I've looked around and not found a complete, start to finish, guide on how to connect to a flat file as a data source in webfocus 5.25.

Here is our situation. We have an excel file with 5 columns on a windows machine and a unix server which is running both webfocus and our webserver. I need to know:

1) What format from excel will work best (.csv from excel does not put in quotes).
2) Where on the server does the flat file need to reside.
3) How do I verify the master file is correct.

I figured out how to make the master file and can write a .fex that can read the master file. The problem is that it returns 0 rows. I suspect that it is not even opening the file to look at it.

As always, any help would be appreciated.

George, you can export from excel to csv, with quotes, the export type you should select is CSV(MS-DOS), its not the first csv choice in the list, look down ****her. The first choice CSV (comma delimited) won't work, you're right. Then all you need is the master. If you have column headers in your export, use SUFFIX=COMT
To read excel directly, you need to have specified a range in the actual worksheet, giving it a range name. There are some articles on the techsupport site on reading excel directly.

quote:

Originally posted by susannah:
[qb] George, you can export from excel to csv, with quotes, the export type you should select is CSV(MS-DOS), its not the first csv choice in the list, look down ****her.
[/qb]

Are you certain? I have excel 2002 and when I output to csv(msdos) and open it in a text editor it still does not have quotation marks.

quote:

[qb]Then all you need is the master.[/qb]

I'm 85% sure my master file is okay.
FILE=phone_dist, SUFFIX=COM
SEGNAME=ROOT_SEG, SEGTYPE=S1, $
FIELD=ACCT_UNIT, ALIAS=ACCT_UNIT, USAGE=I5, ACTUAL=I4, MISSING=ON, TITLE='Account ,Number', DESCRIPTION="Account Number charged for this department's phone charges", $
FIELD=NAME, ALIAS=NAME, USAGE=A20, ACTUAL=A35, MISSING=ON, TITLE='Name', DESCRIPTION='Name of the Account Manager assigned to their account number', $
FIELD=DEPT, ALIAS=DEPT, USAGE=A20, ACTUAL=A50, MISSING=ON, TITLE='Department', DESCRIPTION='Department associated with the account number and account manager', $
FIELD=ROOM, ALIAS=ROOM, USAGE=A20, ACTUAL=A50, TITLE='Room', $
FIELD=CAMPUS, ALIAS=CAMPUS, USAGE=A20, ACTUAL=A10, TITLE='Campus', $
FIELD=EMAIL, ALIAS=EMAIL, USAGE=A20, ACTUAL=A50, MISSING=ON, TITLE='Email,Address', DESCRIPTION='Email Address of the account manager for this account', $

quote:

[qb]To read excel directly, you need to have specified a range in the actual worksheet, giving it a range name. There are some articles on the techsupport site on reading excel directly. [/qb]

Is this possible on a unix server? I've heard that it is not an option in our environment.

well, works for me; but any two environments are gonna be different. Anyway, it shouldn't make a difference.
What about your filedef statement? are you sure its pointing to the .csv file?
Your 0 lines error can happen if your master is fine, but your filedef is pointing to pluto.
issue a check on the mfd
CHECK FILE MYFILE
FILEDEF MYFILE DISK D:\IBI\...\myfile.csv
TABLE FILE MYFILE
PRINT *
END
--caveat: i'm win2k server, don't know beans about unix filedefs

The file def was the issue. I didn't have it pointed in the right area. Thank all of you for your help.