dBASE .DBF File Structure

by Jun 23, 1995

 Technical Information Database

TI838D.txt   dBASE .DBF File Structure
Category   :Database Programming
Platform    :All
Product    :Delphi  All

Description:
Sometimes it is necessary to delve into a dBASE table outside the control
of the Borland Database Engine (BDE). For instance, if the .DBT file (that
contains memo data) for a given table is irretrievably lost, the file will
not be usable because the byte in the file header indicates that there
should be a corresponding memo file. This necessitates toggling this byte
to indicate no such accompanying memo file. Or, you may just want to write
your own data access routine.

Below are the file structures for dBASE table files. Represented are the
file structures as used for various versions of dBASE: dBASE III PLUS 1.1,
dBASE IV 2.0, dBASE 5.0 for DOS, and dBASE 5.0 for Windows.

**************************************************************************
The data file header structure for dBASE III PLUS table file.
**************************************************************************

The table file header:
======================

Byte  Contents    Description
----- --------    --------------------------------------------------------
0     1 byte      Valid dBASE III PLUS table file (03h without a memo 
                  (.DBT file; 83h with a memo).

1-3   3 bytes     Date of last update; in YYMMDD format.
4-7   32-bit      Number of records in the table.
      number
8-9   16-bit      Number of bytes in the header.
      number
10-11 16-bit      Number of bytes in the record.
      number
12-14 3 bytes     Reserved bytes.
15-27 13 bytes    Reserved for dBASE III PLUS on a LAN.
28-31 4 bytes     Reserved bytes.
32-n  32 bytes    Field descriptor array (the structure of this array is
      each        shown below)
n+1   1 byte      0Dh stored as the field terminator.

n above is the last byte in the field descriptor array. The size of the
array depends on the number of fields in the table file.

Table Field Descriptor Bytes
============================

Byte  Contents    Description
----- --------    --------------------------------------------------------
0-10  11 bytes    Field name in ASCII (zero-filled).
11    1 byte      Field type in ASCII (C, D, L, M, or N).
12-15 4 bytes     Field data address (address is set in memory; not useful
                  on disk).
16    1 byte      Field length in binary.
17    1 byte      Field decimal count in binary.
18-19 2 bytes     Reserved for dBASE III PLUS on a LAN.
20    1 byte      Work area ID.
21-22 2 bytes     Reserved for dBASE III PLUS on a LAN.
23    1 byte      SET FIELDS flag.
24-31 1 byte      Reserved bytes.

Table Records
=============

The records follow the header in the table file. Data records are preceded
by one byte, that is, a space (20h) if the record is not deleted, an
asterisk (2Ah) if the record is deleted. Fields are packed into records
without field separators orrecord terminators. The end of the file is
marked by a single byte, with the end-of-file marker, an OEM code page
character value of 26 (1Ah). You can input OEM code page data as indicated
below.

Allowable Input for dBASE Data Types
====================================

Data Type      Data Input
-------------- -----------------------------------------------------------
C (Character)  All OEM code page characters.
D (Date)       Numbers and a character to separate month, day, and year
               (stored internally as 8 digits in YYYYMMDD format).
N (Numeric)    - . 0 1 2 3 4 5 6 7 8 9
L (Logical)    ? Y y N n T t F f (? when not initialized).
M (Memo)       All OEM code page characters (stored internally as 10 
               digits representing a .DBT block number).

Binary, Memo, and OLE Fields And .DBT Files
===========================================

Memo fields store data in .DBT files consisting of blocks numbered 
sequentially (0, 1, 2, and so on). The size of these blocks are internally
set to 512 bytes. The first block in the .DBT file, block 0, is the .DBT
file header.

Memo field of each record in the .DBF file contains the number of the
block (in OEM code page values) where the field's data actually begins. If
a field contains no data, the .DBF file contains blanks (20h) rather than
a number.

When data is changed in a field, the block numbers may also change and the
number in the .DBF may be changed to reflect the new location.

This information is from the Using dBASE III PLUS manual, Appendix C.

**************************************************************************
The data file header structure for dBASE IV 2.0 table file.
**************************************************************************

File Structure:
===============

Byte     Contents       Meaning
-------  ----------     -------------------------------------------------
0        1byte          Valid dBASE IV file; bits 0-2 indicate version
                        number, bit 3 the presence of a dBASE IV memo
                        file, bits 4-6 the presence of an SQL table, bit
                        7 the presence of any memo file (either dBASE III
                        PLUS or dBASE IV).
1-3      3 bytes        Date of last update; formattted as YYMMDD.
4-7      32-bit number  Number of records in the file.
8-9      16-bit number  Number of bytes in the header.
10-11    16-bit number  Number of bytes in the record.
12-13    2 bytes        Reserved; fill with 0.
14       1 byte         Flag indicating incomplete transaction.
15       1 byte         Encryption flag.
16-27    12 bytes       Reserved for dBASE IV in a multi-user environment.
28       1 bytes        Production MDX file flag; 01H if there is an MDX,
                        00H if not.
29       1 byte         Language driver ID.
30-31    2 bytes        Reserved; fill with 0.
32-n*    32 bytes each  Field descriptor array (see below).
n + 1    1 byte         0DH as the field terminator.

* n is the last byte in the field descriptor array. The size of the array
depends on the number of fields in the database file.

The field descriptor array:
===========================

Byte     Contents       Meaning
-------  ------------   --------------------------------------------------
0-10     11 bytes       Field name in ASCII (zero-filled).
11       1 byte         Field type in ASCII (C, D, F, L, M, or N).
12-15    4 bytes        Reserved.
16       1 byte         Field length in binary.
17       1 byte         Field decimal count in binary.
18-19    2 bytes        Reserved.
20       1 byte         Work area ID.
21-30    10 bytes       Reserved.
31       1 byte         Production MDX field flag; 01H if field has an 
                        index tag in the production MDX file, 00H if not.

Database records:
=================

The records follow the header in the database file. Data records are
preceded by one byte; that is, a space (20H) if the record is not deleted,
an asterisk (2AH) if the record is deleted. Fields are packed into
records without field separators or record terminators. The end of the 
file is marked by a single byte, with the end-of-file marker an ASCII 26
(1AH) character.

Allowable Input for dBASE Data Types:
====================================

Data  Type           Data Input
----  ----------     -----------------------------------------------------
C     (Character)    All OEM code page characters.
D     (Date)         Numbers and a character to separate month, day, and
                     year (stored internally as 8 digits in YYYYMMDD
                     format).
F     (Floating      - . 0 1 2 3 4 5 6 7 8 9
      point binary
      numeric)
N     (Binary        - . 0 1 2 3 4 5 6 7 8 9
      coded decimal
      numeric)   
L     (Logical)      ? Y y N n T t F f (? when not initialized).
M     (Memo)         All OEM code page characters (stored internally as 10
                     digits representing a .DBT block number).

Memo Fields And .DBT Files
===========================================

Memo fields store data in .DBT files consisting of blocks numbered 
sequentially (0, 1, 2, and so on). SET BLOCKSIZE determines the size of
each block. The first block in the .DBT file, block 0, is the .DBT file
header.

Each memo field of each record in the .DBF file contains the number of the
block (in OEM code page values) where the field's data actually begins. If
a field contains no data, the .DBF file contains blanks (20h) rather than
a number.

When data is changed in a field, the block numbers may also change and the
number in the .DBF may be changed to reflect the new location.

This information is from the dBASE IV Language Reference manual, Appendix
D.

**************************************************************************
The data file header structure for dBASE 5.0 for DOS table file.
**************************************************************************

The table file header:
======================

Byte  Contents    Description
----- --------    --------------------------------------------------------
0     1 byte      Valid dBASE for Windows table file; bits 0-2 indicate
                  version number; bit 3 indicates presence of a dBASE IV
                  or dBASE for Windows memo file; bits 4-6 indicate the
                  presence of a dBASE IV SQL table; bit 7 indicates the
                  presence of any .DBT memo file (either a dBASE III PLUS
                  type or a dBASE IV or dBASE for Windows memo file).
1-3   3 bytes     Date of last update; in YYMMDD format.
4-7   32-bit      Number of records in the table.
      number
8-9   16-bit      Number of bytes in the header.
      number
10-11 16-bit      Number of bytes in the record.
      number
12-13 2 bytes     Reserved; filled with zeros.
14    1 byte      Flag indicating incomplete dBASE transaction.
15    1 byte      Encryption flag.
16-27 12 bytes    Reserved for multi-user processing.
28    1 byte      Production MDX flag; 01h stored in this byte if a prod-
                  uction .MDX file exists for this table; 00h if no .MDX
                  file exists.
29    1 byte      Language driver ID.
30-31 2 bytes     Reserved; filled with zeros.
32-n  32 bytes    Field descriptor array (the structure of this array is
      each        shown below)
n+1   1 byte      0Dh stored as the field terminator.

n above is the last byte in the field descriptor array. The size of the
array depends on the number of fields in the table file.

Table Field Descriptor Bytes
============================

Byte  Contents    Description
----- --------    --------------------------------------------------------
0-10  11 bytes    Field name in ASCII (zero-filled).
11    1 byte      Field type in ASCII (B, C, D, F, G, L, M, or N).
12-15 4 bytes     Reserved.
16    1 byte      Field length in binary.
17    1 byte      Field decimal count in binary.
18-19 2 bytes     Reserved.
20    1 byte      Work area ID.
21-30 10 bytes    Reserved.
31    1 byte      Production .MDX field flag; 01h if field has an index
                  tag in the production .MDX file; 00h if the field is not
                  indexed.   

Table Records
=============

The records follow the header in the table file. Data records are preceded
by one byte, that is, a space (20h) if the record is not deleted, an
asterisk (2Ah) if the record is deleted. Fields are packed into records
without field separators orrecord terminators. The end of the file is
marked by a single byte, with the end-of-file marker, an OEM code page
character value of 26 (1Ah). You can input OEM code page data as indicated
below.

Allowable Input for dBASE Data Types
====================================

Data Type      Data Input
-------------- -----------------------------------------------------------
C (Character)  All OEM code page characters.
D (Date)       Numbers and a character to separate month, day, and year
               (stored internally as 8 digits in YYYYMMDD format).
F (Floating    - . 0 1 2 3 4 5 6 7 8 9
  point binary
  numeric)
N (Numeric)    - . 0 1 2 3 4 5 6 7 8 9
L (Logical)    ? Y y N n T t F f (? when not initialized).
M (Memo)       All OEM code page characters (stored internally as 10
               digits representing a .DBT block number).

Memo Fields And .DBT Files
===========================================

Memo fields store data in .DBT files consisting of blocks numbered 
sequentially (0, 1, 2, and so on). SET BLOCKSIZE determines the size of
each block. The first block in the .DBT file, block 0, is the .DBT file
header.

Each memo field of each record in the .DBF file contains the number of the
block (in OEM code page values) where the field's data actually begins. If
a field contains no data, the .DBF file contains blanks (20h) rather than
a number.

When data is changed in a field, the block numbers may also change and the
number in the .DBF may be changed to reflect the new location.

Unlike dBASE III PLUS, if you delete text in a memo field, dBASE 5.0 for
DOS may reuse the space from the deleted text when you input new text.
dBASE III PLUS always appends new text to the end of the .DBT file. In
dBASE III PLUS, the .DBT file size grows whenever new text is added, even
if other text in the file is deleted.

This information is from the dBASE for DOS Language Reference manual, 
Appendix C.

**************************************************************************
The data file header structure for dBASE 5.0 for Windows table file.
**************************************************************************

The table file header:
======================

Byte  Contents    Description
----- --------    --------------------------------------------------------
0     1 byte      Valid dBASE for Windows table file; bits 0-2 indicate
                  version number; bit 3 indicates presence of a dBASE IV
                  or dBASE for Windows memo file; bits 4-6 indicate the
                  presence of a dBASE IV SQL table; bit 7 indicates the
                  presence of any .DBT memo file (either a dBASE III PLUS
                  type or a dBASE IV or dBASE for Windows memo file).
1-3   3 bytes     Date of last update; in YYMMDD format.
4-7   32-bit      Number of records in the table.
      number
8-9   16-bit      Number of bytes in the header.
      number
10-11 16-bit      Number of bytes in the record.
      number
12-13 2 bytes     Reserved; filled with zeros.
14    1 byte      Flag indicating incomplete dBASE IV transaction.
15    1 byte      dBASE IV encryption flag.
16-27 12 bytes    Reserved for multi-user processing.
28    1 byte      Production MDX flag; 01h stored in this byte if a prod-
                  uction .MDX file exists for this table; 00h if no .MDX
                  file exists.
29    1 byte      Language driver ID.
30-31 2 bytes     Reserved; filled with zeros.
32-n  32 bytes    Field descriptor array (the structure of this array is
      each        shown below)
n+1   1 byte      0Dh stored as the field terminator.

n above is the last byte in the field descriptor array. The size of the
array depends on the number of fields in the table file.

Table Field Descriptor Bytes
============================

Byte  Contents    Description
----- --------    --------------------------------------------------------
0-10  11 bytes    Field name in ASCII (zero-filled).
11    1 byte      Field type in ASCII (B, C, D, F, G, L, M, or N).
12-15 4 bytes     Reserved.
16    1 byte      Field length in binary.
17    1 byte      Field decimal count in binary.
18-19 2 bytes     Reserved.
20    1 byte      Work area ID.
21-30 10 bytes    Reserved.
31    1 byte      Production .MDX field flag; 01h if field has an index
                  tag in the production .MDX file; 00h if the field is not
                  indexed.   

Table Records
=============

The records follow the header in the table file. Data records are preceded
by one byte, that is, a space (20h) if the record is not deleted, an
asterisk (2Ah) if the record is deleted. Fields are packed into records
without field separators orrecord terminators. The end of the file is
marked by a single byte, with the end-of-file marker, an OEM code page
character value of 26 (1Ah). You can input OEM code page data as indicated
below.

Allowable Input for dBASE Data Types
====================================

Data Type      Data Input
-------------- -----------------------------------------------------------
B (Binary)     All OEM code page characters (stored internally as 10
               digits representing a .DBT block number).
C (Character)  All OEM code page characters.
D (Date)       Numbers and a character to separate month, day, and year
               (stored internally as 8 digits in YYYYMMDD format).
G (General     All OEM code page characters (stored internally as 10
               digits or OLE) representing a .DBT block number).
N (Numeric)    - . 0 1 2 3 4 5 6 7 8 9
L (Logical)    ? Y y N n T t F f (? when not initialized).
M (Memo)       All OEM code page characters (stored internally as 10
               digits representing a .DBT block number).

Binary, Memo, and OLE Fields And .DBT Files
===========================================

Binary, memo, and OLE fields store data in .DBT files consisting of blocks
numbered sequentially (0, 1, 2, and so on). SET BLOCKSIZE determines the
size of each block. The first block in the .DBT file, block 0, is the .DBT
file header.

Each binary, memo, or OLE field of each record in the .DBF file contains
the number of the block (in OEM code page values) where the field's data
actually begins. If a field contains no data, the .DBF file contains
blanks (20h) rather than a number.

When data is changed in a field, the block numbers may also change and the
number in the .DBF may be changed to reflect the new location.

Unlike dBASE III PLUS, if you delete text in a memo field (or binary and
OLE fields), dBASE for Windows (unlike dBASE IV) may reuse the space from
the deleted text when you input new text. dBASE III PLUS always appends
new text to the end of the .DBT file. In dBASE III PLUS, the .DBT file
size grows whenever new text is added, even if other text in the file is
deleted.

This information is from the dBASE for Windows Language Reference manual, 
Appendix C.



Reference:


7/16/98 4:33:55 PM

Article originally contributed by