Date: Thu, 20 Dec 2001 13:09:39 +0100
Reply-To: Peter Crawford <peter.crawford@DB.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Peter Crawford <peter.crawford@DB.COM>
Subject: Re: Reading File
Content-type: text/plain; charset=iso-8859-1
Kathy Samson <Kathy.Samson@eds.com> wrote in
> Ok, so I'm jumping on this bandwagon because I'm looking at the same thing, how to
> easily handle a mixed bag of delimited files. Lets say, MVS with version 6.08 or
> version 8, input & output. Something similar to specifying the delimited option on the
> infile statement. Anything similar on the output side?
On v6 you have to organise your own delimiters, buton v8 there
is an option for the file statement DSD which is similar to the
same option on the infile statement. The default delimiter
(again) is a comma, but can be overridden to semicolon
like
file '.personal.csv.txt' dsd dlm=';' lrecl=2000;
The advantage offered by DSD processing surfaces when
your data might contain a delimiter. If that happens, the normal,
simple handling causes data to appear in the wrong columns.
DSD processing masks a value containing a delimiter with double-
quote (") marks. It is an old standard tolerated by _all_ software
systems, including microsoft, oracle, ibm/lotus.
put ( _all_ ) (:) ;
delivers the values in the data step data vector in varnum order and
presented in their default formats. However, it is probably unsuitable
for binary formats (formats like IB2. RB8. IEEE4. S370FPD8.) which
might contain data mis-interpreted as control characters like end-of-file,
line-feed, vertical-tab, carriage-return etc. on some platforms.
For a generalised solution, it may be enough to format all numbers
with best18. notation, and then all dates with yymmdd10. ... like
format _numeric_ best18. &dates yymmdd10. ;
I recommend yymmdd10. because
it is recognised by semi-intelligent date recognisers (like excel import)
clarifies century
carries no language variations like MAY and MAI (see format eurdfdte. )
is visibly simple to recognise (we are talking data, not web/people )
but you may have another preference.....
(delivering to a non-UK european world of number standards, format best.
may confuse when the decimal fraction separator is expected to be a
comma and the decimal-point/period/dot is considered a separator
for thousands !
On those occasions when the recipient requires this european number
support, best18. is replaced by numx22.8
The DSD option does not place delimiters between quoted strings
in a put statement, like
put "id" 'ac_no' 'name' ;
Clearly this "feature" needs to be overcome to support a heading line
in the .csv file that needs to reflect normal standards. Fortunately
put "id," 'ac_no,' 'name' ;
and
put "id" "," 'ac_no' "," 'name' ;
achieve what is necessary.
When a generated solution is needed from dictionary.columns,
this clip from our sql works
select quote( trim( name ))
into :headings separated by '&dlm'
That &dlm has to be a constant, not only in the sql syntax, but also
when it surfaces in a statement like
put &headings;
So, a suitable value for a tab character between our os/390 and ibm-aix
%let dlm = "05"x
The resolution of this macro variable has to be deferred until &headings
is resolved in the put statement - and not in sql, so single quotes
surround it in the "separated by '&dlm' clause.
that should provide enough of a start point to generate your own
export macro
...........................
Regards
Peter Crawford
Datum: 19/12/2001 23:40
An: SAS-L@LISTSERV.UGA.EDU
Antwort an: "Droogendyk, Harry" <Harry.Droogendyk@CIBC.COM>
Betreff: Re: Reading File
Nachrichtentext:
Kathy:
This solution was posted a couple of days ago for a similar problem, sorry,
can't remember who posted it. The resultant output will be semi-colon
delimited.
data a;
do i = 1 to 5;
do j = 1 to 4;
do k = 1 to 3;
do l = 1 to 2;
output;
end;
end;
end;
end;
run;
data _null_;
set a;
put _n_ (_all_) (";");
run;
-----Original Message-----
From: Paul Choate [mailto:pchoate@JPS.NET]
Sent: December 19, 2001 6:02 PM
To: SAS-L@LISTSERV.UGA.EDU
Subject: Re: Reading File
Hi Kathy,
If you use " put _all_; " the whole data vector is written
with the
variable name and an "=" preceding each. That's almost
delimited.
Of course the Export facility isn't supported on MVS...too
bad.
Kathy Samson <Kathy.Samson@eds.com> wrote in message
news:<3C209B29.6A5A69AC@eds.com>...
> Ok, so I'm jumping on this bandwagon because I'm looking
at the same thing, how to
> easily handle a mixed bag of delimited files. Lets say,
MVS with version 6.08 or
> version 8, input & output. Something similar to
specifying the delimited option on the
> infile statement. Anything similar on the output side?
> Kathy
>
> "Karsten M. Self" wrote:
>
> > on Wed, Dec 19, 2001 at 01:59:39AM +0000, RogersEmail
(rlemay1104@ROGERS.COM) wrote:
> > > I have a delimited file which is constantly written to
by another
> > > program by a multitude of users. I would like to read
it with SAS to
> > > do reports, but when SAS is reading it in, if a user
is trying to
> > > write to the file with a BLAISE application (used for
Computer
> > > Assisted Terminal Interview), it says it is in use and
give me an
> > > error.
> > >
> > > My question is, Is there a way to read a file in SAS
without taking it
> > > exclusively?
> >
> > Probably.
> >
> > Highly dependent on some information you've neglected to
provide, such
> > as the operating system you're using.
> >
> > --
> > Karsten M. Self <kmself@ix.netcom.com>
http://kmself.home.netcom.com/
> > What part of "Gestalt" don't you understand?
Home of the brave
> > http://gestalt-system.sourceforge.net/
Land of the free
> > We freed Dmitry! Boycott Adobe! Repeal the DMCA!
http://www.freesklyarov.org
> > Geek for Hire
http://kmself.home.netcom.com/resume.html
> >
> >
------------------------------------------------------------------------
> > Part 1.2Type: application/pgp-signature
--
Diese E-Mail enthält vertrauliche und/oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet.
This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.
