avro_turf 0.2.0 → 0.3.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/Gemfile +3 -0
- data/lib/avro_turf.rb +44 -44
- data/lib/avro_turf/schema_store.rb +58 -0
- data/lib/avro_turf/version.rb +1 -1
- data/perf/address.avsc +14 -0
- data/perf/encoding_size.rb +26 -0
- data/perf/encoding_speed.rb +30 -0
- data/perf/person.avsc +14 -0
- data/spec/avro_turf_spec.rb +125 -0
- data/spec/schema_store_spec.rb +253 -0
- data/spec/spec_helper.rb +14 -0
- metadata +11 -4
- data/spec/avro_spec.rb +0 -227
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 0845fc0e81baaa72df5abd34b47a0ce6d06a853f
|
4
|
+
data.tar.gz: 434aaa74efb2f96437045a8705c7889348bec9c3
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: ef023ba8d9fb413e69a467c1ecb4a4ef73c6c1f1bb86f1f362b6f351abcd6ed8b2fae965d7b50b4d3ba6e81513c30dcffbe1e5d688044efa872f4b65f85ff39a
|
7
|
+
data.tar.gz: e095b79ce3f37a69efac8537e6d59fcb5c8d66350604eba7f00166911801699eccf945552fcbc5c950223f7c14ddfddd2f58157b70aff2466e6d2ea157485eab
|
data/Gemfile
CHANGED
data/lib/avro_turf.rb
CHANGED
@@ -1,15 +1,17 @@
|
|
1
1
|
require 'avro_turf/version'
|
2
2
|
require 'avro'
|
3
3
|
require 'json'
|
4
|
+
require 'avro_turf/schema_store'
|
4
5
|
|
5
6
|
class AvroTurf
|
6
7
|
class Error < StandardError; end
|
7
8
|
class SchemaError < Error; end
|
9
|
+
class SchemaNotFoundError < Error; end
|
8
10
|
|
9
|
-
def initialize(schemas_path:, namespace: nil)
|
10
|
-
@schemas_path = schemas_path or raise "Please specify a schema path"
|
11
|
-
@schemas = Hash.new
|
11
|
+
def initialize(schemas_path:, namespace: nil, codec: nil)
|
12
12
|
@namespace = namespace
|
13
|
+
@schema_store = SchemaStore.new(path: schemas_path)
|
14
|
+
@codec = codec
|
13
15
|
end
|
14
16
|
|
15
17
|
# Encodes data to Avro using the specified schema.
|
@@ -19,15 +21,31 @@ class AvroTurf
|
|
19
21
|
#
|
20
22
|
# Returns a String containing the encoded data.
|
21
23
|
def encode(data, schema_name:, namespace: @namespace)
|
22
|
-
|
24
|
+
stream = StringIO.new
|
25
|
+
|
26
|
+
encode_to_stream(data, stream: stream, schema_name: schema_name, namespace: namespace)
|
27
|
+
|
28
|
+
stream.string
|
29
|
+
end
|
30
|
+
|
31
|
+
# Encodes data to Avro using the specified schema and writes it to the
|
32
|
+
# specified stream.
|
33
|
+
#
|
34
|
+
# data - The data that should be encoded.
|
35
|
+
# schema_name - The name of a schema in the `schemas_path`.
|
36
|
+
# stream - An IO object that the encoded data should be written to (optional).
|
37
|
+
# codec - The String name of a codec that should be used to compress messages (optional).
|
38
|
+
#
|
39
|
+
# Currently, the only valid codec name is `deflate`.
|
40
|
+
#
|
41
|
+
# Returns nothing.
|
42
|
+
def encode_to_stream(data, schema_name:, stream:, namespace: @namespace)
|
43
|
+
schema = @schema_store.find(schema_name, namespace)
|
23
44
|
writer = Avro::IO::DatumWriter.new(schema)
|
24
45
|
|
25
|
-
|
26
|
-
dw = Avro::DataFile::Writer.new(io, writer, schema)
|
46
|
+
dw = Avro::DataFile::Writer.new(stream, writer, schema, @codec)
|
27
47
|
dw << data
|
28
48
|
dw.close
|
29
|
-
|
30
|
-
io.string
|
31
49
|
end
|
32
50
|
|
33
51
|
# Decodes Avro data.
|
@@ -35,49 +53,31 @@ class AvroTurf
|
|
35
53
|
# encoded_data - A String containing Avro-encoded data.
|
36
54
|
# schema_name - The String name of the schema that should be used to read
|
37
55
|
# the data. If nil, the writer schema will be used.
|
56
|
+
# namespace - The namespace of the Avro schema used to decode the data.
|
38
57
|
#
|
39
58
|
# Returns whatever is encoded in the data.
|
40
59
|
def decode(encoded_data, schema_name: nil, namespace: @namespace)
|
41
|
-
|
42
|
-
|
43
|
-
reader = Avro::IO::DatumReader.new(nil, schema)
|
44
|
-
dr = Avro::DataFile::Reader.new(io, reader)
|
45
|
-
dr.first
|
60
|
+
stream = StringIO.new(encoded_data)
|
61
|
+
decode_stream(stream, schema_name: schema_name, namespace: namespace)
|
46
62
|
end
|
47
63
|
|
48
|
-
|
49
|
-
|
50
|
-
# Resolves and returns a schema.
|
64
|
+
# Decodes Avro data from an IO stream.
|
51
65
|
#
|
52
|
-
#
|
66
|
+
# stream - An IO object containing Avro data.
|
67
|
+
# schema_name - The String name of the schema that should be used to read
|
68
|
+
# the data. If nil, the writer schema will be used.
|
69
|
+
# namespace - The namespace of the Avro schema used to decode the data.
|
53
70
|
#
|
54
|
-
# Returns
|
55
|
-
def
|
56
|
-
|
57
|
-
|
58
|
-
|
59
|
-
|
60
|
-
|
61
|
-
schema_path = File.join(@schemas_path, *namespace, schema_name + ".avsc")
|
62
|
-
schema_json = JSON.parse(File.read(schema_path))
|
63
|
-
schema = Avro::Schema.real_parse(schema_json, @schemas)
|
64
|
-
|
65
|
-
if schema.respond_to?(:fullname) && schema.fullname != fullname
|
66
|
-
raise SchemaError, "expected schema `#{schema_path}' to define type `#{fullname}'"
|
67
|
-
end
|
68
|
-
|
69
|
-
schema
|
70
|
-
rescue ::Avro::SchemaParseError => e
|
71
|
-
# This is a hack in order to figure out exactly which type was missing. The
|
72
|
-
# Avro gem ought to provide this data directly.
|
73
|
-
if e.to_s =~ /"([\w\.]+)" is not a schema we know about/
|
74
|
-
resolve_schema($1)
|
71
|
+
# Returns whatever is encoded in the stream.
|
72
|
+
def decode_stream(stream, schema_name: nil, namespace: @namespace)
|
73
|
+
schema = schema_name && @schema_store.find(schema_name, namespace)
|
74
|
+
reader = Avro::IO::DatumReader.new(nil, schema)
|
75
|
+
dr = Avro::DataFile::Reader.new(stream, reader)
|
76
|
+
dr.first
|
77
|
+
end
|
75
78
|
|
76
|
-
|
77
|
-
|
78
|
-
|
79
|
-
else
|
80
|
-
raise
|
81
|
-
end
|
79
|
+
# Loads all schema definition files in the `schemas_dir`.
|
80
|
+
def load_schemas!
|
81
|
+
@schema_store.load_schemas!
|
82
82
|
end
|
83
83
|
end
|
@@ -0,0 +1,58 @@
|
|
1
|
+
class AvroTurf::SchemaStore
|
2
|
+
def initialize(path:)
|
3
|
+
@path = path or raise "Please specify a schema path"
|
4
|
+
@schemas = Hash.new
|
5
|
+
end
|
6
|
+
|
7
|
+
# Resolves and returns a schema.
|
8
|
+
#
|
9
|
+
# schema_name - The String name of the schema to resolve.
|
10
|
+
#
|
11
|
+
# Returns an Avro::Schema.
|
12
|
+
def find(name, namespace = nil)
|
13
|
+
fullname = Avro::Name.make_fullname(name, namespace)
|
14
|
+
|
15
|
+
return @schemas[fullname] if @schemas.key?(fullname)
|
16
|
+
|
17
|
+
*namespace, schema_name = fullname.split(".")
|
18
|
+
schema_path = File.join(@path, *namespace, schema_name + ".avsc")
|
19
|
+
schema_json = JSON.parse(File.read(schema_path))
|
20
|
+
schema = Avro::Schema.real_parse(schema_json, @schemas)
|
21
|
+
|
22
|
+
if schema.respond_to?(:fullname) && schema.fullname != fullname
|
23
|
+
raise AvroTurf::SchemaError, "expected schema `#{schema_path}' to define type `#{fullname}'"
|
24
|
+
end
|
25
|
+
|
26
|
+
schema
|
27
|
+
rescue ::Avro::SchemaParseError => e
|
28
|
+
# This is a hack in order to figure out exactly which type was missing. The
|
29
|
+
# Avro gem ought to provide this data directly.
|
30
|
+
if e.to_s =~ /"([\w\.]+)" is not a schema we know about/
|
31
|
+
find($1)
|
32
|
+
|
33
|
+
# Re-resolve the original schema now that the dependency has been resolved.
|
34
|
+
@schemas.delete(fullname)
|
35
|
+
find(fullname)
|
36
|
+
else
|
37
|
+
raise
|
38
|
+
end
|
39
|
+
rescue Errno::ENOENT
|
40
|
+
raise AvroTurf::SchemaNotFoundError, "could not find Avro schema at `#{schema_path}'"
|
41
|
+
end
|
42
|
+
|
43
|
+
# Loads all schema definition files in the `schemas_dir`.
|
44
|
+
def load_schemas!
|
45
|
+
pattern = [@path, "**", "*.avsc"].join("/")
|
46
|
+
|
47
|
+
Dir.glob(pattern) do |schema_path|
|
48
|
+
# Remove the path prefix.
|
49
|
+
schema_path.sub!(/^\/#{@path}\//, "")
|
50
|
+
|
51
|
+
# Replace `/` with `.` and chop off the file extension.
|
52
|
+
schema_name = File.basename(schema_path.tr("/", "."), ".avsc")
|
53
|
+
|
54
|
+
# Load and cache the schema.
|
55
|
+
find(schema_name)
|
56
|
+
end
|
57
|
+
end
|
58
|
+
end
|
data/lib/avro_turf/version.rb
CHANGED
data/perf/address.avsc
ADDED
@@ -0,0 +1,26 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
#
|
3
|
+
# Measures the encoded size of messages of increasing size.
|
4
|
+
|
5
|
+
$LOAD_PATH.unshift(File.expand_path("../lib", File.dirname(__FILE__)))
|
6
|
+
|
7
|
+
require 'benchmark'
|
8
|
+
require 'avro_turf'
|
9
|
+
|
10
|
+
sizes = [1, 10, 100, 1_000, 10_000]
|
11
|
+
avro = AvroTurf.new(schemas_path: File.dirname(__FILE__))
|
12
|
+
|
13
|
+
sizes.each do |size|
|
14
|
+
data = {
|
15
|
+
"name" => "John" * size,
|
16
|
+
"address" => {
|
17
|
+
"street" => "1st st." * size,
|
18
|
+
"city" => "Citytown" * size
|
19
|
+
}
|
20
|
+
}
|
21
|
+
|
22
|
+
result = avro.encode(data, schema_name: "person")
|
23
|
+
encoded_size = result.bytesize
|
24
|
+
encode_factor = result.bytesize / size.to_f
|
25
|
+
puts "size #{size}: #{encoded_size} bytes (encoding factor #{encode_factor})"
|
26
|
+
end
|
@@ -0,0 +1,30 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
#
|
3
|
+
# Measures the time it takes to encode messages of increasing size.
|
4
|
+
|
5
|
+
$LOAD_PATH.unshift(File.expand_path("../lib", File.dirname(__FILE__)))
|
6
|
+
|
7
|
+
require 'benchmark'
|
8
|
+
require 'avro_turf'
|
9
|
+
|
10
|
+
# Number of iterations per run.
|
11
|
+
N = 10_000
|
12
|
+
|
13
|
+
Benchmark.bm(15) do |x|
|
14
|
+
sizes = [1, 10, 100, 1_000, 10_000]
|
15
|
+
avro = AvroTurf.new(schemas_path: File.dirname(__FILE__))
|
16
|
+
|
17
|
+
sizes.each do |size|
|
18
|
+
data = {
|
19
|
+
"name" => "John" * size,
|
20
|
+
"address" => {
|
21
|
+
"street" => "1st st." * size,
|
22
|
+
"city" => "Citytown" * size
|
23
|
+
}
|
24
|
+
}
|
25
|
+
|
26
|
+
x.report("size #{size}:") do
|
27
|
+
N.times { avro.encode(data, schema_name: "person") }
|
28
|
+
end
|
29
|
+
end
|
30
|
+
end
|
data/perf/person.avsc
ADDED
@@ -0,0 +1,125 @@
|
|
1
|
+
describe AvroTurf do
|
2
|
+
let(:avro) { AvroTurf.new(schemas_path: "spec/schemas/") }
|
3
|
+
|
4
|
+
before do
|
5
|
+
FileUtils.mkdir_p("spec/schemas")
|
6
|
+
end
|
7
|
+
|
8
|
+
describe "#encode" do
|
9
|
+
before do
|
10
|
+
define_schema "person.avsc", <<-AVSC
|
11
|
+
{
|
12
|
+
"name": "person",
|
13
|
+
"type": "record",
|
14
|
+
"fields": [
|
15
|
+
{
|
16
|
+
"type": "string",
|
17
|
+
"name": "full_name"
|
18
|
+
}
|
19
|
+
]
|
20
|
+
}
|
21
|
+
AVSC
|
22
|
+
end
|
23
|
+
|
24
|
+
it "encodes data with Avro" do
|
25
|
+
data = {
|
26
|
+
"full_name" => "John Doe"
|
27
|
+
}
|
28
|
+
|
29
|
+
encoded_data = avro.encode(data, schema_name: "person")
|
30
|
+
|
31
|
+
expect(avro.decode(encoded_data)).to eq(data)
|
32
|
+
end
|
33
|
+
|
34
|
+
it "allows specifying a codec that should be used to compress messages" do
|
35
|
+
compressed_avro = AvroTurf.new(schemas_path: "spec/schemas/", codec: "deflate")
|
36
|
+
|
37
|
+
data = {
|
38
|
+
"full_name" => "John Doe" * 100
|
39
|
+
}
|
40
|
+
|
41
|
+
uncompressed_data = avro.encode(data, schema_name: "person")
|
42
|
+
compressed_data = compressed_avro.encode(data, schema_name: "person")
|
43
|
+
|
44
|
+
expect(compressed_data.bytesize).to be < uncompressed_data.bytesize
|
45
|
+
expect(compressed_avro.decode(compressed_data)).to eq(data)
|
46
|
+
end
|
47
|
+
end
|
48
|
+
|
49
|
+
describe "#decode" do
|
50
|
+
it "decodes Avro data using the inlined writer's schema" do
|
51
|
+
define_schema "message.avsc", <<-AVSC
|
52
|
+
{
|
53
|
+
"name": "message",
|
54
|
+
"type": "string"
|
55
|
+
}
|
56
|
+
AVSC
|
57
|
+
|
58
|
+
encoded_data = avro.encode("hello, world", schema_name: "message")
|
59
|
+
|
60
|
+
expect(avro.decode(encoded_data)).to eq "hello, world"
|
61
|
+
end
|
62
|
+
|
63
|
+
it "decodes Avro data using a specified reader's schema" do
|
64
|
+
FileUtils.mkdir_p("spec/schemas/reader")
|
65
|
+
|
66
|
+
define_schema "point.avsc", <<-AVSC
|
67
|
+
{
|
68
|
+
"name": "point",
|
69
|
+
"type": "record",
|
70
|
+
"fields": [
|
71
|
+
{ "name": "x", "type": "long" },
|
72
|
+
{ "name": "y", "type": "long" }
|
73
|
+
]
|
74
|
+
}
|
75
|
+
AVSC
|
76
|
+
|
77
|
+
define_schema "reader/point.avsc", <<-AVSC
|
78
|
+
{
|
79
|
+
"name": "point",
|
80
|
+
"type": "record",
|
81
|
+
"fields": [
|
82
|
+
{ "name": "x", "type": "long" }
|
83
|
+
]
|
84
|
+
}
|
85
|
+
AVSC
|
86
|
+
|
87
|
+
encoded_data = avro.encode({ "x" => 42, "y" => 13 }, schema_name: "point")
|
88
|
+
reader_avro = AvroTurf.new(schemas_path: "spec/schemas/reader")
|
89
|
+
|
90
|
+
expect(reader_avro.decode(encoded_data, schema_name: "point")).to eq({ "x" => 42 })
|
91
|
+
end
|
92
|
+
end
|
93
|
+
|
94
|
+
describe "#encode_to_stream" do
|
95
|
+
it "writes encoded data to an existing stream" do
|
96
|
+
define_schema "message.avsc", <<-AVSC
|
97
|
+
{
|
98
|
+
"name": "message",
|
99
|
+
"type": "string"
|
100
|
+
}
|
101
|
+
AVSC
|
102
|
+
|
103
|
+
stream = StringIO.new
|
104
|
+
avro.encode_to_stream("hello", stream: stream, schema_name: "message")
|
105
|
+
|
106
|
+
expect(avro.decode(stream.string)).to eq "hello"
|
107
|
+
end
|
108
|
+
end
|
109
|
+
|
110
|
+
describe "#decode_stream" do
|
111
|
+
it "decodes Avro data from a stream" do
|
112
|
+
define_schema "message.avsc", <<-AVSC
|
113
|
+
{
|
114
|
+
"name": "message",
|
115
|
+
"type": "string"
|
116
|
+
}
|
117
|
+
AVSC
|
118
|
+
|
119
|
+
encoded_data = avro.encode("hello", schema_name: "message")
|
120
|
+
stream = StringIO.new(encoded_data)
|
121
|
+
|
122
|
+
expect(avro.decode_stream(stream)).to eq "hello"
|
123
|
+
end
|
124
|
+
end
|
125
|
+
end
|
@@ -0,0 +1,253 @@
|
|
1
|
+
require 'avro_turf/schema_store'
|
2
|
+
|
3
|
+
describe AvroTurf::SchemaStore do
|
4
|
+
let(:store) { AvroTurf::SchemaStore.new(path: "spec/schemas") }
|
5
|
+
|
6
|
+
before do
|
7
|
+
FileUtils.mkdir_p("spec/schemas")
|
8
|
+
end
|
9
|
+
|
10
|
+
describe "#find" do
|
11
|
+
it "finds schemas on the file system" do
|
12
|
+
define_schema "message.avsc", <<-AVSC
|
13
|
+
{
|
14
|
+
"name": "message",
|
15
|
+
"type": "record",
|
16
|
+
"fields": [
|
17
|
+
{
|
18
|
+
"type": "string",
|
19
|
+
"name": "message"
|
20
|
+
}
|
21
|
+
]
|
22
|
+
}
|
23
|
+
AVSC
|
24
|
+
|
25
|
+
schema = store.find("message")
|
26
|
+
expect(schema.fullname).to eq "message"
|
27
|
+
end
|
28
|
+
|
29
|
+
it "resolves missing references" do
|
30
|
+
define_schema "person.avsc", <<-AVSC
|
31
|
+
{
|
32
|
+
"name": "person",
|
33
|
+
"type": "record",
|
34
|
+
"fields": [
|
35
|
+
{
|
36
|
+
"name": "address",
|
37
|
+
"type": "address"
|
38
|
+
}
|
39
|
+
]
|
40
|
+
}
|
41
|
+
AVSC
|
42
|
+
|
43
|
+
define_schema "address.avsc", <<-AVSC
|
44
|
+
{
|
45
|
+
"type": "record",
|
46
|
+
"name": "address",
|
47
|
+
"fields": []
|
48
|
+
}
|
49
|
+
AVSC
|
50
|
+
|
51
|
+
schema = store.find("person")
|
52
|
+
|
53
|
+
expect(schema.fullname).to eq "person"
|
54
|
+
end
|
55
|
+
|
56
|
+
it "finds namespaced schemas" do
|
57
|
+
FileUtils.mkdir_p("spec/schemas/test/people")
|
58
|
+
|
59
|
+
define_schema "test/people/person.avsc", <<-AVSC
|
60
|
+
{
|
61
|
+
"name": "person",
|
62
|
+
"namespace": "test.people",
|
63
|
+
"type": "record",
|
64
|
+
"fields": [
|
65
|
+
{
|
66
|
+
"name": "address",
|
67
|
+
"type": "test.people.address"
|
68
|
+
}
|
69
|
+
]
|
70
|
+
}
|
71
|
+
AVSC
|
72
|
+
|
73
|
+
define_schema "test/people/address.avsc", <<-AVSC
|
74
|
+
{
|
75
|
+
"name": "address",
|
76
|
+
"namespace": "test.people",
|
77
|
+
"type": "record",
|
78
|
+
"fields": []
|
79
|
+
}
|
80
|
+
AVSC
|
81
|
+
|
82
|
+
schema = store.find("person", "test.people")
|
83
|
+
|
84
|
+
expect(schema.fullname).to eq "test.people.person"
|
85
|
+
end
|
86
|
+
|
87
|
+
it "ignores the namespace when the name contains a dot" do
|
88
|
+
FileUtils.mkdir_p("spec/schemas/test/acme")
|
89
|
+
|
90
|
+
define_schema "test/acme/message.avsc", <<-AVSC
|
91
|
+
{
|
92
|
+
"name": "message",
|
93
|
+
"namespace": "test.acme",
|
94
|
+
"type": "record",
|
95
|
+
"fields": []
|
96
|
+
}
|
97
|
+
AVSC
|
98
|
+
|
99
|
+
schema = store.find("test.acme.message", "test.yolo")
|
100
|
+
|
101
|
+
expect(schema.fullname).to eq "test.acme.message"
|
102
|
+
end
|
103
|
+
|
104
|
+
it "raises AvroTurf::SchemaNotFoundError if there's no schema file matching the name" do
|
105
|
+
expect {
|
106
|
+
store.find("not_there")
|
107
|
+
}.to raise_error(AvroTurf::SchemaNotFoundError, "could not find Avro schema at `spec/schemas/not_there.avsc'")
|
108
|
+
end
|
109
|
+
|
110
|
+
it "raises AvroTurf::SchemaNotFoundError if a type reference cannot be resolved" do
|
111
|
+
define_schema "person.avsc", <<-AVSC
|
112
|
+
{
|
113
|
+
"name": "person",
|
114
|
+
"type": "record",
|
115
|
+
"fields": [
|
116
|
+
{
|
117
|
+
"name": "address",
|
118
|
+
"type": "address"
|
119
|
+
}
|
120
|
+
]
|
121
|
+
}
|
122
|
+
AVSC
|
123
|
+
|
124
|
+
expect {
|
125
|
+
store.find("person")
|
126
|
+
}.to raise_exception(AvroTurf::SchemaNotFoundError)
|
127
|
+
end
|
128
|
+
|
129
|
+
it "raises AvroTurf::SchemaError if the schema's namespace doesn't match the file location" do
|
130
|
+
FileUtils.mkdir_p("spec/schemas/test/people")
|
131
|
+
|
132
|
+
define_schema "test/people/person.avsc", <<-AVSC
|
133
|
+
{
|
134
|
+
"name": "person",
|
135
|
+
"namespace": "yoyoyo.nanana",
|
136
|
+
"type": "record",
|
137
|
+
"fields": []
|
138
|
+
}
|
139
|
+
AVSC
|
140
|
+
|
141
|
+
expect {
|
142
|
+
store.find("test.people.person")
|
143
|
+
}.to raise_error(AvroTurf::SchemaError, "expected schema `spec/schemas/test/people/person.avsc' to define type `test.people.person'")
|
144
|
+
end
|
145
|
+
|
146
|
+
it "handles circular dependencies" do
|
147
|
+
define_schema "a.avsc", <<-AVSC
|
148
|
+
{
|
149
|
+
"name": "a",
|
150
|
+
"type": "record",
|
151
|
+
"fields": [
|
152
|
+
{
|
153
|
+
"type": "b",
|
154
|
+
"name": "b"
|
155
|
+
}
|
156
|
+
]
|
157
|
+
}
|
158
|
+
AVSC
|
159
|
+
|
160
|
+
define_schema "b.avsc", <<-AVSC
|
161
|
+
{
|
162
|
+
"name": "b",
|
163
|
+
"type": "record",
|
164
|
+
"fields": [
|
165
|
+
{
|
166
|
+
"type": "a",
|
167
|
+
"name": "a"
|
168
|
+
}
|
169
|
+
]
|
170
|
+
}
|
171
|
+
AVSC
|
172
|
+
|
173
|
+
schema = store.find("a")
|
174
|
+
expect(schema.fullname).to eq "a"
|
175
|
+
end
|
176
|
+
|
177
|
+
it "caches schemas in memory" do
|
178
|
+
define_schema "person.avsc", <<-AVSC
|
179
|
+
{
|
180
|
+
"name": "person",
|
181
|
+
"type": "record",
|
182
|
+
"fields": [
|
183
|
+
{
|
184
|
+
"type": "string",
|
185
|
+
"name": "full_name"
|
186
|
+
}
|
187
|
+
]
|
188
|
+
}
|
189
|
+
AVSC
|
190
|
+
|
191
|
+
# Warm the schema cache.
|
192
|
+
store.find("person")
|
193
|
+
|
194
|
+
# Force a failure if the schema file is read again.
|
195
|
+
FileUtils.rm("spec/schemas/person.avsc")
|
196
|
+
|
197
|
+
schema = store.find("person")
|
198
|
+
expect(schema.fullname).to eq "person"
|
199
|
+
end
|
200
|
+
end
|
201
|
+
|
202
|
+
describe "#load_schemas!" do
|
203
|
+
it "loads schemas defined in the `schemas_path` directory" do
|
204
|
+
define_schema "person.avsc", <<-AVSC
|
205
|
+
{
|
206
|
+
"name": "person",
|
207
|
+
"type": "record",
|
208
|
+
"fields": [
|
209
|
+
{
|
210
|
+
"type": "string",
|
211
|
+
"name": "full_name"
|
212
|
+
}
|
213
|
+
]
|
214
|
+
}
|
215
|
+
AVSC
|
216
|
+
|
217
|
+
# Warm the schema cache.
|
218
|
+
store.load_schemas!
|
219
|
+
|
220
|
+
# Force a failure if the schema file is read again.
|
221
|
+
FileUtils.rm("spec/schemas/person.avsc")
|
222
|
+
|
223
|
+
schema = store.find("person")
|
224
|
+
expect(schema.fullname).to eq "person"
|
225
|
+
end
|
226
|
+
|
227
|
+
it "recursively finds schema definitions in subdirectories" do
|
228
|
+
FileUtils.mkdir_p("spec/schemas/foo/bar")
|
229
|
+
|
230
|
+
define_schema "foo/bar/person.avsc", <<-AVSC
|
231
|
+
{
|
232
|
+
"name": "foo.bar.person",
|
233
|
+
"type": "record",
|
234
|
+
"fields": [
|
235
|
+
{
|
236
|
+
"type": "string",
|
237
|
+
"name": "full_name"
|
238
|
+
}
|
239
|
+
]
|
240
|
+
}
|
241
|
+
AVSC
|
242
|
+
|
243
|
+
# Warm the schema cache.
|
244
|
+
store.load_schemas!
|
245
|
+
|
246
|
+
# Force a failure if the schema file is read again.
|
247
|
+
FileUtils.rm("spec/schemas/foo/bar/person.avsc")
|
248
|
+
|
249
|
+
schema = store.find("foo.bar.person")
|
250
|
+
expect(schema.fullname).to eq "foo.bar.person"
|
251
|
+
end
|
252
|
+
end
|
253
|
+
end
|
data/spec/spec_helper.rb
CHANGED
@@ -1,2 +1,16 @@
|
|
1
1
|
require 'bundler/setup'
|
2
|
+
require 'fakefs/spec_helpers'
|
2
3
|
require 'avro_turf'
|
4
|
+
|
5
|
+
module Helpers
|
6
|
+
def define_schema(path, content)
|
7
|
+
File.open(File.join("spec/schemas", path), "w") do |f|
|
8
|
+
f.write(content)
|
9
|
+
end
|
10
|
+
end
|
11
|
+
end
|
12
|
+
|
13
|
+
RSpec.configure do |config|
|
14
|
+
config.include FakeFS::SpecHelpers
|
15
|
+
config.include Helpers
|
16
|
+
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: avro_turf
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.
|
4
|
+
version: 0.3.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Daniel Schierbeck
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2015-
|
11
|
+
date: 2015-06-08 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: avro
|
@@ -96,8 +96,14 @@ files:
|
|
96
96
|
- avro_turf.gemspec
|
97
97
|
- circle.yml
|
98
98
|
- lib/avro_turf.rb
|
99
|
+
- lib/avro_turf/schema_store.rb
|
99
100
|
- lib/avro_turf/version.rb
|
100
|
-
-
|
101
|
+
- perf/address.avsc
|
102
|
+
- perf/encoding_size.rb
|
103
|
+
- perf/encoding_speed.rb
|
104
|
+
- perf/person.avsc
|
105
|
+
- spec/avro_turf_spec.rb
|
106
|
+
- spec/schema_store_spec.rb
|
101
107
|
- spec/spec_helper.rb
|
102
108
|
homepage: https://github.com/dasch/avro_turf
|
103
109
|
licenses:
|
@@ -125,6 +131,7 @@ specification_version: 4
|
|
125
131
|
summary: A library that makes it easier to use the Avro serialization format from
|
126
132
|
Ruby
|
127
133
|
test_files:
|
128
|
-
- spec/
|
134
|
+
- spec/avro_turf_spec.rb
|
135
|
+
- spec/schema_store_spec.rb
|
129
136
|
- spec/spec_helper.rb
|
130
137
|
has_rdoc:
|
data/spec/avro_spec.rb
DELETED
@@ -1,227 +0,0 @@
|
|
1
|
-
require 'fakefs/spec_helpers'
|
2
|
-
|
3
|
-
describe AvroTurf do
|
4
|
-
include FakeFS::SpecHelpers
|
5
|
-
|
6
|
-
let(:avro) { AvroTurf.new(schemas_path: "spec/schemas/") }
|
7
|
-
|
8
|
-
before do
|
9
|
-
FileUtils.mkdir_p("spec/schemas")
|
10
|
-
end
|
11
|
-
|
12
|
-
it "encodes and decodes data using a named schema" do
|
13
|
-
define_schema "person.avsc", <<-AVSC
|
14
|
-
{
|
15
|
-
"name": "person",
|
16
|
-
"type": "record",
|
17
|
-
"fields": [
|
18
|
-
{
|
19
|
-
"type": "string",
|
20
|
-
"name": "full_name"
|
21
|
-
},
|
22
|
-
{
|
23
|
-
"name": "address",
|
24
|
-
"type": {
|
25
|
-
"type": "record",
|
26
|
-
"name": "address",
|
27
|
-
"fields": [
|
28
|
-
{
|
29
|
-
"type": "string",
|
30
|
-
"name": "street"
|
31
|
-
},
|
32
|
-
{
|
33
|
-
"type": "string",
|
34
|
-
"name": "city"
|
35
|
-
}
|
36
|
-
]
|
37
|
-
}
|
38
|
-
}
|
39
|
-
]
|
40
|
-
}
|
41
|
-
AVSC
|
42
|
-
|
43
|
-
data = {
|
44
|
-
"full_name" => "John Doe",
|
45
|
-
"address" => {
|
46
|
-
"street" => "Market st. 989",
|
47
|
-
"city" => "San Francisco"
|
48
|
-
}
|
49
|
-
}
|
50
|
-
|
51
|
-
encoded_data = avro.encode(data, schema_name: "person")
|
52
|
-
|
53
|
-
expect(avro.decode(encoded_data, schema_name: "person")).to eq(data)
|
54
|
-
end
|
55
|
-
|
56
|
-
it "resolves named types" do
|
57
|
-
define_schema "person.avsc", <<-AVSC
|
58
|
-
{
|
59
|
-
"name": "person",
|
60
|
-
"type": "record",
|
61
|
-
"fields": [
|
62
|
-
{
|
63
|
-
"type": "string",
|
64
|
-
"name": "full_name"
|
65
|
-
},
|
66
|
-
{
|
67
|
-
"name": "address",
|
68
|
-
"type": "address"
|
69
|
-
}
|
70
|
-
]
|
71
|
-
}
|
72
|
-
AVSC
|
73
|
-
|
74
|
-
define_schema "address.avsc", <<-AVSC
|
75
|
-
{
|
76
|
-
"type": "record",
|
77
|
-
"name": "address",
|
78
|
-
"fields": [
|
79
|
-
{
|
80
|
-
"type": "string",
|
81
|
-
"name": "street"
|
82
|
-
},
|
83
|
-
{
|
84
|
-
"type": "string",
|
85
|
-
"name": "city"
|
86
|
-
}
|
87
|
-
]
|
88
|
-
}
|
89
|
-
AVSC
|
90
|
-
|
91
|
-
data = {
|
92
|
-
"full_name" => "John Doe",
|
93
|
-
"address" => {
|
94
|
-
"street" => "Market st. 989",
|
95
|
-
"city" => "San Francisco"
|
96
|
-
}
|
97
|
-
}
|
98
|
-
|
99
|
-
encoded_data = avro.encode(data, schema_name: "person")
|
100
|
-
|
101
|
-
expect(avro.decode(encoded_data, schema_name: "person")).to eq(data)
|
102
|
-
end
|
103
|
-
|
104
|
-
it "allows decoding without specifying a reader schema" do
|
105
|
-
define_schema "message.avsc", <<-AVSC
|
106
|
-
{
|
107
|
-
"name": "message",
|
108
|
-
"type": "string"
|
109
|
-
}
|
110
|
-
AVSC
|
111
|
-
|
112
|
-
encoded_data = avro.encode("hello, world", schema_name: "message")
|
113
|
-
|
114
|
-
expect(avro.decode(encoded_data)).to eq "hello, world"
|
115
|
-
end
|
116
|
-
|
117
|
-
it "allows using namespaces in schemas" do
|
118
|
-
FileUtils.mkdir_p("spec/schemas/test/people")
|
119
|
-
|
120
|
-
define_schema "test/people/person.avsc", <<-AVSC
|
121
|
-
{
|
122
|
-
"name": "person",
|
123
|
-
"namespace": "test.people",
|
124
|
-
"type": "record",
|
125
|
-
"fields": [
|
126
|
-
{
|
127
|
-
"type": "string",
|
128
|
-
"name": "full_name"
|
129
|
-
},
|
130
|
-
{
|
131
|
-
"name": "address",
|
132
|
-
"type": "test.people.address"
|
133
|
-
}
|
134
|
-
]
|
135
|
-
}
|
136
|
-
AVSC
|
137
|
-
|
138
|
-
define_schema "test/people/address.avsc", <<-AVSC
|
139
|
-
{
|
140
|
-
"name": "address",
|
141
|
-
"namespace": "test.people",
|
142
|
-
"type": "record",
|
143
|
-
"fields": [
|
144
|
-
{
|
145
|
-
"type": "string",
|
146
|
-
"name": "street"
|
147
|
-
},
|
148
|
-
{
|
149
|
-
"type": "string",
|
150
|
-
"name": "city"
|
151
|
-
}
|
152
|
-
]
|
153
|
-
}
|
154
|
-
AVSC
|
155
|
-
|
156
|
-
data = {
|
157
|
-
"full_name" => "John Doe",
|
158
|
-
"address" => {
|
159
|
-
"street" => "Market st. 989",
|
160
|
-
"city" => "San Francisco"
|
161
|
-
}
|
162
|
-
}
|
163
|
-
|
164
|
-
encoded_data = avro.encode(data, schema_name: "person", namespace: "test.people")
|
165
|
-
|
166
|
-
expect(avro.decode(encoded_data, schema_name: "person", namespace: "test.people")).to eq(data)
|
167
|
-
end
|
168
|
-
|
169
|
-
it "raises AvroTurf::SchemaError if the schema's namespace doesn't match the file location" do
|
170
|
-
FileUtils.mkdir_p("spec/schemas/test/people")
|
171
|
-
|
172
|
-
define_schema "test/people/person.avsc", <<-AVSC
|
173
|
-
{
|
174
|
-
"name": "person",
|
175
|
-
"namespace": "yoyoyo.nanana",
|
176
|
-
"type": "record",
|
177
|
-
"fields": [
|
178
|
-
{
|
179
|
-
"type": "string",
|
180
|
-
"name": "full_name"
|
181
|
-
}
|
182
|
-
]
|
183
|
-
}
|
184
|
-
AVSC
|
185
|
-
|
186
|
-
data = { "full_name" => "John Doe" }
|
187
|
-
|
188
|
-
expect {
|
189
|
-
avro.encode(data, schema_name: "test.people.person")
|
190
|
-
}.to raise_error(AvroTurf::SchemaError, "expected schema `spec/schemas/test/people/person.avsc' to define type `test.people.person'")
|
191
|
-
end
|
192
|
-
|
193
|
-
it "caches schemas in memory" do
|
194
|
-
define_schema "person.avsc", <<-AVSC
|
195
|
-
{
|
196
|
-
"name": "person",
|
197
|
-
"type": "record",
|
198
|
-
"fields": [
|
199
|
-
{
|
200
|
-
"type": "string",
|
201
|
-
"name": "full_name"
|
202
|
-
}
|
203
|
-
]
|
204
|
-
}
|
205
|
-
AVSC
|
206
|
-
|
207
|
-
data = {
|
208
|
-
"full_name" => "John Doe"
|
209
|
-
}
|
210
|
-
|
211
|
-
# Warm the schema cache.
|
212
|
-
avro.encode(data, schema_name: "person")
|
213
|
-
|
214
|
-
# Force a failure if the schema file is read again.
|
215
|
-
FileUtils.rm("spec/schemas/person.avsc")
|
216
|
-
|
217
|
-
encoded_data = avro.encode(data, schema_name: "person")
|
218
|
-
|
219
|
-
expect(avro.decode(encoded_data, schema_name: "person")).to eq(data)
|
220
|
-
end
|
221
|
-
|
222
|
-
def define_schema(path, content)
|
223
|
-
File.open(File.join("spec/schemas", path), "w") do |f|
|
224
|
-
f.write(content)
|
225
|
-
end
|
226
|
-
end
|
227
|
-
end
|