[
https://issues.apache.org/jira/browse/ARROW-15051?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Zixi Wang updated ARROW-15051:
------------------------------
Description:
When the record batch schema field's parameter of nullable is set to false, it
supposes to make the field not null, however, it can still have null values.
Code to reproduce:
var memoryAllocator = new NativeMemoryAllocator(alignment: 64);
var stream1 = File.OpenWrite(@"c:\temp\data\test.arrow");
Schema schema = new Schema
(new List<Field>
{ new Field("Column A", StringType.Default, false) }
,
default
);
List<string> rows = new List<string>();
rows.Add("11");
rows.Add("22");
rows.Add(null);
ArrowFileWriter writer = new ArrowFileWriter(stream1, schema);
var recordBatch1 = new RecordBatch.Builder(memoryAllocator)
.Append("Column A", false, col => col.String(array =>
array.AppendRange(rows)))
.Build();
writer.WriteRecordBatch(recordBatch1);
writer.WriteEnd();
stream1.Close();
was:
When the record batch schema field's parameter of nullable is set to false, it
supposes to make the field not null, however, it can still have null values.
Code to produce:
var memoryAllocator = new NativeMemoryAllocator(alignment: 64);
var stream1 = File.OpenWrite(@"c:\temp\data\test.arrow");
Schema schema = new Schema
(new List<Field>
{
new Field("Column A", StringType.Default, false)
},
default
);
List<string> rows = new List<string>();
rows.Add("11");
rows.Add("22");
rows.Add(null);
ArrowFileWriter writer = new ArrowFileWriter(stream1, schema);
var recordBatch1 = new RecordBatch.Builder(memoryAllocator)
.Append("Column A", false, col => col.String(array =>
array.AppendRange(rows)))
.Build();
writer.WriteRecordBatch(recordBatch1);
writer.WriteEnd();
stream1.Close();
> [C#] Not nullable field can have null values
> --------------------------------------------
>
> Key: ARROW-15051
> URL: https://issues.apache.org/jira/browse/ARROW-15051
> Project: Apache Arrow
> Issue Type: Bug
> Components: C#
> Affects Versions: 6.0.1
> Reporter: Zixi Wang
> Priority: Minor
>
> When the record batch schema field's parameter of nullable is set to false,
> it supposes to make the field not null, however, it can still have null
> values.
>
> Code to reproduce:
> var memoryAllocator = new NativeMemoryAllocator(alignment: 64);
> var stream1 = File.OpenWrite(@"c:\temp\data\test.arrow");
>
> Schema schema = new Schema
> (new List<Field>
>
> { new Field("Column A", StringType.Default, false)
> }
> ,
> default
> );
> List<string> rows = new List<string>();
> rows.Add("11");
> rows.Add("22");
> rows.Add(null);
> ArrowFileWriter writer = new ArrowFileWriter(stream1, schema);
> var recordBatch1 = new RecordBatch.Builder(memoryAllocator)
> .Append("Column A", false, col => col.String(array =>
> array.AppendRange(rows)))
> .Build();
> writer.WriteRecordBatch(recordBatch1);
> writer.WriteEnd();
> stream1.Close();
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
