[ https://issues.apache.org/jira/browse/FLINK-19949?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17225244#comment-17225244 ]
Danny Chen commented on FLINK-19949: ------------------------------------ Yes, but the unicode has poor usability. As a user, there would bring in confusion when the same DDL works in the Java code but fails when written in a file. And the unescape way does not cause any regression. > Unescape CSV format line delimiter character > -------------------------------------------- > > Key: FLINK-19949 > URL: https://issues.apache.org/jira/browse/FLINK-19949 > Project: Flink > Issue Type: Improvement > Components: Formats (JSON, Avro, Parquet, ORC, SequenceFile) > Affects Versions: 1.11.2 > Reporter: Danny Chen > Priority: Major > Fix For: 1.12.0 > > > We should unescape the line delimiter characters first because the DDL can be > read from a file. So that the new line "\n" in the DDL options was recognized > as 2 characters. > While what user want is actually the invisible new line character. > {code:sql} > create table t1( > ... > ) with ( > 'format' = 'csv', > 'csv.line-delimiter' = '\n' > ... > ) > {code} -- This message was sent by Atlassian Jira (v8.3.4#803005)