Virgo,

Not quite sure what you mean by "binary that 'snuck' in".

Are you using the REINDEX command?  In cases of corrupted indexes, it is usually recommended that you delete the existing index files and then manually recreate them from scratch.

Andy Blay
a.d.blay@talk21.com

Binary snuck in

I get messages....

file name,  index name     'binary of various types'   of file.

I have been using reindex.  I thought about the fact that I was just compounding the error.  

The indices are all cdx.  I don't know for sure what they were made of and I have no idea at all in terms of the order in which they were put.  

Meanwhile, I have some other programs that are failing because they can't find the appropriate index.  

The two files with the corrputed indices are in the gigabyte range.

if you have an ftp program connect to:
server = ibs.internationalbid.net
username = fpug
password = 73897253

There are programs in the directory \sharedcode\sysindex\fpw that can help you.
The Word Document Sysindex.doc tells you about files.
There is a database (SysIndex.dbf & SysIndex.fpt) there that you must populate.
The Prg DataChk will return a logical if there is a stray binary code in a memory variable.
The Prg DbfCheck runs datachk against every field in in a table in the SysIndex database.
The Prg CdxStru will collect the index information for every file in the sysindex database.
The PrgThe Word Document Sysindex.doc tells you about files. will rebuild the indexes from scratch for every file in the sysindex database.

Warning - Use the programs at you own risk.  Run tests against backup files first. The code is your to use/atler/pass around as you see fit.

There is a VFP Version in the directory \sharedcode\sysindex\vfp.

David W. Grewe
Dave@internationalbid.net

Thank you.  I now have things to try.

Make a backup file and open all the tables and in command window type DISPLAY STATUS. This will display the index structure. Note all the indexes structure and then delete the tag file (*.cdx). Make new index file based on the structure you noted. e.g.

Disp stat


 c:\patient.dbf

STRUCTURAL INDEX patient.cdx

INDEX TAG NAME Primaryindex      alltrim(name)+dtoc(dob)
INDEX TAG NAME Secondkey         alltrim(recno)

Now when you will delete patient.cdx it will delete Primaryindex and secondrykey tags.

Now make new index file by using the same structure shown above.

farrukh

Thank you.