Class: Embulk::Input::GoogleAnalytics::Client

Inherits:
Object
  • Object
show all
Defined in:
lib/embulk/input/google_analytics/client.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(task, is_preview = false) ⇒ Client

Returns a new instance of Client.



16
17
18
19
 
# File 'lib/embulk/input/google_analytics/client.rb', line 16

def initialize(task, is_preview = false)
  @task = task
  @is_preview = is_preview
end

Instance Attribute Details

#task ⇒ Object (readonly)

Returns the value of attribute task.



14
15
16
 
# File 'lib/embulk/input/google_analytics/client.rb', line 14

def task
  @task
end

Instance Method Details

#auth ⇒ Object 



173
174
175
176
177
178
179
180
181
182
 
# File 'lib/embulk/input/google_analytics/client.rb', line 173

def auth
  retryer.with_retry do
    Google::Auth::ServiceAccountCredentials.make_creds(
      json_key_io: StringIO.new(task["json_key_content"]),
      scope: "https://www.googleapis.com/auth/analytics.readonly"
    )
  end
rescue Google::Apis::AuthorizationError => e
  raise ConfigError.new(e.message)
end

#build_report_request(page_token = nil) ⇒ Object 



146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
 
# File 'lib/embulk/input/google_analytics/client.rb', line 146

def build_report_request(page_token = nil)
  query = {
    view_id: view_id,
    dimensions: [{name: task["time_series"]}] + task["dimensions"].map{|d| {name: d}},
    metrics: task["metrics"].map{|m| {expression: m}},
    include_empty_rows: true,
    page_size: preview? ? 10 : 10000,
  }

  if task["start_date"] || task["end_date"]
    query[:date_ranges] = [{
      start_date: task["start_date"],
      end_date: task["end_date"],
    }]
  end

  if page_token
    query[:page_token] = page_token
  end

  [query]
end

#each_report_row(&block) ⇒ Object 



25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
 
# File 'lib/embulk/input/google_analytics/client.rb', line 25

def each_report_row(&block)
  page_token = nil
  Embulk.logger.info "view_id:#{view_id} timezone has been set as '#{get_profile[:timezone]}'"

  loop do
    result = get_reports(page_token)
    report = result.to_h[:reports].first

    unless page_token
      # display for first request only
      Embulk.logger.info "Total: #{report[:data][:row_count]} rows. Fetched first response"
    end

    if !report[:data].has_key?(:rows)
      Embulk.logger.warn "Result doesn't contain rows."
      break
    end

    if report[:data][:rows].empty?
      Embulk.logger.warn "Result has 0 rows."
      break
    end

    dimensions = report[:column_header][:dimensions]
    metrics = report[:column_header][:metric_header][:metric_header_entries].map{|m| m[:name]}
    report[:data][:rows].each do |row|
      dim = dimensions.zip(row[:dimensions]).to_h
      met = metrics.zip(row[:metrics].first[:values]).to_h
      format_row = dim.merge(met)
      raw_time = format_row[task["time_series"]]
      next if too_early_data?(raw_time)
      format_row[task["time_series"]] = time_parse_with_profile_timezone(raw_time)
      format_row["view_id"] = view_id
      block.call format_row
    end

    break if preview?

    unless page_token = report[:next_page_token]
      break
    end
    Embulk.logger.info "Fetching report with page_token: #{page_token}"
  end
end

#get_all_profiles ⇒ Object 



85
86
87
88
89
90
91
92
93
 
# File 'lib/embulk/input/google_analytics/client.rb', line 85

def get_all_profiles
  service = Google::Apis::AnalyticsV3::AnalyticsService.new
  service.authorization = auth

  Embulk.logger.debug "Fetching profile from API"
  retryer.with_retry do
    service.list_profiles("~all", "~all")
  end
end

#get_columns_list ⇒ Object 



124
125
126
 
# File 'lib/embulk/input/google_analytics/client.rb', line 124

def get_columns_list
  get_custom_dimensions + get_metadata_columns
end

#get_custom_dimensions ⇒ Object 



137
138
139
140
141
142
143
144
 
# File 'lib/embulk/input/google_analytics/client.rb', line 137

def get_custom_dimensions
  # https://developers.google.com/analytics/devguides/config/mgmt/v3/mgmtReference/management/customDimensions/list
  service = Google::Apis::AnalyticsV3::AnalyticsService.new
  service.authorization = auth
  retryer.with_retry do
    service.list_custom_dimensions(get_profile[:account_id], get_profile[:web_property_id]).to_h[:items]
  end
end

#get_metadata_columns ⇒ Object 



128
129
130
131
132
133
134
135
 
# File 'lib/embulk/input/google_analytics/client.rb', line 128

def get_metadata_columns
  # https://developers.google.com/analytics/devguides/reporting/metadata/v3/reference/metadata/columns/list
  service = Google::Apis::AnalyticsV3::AnalyticsService.new
  service.authorization = auth
  retryer.with_retry do
    service.list_metadata_columns("ga").to_h[:items]
  end
end

#get_profile ⇒ Object 



70
71
72
73
74
75
76
77
78
79
80
81
82
83
 
# File 'lib/embulk/input/google_analytics/client.rb', line 70

def get_profile
  @profile ||=
    begin
      profile = get_all_profiles.to_h[:items].find do |prof|
        prof[:id] == view_id
      end

      unless profile
        raise Embulk::ConfigError.new("Can't find view_id:#{view_id} profile via Google Analytics API.")
      end

      profile
    end
end

#get_reports(page_token = nil) ⇒ Object 



110
111
112
113
114
115
116
117
118
119
120
121
122
 
# File 'lib/embulk/input/google_analytics/client.rb', line 110

def get_reports(page_token = nil)
  # https://developers.google.com/analytics/devguides/reporting/core/v4/rest/v4/reports/batchGet
  service = Google::Apis::AnalyticsreportingV4::AnalyticsReportingService.new
  service.authorization = auth

  request = Google::Apis::AnalyticsreportingV4::GetReportsRequest.new
  request.report_requests = build_report_request(page_token)

  Embulk.logger.info "Query to Core Report API: #{request.to_json}"
  retryer.with_retry do
    service.batch_get_reports request
  end
end

#preview? ⇒ Boolean

Returns:

  • (Boolean) 


21
22
23
 
# File 'lib/embulk/input/google_analytics/client.rb', line 21

def preview?
  @is_preview
end

#retryer ⇒ Object 



205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
 
# File 'lib/embulk/input/google_analytics/client.rb', line 205

def retryer
  PerfectRetry.new do |config|
    config.limit = task["retry_limit"]
    config.logger = Embulk.logger
    config.log_level = nil

    # https://developers.google.com/analytics/devguides/reporting/core/v4/errors
    # https://developers.google.com/analytics/devguides/reporting/core/v4/limits-quotas#additional_quota
    # https://github.com/google/google-api-ruby-client/blob/master/lib/google/apis/errors.rb
    # https://github.com/google/google-api-ruby-client/blob/0.9.11/lib/google/apis/core/http_command.rb#L33
    config.rescues = Google::Apis::Core::HttpCommand::RETRIABLE_ERRORS
    config.dont_rescues = [Embulk::DataError, Embulk::ConfigError]
    config.sleep = lambda{|n| task["retry_initial_wait_sec"]* (2 ** (n-1)) }
    config.raise_original_error = true
  end
end

#swap_time_zone(&block) ⇒ Object 



184
185
186
187
188
189
190
 
# File 'lib/embulk/input/google_analytics/client.rb', line 184

def swap_time_zone(&block)
  orig_timezone = Time.zone
  Time.zone = get_profile[:timezone]
  yield
ensure
  Time.zone = orig_timezone
end

#time_parse_with_profile_timezone(time_string) ⇒ Object 



95
96
97
98
99
100
101
102
103
104
105
106
107
108
 
# File 'lib/embulk/input/google_analytics/client.rb', line 95

def time_parse_with_profile_timezone(time_string)
  date_format =
    case task["time_series"]
    when "ga:dateHour"
      "%Y%m%d%H"
    when "ga:date"
      "%Y%m%d"
    end
  parts = Date._strptime(time_string, date_format)

  swap_time_zone do
    Time.zone.local(*parts.values_at(:year, :mon, :mday, :hour)).to_time
  end
end

#too_early_data?(time_str) ⇒ Boolean

Returns:

  • (Boolean) 


192
193
194
195
196
197
198
199
200
201
202
203
 
# File 'lib/embulk/input/google_analytics/client.rb', line 192

def too_early_data?(time_str)
  # fetching 20160720 data on 2016-07-20, it is too early fetching
  swap_time_zone do
    now = Time.zone.now
    case task["time_series"]
    when "ga:dateHour"
      time_str.to_i >= now.strftime("%Y%m%d%H").to_i
    when "ga:date"
      time_str.to_i >= now.strftime("%Y%m%d").to_i
    end
  end
end

#view_id ⇒ Object 



169
170
171
 
# File 'lib/embulk/input/google_analytics/client.rb', line 169

def view_id
  task["view_id"]
end