User Defined Type unpacking not working

[Resiak]

Hi,

I've just created an UDT to use into my application but it seems there is something wrong when retrieving the data.
The UDT has a very simple structure:

CREATE TYPE mytype (a int, b text);

I've then run the insert command and I can see (using Datastax DevCenter) that the data in the table looks right, infact in the column I have something like {a: 125, b: '3'}.
When unpacking though I get an array with value string(4) "}"

I've also tried to use my UDT into a map, but I've got a similar result.
I'm using Cassandra 2.1.0

[LarsFronius]

Hey @Resiak - can you send me a bit of code, including the inserts you are doing so I can write a regression test for this and work on it?

[steambao]

Here is an example which I'm having trouble with.

If there are two or more results in the set (like the first insert record), I get a null result.
If there is only one one result in the set (like the second insert record), I get only the first field in the User Defined Type like below:

public 'phone_numbers' =>
array (size=1)
0 => string 'Home' (length=4)

Are User defined types supported? Has anyone had success with them?

---

-- User Defined Type
CREATE TYPE phone_number (
type text,
number text
);

-- User Table
CREATE TABLE IF NOT EXISTS user(
username text,
password text,
phone_numbers set<frozen<phone_number>>,
PRIMARY KEY (username)
);

-- Example Insert Data
INSERT INTO user(username, password, phone_numbers)
VALUES ('test', 'test', {
{
type: 'Home',
number: '000111222'
},
{
type: 'Work',
number: '1111222333'
}
});
INSERT INTO user(username, password, phone_numbers)
VALUES ('test2', 'test2', {
{
type: 'Office',
number: '333222111'
}
});