Обсуждение: Format base - Code contribution

On 23 April 2018 at 08:23, Miles Elam <mileselam@gmail.com> wrote:
> I would like to donate some code to the project, formatting numbers as any
> base from 2 to 64. The FAQ describes contributions to the core code, but
> it's possible contrib is a better target. This is all of course contingent
> on how well received this extension code is of course. Code available at the
> following link.
>
> https://github.com/ttfkam/pg_formatbase

Personally, I think this is a better candidate for being incorporated
directly rather than as a contrib. This sort of utility is much less
useful if you cannot rely on it being present.

I'm not convinced by the wisdom of adding int8 overloads, etc, with a
second argument. I'd rather this be named as a separate function. I
realise that many programming languages do this, but it's IMO less
discoverable this way, and might make our life harder if we later need
to overload these functions in a different way.

We already have to_hex. So to_base seems a reasonable choice. Then
adding a from_hex, from_base seems natural. Bonus points if you add
to/from base64 and oct while you're at it.


We don't seem to have a "from_hex" or "int8_from_hex", which is a
bewildering oversight really, and we don't accept literals:

test=> select int8 0x1234;
ERROR:  syntax error at or near "0"
LINE 1: select int8 0x1234;
                    ^
test=> select int8 '0x1234';
ERROR:  invalid input syntax for integer: "0x1234"
LINE 1: select int8 '0x1234';

... unless you abuse our nonstandard bitstring SQL extension:

test=> select x'1234';
     ?column?
------------------
 0001001000110100
(1 row)

test=> select pg_typeof( x'1234' );
 pg_typeof
-----------
 bit
(1 row)

test=> select x'1234'::int8;
 int8
------
 4660
(1 row)

which won't work for anyone using bind parameters, so it's not that
handy really.


I'm also amused by

test=> select 0x1234;
 x1234
-------
     0
(1 row)

because of our willingness to ignore the whitespace between a value
and the column label.


While I'm on that topic, I've never found anything that unquotes a
literal or identifier without going through the full parser, some sort
of unquote_literal. Guess I should find time to scratch that itch
myself soon.

-- 
 Craig Ringer                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services