Open al3xandru opened 8 years ago
The current version does allow you to use case-sensitive keyspace, table, and column names. For example, you can do -schema 'myks."MyTable"(col1, "Col2")' Does that cover the use case you are looking to address?
That gets unnecessarily cumbersome very fast.
The double quotes standard around upper case keyspaces and columns is a cassandra standard. You have to do it in cqlsh, at the driver, etc.
I see your point about it being a bit cumbersome however and am having trouble thinking of downsides of the change. What happens if there's a table called myTable and another table called mytable. Would this be able to tell the difference?
Based on this https://docs.datastax.com/en/cql/3.3/cql/cql_reference/ucase-lcase_r.html I think requiring double quotes to get case sensitivity is necessary (even if a bit cumbersome).
With JSON, this becomes a bigger issue because now we're requiring the column names to come escaped in the JSON file. Nobody will have this. I think it's time to re-evaluate
Here's a new PR to deal with the JSON issue.
Do all the necessary escaping internally so schema parameter can be treated as case sensitive