Обсуждение: psql code to be obducted by alien (me)

Поиск
Список
Период
Сортировка

psql code to be obducted by alien (me)

От
Peter Eisentraut
Дата:
As authorized by Tom L. himself and barring any loud protests I will start
a code clean-up on psql sometime within the next few days. So if you are
doing something on it or plan on doing so, please let me know. Among the
things on the agenda are:

* Reduce function size, indentation levels, file sizes

* Make use of latest libpq facilities

* Take care of various TODO items, such as NULL display (perhaps the one
recently submitted by me?), \copy issues, possibly more.

* Allow for implementation of a more sophisticated readline TAB
completion. (Not necessarily the one I recently sent in, if you can come
up with a better one. I'll try to keep it general.)

* Have tables vs. views show up correctly, as explained to me by Jan.

* Various enhancements on HTML display.

* A full bag of other ideas which will probably have to be postponed.

I am also tempted to drop the use of libpq's PQprint so that it can be
phased out, since I suspect hardly anyone else uses it and no one is
really happy with it. One could perhaps put a big #ifdef around that code
then. We'll see what happens. If this development would break your heart,
please yell.

Any other suggestions before I lock myself into my room are welcome.

Peter

-- 
Peter Eisentraut - peter_e@gmx.net
http://yi.org/peter-e




Re: [HACKERS] psql code to be obducted by alien (me)

От
Tom Lane
Дата:
Peter Eisentraut <peter_e@gmx.net> writes:
> I am also tempted to drop the use of libpq's PQprint so that it can be
> phased out, since I suspect hardly anyone else uses it and no one is
> really happy with it.

If you write new printing code for psql, please consider making it a
separate module that could be included into libpq so other applications
can use it.

Another part of psql that should be made as independent as possible
is the support for \copy.  I recall a number of people asking in the
past how they can read and write tables to files in their own apps.
There's not that much code involved, but psql is such a mess that it's
hard to point to a chunk of code they can borrow.

BTW, something closely related to \copy that's languishing on the TODO
list is the ability to load the contents of a local file into a Large
Object or write the data out again.  This would be the equivalent of the
server-side operations lo_import and lo_export, but reading or writing a
file in psql's environment instead of the backend's.  Basically a wrapper
around lo_read/lo_write, not much to it but it needs done...

Anyway, I guess the point of all this is that psql should be not only
a useful app in its own right, but a source of how-to examples and
borrowable code for people making their own client apps.  I think you
will move it a long way in that direction just by doing cleanup, but
please keep the notion in mind while you work.
        regards, tom lane


Re: [HACKERS] psql code to be obducted by alien (me)

От
Bruce Momjian
Дата:
Happy coding...  :-)


> As authorized by Tom L. himself and barring any loud protests I will start
> a code clean-up on psql sometime within the next few days. So if you are
> doing something on it or plan on doing so, please let me know. Among the
> things on the agenda are:
> 
> * Reduce function size, indentation levels, file sizes
> 
> * Make use of latest libpq facilities
> 
> * Take care of various TODO items, such as NULL display (perhaps the one
> recently submitted by me?), \copy issues, possibly more.
> 
> * Allow for implementation of a more sophisticated readline TAB
> completion. (Not necessarily the one I recently sent in, if you can come
> up with a better one. I'll try to keep it general.)
> 
> * Have tables vs. views show up correctly, as explained to me by Jan.
> 
> * Various enhancements on HTML display.
> 
> * A full bag of other ideas which will probably have to be postponed.
> 
> I am also tempted to drop the use of libpq's PQprint so that it can be
> phased out, since I suspect hardly anyone else uses it and no one is
> really happy with it. One could perhaps put a big #ifdef around that code
> then. We'll see what happens. If this development would break your heart,
> please yell.
> 
> Any other suggestions before I lock myself into my room are welcome.
> 
> Peter
> 
> -- 
> Peter Eisentraut - peter_e@gmx.net
> http://yi.org/peter-e
> 
> 
> 
> ************
> 


--  Bruce Momjian                        |  http://www.op.net/~candle maillist@candle.pha.pa.us            |  (610)
853-3000+  If your life is a hard drive,     |  830 Blythe Avenue +  Christ can be your backup.        |  Drexel Hill,
Pennsylvania19026