[ https://issues.apache.org/jira/browse/NIFI-4024?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16080370#comment-16080370 ]
ASF GitHub Bot commented on NIFI-4024: -------------------------------------- Github user bbende commented on a diff in the pull request: https://github.com/apache/nifi/pull/1961#discussion_r126431547 --- Diff: nifi-nar-bundles/nifi-hbase-bundle/nifi-hbase-processors/src/main/java/org/apache/nifi/hbase/PutHBaseRecord.java --- @@ -49,10 +51,12 @@ @InputRequirement(InputRequirement.Requirement.INPUT_REQUIRED) @Tags({"hadoop", "hbase", "put", "record"}) @CapabilityDescription("Adds rows to HBase based on the contents of a flowfile using a configured record reader.") +@ReadsAttribute(attribute = "restart.index", description = "Reads restart.index when it needs to replay part of a record set that did not get into HBase.") +@WritesAttribute(attribute = "restart.index", description = "Writes restart.index when a batch fails to be insert into HBase") public class PutHBaseRecord extends AbstractPutHBase { protected static final PropertyDescriptor ROW_FIELD_NAME = new PropertyDescriptor.Builder() - .name("Row Identifier Field Name") + .name("Row Identifier Record Path") --- End diff -- I think I confused things here with my previous comment... If we made this property a record path then it requires more code later on to evaluate the record path against the record, which we aren't doing right now, so lets put this back to "Row Identifier Field Name". Sorry for the confusion. > Create EvaluateRecordPath processor > ----------------------------------- > > Key: NIFI-4024 > URL: https://issues.apache.org/jira/browse/NIFI-4024 > Project: Apache NiFi > Issue Type: New Feature > Reporter: Steve Champagne > Priority: Minor > > With the new RecordPath DSL, it would be nice if there was a processor that > could pull fields into attributes of the flowfile based on a RecordPath. This > would be similar to the EvaluateJsonPath processor that currently exists, > except it could be used to pull fields from arbitrary record formats. My > current use case for it would be pulling fields out of Avro records while > skipping the steps of having to convert Avro to JSON, evaluate JsonPath, and > then converting back to Avro. -- This message was sent by Atlassian JIRA (v6.4.14#64029)