imback82 commented on a change in pull request #29328:
URL: https://github.com/apache/spark/pull/29328#discussion_r468330253
##########
File path:
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/csv/CSVDataSource.scala
##########
@@ -155,7 +155,7 @@ object TextInputCSVDataSource extends CSVDataSource {
sparkSession,
paths = paths,
className = classOf[TextFileFormat].getName,
- options = options.parameters
+ options = options.parameters - "path"
Review comment:
Yes, without this, few tests that involve "inferring" fail.
Basically, when you do `spark.read.csv("/tmp")`, "path" option is populated
since there is only one path parameter in `load()`. Now, when `infer` is
called, `inputPaths: Seq[FileStatus]` is a result of listing files and there is
the "path" option that is being carried over, so `infer` will try to read more
data than needed.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
