trino-client 2.0.1 → 2.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 50686d406d4aedd29016b8b5cfee3024869e797791cc53e099dbf813f6a7d135
-  data.tar.gz: d4d2e2762b8eba6ecaf6c2dd07c19f1db37eee5c8d1ebf4c995d88ec768fe62c
+  metadata.gz: a9ecbdf7d017c29ac4e645d54564a701538724367f8ae62551c5e5dce705f91f
+  data.tar.gz: 78f07db77b0c6231c563c4a39a74e07d186c93b0b0ad03327439c9781e5d577d
 SHA512:
-  metadata.gz: 203db357ca5034c2000d836667ac1bb92d976300d7299663105af90f284657854eb061625b48756a4cba0d21a4d8342d5006c95a7b6eab7e37ca1abbe832c21d
-  data.tar.gz: 5566882e00f6edac34ac3d863d78aea8f3358e829b86d7ffff9f51582b42315b7a2b87e0d71a7b22cfd98eea5ac0fb20ab28f1e219991f1ac62e28256cf914ec
+  metadata.gz: 2c6c3b16e146d2141df38a4bd10e873cedcb18d3a0f6cb8908122382ac2cc44b471a27a4c243981ff9783da450f497ef4219d6bf4a4cc1c4e94a7d2a6ab084f2
+  data.tar.gz: 71bf6ac5e65dada7e5f02f03f7abf13132a81849520720a41641acc2041ef2bfbd4f8000de476d968dfec3c5666cf94c6acff52ec2771aaabb924bfddd228d44

data/ChangeLog.md

@@ -1,5 +1,12 @@
 trino-client-ruby
 ====
+## 2.2.0
+- Add transform_row and scalar_parser documentation and make them easier to use ([#118](https://github.com/treasure-data/trino-client-ruby/issues/118)) [[41ffca7](https://github.com/treasure-data/trino-client-ruby/commit/41ffca7)]
+
+## 2.1.0
+- Add utility for transforming Trino ROW type columns into Ruby hashes ([#117](https://github.com/treasure-data/trino-client-ruby/issues/117)) [[e14251c](https://github.com/treasure-data/trino-client-ruby/commit/e14251c)]
+- Bump actions/checkout from 3 to 4 ([#116](https://github.com/treasure-data/trino-client-ruby/issues/116)) [[ebd9d9e](https://github.com/treasure-data/trino-client-ruby/commit/ebd9d9e)]
+
 ## 2.0.1
 - Reduce published gem size ([#113](https://github.com/treasure-data/trino-client-ruby/issues/113)) [[4e03819](https://github.com/treasure-data/trino-client-ruby/commit/4e03819)]
 - Update standard requirement from ~> 1.24.3 to ~> 1.30.1 ([#112](https://github.com/treasure-data/trino-client-ruby/issues/112)) [[b9f9c0c](https://github.com/treasure-data/trino-client-ruby/commit/b9f9c0c)]

data/README.md

@@ -64,6 +64,19 @@ query = client.query("select * from sys.node")
 query_id = query.query_info.query_id
 query.each_row {|row| ... }  # when a thread is processing the query,
 client.kill(query_id)  # another thread / process can kill the query.
+
+# Use Query#transform_row to parse Trino ROW types into Ruby Hashes.
+# You can also set a scalar_parser to parse scalars how you'd like them.
+scalar_parser = -> (data, type) { (type === 'json') ? JSON.parse(data) : data }
+client.query("select * from sys.node") do |q|
+  q.scalar_parser = scalar_parser
+
+  # get query results. it feeds more rows until
+  # query execution finishes:
+  q.each_row {|row|
+    p q.transform_row(row)
+  }
+end
 ```
 
 ## Build models

data/lib/trino/client/column_value_parser.rb

@@ -0,0 +1,115 @@
+module Trino::Client
+  class ColumnValueParser
+    INSIDE_MATCHING_PARENS_REGEX = /\((?>[^)(]+|\g<0>)*\)/
+
+    attr_reader :name, :type, :scalar_parser
+
+    def initialize(column, scalar_parser = nil)
+      @name = column.name
+      @type = prepare_type_for_parsing(column.type)
+      @scalar_parser = scalar_parser
+    end
+
+    # Public: Parse the value of a row's field by using its column's Trino type.
+    # Trino types can be scalars like VARCHAR and TIMESTAMP or complex types
+    # like ARRAY and ROW. ROW types are treated as objects.
+    # An ARRAY column's type is an array of types as you'd expect. A ROW
+    # column's type is a comma-separated list of space-separated (name, type) tuples.
+    #
+    # data - The value of a row's field. Can be a string, number, an array of those,
+    #        or an arrays of arrays, etc.
+    # dtype - The Trino type string of the column. See above explanation.
+    #
+    # Returns:
+    # - The given value for strings and numbers
+    # - A Time for timestamps
+    # - A Hash of { field1 => value1, field2 => value2, ...etc } for row types
+    # - An array of the above for array types
+    def value(data, dtype = type)
+      # Convert Trino ARRAY elements into Ruby Arrays
+      if starts_with?(dtype, 'array(')
+        return parse_array_element(data, dtype)
+
+      # Convert Trino ROW elements into Ruby Hashes
+      elsif starts_with?(dtype, 'row(')
+        return parse_row_element(data, dtype)
+
+      # If defined, use scalar_parser to convert scalar types
+      elsif !scalar_parser.nil?
+        return scalar_parser.call(data, dtype)
+      end
+
+      # Otherwise, values are returned unaltered
+      data
+    end
+
+    private
+
+    # Private: Remove quotation marks and handle recent versions of
+    # Trino having a 'with time zone' suffix on some fields that breaks
+    # out assumption that types don't have spaces in them.
+    #
+    # Returns a string.
+    def prepare_type_for_parsing(type)
+      type.gsub('"', '').gsub(' with time zone', '_with_time_zone')
+    end
+
+    def parse_array_element(data, dtype)
+      # If the element is empty, return an empty array
+      return [] if blank?(data)
+
+      # Inner data type will be the current dtype with `array(` and `)` chopped off
+      inner_dtype = dtype.match(INSIDE_MATCHING_PARENS_REGEX)[0][1..-2]
+
+      data.map { |inner_data| value(inner_data, inner_dtype) }
+    end
+
+    def parse_row_element(data, dtype)
+      # If the element is empty, return an empty object
+      return {} if blank?(data)
+
+      parsed_row_element = {}
+
+      inner_dtype = dtype.match(INSIDE_MATCHING_PARENS_REGEX)[0][1..-2]
+      elems = inner_dtype.split(' ')
+      num_elems_to_skip = 0
+      field_position = 0
+
+      # Iterate over each datatype of the row and mutate parsed_row_element
+      # to have a key of the field name and value for that field's value.
+      elems.each_with_index do |field, i|
+        # We detected an array or row and are skipping all of the elements within it
+        # since its conversion was handled by calling `value` recursively.
+        if num_elems_to_skip.positive?
+          num_elems_to_skip -= 1
+          next
+        end
+
+        # Field names never have these characters and are never the last element.
+        next if field.include?(',') || field.include?('(') || field.include?(')') || i == elems.length - 1
+
+        type = elems[(i + 1)..].join(' ')
+
+        # If this row has a nested array or row, the type of this field is that array or row's type.
+        if starts_with?(type, 'array(') || starts_with?(type, 'row(')
+          datatype = type.sub(/\(.*/, '')
+          type = "#{datatype}#{type.match(INSIDE_MATCHING_PARENS_REGEX)[0]}"
+          num_elems_to_skip = type.split(' ').length # see above comment about num_elems_to_skip
+        end
+
+        parsed_row_element[field] = value(data[field_position], type)
+        field_position += 1
+      end
+
+      parsed_row_element
+    end
+
+    def blank?(obj)
+      obj.respond_to?(:empty?) ? !!obj.empty? : !obj
+    end
+
+    def starts_with?(str, prefix)
+      prefix.respond_to?(:to_str) && str[0, prefix.length] == prefix
+    end
+  end
+end

data/lib/trino/client/query.rb

@@ -18,6 +18,7 @@ module Trino::Client
   require 'faraday'
   require 'faraday/gzip'
   require 'faraday/follow_redirects'
+  require 'trino/client/column_value_parser'
   require 'trino/client/models'
   require 'trino/client/errors'
   require 'trino/client/faraday_client'
@@ -44,6 +45,21 @@ module Trino::Client
       Trino::Client.faraday_client(options)
     end
 
+    def self.transform_row(column_value_parsers, row)
+      row_object = {}
+
+      row.each_with_index do |element, i|
+        column = column_value_parsers[i]
+        value = column.value(element)
+
+        row_object[column.name] = value
+      end
+
+      row_object
+    end
+
+    attr_accessor :scalar_parser
+
     def initialize(api)
       @api = api
     end
@@ -86,6 +102,21 @@ module Trino::Client
       return @api.current_results.columns
     end
 
+    def column_value_parsers
+      @column_value_parsers ||= {}
+      @column_value_parsers[scalar_parser] ||= columns.map {|column|
+        ColumnValueParser.new(column, scalar_parser)
+      }
+    end
+
+    def transform_rows
+      rows.map { |row| transform_row(row) }
+    end
+
+    def transform_row(row)
+      self.class.transform_row(column_value_parsers, row)
+    end
+
     def rows
       rows = []
       each_row_chunk {|chunk|

data/lib/trino/client/version.rb

@@ -15,6 +15,6 @@
 #
 module Trino
   module Client
-    VERSION = "2.0.1"
+    VERSION = "2.2.0"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: trino-client
 version: !ruby/object:Gem::Version
-  version: 2.0.1
+  version: 2.2.0
 platform: ruby
 authors:
 - Sadayuki Furuhashi
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2023-08-17 00:00:00.000000000 Z
+date: 2023-09-28 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: faraday
@@ -190,6 +190,7 @@ files:
 - lib/trino-client.rb
 - lib/trino/client.rb
 - lib/trino/client/client.rb
+- lib/trino/client/column_value_parser.rb
 - lib/trino/client/errors.rb
 - lib/trino/client/faraday_client.rb
 - lib/trino/client/model_versions/0.149.rb