I want to use SELECT INTO to make a temporary table in one of my functions. SELECT INTO works in SQL but not PL/pgSQL.
This statement creates a table called mytable (If orig_table exists as a relation):
SELECT *
INTO TEMP TABLE mytable
FROM orig_table;
But put this function into PostgreSQL, and you get the error: ERROR: "temp" is not a known variable
CREATE OR REPLACE FUNCTION whatever()
RETURNS void AS $$
BEGIN
SELECT *
INTO TEMP TABLE mytable
FROM orig_table;
END; $$ LANGUAGE plpgsql;
I can SELECT INTO a variable of type record within PL/pgSQL, but then I have to define the structure when getting data out of that record. SELECT INTO is really simple - automatically creating a table of the same structure of the SELECT query. Does anyone have any explanation for why this doesn't work inside a function?
It seems like SELECT INTO works differently in PL/pgSQL, because you can select into the variables you've declared. I don't want to declare my temporary table structure, though. I wish it would just create the structure automatically like it does in SQL.
Try
CREATE TEMP TABLE mytable AS
SELECT *
FROM orig_table;
Per http://www.postgresql.org/docs/current/static/sql-selectinto.html
CREATE TABLE AS is functionally similar to SELECT INTO. CREATE TABLE AS is the recommended syntax, since this form of SELECT INTO is not available in ECPG or PL/pgSQL, because they interpret the INTO clause differently. Furthermore, CREATE TABLE AS offers a superset of the functionality provided by SELECT INTO.
Related
I am struggling a bit on some dynamic postgresql :
I have a table named "list_columns" containing the columns names list with "column_name" as the variable name; those column names come from an input table called "input_table".
[list_columns]
column_name
col_name_a
col_name_b
col_name_c...
[input_table]
col_name_a
col_name_b
col_name_c
col_name_d
col_name_e
value_a_1
value_b_1
value_c_1
value_d_1
value_e_1
value_a_2
value_b_2
value_c_2
value_d_2
value_e_2
...
...
...
...
...
What I'd like to do is dynamically create a new table using that list, something like this:
create table output_table as
select (select distinct(column_name) seperated by "," from list_columns) from input_table;
The resulting table would be
[output_table]
col_name_a
col_name_b
col_name_c
value_a_1
value_b_1
value_c_1
value_a_2
value_b_2
value_c_2
...
...
...
I saw I should use some execute procedures but I can't figure out how to do so.
Note: I know i could directly select the 3 columns; I oversimplied the case.
If someone would be kind enough to help me on this,
Thank you,
Regards,
Jonathan
You need dynamic SQL for this, and for that you need PL/pgSQL.
You need to assemble the CREATE TABLE statement based on the input_table, then run that generated SQL.
do
$$
declare
l_columns text;
l_sql text;
begin
-- this generates the list of columns
select string_agg(distinct column_name, ',')
into l_columns
from list_table;
-- this generates the actual CREATE TABLE statement using the columns
-- from the previous step
l_sql := 'create table output_table as select '||l_columns||' from input_table';
-- this runs the generated SQL, thus creating the output table.
execute l_sql;
end;
$$;
If you need that a lot, you can put that into a stored function (your unsupported Postgres version doesn't support real procedures).
This code:
ALTER TABLE myschema.mytable add column geom geometry (point,4326);
CREATE INDEX mytable_idx on myschema.mytable using GIST(geom);
UPDATE myschema.mytable set geom = st_setsrid(st_point(mytable.long, mytable.lat), 4326);
This works fine when updating a single table. How would you convert it into a dynamic SQL function, with schema and table as parameters?
Since the function input must be an existing table, the simplest safe way would be to use a regclass input parameter like demonstrated here:
Table name as a PostgreSQL function parameter
However, you also need the bare table name for the concatenated index name, so I'll stick with taking text for schema and table separately:
CREATE OR REPLACE FUNCTION create_geom(_sch text, _tab text)
RETURNS void
LANGUAGE plpgsql AS
$func$
BEGIN
EXECUTE format(
'ALTER TABLE %1$I.%2$I ADD COLUMN geom geometry(POINT,4326);
UPDATE %1$I.%2$I SET geom = st_setsrid(st_point(long, lat), 4326);
CREATE INDEX %3$I ON %1$I.%2$I USING gist(geom);'
, _sch, _tab
, _tab || '_geom_gist_idx');
END
$func$;
Call:
SELECT create_geom('myschema', 'mytable');
Use a single EXECUTE, no need for multiple calls.
Just omit table-qualification for columns in the UPDATE. While not joining additional tables, column names are unambiguous. Else, use a table alias, which can be constant. Like:
UPDATE %1$s AS x SET geom = st_setsrid(st_point(x.long, x.lat), 4326);
But it's smarter to populate the column before you build the index. That's a lot faster and produces a balanced index without bloat. So I switched the commands.
Note how I concatenate the index name first (_tab || '_geom_gist_idx'), and then double-quote as required with %3$I. That's the safe way. Something like %I_idx fails with non-standard names.
That said, it's typically a mistake to add columns with redundant information to a table. (What keeps you from changing one or the other? Why bloat the table?) Either just use an expression index instead of all of the above:
CREATE INDEX ON myschema.mytable USING gist (st_setsrid(st_point(long, lat), 4326));
Or drop the now redundant long & lat from the table. Those can be extracted from the new geom cheaply on the fly.
Or, if you need all columns (for special performance reasons?), consider a generated column instead. See:
Computed / calculated / virtual / derived columns in PostgreSQL
Having your queries as SQL templates and using format function for identifiers:
CREATE OR REPLACE FUNCTION public.create_geom(sch text, tab text)
RETURNS void language plpgsql AS $body$
DECLARE
DYNSQLA constant text := 'ALTER TABLE %I.%I add column geom geometry (point,4326)';
DYNSQLB constant text := 'CREATE INDEX %I_idx on %I.%I using GIST(geom);';
DYNSQLC constant text := 'UPDATE %I.%I set geom = st_setsrid(st_point(%I.long, %I.lat), 4326)';
BEGIN
execute format(DYNSQLA, sch, tab);
execute format(DYNSQLB, tab, sch, tab);
execute format(DYNSQLC, sch, tab, tab, tab);
END;
$body$;
SELECT create_geom('myschema','mytable');
I am writing a function which will select and SUM the resulting output into a new table-therefore I attempted to use the INTO function. However, my standalone code works, yet once a place into a function I get an error stating that the new SELECT INTO table is not a defined variable (perhaps I am missing something). Please see code below:
CREATE OR REPLACE FUNCTION rev_1.calculate_costing_layer()
RETURNS trigger AS
$BODY$
BEGIN
-- This will create an intersection between pipelines and sum the cost to a new table for output
-- May need to create individual cost columns- Will also keep infrastructure costing seperated
--DROP table rev_1.costing_layer;
SELECT inyaninga_phases.geom, catchment_e_gravity_lines.name, SUM(catchment_e_gravity_lines.cost) AS gravity_sum
INTO rev_1.costing_layer
FROM rev_1.inyaninga_phases
ON ST_Intersects(catchment_e_gravity_lines.geom,inyaninga_phases.geom)
GROUP BY catchment_e_gravity_lines.name, inyaninga_phases.geom;
RETURN NEW;
END;
$BODY$
language plpgsql
Per the documentation:
CREATE TABLE AS is functionally similar to SELECT INTO. CREATE TABLE AS is the recommended syntax, since this form of SELECT INTO is not available in ECPG or PL/pgSQL, because they interpret the INTO clause differently. Furthermore, CREATE TABLE AS offers a superset of the functionality provided by SELECT INTO.
Use CREATE TABLE AS.
Although SELECT ... INTO new_table is valid PostgreSQL, its use has been deprecated (or, at least, "unrecommended"). It doesn't work at all in PL/PGSQL, because INSERT INTO is used to get results into variables.
If you want to create a new table, you should use instead:
CREATE TABLE rev_1.costing_layer AS
SELECT
inyaninga_phases.geom, catchment_e_gravity_lines.name, SUM(catchment_e_gravity_lines.cost) AS gravity_sum
FROM
rev_1.inyaninga_phases
ON ST_Intersects(catchment_e_gravity_lines.geom,inyaninga_phases.geom)
GROUP BY
catchment_e_gravity_lines.name, inyaninga_phases.geom;
If the table has already been created an you just want to insert a new row in it, you should use:
INSERT INTO
rev_1.costing_layer
(geom, name, gravity_sum)
-- Same select than before
SELECT
inyaninga_phases.geom, catchment_e_gravity_lines.name, SUM(catchment_e_gravity_lines.cost) AS gravity_sum
FROM
rev_1.inyaninga_phases
ON ST_Intersects(catchment_e_gravity_lines.geom,inyaninga_phases.geom)
GROUP BY
catchment_e_gravity_lines.name, inyaninga_phases.geom;
In a trigger function, you're not very likely to create a new table every time, so, my guess is that you want to do the INSERT and not the CREATE TABLE ... AS.
After some aggravation, I found (IMO) odd behavior when a function calls another. If the outer function creates a temporary table, and the inner function creates a temporary table with the same name, the inner function "wins." Is this intended? FWIW, I am proficient at SQL Server, and temporary tables do not act this way. Temporary tables (#temp or #temp) are scoped to the function. So, an equivalent function (SQL Server stored procedure) would return "7890," not "1234."
drop function if exists inner_function();
drop function if exists outer_function();
create function inner_function()
returns integer
as
$$
begin
drop table if exists tempTable;
create temporary table tempTable (
inner_id int
);
insert into tempTable (inner_id) values (1234);
return 56;
end;
$$
language plpgsql;
create function outer_function()
returns table (
return_id integer
)
as
$$
declare intReturn integer;
begin
drop table if exists tempTable; -- note that inner_function() also declares tempTable
create temporary table tempTable (
outer_id integer
);
insert into tempTable (outer_id) values (7890);
intReturn = inner_function(); -- the inner_function() function recreates tempTable
return query
select * from tempTable; -- returns "1234", not "7890" like I expected
end;
$$
language plpgsql;
select * from outer_function(); -- returns "1234", not "7890" like I expected
There are no problem with this behaviour, in PostgreSQL temp table can have two scopes:
- session (default)
- transaction
To use the "transaction" scope you should use "ON COMMIT DROP" at the end of the CREATE TEMP statement, i.e:
CREATE TEMP TABLE foo(bar INT) ON COMMIT DROP;
Anyway your two functions will be executed in one transaction so when you call the inner_function from the outer_function you'll be in the same transaction and PostgreSQL will detect that "tempTable" already exists in the current session and will drop it in "inner_function" and create again...
Is this intended?
Yes, these are tables in the database, similar to permanent tables.
They exist in a special schema, and are automatically dropped at the end of a session or transaction. If you create a temporary table with the same name as a permanent table, then you must prefix the permanent table with its schema name to reference it while the temporary table exists.
If you want to emulate the SQL Server implementation then you might consider using particular prefixes for your temporary tables.
I am using postgreSQL. I want to select data from a table. Such table name contains the current year. such as abc2013. I have tried
select * from concat('abc',date_part('year',current_date))
select *from from concat('abc', extract (year from current_date))
So how to fetch data from such table dynamically?
Please don't do this - look hard at alternatives first, starting with partitioning and constraint exclusion.
If you must use dynamic table names, do it at application level during query generation.
If all else fails you can use a PL/PgSQL procedure like:
CREATE OR REPLACE pleasedont(int year) RETURNS TABLE basetable AS $$
BEGIN
RETURN QUERY EXECUTE format('SELECT col1, col2, col3 FROM %I', 'basetable_'||year);
END;
$$ LANGUAGE plpgsql;
This will only work if you have a base table that has the same structure as the sub-tables. It's also really painful to work with when you start adding qualifiers (where clause constraints, etc), and it prevents any kind of plan caching or effective prepared statement use.