Page 1 of 1

queries with long lists (arrays)

Posted: Sat Aug 18, 2001 1:05 pm
by SMcNaughton
I have been using R&R since at least R&R4. I just upgraded from R&R6 to 9 (xbase). I was surprised that things haven^t really changed that much! ____In R6 I used the RRCGI to write clipper code. I then heavily modified the code ridding all the variables into arrays using DEFINEs and added a User Interface that wasn^t subject to the limits of the current (and then) parameter passing.____The main reason for this was because we present a list of codes to the User for tagging. Some of the code lists are in excess of 15,000 entries. The tagged entries would be read into an array. (The limit is about 4,000 entries for Clipper/vo. Practically, a couple hundred is the norm.)____There are many other User Interface features available to the User that can result in a fairly large query before including this one. Obviously, a simple list of these codes is not possible as it would violate the 1024 character limit imposed by R&R9. ____In my clipper version I just ASCAN(array,<field or calc value>)>0 which is fast and keeps the query size down. Without the RRCGI capability, how is one supposed to accomplish this now??____It would be nice if R&R could read a database of values into an array for scanning as in my example above. The INLIST() function is woefully inadequate for this purpose due to length limitations. I suppose a temp database could be made in the UI and scanned each time, but that will result in a sluggish report in comparison to array based scanning.____If there is no good way to do this now, please place my request in the suggestion box :-)____Oh, one other thing, I also allow the Users to define the report parameters and save them, including their arrays, to a memorization database so they can call up their favorite User defined report, optionally edit the parameters and run it on a regular basis, including batching it with other saved reports. But without the capability above, I will loose this feature in our product!____TIA____Steve McNaughton____

=> RE: queries with long lists (arrays)

Posted: Mon Aug 20, 2001 6:23 am
by kfleming
Version 9 of R&R does have much of the original core that was developed back in DOS but it also has lots of new functionality __as well. Things like parameteRR fields, page x of y numbering, email bursting, a continuing addition to the range of available functions...____Do you really give your users a 15,000 item list for selection?__Creating a temp database of selected items and adding a lookup relation with skip failure action in your report could be used to filter records and would probably not be that inefficient.______Your request for array listing has been added to the enhancment database.____Kathleen__R&R Support ____

==> RE: queries with long lists (arrays)

Posted: Tue Aug 21, 2001 4:46 pm
by SMcNaughton
__>Do you really give your users __>a 15,000 item list for __>selection? ____yes. there^s actually about 18000+ diagnosis codes available for doctors in treating patients. they need to be able to tag and extract from that database at will.____>Creating a temp database of selected __>items and adding a lookup __>relation with skip failure action __>in your report could be __>used to filter records and __>would probably not be that __>inefficient. ____Well, that^s an alternative, but the relation can be to SEVERAL fields in a all-ready-related database. This action over a network can be very data intensive. Some of my sites run 50+ terminals.____>Your request for array listing has __>been added to the enhancment __>database. ____Thank you. I^m sure that others would use it as opposed to list boxes if they can load it from another database easily. I use the native feature of Clipper 5.3 and VO (based on flexfile) to store these arrays. I realize that R&R is not just for those 2 products, so I didn^t ask for that feature :)____Take care.____Steve McNaughton__