affiliate_window_etl 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/README.md +109 -0
- data/lib/affiliate_window/etl.rb +11 -0
- data/lib/affiliate_window/etl/base.rb +78 -0
- data/lib/affiliate_window/etl/config.rb +37 -0
- data/lib/affiliate_window/etl/database.rb +89 -0
- data/lib/affiliate_window/etl/extracter.rb +183 -0
- data/lib/affiliate_window/etl/loader.rb +25 -0
- data/lib/affiliate_window/etl/normaliser.rb +22 -0
- data/lib/affiliate_window/etl/scheduler.rb +67 -0
- data/lib/affiliate_window/etl/tasks.rake +12 -0
- data/lib/affiliate_window/etl/transformer.rb +135 -0
- data/lib/affiliate_window/etl/version.rb +5 -0
- data/lib/affiliate_window_etl.rb +1 -0
- metadata +211 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: 3c47d9293527c807d7c97c7c549ae5b29604a6a9
|
4
|
+
data.tar.gz: 07b406b3253d7b040b2c925167ac9ef5fd0c387e
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 65c1afd3e2086c82ca1f383e468e5f6522fb8d3372304f441beeee7e66beed14647040301f10681bdb90c3214343098c222bfbe2036504b70adb6f6dfb663be2
|
7
|
+
data.tar.gz: 83334372aef519037ca82a0920bbbb1253ec0642e583b6ec150759bf8afca3a93b442032aa178a196e9d5d73454ce985e462ded16fa78d1c9b6c4f336b87c694
|
data/README.md
ADDED
@@ -0,0 +1,109 @@
|
|
1
|
+
##Affiliate Window ETL
|
2
|
+
|
3
|
+
[![Build Status](https://travis-ci.org/reevoo/affiliate_window_etl.svg?branch=master)](https://travis-ci.org/reevoo/affiliate_window_etl)
|
4
|
+
|
5
|
+
This gem provides an extract-transform-load process for retrieving records from
|
6
|
+
the [Affiliate Window](http://www.affiliatewindow.com/) API and loading them
|
7
|
+
into a Postgres database. It works incrementally, updating existing records
|
8
|
+
rather than creating duplicates. It handles its own scheduling and determines
|
9
|
+
which records need to be retrieved from the API based on the current state of
|
10
|
+
the database.
|
11
|
+
|
12
|
+
##How to run it
|
13
|
+
|
14
|
+
You can either clone this repository and run the ETL with rake or add it as a
|
15
|
+
dependency of another application.
|
16
|
+
|
17
|
+
**To run from this repository:**
|
18
|
+
|
19
|
+
1. Create the database: `createdb affiliate_window`
|
20
|
+
2. Migrate the database: `rake db:migrate`
|
21
|
+
3. Run the ETL: `rake run`
|
22
|
+
|
23
|
+
**To run as a dependency:**
|
24
|
+
|
25
|
+
```ruby
|
26
|
+
require "affiliate_window/etl"
|
27
|
+
etl = AffiliateWindow::ETL.new
|
28
|
+
|
29
|
+
etl.migrate
|
30
|
+
etl.run
|
31
|
+
```
|
32
|
+
|
33
|
+
By default, it will inherit config from your environment variables. If you don't
|
34
|
+
want it do this, you can pass in a hash:
|
35
|
+
|
36
|
+
```ruby
|
37
|
+
AffiliateWindow::ETL.new(
|
38
|
+
"START_DATE" => "2016-01-01",
|
39
|
+
"END_DATE" => "2016-01-07",
|
40
|
+
# ...
|
41
|
+
)
|
42
|
+
```
|
43
|
+
|
44
|
+
If you would like, you can add the gem's rake tasks to your app:
|
45
|
+
|
46
|
+
```ruby
|
47
|
+
# Rakefile
|
48
|
+
namespace :etl do
|
49
|
+
spec = Gem::Specification.find_by_name "affiliate_window_etl"
|
50
|
+
load "#{spec.gem_dir}/lib/affiliate_window/etl/tasks.rake"
|
51
|
+
end
|
52
|
+
```
|
53
|
+
|
54
|
+
If you are incorporating the ETL into an app that has its own migrations, it is
|
55
|
+
recommended you copy the migrations into the container app:
|
56
|
+
|
57
|
+
```ruby
|
58
|
+
etl.migration_filenames.each do |filename|
|
59
|
+
# copy to db/migrate/
|
60
|
+
end
|
61
|
+
```
|
62
|
+
|
63
|
+
##How to configure it
|
64
|
+
|
65
|
+
The library tries to be [twelve-factor](https://12factor.net/) compliant and is
|
66
|
+
configured via environment variables. When not specified, the configuration
|
67
|
+
defaults to something suitable for development purposes. If the library is used
|
68
|
+
as a dependency of something else, this configuration can be passed into the
|
69
|
+
`AffiliateWindow::ETL` initializer.
|
70
|
+
|
71
|
+
`ACCOUNT_ID`
|
72
|
+
|
73
|
+
The id of the Affiliate Window account for which to retrieve records.
|
74
|
+
|
75
|
+
`AFFILIATE_API_PASSWORD`
|
76
|
+
|
77
|
+
The API token of for the Publisher Service. Can be retrieved from
|
78
|
+
[this page](https://www.affiliatewindow.com/affiliates/accountdetails.php).
|
79
|
+
|
80
|
+
`DATABASE_URL`
|
81
|
+
|
82
|
+
The connection to the database may be configured with `DATABASE_URL`.
|
83
|
+
|
84
|
+
e.g. `postgres://user:password@database_host:1234/my-database?pool=5&encoding=unicode`
|
85
|
+
|
86
|
+
`LAST_N_DAYS`
|
87
|
+
|
88
|
+
The number of days to retrieve. When the ETL runs, it will fetch this many days
|
89
|
+
of data, prior to today. It can take a while for transactions to appear in the
|
90
|
+
API. It is recommended this be set to 60 in production. Defaults to 7.
|
91
|
+
|
92
|
+
`DEBUG_STREAM`
|
93
|
+
|
94
|
+
The stream to write debug output. Defaults to `stdout`.
|
95
|
+
|
96
|
+
Valid options are: `stdout`, `stderr` and `none`.
|
97
|
+
|
98
|
+
## How to contribute
|
99
|
+
|
100
|
+
Bug reports and pull requests are welcome on
|
101
|
+
[GitHub](https://github.com/reevoo/affiliate_window_etl). This project is
|
102
|
+
intended to be a safe, welcoming space for collaboration, and contributors are
|
103
|
+
expected to adhere to the
|
104
|
+
[Contributor Covenant](http://contributor-covenant.org/) code of conduct.
|
105
|
+
|
106
|
+
## License
|
107
|
+
|
108
|
+
The gem is available as open-source under the terms of the
|
109
|
+
[MIT License](http://opensource.org/licenses/MIT).
|
@@ -0,0 +1,11 @@
|
|
1
|
+
require "active_record"
|
2
|
+
require "affiliate_window"
|
3
|
+
|
4
|
+
require "affiliate_window/etl/base"
|
5
|
+
require "affiliate_window/etl/scheduler"
|
6
|
+
require "affiliate_window/etl/database"
|
7
|
+
require "affiliate_window/etl/extracter"
|
8
|
+
require "affiliate_window/etl/transformer"
|
9
|
+
require "affiliate_window/etl/normaliser"
|
10
|
+
require "affiliate_window/etl/loader"
|
11
|
+
require "affiliate_window/etl/config"
|
@@ -0,0 +1,78 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
attr_accessor :config
|
4
|
+
|
5
|
+
def initialize(env: ENV)
|
6
|
+
self.config = Config.new(env: env)
|
7
|
+
end
|
8
|
+
|
9
|
+
def run
|
10
|
+
database.connect!
|
11
|
+
|
12
|
+
scheduler.jobs.each do |job|
|
13
|
+
extracter.extract(job.type, job.args).each do |record|
|
14
|
+
transformer.transform(record).each do |transformed_record|
|
15
|
+
loader.load(transformed_record)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
end
|
19
|
+
end
|
20
|
+
|
21
|
+
def migrate
|
22
|
+
database.connect!
|
23
|
+
ActiveRecord::Migrator.migrate(migrations_path)
|
24
|
+
end
|
25
|
+
|
26
|
+
def migration_filenames
|
27
|
+
path = File.expand_path("../../../db/migrate/*", File.dirname(__FILE__))
|
28
|
+
Dir[path].each
|
29
|
+
end
|
30
|
+
|
31
|
+
private
|
32
|
+
|
33
|
+
def scheduler
|
34
|
+
Scheduler.new(
|
35
|
+
database: database,
|
36
|
+
last_n_days: config.last_n_days,
|
37
|
+
)
|
38
|
+
end
|
39
|
+
|
40
|
+
def extracter
|
41
|
+
Extracter.new(
|
42
|
+
client: client,
|
43
|
+
output: config.output_stream,
|
44
|
+
)
|
45
|
+
end
|
46
|
+
|
47
|
+
def transformer
|
48
|
+
Transformer.new(normaliser: normaliser)
|
49
|
+
end
|
50
|
+
|
51
|
+
def loader
|
52
|
+
Loader.new(database: database)
|
53
|
+
end
|
54
|
+
|
55
|
+
def database
|
56
|
+
Database.new(config.database_url)
|
57
|
+
end
|
58
|
+
|
59
|
+
def client
|
60
|
+
AffiliateWindow.login(
|
61
|
+
account_id: config.account_id,
|
62
|
+
affiliate_api_password: config.affiliate_api_password,
|
63
|
+
)
|
64
|
+
end
|
65
|
+
|
66
|
+
def schema
|
67
|
+
Schema.new
|
68
|
+
end
|
69
|
+
|
70
|
+
def normaliser
|
71
|
+
Normaliser.new
|
72
|
+
end
|
73
|
+
|
74
|
+
def migrator
|
75
|
+
Migrator.new
|
76
|
+
end
|
77
|
+
end
|
78
|
+
end
|
@@ -0,0 +1,37 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Config
|
4
|
+
attr_accessor :env
|
5
|
+
|
6
|
+
def initialize(env:)
|
7
|
+
self.env = env
|
8
|
+
end
|
9
|
+
|
10
|
+
def database_url
|
11
|
+
env.fetch(
|
12
|
+
"DATABASE_URL",
|
13
|
+
"postgres://#{`whoami`.strip}@localhost:5432/affiliate_window?pool=5&encoding=unicode",
|
14
|
+
)
|
15
|
+
end
|
16
|
+
|
17
|
+
def account_id
|
18
|
+
env.fetch("ACCOUNT_ID", 1234)
|
19
|
+
end
|
20
|
+
|
21
|
+
def affiliate_api_password
|
22
|
+
env.fetch("AFFILIATE_API_PASSWORD", "password")
|
23
|
+
end
|
24
|
+
|
25
|
+
def last_n_days
|
26
|
+
env.fetch("LAST_N_DAYS", "7").to_i
|
27
|
+
end
|
28
|
+
|
29
|
+
def output_stream
|
30
|
+
name = env.fetch("DEBUG_STREAM", "stdout")
|
31
|
+
name = name.downcase.to_sym
|
32
|
+
|
33
|
+
{ stdout: $stdout, stderr: $stderr, none: nil }.fetch(name)
|
34
|
+
end
|
35
|
+
end
|
36
|
+
end
|
37
|
+
end
|
@@ -0,0 +1,89 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Database
|
4
|
+
attr_accessor :database_url
|
5
|
+
|
6
|
+
def initialize(database_url)
|
7
|
+
self.database_url = database_url
|
8
|
+
end
|
9
|
+
|
10
|
+
def connect!
|
11
|
+
ActiveRecord::Base.establish_connection(database_url)
|
12
|
+
end
|
13
|
+
|
14
|
+
def model(record_type)
|
15
|
+
MODELS.fetch(record_type)
|
16
|
+
end
|
17
|
+
|
18
|
+
class ClickStat < ActiveRecord::Base
|
19
|
+
def self.identity
|
20
|
+
[:date, :merchant_name]
|
21
|
+
end
|
22
|
+
end
|
23
|
+
|
24
|
+
class CommissionGroup < ActiveRecord::Base
|
25
|
+
def self.identity
|
26
|
+
[:merchant_id, :commission_group_code]
|
27
|
+
end
|
28
|
+
end
|
29
|
+
|
30
|
+
class CommissionRange < ActiveRecord::Base
|
31
|
+
self.inheritance_column = :_disabled
|
32
|
+
|
33
|
+
def self.identity
|
34
|
+
[:merchant_id]
|
35
|
+
end
|
36
|
+
end
|
37
|
+
|
38
|
+
class ImpressionStat < ActiveRecord::Base
|
39
|
+
def self.identity
|
40
|
+
[:date, :merchant_name]
|
41
|
+
end
|
42
|
+
end
|
43
|
+
|
44
|
+
class Merchant < ActiveRecord::Base
|
45
|
+
def self.identity
|
46
|
+
[:id]
|
47
|
+
end
|
48
|
+
end
|
49
|
+
|
50
|
+
class MerchantSector < ActiveRecord::Base
|
51
|
+
def self.identity
|
52
|
+
[:merchant_id, :sector_id]
|
53
|
+
end
|
54
|
+
end
|
55
|
+
|
56
|
+
class Transaction < ActiveRecord::Base
|
57
|
+
self.inheritance_column = :_disabled
|
58
|
+
|
59
|
+
def self.identity
|
60
|
+
[:id]
|
61
|
+
end
|
62
|
+
end
|
63
|
+
|
64
|
+
class TransactionPart < ActiveRecord::Base
|
65
|
+
def self.identity
|
66
|
+
[:transaction_id, :commission_group_name]
|
67
|
+
end
|
68
|
+
end
|
69
|
+
|
70
|
+
class TransactionProduct < ActiveRecord::Base
|
71
|
+
def self.identity
|
72
|
+
[:id]
|
73
|
+
end
|
74
|
+
end
|
75
|
+
|
76
|
+
MODELS = {
|
77
|
+
click_stat: ClickStat,
|
78
|
+
commission_group: CommissionGroup,
|
79
|
+
commission_range: CommissionRange,
|
80
|
+
impression_stat: ImpressionStat,
|
81
|
+
merchant: Merchant,
|
82
|
+
merchant_sector: MerchantSector,
|
83
|
+
transaction: Transaction,
|
84
|
+
transaction_part: TransactionPart,
|
85
|
+
transaction_product: TransactionProduct,
|
86
|
+
}
|
87
|
+
end
|
88
|
+
end
|
89
|
+
end
|
@@ -0,0 +1,183 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Extracter # rubocop:disable Metrics/ClassLength
|
4
|
+
CHUNK_SIZE = 100
|
5
|
+
|
6
|
+
attr_accessor :client, :output
|
7
|
+
|
8
|
+
def initialize(client:, output: nil)
|
9
|
+
self.client = client
|
10
|
+
self.output = output
|
11
|
+
end
|
12
|
+
|
13
|
+
def extract(type, params = {})
|
14
|
+
Enumerator.new do |yielder|
|
15
|
+
public_send("extract_#{type}", yielder, params)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
|
19
|
+
def extract_merchants(yielder, _params) # rubocop:disable Metrics/AbcSize
|
20
|
+
response = client.get_merchant_list
|
21
|
+
merchants = response.fetch(:merchant)
|
22
|
+
merchant_ids = merchants.map { |m| m.fetch(:i_id) }
|
23
|
+
|
24
|
+
count = 0
|
25
|
+
merchant_ids.each_slice(CHUNK_SIZE) do |ids|
|
26
|
+
response = client.get_merchant(merchant_ids: ids)
|
27
|
+
merchants = response.fetch(:merchant)
|
28
|
+
|
29
|
+
merchants.each do |record|
|
30
|
+
yielder.yield(record.merge(record_type: :merchant))
|
31
|
+
end
|
32
|
+
|
33
|
+
count += [CHUNK_SIZE, ids.count].min
|
34
|
+
write "Extracted #{count} / #{merchant_ids.count} merchants"
|
35
|
+
end
|
36
|
+
|
37
|
+
extract_commission_groups(yielder, merchant_ids: merchant_ids)
|
38
|
+
end
|
39
|
+
|
40
|
+
def extract_commission_groups(yielder, merchant_ids:)
|
41
|
+
merchant_ids.each.with_index do |id, index|
|
42
|
+
maybe_response = catch_invalid_relationship_error do
|
43
|
+
client.get_commission_group_list(merchant_id: id)
|
44
|
+
end
|
45
|
+
|
46
|
+
next unless maybe_response
|
47
|
+
response = maybe_response
|
48
|
+
|
49
|
+
commission_groups = [response.fetch(:commission_group)].flatten
|
50
|
+
|
51
|
+
commission_groups.each do |record|
|
52
|
+
yielder.yield(record.merge(
|
53
|
+
record_type: :commission_group,
|
54
|
+
merchant_id: id,
|
55
|
+
))
|
56
|
+
end
|
57
|
+
|
58
|
+
write "Extracted commission groups for #{index + 1} / #{merchant_ids.count} merchants"
|
59
|
+
end
|
60
|
+
end
|
61
|
+
|
62
|
+
def extract_daily_transactions(yielder, date:)
|
63
|
+
response = client.get_transaction_list(
|
64
|
+
start_date: "#{date}T00:00:00",
|
65
|
+
end_date: "#{date}T23:59:59",
|
66
|
+
date_type: "transaction",
|
67
|
+
)
|
68
|
+
results = response.fetch(:results)
|
69
|
+
pagination = response.fetch(:pagination)
|
70
|
+
|
71
|
+
check_all_records_received!(pagination)
|
72
|
+
|
73
|
+
transactions = results.fetch(:transaction)
|
74
|
+
transactions.each do |record|
|
75
|
+
yielder.yield(record.merge(record_type: :transaction))
|
76
|
+
end
|
77
|
+
|
78
|
+
write "Extracted #{transactions.count} transactions for #{date}"
|
79
|
+
|
80
|
+
transaction_ids = transactions.map { |t| t.fetch(:i_id) }
|
81
|
+
extract_transaction_products(yielder, transaction_ids: transaction_ids)
|
82
|
+
end
|
83
|
+
|
84
|
+
def extract_transactions(yielder, transaction_ids:)
|
85
|
+
count = 0
|
86
|
+
transaction_ids.each_slice(CHUNK_SIZE) do |ids|
|
87
|
+
response = client.get_transaction(transaction_ids: ids)
|
88
|
+
|
89
|
+
transactions = response.fetch(:transaction)
|
90
|
+
transactions.each do |record|
|
91
|
+
yielder.yield(record.merge(record_type: :transaction))
|
92
|
+
end
|
93
|
+
|
94
|
+
count += [CHUNK_SIZE, ids.count].min
|
95
|
+
write "Extracted #{count} / #{transaction_ids.count} transactions"
|
96
|
+
end
|
97
|
+
|
98
|
+
extract_transaction_products(yielder, transaction_ids: transaction_ids)
|
99
|
+
end
|
100
|
+
|
101
|
+
def extract_transaction_products(yielder, transaction_ids:)
|
102
|
+
count = 0
|
103
|
+
transaction_ids.each_slice(CHUNK_SIZE) do |ids|
|
104
|
+
response = client.get_transaction_product(transaction_ids: ids)
|
105
|
+
transaction_products = [response.fetch(:transaction_product)].flatten
|
106
|
+
|
107
|
+
transaction_products.each do |record|
|
108
|
+
yielder.yield(record.merge(record_type: :transaction_product))
|
109
|
+
end
|
110
|
+
|
111
|
+
count += [CHUNK_SIZE, ids.count].min
|
112
|
+
write "Extracted #{transaction_products.count} products for #{count} / #{transaction_ids.count} transactions"
|
113
|
+
end
|
114
|
+
end
|
115
|
+
|
116
|
+
def extract_daily_clicks(yielder, date:)
|
117
|
+
response = client.get_click_stats(
|
118
|
+
start_date: "#{date}T00:00:00",
|
119
|
+
end_date: "#{date}T23:59:59",
|
120
|
+
date_type: "transaction",
|
121
|
+
)
|
122
|
+
results = response.fetch(:results)
|
123
|
+
pagination = response.fetch(:pagination)
|
124
|
+
|
125
|
+
check_all_records_received!(pagination)
|
126
|
+
|
127
|
+
click_stats = results.fetch(:click_stats)
|
128
|
+
click_stats.each do |record|
|
129
|
+
yielder.yield(record.merge(
|
130
|
+
record_type: :click_stat,
|
131
|
+
date: date,
|
132
|
+
))
|
133
|
+
end
|
134
|
+
|
135
|
+
write "Extracted #{click_stats.count} click stats for #{date}"
|
136
|
+
end
|
137
|
+
|
138
|
+
def extract_daily_impressions(yielder, date:)
|
139
|
+
response = client.get_impression_stats(
|
140
|
+
start_date: "#{date}T00:00:00",
|
141
|
+
end_date: "#{date}T23:59:59",
|
142
|
+
date_type: "transaction",
|
143
|
+
)
|
144
|
+
results = response.fetch(:results)
|
145
|
+
pagination = response.fetch(:pagination)
|
146
|
+
|
147
|
+
check_all_records_received!(pagination)
|
148
|
+
|
149
|
+
impression_stats = results.fetch(:impression_stats)
|
150
|
+
impression_stats.each do |record|
|
151
|
+
yielder.yield(record.merge(
|
152
|
+
record_type: :impression_stat,
|
153
|
+
date: date,
|
154
|
+
))
|
155
|
+
end
|
156
|
+
|
157
|
+
write "Extracted #{impression_stats.count} impression stats for #{date}"
|
158
|
+
end
|
159
|
+
|
160
|
+
def check_all_records_received!(pagination)
|
161
|
+
retrieved = pagination.fetch(:i_rows_returned)
|
162
|
+
total = pagination.fetch(:i_rows_available)
|
163
|
+
|
164
|
+
fail "Did not receive all records: #{retrieved} retrieved out of #{total}" unless total == retrieved
|
165
|
+
end
|
166
|
+
|
167
|
+
# If the current account is not affiliated with the merchant, the API does
|
168
|
+
# not let you retrieve commission groups for that merchant.
|
169
|
+
def catch_invalid_relationship_error(&block)
|
170
|
+
block.call
|
171
|
+
rescue AffiliateWindow::Error => e
|
172
|
+
raise unless e.message.match(/Invalid merchant \/ affiliate relationship/)
|
173
|
+
nil
|
174
|
+
end
|
175
|
+
|
176
|
+
def write(message)
|
177
|
+
return unless output
|
178
|
+
message_with_quota = "[quota:#{client.remaining_quota}] #{message}"
|
179
|
+
output.puts(message_with_quota)
|
180
|
+
end
|
181
|
+
end
|
182
|
+
end
|
183
|
+
end
|
@@ -0,0 +1,25 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Loader
|
4
|
+
attr_accessor :database
|
5
|
+
|
6
|
+
def initialize(database:)
|
7
|
+
self.database = database
|
8
|
+
end
|
9
|
+
|
10
|
+
def load(attributes)
|
11
|
+
record_type = attributes.delete(:record_type)
|
12
|
+
attributes.delete_if { |_, v| v.nil? }
|
13
|
+
|
14
|
+
model = database.model(record_type)
|
15
|
+
identity = attributes.slice(*model.identity)
|
16
|
+
|
17
|
+
if (record = model.find_by(identity))
|
18
|
+
record.update!(attributes)
|
19
|
+
else
|
20
|
+
model.create!(attributes)
|
21
|
+
end
|
22
|
+
end
|
23
|
+
end
|
24
|
+
end
|
25
|
+
end
|
@@ -0,0 +1,22 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Normaliser
|
4
|
+
def normalise!(record, field_name:, nested_name:, foreign_name:, id_name: :i_id, record_type: nil) # rubocop:disable Metrics/ParameterLists, Metrics/LineLength
|
5
|
+
record_type ||= nested_name
|
6
|
+
|
7
|
+
value = record.delete(field_name)
|
8
|
+
return [] unless value
|
9
|
+
|
10
|
+
elements = [value.fetch(nested_name)].flatten
|
11
|
+
foreign_id = record.fetch(id_name)
|
12
|
+
|
13
|
+
elements.map do |attributes|
|
14
|
+
attributes.merge(
|
15
|
+
record_type: record_type,
|
16
|
+
foreign_name => foreign_id,
|
17
|
+
)
|
18
|
+
end
|
19
|
+
end
|
20
|
+
end
|
21
|
+
end
|
22
|
+
end
|
@@ -0,0 +1,67 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Scheduler
|
4
|
+
attr_accessor :database, :last_n_days
|
5
|
+
|
6
|
+
def initialize(database:, last_n_days:)
|
7
|
+
self.database = database
|
8
|
+
self.last_n_days = last_n_days
|
9
|
+
end
|
10
|
+
|
11
|
+
def jobs
|
12
|
+
jobs = []
|
13
|
+
|
14
|
+
schedule_last_n_days(:daily_transactions, jobs)
|
15
|
+
schedule_last_n_days(:daily_clicks, jobs)
|
16
|
+
schedule_last_n_days(:daily_impressions, jobs)
|
17
|
+
schedule_old_pending_transactions(jobs)
|
18
|
+
schedule_merchants(jobs)
|
19
|
+
|
20
|
+
jobs
|
21
|
+
end
|
22
|
+
|
23
|
+
private
|
24
|
+
|
25
|
+
def schedule_last_n_days(type, jobs)
|
26
|
+
today = Date.today
|
27
|
+
|
28
|
+
(n_days_ago..today).each do |date|
|
29
|
+
job = Job.new(type, date: date.to_s)
|
30
|
+
jobs.push(job)
|
31
|
+
end
|
32
|
+
end
|
33
|
+
|
34
|
+
def schedule_old_pending_transactions(jobs)
|
35
|
+
model = database.model(:transaction)
|
36
|
+
pending = model.where(status: "pending")
|
37
|
+
old_pending = pending.where("transaction_date < ?", n_days_ago)
|
38
|
+
|
39
|
+
transaction_ids = old_pending.pluck(:id)
|
40
|
+
|
41
|
+
job = Job.new(:transactions, transaction_ids: transaction_ids)
|
42
|
+
jobs.push(job)
|
43
|
+
end
|
44
|
+
|
45
|
+
def schedule_merchants(jobs)
|
46
|
+
jobs.push(Job.new(:merchants))
|
47
|
+
end
|
48
|
+
|
49
|
+
def n_days_ago
|
50
|
+
Date.today - last_n_days + 1
|
51
|
+
end
|
52
|
+
|
53
|
+
class Job
|
54
|
+
attr_accessor :type, :args
|
55
|
+
|
56
|
+
def initialize(type, args = {})
|
57
|
+
self.type = type
|
58
|
+
self.args = args
|
59
|
+
end
|
60
|
+
|
61
|
+
def ==(other)
|
62
|
+
type == other.type && args == other.args
|
63
|
+
end
|
64
|
+
end
|
65
|
+
end
|
66
|
+
end
|
67
|
+
end
|
@@ -0,0 +1,135 @@
|
|
1
|
+
class AffiliateWindow
|
2
|
+
class ETL
|
3
|
+
class Transformer # rubocop:disable Metrics/ClassLength
|
4
|
+
attr_accessor :normaliser
|
5
|
+
|
6
|
+
def initialize(normaliser:)
|
7
|
+
self.normaliser = normaliser
|
8
|
+
end
|
9
|
+
|
10
|
+
def transform(record)
|
11
|
+
record = record.dup
|
12
|
+
record_type = record.fetch(:record_type)
|
13
|
+
|
14
|
+
transformed_records = []
|
15
|
+
|
16
|
+
case record_type
|
17
|
+
when :merchant
|
18
|
+
normalise_commision_ranges!(record, transformed_records)
|
19
|
+
normalise_sectors!(record, transformed_records)
|
20
|
+
when :transaction
|
21
|
+
normalise_transaction_parts!(record, transformed_records)
|
22
|
+
when :transaction_product
|
23
|
+
normalise_transaction_products!(record, transformed_records)
|
24
|
+
|
25
|
+
# transaction_product has no other top-level attributes
|
26
|
+
return transformed_records
|
27
|
+
end
|
28
|
+
|
29
|
+
attributes = infer_field_names(record)
|
30
|
+
transformed_records.push(attributes)
|
31
|
+
end
|
32
|
+
|
33
|
+
private
|
34
|
+
|
35
|
+
def normalise_commision_ranges!(record, transformed_records)
|
36
|
+
commision_ranges = normaliser.normalise!(
|
37
|
+
record,
|
38
|
+
field_name: :a_commission_ranges,
|
39
|
+
nested_name: :commission_range,
|
40
|
+
foreign_name: :merchant_id,
|
41
|
+
)
|
42
|
+
|
43
|
+
commision_ranges.each do |commision_range|
|
44
|
+
attributes = infer_field_names(commision_range)
|
45
|
+
transformed_records.push(attributes)
|
46
|
+
end
|
47
|
+
end
|
48
|
+
|
49
|
+
def normalise_sectors!(record, transformed_records)
|
50
|
+
sectors = normaliser.normalise!(
|
51
|
+
record,
|
52
|
+
field_name: :a_sectors,
|
53
|
+
nested_name: :merchant_sector,
|
54
|
+
foreign_name: :merchant_id,
|
55
|
+
)
|
56
|
+
|
57
|
+
sectors.each do |sector|
|
58
|
+
attributes = infer_field_names(sector)
|
59
|
+
transformed_records.push(attributes)
|
60
|
+
end
|
61
|
+
end
|
62
|
+
|
63
|
+
def normalise_transaction_parts!(record, transformed_records)
|
64
|
+
transaction_parts = normaliser.normalise!(
|
65
|
+
record,
|
66
|
+
field_name: :a_transaction_parts,
|
67
|
+
nested_name: :transaction_part,
|
68
|
+
foreign_name: :transaction_id,
|
69
|
+
)
|
70
|
+
|
71
|
+
transaction_parts.each do |transaction_part|
|
72
|
+
attributes = infer_field_names(transaction_part)
|
73
|
+
transformed_records.push(attributes)
|
74
|
+
end
|
75
|
+
end
|
76
|
+
|
77
|
+
def normalise_transaction_products!(record, transformed_records)
|
78
|
+
transaction_products = normaliser.normalise!(
|
79
|
+
record,
|
80
|
+
field_name: :a_products,
|
81
|
+
nested_name: :product,
|
82
|
+
foreign_name: :transaction_id,
|
83
|
+
id_name: :i_transaction_id,
|
84
|
+
record_type: :transaction_product,
|
85
|
+
)
|
86
|
+
|
87
|
+
transaction_products.each do |transaction_product|
|
88
|
+
attributes = infer_field_names(transaction_product)
|
89
|
+
transformed_records.push(attributes)
|
90
|
+
end
|
91
|
+
end
|
92
|
+
|
93
|
+
def infer_field_names(record, prefix = nil)
|
94
|
+
record.keys.each.with_object({}) do |field_name, hash|
|
95
|
+
value = record.fetch(field_name)
|
96
|
+
new_name = new_field_name(field_name)
|
97
|
+
|
98
|
+
case value
|
99
|
+
when Hash
|
100
|
+
sub_record = record.fetch(field_name)
|
101
|
+
sub_prefix = "#{prefix}#{new_name}_"
|
102
|
+
sub_attributes = infer_field_names(sub_record, sub_prefix)
|
103
|
+
|
104
|
+
hash.merge!(sub_attributes)
|
105
|
+
when Array
|
106
|
+
fail arrays_unsupported_error(field_name, value)
|
107
|
+
else
|
108
|
+
new_name = "#{prefix}#{new_name}".to_sym
|
109
|
+
attributes = record.fetch(field_name, nil)
|
110
|
+
|
111
|
+
hash.merge!(new_name => attributes)
|
112
|
+
end
|
113
|
+
end
|
114
|
+
end
|
115
|
+
|
116
|
+
def new_field_name(field_name)
|
117
|
+
if field_name.to_s[1] == "_"
|
118
|
+
field_name.to_s[2..-1]
|
119
|
+
else
|
120
|
+
field_name
|
121
|
+
end
|
122
|
+
end
|
123
|
+
|
124
|
+
def arrays_unsupported_error(field_name, _array)
|
125
|
+
message = "Unable to transform '#{field_name}' because its value is an array.\n"
|
126
|
+
message += "To cope with this, normalise elements of the array into a separate table.\n"
|
127
|
+
message += "Then, add a foreign key from the normalised record to this one."
|
128
|
+
|
129
|
+
TypeError.new(message)
|
130
|
+
end
|
131
|
+
|
132
|
+
class TypeError < StandardError; end
|
133
|
+
end
|
134
|
+
end
|
135
|
+
end
|
@@ -0,0 +1 @@
|
|
1
|
+
require "affiliate_window/etl"
|
metadata
ADDED
@@ -0,0 +1,211 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: affiliate_window_etl
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Reevoo Developers
|
8
|
+
autorequire:
|
9
|
+
bindir: bin
|
10
|
+
cert_chain: []
|
11
|
+
date: 2016-10-12 00:00:00.000000000 Z
|
12
|
+
dependencies:
|
13
|
+
- !ruby/object:Gem::Dependency
|
14
|
+
name: affiliate_window
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
16
|
+
requirements:
|
17
|
+
- - "~>"
|
18
|
+
- !ruby/object:Gem::Version
|
19
|
+
version: '0.1'
|
20
|
+
type: :runtime
|
21
|
+
prerelease: false
|
22
|
+
version_requirements: !ruby/object:Gem::Requirement
|
23
|
+
requirements:
|
24
|
+
- - "~>"
|
25
|
+
- !ruby/object:Gem::Version
|
26
|
+
version: '0.1'
|
27
|
+
- !ruby/object:Gem::Dependency
|
28
|
+
name: activerecord
|
29
|
+
requirement: !ruby/object:Gem::Requirement
|
30
|
+
requirements:
|
31
|
+
- - "~>"
|
32
|
+
- !ruby/object:Gem::Version
|
33
|
+
version: '5.0'
|
34
|
+
type: :runtime
|
35
|
+
prerelease: false
|
36
|
+
version_requirements: !ruby/object:Gem::Requirement
|
37
|
+
requirements:
|
38
|
+
- - "~>"
|
39
|
+
- !ruby/object:Gem::Version
|
40
|
+
version: '5.0'
|
41
|
+
- !ruby/object:Gem::Dependency
|
42
|
+
name: pg
|
43
|
+
requirement: !ruby/object:Gem::Requirement
|
44
|
+
requirements:
|
45
|
+
- - "~>"
|
46
|
+
- !ruby/object:Gem::Version
|
47
|
+
version: '0.19'
|
48
|
+
type: :runtime
|
49
|
+
prerelease: false
|
50
|
+
version_requirements: !ruby/object:Gem::Requirement
|
51
|
+
requirements:
|
52
|
+
- - "~>"
|
53
|
+
- !ruby/object:Gem::Version
|
54
|
+
version: '0.19'
|
55
|
+
- !ruby/object:Gem::Dependency
|
56
|
+
name: rspec
|
57
|
+
requirement: !ruby/object:Gem::Requirement
|
58
|
+
requirements:
|
59
|
+
- - "~>"
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: '3.5'
|
62
|
+
type: :development
|
63
|
+
prerelease: false
|
64
|
+
version_requirements: !ruby/object:Gem::Requirement
|
65
|
+
requirements:
|
66
|
+
- - "~>"
|
67
|
+
- !ruby/object:Gem::Version
|
68
|
+
version: '3.5'
|
69
|
+
- !ruby/object:Gem::Dependency
|
70
|
+
name: pry
|
71
|
+
requirement: !ruby/object:Gem::Requirement
|
72
|
+
requirements:
|
73
|
+
- - "~>"
|
74
|
+
- !ruby/object:Gem::Version
|
75
|
+
version: '0.10'
|
76
|
+
type: :development
|
77
|
+
prerelease: false
|
78
|
+
version_requirements: !ruby/object:Gem::Requirement
|
79
|
+
requirements:
|
80
|
+
- - "~>"
|
81
|
+
- !ruby/object:Gem::Version
|
82
|
+
version: '0.10'
|
83
|
+
- !ruby/object:Gem::Dependency
|
84
|
+
name: rake
|
85
|
+
requirement: !ruby/object:Gem::Requirement
|
86
|
+
requirements:
|
87
|
+
- - "~>"
|
88
|
+
- !ruby/object:Gem::Version
|
89
|
+
version: '11.3'
|
90
|
+
type: :development
|
91
|
+
prerelease: false
|
92
|
+
version_requirements: !ruby/object:Gem::Requirement
|
93
|
+
requirements:
|
94
|
+
- - "~>"
|
95
|
+
- !ruby/object:Gem::Version
|
96
|
+
version: '11.3'
|
97
|
+
- !ruby/object:Gem::Dependency
|
98
|
+
name: timecop
|
99
|
+
requirement: !ruby/object:Gem::Requirement
|
100
|
+
requirements:
|
101
|
+
- - "~>"
|
102
|
+
- !ruby/object:Gem::Version
|
103
|
+
version: '0.8'
|
104
|
+
type: :development
|
105
|
+
prerelease: false
|
106
|
+
version_requirements: !ruby/object:Gem::Requirement
|
107
|
+
requirements:
|
108
|
+
- - "~>"
|
109
|
+
- !ruby/object:Gem::Version
|
110
|
+
version: '0.8'
|
111
|
+
- !ruby/object:Gem::Dependency
|
112
|
+
name: bundler-audit
|
113
|
+
requirement: !ruby/object:Gem::Requirement
|
114
|
+
requirements:
|
115
|
+
- - "~>"
|
116
|
+
- !ruby/object:Gem::Version
|
117
|
+
version: '0.5'
|
118
|
+
type: :development
|
119
|
+
prerelease: false
|
120
|
+
version_requirements: !ruby/object:Gem::Requirement
|
121
|
+
requirements:
|
122
|
+
- - "~>"
|
123
|
+
- !ruby/object:Gem::Version
|
124
|
+
version: '0.5'
|
125
|
+
- !ruby/object:Gem::Dependency
|
126
|
+
name: reevoocop
|
127
|
+
requirement: !ruby/object:Gem::Requirement
|
128
|
+
requirements:
|
129
|
+
- - "~>"
|
130
|
+
- !ruby/object:Gem::Version
|
131
|
+
version: 0.0.8
|
132
|
+
type: :development
|
133
|
+
prerelease: false
|
134
|
+
version_requirements: !ruby/object:Gem::Requirement
|
135
|
+
requirements:
|
136
|
+
- - "~>"
|
137
|
+
- !ruby/object:Gem::Version
|
138
|
+
version: 0.0.8
|
139
|
+
- !ruby/object:Gem::Dependency
|
140
|
+
name: simplecov
|
141
|
+
requirement: !ruby/object:Gem::Requirement
|
142
|
+
requirements:
|
143
|
+
- - "~>"
|
144
|
+
- !ruby/object:Gem::Version
|
145
|
+
version: '0.11'
|
146
|
+
type: :development
|
147
|
+
prerelease: false
|
148
|
+
version_requirements: !ruby/object:Gem::Requirement
|
149
|
+
requirements:
|
150
|
+
- - "~>"
|
151
|
+
- !ruby/object:Gem::Version
|
152
|
+
version: '0.11'
|
153
|
+
- !ruby/object:Gem::Dependency
|
154
|
+
name: simplecov-console
|
155
|
+
requirement: !ruby/object:Gem::Requirement
|
156
|
+
requirements:
|
157
|
+
- - "~>"
|
158
|
+
- !ruby/object:Gem::Version
|
159
|
+
version: '0.3'
|
160
|
+
type: :development
|
161
|
+
prerelease: false
|
162
|
+
version_requirements: !ruby/object:Gem::Requirement
|
163
|
+
requirements:
|
164
|
+
- - "~>"
|
165
|
+
- !ruby/object:Gem::Version
|
166
|
+
version: '0.3'
|
167
|
+
description: An ETL for retrieving records from the Affiliate Window APIand loading
|
168
|
+
them into a Postgres database.
|
169
|
+
email: developers@reevoo.com
|
170
|
+
executables: []
|
171
|
+
extensions: []
|
172
|
+
extra_rdoc_files: []
|
173
|
+
files:
|
174
|
+
- README.md
|
175
|
+
- lib/affiliate_window/etl.rb
|
176
|
+
- lib/affiliate_window/etl/base.rb
|
177
|
+
- lib/affiliate_window/etl/config.rb
|
178
|
+
- lib/affiliate_window/etl/database.rb
|
179
|
+
- lib/affiliate_window/etl/extracter.rb
|
180
|
+
- lib/affiliate_window/etl/loader.rb
|
181
|
+
- lib/affiliate_window/etl/normaliser.rb
|
182
|
+
- lib/affiliate_window/etl/scheduler.rb
|
183
|
+
- lib/affiliate_window/etl/tasks.rake
|
184
|
+
- lib/affiliate_window/etl/transformer.rb
|
185
|
+
- lib/affiliate_window/etl/version.rb
|
186
|
+
- lib/affiliate_window_etl.rb
|
187
|
+
homepage: https://github.com/reevoo/affiliate_window_etl
|
188
|
+
licenses:
|
189
|
+
- MIT
|
190
|
+
metadata: {}
|
191
|
+
post_install_message:
|
192
|
+
rdoc_options: []
|
193
|
+
require_paths:
|
194
|
+
- lib
|
195
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
196
|
+
requirements:
|
197
|
+
- - ">="
|
198
|
+
- !ruby/object:Gem::Version
|
199
|
+
version: '0'
|
200
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
201
|
+
requirements:
|
202
|
+
- - ">="
|
203
|
+
- !ruby/object:Gem::Version
|
204
|
+
version: '0'
|
205
|
+
requirements: []
|
206
|
+
rubyforge_project:
|
207
|
+
rubygems_version: 2.6.7
|
208
|
+
signing_key:
|
209
|
+
specification_version: 4
|
210
|
+
summary: Affiliate Window ETL
|
211
|
+
test_files: []
|