This is an automated email from the ASF dual-hosted git repository.
kou pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/arrow.git
The following commit(s) were added to refs/heads/main by this push:
new a64e58313a GH-48382: [Ruby] Add support for reading struct array
(#48383)
a64e58313a is described below
commit a64e58313ac2354b8a888036523cf90bdb11e634
Author: Sutou Kouhei <[email protected]>
AuthorDate: Mon Dec 8 06:16:40 2025 +0900
GH-48382: [Ruby] Add support for reading struct array (#48383)
### Rationale for this change
It's a nested array.
### What changes are included in this PR?
* Add `ArrowFormat::StructType`
* Add `ArrowFormat::StructArray`
### Are these changes tested?
Yes.
### Are there any user-facing changes?
Yes.
* GitHub Issue: #48382
Authored-by: Sutou Kouhei <[email protected]>
Signed-off-by: Sutou Kouhei <[email protected]>
---
ruby/red-arrow-format/lib/arrow-format/array.rb | 17 +++++++++++++++++
.../lib/arrow-format/file-reader.rb | 9 +++++++++
ruby/red-arrow-format/lib/arrow-format/type.rb | 12 ++++++++++++
ruby/red-arrow-format/test/test-file-reader.rb | 21 +++++++++++++++++++++
4 files changed, 59 insertions(+)
diff --git a/ruby/red-arrow-format/lib/arrow-format/array.rb
b/ruby/red-arrow-format/lib/arrow-format/array.rb
index 41a553bcd9..c5b636dc75 100644
--- a/ruby/red-arrow-format/lib/arrow-format/array.rb
+++ b/ruby/red-arrow-format/lib/arrow-format/array.rb
@@ -180,4 +180,21 @@ module ArrowFormat
apply_validity(values)
end
end
+
+ class StructArray < Array
+ def initialize(type, size, validity_buffer, children)
+ super(type, size, validity_buffer)
+ @children = children
+ end
+
+ def to_a
+ if @children.empty?
+ values = [[]] * @size
+ else
+ children_values = @children.collect(&:to_a)
+ values = children_values[0].zip(*children_values[1..-1])
+ end
+ apply_validity(values)
+ end
+ end
end
diff --git a/ruby/red-arrow-format/lib/arrow-format/file-reader.rb
b/ruby/red-arrow-format/lib/arrow-format/file-reader.rb
index 72c5c9bca8..79ea86cd33 100644
--- a/ruby/red-arrow-format/lib/arrow-format/file-reader.rb
+++ b/ruby/red-arrow-format/lib/arrow-format/file-reader.rb
@@ -33,6 +33,7 @@ require_relative "org/apache/arrow/flatbuf/message"
require_relative "org/apache/arrow/flatbuf/null"
require_relative "org/apache/arrow/flatbuf/precision"
require_relative "org/apache/arrow/flatbuf/schema"
+require_relative "org/apache/arrow/flatbuf/struct_"
require_relative "org/apache/arrow/flatbuf/utf8"
module ArrowFormat
@@ -159,6 +160,9 @@ module ArrowFormat
end
when Org::Apache::Arrow::Flatbuf::List
type = ListType.new(read_field(fb_field.children[0]))
+ when Org::Apache::Arrow::Flatbuf::Struct
+ children = fb_field.children.collect {|child| read_field(child)}
+ type = StructType.new(children)
when Org::Apache::Arrow::Flatbuf::Binary
type = BinaryType.singleton
when Org::Apache::Arrow::Flatbuf::LargeBinary
@@ -200,6 +204,11 @@ module ArrowFormat
offsets = body.slice(offsets_buffer.offset, offsets_buffer.length)
child = read_column(field.type.child, nodes, buffers, body)
field.type.build_array(length, validity, offsets, child)
+ when StructType
+ children = field.type.children.collect do |child|
+ read_column(child, nodes, buffers, body)
+ end
+ field.type.build_array(length, validity, children)
when VariableSizeBinaryType
offsets_buffer = buffers.shift
values_buffer = buffers.shift
diff --git a/ruby/red-arrow-format/lib/arrow-format/type.rb
b/ruby/red-arrow-format/lib/arrow-format/type.rb
index c7cebcd745..179a89ff4c 100644
--- a/ruby/red-arrow-format/lib/arrow-format/type.rb
+++ b/ruby/red-arrow-format/lib/arrow-format/type.rb
@@ -206,4 +206,16 @@ module ArrowFormat
ListArray.new(self, size, validity_buffer, offsets_buffer, child)
end
end
+
+ class StructType < Type
+ attr_reader :children
+ def initialize(children)
+ super("Struct")
+ @children = children
+ end
+
+ def build_array(size, validity_buffer, children)
+ StructArray.new(self, size, validity_buffer, children)
+ end
+ end
end
diff --git a/ruby/red-arrow-format/test/test-file-reader.rb
b/ruby/red-arrow-format/test/test-file-reader.rb
index b39d7b1fff..48c20d1f3f 100644
--- a/ruby/red-arrow-format/test/test-file-reader.rb
+++ b/ruby/red-arrow-format/test/test-file-reader.rb
@@ -139,4 +139,25 @@ class TestFileReader < Test::Unit::TestCase
read)
end
end
+
+ sub_test_case("Struct") do
+ def build_array
+ data_type = Arrow::StructDataType.new(count: :int8,
+ visible: :boolean)
+ Arrow::StructArray.new(data_type, [[-128, nil], nil, [nil, true]])
+ end
+
+ def test_read
+ assert_equal([
+ {
+ "value" => [
+ [-128, nil],
+ nil,
+ [nil, true],
+ ],
+ },
+ ],
+ read)
+ end
+ end
end
