Faster Way to Check for Duplicates?

[johngtatp]

Parent Record includes a Reservation# (resnum)

Child Records (linked by resnum) can have an Arrival, Departure, or Towing which is a single character field using an "A" or "D" or "T".

I can have no more than one of each type for a Reservation

I need to check for duplicate on a button press when the user is adding one of the above child records. The data entry is on a form using a set where the parent/child relationship is used.

I currently maintain 3 indexes which are both calculated and filtered and I would like to get rid of them if possible

Arrival example dupentrya index is calculated: ALLTRIM(STR(RESNUM)+"A") and filtered: Type="A"

I can then check for an arrival duplicate with: exist(alltrim(str(Resnum)+"A"),"res_detail","dupentrya")

My thought is that maintaining the three indexes is a lot of overhead given my network and heavy use of changes that are applied to the file every minute, seems like it would give me a boost. I thought that if I am in a form that has the reservation open as the parent, and have a browse on the same form showing the children, that perhaps there is some way to reference what is already open on the screen to check for a duplicate than having to use the index. I thank all in advance for any ideas.

[Tom Henkel]

Re: Faster Way to Check for Duplicates?

I don't understand why you would need 3 indexes. A single index of resnum+type would give you the same results. NO filter. then if you do the check when they enter or select the type, it would show if the record already exists. You could get really intense with this and build a dynamic drop down list of available types when the user enters the resnum only showing those types that are not already used, but that would require some reasonable code.

Tom

[johngtatp]

Re: Faster Way to Check for Duplicates?

You are right! There was a rationale at one point for another functionality, but not necessary now.

However, could I possibly get rid of all of them though by just checking against child records already on the open form?

[MoGrace]

Re: Faster Way to Check for Duplicates?

If you fetch thru the linked child records and create a list of resnum and type combinations then a can save event could warn the user if that combo already exists in the current record.

[Stan Mathews]

Re: Faster Way to Check for Duplicates?

Might not have to fetch.

word_exists(table.current(#).record_content_get("resnum+type"),value_combination_to_check,crlf())

should yield .T. if the value_combination_to_check matched any of the entries in the current linked records.

# is the position of the child table in the set.

[MoGrace]

Re: Faster Way to Check for Duplicates?

Stan - you are a treasure here for sure!

[Stan Mathews]

Re: Faster Way to Check for Duplicates?

Sometimes table.current().some_method() works and sometimes Alpha doesn't like it. Usually worth a try.

I don't currently have anything to test this one on.

Seems to me it works in this instance.

aset.zip

I did not test on a large table or verify the speed. Only testing the principle.

[johngtatp]

Re: Faster Way to Check for Duplicates?

Stan,

Thank you. Never tried this, and not even a thought before. I will give this a whirl, thanks. Could let me let go of an index...