Przeglądaj źródła

BigQuery cleanup

Andrew Kane 7 lat temu
rodzic
commit
325207c943

+ 4 - 0
CHANGELOG.md

@@ -1,3 +1,7 @@
+## 1.7.10 [unreleased]
+
+- Added support for Google BigQuery
+
 ## 1.7.9
 
 - Added beta support for Apache Drill

+ 4 - 5
README.md

@@ -392,7 +392,7 @@ data_sources:
 - [Redshift](#redshift)
 - [Presto](#presto)
 - [Apache Drill](#apache-drill)
-- [Google BigQuery](#google-bigquery)
+- [Google BigQuery](#google-bigquery-master)
 - [MongoDB](#mongodb-1)
 - [Elasticsearch](#elasticsearch) [beta]
 
@@ -485,7 +485,7 @@ data_sources:
     url: http://hostname:8047
 ```
 
-### Google BigQuery
+### Google BigQuery [master]
 
 Add [google-cloud-bigquery](https://github.com/GoogleCloudPlatform/google-cloud-ruby/tree/master/google-cloud-bigquery) to your Gemfile and set:
 
@@ -493,9 +493,8 @@ Add [google-cloud-bigquery](https://github.com/GoogleCloudPlatform/google-cloud-
 data_sources:
   my_source:
     adapter: bigquery
-    url: "" # required for blazer
-    project: "your_project"
-    keyfile: "json credentials file"
+    project: your-project
+    keyfile: path/to/keyfile.json
 ```
 
 ### MongoDB

+ 1 - 1
app/views/blazer/queries/show.html.erb

@@ -62,7 +62,7 @@
   </script>
 <% end %>
 
-<% if %w[sql presto drill].include?(Blazer.data_sources[@query.data_source].adapter) %>
+<% if %w[sql presto drill bigquery].include?(Blazer.data_sources[@query.data_source].adapter) %>
   <script>
     // do not highlight really long queries
     // this can lead to performance issues

+ 17 - 30
lib/blazer/adapters/bigquery_adapter.rb

@@ -5,28 +5,30 @@ module Blazer
         columns = []
         rows = []
         error = nil
+
         begin
-          results = bigquery.query(statement, timeout: timeout_ms)
+          results = bigquery.query(statement, timeout: 30000) # ms
           columns = results.first.keys.map(&:to_s) if results.size > 0
           rows = results.map(&:values)
-        rescue StandardError => e
+        rescue => e
           error = e.message
         end
+
         [columns, rows, error]
       end
 
       def tables
-        table_refs.map{|t| "#{t.project_id}.#{t.dataset_id}.#{t.table_id}" }
+        table_refs.map { |t| "#{t.project_id}.#{t.dataset_id}.#{t.table_id}" }
       end
 
       def schema
-        table_refs.map{|table_ref| 
+        table_refs.map do |table_ref|
           {
             schema: table_ref.dataset_id,
             table: table_ref.table_id,
             columns: table_columns(table_ref)
           }
-        }
+        end
       end
 
       def preview_statement
@@ -36,38 +38,23 @@ module Blazer
       private
 
       def bigquery
-        @bigquery ||= connect!
-      end
-
-      def connect!
-        require "google/cloud/bigquery"
-        params = { project: settings["project"], keyfile: settings["keyfile"] }
-        @bigquery = Google::Cloud::Bigquery.new(params)
-        ::Google::Apis.logger.level = Logger::INFO
-        @bigquery
+        @bigquery ||= begin
+          require "google/cloud/bigquery"
+          Google::Cloud::Bigquery.new(
+            project: settings["project"],
+            keyfile: settings["keyfile"]
+          )
+        end
       end
 
       def table_refs
-        bigquery
-          .datasets
-          .map(&:tables)
-          .flat_map { |table_list| table_list.map(&:table_ref) }
+        bigquery.datasets.map(&:tables).flat_map { |table_list| table_list.map(&:table_ref) }
       end
 
       def table_columns(table_ref)
-        schema = 
-          bigquery
-          .service
-          .get_table(table_ref.dataset_id, table_ref.table_id)
-          .schema
+        schema = bigquery.service.get_table(table_ref.dataset_id, table_ref.table_id).schema
         return [] if schema.nil?
-        schema
-          .fields
-          .map { |field| { name: field.name, data_type: field.type} }
-      end
-
-      def timeout_ms
-        30 * 1000 # 30 seconds
+        schema.fields.map { |field| {name: field.name, data_type: field.type} }
       end
     end
   end

+ 1 - 1
lib/blazer/data_source.rb

@@ -12,7 +12,7 @@ module Blazer
       @id = id
       @settings = settings
 
-      unless settings["url"] || Rails.env.development?
+      unless settings["url"] || Rails.env.development? || settings["adapter"] == "bigquery"
         raise Blazer::Error, "Empty url for data source: #{id}"
       end