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


Groups > comp.lang.clipper.visual-objects > #10926 > unrolled thread

Wrong Sort Order of Index

Started byJürgen Knauf <trotterklaus@googlemail.com>
First post2017-08-24 06:04 -0700
Last post2017-08-25 09:16 -0400
Articles 14 — 7 participants

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


Contents

  Wrong Sort Order of Index Jürgen Knauf <trotterklaus@googlemail.com> - 2017-08-24 06:04 -0700
    Re: Wrong Sort Order of Index dlzc <dlzc1@cox.net> - 2017-08-24 08:49 -0700
    Re: Wrong Sort Order of Index "Jamal" <noooospam@example.com> - 2017-08-24 15:11 -0400
      Re: Wrong Sort Order of Index "Karl-Heinz" <invalid@invalid.invalid> - 2017-08-25 09:26 +0200
        Re: Wrong Sort Order of Index "Wolfgang Riedmann" <wriedmann@gmail.com> - 2017-08-25 15:18 +0200
    Re: Wrong Sort Order of Index "Michael Hagl" <haglmi@hagl.de> - 2017-08-25 11:49 +0200
      Re: Wrong Sort Order of Index Jürgen Knauf <trotterklaus@googlemail.com> - 2017-08-25 03:13 -0700
      Re: Wrong Sort Order of Index "Karl-Heinz" <invalid@invalid.invalid> - 2017-08-25 12:58 +0200
        Re: Wrong Sort Order of Index Jürgen Knauf <trotterklaus@googlemail.com> - 2017-08-25 04:32 -0700
          Re: Wrong Sort Order of Index Jürgen Knauf <trotterklaus@googlemail.com> - 2017-08-25 04:37 -0700
        Re: Wrong Sort Order of Index Jürgen Knauf <trotterklaus@googlemail.com> - 2017-08-25 04:38 -0700
          Re: Wrong Sort Order of Index "Michael Hagl" <haglmi@hagl.de> - 2017-08-25 15:14 +0200
        Re: Wrong Sort Order of Index "Michael Hagl" <haglmi@hagl.de> - 2017-08-25 14:30 +0200
          Re: Wrong Sort Order of Index James Martin <james@northwoodlumber.com> - 2017-08-25 09:16 -0400

#10926 — Wrong Sort Order of Index

FromJürgen Knauf <trotterklaus@googlemail.com>
Date2017-08-24 06:04 -0700
SubjectWrong Sort Order of Index
Message-ID<d1f59ccc-5ed8-40b0-8cec-55b885373f14@googlegroups.com>
There is a strange bug in Advantage Database Server 12.

When there is the special Character "-" in a string of a data field with an Index,
Advantage Database Server do not find the correct records. The sort order of the  index "UPPER(ARTICELNO)" is wrong:

Example:
aaa-bab
aaa-bbc
aaa-bcb
aaa
abc
abd

What can I do?

Jürgen Knauf

[toc] | [next] | [standalone]


#10927

Fromdlzc <dlzc1@cox.net>
Date2017-08-24 08:49 -0700
Message-ID<af37c772-a4f1-4bfb-a7f3-704056302209@googlegroups.com>
In reply to#10926
Dear Jürgen Knauf:

On Thursday, August 24, 2017 at 6:04:29 AM UTC-7, Jürgen Knauf wrote:
> There is a strange bug in Advantage Database Server 12.
> 
> When there is the special Character "-" in a string of
> a data field with an Index, Advantage Database Server
> do not find the correct records. The sort order of the
> index "UPPER(ARTICELNO)" is wrong:
> 
> Example:
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
> 
> What can I do?

What is the code-page in effect in the database, and in the application?  The "upper()" function has different meanings / actions in different code bases.

The comparison operator may stop at the length of the shorter of two strings, so I'd not be surprised if you occasionally saw:

> aaa-bab
> aaa
> aaa-bbc
> aaa-bcb

... depending on the binary tree algorithm.  Not that this gives you a work-around, but you might try UPPER(TRIM(ARTICELNO)+" ") and see if that doesn't help.  (It is still an ASCII character, so it may map higher than intended...)

David A. Smith

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


#10928

From"Jamal" <noooospam@example.com>
Date2017-08-24 15:11 -0400
Message-ID<onn8du$jdu$1@news.albasani.net>
In reply to#10926
May be you need to add SetCollation(#CLIPPER) to your app's Start method.

I suggest you look at the docs to see if this setting has any side effects 
or if it works with ADS.

Jamal


"Jürgen Knauf"  wrote in message 
news:d1f59ccc-5ed8-40b0-8cec-55b885373f14@googlegroups.com...

There is a strange bug in Advantage Database Server 12.

When there is the special Character "-" in a string of a data field with an 
Index,
Advantage Database Server do not find the correct records. The sort order of 
the  index "UPPER(ARTICELNO)" is wrong:

Example:
aaa-bab
aaa-bbc
aaa-bcb
aaa
abc
abd

What can I do?

Jürgen Knauf 

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


#10929

From"Karl-Heinz" <invalid@invalid.invalid>
Date2017-08-25 09:26 +0200
Message-ID<f0a1qeFu224U1@mid.individual.net>
In reply to#10928
Hi Jamal,

> Example:
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
>
>> May be you need to add SetCollation(#CLIPPER) to your app's Start method.

yes, the sorting results proof that the - default - setting
setcollation(#windows) is active.

regards
Karl-Heinz 

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


#10940

From"Wolfgang Riedmann" <wriedmann@gmail.com>
Date2017-08-25 15:18 +0200
Message-ID<f0ame5F49scU1@mid.individual.net>
In reply to#10929
Hi Karl-Heinz,

> > > May be you need to add SetCollation(#CLIPPER) to your app's Start
> > > method.
> 
> yes, the sorting results proof that the - default - setting
> setcollation(#windows) is active.

ADS needs SetCollation( #Clipper ).

Wolfgang



-- 

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


#10930

From"Michael Hagl" <haglmi@hagl.de>
Date2017-08-25 11:49 +0200
Message-ID<f0aa3vF1f79U1@mid.individual.net>
In reply to#10926
Hi Jürgen,

I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to 
GERMAN),l but xHarbour

Sort result "Upper(text1)" as it should

aaa
aaa-bab
aaa-bbc
aaa-bcb
abc
adb

I dont think that "-" Chr(45) is a special character and charset or 
collation doesn´t matter.

Are you sure that the last char behind aaa ist a space (Chr(32))?

Michael Hagl


"Jürgen Knauf" <trotterklaus@googlemail.com> schrieb im Newsbeitrag 
news:d1f59ccc-5ed8-40b0-8cec-55b885373f14@googlegroups.com...
There is a strange bug in Advantage Database Server 12.

When there is the special Character "-" in a string of a data field with an 
Index,
Advantage Database Server do not find the correct records. The sort order of 
the  index "UPPER(ARTICELNO)" is wrong:

Example:
aaa-bab
aaa-bbc
aaa-bcb
aaa
abc
abd

What can I do?

Jürgen Knauf 

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


#10931

FromJürgen Knauf <trotterklaus@googlemail.com>
Date2017-08-25 03:13 -0700
Message-ID<43283fb2-412c-4f10-adc0-c9b898b3209d@googlegroups.com>
In reply to#10930
Am Freitag, 25. August 2017 11:48:50 UTC+2 schrieb Michael Hagl:
> Hi Jürgen,
> 
> I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to 
> GERMAN),l but xHarbour
> 
> Sort result "Upper(text1)" as it should
> 
> aaa
> aaa-bab
> aaa-bbc
> aaa-bcb
> abc
> adb
> 
> I dont think that "-" Chr(45) is a special character and charset or 
> collation doesn´t matter.
> 
> Are you sure that the last char behind aaa ist a space (Chr(32))?
> 
> Michael Hagl
> 
> 
> "Jürgen Knauf" <trotterklaus@googlemail.com> schrieb im Newsbeitrag 
> news:d1f59ccc-5ed8-40b0-8cec-55b885373f14@googlegroups.com...
> There is a strange bug in Advantage Database Server 12.
> 
> When there is the special Character "-" in a string of a data field with an 
> Index,
> Advantage Database Server do not find the correct records. The sort order of 
> the  index "UPPER(ARTICELNO)" is wrong:
> 
> Example:
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
> 
> What can I do?
> 
> Jürgen Knauf

Hi Michael,
I can reproduce it in the Adavantage database creator.

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


#10933

From"Karl-Heinz" <invalid@invalid.invalid>
Date2017-08-25 12:58 +0200
Message-ID<f0ae74F2dqjU1@mid.individual.net>
In reply to#10930
Hi Michael,

> I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
> GERMAN),l but xHarbour

i made a small VO-Test. Depending on that code i get these results:

FUNCTION Start()
LOCAL oDB AS DBServer
LOCAL cPfad AS STRING


// makes no differences if sortorder.dbf ist a OEM-Dbf or not

//    SetCollation ( #windows )
    SetCollation ( #clipper )

    RddSetDefault ( "DBFCDX" )

    cPfad := "D:\test\"


    IF File ( cPfad + "SOrtorder.cdx" )
        FErase ( cPfad + "SOrtorder.cdx" )

    ENDIF

    oDB:= DbServer { cPfad + "SOrtorder.dbf" }
    oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )

    oDB:Close()

   RETURN NIL


#clipper

aaa
aaa-bab
aaa-bbc
aaa-bcb
abc
abd


#Windows

aaa-bab
aaa-bbc
aaa-bcb
aaa
abc
abd

But there might be another problem.No matter if the index is created using
#clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
the index, Jürgen switches from #windows to #clipper ? Didn´t try that, but
i guess that a seek fails - maybe not always -  in such cases ?

regards
Karl-Heinz 

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


#10934

FromJürgen Knauf <trotterklaus@googlemail.com>
Date2017-08-25 04:32 -0700
Message-ID<43d5c3ac-4db3-42b7-a211-eef0012ded06@googlegroups.com>
In reply to#10933
Am Freitag, 25. August 2017 12:58:47 UTC+2 schrieb Karl-Heinz:
> Hi Michael,
> 
> > I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
> > GERMAN),l but xHarbour
> 
> i made a small VO-Test. Depending on that code i get these results:
> 
> FUNCTION Start()
> LOCAL oDB AS DBServer
> LOCAL cPfad AS STRING
> 
> 
> // makes no differences if sortorder.dbf ist a OEM-Dbf or not
> 
> //    SetCollation ( #windows )
>     SetCollation ( #clipper )
> 
>     RddSetDefault ( "DBFCDX" )
> 
>     cPfad := "D:\test\"
> 
> 
>     IF File ( cPfad + "SOrtorder.cdx" )
>         FErase ( cPfad + "SOrtorder.cdx" )
> 
>     ENDIF
> 
>     oDB:= DbServer { cPfad + "SOrtorder.dbf" }
>     oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )
> 
>     oDB:Close()
> 
>    RETURN NIL
> 
> 
> #clipper
> 
> aaa
> aaa-bab
> aaa-bbc
> aaa-bcb
> abc
> abd
> 
> 
> #Windows
> 
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
> 
> But there might be another problem.No matter if the index is created using
> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
> the index, Jürgen switches from #windows to #clipper ? Didn´t try that, but
> i guess that a seek fails - maybe not always -  in such cases ?
> 
> regards
> Karl-Heinz

I tried with SetCollation(#CLIPPER).
But the same problem.

 Search:
 cSuche = "aaaa"
 lFound := oglServer:Seek(cSuche)

 result:
 lFound => true
 found record => aaa-bab

It only works with oglServer:Seek(PadR(Upper(cSuche),30)).

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


#10935

FromJürgen Knauf <trotterklaus@googlemail.com>
Date2017-08-25 04:37 -0700
Message-ID<6159dac4-941f-460c-b56c-690df267e777@googlegroups.com>
In reply to#10934
Am Freitag, 25. August 2017 13:32:34 UTC+2 schrieb Jürgen Knauf:
> Am Freitag, 25. August 2017 12:58:47 UTC+2 schrieb Karl-Heinz:
> > Hi Michael,
> > 
> > > I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
> > > GERMAN),l but xHarbour
> > 
> > i made a small VO-Test. Depending on that code i get these results:
> > 
> > FUNCTION Start()
> > LOCAL oDB AS DBServer
> > LOCAL cPfad AS STRING
> > 
> > 
> > // makes no differences if sortorder.dbf ist a OEM-Dbf or not
> > 
> > //    SetCollation ( #windows )
> >     SetCollation ( #clipper )
> > 
> >     RddSetDefault ( "DBFCDX" )
> > 
> >     cPfad := "D:\test\"
> > 
> > 
> >     IF File ( cPfad + "SOrtorder.cdx" )
> >         FErase ( cPfad + "SOrtorder.cdx" )
> > 
> >     ENDIF
> > 
> >     oDB:= DbServer { cPfad + "SOrtorder.dbf" }
> >     oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )
> > 
> >     oDB:Close()
> > 
> >    RETURN NIL
> > 
> > 
> > #clipper
> > 
> > aaa
> > aaa-bab
> > aaa-bbc
> > aaa-bcb
> > abc
> > abd
> > 
> > 
> > #Windows
> > 
> > aaa-bab
> > aaa-bbc
> > aaa-bcb
> > aaa
> > abc
> > abd
> > 
> > But there might be another problem.No matter if the index is created using
> > #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
> > the index, Jürgen switches from #windows to #clipper ? Didn´t try that, but
> > i guess that a seek fails - maybe not always -  in such cases ?
> > 
> > regards
> > Karl-Heinz
> 
> I tried with SetCollation(#CLIPPER).
> But the same problem.
> 
>  Search:
>  cSuche = "aaaa"
>  lFound := oglServer:Seek(cSuche)
> 
>  result:
>  lFound => true
>  found record => aaa-bab
> 
> It only works with oglServer:Seek(PadR(Upper(cSuche),30)).

 lFound := oglServer:Seek(Upper(cSuche)) do not work.

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


#10936

FromJürgen Knauf <trotterklaus@googlemail.com>
Date2017-08-25 04:38 -0700
Message-ID<505743cc-e730-4a81-a725-3679086c7d59@googlegroups.com>
In reply to#10933
Am Freitag, 25. August 2017 12:58:47 UTC+2 schrieb Karl-Heinz:
> Hi Michael,
> 
> > I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
> > GERMAN),l but xHarbour
> 
> i made a small VO-Test. Depending on that code i get these results:
> 
> FUNCTION Start()
> LOCAL oDB AS DBServer
> LOCAL cPfad AS STRING
> 
> 
> // makes no differences if sortorder.dbf ist a OEM-Dbf or not
> 
> //    SetCollation ( #windows )
>     SetCollation ( #clipper )
> 
>     RddSetDefault ( "DBFCDX" )
> 
>     cPfad := "D:\test\"
> 
> 
>     IF File ( cPfad + "SOrtorder.cdx" )
>         FErase ( cPfad + "SOrtorder.cdx" )
> 
>     ENDIF
> 
>     oDB:= DbServer { cPfad + "SOrtorder.dbf" }
>     oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )
> 
>     oDB:Close()
> 
>    RETURN NIL
> 
> 
> #clipper
> 
> aaa
> aaa-bab
> aaa-bbc
> aaa-bcb
> abc
> abd
> 
> 
> #Windows
> 
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
> 
> But there might be another problem.No matter if the index is created using
> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
> the index, Jürgen switches from #windows to #clipper ? Didn´t try that, but
> i guess that a seek fails - maybe not always -  in such cases ?
> 
> regards
> Karl-Heinz

I tried with SetCollation(#CLIPPER).
But the same problem.

 Search:
 cSuche = "aaaa"
 lFound := oglServer:Seek(Upper(cSuche))

 result:
 lFound => true
 found record => aaa-bab

It only works with oglServer:Seek(PadR(Upper(cSuche),30)).

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


#10938

From"Michael Hagl" <haglmi@hagl.de>
Date2017-08-25 15:14 +0200
Message-ID<f0am5jF47vsU1@mid.individual.net>
In reply to#10936
Hi,

I have done some tests.
When change collation to ADS_ANSI "AdsSetCharType(ADS_ANSI)" and do a 
reindex, I get the same result with wrong sorting order and unwanted seek 
result.

Maybe the ads sorting order can be changed by function AdsSetCollation(), 
see ADS help file?

Michael Hagl



"Jürgen Knauf" <trotterklaus@googlemail.com> schrieb im Newsbeitrag 
news:505743cc-e730-4a81-a725-3679086c7d59@googlegroups.com...
Am Freitag, 25. August 2017 12:58:47 UTC+2 schrieb Karl-Heinz:
> Hi Michael,
>
> > I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
> > GERMAN),l but xHarbour
>
> i made a small VO-Test. Depending on that code i get these results:
>
> FUNCTION Start()
> LOCAL oDB AS DBServer
> LOCAL cPfad AS STRING
>
>
> // makes no differences if sortorder.dbf ist a OEM-Dbf or not
>
> //    SetCollation ( #windows )
>     SetCollation ( #clipper )
>
>     RddSetDefault ( "DBFCDX" )
>
>     cPfad := "D:\test\"
>
>
>     IF File ( cPfad + "SOrtorder.cdx" )
>         FErase ( cPfad + "SOrtorder.cdx" )
>
>     ENDIF
>
>     oDB:= DbServer { cPfad + "SOrtorder.dbf" }
>     oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )
>
>     oDB:Close()
>
>    RETURN NIL
>
>
> #clipper
>
> aaa
> aaa-bab
> aaa-bbc
> aaa-bcb
> abc
> abd
>
>
> #Windows
>
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
>
> But there might be another problem.No matter if the index is created using
> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
> the index, Jürgen switches from #windows to #clipper ? Didn´t try that, 
> but
> i guess that a seek fails - maybe not always -  in such cases ?
>
> regards
> Karl-Heinz

I tried with SetCollation(#CLIPPER).
But the same problem.

 Search:
 cSuche = "aaaa"
 lFound := oglServer:Seek(Upper(cSuche))

 result:
 lFound => true
 found record => aaa-bab

It only works with oglServer:Seek(PadR(Upper(cSuche),30)). 

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


#10937

From"Michael Hagl" <haglmi@hagl.de>
Date2017-08-25 14:30 +0200
Message-ID<f0ajh3F3l9cU1@mid.individual.net>
In reply to#10933
Hi Karl-Heinz,


> #Windows
>
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
>
> But there might be another problem.No matter if the index is created using
> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
> the index, Jürgen switches from #windows to #clipper ? Didn´t try that, 
> but
> i guess that a seek fails - maybe not always -  in such cases ?
>

At this sorting order a softseek for "aaa" and find aaa-bab is IMHO correct. 
The fist occurence of  "aaa" has been found. I think, this sorting order is 
wrong. This collation sorts the "-" before the Space

Michael Hagl


"Karl-Heinz" <invalid@invalid.invalid> schrieb im Newsbeitrag 
news:f0ae74F2dqjU1@mid.individual.net...
> Hi Michael,
>
>> I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
>> GERMAN),l but xHarbour
>
> i made a small VO-Test. Depending on that code i get these results:
>
> FUNCTION Start()
> LOCAL oDB AS DBServer
> LOCAL cPfad AS STRING
>
>
> // makes no differences if sortorder.dbf ist a OEM-Dbf or not
>
> //    SetCollation ( #windows )
>    SetCollation ( #clipper )
>
>    RddSetDefault ( "DBFCDX" )
>
>    cPfad := "D:\test\"
>
>
>    IF File ( cPfad + "SOrtorder.cdx" )
>        FErase ( cPfad + "SOrtorder.cdx" )
>
>    ENDIF
>
>    oDB:= DbServer { cPfad + "SOrtorder.dbf" }
>    oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )
>
>    oDB:Close()
>
>   RETURN NIL
>
>
> #clipper
>
> aaa
> aaa-bab
> aaa-bbc
> aaa-bcb
> abc
> abd
>
>
> #Windows
>
> aaa-bab
> aaa-bbc
> aaa-bcb
> aaa
> abc
> abd
>
> But there might be another problem.No matter if the index is created using
> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
> the index, Jürgen switches from #windows to #clipper ? Didn´t try that, 
> but
> i guess that a seek fails - maybe not always -  in such cases ?
>
> regards
> Karl-Heinz 

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


#10939

FromJames Martin <james@northwoodlumber.com>
Date2017-08-25 09:16 -0400
Message-ID<onp7o7$inl$1@dont-email.me>
In reply to#10937
But space is ascii 32 and dash is ascii 45.

James

On 2017-08-25 08:30 AM, Michael Hagl wrote:
> Hi Karl-Heinz,
> 
> 
>> #Windows
>>
>> aaa-bab
>> aaa-bbc
>> aaa-bcb
>> aaa
>> abc
>> abd
>>
>> But there might be another problem.No matter if the index is created using
>> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
>> the index, Jürgen switches from #windows to #clipper ? Didn´t try that,
>> but
>> i guess that a seek fails - maybe not always -  in such cases ?
>>
> 
> At this sorting order a softseek for "aaa" and find aaa-bab is IMHO correct.
> The fist occurence of  "aaa" has been found. I think, this sorting order is
> wrong. This collation sorts the "-" before the Space
> 
> Michael Hagl
> 
> 
> "Karl-Heinz" <invalid@invalid.invalid> schrieb im Newsbeitrag
> news:f0ae74F2dqjU1@mid.individual.net...
>> Hi Michael,
>>
>>> I have do some tests: (ADS-12, OEM- und ANSI-Character Set is setting to
>>> GERMAN),l but xHarbour
>>
>> i made a small VO-Test. Depending on that code i get these results:
>>
>> FUNCTION Start()
>> LOCAL oDB AS DBServer
>> LOCAL cPfad AS STRING
>>
>>
>> // makes no differences if sortorder.dbf ist a OEM-Dbf or not
>>
>> //    SetCollation ( #windows )
>>     SetCollation ( #clipper )
>>
>>     RddSetDefault ( "DBFCDX" )
>>
>>     cPfad := "D:\test\"
>>
>>
>>     IF File ( cPfad + "SOrtorder.cdx" )
>>         FErase ( cPfad + "SOrtorder.cdx" )
>>
>>     ENDIF
>>
>>     oDB:= DbServer { cPfad + "SOrtorder.dbf" }
>>     oDB:Createorder ( "OrderID", cPfad + "SOrtorder.cdx", "Upper(ID)" )
>>
>>     oDB:Close()
>>
>>    RETURN NIL
>>
>>
>> #clipper
>>
>> aaa
>> aaa-bab
>> aaa-bbc
>> aaa-bcb
>> abc
>> abd
>>
>>
>> #Windows
>>
>> aaa-bab
>> aaa-bbc
>> aaa-bcb
>> aaa
>> abc
>> abd
>>
>> But there might be another problem.No matter if the index is created using
>> #clipper or #windows a seek shouldn´t fail. Maybe later on, after creating
>> the index, Jürgen switches from #windows to #clipper ? Didn´t try that,
>> but
>> i guess that a seek fails - maybe not always -  in such cases ?
>>
>> regards
>> Karl-Heinz
> 
> 

[toc] | [prev] | [standalone]


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


csiph-web