- 23 2月, 2019 3 次提交
-
-
由 Timo Walther 提交于
It adds CsvRowSerializationSchema, CsvRowDeserializationSchema, a new CSV descriptor, CsvRowTableFormatFactory, documentation and tests The format integrates nicely with most SQL types. It deprecates the "old CSV" descriptor stack and prepares also for FLINK-7050 (#4660). The old CSV descriptor is still available under as "OldCsv". This closes #7777.
-
由 wind 提交于
This closes #6541.
-
由 Stefan Richter 提交于
This closes #7768.
-
- 22 2月, 2019 36 次提交
-
-
由 Congxian Qiu 提交于
[FLINK-11704][tests] Improve AbstractCheckpointStateOutputStreamTestBase by using JUnit's Parameterized This closes #7784.
-
由 TANG Wen-hui 提交于
This closes #7806.
-
由 TANG Wen-hui 提交于
-
由 TANG Wen-hui 提交于
-
由 tison 提交于
-
由 tison 提交于
-
由 Aljoscha Krettek 提交于
-
由 Alexey Trenikhin 提交于
We now directly use the ConsumerRecord from the Kafka API instead of trying to forward what we need to the deserialization schema ourselves. This makes it more future-proof, if Kafka adds new fields to the ConsumerRecord. The previously used KeyedDeserializationSchema now extends KafkaDeserializationSchema and has a default method to bridge the interface. This way existing uses of KeyedDeserializationSchema still work.
-
由 Alexey Trenikhin 提交于
-
由 Alexey Trenikhin 提交于
This moves the code to a method in the base class.
-
由 Kostas Kloudas 提交于
-
由 ZiLi Chen 提交于
-
由 Gary Yao 提交于
-
由 tison 提交于
-
由 zentol 提交于
-
由 Tzu-Li (Gordon) Tai 提交于
In this case, there is no specific need to let that cache be specifically a HashMap. Changing this to the generic Map does not break contracts with how the PojoSerializer works. It also doesn't break compatibility w.r.t. Java serialization of the PojoSerializer.
-
由 Tzu-Li (Gordon) Tai 提交于
In the PojoSerializerSnapshot#resolveSchemaCompatibility method we call for each field the findField(...) method in the PojoSerializer, which performs a linear scan of the array of fields. This effectively has O(n^2) complexity. This is improved by this commit by building an index in the resolveSchemaCompatibility method for faster field serializer lookups.
-
由 Tzu-Li (Gordon) Tai 提交于
-
由 Tzu-Li (Gordon) Tai 提交于
-
由 Tzu-Li (Gordon) Tai 提交于
The following tests have been adjusted to new scope of functionality related to upgrading the PojoSerializer: PojoSerializerTest and PojoSerializerUpgradeTest. In PojoSerializerTest, assertions that were supposed to check for "isCompatibleWithReconfiguredSerializer" cases have been adjusted to correctly do so. Also, a test where changing field order was tested was removed, since the new PojoSerializerSnapshot always assumes that fields have been sorted already by the type extractor. In PojoSerializerUpgradeTest, tests which remove / add fields to POJO classes were updated to no longer expect failures. They were expecting failures because POJOs were not evolvable in the past. The updates to these tests also provide coverage for the POJO schema evolution feature. This closes #7759.
-
由 Tzu-Li (Gordon) Tai 提交于
-
由 Tzu-Li (Gordon) Tai 提交于
-
由 Tzu-Li (Gordon) Tai 提交于
[FLINK-11485] [core] Refactor multiple nested serializer compatibility resolution logic to CompositeTypeSerializerUtil This commit refactors the logic of resolving overall compatibility results across multiple nested serializers of a composite serializer out of the CompositeTypeSerializerSnapshot class. This allows us to reuse this functionality when implementing the compatibility check for the PojoSerializerSnapshot
-
由 Tzu-Li (Gordon) Tai 提交于
Having "Linked" in the name better reflects the fact that the map preserves entry ordering.
-
由 Tzu-Li (Gordon) Tai 提交于
This utility class may commonly be used by the KryoSerializer and the PojoSerializer. This commit moves the utility to a more suitable package and adjusts usage access.
-
由 Igal Shilman 提交于
This closes #7496.
-
由 Igal Shilman 提交于
-
由 Igal Shilman 提交于
-
由 Igal Shilman 提交于
This commit adds two more tests to TypeSerializerSnapshotMigrationTestBase that validates that the new serializer are able to read previously stored data (perhaps after reconfiguration).
-
由 Igal Shilman 提交于
This commit adds a new KryoSerializerSnapshot class and uses that instead of the now deprecated KryoSerializerConfigSnapshot. It also adds migration tests for the new KryoSerializerSnapshot, extending the TypeSerializerSnapshotMigrationTestBase.
-
由 Igal Shilman 提交于
[FLINK-11323] [tests] Adjust serializer tests to respect new COMPATIBLE_WITH_RECONFIGURED_SERIALIZER option Adjust serializer tests to use the reconfigured serializer if such serializer was provided.
-
由 Igal Shilman 提交于
An OptinalMap is a map that allows null values and null keys (as long as a name is assosicated with them. This would replace the usage of DummyUnloadableClass and DummyUnloadableSerializer within KryoSerializerSnapshot.
-
由 Tzu-Li (Gordon) Tai 提交于
The test enum type written in test resource files has different serialVersionUIDs generated between Scala 2.11 / 2.12. The test fails in 2.12 because we wrote the test files with 2.11. This commit fixes that by explicitly setting a fixed serialVersionUID for the test enum type.
-
由 zentol 提交于
-
由 Chesnay Schepler 提交于
-
由 Gary Yao 提交于
-
- 21 2月, 2019 1 次提交
-
-
由 zentol 提交于
-