Hi all,

First, I never used either pervasive not btrieve, so some of my
questions might seems stupid or at least naive - please excuse this

I need to migrate a database from an old application that use pervasive
database to Access / SQL server
The old application is a DOS app and it uses Pervasive Sql / Btrieve
(?!?!)
Actually, it uses btrieve (the app is started froma bat which starts
btrieve first)
I understood Pervasive uses the same file format (?!?!) as btrieve...

The old dos app works on WinXP in dos prompt, but it works ONLY if the
pc is also running Pervasive Sql 2000.

The problem is I don't know what exactly are database files and how to
connect to them with ODBC/OleDB/whatever, to extract old data to
convert to new format

The old DOS app has several files with extension BST, which I belive
are the database
But when I open Pervasive Control Center and try to create a new
database I see it create the files with different extension.

Finally, I managed to open the BST file with Rebuild utility - it told
me it processed NNN records. So BST file is actually recognized.
Still, I have no idea how to attach it to a database, or open it as
database.

Please help.

Regards,
Marius

In article <1164754595.315271.180480 (AT) n67g2000cwd (DOT) googlegroups.com>,
bzamfir (AT) gmail (DOT) com says...
and indexes. Then you can read the files in Access through ODBC. We do
this all the time.

I use 'DDF Editor' to create and maintain these files, but if you don't
already have them you'll need to know the Btrieve file formats, field by
field.
--
Regards, John
news (AT) johnkane (DOT) clara.co.uk

Try and locate the DDF files. These include the file.ddf, field.ddf
and index.ddf. These will be required to extract the data. If you do
not have them and need to build these you can download BtSearch at
www.nssdd.com. It allows you to analyze and build the DDF files
afterwhich it will export out to CSV or other formats.

Gil

John Kane wrote:

Hi,

Thank hyou all for answers


nelsonsoft a scris:
Unfortunately I have no DDF filles, and the customer (which is end-
user) has no idea what data base structure means.

I opened the data file with btSearch and it reported me that the data
file has variable record length.
I guess this means it contains fields like memo or varchar-like (in
Oracle, sql server)?
Since the field does not have a fixed length size, is there any way to
know the know the field limit? Like a field terminator char /
delimited? Is it such thing?

Also, when I processed the BST file with Rebuild utility - it told
me it processed NNN records. So somehow does it recognized at least the
records? Is there any record delimiter?

Is there any utility to automatocally recoved table structure?

Thanks
Marius

Variable length records are in the hands of the developers. I have
seen them used in many ways. Some use it for a note field. Others
have used it to store fields that are of fixed sizes thus you can still
build a DDF for them. Then others have used packed fields in which
case you cannot build a DDF for those. However you can always build it
on the fixed portion of the file which also includes any indexed
fields. So until you start getting into the record structure you will
not know about the variable length portion of the file.

As far as the current record count this is stored in the header. But
the record structure is not stored in the files other than some of the
index information. Email me if you have additional questions.

Gil


bzamfir (AT) gmail (DOT) com wrote: