is there any way to flexibly print with 'OVER' according to a value in a column of a table? eg:
-set &overtxt = 'over';
TABLE FILE XXX
PRINT COL1
COL2 &overtxt
COL3

I want to flexibly change that 'OVER' text to '' depending on the value of another column, say, col5?

Mike

The simple answer is NO. You cannot mix Dialogue Manager and TABLE in this way.

The complex answer is a question. What do you hope to achieve? Can you show a layout you want based on, say, the CAR file. If we can understand your requirements, then a solution may be possible.

it is a long story to this and I cannot use the car file as a sample. The basic idea is this:
there is this column in this audit table. It is variable length, containing changed info for some other table. It can be for a row add, a row delete, or an row updat row.

For added/deleted row, it starts with add/del status and strings all the column values together with a delimiter. For update row, it starts with upd status and strings with old and new values of the changed (only) columns together with a delimiter. The total possible columns are 86. What we are required to do is: for add/del, we print two columns in one row. For update, we print the changed column in one row with old and new values. So, when I map out the table file print with

print col1 col2 over
col3 col4 over
col5 col6 over...

the add/del takes up two pages, but the updates will usually have only one page's worth of data (only a few valus changes), leaving the second page blank. So, I want to stop that 'over' after a certain place so that blank can overlay blank and no second blank page generated under 'update'.

Mike, Inspite of what others have said, YES you can do it and with DIALOG MANAGER but NOT in the way that you are trying to do it, BUT with one extra step. It will also depend upon the TIME that you have to render the report. i.e. must it render in 5 seconds or 6 seconds

What I do AND I SHALL BE SPEAKING ON THE SUBJECT AT THE SUMMIT, is to create a focexec ON THE FLY,
What you do is to
Create a focexec that you will call in your main program (SEE FILEDEF below)
FILEDEF mydynamicfile DISK C:\ibi\apps\financialreports\mydynamicfile.fex
-RUN
-SET &MYVARIABLE = 'E07';
This is your variable field (E07) that YOU MENTIONED

NOW FROM THE MAIN PROGRAM DO A

-WRITE MYDYNAMICFILE TABLE FILE XXX
-WRITE MYDYNAMICFILE PRINT E01 E02 &MYVARIABLE
-WRITE MYDYNAMICFILE END
THEN
further down in your code you do
-INCLUDE MYDYNAMICFILE


Please give that a try and if you need more help contact me at njm870t@hotmail.com
sincerest regards Steve

This message has been edited. Last edited by: Stefaans,

Hi Steve: the problem is that the variable value has to depend on the value of a column in the table itself during the TABLE File command. So, the setting of the var value has to done during the TABLE file itself.

I have a stripped-down of the fex below for your reference. I am talking about to variate the 'OVER' text at the last compute statement depending on the DATYPE.

DEFINE FILE AUSALSTA
DATYPE/A1 = SUBSTR(2400, COL_DATA, 1, 1, 1, DATYPE);
END

TABLE FILE AUSALSTA
PRINT
COMPUTE DATALEN/I4 = POSIT(COL_DATA, 2400, '&|$&|', 5, DATALEN); NOPRINT
COMPUTE SUB_DATA/A2400V = SUBSTR(2400, COL_DATA, 10, DATALEN, 2400, SUB_DATA); NOPRINT

-SET &NUM = 0;
-REPEAT LOOP 42 TIMES

-SET &NUM = &NUM + 1;
-SET &NXT = &NUM + 1;
COMPUTE TOKEN_&NUM.EVAL/A40 = GETTOK(SUB_DATA, 2400, &NUM, '%', 40, TOKEN_&NUM.EVAL); NOPRINT
COMPUTE VSTR_&NUM.EVAL/A37 = SUBSTR(40, TOKEN_&NUM.EVAL, 5, 40, 37, VSTR_&NUM.EVAL); IN 20 AS ''

COMPUTE TOKEN_&NXT.EVAL/A40 = GETTOK(SUB_DATA, 2400, &NXT, '%', 40, TOKEN_&NXT.EVAL); NOPRINT
COMPUTE VSTR_&NXT.EVAL/A37 = SUBSTR(40, TOKEN_&NXT.EVAL, 5, 40, 37, VSTR_&NXT.EVAL); IN 57 AS '' OVER

TABLE_STATUS NOPRINT

-SET &NUM = &NUM + 1;
-LOOP

ON TABLE PCHOLD FORMAT PDF

END

Hey Mike
Great to hear from you.
Please forgive me if I am missing something.

Indications are that your first character in the record determines the data type. Great.
1) Are all the records of one type say per run?
2) I assume that they are as the overtext is listed once
why not table out the record to a save or hold file, then read the file to asceratin the DATYPE.
The -SET &MYOVERTEXT = IF DATTYPE EQ 'A'
- THEN 'AddaRecord' ELSE
- IF....
Then go the route that I suggested in my earlier note.
It would be easier if I new exactly what you were trying to do, because I am sure that I have done something similar in past.

If you would like to take this one offline, contact me by email and perhaps we can have a quick chat.
regards Steve

This message has been edited. Last edited by: Stefaans,

Steve

When you say:

quote:

Mike, Inspite of what others have said, YES you can do it and with DIALOG MANAGER but NOT in the way that you are trying to do it

Please read what I actually said:

quote:

The simple answer is NO. You cannot mix Dialogue Manager and TABLE in this way.

I, and many others here, are aware that a focexec can be created in DM, which is why I asked for a bit more explanation before trying to provide an answer.

Mike

I would try a different approach altogether. From the focexec snippet you appear to have a file delimited with a % sign. Look at creating an MFD over it with SUFFIX=DFIX, a delimter of % and taking advantage of the MISSING attribute, using fieldnames such as F1,F2 etc to facilitate the build of the focexec in DM.

Then construct the focexec using a CNTR for each record and use syntax like

ON CNTR SUBFOOT
"<F1 <F2 "
WHEN F2 NE MISSING;
ON CNTR SUBFOOT
"<F3 <F4 "
WHEN F4 NE MISSING;

etc.

The only issues are aligning the subfoot items and getting rid of the line delimiter, but those should be doable.

This message has been edited. Last edited by: Alan B,

Hi Alan: that is an interesting concept. The subfoot trick is worthy of exploring. I need to clarify that it is not a 'file' delimited with a % sign. It is a column in a table delimited with a % sign. It looks a file to you because it is a snippet and I pretty much took out the code to unstring them. So, I am not sure if I need the MFD part. However, before I try it, I would need clarification from you about this:

creating an MFD over it with SUFFIX=DFIX, a delimter of %

what does "SUFFIX=DFIX, a delimter of %" mean? what does it look like actually in code? Thanks in advance.

Mike

Try this link

Basically it is a method of reading a non-fixed format file base on a delimiter between each field. It also seems to cope with variable number of fields, or has done in my tests.

Hi Alan: read the link doc. I don't know if I can do a MFD on it. Our current MFD looks like this (simplified):
FILENAME=AUDITINFO, SUFFIX=DB2 , $
SEGMENT=AUDITINFO, SEGTYPE=S0, $
FIELDNAME=TBL_CODE, ALIAS=TBL_CODE, USAGE=A2, ACTUAL=A2, $
FIELDNAME=PROCESS_DATE_TIME, ALIAS=PROCESS_DATE_TIME, USAGE=HYYMDm, ACTUAL=HYYMDm, $
FIELDNAME=COL_DATA, ALIAS=COL_DATA, USAGE=A2400V, ACTUAL=A2400V, $


Note the last field, col_data. That is where everything is delimited into and each delimited value can be different format(date, int, etc) and it can be any different combination sequence.
I deduct the format of each delimited value by the first two chars during the TABLE FILE command. So, I can't beforehand have a fixed format of definition of the col_data.

Mike

Perhaps you need to do this in 2 passes of the data then.

You were successfully (I presume) getting the data out.

Can you hold the data delimited (not actually fixed format) from the first run, apply a DFIX master over it, and then run the report to output. How does that sound?

I am aware of the two pass possibility. I am just hung up on how to do a consistent MFD for the hold file because on the first record, the record layout can be:
fld1 char (10)
fld2 char (20)
fld3 char (20)


but the second record can be
fld1 char (16)
fld2 char (10)
and so forth
It all depends on the add/update status and on which column fields the update to the table was done (this is basically an audit trace record). There is no uniform pattern to code a MFD for the hold file.... But... I suppose I can allow for max length and max num of fields though that will substantially increase my hold file size. Maybe performance impact? hummm..
Maybe I can twig it somehow using subfoor without hold file two pass?

This message has been edited. Last edited by: mike,

Mike

Declare the field to be the maximum it can be for either record, DFIX will deal with it. The file can have variable length records. Perhaps the name of DFIX is putting you off, there needs to be nothing FIXed about the data file.

FILEDEF MASTER DISK TEST450.MAS
-RUN

-WRITE MASTER FILENAME=TEST450, SUFFIX=DFIX
-WRITE MASTER SEGNAME=TEST450
-WRITE MASTER FIELDNAME=F1       ,ALIAS=F1  ,USAGE=A40 ,ACTUAL=A40 ,$
-WRITE MASTER FIELDNAME=F2       ,ALIAS=F2  ,USAGE=A40 ,ACTUAL=A40 ,MISSING=ON, $
-WRITE MASTER FIELDNAME=F3       ,ALIAS=F3  ,USAGE=A40 ,ACTUAL=A40 ,MISSING=ON, $
-WRITE MASTER FIELDNAME=F4       ,ALIAS=F4  ,USAGE=A40 ,ACTUAL=A40 ,MISSING=ON, $
-WRITE MASTER FIELDNAME=F5       ,ALIAS=F5  ,USAGE=A40 ,ACTUAL=A40 ,MISSING=ON, $
-WRITE MASTER FIELDNAME=F6       ,ALIAS=F6  ,USAGE=A40 ,ACTUAL=A40 ,MISSING=ON, $
-WRITE MASTER FIELDNAME=DELIMITER ,ALIAS='%' ,USAGE=A1  ,ACTUAL=A1  ,$
FILEDEF TEST450 DISK TEST450.DAT
-RUN

-WRITE TEST450 1Lorem ipsum %dolor sit amet,% consectetur adipiscing e%lit, set eiusmod te%mpor incidunt et labore %et dolore magna
-WRITE TEST450 2Plloaso mako nuto u%f cakso do%dtos anr koop a cupy uf cak v%ux noaw yerw phuno

DEFINE FILE TEST450
DATYPE/A1 = SUBSTR(40,F1, 1, 1, 1, DATYPE);
CNTR/I4 WITH F1 = CNTR+1;
END

TABLE FILE TEST450
PRINT DATYPE NOPRINT
BY CNTR NOPRINT
ON CNTR PAGE-BREAK
ON CNTR SUBFOOT
"<F1 =====<F2 "
WHEN F2 NE MISSING;
ON CNTR SUBFOOT
"<F3 =====<F4 "
WHEN F4 NE MISSING;
ON CNTR SUBFOOT
"<F5 =====<F6 "
WHEN F6 NE MISSING;
ON TABLE PCHOLD FORMAT PDF
END

Thanks, Alan! That works like a charm! By the way, I was also able to work around it without a two-pass hold file! Thanks!