activewarehouse-etl-sgonyea 0.9.6
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +9 -0
- data/0.9-UPGRADE +6 -0
- data/CHANGELOG +236 -0
- data/Gemfile +4 -0
- data/HOW_TO_RELEASE +13 -0
- data/LICENSE +7 -0
- data/README.textile +111 -0
- data/Rakefile +103 -0
- data/TODO +28 -0
- data/active_support_logger.patch +78 -0
- data/activewarehouse-etl.gemspec +36 -0
- data/bin/etl +28 -0
- data/bin/etl.cmd +8 -0
- data/examples/database.example.yml +16 -0
- data/lib/etl.rb +97 -0
- data/lib/etl/batch.rb +2 -0
- data/lib/etl/batch/batch.rb +111 -0
- data/lib/etl/batch/directives.rb +65 -0
- data/lib/etl/builder.rb +2 -0
- data/lib/etl/builder/date_dimension_builder.rb +96 -0
- data/lib/etl/builder/time_dimension_builder.rb +31 -0
- data/lib/etl/commands/etl.rb +89 -0
- data/lib/etl/control.rb +3 -0
- data/lib/etl/control/control.rb +405 -0
- data/lib/etl/control/destination.rb +438 -0
- data/lib/etl/control/destination/csv_destination.rb +113 -0
- data/lib/etl/control/destination/database_destination.rb +97 -0
- data/lib/etl/control/destination/excel_destination.rb +91 -0
- data/lib/etl/control/destination/file_destination.rb +126 -0
- data/lib/etl/control/destination/insert_update_database_destination.rb +136 -0
- data/lib/etl/control/destination/update_database_destination.rb +109 -0
- data/lib/etl/control/destination/yaml_destination.rb +74 -0
- data/lib/etl/control/source.rb +132 -0
- data/lib/etl/control/source/database_source.rb +224 -0
- data/lib/etl/control/source/enumerable_source.rb +11 -0
- data/lib/etl/control/source/file_source.rb +90 -0
- data/lib/etl/control/source/model_source.rb +39 -0
- data/lib/etl/core_ext.rb +1 -0
- data/lib/etl/core_ext/time.rb +5 -0
- data/lib/etl/core_ext/time/calculations.rb +42 -0
- data/lib/etl/engine.rb +582 -0
- data/lib/etl/execution.rb +19 -0
- data/lib/etl/execution/base.rb +8 -0
- data/lib/etl/execution/batch.rb +10 -0
- data/lib/etl/execution/job.rb +8 -0
- data/lib/etl/execution/migration.rb +90 -0
- data/lib/etl/generator.rb +2 -0
- data/lib/etl/generator/generator.rb +20 -0
- data/lib/etl/generator/surrogate_key_generator.rb +39 -0
- data/lib/etl/http_tools.rb +139 -0
- data/lib/etl/parser.rb +11 -0
- data/lib/etl/parser/apache_combined_log_parser.rb +49 -0
- data/lib/etl/parser/csv_parser.rb +93 -0
- data/lib/etl/parser/excel_parser.rb +112 -0
- data/lib/etl/parser/fixed_width_parser.rb +65 -0
- data/lib/etl/parser/nokogiri_xml_parser.rb +83 -0
- data/lib/etl/parser/parser.rb +41 -0
- data/lib/etl/parser/sax_parser.rb +218 -0
- data/lib/etl/parser/xml_parser.rb +65 -0
- data/lib/etl/processor.rb +11 -0
- data/lib/etl/processor/block_processor.rb +14 -0
- data/lib/etl/processor/bulk_import_processor.rb +94 -0
- data/lib/etl/processor/check_exist_processor.rb +80 -0
- data/lib/etl/processor/check_unique_processor.rb +39 -0
- data/lib/etl/processor/copy_field_processor.rb +26 -0
- data/lib/etl/processor/database_join_processor.rb +82 -0
- data/lib/etl/processor/encode_processor.rb +55 -0
- data/lib/etl/processor/ensure_fields_presence_processor.rb +24 -0
- data/lib/etl/processor/escape_csv_processor.rb +77 -0
- data/lib/etl/processor/filter_row_processor.rb +51 -0
- data/lib/etl/processor/ftp_downloader_processor.rb +68 -0
- data/lib/etl/processor/ftp_uploader_processor.rb +65 -0
- data/lib/etl/processor/hierarchy_exploder_processor.rb +55 -0
- data/lib/etl/processor/imapattachment_downloader_processor.rb +91 -0
- data/lib/etl/processor/pop3attachment_downloader_processor.rb +90 -0
- data/lib/etl/processor/print_row_processor.rb +12 -0
- data/lib/etl/processor/processor.rb +25 -0
- data/lib/etl/processor/rename_processor.rb +24 -0
- data/lib/etl/processor/require_non_blank_processor.rb +26 -0
- data/lib/etl/processor/row_processor.rb +27 -0
- data/lib/etl/processor/sequence_processor.rb +23 -0
- data/lib/etl/processor/sftp_downloader_processor.rb +63 -0
- data/lib/etl/processor/sftp_uploader_processor.rb +63 -0
- data/lib/etl/processor/surrogate_key_processor.rb +53 -0
- data/lib/etl/processor/truncate_processor.rb +40 -0
- data/lib/etl/processor/zip_file_processor.rb +27 -0
- data/lib/etl/row.rb +20 -0
- data/lib/etl/screen.rb +14 -0
- data/lib/etl/screen/row_count_screen.rb +20 -0
- data/lib/etl/transform.rb +2 -0
- data/lib/etl/transform/block_transform.rb +13 -0
- data/lib/etl/transform/calculation_transform.rb +71 -0
- data/lib/etl/transform/date_to_string_transform.rb +20 -0
- data/lib/etl/transform/decode_transform.rb +51 -0
- data/lib/etl/transform/default_transform.rb +20 -0
- data/lib/etl/transform/foreign_key_lookup_transform.rb +211 -0
- data/lib/etl/transform/hierarchy_lookup_transform.rb +49 -0
- data/lib/etl/transform/md5_transform.rb +13 -0
- data/lib/etl/transform/ordinalize_transform.rb +14 -0
- data/lib/etl/transform/sha1_transform.rb +13 -0
- data/lib/etl/transform/split_fields_transform.rb +27 -0
- data/lib/etl/transform/string_to_date_time_transform.rb +14 -0
- data/lib/etl/transform/string_to_date_transform.rb +16 -0
- data/lib/etl/transform/string_to_time_transform.rb +11 -0
- data/lib/etl/transform/transform.rb +61 -0
- data/lib/etl/transform/trim_transform.rb +26 -0
- data/lib/etl/transform/type_transform.rb +35 -0
- data/lib/etl/util.rb +59 -0
- data/lib/etl/version.rb +3 -0
- data/test-matrix.yml +10 -0
- data/test/.gitignore +1 -0
- data/test/.ignore +2 -0
- data/test/all.ebf +6 -0
- data/test/apache_combined_log.ctl +11 -0
- data/test/batch_test.rb +41 -0
- data/test/batch_with_error.ebf +6 -0
- data/test/batched1.ctl +0 -0
- data/test/batched2.ctl +0 -0
- data/test/block_processor.ctl +6 -0
- data/test/block_processor_error.ctl +1 -0
- data/test/block_processor_pre_post_process.ctl +4 -0
- data/test/block_processor_remove_rows.ctl +5 -0
- data/test/block_processor_test.rb +38 -0
- data/test/check_exist_processor_test.rb +92 -0
- data/test/check_unique_processor_test.rb +40 -0
- data/test/config/Gemfile.rails-2.3.x +3 -0
- data/test/config/Gemfile.rails-2.3.x.lock +53 -0
- data/test/config/Gemfile.rails-3.0.x +3 -0
- data/test/config/Gemfile.rails-3.0.x.lock +61 -0
- data/test/config/common.rb +29 -0
- data/test/connection/mysql/connection.rb +9 -0
- data/test/connection/mysql/schema.sql +37 -0
- data/test/connection/postgresql/connection.rb +13 -0
- data/test/connection/postgresql/schema.sql +40 -0
- data/test/control_test.rb +43 -0
- data/test/data/apache_combined_log.txt +3 -0
- data/test/data/bulk_import.txt +3 -0
- data/test/data/bulk_import_with_empties.txt +3 -0
- data/test/data/decode.txt +3 -0
- data/test/data/delimited.txt +3 -0
- data/test/data/encode_source_latin1.txt +2 -0
- data/test/data/excel.xls +0 -0
- data/test/data/excel2.xls +0 -0
- data/test/data/fixed_width.txt +3 -0
- data/test/data/multiple_delimited_1.txt +3 -0
- data/test/data/multiple_delimited_2.txt +3 -0
- data/test/data/nokogiri.xml +38 -0
- data/test/data/people.txt +3 -0
- data/test/data/sax.xml +14 -0
- data/test/data/xml.xml +16 -0
- data/test/database_join_processor_test.rb +43 -0
- data/test/date_dimension_builder_test.rb +96 -0
- data/test/delimited.ctl +30 -0
- data/test/delimited_absolute.ctl +31 -0
- data/test/delimited_destination_db.ctl +23 -0
- data/test/delimited_excel.ctl +31 -0
- data/test/delimited_insert_update.ctl +34 -0
- data/test/delimited_update.ctl +34 -0
- data/test/delimited_with_bulk_load.ctl +34 -0
- data/test/destination_test.rb +275 -0
- data/test/directive_test.rb +23 -0
- data/test/encode_processor_test.rb +32 -0
- data/test/engine_test.rb +78 -0
- data/test/ensure_fields_presence_processor_test.rb +28 -0
- data/test/errors.ctl +24 -0
- data/test/etl_test.rb +42 -0
- data/test/excel.ctl +24 -0
- data/test/excel2.ctl +25 -0
- data/test/fixed_width.ctl +35 -0
- data/test/foreign_key_lookup_transform_test.rb +50 -0
- data/test/generator_test.rb +14 -0
- data/test/inline_parser.ctl +17 -0
- data/test/mocks/mock_destination.rb +26 -0
- data/test/mocks/mock_source.rb +25 -0
- data/test/model_source.ctl +14 -0
- data/test/multiple_delimited.ctl +22 -0
- data/test/multiple_source_delimited.ctl +39 -0
- data/test/nokogiri_all.ctl +35 -0
- data/test/nokogiri_select.ctl +35 -0
- data/test/nokogiri_test.rb +35 -0
- data/test/parser_test.rb +224 -0
- data/test/performance/delimited.ctl +30 -0
- data/test/processor_test.rb +44 -0
- data/test/row_processor_test.rb +17 -0
- data/test/sax.ctl +26 -0
- data/test/scd/1.txt +1 -0
- data/test/scd/2.txt +1 -0
- data/test/scd/3.txt +1 -0
- data/test/scd_test.rb +257 -0
- data/test/scd_test_type_1.ctl +43 -0
- data/test/scd_test_type_2.ctl +34 -0
- data/test/screen_test.rb +9 -0
- data/test/screen_test_error.ctl +3 -0
- data/test/screen_test_fatal.ctl +3 -0
- data/test/source_test.rb +154 -0
- data/test/test_helper.rb +37 -0
- data/test/transform_test.rb +101 -0
- data/test/truncate_processor_test.rb +37 -0
- data/test/xml.ctl +31 -0
- metadata +370 -0
@@ -0,0 +1,438 @@
|
|
1
|
+
module ETL #:nodoc:
|
2
|
+
module Control #:nodoc:
|
3
|
+
# Base class for destinations.
|
4
|
+
class Destination
|
5
|
+
# Read-only accessor for the ETL::Control::Control instance
|
6
|
+
attr_reader :control
|
7
|
+
|
8
|
+
# Read-only accessor for the configuration Hash
|
9
|
+
attr_reader :configuration
|
10
|
+
|
11
|
+
# Read-only accessor for the destination mapping Hash
|
12
|
+
attr_reader :mapping
|
13
|
+
|
14
|
+
# Accessor to the buffer size
|
15
|
+
attr_accessor :buffer_size
|
16
|
+
|
17
|
+
# Unique flag.
|
18
|
+
attr_accessor :unique
|
19
|
+
|
20
|
+
# A condition for writing
|
21
|
+
attr_accessor :condition
|
22
|
+
|
23
|
+
# An array of rows to append to the destination
|
24
|
+
attr_accessor :append_rows
|
25
|
+
|
26
|
+
class << self
|
27
|
+
# Get the destination class for the specified name.
|
28
|
+
#
|
29
|
+
# For example if name is :database or 'database' then the
|
30
|
+
# DatabaseDestination class is returned
|
31
|
+
def class_for_name(name)
|
32
|
+
ETL::Control.const_get("#{name.to_s.camelize}Destination")
|
33
|
+
end
|
34
|
+
end
|
35
|
+
|
36
|
+
# Initialize the destination
|
37
|
+
#
|
38
|
+
# Arguments:
|
39
|
+
# * <tt>control</tt>: The ETL::Control::Control instance
|
40
|
+
# * <tt>configuration</tt>: The configuration Hash
|
41
|
+
# * <tt>mapping</tt>: The mapping Hash
|
42
|
+
#
|
43
|
+
# Options:
|
44
|
+
# * <tt>:buffer_size</tt>: The output buffer size (default 1000 records)
|
45
|
+
# * <tt>:condition</tt>: A conditional proc that must return true for the
|
46
|
+
# row to be written
|
47
|
+
# * <tt>:append_rows</tt>: An array of rows to append
|
48
|
+
def initialize(control, configuration, mapping)
|
49
|
+
@control = control
|
50
|
+
@configuration = configuration
|
51
|
+
@mapping = mapping
|
52
|
+
@buffer_size = configuration[:buffer_size] ||= 100
|
53
|
+
@condition = configuration[:condition]
|
54
|
+
@append_rows = configuration[:append_rows]
|
55
|
+
end
|
56
|
+
|
57
|
+
# Get the current row number
|
58
|
+
def current_row
|
59
|
+
@current_row ||= 1
|
60
|
+
end
|
61
|
+
|
62
|
+
# Write the given row
|
63
|
+
def write(row)
|
64
|
+
if @condition.nil? || @condition.call(row)
|
65
|
+
process_change(row)
|
66
|
+
end
|
67
|
+
flush if buffer.length >= buffer_size
|
68
|
+
end
|
69
|
+
|
70
|
+
# Abstract method
|
71
|
+
def flush
|
72
|
+
raise NotImplementedError, "flush method must be implemented by subclasses"
|
73
|
+
end
|
74
|
+
|
75
|
+
# Abstract method
|
76
|
+
def close
|
77
|
+
raise NotImplementedError, "close method must be implemented by subclasses"
|
78
|
+
end
|
79
|
+
|
80
|
+
def errors
|
81
|
+
@errors ||= []
|
82
|
+
end
|
83
|
+
|
84
|
+
protected
|
85
|
+
# Access the buffer
|
86
|
+
def buffer
|
87
|
+
@buffer ||= []
|
88
|
+
end
|
89
|
+
|
90
|
+
# Access the generators map
|
91
|
+
def generators
|
92
|
+
@generators ||= {}
|
93
|
+
end
|
94
|
+
|
95
|
+
# Get the order of elements from the source order
|
96
|
+
def order_from_source
|
97
|
+
order = []
|
98
|
+
control.sources.first.definition.each do |item|
|
99
|
+
case item
|
100
|
+
when Hash
|
101
|
+
order << item[:name]
|
102
|
+
else
|
103
|
+
order << item
|
104
|
+
end
|
105
|
+
end
|
106
|
+
order
|
107
|
+
end
|
108
|
+
|
109
|
+
# Return true if the row is allowed. The row will not be allowed if the
|
110
|
+
# :unique option is specified in the configuration and the compound key
|
111
|
+
# already exists
|
112
|
+
def row_allowed?(row)
|
113
|
+
if unique
|
114
|
+
key = (unique.collect { |k| row[k] }).join('|')
|
115
|
+
return false if compound_key_constraints[key]
|
116
|
+
compound_key_constraints[key] = 1
|
117
|
+
end
|
118
|
+
return true
|
119
|
+
end
|
120
|
+
|
121
|
+
# Get a hash of compound key contraints. This is used to determine if a
|
122
|
+
# row can be written when the unique option is specified
|
123
|
+
def compound_key_constraints
|
124
|
+
@compound_key_constraints ||= {}
|
125
|
+
end
|
126
|
+
|
127
|
+
# Return fields which are Slowly Changing Dimension fields.
|
128
|
+
# Uses the scd_fields specified in the configuration. If that's
|
129
|
+
# missing, uses all of the row's fields.
|
130
|
+
def scd_fields(row)
|
131
|
+
@scd_fields ||= configuration[:scd_fields] || row.keys
|
132
|
+
ETL::Engine.logger.debug "@scd_fields is: #{@scd_fields.inspect}"
|
133
|
+
@scd_fields
|
134
|
+
end
|
135
|
+
|
136
|
+
# returns the fields that are required to identify an SCD
|
137
|
+
def scd_required_fields
|
138
|
+
if scd? and scd_type == 2
|
139
|
+
[scd_effective_date_field, scd_end_date_field, scd_latest_version_field]
|
140
|
+
else
|
141
|
+
[]
|
142
|
+
end
|
143
|
+
end
|
144
|
+
|
145
|
+
def non_scd_fields(row)
|
146
|
+
@non_scd_fields ||= row.keys - natural_key - scd_fields(row) - [primary_key] - scd_required_fields
|
147
|
+
ETL::Engine.logger.debug "@non_scd_fields is: #{@non_scd_fields.inspect}"
|
148
|
+
@non_scd_fields
|
149
|
+
end
|
150
|
+
|
151
|
+
def non_evolving_fields
|
152
|
+
(Array(configuration[:scd][:non_evolving_fields]) << primary_key).uniq
|
153
|
+
end
|
154
|
+
|
155
|
+
def scd?
|
156
|
+
!configuration[:scd].nil?
|
157
|
+
end
|
158
|
+
|
159
|
+
def scd_type
|
160
|
+
scd? ? configuration[:scd][:type] : nil
|
161
|
+
end
|
162
|
+
|
163
|
+
# Get the Slowly Changing Dimension effective date field. Defaults to
|
164
|
+
# 'effective_date'.
|
165
|
+
def scd_effective_date_field
|
166
|
+
configuration[:scd][:effective_date_field] || :effective_date if scd?
|
167
|
+
end
|
168
|
+
|
169
|
+
# Get the Slowly Changing Dimension end date field. Defaults to
|
170
|
+
# 'end_date'.
|
171
|
+
def scd_end_date_field
|
172
|
+
configuration[:scd][:end_date_field] || :end_date if scd?
|
173
|
+
end
|
174
|
+
|
175
|
+
# Get the Slowly Changing Dimension latest version field. Defaults to
|
176
|
+
# 'latest_version'.
|
177
|
+
def scd_latest_version_field
|
178
|
+
configuration[:scd][:latest_version_field] || :latest_version if scd?
|
179
|
+
end
|
180
|
+
|
181
|
+
# Return the natural key field names, defaults to []
|
182
|
+
def natural_key
|
183
|
+
@natural_key ||= determine_natural_key
|
184
|
+
end
|
185
|
+
|
186
|
+
# Get the dimension table if specified
|
187
|
+
def dimension_table
|
188
|
+
@dimension_table ||= if scd?
|
189
|
+
ETL::Engine.table(configuration[:scd][:dimension_table], dimension_target) or raise ConfigurationError, "dimension_table setting required"
|
190
|
+
end
|
191
|
+
end
|
192
|
+
|
193
|
+
# Get the dimension target if specified
|
194
|
+
def dimension_target
|
195
|
+
@dimension_target ||= if scd?
|
196
|
+
configuration[:scd][:dimension_target] or raise ConfigurationError, "dimension_target setting required"
|
197
|
+
end
|
198
|
+
end
|
199
|
+
|
200
|
+
# Process a row to determine the change type
|
201
|
+
def process_change(row)
|
202
|
+
ETL::Engine.logger.debug "Processing row: #{row.inspect}"
|
203
|
+
return unless row
|
204
|
+
|
205
|
+
# Change processing can only occur if the natural key exists in the row
|
206
|
+
ETL::Engine.logger.debug "Checking for natural key existence"
|
207
|
+
unless has_natural_key?(row)
|
208
|
+
buffer << row
|
209
|
+
return
|
210
|
+
end
|
211
|
+
|
212
|
+
@timestamp = Time.now
|
213
|
+
|
214
|
+
# See if the scd_fields of the current record have changed
|
215
|
+
# from the last time this record was loaded into the data
|
216
|
+
# warehouse. If they match then throw away this row (no need
|
217
|
+
# to process). If they do not match then the record is an
|
218
|
+
# 'update'. If the record doesn't exist then it is an 'insert'
|
219
|
+
ETL::Engine.logger.debug "Checking record for SCD change"
|
220
|
+
if @existing_row = preexisting_row(row)
|
221
|
+
if has_scd_field_changes?(row)
|
222
|
+
process_scd_change(row)
|
223
|
+
else
|
224
|
+
process_scd_match(row)
|
225
|
+
end
|
226
|
+
else
|
227
|
+
schedule_new_record(row)
|
228
|
+
end
|
229
|
+
end
|
230
|
+
|
231
|
+
# Add any virtual fields to the row. Virtual rows will get their value
|
232
|
+
# from one of the following:
|
233
|
+
# * If the mapping is a Class, then an object which implements the next
|
234
|
+
# method
|
235
|
+
# * If the mapping is a Symbol, then the XGenerator where X is the
|
236
|
+
# classified symbol
|
237
|
+
# * If the mapping is a Proc, then it will be called with the row
|
238
|
+
# * Otherwise the value itself will be assigned to the field
|
239
|
+
def add_virtuals!(row)
|
240
|
+
if mapping[:virtual]
|
241
|
+
mapping[:virtual].each do |key,value|
|
242
|
+
# If the row already has the virtual set, assume that's correct
|
243
|
+
next if row[key]
|
244
|
+
# Engine.logger.debug "Mapping virtual #{key}/#{value} for row #{row}"
|
245
|
+
case value
|
246
|
+
when Class
|
247
|
+
generator = generators[key] ||= value.new
|
248
|
+
row[key] = generator.next
|
249
|
+
when Symbol
|
250
|
+
generator = generators[key] ||= ETL::Generator::Generator.class_for_name(value).new(options)
|
251
|
+
row[key] = generator.next
|
252
|
+
when Proc, Method
|
253
|
+
row[key] = value.call(row)
|
254
|
+
else
|
255
|
+
if value.is_a?(ETL::Generator::Generator)
|
256
|
+
row[key] = value.next
|
257
|
+
else
|
258
|
+
row[key] = value
|
259
|
+
end
|
260
|
+
end
|
261
|
+
end
|
262
|
+
end
|
263
|
+
end
|
264
|
+
|
265
|
+
private
|
266
|
+
|
267
|
+
# Determine the natural key. This method will always return an array
|
268
|
+
# of symbols. The default value is [].
|
269
|
+
def determine_natural_key
|
270
|
+
Array(configuration[:natural_key]).collect(&:to_sym)
|
271
|
+
end
|
272
|
+
|
273
|
+
# Check whether a natural key has been defined, and if so, whether
|
274
|
+
# this row has enough information to do searches based on that natural
|
275
|
+
# key.
|
276
|
+
#
|
277
|
+
# TODO: This should be factored out into
|
278
|
+
# ETL::Row#has_all_fields?(field_array) But that's not possible
|
279
|
+
# until *all* sources cast to ETL::Row, instead of sometimes
|
280
|
+
# using Hash
|
281
|
+
def has_natural_key?(row)
|
282
|
+
natural_key.any? && natural_key.all? { |key| row.has_key?(key) }
|
283
|
+
end
|
284
|
+
|
285
|
+
# Helper for generating the SQL where clause that allows searching
|
286
|
+
# by a natural key
|
287
|
+
def natural_key_equality_for_row(row)
|
288
|
+
statement = []
|
289
|
+
values = []
|
290
|
+
natural_key.each do |nk|
|
291
|
+
statement << "#{nk} = #{ActiveRecord::Base.send(:quote_bound_value, row[nk], connection)}"
|
292
|
+
end
|
293
|
+
statement = statement.join(" AND ")
|
294
|
+
return statement
|
295
|
+
end
|
296
|
+
|
297
|
+
# Do all the steps required when a SCD *has* changed. Exact steps
|
298
|
+
# depend on what type of SCD we're handling.
|
299
|
+
def process_scd_change(row)
|
300
|
+
ETL::Engine.logger.debug "SCD fields do not match"
|
301
|
+
|
302
|
+
if scd_type == 2
|
303
|
+
# SCD Type 2: new row should be added and old row should be updated
|
304
|
+
ETL::Engine.logger.debug "type 2 SCD"
|
305
|
+
|
306
|
+
# To update the old row, we delete the version in the database
|
307
|
+
# and insert a new expired version
|
308
|
+
|
309
|
+
# If there is no truncate then the row will exist twice in the database
|
310
|
+
delete_outdated_record
|
311
|
+
|
312
|
+
ETL::Engine.logger.debug "expiring original record"
|
313
|
+
@existing_row[scd_end_date_field] = @timestamp
|
314
|
+
@existing_row[scd_latest_version_field] = false
|
315
|
+
|
316
|
+
buffer << @existing_row
|
317
|
+
|
318
|
+
elsif scd_type == 1
|
319
|
+
# SCD Type 1: only the new row should be added
|
320
|
+
ETL::Engine.logger.debug "type 1 SCD"
|
321
|
+
|
322
|
+
# Copy primary key, and other non-evolving fields over from
|
323
|
+
# original version of record
|
324
|
+
non_evolving_fields.each do |non_evolving_field|
|
325
|
+
row[non_evolving_field] = @existing_row[non_evolving_field]
|
326
|
+
end
|
327
|
+
|
328
|
+
# If there is no truncate then the row will exist twice in the database
|
329
|
+
delete_outdated_record
|
330
|
+
else
|
331
|
+
# SCD Type 3: not supported
|
332
|
+
ETL::Engine.logger.debug "SCD type #{scd_type} not supported"
|
333
|
+
end
|
334
|
+
|
335
|
+
# In all cases, the latest, greatest version of the record
|
336
|
+
# should go into the load
|
337
|
+
schedule_new_record(row)
|
338
|
+
end
|
339
|
+
|
340
|
+
# Do all the steps required when a SCD has *not* changed. Exact
|
341
|
+
# steps depend on what type of SCD we're handling.
|
342
|
+
def process_scd_match(row)
|
343
|
+
ETL::Engine.logger.debug "SCD fields match"
|
344
|
+
|
345
|
+
if scd_type == 2 && has_non_scd_field_changes?(row)
|
346
|
+
ETL::Engine.logger.debug "Non-SCD field changes"
|
347
|
+
# Copy important data over from original version of record
|
348
|
+
row[primary_key] = @existing_row[primary_key]
|
349
|
+
row[scd_end_date_field] = @existing_row[scd_end_date_field]
|
350
|
+
row[scd_effective_date_field] = @existing_row[scd_effective_date_field]
|
351
|
+
row[scd_latest_version_field] = @existing_row[scd_latest_version_field]
|
352
|
+
|
353
|
+
# If there is no truncate then the row will exist twice in the database
|
354
|
+
delete_outdated_record
|
355
|
+
|
356
|
+
buffer << row
|
357
|
+
else
|
358
|
+
# The record is totally the same, so skip it
|
359
|
+
end
|
360
|
+
end
|
361
|
+
|
362
|
+
# Find the version of this row that already exists in the datawarehouse.
|
363
|
+
def preexisting_row(row)
|
364
|
+
q = "SELECT * FROM #{dimension_table} WHERE #{natural_key_equality_for_row(row)}"
|
365
|
+
q << " AND #{scd_latest_version_field}" if scd_type == 2
|
366
|
+
|
367
|
+
ETL::Engine.logger.debug "looking for original record"
|
368
|
+
result = connection.select_one(q)
|
369
|
+
|
370
|
+
ETL::Engine.logger.debug "Result: #{result.inspect}"
|
371
|
+
|
372
|
+
result ? ETL::Row[result.symbolize_keys!] : nil
|
373
|
+
end
|
374
|
+
|
375
|
+
# Check whether non-scd fields have changed since the last
|
376
|
+
# load of this record.
|
377
|
+
def has_scd_field_changes?(row)
|
378
|
+
scd_fields(row).any? { |csd_field|
|
379
|
+
ETL::Engine.logger.debug "Row: #{row.inspect}"
|
380
|
+
ETL::Engine.logger.debug "Existing Row: #{@existing_row.inspect}"
|
381
|
+
ETL::Engine.logger.debug "comparing: #{row[csd_field].to_s} != #{@existing_row[csd_field].to_s}"
|
382
|
+
x=row[csd_field].to_s != @existing_row[csd_field].to_s
|
383
|
+
ETL::Engine.logger.debug x
|
384
|
+
x
|
385
|
+
}
|
386
|
+
end
|
387
|
+
|
388
|
+
# Check whether non-scd fields have changed since the last
|
389
|
+
# load of this record.
|
390
|
+
def has_non_scd_field_changes?(row)
|
391
|
+
non_scd_fields(row).any? { |non_csd_field| row[non_csd_field].to_s != @existing_row[non_csd_field].to_s }
|
392
|
+
end
|
393
|
+
|
394
|
+
# Grab, or re-use, a database connection for running queries directly
|
395
|
+
# during the destination processing.
|
396
|
+
def connection
|
397
|
+
@conn ||= ETL::Engine.connection(dimension_target)
|
398
|
+
end
|
399
|
+
|
400
|
+
# Utility for removing a row that has outdated information. Note
|
401
|
+
# that this deletes directly from the database, even if this is a file
|
402
|
+
# destination. It needs to do this because you can't do deletes in a
|
403
|
+
# bulk load.
|
404
|
+
def delete_outdated_record
|
405
|
+
ETL::Engine.logger.debug "deleting old row"
|
406
|
+
|
407
|
+
q = "DELETE FROM #{dimension_table} WHERE #{primary_key} = #{@existing_row[primary_key]}"
|
408
|
+
connection.delete(q)
|
409
|
+
end
|
410
|
+
|
411
|
+
# Schedule the latest, greatest version of the row for insertion
|
412
|
+
# into the database
|
413
|
+
def schedule_new_record(row)
|
414
|
+
ETL::Engine.logger.debug "writing new record"
|
415
|
+
if scd_type == 2
|
416
|
+
row[scd_effective_date_field] = @timestamp
|
417
|
+
row[scd_end_date_field] = '9999-12-31 00:00:00'
|
418
|
+
row[scd_latest_version_field] = true
|
419
|
+
end
|
420
|
+
buffer << row
|
421
|
+
end
|
422
|
+
|
423
|
+
# Get the name of the primary key for this table. Asks the dimension
|
424
|
+
# model class for this information, but if that class hasn't been
|
425
|
+
# defined, just defaults to :id.
|
426
|
+
def primary_key
|
427
|
+
return @primary_key if @primary_key
|
428
|
+
@primary_key = dimension_table.to_s.camelize.constantize.primary_key.to_sym
|
429
|
+
rescue NameError => e
|
430
|
+
ETL::Engine.logger.debug "couldn't get primary_key from dimension model class, using default :id"
|
431
|
+
@primary_key = :id
|
432
|
+
end
|
433
|
+
|
434
|
+
end
|
435
|
+
end
|
436
|
+
end
|
437
|
+
|
438
|
+
Dir[File.dirname(__FILE__) + "/destination/*.rb"].each { |file| require(file) }
|
@@ -0,0 +1,113 @@
|
|
1
|
+
# This source file contains the ETL::Control::CsvDestination
|
2
|
+
|
3
|
+
module ETL #:nodoc:
|
4
|
+
module Control #:nodoc:
|
5
|
+
# CSV File as the final destination.
|
6
|
+
class CsvDestination < Destination
|
7
|
+
# The File to write to
|
8
|
+
attr_reader :file
|
9
|
+
|
10
|
+
# The output order
|
11
|
+
attr_reader :order
|
12
|
+
|
13
|
+
# Flag which indicates to append (default is to overwrite)
|
14
|
+
attr_accessor :append
|
15
|
+
|
16
|
+
# The separator
|
17
|
+
attr_accessor :separator
|
18
|
+
|
19
|
+
# The end of line marker
|
20
|
+
attr_accessor :eol
|
21
|
+
|
22
|
+
# The enclosure character
|
23
|
+
attr_accessor :enclose
|
24
|
+
|
25
|
+
# Initialize the object.
|
26
|
+
# * <tt>control</tt>: The Control object
|
27
|
+
# * <tt>configuration</tt>: The configuration map
|
28
|
+
# * <tt>mapping</tt>: The output mapping
|
29
|
+
#
|
30
|
+
# Configuration options:
|
31
|
+
# * <tt>:file<tt>: The file to write to (REQUIRED)
|
32
|
+
# * <tt>:append</tt>: Set to true to append to the file (default is to overwrite)
|
33
|
+
# * <tt>:separator</tt>: Record separator (default is a comma)
|
34
|
+
# * <tt>:eol</tt>: End of line marker (default is \n)
|
35
|
+
# * <tt>:enclose</tt>: Set to true of false
|
36
|
+
# * <tt>:unique</tt>: Set to true to only write unique records
|
37
|
+
# * <tt>:append_rows</tt>: Array of rows to append
|
38
|
+
#
|
39
|
+
# Mapping options:
|
40
|
+
# * <tt>:order</tt>: The order array
|
41
|
+
def initialize(control, configuration, mapping={})
|
42
|
+
super
|
43
|
+
path = Pathname.new(configuration[:file])
|
44
|
+
@file = path.absolute? ? path : Pathname.new(File.dirname(File.expand_path(control.file))) + path
|
45
|
+
@append = configuration[:append] ||= false
|
46
|
+
@separator = configuration[:separator] ||= ','
|
47
|
+
@eol = configuration[:eol] ||= "\n"
|
48
|
+
@enclose = true & configuration[:enclose]
|
49
|
+
@unique = configuration[:unique] ? configuration[:unique] + scd_required_fields : configuration[:unique]
|
50
|
+
@unique.uniq! unless @unique.nil?
|
51
|
+
@order = mapping[:order] ? mapping[:order] + scd_required_fields : order_from_source
|
52
|
+
@order.uniq! unless @order.nil?
|
53
|
+
raise ControlError, "Order required in mapping" unless @order
|
54
|
+
end
|
55
|
+
|
56
|
+
# Close the destination. This will flush the buffer and close the underlying stream or connection.
|
57
|
+
def close
|
58
|
+
buffer << append_rows if append_rows
|
59
|
+
flush
|
60
|
+
f.close
|
61
|
+
end
|
62
|
+
|
63
|
+
# Flush the destination buffer
|
64
|
+
def flush
|
65
|
+
#puts "Flushing buffer (#{file}) with #{buffer.length} rows"
|
66
|
+
buffer.flatten.each do |row|
|
67
|
+
#puts "row change type: #{row.change_type}"
|
68
|
+
# check to see if this row's compound key constraint already exists
|
69
|
+
# note that the compound key constraint may not utilize virtual fields
|
70
|
+
next unless row_allowed?(row)
|
71
|
+
|
72
|
+
# add any virtual fields
|
73
|
+
add_virtuals!(row)
|
74
|
+
|
75
|
+
# collect all of the values using the order designated in the configuration
|
76
|
+
values = order.collect do |name|
|
77
|
+
value = row[name]
|
78
|
+
case value
|
79
|
+
when Date, Time, DateTime
|
80
|
+
value.to_s(:db)
|
81
|
+
else
|
82
|
+
value.to_s
|
83
|
+
end
|
84
|
+
end
|
85
|
+
|
86
|
+
f << values
|
87
|
+
end
|
88
|
+
f.flush
|
89
|
+
buffer.clear
|
90
|
+
#puts "After flush there are #{buffer.length} rows"
|
91
|
+
end
|
92
|
+
|
93
|
+
private
|
94
|
+
# Get the open file stream
|
95
|
+
def f
|
96
|
+
@f ||= FasterCSV.open(file, mode, options)
|
97
|
+
end
|
98
|
+
|
99
|
+
def options
|
100
|
+
@options ||= {
|
101
|
+
:col_sep => separator,
|
102
|
+
:row_sep => eol,
|
103
|
+
:force_quotes => enclose
|
104
|
+
}
|
105
|
+
end
|
106
|
+
|
107
|
+
# Get the appropriate mode to open the file stream
|
108
|
+
def mode
|
109
|
+
append ? 'a' : 'w'
|
110
|
+
end
|
111
|
+
end
|
112
|
+
end
|
113
|
+
end
|