Github user HyukjinKwon commented on a diff in the pull request:
https://github.com/apache/spark/pull/18304#discussion_r122141063
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/csv/CSVOptions.scala
---
@@ -90,6 +90,7 @@ class CSVOptions(
val quote = getChar("quote", '\"')
val escape = getChar("escape", '\\')
val comment = getChar("comment", '\u0000')
+ val lineSeparator = parameters.getOrElse("lineSeparator", "\n")
--- End diff --
Yea, that's why I haven't proposed this change so far (for both hard corded
`\n` and configurable line separator).
For the hard corded `\n`, IMHO, reviving hard corded `\n` is probably fine
for now as consistent with JSON/TEXT datasources because It uses, after the
change I described above, OS-dependent newlline by default in Univocity -
https://github.com/uniVocity/univocity-parsers/blob/e34a149077147de7cafcafbbaf8a4310e6297584/src/main/java/com/univocity/parsers/common/Format.java#L78,
which is machine-dependent. If we are worried of changing this, we could use
`System.getProperty("line.separator")` or just not set via `Option`. This
probably does not block this PR.
For the configurable line separator, we are not able to read it back when
`wholeFile` (or `miltiLine` now) option is disabled (it's default and this
should be primary) because we are dependent of `LineRecordReader` which only
deals with `\n` or `\r\n`. We really should be able to read it back.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
