Repository navigation
Conversation
d38f5ba to
cbccfdb
Compare
|
@jorisvandenbossche @pitrou I would appreciate your review on this PR when you get a chance. |
2285758 to
4542cc7
Compare
|
@edponce Can you rebase to try and get a clean CI build? |
There was a problem hiding this comment.
Not sure why you're changing this test?
There was a problem hiding this comment.
In C++, a ChunkedArray has to have all Arrays of the same type, but the Python API allowed coercion of None to the first inferred type
["a", None] --> StringType
[None] --> NullType
[] --> NullTypeMaybe we should allow such coercion in C++ as well.
There was a problem hiding this comment.
Hmm, this could be a slight API break. @jorisvandenbossche What do you think?
There was a problem hiding this comment.
Let me double check carefully the current C++ and Python behavior before jumping to conclusions.
There was a problem hiding this comment.
For the record, are the error conditions checked by the test suite?
c2fc63f to
3e6f309
Compare
|
@milesgranger Would you like to take this up perhaps? |
…13950) Supersedes and will close #12096 [ARROW-15277](https://issues.apache.org/jira/browse/ARROW-15277) Lead-authored-by: Miles Granger <[email protected]> Co-authored-by: Eduardo Ponce <[email protected]> Co-authored-by: Antoine Pitrou <[email protected]> Signed-off-by: Antoine Pitrou <[email protected]>
…pache#13950) Supersedes and will close apache#12096 [ARROW-15277](https://issues.apache.org/jira/browse/ARROW-15277) Lead-authored-by: Miles Granger <[email protected]> Co-authored-by: Eduardo Ponce <[email protected]> Co-authored-by: Antoine Pitrou <[email protected]> Signed-off-by: Antoine Pitrou <[email protected]>
Replace Python explicit type/size checks for ChunkedArray with C++
ChunkedArray::Make().