parqueteur 1.0.3 → 1.3.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.gitignore +2 -0
- data/Gemfile.lock +1 -1
- data/README.md +43 -8
- data/examples/convert-and-compression.rb +56 -0
- data/examples/convert-methods.rb +54 -0
- data/examples/convert-to-io.rb +52 -0
- data/examples/convert-with-chunks.rb +54 -0
- data/examples/convert-without-compression.rb +52 -0
- data/examples/hello-world.rb +56 -0
- data/lib/parqueteur/column.rb +3 -20
- data/lib/parqueteur/column_collection.rb +8 -0
- data/lib/parqueteur/converter.rb +110 -67
- data/lib/parqueteur/input.rb +12 -27
- data/lib/parqueteur/struct.rb +25 -0
- data/lib/parqueteur/type.rb +21 -0
- data/lib/parqueteur/type_resolver.rb +44 -48
- data/lib/parqueteur/types/array_type.rb +21 -0
- data/lib/parqueteur/types/boolean_type.rb +15 -0
- data/lib/parqueteur/types/date32_type.rb +15 -0
- data/lib/parqueteur/types/date64_type.rb +15 -0
- data/lib/parqueteur/types/decimal128_type.rb +18 -0
- data/lib/parqueteur/types/decimal256_type.rb +18 -0
- data/lib/parqueteur/types/int32_type.rb +23 -0
- data/lib/parqueteur/types/int64_type.rb +23 -0
- data/lib/parqueteur/types/map_type.rb +36 -0
- data/lib/parqueteur/types/string_type.rb +20 -0
- data/lib/parqueteur/types/struct_type.rb +35 -0
- data/lib/parqueteur/types/time32_type.rb +19 -0
- data/lib/parqueteur/types/time64_type.rb +19 -0
- data/lib/parqueteur/types/timestamp_type.rb +24 -0
- data/lib/parqueteur/version.rb +1 -1
- data/lib/parqueteur.rb +24 -7
- data/parqueteur.gemspec +2 -2
- data/scripts/apache-arrow-ubuntu-install.sh +18 -0
- metadata +27 -8
- data/example.rb +0 -20
- data/lib/parqueteur/chunked_converter.rb +0 -28
- data/lib/parqueteur/value_array_builder.rb +0 -59
- data/test.json +0 -1
@@ -0,0 +1,19 @@
|
|
1
|
+
# frozen_string_literal: true
|
2
|
+
|
3
|
+
module Parqueteur
|
4
|
+
module Types
|
5
|
+
class Time32Type < Parqueteur::Type
|
6
|
+
def build_value_array(values)
|
7
|
+
Arrow::Time32Array.new(
|
8
|
+
@options.fetch(:precision, :second), values
|
9
|
+
)
|
10
|
+
end
|
11
|
+
|
12
|
+
def arrow_type_builder
|
13
|
+
Arrow::Time32DataType.new(
|
14
|
+
options.fetch(:unit, :second)
|
15
|
+
)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
end
|
19
|
+
end
|
@@ -0,0 +1,19 @@
|
|
1
|
+
# frozen_string_literal: true
|
2
|
+
|
3
|
+
module Parqueteur
|
4
|
+
module Types
|
5
|
+
class Time64Type < Parqueteur::Type
|
6
|
+
def build_value_array(values)
|
7
|
+
Arrow::Time64Array.new(
|
8
|
+
@options.fetch(:precision, :second), values
|
9
|
+
)
|
10
|
+
end
|
11
|
+
|
12
|
+
def arrow_type_builder
|
13
|
+
Arrow::Time64DataType.new(
|
14
|
+
options.fetch(:unit, :second)
|
15
|
+
)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
end
|
19
|
+
end
|
@@ -0,0 +1,24 @@
|
|
1
|
+
# frozen_string_literal: true
|
2
|
+
|
3
|
+
# when :timestamp
|
4
|
+
# Arrow::TimestampDataType.new(
|
5
|
+
# options.fetch(:unit, :second)
|
6
|
+
# )
|
7
|
+
|
8
|
+
module Parqueteur
|
9
|
+
module Types
|
10
|
+
class TimestampType < Parqueteur::Type
|
11
|
+
def build_value_array(values)
|
12
|
+
Arrow::TimestampArray.new(
|
13
|
+
options.fetch(:unit, :second), values
|
14
|
+
)
|
15
|
+
end
|
16
|
+
|
17
|
+
def arrow_type_builder
|
18
|
+
Arrow::TimestampDataType.new(
|
19
|
+
options.fetch(:unit, :second)
|
20
|
+
)
|
21
|
+
end
|
22
|
+
end
|
23
|
+
end
|
24
|
+
end
|
data/lib/parqueteur/version.rb
CHANGED
data/lib/parqueteur.rb
CHANGED
@@ -1,17 +1,34 @@
|
|
1
1
|
# frozen_string_literal: true
|
2
2
|
|
3
|
-
|
4
|
-
require '
|
3
|
+
require 'json'
|
4
|
+
require 'singleton'
|
5
|
+
require 'tempfile'
|
6
|
+
require 'parquet'
|
7
|
+
|
8
|
+
require_relative 'parqueteur/version'
|
5
9
|
require 'parqueteur/column'
|
6
10
|
require 'parqueteur/column_collection'
|
7
11
|
require 'parqueteur/converter'
|
8
|
-
require 'parqueteur/chunked_converter'
|
9
12
|
require 'parqueteur/input'
|
10
|
-
require 'parqueteur/
|
11
|
-
require '
|
12
|
-
require '
|
13
|
+
require 'parqueteur/struct'
|
14
|
+
require 'parqueteur/type'
|
15
|
+
require 'parqueteur/type_resolver'
|
16
|
+
require 'parqueteur/types/array_type'
|
17
|
+
require 'parqueteur/types/boolean_type'
|
18
|
+
require 'parqueteur/types/date32_type'
|
19
|
+
require 'parqueteur/types/date64_type'
|
20
|
+
require 'parqueteur/types/decimal128_type'
|
21
|
+
require 'parqueteur/types/decimal256_type'
|
22
|
+
require 'parqueteur/types/int32_type'
|
23
|
+
require 'parqueteur/types/int64_type'
|
24
|
+
require 'parqueteur/types/map_type'
|
25
|
+
require 'parqueteur/types/string_type'
|
26
|
+
require 'parqueteur/types/struct_type'
|
27
|
+
require 'parqueteur/types/time32_type'
|
28
|
+
require 'parqueteur/types/time64_type'
|
29
|
+
require 'parqueteur/types/timestamp_type'
|
13
30
|
|
14
31
|
module Parqueteur
|
15
32
|
class Error < StandardError; end
|
16
|
-
|
33
|
+
class TypeNotFound < Error; end
|
17
34
|
end
|
data/parqueteur.gemspec
CHANGED
@@ -8,8 +8,8 @@ Gem::Specification.new do |spec|
|
|
8
8
|
spec.authors = ["Julien D."]
|
9
9
|
spec.email = ["julien@pocketsizesun.com"]
|
10
10
|
spec.license = 'Apache-2.0'
|
11
|
-
spec.summary = 'Parqueteur - A Ruby gem that convert
|
12
|
-
spec.description = 'Convert
|
11
|
+
spec.summary = 'Parqueteur - A Ruby gem that convert data to Parquet'
|
12
|
+
spec.description = 'Convert data to Parquet'
|
13
13
|
spec.homepage = 'https://github.com/pocketsizesun/parqueteur-ruby'
|
14
14
|
spec.required_ruby_version = Gem::Requirement.new(">= 2.3.0")
|
15
15
|
|
@@ -0,0 +1,18 @@
|
|
1
|
+
#!/bin/sh
|
2
|
+
|
3
|
+
if [ $(dpkg-query -W -f='${Status}' apache-arrow-apt-source 2>/dev/null | grep -c "ok installed") -eq 1 ]
|
4
|
+
then
|
5
|
+
exit 0
|
6
|
+
fi
|
7
|
+
|
8
|
+
LSB_RELEASE_CODENAME_SHORT=$(lsb_release --codename --short)
|
9
|
+
|
10
|
+
apt-get update
|
11
|
+
apt-get install -y -V ca-certificates lsb-release wget
|
12
|
+
wget https://apache.jfrog.io/artifactory/arrow/$(lsb_release --id --short | tr 'A-Z' 'a-z')/apache-arrow-apt-source-latest-$(lsb_release --codename --short).deb
|
13
|
+
apt-get install -y -V ./apache-arrow-apt-source-latest-${LSB_RELEASE_CODENAME_SHORT}.deb
|
14
|
+
rm ./apache-arrow-apt-source-latest-${LSB_RELEASE_CODENAME_SHORT}.deb
|
15
|
+
apt-get update
|
16
|
+
apt-get install -y libgirepository1.0-dev libarrow-dev libarrow-glib-dev libparquet-dev libparquet-glib-dev
|
17
|
+
|
18
|
+
exit 0
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: parqueteur
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.0
|
4
|
+
version: 1.3.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Julien D.
|
8
8
|
autorequire:
|
9
9
|
bindir: exe
|
10
10
|
cert_chain: []
|
11
|
-
date: 2021-10-
|
11
|
+
date: 2021-10-03 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: red-parquet
|
@@ -24,7 +24,7 @@ dependencies:
|
|
24
24
|
- - "~>"
|
25
25
|
- !ruby/object:Gem::Version
|
26
26
|
version: '5.0'
|
27
|
-
description: Convert
|
27
|
+
description: Convert data to Parquet
|
28
28
|
email:
|
29
29
|
- julien@pocketsizesun.com
|
30
30
|
executables: []
|
@@ -38,18 +38,37 @@ files:
|
|
38
38
|
- Rakefile
|
39
39
|
- bin/console
|
40
40
|
- bin/setup
|
41
|
-
-
|
41
|
+
- examples/convert-and-compression.rb
|
42
|
+
- examples/convert-methods.rb
|
43
|
+
- examples/convert-to-io.rb
|
44
|
+
- examples/convert-with-chunks.rb
|
45
|
+
- examples/convert-without-compression.rb
|
46
|
+
- examples/hello-world.rb
|
42
47
|
- lib/parqueteur.rb
|
43
|
-
- lib/parqueteur/chunked_converter.rb
|
44
48
|
- lib/parqueteur/column.rb
|
45
49
|
- lib/parqueteur/column_collection.rb
|
46
50
|
- lib/parqueteur/converter.rb
|
47
51
|
- lib/parqueteur/input.rb
|
52
|
+
- lib/parqueteur/struct.rb
|
53
|
+
- lib/parqueteur/type.rb
|
48
54
|
- lib/parqueteur/type_resolver.rb
|
49
|
-
- lib/parqueteur/
|
55
|
+
- lib/parqueteur/types/array_type.rb
|
56
|
+
- lib/parqueteur/types/boolean_type.rb
|
57
|
+
- lib/parqueteur/types/date32_type.rb
|
58
|
+
- lib/parqueteur/types/date64_type.rb
|
59
|
+
- lib/parqueteur/types/decimal128_type.rb
|
60
|
+
- lib/parqueteur/types/decimal256_type.rb
|
61
|
+
- lib/parqueteur/types/int32_type.rb
|
62
|
+
- lib/parqueteur/types/int64_type.rb
|
63
|
+
- lib/parqueteur/types/map_type.rb
|
64
|
+
- lib/parqueteur/types/string_type.rb
|
65
|
+
- lib/parqueteur/types/struct_type.rb
|
66
|
+
- lib/parqueteur/types/time32_type.rb
|
67
|
+
- lib/parqueteur/types/time64_type.rb
|
68
|
+
- lib/parqueteur/types/timestamp_type.rb
|
50
69
|
- lib/parqueteur/version.rb
|
51
70
|
- parqueteur.gemspec
|
52
|
-
-
|
71
|
+
- scripts/apache-arrow-ubuntu-install.sh
|
53
72
|
homepage: https://github.com/pocketsizesun/parqueteur-ruby
|
54
73
|
licenses:
|
55
74
|
- Apache-2.0
|
@@ -73,5 +92,5 @@ requirements: []
|
|
73
92
|
rubygems_version: 3.2.3
|
74
93
|
signing_key:
|
75
94
|
specification_version: 4
|
76
|
-
summary: Parqueteur - A Ruby gem that convert
|
95
|
+
summary: Parqueteur - A Ruby gem that convert data to Parquet
|
77
96
|
test_files: []
|
data/example.rb
DELETED
@@ -1,20 +0,0 @@
|
|
1
|
-
require 'bundler/setup'
|
2
|
-
require 'parqueteur'
|
3
|
-
|
4
|
-
class Foo < Parqueteur::Converter
|
5
|
-
column :id, :long
|
6
|
-
column :reference, :string
|
7
|
-
column :hash, :map, key: :string, value: :string
|
8
|
-
column :valid, :boolean
|
9
|
-
column :total, :integer
|
10
|
-
end
|
11
|
-
|
12
|
-
LETTERS = ('a'..'z').to_a
|
13
|
-
|
14
|
-
data = 1000.times.collect do |i|
|
15
|
-
{ 'id' => i + 1, 'reference' => "coucou:#{i}", 'hash' => { 'a' => LETTERS.sample }, 'valid' => rand < 0.5, 'total' => rand(100..500) }
|
16
|
-
end
|
17
|
-
|
18
|
-
chunked_converter = Parqueteur::ChunkedConverter.new(data, Foo)
|
19
|
-
pp chunked_converter.write_files('test')
|
20
|
-
# puts Foo.convert(data, output: 'test.parquet')
|
@@ -1,28 +0,0 @@
|
|
1
|
-
# frozen_string_literal: true
|
2
|
-
|
3
|
-
module Parqueteur
|
4
|
-
class ChunkedConverter
|
5
|
-
attr_reader :schema
|
6
|
-
|
7
|
-
def initialize(input, converter, chunk_size = 200)
|
8
|
-
@input = Parqueteur::Input.from(input)
|
9
|
-
@converter = converter
|
10
|
-
@chunk_size = chunk_size
|
11
|
-
end
|
12
|
-
|
13
|
-
def chunks
|
14
|
-
Enumerator.new do |arr|
|
15
|
-
@input.each_slice(@chunk_size) do |chunk|
|
16
|
-
local_converter = @converter.new(chunk)
|
17
|
-
arr << local_converter.to_io
|
18
|
-
end
|
19
|
-
end
|
20
|
-
end
|
21
|
-
|
22
|
-
def write_files(prefix)
|
23
|
-
chunks.each_with_index do |chunk, idx|
|
24
|
-
File.write("#{prefix}.#{idx}.parquet", chunk.read)
|
25
|
-
end
|
26
|
-
end
|
27
|
-
end
|
28
|
-
end
|
@@ -1,59 +0,0 @@
|
|
1
|
-
# frozen_string_literal: true
|
2
|
-
|
3
|
-
module Parqueteur
|
4
|
-
class ValueArrayBuilder
|
5
|
-
attr_reader :type, :options, :arrow_type
|
6
|
-
|
7
|
-
def self.build(input, type, options)
|
8
|
-
new(type, options).build(input)
|
9
|
-
end
|
10
|
-
|
11
|
-
def initialize(type, options)
|
12
|
-
@type = type
|
13
|
-
@options = options
|
14
|
-
@arrow_type = Parqueteur::TypeResolver.resolve(type, options)
|
15
|
-
end
|
16
|
-
|
17
|
-
def build(input)
|
18
|
-
return if input.nil?
|
19
|
-
|
20
|
-
case type
|
21
|
-
when :array
|
22
|
-
Arrow::ListArrayBuilder.build(arrow_type, input)
|
23
|
-
when :map
|
24
|
-
builder = Arrow::MapArrayBuilder.new(arrow_type)
|
25
|
-
input.each do |entry|
|
26
|
-
builder.append_value
|
27
|
-
next if entry.nil?
|
28
|
-
|
29
|
-
entry.each do |k, v|
|
30
|
-
builder.key_builder.append(k)
|
31
|
-
builder.item_builder.append(v)
|
32
|
-
end
|
33
|
-
end
|
34
|
-
|
35
|
-
builder.finish
|
36
|
-
when :boolean
|
37
|
-
Arrow::BooleanArray.new(input)
|
38
|
-
when :integer
|
39
|
-
if options.fetch(:unsigned, false) == true
|
40
|
-
Arrow::UInt32Array.new(input)
|
41
|
-
else
|
42
|
-
Arrow::Int32Array.new(input)
|
43
|
-
end
|
44
|
-
when :long
|
45
|
-
if options.fetch(:unsigned, false) == true
|
46
|
-
Arrow::UInt64Array.new(input)
|
47
|
-
else
|
48
|
-
Arrow::Int64Array.new(input)
|
49
|
-
end
|
50
|
-
when :string
|
51
|
-
Arrow::StringArray.new(input)
|
52
|
-
when :timestamp
|
53
|
-
Arrow::TimestampArray.new(input)
|
54
|
-
else
|
55
|
-
raise Error, "unknown type: #{type}"
|
56
|
-
end
|
57
|
-
end
|
58
|
-
end
|
59
|
-
end
|
data/test.json
DELETED
@@ -1 +0,0 @@
|
|
1
|
-
[{"id":1,"reference":"coucou","hash":{"a":"b"},"valid":true,"hash2":{},"numbers":[1,2,3],"map_array":[]},{"id":2,"reference":"coucou","hash":{"c":"d"},"valid":false,"hash2":{},"numbers":[4,5,6],"map_array":[]},{"id":3,"reference":"coucou","hash":{"e":"f"},"valid":true,"hash2":{"x":[1,2,3]},"numbers":[7,8,9],"map_array":[{"x":"y"}]}]
|