avro_turf 1.3.0 → 1.3.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ccc550653b8422953e55c1319f37f478d1851afab78b54c716ba23fb0e59af17
-  data.tar.gz: a5f4721aa89807ee5b401c9a55d69d030cc2c0b7d54764b75e350f50aec9c118
+  metadata.gz: a4e9638703b961c348d06adcfa3e34ac2bd908e68f9cf7762a550ae4ef453b8a
+  data.tar.gz: 6d53b1f9b5e4b9e5a3aaeb13207bf350e2b56c8db138cb499643180a94a684b7
 SHA512:
-  metadata.gz: 8e8e81e7e00c1fb40d43bc8ef03c64af8b4d79c921d5eaac9b32ea0039e4c877956612e98061948d7dbc0b88b320a96858b7d31d6f9f9b7bd30bead3da985035
-  data.tar.gz: 3301e480eb51017cfd58fad552afc2b7c60b44545518fcaa8f892f90c15c7c03eb176e56af1447692d660c912cba1b22a20044f15d23d7ba949af3d075eed23b
+  metadata.gz: 7045f852f20d3ddeca3256724c918ed255722a1556f662d159b805921670d0a2af3e574e3c967bee7622d5d3e05f5348d6023b250006c3651cb7b1f9cec059bf
+  data.tar.gz: 50ca19fb058246ba19b28472e03bedcfbc09ddcd336d46deb29dec4b406d81a65c2b0f6ca84ec36d5227da035906976af133cf757cbfc2f20abda116c8e17d8c

data/.github/workflows/ruby.yml

@@ -6,6 +6,10 @@ jobs:
   build:
 
     runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        ruby-version: [1.8.x, 1.9.x, 2.0.x, 2.1.x, 2.2.x,
+          2.3.x, 2.4.x, 2.5.x, 2.6.x, 2.7.x, 3.0.x]
 
     steps:
     - uses: actions/checkout@v1

data/CHANGELOG.md

@@ -2,6 +2,11 @@
 
 ## Unreleased
 
+## v1.3.1
+
+- Prevent CachedConfluentSchemaRegistry from caching the 'latest' version (#140)
+- Fix issue with zero length schema cache file (#138)
+
 ## v1.3.0
 
 - Add support for plain user/password auth to ConfluentSchemaRegistry (#120)

data/lib/avro_turf/cached_confluent_schema_registry.rb

@@ -33,6 +33,8 @@ class AvroTurf::CachedConfluentSchemaRegistry
   end
 
   def subject_version(subject, version = 'latest')
+    return @upstream.subject_version(subject, version) if version == 'latest'
+    
     @cache.lookup_by_version(subject, version) ||
       @cache.store_by_version(subject, version, @upstream.subject_version(subject, version))
   end

data/lib/avro_turf/confluent_schema_registry.rb

@@ -19,7 +19,7 @@ class AvroTurf::ConfluentSchemaRegistry
     headers = {
       "Content-Type" => CONTENT_TYPE
     }
-    headers[:proxy] = proxy if proxy&.present?
+    headers[:proxy] = proxy unless proxy.nil?
     @connection = Excon.new(
       url,
       headers: headers,

data/lib/avro_turf/disk_cache.rb

@@ -2,15 +2,19 @@
 # Extends the InMemoryCache to provide a write-thru to disk for persistent cache.
 class AvroTurf::DiskCache < AvroTurf::InMemoryCache
 
-  def initialize(disk_path)
+  def initialize(disk_path, logger: Logger.new($stdout))
     super()
 
+    @logger = logger
+
     # load the write-thru cache on startup, if it exists
     @schemas_by_id_path = File.join(disk_path, 'schemas_by_id.json')
-    @schemas_by_id = JSON.parse(File.read(@schemas_by_id_path)) if File.exist?(@schemas_by_id_path)
+    hash = read_from_disk_cache(@schemas_by_id_path)
+    @schemas_by_id = hash if hash
 
     @ids_by_schema_path = File.join(disk_path, 'ids_by_schema.json')
-    @ids_by_schema = JSON.parse(File.read(@ids_by_schema_path)) if File.exist?(@ids_by_schema_path)
+    hash = read_from_disk_cache(@ids_by_schema_path)
+    @ids_by_schema = hash if hash
 
     @schemas_by_subject_version_path = File.join(disk_path, 'schemas_by_subject_version.json')
     @schemas_by_subject_version = {}
@@ -31,12 +35,18 @@ class AvroTurf::DiskCache < AvroTurf::InMemoryCache
     return value
   end
 
-  # override to include write-thru cache after storing result from upstream
+  # override to use a json serializable cache key
+  def lookup_by_schema(subject, schema)
+    key = "#{subject}#{schema}"
+    @ids_by_schema[key]
+  end
+
+  # override to use a json serializable cache key and update the file cache
   def store_by_schema(subject, schema, id)
-    # must return the value from storing the result (i.e. do not return result from file write)
-    value = super
+    key = "#{subject}#{schema}"
+    @ids_by_schema[key] = id
     File.write(@ids_by_schema_path, JSON.pretty_generate(@ids_by_schema))
-    return value
+    id
   end
 
   # checks instance var (in-memory cache) for schema
@@ -49,7 +59,7 @@ class AvroTurf::DiskCache < AvroTurf::InMemoryCache
 
     return schema unless schema.nil?
 
-    hash = JSON.parse(File.read(@schemas_by_subject_version_path)) if File.exist?(@schemas_by_subject_version_path)
+    hash = read_from_disk_cache(@schemas_by_subject_version_path)
     if hash
       @schemas_by_subject_version = hash
       @schemas_by_subject_version[key]
@@ -63,7 +73,7 @@ class AvroTurf::DiskCache < AvroTurf::InMemoryCache
   # update instance var (in memory-cache) to match
   def store_by_version(subject, version, schema)
     key = "#{subject}#{version}"
-    hash = JSON.parse(File.read(@schemas_by_subject_version_path)) if File.exist?(@schemas_by_subject_version_path)
+    hash = read_from_disk_cache(@schemas_by_subject_version_path)
     hash = if hash
              hash[key] = schema
              hash
@@ -77,6 +87,19 @@ class AvroTurf::DiskCache < AvroTurf::InMemoryCache
     @schemas_by_subject_version[key]
   end
 
+  # Parse the file from disk, if it exists and is not zero length
+  private def read_from_disk_cache(path)
+    if File.exist?(path)
+      if File.size(path)!=0
+        return JSON.parse(File.read(path))
+      else
+        # just log a message if skipping zero length file
+        @logger.warn "skipping JSON.parse of zero length file at #{path}"
+      end
+    end
+    return nil
+  end
+
   private def write_to_disk_cache(path, hash)
     File.write(path, JSON.pretty_generate(hash))
   end

data/lib/avro_turf/in_memory_cache.rb

@@ -17,12 +17,12 @@ class AvroTurf::InMemoryCache
   end
 
   def lookup_by_schema(subject, schema)
-    key = subject + schema.to_s
+    key = [subject, schema.to_s]
     @ids_by_schema[key]
   end
 
   def store_by_schema(subject, schema, id)
-    key = subject + schema.to_s
+    key = [subject, schema.to_s]
     @ids_by_schema[key] = id
   end
 

data/lib/avro_turf/messaging.rb

@@ -124,7 +124,7 @@ class AvroTurf
       writer.write(message, encoder)
 
       stream.string
-    rescue Excon::Error::NotFound
+    rescue Excon::Errors::NotFound
       if schema_id
         raise SchemaNotFoundError.new("Schema with id: #{schema_id} is not found on registry")
       else
@@ -194,8 +194,10 @@ class AvroTurf
 
     # Fetch the schema from registry with the provided schema_id.
     def fetch_schema_by_id(schema_id)
-      schema_json = @registry.fetch(schema_id)
-      schema = Avro::Schema.parse(schema_json)
+      schema = @schemas_by_id.fetch(schema_id) do
+        schema_json = @registry.fetch(schema_id)
+        Avro::Schema.parse(schema_json)
+      end
       [schema, schema_id]
     end
 

data/lib/avro_turf/schema_store.rb

@@ -22,7 +22,7 @@ class AvroTurf::SchemaStore
       # Still need to check is the schema already loaded
       return @schemas[fullname] if @schemas.key?(fullname)
 
-      load_schema!(fullname, namespace)
+      load_schema!(fullname)
     end
   end
 
@@ -42,13 +42,12 @@ class AvroTurf::SchemaStore
     end
   end
 
-  private
+  protected
 
   # Loads single schema
   # Such method is not thread-safe, do not call it of from mutex synchronization routine
-  def load_schema!(fullname, namespace = nil, local_schemas_cache = {})
-    *namespace, schema_name = fullname.split(".")
-    schema_path = File.join(@path, *namespace, schema_name + ".avsc")
+  def load_schema!(fullname, local_schemas_cache = {})
+    schema_path = build_schema_path(fullname)
     schema_json = JSON.parse(File.read(schema_path))
 
     schema = Avro::Schema.real_parse(schema_json, local_schemas_cache)
@@ -78,17 +77,22 @@ class AvroTurf::SchemaStore
       # Try to first resolve a referenced schema from disk.
       # If this is successful, the Avro gem will have mutated the
       # local_schemas_cache, adding all the new schemas it found.
-      load_schema!($1, nil, local_schemas_cache)
+      load_schema!($1, local_schemas_cache)
 
       # Attempt to re-parse the original schema now that the dependency
       # has been resolved and use the now-updated local_schemas_cache to
       # pick up where we left off.
       local_schemas_cache.delete(fullname)
-      load_schema!(fullname, nil, local_schemas_cache)
+      load_schema!(fullname, local_schemas_cache)
     else
       raise
     end
   rescue Errno::ENOENT, Errno::ENAMETOOLONG
     raise AvroTurf::SchemaNotFoundError, "could not find Avro schema at `#{schema_path}'"
   end
+
+  def build_schema_path(fullname)
+    *namespace, schema_name = fullname.split(".")
+    schema_path = File.join(@path, *namespace, schema_name + ".avsc")
+  end
 end

data/lib/avro_turf/version.rb

@@ -1,3 +1,3 @@
 class AvroTurf
-  VERSION = "1.3.0"
+  VERSION = "1.3.1"
 end

data/spec/disk_cached_confluent_schema_registry_spec.rb

@@ -4,7 +4,8 @@ require 'avro_turf/test/fake_confluent_schema_registry_server'
 
 describe AvroTurf::CachedConfluentSchemaRegistry do
   let(:upstream) { instance_double(AvroTurf::ConfluentSchemaRegistry) }
-  let(:cache)    { AvroTurf::DiskCache.new("spec/cache")}
+  let(:logger_io) { StringIO.new }
+  let(:cache)    { AvroTurf::DiskCache.new("spec/cache", logger: Logger.new(logger_io))}
   let(:registry) { described_class.new(upstream, cache: cache) }
   let(:id) { rand(999) }
   let(:schema) do
@@ -80,6 +81,40 @@ describe AvroTurf::CachedConfluentSchemaRegistry do
     end
   end
 
+  describe "#fetch (zero length cache file)" do
+    let(:cache_after) do
+      {
+        "#{id}" => "#{schema}"
+      }
+    end
+
+    before do
+      # setup the disk cache with a zero length file
+      File.write(File.join("spec/cache", "schemas_by_id.json"), '')
+    end
+
+    it "skips zero length disk cache" do
+      # multiple calls return same result, with only one upstream call
+      allow(upstream).to receive(:fetch).with(id).and_return(schema)
+      expect(registry.fetch(id)).to eq(schema)
+      expect(registry.fetch(id)).to eq(schema)
+      expect(upstream).to have_received(:fetch).exactly(1).times
+      expect(load_cache("schemas_by_id.json")).to eq cache_after
+      expect(logger_io.string).to include("zero length file at spec/cache/schemas_by_id.json")
+    end
+  end
+
+  describe "#fetch (corrupt cache file)" do
+    before do
+      # setup the disk cache with a corrupt file (i.e. not json)
+      File.write(File.join("spec/cache", "schemas_by_id.json"), 'NOTJSON')
+    end
+
+    it "raises error on corrupt cache file" do
+      expect{registry.fetch(id)}.to raise_error(JSON::ParserError, /unexpected token/)
+    end
+  end
+
   describe "#register" do
     let(:subject_name) { "a_subject" }
     let(:cache_before) do
@@ -120,6 +155,41 @@ describe AvroTurf::CachedConfluentSchemaRegistry do
     end
   end
 
+  describe "#register (zero length cache file)" do
+    let(:subject_name) { "a_subject" }
+    let(:cache_after) do
+      {
+        "#{subject_name}#{schema}" => id
+      }
+    end
+
+    before do
+      # setup the disk cache with a zero length file
+      File.write(File.join("spec/cache", "ids_by_schema.json"), '')
+    end
+
+    it "skips zero length disk cache" do
+      # multiple calls return same result, with only one upstream call
+      allow(upstream).to receive(:register).with(subject_name, schema).and_return(id)
+      expect(registry.register(subject_name, schema)).to eq(id)
+      expect(registry.register(subject_name, schema)).to eq(id)
+      expect(upstream).to have_received(:register).exactly(1).times
+      expect(load_cache("ids_by_schema.json")).to eq cache_after
+      expect(logger_io.string).to include("zero length file at spec/cache/ids_by_schema.json")
+    end
+  end
+
+  describe "#register (corrupt cache file)" do
+    before do
+      # setup the disk cache with a corrupt file (i.e. not json)
+      File.write(File.join("spec/cache", "ids_by_schema.json"), 'NOTJSON')
+    end
+
+    it "raises error on corrupt cache file" do
+      expect{registry.register(subject_name, schema)}.to raise_error(JSON::ParserError, /unexpected token/)
+    end
+  end
+
   describe "#subject_version" do
     it "writes thru to disk cache" do
       # multiple calls return same result, with zero upstream calls

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: avro_turf
 version: !ruby/object:Gem::Version
-  version: 1.3.0
+  version: 1.3.1
 platform: ruby
 authors:
 - Daniel Schierbeck
-autorequire:
+autorequire: 
 bindir: bin
 cert_chain: []
-date: 2020-09-10 00:00:00.000000000 Z
+date: 2021-02-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: avro
@@ -156,14 +156,13 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-description:
+description: 
 email:
 - dasch@zendesk.com
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- ".circleci/config.yml"
 - ".github/workflows/ruby.yml"
 - ".github/workflows/stale.yml"
 - ".gitignore"
@@ -251,7 +250,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubygems_version: 3.1.2
-signing_key:
+signing_key: 
 specification_version: 4
 summary: A library that makes it easier to use the Avro serialization format from
   Ruby

data/.circleci/config.yml

@@ -1,36 +0,0 @@
-version: 2
-jobs:
-  build:
-    environment:
-      CIRCLE_ARTIFACTS: /tmp/circleci-artifacts
-      CIRCLE_TEST_REPORTS: /tmp/circleci-test-results
-    docker:
-    - image: circleci/ruby:2.6.2
-    steps:
-    - checkout
-    - run: mkdir -p $CIRCLE_ARTIFACTS $CIRCLE_TEST_REPORTS
-    - restore_cache:
-        keys:
-        # This branch if available
-        - v1-dep-{{ .Branch }}-
-        # Default branch if not
-        - v1-dep-master-
-        # Any branch if there are none on the default branch - this should be unnecessary if you have your default branch configured correctly
-        - v1-dep-
-    - run: gem install bundler --no-document
-    - run: 'bundle check --path=vendor/bundle || bundle install --path=vendor/bundle --jobs=4 --retry=3'
-    # Save dependency cache
-    - save_cache:
-        key: v1-dep-{{ .Branch }}-{{ epoch }}
-        paths:
-        - vendor/bundle
-        - ~/.bundle
-    - run: mkdir -p $CIRCLE_TEST_REPORTS/rspec
-    - run:
-        command: bundle exec rspec --color --require spec_helper --format progress
-    - store_test_results:
-        path: /tmp/circleci-test-results
-    - store_artifacts:
-        path: /tmp/circleci-artifacts
-    - store_artifacts:
-        path: /tmp/circleci-test-results