Github user kevinyu98 commented on the issue:
https://github.com/apache/spark/pull/21285
@HyukjinKwon thanks for reviewing this pr. I didn't notice that pr until
you point out. If we plan to support wildcard in the LOAD DATA command, then we
can close this PR.
But with his current code, the problem reported by this JIRA still exists,
because for the non-local case, the Path will be truncate after `val srcPath =
new Path(loadPath)`. I download his code, and it still have the same issue as
this pr reported.
I create text1.txt on my local machine, then run LOAD DATA
`load data inpath '/Users/qianyangyu/IdeaProjects/spark/??xt1.txt' into
table foo1;' ` successful, but it didn't load data into the table
`load data inpath '/Users/qianyangyu/IdeaProjects/spark/t?xt1.txt' into
table foo1;` failed
`spark-sql> load data inpath
'/Users/qianyangyu/IdeaProjects/spark/??xt1.txt' into table foo1;
Time taken: 0.112 seconds
spark-sql> select * from foo1;
18/05/09 23:23:51 INFO DAGScheduler: Job 1 finished: processCmd at
CliDriver.java:376, took 0.000029 s
Time taken: 0.056 seconds
18/05/09 23:23:51 INFO SparkSQLCLIDriver: Time taken: 0.056 seconds
spark-sql> load data inpath
'/Users/qianyangyu/IdeaProjects/spark/t?xt1.txt' into table foo1;
Error in query: LOAD DATA input path does not exist:
/Users/qianyangyu/IdeaProjects/spark/t?xt1.txt;
`
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
