avro_turf 0.8.1 → 0.11.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.circleci/config.yml +36 -0
- data/.github/workflows/ruby.yml +20 -0
- data/CHANGELOG.md +20 -0
- data/Gemfile +0 -3
- data/README.md +20 -0
- data/avro_turf.gemspec +6 -5
- data/lib/avro_turf.rb +10 -1
- data/lib/avro_turf/cached_confluent_schema_registry.rb +18 -6
- data/lib/avro_turf/confluent_schema_registry.rb +23 -4
- data/lib/avro_turf/disk_cache.rb +83 -0
- data/lib/avro_turf/in_memory_cache.rb +38 -0
- data/lib/avro_turf/messaging.rb +109 -16
- data/lib/avro_turf/schema_store.rb +35 -20
- data/lib/avro_turf/test/fake_confluent_schema_registry_server.rb +15 -3
- data/lib/avro_turf/version.rb +1 -1
- data/spec/cached_confluent_schema_registry_spec.rb +24 -2
- data/spec/confluent_schema_registry_spec.rb +13 -1
- data/spec/disk_cached_confluent_schema_registry_spec.rb +159 -0
- data/spec/messaging_spec.rb +205 -17
- data/spec/schema_store_spec.rb +36 -0
- data/spec/spec_helper.rb +8 -0
- data/spec/support/confluent_schema_registry_context.rb +8 -5
- data/spec/test/fake_confluent_schema_registry_server_spec.rb +40 -0
- metadata +34 -13
- data/circle.yml +0 -4
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 1e2ee4d9598bcaa2ec5738a1130fae0b19be7b5e7250f27540313b7471f97e23
|
4
|
+
data.tar.gz: 0b441cb30a153958c2ea283300a1a05c26218e2a97cb807fce36f8ad9d0240da
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 12779eac5c325752cfa1be34da94ef5f332490cda4ff0aef29529a00557008cecf39592396a3f3525a2a12cb67a46744800781e0da69d4bf02511f5a2284e5e7
|
7
|
+
data.tar.gz: a2e4c84fb338d62296aefb8ae8c206c262d756475ec11bd9cead17bfe6015ea069ff36a891df3381385fa650cb4b5ea1584855fccc5294f5910ab34563ad973a
|
@@ -0,0 +1,36 @@
|
|
1
|
+
version: 2
|
2
|
+
jobs:
|
3
|
+
build:
|
4
|
+
environment:
|
5
|
+
CIRCLE_ARTIFACTS: /tmp/circleci-artifacts
|
6
|
+
CIRCLE_TEST_REPORTS: /tmp/circleci-test-results
|
7
|
+
docker:
|
8
|
+
- image: circleci/ruby:2.6.2
|
9
|
+
steps:
|
10
|
+
- checkout
|
11
|
+
- run: mkdir -p $CIRCLE_ARTIFACTS $CIRCLE_TEST_REPORTS
|
12
|
+
- restore_cache:
|
13
|
+
keys:
|
14
|
+
# This branch if available
|
15
|
+
- v1-dep-{{ .Branch }}-
|
16
|
+
# Default branch if not
|
17
|
+
- v1-dep-master-
|
18
|
+
# Any branch if there are none on the default branch - this should be unnecessary if you have your default branch configured correctly
|
19
|
+
- v1-dep-
|
20
|
+
- run: gem install bundler --no-document
|
21
|
+
- run: 'bundle check --path=vendor/bundle || bundle install --path=vendor/bundle --jobs=4 --retry=3'
|
22
|
+
# Save dependency cache
|
23
|
+
- save_cache:
|
24
|
+
key: v1-dep-{{ .Branch }}-{{ epoch }}
|
25
|
+
paths:
|
26
|
+
- vendor/bundle
|
27
|
+
- ~/.bundle
|
28
|
+
- run: mkdir -p $CIRCLE_TEST_REPORTS/rspec
|
29
|
+
- run:
|
30
|
+
command: bundle exec rspec --color --require spec_helper --format progress
|
31
|
+
- store_test_results:
|
32
|
+
path: /tmp/circleci-test-results
|
33
|
+
- store_artifacts:
|
34
|
+
path: /tmp/circleci-artifacts
|
35
|
+
- store_artifacts:
|
36
|
+
path: /tmp/circleci-test-results
|
@@ -0,0 +1,20 @@
|
|
1
|
+
name: Ruby
|
2
|
+
|
3
|
+
on: [push]
|
4
|
+
|
5
|
+
jobs:
|
6
|
+
build:
|
7
|
+
|
8
|
+
runs-on: ubuntu-latest
|
9
|
+
|
10
|
+
steps:
|
11
|
+
- uses: actions/checkout@v1
|
12
|
+
- name: Set up Ruby 2.6
|
13
|
+
uses: actions/setup-ruby@v1
|
14
|
+
with:
|
15
|
+
ruby-version: 2.6.x
|
16
|
+
- name: Build and test with RSpec
|
17
|
+
run: |
|
18
|
+
gem install bundler
|
19
|
+
bundle install --jobs 4 --retry 3
|
20
|
+
bundle exec rspec
|
data/CHANGELOG.md
CHANGED
@@ -1,5 +1,25 @@
|
|
1
1
|
# avro_turf
|
2
2
|
|
3
|
+
## Unreleased
|
4
|
+
|
5
|
+
## v0.11.0
|
6
|
+
|
7
|
+
- Add proxy support (#107)
|
8
|
+
- Adding support for client certs (#109)
|
9
|
+
|
10
|
+
## v0.10.0
|
11
|
+
|
12
|
+
- Add more disk caching (#103)
|
13
|
+
- Include schema information when decoding (#100, #101, #104)
|
14
|
+
|
15
|
+
## v0.9.0
|
16
|
+
|
17
|
+
- Compatibility with Avro v1.9.0 (#94)
|
18
|
+
- Disable the auto registeration of schema (#95)
|
19
|
+
- abstracted caching from CachedConfluentSchemaRegistry (#74)
|
20
|
+
- Load avro-patches if installed to silence deprecation errors (#85)
|
21
|
+
- Make schema store to be thread safe (#92)
|
22
|
+
|
3
23
|
## v0.8.1
|
4
24
|
|
5
25
|
- Allow accessing schema store from outside AvroTurf (#68).
|
data/Gemfile
CHANGED
data/README.md
CHANGED
@@ -124,9 +124,29 @@ avro = AvroTurf::Messaging.new(registry_url: "http://my-registry:8081/")
|
|
124
124
|
# time a schema is used.
|
125
125
|
data = avro.encode({ "title" => "hello, world" }, schema_name: "greeting")
|
126
126
|
|
127
|
+
# If you don't want to automatically register new schemas, you can pass explicitly
|
128
|
+
# subject and version to specify which schema should be used for encoding.
|
129
|
+
# It will fetch that schema from the registry and cache it. Subsequent instances
|
130
|
+
# of the same schema version will be served by the cache.
|
131
|
+
data = avro.encode({ "title" => "hello, world" }, subject: 'greeting', version: 1)
|
132
|
+
|
133
|
+
# You can also pass explicitly schema_id to specify which schema
|
134
|
+
# should be used for encoding.
|
135
|
+
# It will fetch that schema from the registry and cache it. Subsequent instances
|
136
|
+
# of the same schema version will be served by the cache.
|
137
|
+
data = avro.encode({ "title" => "hello, world" }, schema_id: 2)
|
138
|
+
|
127
139
|
# When decoding, the schema will be fetched from the registry and cached. Subsequent
|
128
140
|
# instances of the same schema id will be served by the cache.
|
129
141
|
avro.decode(data) #=> { "title" => "hello, world" }
|
142
|
+
|
143
|
+
# If you want to get decoded message as well as the schema used to encode the message,
|
144
|
+
# you can use `#decode_message` method.
|
145
|
+
result = avro.decode_message(data)
|
146
|
+
result.message #=> { "title" => "hello, world" }
|
147
|
+
result.schema_id #=> 3
|
148
|
+
result.writer_schema #=> #<Avro::Schema: ...>
|
149
|
+
result.reader_schema #=> nil
|
130
150
|
```
|
131
151
|
|
132
152
|
### Confluent Schema Registry Client
|
data/avro_turf.gemspec
CHANGED
@@ -17,16 +17,17 @@ Gem::Specification.new do |spec|
|
|
17
17
|
spec.test_files = spec.files.grep(%r{^(test|spec|features)/})
|
18
18
|
spec.require_paths = ["lib"]
|
19
19
|
|
20
|
-
spec.add_dependency "avro", ">= 1.7.7", "< 1.
|
20
|
+
spec.add_dependency "avro", ">= 1.7.7", "< 1.10"
|
21
21
|
spec.add_dependency "excon", "~> 0.45"
|
22
22
|
|
23
|
-
spec.add_development_dependency "bundler", "~>
|
24
|
-
spec.add_development_dependency "rake", "~>
|
25
|
-
spec.add_development_dependency "rspec", "~> 3.2
|
26
|
-
spec.add_development_dependency "fakefs", "~> 0.
|
23
|
+
spec.add_development_dependency "bundler", "~> 2.0"
|
24
|
+
spec.add_development_dependency "rake", "~> 13.0"
|
25
|
+
spec.add_development_dependency "rspec", "~> 3.2"
|
26
|
+
spec.add_development_dependency "fakefs", "~> 0.20.0"
|
27
27
|
spec.add_development_dependency "webmock"
|
28
28
|
spec.add_development_dependency "sinatra"
|
29
29
|
spec.add_development_dependency "json_spec"
|
30
|
+
spec.add_development_dependency "rack-test"
|
30
31
|
|
31
32
|
spec.post_install_message = %{
|
32
33
|
avro_turf v0.8.0 deprecates the names AvroTurf::SchemaRegistry,
|
data/lib/avro_turf.rb
CHANGED
@@ -1,9 +1,18 @@
|
|
1
|
+
begin
|
2
|
+
require 'avro-patches'
|
3
|
+
rescue LoadError
|
4
|
+
false
|
5
|
+
end
|
1
6
|
require 'avro_turf/version'
|
2
7
|
require 'avro'
|
3
8
|
require 'json'
|
4
9
|
require 'avro_turf/schema_store'
|
5
10
|
require 'avro_turf/core_ext'
|
6
|
-
|
11
|
+
|
12
|
+
# check for something that indicates Avro v1.9.0 or later
|
13
|
+
unless defined?(::Avro::LogicalTypes)
|
14
|
+
require 'avro_turf/schema_to_avro_patch'
|
15
|
+
end
|
7
16
|
|
8
17
|
class AvroTurf
|
9
18
|
class Error < StandardError; end
|
@@ -1,16 +1,23 @@
|
|
1
1
|
require 'avro_turf/confluent_schema_registry'
|
2
|
+
require 'avro_turf/in_memory_cache'
|
3
|
+
require 'avro_turf/disk_cache'
|
2
4
|
|
3
5
|
# Caches registrations and lookups to the schema registry in memory.
|
4
6
|
class AvroTurf::CachedConfluentSchemaRegistry
|
5
7
|
|
6
|
-
|
8
|
+
# Instantiate a new CachedConfluentSchemaRegistry instance with the given configuration.
|
9
|
+
# By default, uses a provided InMemoryCache to prevent repeated calls to the upstream registry.
|
10
|
+
#
|
11
|
+
# upstream - The upstream schema registry object that fully responds to all methods in the
|
12
|
+
# AvroTurf::ConfluentSchemaRegistry interface.
|
13
|
+
# cache - Optional user provided Cache object that responds to all methods in the AvroTurf::InMemoryCache interface.
|
14
|
+
def initialize(upstream, cache: nil)
|
7
15
|
@upstream = upstream
|
8
|
-
@
|
9
|
-
@ids_by_schema = {}
|
16
|
+
@cache = cache || AvroTurf::InMemoryCache.new()
|
10
17
|
end
|
11
18
|
|
12
19
|
# Delegate the following methods to the upstream
|
13
|
-
%i(subjects subject_versions
|
20
|
+
%i(subjects subject_versions check compatible?
|
14
21
|
global_config update_global_config subject_config update_subject_config).each do |name|
|
15
22
|
define_method(name) do |*args|
|
16
23
|
instance_variable_get(:@upstream).send(name, *args)
|
@@ -18,10 +25,15 @@ class AvroTurf::CachedConfluentSchemaRegistry
|
|
18
25
|
end
|
19
26
|
|
20
27
|
def fetch(id)
|
21
|
-
@
|
28
|
+
@cache.lookup_by_id(id) || @cache.store_by_id(id, @upstream.fetch(id))
|
22
29
|
end
|
23
30
|
|
24
31
|
def register(subject, schema)
|
25
|
-
@
|
32
|
+
@cache.lookup_by_schema(subject, schema) || @cache.store_by_schema(subject, schema, @upstream.register(subject, schema))
|
33
|
+
end
|
34
|
+
|
35
|
+
def subject_version(subject, version = 'latest')
|
36
|
+
@cache.lookup_by_version(subject, version) ||
|
37
|
+
@cache.store_by_version(subject, version, @upstream.subject_version(subject, version))
|
26
38
|
end
|
27
39
|
end
|
@@ -3,11 +3,30 @@ require 'excon'
|
|
3
3
|
class AvroTurf::ConfluentSchemaRegistry
|
4
4
|
CONTENT_TYPE = "application/vnd.schemaregistry.v1+json".freeze
|
5
5
|
|
6
|
-
def initialize(
|
6
|
+
def initialize(
|
7
|
+
url,
|
8
|
+
logger: Logger.new($stdout),
|
9
|
+
proxy: nil,
|
10
|
+
client_cert: nil,
|
11
|
+
client_key: nil,
|
12
|
+
client_key_pass: nil,
|
13
|
+
client_cert_data: nil,
|
14
|
+
client_key_data: nil
|
15
|
+
)
|
7
16
|
@logger = logger
|
8
|
-
|
9
|
-
"Content-Type" => CONTENT_TYPE
|
10
|
-
}
|
17
|
+
headers = {
|
18
|
+
"Content-Type" => CONTENT_TYPE
|
19
|
+
}
|
20
|
+
headers[:proxy] = proxy if proxy&.present?
|
21
|
+
@connection = Excon.new(
|
22
|
+
url,
|
23
|
+
headers: headers,
|
24
|
+
client_cert: client_cert,
|
25
|
+
client_key: client_key,
|
26
|
+
client_key_pass: client_key_pass,
|
27
|
+
client_cert_data: client_cert_data,
|
28
|
+
client_key_data: client_key_data
|
29
|
+
)
|
11
30
|
end
|
12
31
|
|
13
32
|
def fetch(id)
|
@@ -0,0 +1,83 @@
|
|
1
|
+
# A cache for the CachedConfluentSchemaRegistry.
|
2
|
+
# Extends the InMemoryCache to provide a write-thru to disk for persistent cache.
|
3
|
+
class AvroTurf::DiskCache < AvroTurf::InMemoryCache
|
4
|
+
|
5
|
+
def initialize(disk_path)
|
6
|
+
super()
|
7
|
+
|
8
|
+
# load the write-thru cache on startup, if it exists
|
9
|
+
@schemas_by_id_path = File.join(disk_path, 'schemas_by_id.json')
|
10
|
+
@schemas_by_id = JSON.parse(File.read(@schemas_by_id_path)) if File.exist?(@schemas_by_id_path)
|
11
|
+
|
12
|
+
@ids_by_schema_path = File.join(disk_path, 'ids_by_schema.json')
|
13
|
+
@ids_by_schema = JSON.parse(File.read(@ids_by_schema_path)) if File.exist?(@ids_by_schema_path)
|
14
|
+
|
15
|
+
@schemas_by_subject_version_path = File.join(disk_path, 'schemas_by_subject_version.json')
|
16
|
+
@schemas_by_subject_version = {}
|
17
|
+
end
|
18
|
+
|
19
|
+
# override
|
20
|
+
# the write-thru cache (json) does not store keys in numeric format
|
21
|
+
# so, convert id to a string for caching purposes
|
22
|
+
def lookup_by_id(id)
|
23
|
+
super(id.to_s)
|
24
|
+
end
|
25
|
+
|
26
|
+
# override to include write-thru cache after storing result from upstream
|
27
|
+
def store_by_id(id, schema)
|
28
|
+
# must return the value from storing the result (i.e. do not return result from file write)
|
29
|
+
value = super(id.to_s, schema)
|
30
|
+
File.write(@schemas_by_id_path, JSON.pretty_generate(@schemas_by_id))
|
31
|
+
return value
|
32
|
+
end
|
33
|
+
|
34
|
+
# override to include write-thru cache after storing result from upstream
|
35
|
+
def store_by_schema(subject, schema, id)
|
36
|
+
# must return the value from storing the result (i.e. do not return result from file write)
|
37
|
+
value = super
|
38
|
+
File.write(@ids_by_schema_path, JSON.pretty_generate(@ids_by_schema))
|
39
|
+
return value
|
40
|
+
end
|
41
|
+
|
42
|
+
# checks instance var (in-memory cache) for schema
|
43
|
+
# checks disk cache if in-memory cache doesn't exists
|
44
|
+
# if file exists but no in-memory cache, read from file and sync in-memory cache
|
45
|
+
# finally, if file doesn't exist return nil
|
46
|
+
def lookup_by_version(subject, version)
|
47
|
+
key = "#{subject}#{version}"
|
48
|
+
schema = @schemas_by_subject_version[key]
|
49
|
+
|
50
|
+
return schema unless schema.nil?
|
51
|
+
|
52
|
+
hash = JSON.parse(File.read(@schemas_by_subject_version_path)) if File.exist?(@schemas_by_subject_version_path)
|
53
|
+
if hash
|
54
|
+
@schemas_by_subject_version = hash
|
55
|
+
@schemas_by_subject_version[key]
|
56
|
+
end
|
57
|
+
end
|
58
|
+
|
59
|
+
# check if file exists and parse json into a hash
|
60
|
+
# if file exists take json and overwite/insert schema at key
|
61
|
+
# if file doesn't exist create new hash
|
62
|
+
# write the new/updated hash to file
|
63
|
+
# update instance var (in memory-cache) to match
|
64
|
+
def store_by_version(subject, version, schema)
|
65
|
+
key = "#{subject}#{version}"
|
66
|
+
hash = JSON.parse(File.read(@schemas_by_subject_version_path)) if File.exist?(@schemas_by_subject_version_path)
|
67
|
+
hash = if hash
|
68
|
+
hash[key] = schema
|
69
|
+
hash
|
70
|
+
else
|
71
|
+
{key => schema}
|
72
|
+
end
|
73
|
+
|
74
|
+
write_to_disk_cache(@schemas_by_subject_version_path, hash)
|
75
|
+
|
76
|
+
@schemas_by_subject_version = hash
|
77
|
+
@schemas_by_subject_version[key]
|
78
|
+
end
|
79
|
+
|
80
|
+
private def write_to_disk_cache(path, hash)
|
81
|
+
File.write(path, JSON.pretty_generate(hash))
|
82
|
+
end
|
83
|
+
end
|
@@ -0,0 +1,38 @@
|
|
1
|
+
# A cache for the CachedConfluentSchemaRegistry.
|
2
|
+
# Simply stores the schemas and ids in in-memory hashes.
|
3
|
+
class AvroTurf::InMemoryCache
|
4
|
+
|
5
|
+
def initialize
|
6
|
+
@schemas_by_id = {}
|
7
|
+
@ids_by_schema = {}
|
8
|
+
@schema_by_subject_version = {}
|
9
|
+
end
|
10
|
+
|
11
|
+
def lookup_by_id(id)
|
12
|
+
@schemas_by_id[id]
|
13
|
+
end
|
14
|
+
|
15
|
+
def store_by_id(id, schema)
|
16
|
+
@schemas_by_id[id] = schema
|
17
|
+
end
|
18
|
+
|
19
|
+
def lookup_by_schema(subject, schema)
|
20
|
+
key = subject + schema.to_s
|
21
|
+
@ids_by_schema[key]
|
22
|
+
end
|
23
|
+
|
24
|
+
def store_by_schema(subject, schema, id)
|
25
|
+
key = subject + schema.to_s
|
26
|
+
@ids_by_schema[key] = id
|
27
|
+
end
|
28
|
+
|
29
|
+
def lookup_by_version(subject, version)
|
30
|
+
key = "#{subject}#{version}"
|
31
|
+
@schema_by_subject_version[key]
|
32
|
+
end
|
33
|
+
|
34
|
+
def store_by_version(subject, version, schema)
|
35
|
+
key = "#{subject}#{version}"
|
36
|
+
@schema_by_subject_version[key] = schema
|
37
|
+
end
|
38
|
+
end
|
data/lib/avro_turf/messaging.rb
CHANGED
@@ -21,21 +21,53 @@ class AvroTurf
|
|
21
21
|
# 1: https://github.com/confluentinc/schema-registry
|
22
22
|
class Messaging
|
23
23
|
MAGIC_BYTE = [0].pack("C").freeze
|
24
|
+
DecodedMessage = Struct.new(:schema_id, :writer_schema, :reader_schema, :message)
|
25
|
+
private_constant(:DecodedMessage)
|
24
26
|
|
25
27
|
# Instantiate a new Messaging instance with the given configuration.
|
26
28
|
#
|
27
|
-
# registry
|
28
|
-
#
|
29
|
-
# registry_url
|
30
|
-
# schema_store
|
31
|
-
# schemas_path
|
32
|
-
# namespace
|
33
|
-
# logger
|
34
|
-
|
29
|
+
# registry - A schema registry object that responds to all methods in the
|
30
|
+
# AvroTurf::ConfluentSchemaRegistry interface.
|
31
|
+
# registry_url - The String URL of the schema registry that should be used.
|
32
|
+
# schema_store - A schema store object that responds to #find(schema_name, namespace).
|
33
|
+
# schemas_path - The String file system path where local schemas are stored.
|
34
|
+
# namespace - The String default schema namespace.
|
35
|
+
# logger - The Logger that should be used to log information (optional).
|
36
|
+
# proxy - Forward the request via proxy (optional).
|
37
|
+
# client_cert - Name of file containing client certificate (optional).
|
38
|
+
# client_key - Name of file containing client private key to go with client_cert (optional).
|
39
|
+
# client_key_pass - Password to go with client_key (optional).
|
40
|
+
# client_cert_data - In-memory client certificate (optional).
|
41
|
+
# client_key_data - In-memory client private key to go with client_cert_data (optional).
|
42
|
+
def initialize(
|
43
|
+
registry: nil,
|
44
|
+
registry_url: nil,
|
45
|
+
schema_store: nil,
|
46
|
+
schemas_path: nil,
|
47
|
+
namespace: nil,
|
48
|
+
logger: nil,
|
49
|
+
proxy: nil,
|
50
|
+
client_cert: nil,
|
51
|
+
client_key: nil,
|
52
|
+
client_key_pass: nil,
|
53
|
+
client_cert_data: nil,
|
54
|
+
client_key_data: nil
|
55
|
+
)
|
35
56
|
@logger = logger || Logger.new($stderr)
|
36
57
|
@namespace = namespace
|
37
58
|
@schema_store = schema_store || SchemaStore.new(path: schemas_path || DEFAULT_SCHEMAS_PATH)
|
38
|
-
@registry = registry || CachedConfluentSchemaRegistry.new(
|
59
|
+
@registry = registry || CachedConfluentSchemaRegistry.new(
|
60
|
+
ConfluentSchemaRegistry.new(
|
61
|
+
registry_url,
|
62
|
+
logger: @logger,
|
63
|
+
proxy: proxy,
|
64
|
+
client_cert: client_cert,
|
65
|
+
client_key: client_key,
|
66
|
+
client_key_pass: client_key_pass,
|
67
|
+
client_cert_data: client_cert_data,
|
68
|
+
client_key_data: client_key_data
|
69
|
+
)
|
70
|
+
)
|
39
71
|
@schemas_by_id = {}
|
40
72
|
end
|
41
73
|
|
@@ -46,14 +78,24 @@ class AvroTurf
|
|
46
78
|
# schema_name - The String name of the schema that should be used to encode
|
47
79
|
# the data.
|
48
80
|
# namespace - The namespace of the schema (optional).
|
81
|
+
# subject - The subject name the schema should be registered under in
|
82
|
+
# the schema registry (optional).
|
83
|
+
# version - The integer version of the schema that should be used to decode
|
84
|
+
# the data. Must match the schema used when encoding (optional).
|
85
|
+
# schema_id - The integer id of the schema that should be used to encode
|
86
|
+
# the data.
|
49
87
|
#
|
50
88
|
# Returns the encoded data as a String.
|
51
|
-
def encode(message, schema_name: nil, namespace: @namespace, subject: nil)
|
52
|
-
schema =
|
53
|
-
|
54
|
-
|
55
|
-
|
56
|
-
|
89
|
+
def encode(message, schema_name: nil, namespace: @namespace, subject: nil, version: nil, schema_id: nil)
|
90
|
+
schema_id, schema = if schema_id
|
91
|
+
fetch_schema_by_id(schema_id)
|
92
|
+
elsif subject && version
|
93
|
+
fetch_schema(subject, version)
|
94
|
+
elsif schema_name
|
95
|
+
register_schema(subject, schema_name, namespace)
|
96
|
+
else
|
97
|
+
raise ArgumentError.new('Neither schema_name nor schema_id nor subject + version provided to determine the schema.')
|
98
|
+
end
|
57
99
|
|
58
100
|
stream = StringIO.new
|
59
101
|
writer = Avro::IO::DatumWriter.new(schema)
|
@@ -69,6 +111,12 @@ class AvroTurf
|
|
69
111
|
writer.write(message, encoder)
|
70
112
|
|
71
113
|
stream.string
|
114
|
+
rescue Excon::Error::NotFound
|
115
|
+
if schema_id
|
116
|
+
raise SchemaNotFoundError.new("Schema with id: #{schema_id} is not found on registry")
|
117
|
+
else
|
118
|
+
raise SchemaNotFoundError.new("Schema with subject: `#{subject}` version: `#{version}` is not found on registry")
|
119
|
+
end
|
72
120
|
end
|
73
121
|
|
74
122
|
# Decodes data into the original message.
|
@@ -80,6 +128,20 @@ class AvroTurf
|
|
80
128
|
#
|
81
129
|
# Returns the decoded message.
|
82
130
|
def decode(data, schema_name: nil, namespace: @namespace)
|
131
|
+
decode_message(data, schema_name: schema_name, namespace: namespace).message
|
132
|
+
end
|
133
|
+
|
134
|
+
# Decodes data into the original message.
|
135
|
+
#
|
136
|
+
# data - A String containing encoded data.
|
137
|
+
# schema_name - The String name of the schema that should be used to decode
|
138
|
+
# the data. Must match the schema used when encoding (optional).
|
139
|
+
# namespace - The namespace of the schema (optional).
|
140
|
+
#
|
141
|
+
# Returns Struct with the next attributes:
|
142
|
+
# schema_id - The integer id of schema used to encode the message
|
143
|
+
# message - The decoded message
|
144
|
+
def decode_message(data, schema_name: nil, namespace: @namespace)
|
83
145
|
readers_schema = schema_name && @schema_store.find(schema_name, namespace)
|
84
146
|
stream = StringIO.new(data)
|
85
147
|
decoder = Avro::IO::BinaryDecoder.new(stream)
|
@@ -100,7 +162,38 @@ class AvroTurf
|
|
100
162
|
end
|
101
163
|
|
102
164
|
reader = Avro::IO::DatumReader.new(writers_schema, readers_schema)
|
103
|
-
reader.read(decoder)
|
165
|
+
message = reader.read(decoder)
|
166
|
+
|
167
|
+
DecodedMessage.new(schema_id, writers_schema, readers_schema, message)
|
168
|
+
rescue Excon::Error::NotFound
|
169
|
+
raise SchemaNotFoundError.new("Schema with id: #{schema_id} is not found on registry")
|
170
|
+
end
|
171
|
+
|
172
|
+
private
|
173
|
+
|
174
|
+
# Providing subject and version to determine the schema,
|
175
|
+
# which skips the auto registeration of schema on the schema registry.
|
176
|
+
# Fetch the schema from registry with the provided subject name and version.
|
177
|
+
def fetch_schema(subject, version)
|
178
|
+
schema_data = @registry.subject_version(subject, version)
|
179
|
+
schema_id = schema_data.fetch('id')
|
180
|
+
schema = Avro::Schema.parse(schema_data.fetch('schema'))
|
181
|
+
[schema_id, schema]
|
182
|
+
end
|
183
|
+
|
184
|
+
# Fetch the schema from registry with the provided schema_id.
|
185
|
+
def fetch_schema_by_id(schema_id)
|
186
|
+
schema_json = @registry.fetch(schema_id)
|
187
|
+
schema = Avro::Schema.parse(schema_json)
|
188
|
+
[schema_id, schema]
|
189
|
+
end
|
190
|
+
|
191
|
+
# Schemas are registered under the full name of the top level Avro record
|
192
|
+
# type, or `subject` if it's provided.
|
193
|
+
def register_schema(subject, schema_name, namespace)
|
194
|
+
schema = @schema_store.find(schema_name, namespace)
|
195
|
+
schema_id = @registry.register(subject || schema.fullname, schema)
|
196
|
+
[schema_id, schema]
|
104
197
|
end
|
105
198
|
end
|
106
199
|
end
|