Object 



238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
 
# File 'lib/dataduck/commands.rb', line 238

def self.quickstart_create_table(table_name, db)
  columns = []
  schema = db.schema(table_name)
  schema.each do |property_schema|
    property_name = property_schema[0]
    property_type = property_schema[1][:type]
    commented_out = ['ssn', 'socialsecurity', 'password', 'encrypted_password', 'salt', 'password_salt', 'pw'].include?(property_name.to_s.downcase)
    columns << [property_name.to_s, property_type.to_s, commented_out]
  end

  columns.sort! { |a, b| a[0] <=> b[0] }

  table_name = table_name.to_s.downcase
  table_name_camelcased = table_name.split('_').collect(&:capitalize).join
  namespace = Namespace.new(table_name_camelcased: table_name_camelcased, table_name: table_name, columns: columns)
  template = File.open("#{ DataDuck.gem_root }/lib/templates/quickstart/table.rb.erb", 'r').read
  result = ERB.new(template).result(namespace.get_binding)
  DataDuck::Commands.quickstart_save_file("#{ DataDuck.project_root }/src/tables/#{ table_name }.rb", result)
end

.quickstart_save_file(output_path_full, contents) ⇒ Object 



258
259
260
261
262
263
264
265
266
 
# File 'lib/dataduck/commands.rb', line 258

def self.quickstart_save_file(output_path_full, contents)
  *output_path, output_filename = output_path_full.split('/')
  output_path = output_path.join("/")
  FileUtils::mkdir_p(output_path)

  output = File.open(output_path_full, "w")
  output << contents
  output.close
end

.quickstart_save_main ⇒ Object 



268
269
270
271
272
273
 
# File 'lib/dataduck/commands.rb', line 268

def self.quickstart_save_main
  namespace = Namespace.new
  template = File.open("#{ DataDuck.gem_root }/lib/templates/quickstart/main.rb.erb", 'r').read
  result = ERB.new(template).result(namespace.get_binding)
  DataDuck::Commands.quickstart_save_file("#{ DataDuck.project_root }/src/main.rb", result)
end

.quickstart_update_gitignore ⇒ Object 



227
228
229
230
231
232
233
234
235
236
 
# File 'lib/dataduck/commands.rb', line 227

def self.quickstart_update_gitignore
  main_gitignore_path = "#{ DataDuck.project_root }/.gitignore"
  FileUtils.touch(main_gitignore_path)

  secret_gitignore_path = "#{ DataDuck.project_root }/config/secret/.gitignore"
  FileUtils.touch(secret_gitignore_path)
  output = File.open(secret_gitignore_path, "w")
  output << '[^.]*'
  output.close
end

.route_command(args) ⇒ Object 



38
39
40
41
42
43
44
45
46
47
48
49
50
 
# File 'lib/dataduck/commands.rb', line 38

def self.route_command(args)
  if args.length == 0
    return DataDuck::Commands.help
  end

  command = args[0]
  if !Commands.acceptable_commands.include?(command)
    puts "No such command: #{ command }"
    return DataDuck::Commands.help
  end

  DataDuck::Commands.public_send(command, *args[1..-1])
end

.show(table_name = nil) ⇒ Object 



113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
 
# File 'lib/dataduck/commands.rb', line 113

def self.show(table_name = nil)
  if table_name.nil?
    Dir[DataDuck.project_root + "/src/tables/*.rb"].each do |file|
      table_name_underscores = file.split("/").last.gsub(".rb", "")
      table_name_camelized = DataDuck::Util.underscore_to_camelcase(table_name_underscores)
      require file
      table = Object.const_get(table_name_camelized)
      if table <= DataDuck::Table
        puts table_name_underscores
      end
    end
  else
    table_name_camelized = DataDuck::Util.underscore_to_camelcase(table_name)
    require DataDuck.project_root + "/src/tables/#{ table_name }.rb"
    table_class = Object.const_get(table_name_camelized)
    if !(table_class <= DataDuck::Table)
      raise Exception.new("Table class #{ table_name_camelized } must inherit from DataDuck::Table")
    end

    table = table_class.new
    table.show
  end
end