Tyler Hobbs created CASSANDRA-7792:
--------------------------------------
Summary: Fix cqlsh null handling in COPY FROM
Key: CASSANDRA-7792
URL: https://issues.apache.org/jira/browse/CASSANDRA-7792
Project: Cassandra
Issue Type: Bug
Components: Tools
Reporter: Tyler Hobbs
Assignee: Tyler Hobbs
Priority: Minor
Fix For: 2.1.0
In {{create_insert_statement()}}, we have the following logic:
{noformat}
if value != nullval:
if cqltype in ('ascii', 'text', 'timestamp', 'inet'):
rowmap[name] = protect_value(value)
else:
rowmap[name] = value
elif name in clustering_key_columns and not type.empty_binary_ok:
rowmap[name] = 'blobAs%s(0x)' % cqltype.title()
else:
rowmap[name] = 'null'
{noformat}
The first problem is that we're inserting empty binaries when the type should
_not_ have empty binaries.
However, it could also be considered incorrect that we insert empty binaries
for clustering columns when we have something that matches the "null value".
This is to avoid inserting nulls for clustering columns. Of course, users may
actually want to insert empty strings for columns, but I think that if they
want that behavior, changing the null value is the correct way to do that.
I propose changing the "elif" statement to something like:
{noformat}
elif name in primary_key_columns:
raise Exception("Nulls are not allowed in primary key columns")
{noformat}
I set the fix version for this to 2.1.0. If we're not comfortable making this
entire change in 2.1.0, we should at least remove the "not" in the elif and
make the other changes for 2.1.1
--
This message was sent by Atlassian JIRA
(v6.2#6252)
