Date: Fri, 20 Apr 2001 09:14:49 +0100
Reply-To: Peter Crawford <Peter@CRAWFORDSOFTWARE.DEMON.CO.UK>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Peter Crawford <Peter@CRAWFORDSOFTWARE.DEMON.CO.UK>
Organization: Crawford Software Consultancy Limited
Subject: Re: flat files and end-of-file characters
This is probably a case for that recent technical document from SI about
reading mvs files over FTP. Download it from the SI with FTP link
Reading EBCDIC Files on ASCII Systems
This Technical document goes into details about special solutions for
reading variable length data files which have to be read without the
EBCDIC to ASCII conversion provided by default FTP downloads (I'm
talking binary or type= image).
There seems to be quite a bit more to it than just using binary transfer
and S370Fxxx and $EBCDIC informats.
Fehd, Ronald J. <rjf2@CDC.GOV> writes
>> From: Miller, Jeremy T. [mailto:zyp9@CDC.GOV]
>> I am reading in some very large flat file with SAS. For
>> example, maybe one
>> file has 400k records, however, SAS is only getting maybe 1/3
>> of that. Just
>> to check, I open the flat file with Ultra-Edit and look at
>> the last record
>> SAS read in. It seems that there are end-of-file characters
>> embedded within
>> the ASCII data (e.g., G 07CC107091D1). So, far, I have just
>> been replacing
>> them within Ultra-Edit and updating the file. Is there any
>> way to have SAS
>> ignore the squares when reading in the data?
>another happy UE user! Hi!
>I'm not sure about getting past the end-of-file
>on WinDoze you'll want to be using the infile options
>pad lrecl=<your logical record length>
>you can get the lrecl after the first time you run your job by looking at
>the notes sections
>SAS tells you the longest and shortest records.
>set lrecl to the max value seen.
>so aside from kludging the data files your self
>why are you accepting, let alone receiving, such cobbled files?