From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Demitri Muna <postgresql(at)demitri(dot)com> |
Cc: | pgsql-general(at)lists(dot)postgresql(dot)org |
Subject: | Re: Getting a primitive numeric value from "DatumGetNumeric"? |
Date: | 2018-02-21 03:54:41 |
Message-ID: | 21871.1519185281@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Demitri Muna <postgresql(at)demitri(dot)com> writes:
> I’m writing a C extension for PostgreSQL. One possible input datatype for my function is a numeric array, e.g. ARRAY[[1.5,2.5],[3.5,4.5]]. I can use “DatumGetNumeric” to extract a “Numeric” data type from the data, but at some point I need to convert this to a number (e.g. double) so that I can do mathy things with it. How does one convert a “Numeric” to, say, a double?
If you want to work with doubles, why don't you declare the function as
taking doubles?
> I have a workaround in that I can pass this to my function:
> ARRAY[[1.5,2.5],[3.5,4.5]]::float8[]
> but I’d rather have the code do that instead of bothering the user to remember that.
Well, the implicit coercions work in your favor in this particular case.
You can just do, eg,
regression=# create function foo(float8[]) returns float8 as
regression-# 'select $1[1]' language sql;
CREATE FUNCTION
regression=# select foo(array[1.1,1.2]);
foo
-----
1.1
(1 row)
or to emphasize that it is doing a conversion:
regression=# select foo(array[1.1,1.2]::numeric[]);
foo
-----
1.1
(1 row)
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Alexander Farber | 2018-02-21 12:16:10 | Not sure if I should CREATE INDEX for text columns on which I plan to filter later |
Previous Message | Demitri Muna | 2018-02-21 00:11:49 | Getting a primitive numeric value from "DatumGetNumeric"? |