[ https://issues.apache.org/jira/browse/FLINK-28528?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Xuannan Su updated FLINK-28528: ------------------------------- Description: The bug can be reproduced with the following test. The test can pass if we use the commented way to define the watermark. {code:python} def test_flink_2(self): env = StreamExecutionEnvironment.get_execution_environment() t_env = StreamTableEnvironment.create(env) table = t_env.from_descriptor( TableDescriptor.for_connector("filesystem") .schema( Schema.new_builder() .column("name", DataTypes.STRING()) .column("cost", DataTypes.INT()) .column("distance", DataTypes.INT()) .column("time", DataTypes.TIMESTAMP(3)) .watermark("time", expr.col("time") - expr.lit(60).seconds) # .watermark("time", "`time` - INTERVAL '60' SECOND") .build() ) .format("csv") .option("path", "./input.csv") .build() ) print(table.get_schema()) {code} It causes the following exception {code:none} E pyflink.util.exceptions.TableException: org.apache.flink.table.api.TableException: Expression 'minus(time, 60000)' is not string serializable. Currently, only expressions that originated from a SQL expression have a well-defined string representation. E at org.apache.flink.table.expressions.ResolvedExpression.asSerializableString(ResolvedExpression.java:51) E at org.apache.flink.table.api.TableSchema.lambda$fromResolvedSchema$13(TableSchema.java:455) E at java.util.Collections$SingletonList.forEach(Collections.java:4824) E at org.apache.flink.table.api.TableSchema.fromResolvedSchema(TableSchema.java:451) E at org.apache.flink.table.api.Table.getSchema(Table.java:101) E at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) E at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) E at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) E at java.lang.reflect.Method.invoke(Method.java:498) E at org.apache.flink.api.python.shaded.py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244) E at org.apache.flink.api.python.shaded.py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357) E at org.apache.flink.api.python.shaded.py4j.Gateway.invoke(Gateway.java:282) E at org.apache.flink.api.python.shaded.py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132) E at org.apache.flink.api.python.shaded.py4j.commands.CallCommand.execute(CallCommand.java:79) E at org.apache.flink.api.python.shaded.py4j.GatewayConnection.run(GatewayConnection.java:238) E at java.lang.Thread.run(Thread.java:748) {code} was: The bug can be reproduced with the following test. The test can pass if we use the commented way to define the watermark. {code:python} def test_flink_2(self): env = StreamExecutionEnvironment.get_execution_environment() t_env = StreamTableEnvironment.create(env) table = t_env.from_descriptor( TableDescriptor.for_connector("filesystem") .schema( Schema.new_builder() .column("name", DataTypes.STRING()) .column("cost", DataTypes.INT()) .column("distance", DataTypes.INT()) .column("time", DataTypes.TIMESTAMP(3)) .watermark("time", expr.col("time") - expr.lit(60).seconds) # .watermark("time", "`time` - INTERVAL '60' SECOND") .build() ) .format("csv") .option("path", "./input.csv") .build() ) print(table.get_schema()) {code} It causes the following exception {code:none} // Some comments here E pyflink.util.exceptions.TableException: org.apache.flink.table.api.TableException: Expression 'minus(time, 60000)' is not string serializable. Currently, only expressions that originated from a SQL expression have a well-defined string representation. E at org.apache.flink.table.expressions.ResolvedExpression.asSerializableString(ResolvedExpression.java:51) E at org.apache.flink.table.api.TableSchema.lambda$fromResolvedSchema$13(TableSchema.java:455) E at java.util.Collections$SingletonList.forEach(Collections.java:4824) E at org.apache.flink.table.api.TableSchema.fromResolvedSchema(TableSchema.java:451) E at org.apache.flink.table.api.Table.getSchema(Table.java:101) E at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) E at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) E at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) E at java.lang.reflect.Method.invoke(Method.java:498) E at org.apache.flink.api.python.shaded.py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244) E at org.apache.flink.api.python.shaded.py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357) E at org.apache.flink.api.python.shaded.py4j.Gateway.invoke(Gateway.java:282) E at org.apache.flink.api.python.shaded.py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132) E at org.apache.flink.api.python.shaded.py4j.commands.CallCommand.execute(CallCommand.java:79) E at org.apache.flink.api.python.shaded.py4j.GatewayConnection.run(GatewayConnection.java:238) E at java.lang.Thread.run(Thread.java:748) {code} > Table.getSchema fails on table with watermark > --------------------------------------------- > > Key: FLINK-28528 > URL: https://issues.apache.org/jira/browse/FLINK-28528 > Project: Flink > Issue Type: Bug > Components: API / Python > Affects Versions: 1.15.1 > Reporter: Xuannan Su > Priority: Major > > The bug can be reproduced with the following test. The test can pass if we > use the commented way to define the watermark. > {code:python} > def test_flink_2(self): > env = StreamExecutionEnvironment.get_execution_environment() > t_env = StreamTableEnvironment.create(env) > table = t_env.from_descriptor( > TableDescriptor.for_connector("filesystem") > .schema( > Schema.new_builder() > .column("name", DataTypes.STRING()) > .column("cost", DataTypes.INT()) > .column("distance", DataTypes.INT()) > .column("time", DataTypes.TIMESTAMP(3)) > .watermark("time", expr.col("time") - expr.lit(60).seconds) > # .watermark("time", "`time` - INTERVAL '60' SECOND") > .build() > ) > .format("csv") > .option("path", "./input.csv") > .build() > ) > print(table.get_schema()) > {code} > It causes the following exception > {code:none} > E pyflink.util.exceptions.TableException: > org.apache.flink.table.api.TableException: Expression 'minus(time, 60000)' is > not string serializable. Currently, only expressions that originated from a > SQL expression have a well-defined string representation. > E at > org.apache.flink.table.expressions.ResolvedExpression.asSerializableString(ResolvedExpression.java:51) > E at > org.apache.flink.table.api.TableSchema.lambda$fromResolvedSchema$13(TableSchema.java:455) > E at > java.util.Collections$SingletonList.forEach(Collections.java:4824) > E at > org.apache.flink.table.api.TableSchema.fromResolvedSchema(TableSchema.java:451) > E at org.apache.flink.table.api.Table.getSchema(Table.java:101) > E at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) > E at > sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) > E at > sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) > E at java.lang.reflect.Method.invoke(Method.java:498) > E at > org.apache.flink.api.python.shaded.py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244) > E at > org.apache.flink.api.python.shaded.py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357) > E at > org.apache.flink.api.python.shaded.py4j.Gateway.invoke(Gateway.java:282) > E at > org.apache.flink.api.python.shaded.py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132) > E at > org.apache.flink.api.python.shaded.py4j.commands.CallCommand.execute(CallCommand.java:79) > E at > org.apache.flink.api.python.shaded.py4j.GatewayConnection.run(GatewayConnection.java:238) > E at java.lang.Thread.run(Thread.java:748) > {code} -- This message was sent by Atlassian Jira (v8.20.10#820010)