PDA

View Full Version : strange .dbf format


pirmin
01-07-03, 18:22
Hi, I've got a .dbf file with the first byte h03 and more than 32 fields. The MS VFP ODBC driver refuses to read it. If I make it a dBase II file by changing the first byte to h02 ODBC will crash. Im using the latest driver from Microsoft and have tried all other .dbf capable drivers I could find with no success. What can I do? Any hints?
The .dbf format would perfectly match the dBase II layout just except the field count exceeding 32. There are other .dbf files in the same directory with less than 32 fields and the first byte h02 and they work.

lg48
01-08-03, 19:08
You propably have a Clipper or a dBase III Plus DBF. h03 means it's a DBF without a Memo file. If the first byte is h83 there should be a Memo file with the DBF file. If it's a Clipper DBF the maximum fields per record is 1024 and 255 for dBase III.

If you are using ODBC try to use the dBase III driver instead. Anyway, you should be able to open the DBF in MS Excel. If you can not open it with Excel it might be corrupted.

/LG

Originally posted by pirmin
Hi, I've got a .dbf file with the first byte h03 and more than 32 fields. The MS VFP ODBC driver refuses to read it. If I make it a dBase II file by changing the first byte to h02 ODBC will crash. Im using the latest driver from Microsoft and have tried all other .dbf capable drivers I could find with no success. What can I do? Any hints?
The .dbf format would perfectly match the dBase II layout just except the field count exceeding 32. There are other .dbf files in the same directory with less than 32 fields and the first byte h02 and they work.

pirmin
01-08-03, 19:28
Originally posted by lg48
You propably have a Clipper or a dBase III Plus DBF. h03 means it's a /LG

Well, I finally extracted the data with a hex-editor and shell scripts. The header layout was of dBase II style exactly as described in
http://www.e-bachmann.dk/docs/xbase.htm
except the first byte and that it had too much fields.