[
https://issues.apache.org/jira/browse/CASSANDRA-10854?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15069637#comment-15069637
]
Stefania commented on CASSANDRA-10854:
--------------------------------------
We need to keep the current approach in case users want to insert an empty
string in one of the primary key components, for example a clustering key. It
works as follows:
# if the value is not equal to the NULL option value then we parse it and in
this case we get {{invalid literal for int()}} because the key is an integer
but for a text key it would have been a valid value
# If the value is equal to the NULL option value (by default an empty string)
then we insert {{None}} for any column except for primary keys in which case we
complain with {{Cannot insert null value for primary key column ...}}.
# The reason the default NULL option is an empty string is because the csv
parser returns an empty string for missing values.
> cqlsh COPY FROM csv having line with more than one consecutive ',' delimiter
> is throwing 'list index out of range'
> --------------------------------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-10854
> URL: https://issues.apache.org/jira/browse/CASSANDRA-10854
> Project: Cassandra
> Issue Type: Bug
> Components: Tools
> Environment: cqlsh 5.0.1 | Cassandra 2.1.11.969 | DSE 4.8.3 | CQL
> spec 3.2.1
> Reporter: Puspendu Banerjee
> Assignee: Stefania
> Priority: Minor
>
> cqlsh COPY FROM csv having line with more than one consecutive ',' delimiter
> is throwing 'list index out of range'
> Steps to re-produce:
> {code}
> CREATE TABLE tracks_by_album (
> album_title TEXT,
> album_year INT,
> performer TEXT STATIC,
> album_genre TEXT STATIC,
> track_number INT,
> track_title TEXT,
> PRIMARY KEY ((album_title, album_year), track_number)
> );
> {code}
> Create a file: tracks_by_album.csv having following 2 lines :
> {code}
> album,year,performer,genre,number,title
> a,2015,b c d,e f g,,
> {code}
> {code}
> cqlsh> COPY music.tracks_by_album
> (album_title, album_year, performer, album_genre, track_number,
> track_title)
> FROM '~/tracks_by_album.csv'
> WITH HEADER = 'true';
> Error :
> Starting copy of music.tracks_by_album with columns ['album_title',
> 'album_year', 'performer', 'album_genre', 'track_number', 'track_title'].
> list index out of range
> Aborting import at record #1. Previously inserted records are still present,
> and some records after that may be present as well.
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
