Icon View Thread

The following is the text of the current message along with any replies.
Messages 1 to 9 of 9 total
Thread Very odd Locate phenomenon in V3
Fri, Jan 20 2006 12:50 PMPermanent Link

"Dominic Willems"
I haven't investigated it in any real depth but I got a phonecall from a
customer reporting a very odd phenomenon.

Apparently when accented text is being used in one field, such as twice
"é" in this case, the locate on another just never finds a hit. Has this
been observed before...?

Fri, Jan 20 2006 4:26 PMPermanent Link

Tim Young [Elevate Software]

Elevate Software, Inc.

Avatar

Email timyoung@elevatesoft.com

Dominic,

<< Apparently when accented text is being used in one field, such as twice
"é" in this case, the locate on another just never finds a hit. Has this
been observed before...?  >>

When you say "on another", do you mean "on another field" ?  If so, then I
doubt that this is the exact cause.  The two shouldn't have anything to do
with each other.

--
Tim Young
Elevate Software
www.elevatesoft.com

Fri, Jan 20 2006 9:21 PMPermanent Link

"Dominic Willems"
Tim Young [Elevate Software] wrote:
> When you say "on another", do you mean "on another field" ?

Yes.

> If so,
> then I doubt that this is the exact cause.  The two shouldn't have
> anything to do with each other.

I can't agree with you more, but that is what appears to have happened.
I was pulling my two remaining hairs out trying to find the cause why a
locate wasn't working while the record certainly existed. Their system
has been working flawlessly for years now, until I got this phonecall.

The only thing I saw in the afflicted record that was different with
other records was that one field (the name field) contained "é"
characters, while none of the other records had them. The locate,
however, was being carried out on an entirely different field inside the
record (also a string field).

I then proceeded to remove the "é" characters (I replaced them by "e"s)
and tried again. Bingo, it worked. Go figure!

Fri, Jan 20 2006 9:37 PMPermanent Link

"Dominic Willems"
OMG! Had time to look a bit deeper and noticed there was a filter active
in the background. This filter (event) assigned the following to the
filter string:

F := format('pos("%s" in upper(Name) >
0',[F,Uppercase(edCustomerName.Text)])

The Uppercase function turns "é" into "é", not into "É", which I suppose
DBISAM's Upper is doing?

This is one of those things...

Anyhow, nothing to do with DBISAM. All to do with the Uppercase
function! Load of my mind.

Thanks for bearing with me and cheers, (and sorry for scaring you)
Dom

Fri, Jan 20 2006 9:39 PMPermanent Link

"Dominic Willems"
Dominic Willems wrote:
> F := format('pos("%s" in upper(Name) >
> 0',[F,Uppercase(edCustomerName.Text)])

This should read

F := format('pos("%s" in upper(Name) >
0',[Uppercase(edCustomerName.Text)])

(had simplified the snippet for essence but had forgotten about that F)

Fri, Jan 20 2006 9:44 PMPermanent Link

"Dominic Willems"
Using AnsiUppercase now. That works.

Cheers,
Dom

Mon, Jan 23 2006 5:45 AMPermanent Link

Thanks for letting us all know!

/Matthew Jones/
Mon, Jan 23 2006 5:18 PMPermanent Link

Tim Young [Elevate Software]

Elevate Software, Inc.

Avatar

Email timyoung@elevatesoft.com

Dominic,

<< The Uppercase function turns "é" into "é", not into "É", which I suppose
DBISAM's Upper is doing? >>

Correct.  The upper-casing is done according to the full ANSI character set,
not the lower 127 ASCII character set.

--
Tim Young
Elevate Software
www.elevatesoft.com

Mon, Jan 23 2006 5:47 PMPermanent Link

"Dominic Willems"
Tim Young [Elevate Software] wrote:
> Correct.  The upper-casing is done according to the full ANSI
> character set, not the lower 127 ASCII character set.

Which is as it should be. Thanks!

Image