old pgindent change

Поиск
Список
Период
Сортировка
От Nigel J. Andrews
Тема old pgindent change
Дата
Msg-id Pine.LNX.4.21.0309222248030.29572-200000@ponder.fairway2k.co.uk
обсуждение исходный текст
Ответы Re: old pgindent change  ("Nigel J. Andrews" <nandrews@investsystems.co.uk>)
Re: old pgindent change  (Bruce Momjian <pgman@candle.pha.pa.us>)
Список pgsql-hackers
There was a simple change commited in revision 1.47 of pgindent, listed as
being "More updates for GNU indent".

The questions are: why? and surely I can't be the only one whose hit this
problem since November 2001?

On a debian (woody or potato, which ever one had a 2.2 series kernal) using
GNU bash 2.03.0 via /bin/sh (in the pgindent script) I get:


which indent = /usr/local/bin/indent
indent -version = Berkeley indent 5.17

status = 0
You do not appear to have 'indent' installed on your system.

By adding appropiate echo commands before and after the indent -version line in
the script, I've attached my slightly modified version for completeness. I can
never remember the way around the test of $? succeds or fails but the above
experiment plus a look at the near by tests of $? in pgindent seem to show that
there is a problem here.

A simple test of 2.05a.0 on a newer system using:

ls
if [ $? -eq 0 ]; then echo success aaa; fi
lls
if [ $? -eq 0 ]; then echo success bbb; fi

shows that the lls (non-existant executable) status fails the test as expected
while the plain ls status passes.

I'm obviously missing something very significant here or there's a very strange
oddity that's been there, and specifically placed there, for nearly 2 years.


-- 
Nigel J. Andrews

В списке pgsql-hackers по дате отправления:

Предыдущее
От: markw@osdl.org
Дата:
Сообщение: More Prelimiary DBT-2 Test Results with PostgreSQL 7.3.4
Следующее
От: "Nigel J. Andrews"
Дата:
Сообщение: Re: old pgindent change