The documentation for SQL-based UDFs is sparse. I'm wondering if it's possible to write a full-fledged SELECT clause, using the UDF parameters in the query. So in effect, each invocation of the UDF would result in a subquery.
Contrived example:
CREATE TEMP FUNCTION foo(bar STRING) AS (
SELECT * FROM `example.latest` WHERE thing = bar
);
SELECT foo('abc')
BigQuery gives the error "Syntax error: Unexpected keyword SELECT; failed to parse CREATE [TEMP] FUNCTION statement" so I assume it's not possible, but would love to get confirmation.
SELECT is supported in general but unfortunatelly you cannot reference table(s) in UDF!
See UDF Limitations for more
Related
I tried to query data in TDengine with the following sql clause:
SELECT * from TABLE where COLUMN < FUNCTION();
and I got illegal condition expression error.
Does anyone know if using function in where condition is illegal usage in SQL convention or I need to resort to other syntax?
Using function in where condition is illegal. Maybe it will be supported later.
It may seem a little strange, but there are already tables with names for each date.
In my project, I have tables for each date to make statistics easier to handle.
Of course, I don't think this is always the best way, but this is the table structure for my project.
(It's a common technique in Google BigQuery and Amazon Athena. This question is about Google BigQuery)
So to get the data, I want to generate today's date. If I use TODAY, I can get the data of the latest day without rewriting the code even if it is the next day.
I tried, but the code didn't work.
Not work 1:
CONCAT in FROM
SELECT
*
FROM
CONCAT('foo_', FORMAT_TIMESTAMP('%Y%m%d', CURRENT_TIMESTAMP(), 'Asia/Tokyo'))
Error:
Table-valued function not found: CONCAT at [4:3]
Not work 2:
create temporary function:
create temporary function getTableName() as (CONCAT('foo_', FORMAT_TIMESTAMP('%Y%m%d', CURRENT_TIMESTAMP(), 'Asia/Tokyo')));
Error:
CREATE TEMPORARY FUNCTION statements must be followed by an actual query.
Question
How do I generate a table name that contains TODAY's date?
In this case, I would recommend you to use Wild tables in BigQuery, which allows you to use some features in Standard SQL.
With Wild Tables you can use _TABLE_SUFFIX, it grants you the ability to filter/scan tables containing this parameter. The syntax would be as follows:
SELECT *
FROM `test-proj-261014.sample.test_*`
where _TABLE_SUFFIX = FORMAT_DATE('%Y%m%d', CURRENT_DATE)
I hope it helps.
Your first query should go like this:
select CONCAT('foo_', FORMAT_TIMESTAMP('%Y%m%d', CURRENT_TIMESTAMP(), 'Asia/Tokyo'))
For creating temporary function, use the below code:
create temp function getTableName() as
((select CONCAT('foo_', FORMAT_TIMESTAMP('%Y%m%d', CURRENT_TIMESTAMP(), 'Asia/Tokyo'))
));
select getTableName()
The error "CREATE TEMPORARY FUNCTION statements must be followed by an actual query." is because once the temporary functions are defined then you have to use the actual query to use that function and then the validity of function dies out. To define persistent UDFs and use them in multiple queries please go through the link to define permanent functions.You can reuse persistent UDFs across multiple queries, whereas you can only use temporary UDFs in a single query.
I'm trying to do pagination in DB2. I wouldn't like to do it with subquery, but OFFSET is not working with TIMESTAMP_FORMAT.
Use of function TIMESTAMP_FORMAT in QSYS2 not valid. Data mapping error on member
I've found this question, but seems here the problem is with content of column and it's not my case, as values are alright and TIMESTAMP_FORMAT works without OFFSET.
I didn't look for some other way to not use TIMESTAMP_FORMAT, as I need to create pagination on queries written not by me, but by client.
The query looks like this.
SELECT DATE(TIMESTAMP_FORMAT(CHAR("tablename"."date"),'YYMMDD'))
FROM tableName
OFFSET 10 ROWS
I get
"[SQL0583] Use of function TIMESTAMP_FORMAT in QSYS2 not valid."
I'm not sure how OFFSET can relate to TIMESTAMP_FORMAT, but when I replace the select with select * it works fine.
I wonder why there is a conflict between OFFSET and TIMESTAMP_FORMAT and is there a way to bypass this without subquery.
From Listing of SQL Messages:
SQL0583
Function &1 in &2 cannot be invoked where specified because it is
defined to be not deterministic or contains an external action.
Functions that are not deterministic cannot be specified in a GROUP BY
clause or in a JOIN clause, or in the default clause for a global
variable.
Functions that are not deterministic or contain an external
action cannot be specified in a PARTITION BY clause or an ORDER BY
clause for an OLAP function and cannot be specified in the select list
of a query that contains an OFFSET clause.
The RAISE_ERROR function
cannot be specified in a GROUP BY or HAVING clause.
I don't know how to check these properties for the QSYS2.TIMESTAMP_FORMAT function (there is no its definition in the QSYS2.SYSROUTINES table), but it looks like improper definition of this function - there is no reason to create it as not deterministic or external action.
You can "deceive" DB2 like this:
CREATE FUNCTION MYSCHEMA.TIMESTAMP_FORMAT(str VARCHAR(4000), fmt VARCHAR(128))
RETURNS TIMESTAMP
DETERMINISTIC
CONTAINS SQL
NO EXTERNAL ACTION
RETURN QSYS2.TIMESTAMP_FORMAT(str, fmt);
SELECT
DATE(MYSCHEMA.TIMESTAMP_FORMAT(CHAR(tablename.date), 'YYMMDD'))
--DATE(QSYS2.TIMESTAMP_FORMAT(CHAR(tablename.date), 'YYMMDD'))
FROM table(values '190412') tableName(date)
OFFSET 10 ROWS;
And use this function instead. It works on my 7.3 at least.
It's a harmless deception, and you may ask IBM support to clarify such a "feature" of QSYS2.TIMESTAMP_FORMAT...
I suspect your problem is bad data...
The default for the IBM interactive tools, STRSQL and ACS Run SQL Scripts, is OPTIMIZE(*FIRSTIO) meaning get the first few rows back as quickly as possible...
With the OFFSET 10 clause you're probably accessing rows initially that you didn't before.
Try the following
create table mytest as (
SELECT DATE(TIMESTAMP_FORMAT(CHAR("tablename"."date"),'YYMMDD')) as mydate
FROM tableName
) with data
If that doesn't error, then yes you've found a bug, open a PMR.
Otherwise, you can see how far along the DB got by looking at the rows in the new table and track down the record with bad data.
i want to sort my result based on parameter:
select ... group by some_column :paramter
and the parameter should be asc or desc. but when i try it, i get error:
ERROR: syntax error at or near "$6"
is it possible in postgres? if i send the query from the server, i can just concatenate the query but what if it's inside a stored procedure? i would prefer to avoid concatenating strings to build a query
SQL parameters cannot be placed in that location. They are meant to apply scalar values to the SQL template, but not to modify the SQL query structure itself.
What you need is to implement "Dynamic SQL" here, but that's a different strategy.
I'm attempting to write an extension for SQL Developer to better support Postgres. These are just XML configuration files with SQL snippets in them. To display the values for a postgres sequence, I need to run a simple query of the following form:
select * from schema.sequence
The trouble with this is that the Oracle SQL Developer environment provides the correct schema and node (sequence) name as bind variables. This would mean that I should format the query as:
select * from :SCHEMA.:NAME
The trouble with this is that bind variables are only valid in the select clause or the where clause (as far as I'm aware), and using this form of the query returns a "syntax error at or near "$1" error message.
Is there a way to return the values in the sequence object without directly selecting them from the sequence? Perhaps some obtuse joined statement from pg_catalog tables?
Try this:
select *
from information_schema.sequences
where sequence_name = :name
and sequence_schema = :schema;
It's not exactly the same thing as doing a select from the sequence, but the basic information is there.