How to import text with space delimiter and space is valid in the field?

[neilxu]

I have the following text:

id1 string1 comment string2
001 001 qwer asd 5
002 002 asdfyuiop 6
003 003 zxcv 5

The problem is that space is the delimiter and also is the valid in comments. How can I import these?

Thanks

[sbb]

You will need to use a DATA step and parse the first, second fields, and then after finding the offset character to the third field using INDEXC possibly, then use SUBSTR along with the SAS special variable _INFILE_ to assign the final variable.

Example:

DATA ;
INFILE /* additional parameters as needed */ ;
* LOAD THE INPUT BUFFER. ;
INPUT ;
* LENGTH OR ATTRIB statements go here. ;
F1 = SCAN(_INFILE_,1," ");
F2 = SCAN(_INFILE_,2," ");
OFFSET = INDEXC(_INFILE_," ");
OFFSET = INDEXC(SUBSTR(_INFILE_,OFFSET+1)," ");
F3 = SUBSTR(_INFILE_,OFFSET+1);
PUTLOG _ALL_;
RUN;

Scott Barry
SBBWorks, Inc.

[Ksharp]

Hi.
You can input 'comment' and 'string2' as one variable firstly, then parse this variable into two variables you need. try it.
[pre]
data temp;
input d1 $ string1 $ comment & $50.;
_comment=substr(comment,1,length(comment)-2);
string2=substr(comment,length(comment)-1);
drop comment;
datalines;
001 001 qwer asd 5
002 002 asdfyuiop 6
003 003 zxcv 5
;
run;

proc print noobs;
run;

[/pre]



Ksharp

[sbb]

INPUT logic in Ksharp reply supports a max of 1 blank-space between words in the comments.

Scott Barry
SBBWorks, Inc.

[Ksharp]

Emmm.Yes.
If you have more than two blanks, you can use formatted input method.


[pre]
data temp;
infile datalines truncover;
input d1 $ string1 $ comment $50.;
_comment=substr(comment,1,anydigit(comment)-1);
string2=scan(comment,-1);
drop comment;
datalines;
001 001 qwer asd 5
002 002 asdfyuiop 6
003 003 zxcv sdit 5
;
run;

proc print noobs;
run;

[/pre]


P.S : The company is the one I worked for , not mine. 🙂


Ksharp Message was edited by: Ksharp

[Peter_C]

neilxu

what are the rules that define the end of "comment"?

peterC