RE: Speaking of New Features
- From: "Powell, Mark D" <mark.powell@xxxxxxx>
- To: "oracle-l" <oracle-l@xxxxxxxxxxxxx>
- Date: Thu, 10 Sep 2009 15:41:47 -0400
I think the pl/sql parameter list format would be better since with
comments in front of the values clause you do not eliminate the need to
also have a separate column name list so you can insert the values in a
non-default order or omit some of them and accept null for their value.
Worse a typing error could result in the comments and the values not
matching up properly.
-- Mark D Powell --
Phone (313) 592-5148
________________________________
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of chet justice
Sent: Thursday, September 10, 2009 2:46 PM
To: daniel.fink@xxxxxxxxxxxxxx
Cc: oracle-l
Subject: Re: Speaking of New Features
True, that would work. It's just ugly though...in my opinion
anyway. :)
On Thu, Sep 10, 2009 at 2:38 PM, Daniel Fink
<daniel.fink@xxxxxxxxxxxxxx> wrote:
In the absence of actual implementation, comments are
your friend. Why not use comments to indicate which column you are
referencing? Granted it does not totally address the situation of
specifying a limited number of columns or and independent order, but it
would help when inserting 100 columns or so.
Pre-column
DEMO@dwf10gr2> insert into t2
2 values ( /* c1 */ 12,
3 /* c2 */ 42
4 )
5 /
1 row created.
Post-column
DEMO@dwf10gr2> insert into t2
2 values ( 12, -- c1
3 42 -- c2
4* )
DEMO@dwf10gr2> /
1 row created.
Regards,
Daniel Fink
--
Daniel Fink
OptimalDBA http://www.optimaldba.com
Oracle Blog http://optimaldba.blogspot.com
Lost Data? http://www.ora600.be/
chet justice wrote:
I think I would require the use of the
correct column name instead of any type of positional col-n style
labeling.
Agreed. That was just an example, those are
the actual column names.
On Thu, Sep 10, 2009 at 1:47 PM, Powell, Mark D
<mark.powell@xxxxxxx> wrote:
Well, the suggested syntax below would
make matching up a long column list to the provided values/variables a
lot easier and would likely help prevent listing 100 columns to be
inserted but only including 99 variables in the values list. I think I
would require the use of the correct column name instead of any type of
positional col-n style labeling.
-- Mark D Powell --
Phone (313) 592-5148
________________________________
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Jared Still
Sent: Thursday, September 10, 2009 12:04
PM
To: chet.justice@xxxxxxxxx
Cc: oracle-l
Subject: Re: Speaking of New Features
Very Perlish.
I like it. :)
Jared Still
Certifiable Oracle DBA and Part Time
Perl Evangelist
On Wed, Sep 9, 2009 at 6:37 PM, chet
justice <chet.justice@xxxxxxxxx> wrote:
Any thoughts on the "new" syntax for
INSERT statements below?
INSERT INTO my_table
( id => seq.nexval,
create_date => SYSDATE,
update_date => SYSDATE,
col1 => 'A',
col2 => 'SOMETHING',
col3 => 'SOMETHING',
col4 => 'SOMETHING',
col5 => 'SOMETHING',
col6 => 'SOMETHING',
col7 => 'SOMETHING',
col8 => 'SOMETHING',
col9 => 'SOMETHING',
col10 => 'SOMETHING',
col11 => 'SOMETHING',
col12 => 'SOMETHING',
col13 => 'SOMETHING',
col14 => 'SOMETHING' );
Thought of one day while trying to clean
up (make human readable) someone else's code. I would either get too
many values or not enough. After copying the INSERT columns and
subsequent VALUES clause into an Excel spreadsheet to compare them side
by side, I thought, hey, what about named notation?
Anyway, I created the "Idea" on Oracle
Mix here <https://mix.oracle.com/ideas/94278-position-insert-syntax> if
you are inclined to, one way or another, to vote.
chet
--
chet justice
www.oraclenerd.com
--
chet justice
www.oraclenerd.com
--
chet justice
www.oraclenerd.com
Other related posts:
