[jira] [Commented] (NIFI-5552) CSVReader Can't Derive Schema from Quoted Headers

["ASF GitHub Bot (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/NIFI-5552?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16643496#comment-16643496 ] 

ASF GitHub Bot commented on NIFI-5552:
--------------------------------------

Github user mattyb149 commented on the issue:

    https://github.com/apache/nifi/pull/2966
  
    I'm a little leery of automatically normalizing the names. In [NIFI-4612](https://issues.apache.org/jira/browse/NIFI-4612) we added the ability to disable name validation for the AvroSchemaRegistry, so there could be Avro-illegal field names but they'd be preserved. I think that came up for things like PutDatabaseRecord where the field names needed to be preserved but were not Avro-valid. If we automatically normalize them, there might be unintended (and undesired) consequence to things like PutParquet and PutDatabaseRecord. 
    
    Perhaps instead of (or in addition to) normalization, we provide the option to "disable schema validation". This isn't technically Avro-specific, but would be used while we are so coupled to Avro schemas in the meantime?


> CSVReader Can't Derive Schema from Quoted Headers
> -------------------------------------------------
>
>                 Key: NIFI-5552
>                 URL: https://issues.apache.org/jira/browse/NIFI-5552
>             Project: Apache NiFi
>          Issue Type: Bug
>          Components: Extensions
>    Affects Versions: 1.7.1
>            Reporter: Shawn Weeks
>            Assignee: Pierre Villard
>            Priority: Minor
>
> When deriving the schema from a CSV File Header NiFi is unable to generate a valid schema if the Header Columns are Double Quoted even though the CSVReader is set to handle quotes. Using the nile.csv sample file from https://people.sc.fsu.edu/~jburkardt/data/csv/csv.html results in an Illegal initial character exception in the Avro Schema generator. In this specific case the header did not contain any spaces or special characters though it was case sensitive.
> {code:java}
> org.apache.avro.SchemaParseException: Illegal initial character: "Flood"
> at org.apache.avro.Schema.validateName(Schema.java:1147)
> at org.apache.avro.Schema.access$200(Schema.java:81)
> at org.apache.avro.Schema$Field.<init>(Schema.java:403)
> at org.apache.avro.Schema$Field.<init>(Schema.java:423)
> at org.apache.avro.Schema$Field.<init>(Schema.java:415)
> at org.apache.nifi.avro.AvroTypeUtil.buildAvroField(AvroTypeUtil.java:123)
> at org.apache.nifi.avro.AvroTypeUtil.buildAvroSchema(AvroTypeUtil.java:114)
> at org.apache.nifi.avro.AvroTypeUtil.extractAvroSchema(AvroTypeUtil.java:94)
> at org.apache.nifi.schema.access.WriteAvroSchemaAttributeStrategy.getAttributes(WriteAvroSchemaAttributeStrategy.java:58)
> at org.apache.nifi.json.WriteJsonResult.writeRecord(WriteJsonResult.java:137)
> at org.apache.nifi.serialization.AbstractRecordSetWriter.write(AbstractRecordSetWriter.java:59)
> at org.apache.nifi.processors.standard.AbstractRecordProcessor$1.process(AbstractRecordProcessor.java:122)
> at org.apache.nifi.controller.repository.StandardProcessSession.write(StandardProcessSession.java:2885)
> at org.apache.nifi.processors.standard.AbstractRecordProcessor.onTrigger(AbstractRecordProcessor.java:109)
> at org.apache.nifi.processor.AbstractProcessor.onTrigger(AbstractProcessor.java:27)
> at org.apache.nifi.controller.StandardProcessorNode.onTrigger(StandardProcessorNode.java:1165)
> at org.apache.nifi.controller.tasks.ConnectableTask.invoke(ConnectableTask.java:203)
> at org.apache.nifi.controller.scheduling.TimerDrivenSchedulingAgent$1.run(TimerDrivenSchedulingAgent.java:117)
> at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
> at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
> at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
> at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748){code}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)