Date:         Fri, 19 Nov 1999 16:45:09 -0500
Reply-To:     Schechter Robert RS <robert.schecter@PHWILM.ZENECA.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
Comments:     To: "WHITLOI1@westat.com" <WHITLOI1@westat.com>
From:         Schechter Robert RS <robert.schecter@PHWILM.ZENECA.COM>
Subject:      Re: I'll take SAS History for 200, Alex.
Comments: To: SAS-L@LISTSERV.VT.EDU
Content-Type: text/plain; charset="iso-8859-1"

Yes, it is only the presents of the semicolon that counts.
1    Data; input x; cards;

NOTE: The data set WORK.DATA1 has 3 observations and 1 variables.
NOTE: The DATA statement used 1.42 seconds.


5     /* Or no statement; at all */
6    proc
7    print; run;

NOTE: The PROCEDURE PRINT used 0.44 seconds.

> ----------
> From:         WHITLOI1@westat.com[SMTP:WHITLOI1@westat.com]
> Sent:         Friday, November 19, 1999 4:29 PM
> To:   Schechter Robert RS
> Subject:      Re[2]: I'll take SAS History for 200, Alex.
>
> Robert,
>
> This follows your rule but isn't a statement.  Put it in quotes and you
> could
> claim it isn't a semi-colon, but just a quoted semi-colon.
>
> 1222
> 1223  data w ; input ; cards ;
>
> NOTE: The data set WORK.W has 0 observations and 0 variables.
> NOTE: DATA statement used:
>       real time           0.06 seconds
>
> 1224  /* Or no statement; at all */
>
> Ian Whitlock <whitloi1@westat.com>
>
>
> ____________________Reply Separator____________________
> Subject:    Re: I'll take SAS History for 200, Alex.
> Author: Schechter Robert RS <robert.schecter@PHWILM.ZENECA.COM>
> Date:       11/19/1999 3:57 PM
>
> Yes, I believe you're missing a lot. It could include any of
> Ron's step boundaries as well:
>
>
> 1. run;
> 2. semicolon(s) after Cards statement: [at least] one after Cards, four
> after Cards4 (two if by sea)
> 3. data [xxxx];
> 4. proc [xxxx];
> 5. endsas;
> 6. end of input file in noninteractive or batch
>
> My only point was the rule is it needs a semicolon on the next line, or as
> I'm corrected an end of input file in noninteractive or batch
>
>
>
> PROC xxxx;
> DATA  xxxx;
> %macro name;
>
> > ----------
> > From:         Lund, Pete[SMTP:Peter.Lund@cfc.wa.gov]
> > Sent:         Friday, November 19, 1999 3:37 PM
> > To:   Schechter Robert RS; SAS-L@LISTSERV.UGA.EDU
> > Subject:      RE: I'll take SAS History for 200, Alex.
> >
> > Bob-
> >   Good point of clarification.  Data following a CARDS statement is
> > terminated by a line with a ; on it.  AFAIK, that can only be a null
> line
> > (;), a RUN; or a *...; comment.  Any others I'm missing?
> >
> > Pete Lund
> > WA State Caseload Forecast Council
> > (360) 902-0086     voice
> > (360) 902-0084     fax
> > peter.lund@cfc.wa.gov
> >
> >
> > -----Original Message-----
> > From: Schechter Robert RS [mailto:robert.schecter@PHWILM.ZENECA.COM]
> > Sent: Friday, November 19, 1999 11:51 AM
> > To: SAS-L@LISTSERV.UGA.EDU
> > Subject: Re: I'll take SAS History for 200, Alex.
> >
> >
> > No I believe your wrong. The rule is having the ; on the line, not
> > necessarily a step boundary. the SAS comment:
> > * this is the end of my data;
> > while not a step boundary would work just fine because of the obligatory
> > semi-colon.
> >
> > > ----------
> > > From:         Lund, Pete[SMTP:Peter.Lund@CFC.WA.GOV]
> > > Sent:         Friday, November 19, 1999 2:34 PM
> > > To:   SAS-L@LISTSERV.UGA.EDU
> > > Subject:      Re: I'll take SAS History for 200, Alex.
> > >
> > > There is a little "gothca" to the lack of a semicolon following
> > > CARDS/DATALINES data.  Usually, the combination of the two tokens RUN
> > and
> > > ;
> > > form a step boundary.  As long as they occur sequentially it doesn't
> > > matter
> > > if there are spaces or lines between them.  For example, the following
> > > code
> > > works just fine:
> > >
> > >    data test;
> > >      put 'this is a test making test';
> > >      x = _n_;
> > >    run
> > >
> > >    ;
> > >
> > > However, if the RUN is coming after CARDS/DATALINES data, the RUN and
> > the
> > > ;
> > > *must* be on the same physical line.  For example:
> > >
> > >    data test;
> > >      input x;
> > >      cards;
> > >    1
> > >    2
> > >    3
> > >    run
> > >    ;
> > >
> > > This causes an error as the value "run" is trying to be read into the
> > > numeric variable x.  Putting the RUN and ; on the same line solves the
> > > "problem" as does inserting a null line (;) between 3 and RUN.
> > >
> > > If I had to take my shot at Ron's question - the six step boundaries
> are
> > > RUN;, PROC, DATA, %MACRO, %MEND;, and ; when following CARDS or
> > DATALINES
> > > data.
> > >
> > > Pete Lund
> > > WA State Caseload Forecast Council
> > > (360) 902-0086     voice
> > > (360) 902-0084     fax
> > > peter.lund@cfc.wa.gov
> > >
> > >
> > > -----Original Message-----
> > > From: Fehd, Ronald J. [mailto:rjf2@CDC.GOV]
> > > Sent: Friday, November 19, 1999 11:01 AM
> > > To: SAS-L@LISTSERV.UGA.EDU
> > > Subject: Re: I'll take SAS History for 200, Alex.
> > >
> > >
> > > From: Lambert, Bob [mailto:Bob_Lambert@AFCC.COM]
> > > /Why is semicolon needed at end of cards statement?
> > >
> > > What SAS statements are step boundaries? [hint: six items]
> > >
> > > Ya gotta know this stuff in order to do macros.
> > >
> > > Ron Fehd  that maven  CDC Atlanta GA
> > >
> >
>