Comments for SimonOnSoftware

Comment on Random String in PostgreSQL by Szymon

Szymon — Mon, 16 Apr 2012 11:50:23 +0000

That’s not surprising, as this function does nothing with strings, it just uses arrays of characters, it really doesn’t matter what the characters are.

Comment on Random String in PostgreSQL by Roman

Roman — Mon, 16 Apr 2012 11:38:10 +0000

– FOR NATIONAL WORDS SAMPLE:
CREATE OR REPLACE FUNCTION random_string_bis(length INT) RETURNS TEXT
LANGUAGE PLPGSQL
AS $$
DECLARE aAl text[] := ‘{0,1,2,3,4,5,6,7,8,9,А,Б,В,Г,Д,Е,Ё,Ж,З,И,Й,К,Л,М,Н,О,П,Р,С,Т,В,У,Ф,Х,Ц,Ч,Ш,Щ,Ъ,Ы,Ь,Э,Ю,Я}’;
res text;
begin
res := (select string_agg(aAl[(random()*44)::int], ”) FROM generate_series(1,length) t(a));
return res;
end
$$;
select random_string_bis(50);

DROP FUNCTION IF EXISTS random_string_stat (INT, INT);
CREATE OR REPLACE FUNCTION random_string_stat(iLength INT, iLoops int) RETURNS interval
LANGUAGE PLPGSQL
AS $$
DECLARE sRes text;
tsStart timestamp;
tsFinish timestamp;
begin
tsStart := clock_timestamp();
FOR i IN 1..iLoops LOOP
sRes := random_string_bis(iLength);
– sRes := random_string_pavel(iLength);
– sRes := random_text_md5(iLength);
– sRes := random_text_md5_v2(iLength);
END LOOP;
tsFinish := clock_timestamp();
RETURN (tsFinish – tsStart)/iLoops;
end
$$;

Comment on Virtualenv Tutorial by telminov

telminov — Mon, 09 Apr 2012 10:34:00 +0000

It’s good quality tutorial. Thank you!

Comment on Why Ruby on Rails Migrations Don’t Work by Guest

Guest — Fri, 17 Feb 2012 16:48:26 +0000

Just a warning to future readers. Contrary to the vibe I get from this article, You should ALWAYS use uniqueness constraints on the DB as well as in the model.
This, “No unique indexes in databases because they are not needed,” is not a true statement. The model will work most of the time but it WILL NOT check for concurrent writes.

That is (at the same time):
user_a = User.new(name => “foobar”)
user_b = User.new(name => “foobar”)
will validate because neither of these have been saved to the DB yet thus they are both past the model validation.

so, without a DB unique contraint:
user_a.save
user_b.save
will result in both of these in the DB even though they identical.

It WILL happen (although probably very rarely) if you don’t have unique constraints on the DB as well as the model.

Comment on PostgreSQL Versioning Policy by Simon

Simon — Mon, 21 Nov 2011 18:12:16 +0000

@Filip yes indeed thanks for the link

Comment on PostgreSQL Versioning Policy by Filip

Filip — Mon, 21 Nov 2011 18:05:37 +0000

obligatory rading: http://semver.org/

Comment on How to Ensure There is One Row… part 2 by Simon

Simon — Thu, 10 Nov 2011 21:27:46 +0000

Yep, indeed, much better solution
Thanks

Comment on How to Ensure There is One Row… part 2 by daniel

daniel — Wed, 09 Nov 2011 23:25:37 +0000

Better (becouse smaller):
CREATE UNIQUE INDEX i_test ON test (name) WHERE value = ‘a’;

Comment on Random String in PostgreSQL by Svend Hansen

Svend Hansen — Thu, 20 Oct 2011 14:35:31 +0000

Hi,
I like the pure SQL version, but it seems that when you added the numbers to the seed String array ([0-9][A-Z]), you forgot to change the ’26′ to ’36′ when randomly indexing into it (to take advantage of the full string)
Cheers,
Svend.

Comment on PostGIS and Small Number of Huge Geometries by Simon

Simon — Wed, 12 Oct 2011 16:03:20 +0000

I’ve already done that