I am trying to convert SQL Server stored function to PostgreSQL stored function I am getting one syntactical error at declare #table1 table
CREATE OR REPLACE FUNCTION ETL_GetBuildingDetailsByUserID ( p_nInstID numeric=0)
RETURNS Boolean
AS $$
declare #table1 table
(
nbuilding numeric(18, 0) NOT NULL,
sbuild_name varchar(50) NULL,
sclient_build_id varchar(50) NULL,
nbuilding_inst_id numeric(18, 0) NOT NULL,
ntemp_building_id numeric(18,0) NULL,
nno_of_floors numeric(18,0) NULL
)
declare v_strDeptIds text;
v_strSubDeptIds text;
BEGIN
v_strsql := 'SELECT building.*
FROM building
WHERE (building.nbuilding_inst_id = '|| cast(p_nInstID as varchar(1)) ||')
';
print v_strsql
v_strsql1 text;
v_strsql1 := v_strsql
Insert into #table1; execute sp_executesql; v_strsql1
select * from #table1;
Return true;
END;
$$ LANGUAGE plpgsql;
Error
ERROR: syntax error at or near "#"
LINE 4: declare #table1 table
Can any one please tell what I am doing wrong?
It seems your function actually returns the result of a SELECT query, not a boolean value, so returns boolean is wrong to begin with.
To return a result, you need to declare the function as returns table(). But as you seem to simply return rows from the building table you can define it as returns setof building.
Then remove the useless dynamic SQL which seems completely unnecessary.
In PL/pgSQL there are no table variables, and copying the result of a query into one before returning that result from that table seems to be an unnecessary step which only slows down things. In Postgres you simply return the result of the query, there is no need to store it locally.
Additionally: rather than casting a parameter to another type inside the function it's better to declare that parameter with the type you expect.
So the simplified version of that function in PostgreSQL would be:
CREATE OR REPLACE FUNCTION ETL_GetBuildingDetailsByUserID ( p_nInstID text)
RETURNS setof building
AS $$
select building.*
from building
WHERE building.nbuilding_inst_id = p_nInstID
$$ LANGUAGE sql;
You can use it like this:
select *
from ETL_GetBuildingDetailsByUserID ('42');
Unrelated, but: using numeric(18,0) for columns that store values without decimals is overkill. You should define those columns as bigint. Much faster and uses less space than numeric.
Related
I have a Postgres table bearing the following form
CREATE TABLE "public"."days"
(
"id" integer NOT NULL,
"day" character varying(9) NOT NULL,
"visits" bigint[] NOT NULL,
"passes" bigint[] NOT NULL
);
I would like to write a function that allows me to return the visits or the passees column as its result for a specified id. My first attempt goes as follows
CREATE OR REPLACE FUNCTION day_entries(INT,TEXT) RETURNS BIGINT[] LANGUAGE sql AS
'SELECT $2 FROM days WHERE id = $1;'
which fails with an error along the lines of
return type mismatch in function declared to return bigint[]
DETAIL: Actual return type is text.
If I put in visits in place of the $2 things work just as expected. It would make little sense to define several functions to match different columns from the days table. Is there a way to pass the actual column name as a parameter while still keeping Postgres happy?
You can't use parameters as identifiers (=column name), you need dynamic SQL for that. And that requires PL/pgSQL:
CREATE OR REPLACE FUNCTION day_entries(p_id int, p_column text)
RETURNS BIGINT[]
AS
$$
declare
l_result bigint[];
begin
execute format('SELECT %I FROM days WHERE id = $1', p_column)
using p_id
into l_result;
return l_result;
end;
$$
LANGUAGE plpgsql;
format() properly deals with identifiers when building dynamic SQL. The $1 is a parameter placeholder and the value for that is passed with the using p_id clause of the execute statement.
I have the table with some columns:
--table
create table testz
(
ID uuid,
name text
);
Note: I want to insert ID values by passing as a parameter to the function. Because I am generating the ID value
in the front end by using uuid_generate_v4(). So I need to pass the generated value to the function to insert
into the table
My bad try:
--function
CREATE OR REPLACE FUNCTION testz
(
p_id varchar(50),
p_name text
)
RETURNS VOID AS
$BODY$
BEGIN
INSERT INTO testz values(p_id,p_name);
END;
$BODY$
LANGUAGE PLPGSQL;
--EXECUTE FUNCTION
SELECT testz('24f9aa53-e15c-4813-8ec3-ede1495e05f1','Abc');
Getting an error:
ERROR: column "id" is of type uuid but expression is of type character varying
LINE 1: INSERT INTO testz values(p_id,p_name)
You need a simple cast to make sure PostgreSQL understands, what you want to insert:
INSERT INTO testz values(p_id::uuid, p_name); -- or: CAST(p_id AS uuid)
Or (preferably) you need a function, with exact parameter types, like:
CREATE OR REPLACE FUNCTION testz(p_id uuid, p_name text)
RETURNS VOID AS
$BODY$
BEGIN
INSERT INTO testz values(p_id, p_name);
END;
$BODY$
LANGUAGE PLPGSQL;
With this, a cast may be needed at the calling side (but PostgreSQL usually do better automatic casts with function arguments than inside INSERT statements).
SQLFiddle
If your function is that simple, you can use SQL functions too:
CREATE OR REPLACE FUNCTION testz(uuid, text) RETURNS VOID
LANGUAGE SQL AS 'INSERT INTO testz values($1, $2)';
I've setup a Stored Procedure in PL/Proxy to make a query, and receive some RECORDs back.
In PL/Proxy:
CREATE OR REPLACE FUNCTION query_autocomplete(q text, i_id bigint)
RETURNS SETOF RECORD AS $$
CLUSTER 'autocompletecluster';
RUN ON i_id;
$$ LANGUAGE plproxy;
In each Partition:
CREATE OR REPLACE FUNCTION query_autocomplete(q text, i_id bigint)
RETURNS SETOF RECORD AS $$
DECLARE
rec RECORD;
BEGIN
FOR rec IN EXECUTE q
LOOP
RETURN NEXT rec;
END LOOP;
RETURN;
END;
$$ LANGUAGE plpgsql;
As you've likely guessed, this is hitting a defined SERVER in PGSQL called 'autocompletecluster'. The query string that I'm sending through is as follows:
$sql = "SELECT * FROM autocomplete WHERE member_id = :memberId";
$query = $this->db->prepare("SELECT query_autocomplete('{$sql}',1234");
It's returning the following:
SQLSTATE[XX000]: Internal error: 7 ERROR: PL/Proxy function public.query_autocomplete(0): unsupported type
The table that query is hitting is defined as such:
CREATE TABLE autocomplete (
id character varying(100) NOT NULL,
extra_data hstore,
username character varying(254),
member_id bigint
);
What am I doing wrong?
The error strongly suggests that PL/Proxy doesn't support SETOF RECORD. Try instead defining your functions to return autocomplete%rowtype or, failing that, RETURNS TABLE (...) with a matching columns-set.
I have an error, but I don't know what the problem is.
I want execute a function and return a value from a column filled in by the column default, a sequence - the equivalent of currval(sequence).
I use:
PostgreSQL 9.0
pgAdmin III
CREATE OR REPLACE FUNCTION name_function(in param_1 character varying
, out param_2 bigint)
AS
$$
BEGIN
INSERT INTO table (collumn_seq,param_1) VALUES (DEFAULT,param_1)
returning collumn_seq;
--where:collumn_seq reference a collumn serial..
END;
$$
LANGUAGE plpgsql VOLATILE;
I can create the function without error but when trying to execute, the following error is returned:
SELECT name_function('GHGHGH');
ERROR: The query has no destination for result data
It would work like this:
CREATE OR REPLACE FUNCTION name_function(param_1 varchar
, OUT param_2 bigint)
LANGUAGE plpgsql AS
$func$
BEGIN
INSERT INTO table (collumn_seq, param_1) -- "param_1" also the column name?
VALUES (DEFAULT, param_1)
RETURNING collumn_seq
INTO param2;
END
$func$;
Normally, you would add a RETURN statement, but with OUT parameters, this is optional.
Refer to the manual for more details:
Returning from a function
Executing a Query with a Single-row Result
The simple case can be covered with a plain SQL function.
And you can omit the target column that shall get its DEFAULT value.
And you can just as well use a RETURNS clause in this case:
CREATE OR REPLACE FUNCTION name_function(param_1 varchar)
RETURNS bigint
LANGUAGE sql AS
$func$
INSERT INTO table (param_1) -- "param_1" also the column name?
VALUES (param_1)
RETURNING collumn_seq;
$func$;
I want create a function:
CREATE OR REPLACE FUNCTION medibv.delAuto(tableName nvarchar(50), columnName nvarchar(100),value
nvarchar(100))
RETURNS void AS
$BODY$
begin
DELETE from tableName where columnName=value
end;
$BODY$
LANGUAGE plpgsql VOLATILE;
I have these parameters: tableName, columnName, value.
I want tableName as table in PostgreSQL.
CREATE OR REPLACE FUNCTION medibv.delauto(tbl regclass, col text, val text
,OUT success bool)
RETURNS bool AS
$func$
BEGIN
EXECUTE format('
DELETE FROM %s
WHERE %I = $1
RETURNING TRUE', tbl, col)
USING val
INTO success;
RETURN; -- optional in this case
END
$func$ LANGUAGE plpgsql;
Call:
SELECT medibv.delauto('myschema.mytbl', 'my_txt_col', 'foo');
Returns TRUE or NULL.
There is no nvarchar type in Postgres. You may be thinking of SQL Server. The equivalent would be varchar, but most of the time you can simply use text.
regclass is a specialized type for registered table names. It's perfect for the case an prevents SQL injection for the table name automatically and most effectively. More in the related answer below.
The column name is still prone to SQL injection. I sanitize the function with format(%I).
format() requires PostgreSQL 9.1+.
Your function did not report what happened. One or more rows may be found and deleted. Or none at all. As a bare minimum I added a boolean OUT column which will be TRUE if one or more rows were deleted. Because (quoting the manual here):
If multiple rows are returned, only the first will be assigned to the INTO variable.
Lastly, use USING with EXECUTE to pass in values. Don't cast back and forth. This is inefficient and prone to errors and to SQLi once more.
Find more explanation and links in this closely related answer:
Table name as a PostgreSQL function parameter
Use EXECUTE to run dynamic commands:
CREATE OR REPLACE FUNCTION medibv.delAuto(tableName nvarchar(50), columnName nvarchar(100),value
nvarchar(100))
RETURNS void AS
$BODY$
begin
EXECUTE 'DELETE FROM ' || tableName || ' WHERE ' || columnName || '=' || value;
end;
$BODY$
LANGUAGE plpgsql VOLATILE;