Im trying to write a stored procedure in which I can upsert a row even if one of the values in the key is null. I've read the documentation and found that Postgres doesn't work with comparing equality of null values.
I've also read other forum posts and noticed that what I want to achieve can be done through a partial index. I'm able to successfully get a constraint violation, however my "on conflict" never gets hit when i pass in a value that has a null birthday.
I want to be able to pass in a null birthday and update an ID for a person even if their birthday is null.
(
id bigint not null,
name text,
birthday date
);
I create an index and partial index so that it allows birthday to be null
CREATE UNIQUE INDEX name_birthday_key on people (name, birthday);
CREATE UNIQUE INDEX birthday_null_key on people (name) where birthday is null;
create or replace procedure store_person(_identifier bigint, _name character varying, _birthday date)
language plpgsql
as
$$
begin
insert into people (
id, name, birthday
)
values (
_identifier, _name, _birthday
)
on conflict (name, birthday)
do update
set
id = _identifier
where people.birthday = _date and people.name = _name;
end
$$;
if I run:
call public.store_person(1, 'Bob', '1955-01-09')
call public.store_person(2, 'Bob', '1955-01-09')
i successfully see that the only row in the DB is Bob with an ID of 2.
however, if i run
call public.store_person(3, 'Joe', null)
call public.store_person(4, 'Joe', null)
the only row i get is ID 3. the second insert for ID 4 never updates the existing row. I do get a violation error but the "on conflict" update never is hit.
can someone point me in the right direction of how to do this?
The CONFLICT doesn't match because NULLis not equal to NULL. This is not a PostgreSQL thing, it's defined in SQL standard.
Use something like COALESCE(birthday, '0001-01-01') when inserting your data, that will match; and remove the partial index.
Your code has an error, in DO UPDATE...WHERE: there's nothing named _date, should be _birthday.
Related
Given the following (simplified) schema:
CREATE TABLE period (
id UUID NOT NULL DEFAULT uuid_generate_v4(),
name TEXT
);
CREATE TABLE course (
id UUID NOT NULL DEFAULT uuid_generate_v4(),
name TEXT
);
CREATE TABLE registration (
id UUID NOT NULL DEFAULT uuid_generate_v4(),
period_id UUID NOT NULL REFERENCES period(id),
course_id UUID NOT NULL REFERENCES course(id),
inserted_at timestamptz NOT NULL DEFAULT now()
);
I now want to add a new column client_ref, which identifies a registration unique within a period, but consists of only a 4-character string. I want to use pg_hashids - which requires a unique integer input - to base the column value on.
I was thinking of setting up a trigger on the registration table that runs on inserting a new row. I came up with the following:
CREATE OR REPLACE FUNCTION set_client_ref()
RETURNS trigger
LANGUAGE plpgsql
AS $function$
DECLARE
next_row_number integer;
BEGIN
WITH rank AS (
SELECT
period.id AS period_id,
row_number() OVER (PARTITION BY period.id ORDER BY registration.inserted_at)
FROM
registration
JOIN period ON registration.period_id = period.id ORDER BY
period.id,
row_number
)
SELECT
COALESCE(rank.row_number, 0) + 1 INTO next_row_number
FROM
period
LEFT JOIN rank ON (rank.period_id = period.id)
WHERE
period.id = NEW.period_id
ORDER BY
rank.row_number DESC
LIMIT 1;
NEW.client_ref = id_encode (next_row_number);
RETURN NEW;
END
$function$
;
The trigger is set-up like: CREATE TRIGGER set_client_ref BEFORE INSERT ON registration FOR EACH ROW EXECUTE FUNCTION set_client_ref();
This works as expected when inserting a single row to registration, but if I insert multiple within one statement, they end up having the same client_ref. I can reason about why this happens (the rows don't know about each other's existence, so they assume they're all just next in line when retrieving their row_order), but I am not sure what a way is to prevent this. I tried setting up the trigger as an AFTER trigger, but it resulted in the same (duplicated) behaviour.
What would be a better way to get the lowest possible, unique integer for the rows to be inserted (to base the hash function on) that also works when inserting multiple rows?
I've being trying to update a row with a value from another table but I keep getting the same error:
table name "chats" specified more than once
I'm trying to insert a row into the messages table, and then use that timestamp (message_timestamp) to update the last_updated field in the chats table.
Any help would be appreciated! Been stuck on this for a day now:
WITH result AS
(INSERT INTO messages (user_id, chat_id, message_timestamp, users_read, message_text)
VALUES ($1, $2, NOW(), '{}', $3) RETURNING message_timestamp, chat_id)
UPDATE chats SET chats.last_updated=result.message_timestamp FROM result, chats WHERE chats.id=result.chat_id;
Edit:
On the other hand, removing chats from my FROM clause as so:
UPDATE chats SET chats.last_updated=result.message_timestamp FROM result WHERE chats.id=result.chat_id;
results in a different error:
column "chats" of relation "chats" does not exist
Which is weird considering I never call chats.chats
Edit 2: Create statement for the chat table:
CREATE TABLE IF NOT EXISTS chats (
id UUID PRIMARY KEY DEFAULT uuid_generate_v4(),
chat_name VARCHAR,
last_message TIMESTAMP NOT NULL
);
From UPDATE:
Do not include the table's name in the specification of a target
column
So the column last_updated that you want to update should not be qualified with the table's name like chats.last_updated:
UPDATE chats
SET last_updated = result.message_timestamp
FROM result
WHERE chats.id = result.chat_id;
See a simplified demo.
I have a demo table
CREATE TABLE items (
id SERIAL primary key,
user_id integer,
name character varying,
created timestamp with time zone default now()
);
And I want a single query to run and first insert data, then return primary key using returning id and then update the same table with the returned id.
INSERT INTO items (name) values ('pen') RETURNING id as idd
update items set user_id=(select idd) where id=(select idd)
but the above command doesn't work and throws syntax error.
Any help will be appriciated.
You can do that right within the INSERT statement:
INSERT INTO items
(name, user_id)
values
('pen', currval(pg_get_serial_sequence('items','id')));
Online example
You can try this way also :
create temp table insert_item as
with insert_item_cte as (
INSERT INTO items (name)
values ('pen') returning id
)
select id from insert_item_cte;
update items set user_id = items.id
from insert_item ii
where ii.id = items.id;
Online Demo
I have a database with companies and their products, I want for each
company to have a separate product id sequence.
I know that postgresql can't do this, the only way is to have a separate sequence for each company but this is cumbersome.
I thought about a solution to have a separate table to hold the sequences
CREATE TABLE "sequence"
(
"table" character varying(25),
company_id integer DEFAULT 0,
"value" integer
)
"table" will be holt the table name for the sequence, such as products, categories etc.
and value will hold the actual sequence data that will be used for product_id on inserts
I will use UPDATE ... RETURNING value; to get a product id
I was wondering is this solution efficient?
With row level locking, only users of same company adding rows in the same table will have to wait to get a lock and I think that reduces race condition problems.
Is there a better way to solve this problem?
I don't want to use a sequence for products table for all companies because the difference between product id's will be to big, I want to keep it simple for the users.
You could just embed a counter in your companies table:
CREATE TABLE companies (
id SERIAL PRIMARY KEY,
name TEXT,
product_id INT DEFAULT 0
);
CREATE TABLE products (
company INT REFERENCES companies(id),
product_id INT,
PRIMARY KEY (company, product_id),
name TEXT
);
INSERT INTO companies (id, name) VALUES (1, 'Acme Corporation');
INSERT INTO companies (id, name) VALUES (2, 'Umbrella Corporation');
Then, use UPDATE ... RETURNING to get the next product ID for a given company:
> INSERT INTO products VALUES (1, (UPDATE companies SET product_id = product_id+1 WHERE id=$1 RETURNING product_id), 'Anvil');
ERROR: syntax error at or near "companies"
LINE 1: INSERT INTO products VALUES (1, (UPDATE companies SET produc...
^
Oh noes! It seems you can't (as of PostgreSQL 9.1devel) use UPDATE ... RETURNING as a subquery.
The good news is, it's not a problem! Just create a stored procedure that does the increment/return part:
CREATE FUNCTION next_product_id(company INT) RETURNS INT
AS $$
UPDATE companies SET product_id = product_id+1 WHERE id=$1 RETURNING product_id
$$ LANGUAGE 'sql';
Now insertion is a piece of cake:
INSERT INTO products VALUES (1, next_product_id(1), 'Anvil');
INSERT INTO products VALUES (1, next_product_id(1), 'Dynamite');
INSERT INTO products VALUES (2, next_product_id(2), 'Umbrella');
INSERT INTO products VALUES (1, next_product_id(1), 'Explosive tennis balls');
Be sure to use the same company ID in both the product value and the argument to next_product_id(company INT).
Depending on how many companies you have, you could create a sequence for each company. Query it by a function which is set as a default on your product_id column.
Alternatively this function could simply do a SELECT FOR UPDATE and update the values of your table. Should be pretty performant I think.
To automatically add a column in a second table to tie it to the first table via a unique index, I have a rule such as follows:
CREATE OR REPLACE RULE auto_insert AS ON INSERT TO user DO ALSO
INSERT INTO lastlogin (id) VALUES (NEW.userid);
This works fine if user.userid is an integer. However, if it is a sequence (e.g., type serial or bigserial), what is inserted into table lastlogin is the next sequence id. So this command:
INSERT INTO user (username) VALUES ('john');
would insert column [1, 'john', ...] into user but column [2, ...] into lastlogin. The following 2 workarounds do work except that the second one consumes twice as many serials since the sequence is still auto-incrementing:
CREATE OR REPLACE RULE auto_insert AS ON INSERT TO user DO ALSO
INSERT INTO lastlogin (id) VALUES (lastval());
CREATE OR REPLACE RULE auto_insert AS ON INSERT TO user DO ALSO
INSERT INTO lastlogin (id) VALUES (NEW.userid-1);
Unfortunately, the workarounds do not work if I'm inserting multiple rows:
INSERT INTO user (username) VALUES ('john'), ('mary');
The first workaround would use the same id, and the second workaround is all kind of screw-up.
Is it possible to do this via postgresql rules or should I simply do the 2nd insertion into lastlogin myself or use a row trigger? Actually, I think the row trigger would also auto-increment the sequence when I access NEW.userid.
Forget rules altogether. They're bad.
Triggers are way better for you. And in 99% of cases when someone thinks he needs a rule. Try this:
create table users (
userid serial primary key,
username text
);
create table lastlogin (
userid int primary key references users(userid),
lastlogin_time timestamp with time zone
);
create or replace function lastlogin_create_id() returns trigger as $$
begin
insert into lastlogin (userid) values (NEW.userid);
return NEW;
end;
$$
language plpgsql volatile;
create trigger lastlogin_create_id
after insert on users for each row execute procedure lastlogin_create_id();
Then:
insert into users (username) values ('foo'),('bar');
select * from users;
userid | username
--------+----------
1 | foo
2 | bar
(2 rows)
select * from lastlogin;
userid | lastlogin_time
--------+----------------
1 |
2 |
(2 rows)