Daniel Davies created SPARK-39937:
-------------------------------------
Summary: Substring
Key: SPARK-39937
URL: https://issues.apache.org/jira/browse/SPARK-39937
Project: Spark
Issue Type: New Feature
Components: Spark Core
Affects Versions: 3.2.2
Reporter: Daniel Davies
Hello,
We hit a potential limitation of the substring function today that we think
might be useful to contribute back to spark, relating to that fact that
[substring
|https://github.com/apache/spark/blob/775b5ced70c87d6a8709f44eb4a5de48c286a51d/sql/core/src/main/scala/org/apache/spark/sql/functions.scala#L2952]
only takes integers for it's pos and len parameters.
Consider the case where you have a list of ID's, and you need to find
substrings in these IDs that are dependent on another column (for example, a
column expressing the kind of product that the ID represents).
e.g.
{{Table:}}
{{}}
{{}}
ID | Type
+-----------+--------+
ABCD12345 | Water
+-----------+--------+
ABCD12 | Soda
+-----------+--------+
{{}}
Desired:
ID | Type
+-----------+--------+
12345 | Water
+-----------+--------+
12 | Soda
+-----------+--------+
At the moment, you can only extract substrings of fixed position and length-
but if IDs have a variable suffix, it may be desirable to run a substring of
different lengths based on the 'Type' column.
In this case, it's fairly straightforward to solve with a chain of .when
statements; but if there are thousands of types, it may not be as easy.
This seems fairly niche, but I'd be happy to give the implementation a go if
this is something that the community would find useful/ has been seen before.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
