Hi
Before I post the sources of this project, I would like to ask for some testing.
The attachment has two files, the database 64-bit Unicode application and a sample.dbf file.
The application is a database explorer (better name is wanted) for dbf version 3.0 files. It has some additional buildin types, but I disabled them in this test version. Memo files are also disabled.
It supports ANSI, wide and UFT8 char, numeric, timestamps, date, logic, integer (32 bit), real4, real8, real10 and auto-increment fields. This is more or less known stuff. The core is its indexing capability for all these types. A compact form of a B-Tree+ is used that does not waste as much space as other commercial products. In terms of speed, it is fast enough. On my computer, 2 million unsorted records can be indexed in about 10 seconds.
An interesting feature is the ability to mutate the database. You can change the definition of a field type and its position and mutate the database content accordingly. Changing the field settings for existing data must be done with care to avoid data loss.
Since queries are database-dependent, they are implemented as APIs of the database engine.
The source code can be compiled for 32 or 64 bit, ANSI or Unicode.
If you have databases in DBF format, please try it out. Before making changes to the database,
please make a copy first.
Biterider
