Date:         Wed, 8 Aug 2007 18:54:09 -0400
Reply-To:     Arthur Tabachneck <art297@NETSCAPE.NET>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Arthur Tabachneck <art297@NETSCAPE.NET>
Subject:      Re: no error msg for dlm=dsd
Comments: To: "Howard Schreier <hs AT dc-sug DOT org>" <nospam@HOWLES.COM>

I agree with Howard's interpretation that the characters following the DLM
option are interpreted as representing a variable.

Try the following test:

data test1;
  infile cards dlm=ABC;
  input x $ y;
  cards;
ABCDEF 1
BCDEFG 2
ABC 3
;
run;

data test2;
  infile cards dlm=ABC;
  input x $ y;
  ABC=' ';
  cards;
ABCDEF 1
BCDEFG 2
ABC 3
;
run;

data test3;
  infile cards dlm=ABC;
  input x $ y;
  ABC='B';
  cards;
ABCDEF 1
BCDEFG 2
ABC 3
;
run;

Art
-------
On Wed, 8 Aug 2007 18:27:29 -0400, Howard Schreier <hs AT dc-sug DOT org>
<nospam@HOWLES.COM> wrote:

>On Wed, 8 Aug 2007 14:50:21 -0700, Jack Hamilton <jfh@STANFORDALUMNI.ORG>
wrote:
>
>>Probably a bug.
>>
>>You can specify dlm='dsd' if you want those characters to be delimiters.
>>
>>
>>
>>adac1121@GMAIL.COM wrote:
>>> I'd like to know why dlm=dsd didn't get error message
>>> dsd is an option not one of dlm= value should be '  ', '09'x, ','
>>> should be...dlm='09'x dsd;
>>> Thanks,
>
>The syntax is OK without the quotes because the DLM= option can designate
a
>variable instead of a literal.
>
>I ran (using 9.1.3)
>
>   data _null_;
>   infile cards dlm=dsd;
>   input a b c;
>   put _all_;
>   cards;
>   1 2 3
>   ;
>
>Result: dsd=  a=1 b=2 c=3 _ERROR_=0 _N_=1
>
>I was surprised at the absence of a variable-is-uninitialized note. I
would
>expect variable DSD to be created as a *numeric*, possibly making the a
>period (derived from the missing value) one of the delimiters, or
triggering
>some kind of ugliness.