Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.clipper.visual-objects > #4087 > unrolled thread

Our occasional Reportpro issues

Started byMartin <spam@spam.spam>
First post2012-09-07 14:20 +0100
Last post2012-09-21 12:21 +0100
Articles 3 — 2 participants

Back to article view | Back to comp.lang.clipper.visual-objects


Contents

  Our occasional Reportpro issues Martin <spam@spam.spam> - 2012-09-07 14:20 +0100
    Re: Our occasional Reportpro issues "Stephen Quinn" <stevejqNO@bigpondSPAM.net.au> - 2012-09-08 11:52 +1000
      Re: Our occasional Reportpro issues Martin <spam@spam.spam> - 2012-09-21 12:21 +0100

#4087 — Our occasional Reportpro issues

FromMartin <spam@spam.spam>
Date2012-09-07 14:20 +0100
SubjectOur occasional Reportpro issues
Message-ID<owm2s.518446$mc1.49979@fx12.am4>
Well I have done some more digging and have the following queries

firstly the raw line loading function as called by reportpro is over 
1300 lines long, this creates a lot of text blocks which are then 
returned by other functions called by reportpro in psuedo code....

FUNCTION myprintdesc

if !sameline
    loadtheline()
endif

return desctext
FUNCTION myprintdesc2

if !sameline
    loadtheline()
endif

return desctext2

usually for a page the loader gets called twice per line 1 to n then 1 
to n again.

Is there an issue using such a big function to do my text?

Thanks

Martin

[toc] | [next] | [standalone]


#4088

From"Stephen Quinn" <stevejqNO@bigpondSPAM.net.au>
Date2012-09-08 11:52 +1000
Message-ID<MEx2s.1170$tH2.502@viwinnwfe01.internal.bigpond.com>
In reply to#4087
Martin

What you haven't said is if the reports are hand coded or built in the 
designer.
If hand coded then show some actual code (eg LoadTheLine()).

Looks from your description so far is that some controlling variable isn't 
being set/reset and you don't step onto the next 'batch' record but to the 
top of the same SCOPEd record.

Is it repeatable across the same reports?
Have you tried debugging the 'faulty' record outside the 'batch' situation 
to see what's happening?

A 1300 line method is definately overkill IMHO, but if it works for you<g>.

CYA
Steve 

[toc] | [prev] | [next] | [standalone]


#4122

FromMartin <spam@spam.spam>
Date2012-09-21 12:21 +0100
Message-ID<x4Y6s.229640$pn2.225712@fx12.am4>
In reply to#4088
We are using designer reports

The loading method creates a lot of text blocks which are then used, 
written to work fast as well, so only one load per line rather than many

Line record

description from another file 1
description from another file 2
description from another file 1
description from another file 3
description from another file 3

Would be a block of text.

On 08/09/2012 02:52, Stephen Quinn wrote:

> Martin

>
> What you haven't said is if the reports are hand coded or built in the
> designer.
> If hand coded then show some actual code (eg LoadTheLine()).
>
> Looks from your description so far is that some controlling variable isn't
> being set/reset and you don't step onto the next 'batch' record but to the
> top of the same SCOPEd record.
>
> Is it repeatable across the same reports?
> Have you tried debugging the 'faulty' record outside the 'batch' situation
> to see what's happening?
>
> A 1300 line method is definately overkill IMHO, but if it works for you<g>.
>
> CYA
> Steve
>
>

[toc] | [prev] | [standalone]


Back to top | Article view | comp.lang.clipper.visual-objects


csiph-web