MetaKit for Python 0.2 - structured storage 
Author Message
 MetaKit for Python 0.2 - structured storage

This is the first public preview release of MetaKit for python (v 0.2).
It is intended as early test of the combination of MetaKit and Python.

MetaKit is a cross-platform, highly dynamic database library providing
persistent storage for all your application data.  MetaKit allows you
to easily store complex data structures, and supports changing all data
structures on the fly.  Datafiles are fully portable across all major
platforms, including Windows, Macintosh and Unix.


MetaKit offers features such as commit/rollback and high-performance
memory-mapped file paging (on platforms that support this) that you
would usually only expect in full-blown database packages and servers.

Yet the complete library, including data manipulation functions such as
sorting, selection, join, group by, and set operations is around 90 Kb
(on Win32 as DLL).  The full-featured C++ API is based on a set of just
6 core classes and encapsulates all complexity to simplify its use.


The most unique feature is that existing datafiles can be restructured
while open, yet with full commit/rollback.  Due to an innovative design,
restructuring is virtually instant and independent of database size.
These versioning capabilities greatly simplify the task of extending
applications which must maintain full backward datafile compatibility.

Datafiles are self-describing and use an extremely compact format.  All
the usual datatypes are supported (ints, floats, variable strings and
binary data).  Adaptive integer sizing automatically uses 1..32 bits per
value (based on magnitude), all byte-order issues are taken care of.

Each datafile can hold any number of independent tables ("views"), which
can be structured and contain subviews to store hierarchically organized
data (the hierarchy must be homogenous in structure).


This first release of MetaKit for Python demonstrates the flexibility of
run-time data structures.  It allows you to create datafiles, alter
their structure, and efficiently store large amounts of variable data.

This 0.2 release implements a few new datatypes for Python and achieves
excellent performance.  As "proof of concept", only the Win32 DLL has
been built and tested - but many more ports will be added once the API
has stabilized and more experience is gained with actual use in Python.
A "1.0" release will be announced once it is robust enough, and tested.

This initial release does not yet expose some of the more advanced data
manipulation features available in MetaKit 1.8 (the latest C++ release).


You are kindly invited to check out the flexibility and performance of
this 80 Kb download - but be warned that it may radically change the way
you look at database- and persistence-libraries, maybe even at Python...


The MetaKit library is freely available for personal and non-commercial
use. Licenses for royalty-free commercial use are available for $165,
this includes the complete MetaKit library source code.

All feedback, suggestions, and comments are most welcome (*)

-- Jean-Claude

(*)  I will be out of the office until August 10th.  This trial was
     built in the very last minute to offer a preview of things ahead.
Jean-Claude Wippler  - MetaKit home page - http://www.*-*-*.com/
Equi4 Software       - "Portable database software for a changing world"

-------- comp.lang.python.announce (moderated) --------

Python Language Home Page:   http://www.*-*-*.com/

Sat, 30 Dec 2000 03:00:00 GMT  
 [ 1 post ] 

 Relevant Pages 

1. MetaKit for Python 0.2 - structured storage

2. ANN: MetaKit for Tcl 1.0 - structured storage

3. UPD: MetaKit for Tcl 0.4 - structured storage

4. ANN: MetaKit for Tcl 0.2 - Unix release

5. ANN: HAP Python Remote Debugger 0.2 Released

6. IDLE 0.2 -- Integrated DeveLopment Environment for Python

7. ANN: RPy (R from Python) 0.2

8. Webware for Python 0.2

9. IDLE 0.2 -- Integrated DeveLopment Environment for Python

10. ANN: Python Benchmark Suite 0.2


Powered by phpBB® Forum Software