CODEWITHSUNDEEP
hadoopexpressionapache-nifi
lengthy_preamble
lengthy_preamble

Reputation: 434

Nifi: capturing the middle section of a filename

I have a number of files going to HDFS and the naming convention is something like this:

I want to use the segment between the underscores as a variable to make the HDFS path, so it will look something like this:

/my/hdfs/directory/sponsor/2019/

I found a way to do this in two steps, but I think there must be a way to do it in one. For the first step, I have an "Update Attribute" processor that creates a variable "file_src" with the following value:

${filename:substringAfter('_')}

So now it sees the filename as being "beneficiary_20190820", etc. After this, I have another "Update Attribute" processor with a variable named "dest" with the following value:

${file_src:substringBefore('_'):toLower()}

so now my hdfs directory can be something like this:

/my/hdfs/directory/${dest}/2019

It works, but it feels clunky. Is there a way to everything in one step? I feel like maybe these expressions could be nested or something. Thanks in advance for any help.

Upvotes: 0

Views: 723

Answers (1)

daggett
daggett

Reputation: 28564

put everything into one expression

${filename:substringAfter('_'):substringBefore('_'):toLower()}

you even could use this expression directly in hdfs directory expression without UpdateAttribute:

/my/hdfs/directory/${filename:substringAfter('_'):substringBefore('_'):toLower()}/2019

Upvotes: 2

Related Questions