egis 1.2.1 → 1.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e1ca49c2a770af01b7ff35956785bd77ca5df10446e229620f7cb21892b65e1f
-  data.tar.gz: 450b5de96addf904a43f7b14bc5b55c3f876c982977e6b8ae51a6e58a0466a04
+  metadata.gz: 11391a5ed53240c217c0bdd8db15ddfee6de259341a11f7109475149d7486e52
+  data.tar.gz: 117fb43d54ec8363a44c70924cb4e94c194b234a710ec806fd1d9ae0e534d550
 SHA512:
-  metadata.gz: f542064733b5d9805a27dd40b4628527e3faf932788f96b7911db0726f729f8f89835c06ba767d5532cab603633838146113ae46dfd8b1089d70ba3288806312
-  data.tar.gz: d833d875b4cee9b4d435aa0394908b1466ff1588bb9620fe237c72b95e16965ebf34299a8e5a63d1ce70bf1642ebfd053f6089688ffd83320dadcbb10cc2ca8d
+  metadata.gz: db5df8ea30e0a55dcd4988859ea13ed5e674491bf545a468630e9fe0821c8c36cd360b2839d13e609e2007bc32b5aa89e3a58870bdcdbbaca6b5cf448ac6d59f
+  data.tar.gz: 9aa20b80ff73d128b1f057727fe9ce28367b77500e2c1724d1437b8228d900906f3ec3138d98e8d5c00d11a0e4d1f1674319ddf9fac0c3079ebff7b9b8ca9979

data/lib/egis/database.rb

@@ -28,7 +28,8 @@ module Egis
     # @param [String] table_name
     # @param [Egis::TableSchema] table_schema
     # @param [String] table_location S3 URL with table location (e.g. `s3://s3_bucket/table/location/`)
-    # @param [:tsv, :csv, :orc] format Table format (defaults to :tsv)
+    # @param [:tsv, :csv, :orc, {serde: 'SerdeClass', serde_properties: {property: value}}] format Table format
+    # (defaults to :tsv)
     # @return [Egis::Table]
 
     def table(table_name, table_schema, table_location, **options)

data/lib/egis/table_ddl_generator.rb

@@ -9,7 +9,7 @@ module Egis
           #{column_definition_sql(table.schema.columns)}
         )
         #{partition_statement(table.schema)}
-        #{format_statement(table.format)}
+        #{row_format_statement(table.format)}
         LOCATION '#{table.location}';
       SQL
     end
@@ -34,7 +34,30 @@ module Egis
       columns.map { |column| "`#{column.name}` #{column.type}" }.join(",\n")
     end
 
-    def format_statement(format)
+    def serde?(format)
+      format.is_a?(Hash) && format.key?(:serde)
+    end
+
+    def row_format_statement(format)
+      return serde_row_format_statement(format) if serde?(format)
+
+      delimited_row_format_statement(format)
+    end
+
+    def serde_row_format_statement(format)
+      row_format = "ROW FORMAT SERDE '#{format[:serde]}'"
+      return row_format unless format.key?(:serde_properties)
+
+      serde_properties = format[:serde_properties].map { |property, value| "'#{property}' = '#{value}'" }
+      <<-SQL
+        #{row_format}
+        WITH SERDEPROPERTIES (
+          #{serde_properties.join(",\n")}
+        )
+      SQL
+    end
+
+    def delimited_row_format_statement(format)
       case format
       when :csv
         "ROW FORMAT DELIMITED FIELDS TERMINATED BY ','"

data/lib/egis/testing.rb

@@ -43,6 +43,6 @@ module Egis # rubocop:disable Style/Documentation
     yield
   ensure
     @mode = previous_mode
-    test_mode.cleanup if test_mode
+    test_mode&.cleanup
   end
 end

data/lib/egis/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Egis
-  VERSION = '1.2.1'
+  VERSION = '1.3.0'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: egis
 version: !ruby/object:Gem::Version
-  version: 1.2.1
+  version: 1.3.0
 platform: ruby
 authors:
 - Agnieszka Czereba
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2020-05-29 00:00:00.000000000 Z
+date: 2020-07-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: aws-sdk-athena