Date:         Sat, 13 Jan 2001 17:17:37 -0000
Reply-To:     DKVJ Cons Co UK <secretary@dkvj-cons.co.uk>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         DKVJ Cons Co UK <dkv-j_consultancies@TOTALISE.CO.UK>
Organization: DKV-J Consultancies
Subject:      Reading MS OE5 email folder with SAS.
Content-Type: text/plain; charset="iso-8859-1"

I receive email on a regular basis, and have a SAS program that can parse
the text contents of each message to interpret the information sent.

I now want to create a fully automated process to read incoming messages and
act appropriately.

The email client is MicroSoft Outlook Express 5. The messages are identified
by a 'rule' or 'agent' process and passed to a particular folder, where they
are stored in a system folder with type 'dbx'.

I haven't been able to import this DBX file into anything except Outlook
Express, and the only File Export functions OE5 has for message folders is
for Outlook or MSMail.  So, no option to create a text file!!

I have tried to read and parse the file using the varying length parameters
on a data step 'Infile' process.  The read, however stops after 13k bytes
due to what is probably a character SAS interprets as an end of file marker.
The file is in fact 10MB in size.

Surely someone has successfully read email into SAS, and possible even with
MS email applications.  Have you successfully exported the folder to a
standard database file?  Or have you read the file in and managed to skip
the rogue character that is stopping my read?

Incidentally, the Import Wizard fails in both versions 6 and 8 with these
problems:
.    rename as DBF and import as DBF: fails with insufficient memory,
despite NT4 and 320MB RAM available.
.    import as special file: only the first 13k or so characters appear in
the buffer for definition of the data set vector.

Some code and the log below for those interested:

Regards

David


75   FileName EMLFLDER 'E:\Archive\Outlook mail\*********.dbx';
76
77   Data READJSV;
78     InFile EMLFLDER  TruncOver  RECFM = V  LENGTH = RLENGTH   LRECL =
13322;
79     Input @1 STRING $1 @;
80     Put 'Is the LRECL set correctly? 'RLENGTH=;
81     Do Loop = 1 To RLENGTH  By 1;
82       Input @ LOOP   STRING01  $Char1. @;
83       I ++ 1;
84       BYTEVAL = Rank( STRING01);
85       If  I > 13300  Then
86           Put  LOOP=  STRING01 $Char1.  ' '  BYTEVAL=;
87       Output;
88     End;
89   Run;

NOTE: The infile EMLFLDER is:
      FILENAME=E:\Archive\Outlook mail\*********.dbx,
      RECFM=V,LRECL=13322

Is the LRECL set correctly? RLENGTH=13321
LOOP=13301   BYTEVAL=18
LOOP=13302 { BYTEVAL=123
LOOP=13303   BYTEVAL=0
LOOP=13304   BYTEVAL=0
LOOP=13305   BYTEVAL=19
LOOP=13306 f BYTEVAL=131
LOOP=13307   BYTEVAL=0
LOOP=13308   BYTEVAL=0
LOOP=13309   BYTEVAL=20
LOOP=13310 " BYTEVAL=147
LOOP=13311   BYTEVAL=0
LOOP=13312   BYTEVAL=0
LOOP=13313   BYTEVAL=28
LOOP=13314 ¥ BYTEVAL=165
LOOP=13315   BYTEVAL=0
LOOP=13316   BYTEVAL=0
LOOP=13317   BYTEVAL=0
LOOP=13318   BYTEVAL=22
LOOP=13319 D BYTEVAL=68
LOOP=13320   BYTEVAL=16
LOOP=13321 a BYTEVAL=97
NOTE: 1 record was read from the infile EMLFLDER.
      The minimum record length was 13321.
      The maximum record length was 13321.
NOTE: The data set WORK.READJSV has 13321 observations and 5 variables.
NOTE: The DATA statement used 0.79 seconds.