small, fast and cross-platform flat-file database for python

[mir nazim]

hi.
i want to know if there is any flat-file relational database system
available for python.
i require it in a project. it should have following capabilities:

1. should be small and compact on system requirments.
2. should be roubust and fast.
3. must be able to handle a bit large tables( upto 5000 rows)
4. should be available for both linux and ms windows 9x/xp/2000
platforms.
5. it will be nice if it supports sql-92.

hopping to find some real good help soon
bye.
blesess-n-luck

 

[Peter Hansen]

i want to know if there is any flat-file relational database system

^^^^^^^^^ ^^^^^^^^^^

I'm no database expert, but aren't these two terms contradictory in
some important way?

-Peter

 

[Eric Baker]

^^^^^^^^^ ^^^^^^^^^^

I'm no database expert, but aren't these two terms contradictory in
some important way?

-Peter

Depends what your definition if the word "is" is

In my early perl days, i used a simple text file based SQL "database".
http://search.cpan.org/~shgun/Sprite-3.21/Sprite.pm

Now with Python, i would probably use SQLite, which isnt exactly a "flat
text file", but still a nice lightweight database for *nix and win32 IMHO.
http://www.hwaci.com/sw/sqlite/

Eric

 

[Peter Hansen]

Depends what your definition if the word "is" is

In my early perl days, i used a simple text file based SQL "database".
http://search.cpan.org/~shgun/Sprite-3.21/Sprite.pm

Now with Python, i would probably use SQLite, which isnt exactly a "flat
text file", but still a nice lightweight database for *nix and win32 IMHO.
http://www.hwaci.com/sw/sqlite/

I thought it depended more on the definition of "relational", which I
understood to refer to operations such as "joins" between different
tables. I suppose it could just be operations between different
searches of the same table, however, in which case I suppose it's
a valid term.

-Peter

 

[Mike Rovner]

^^^^^^^^^ ^^^^^^^^^^

I'm no database expert, but aren't these two terms contradictory in
some important way?

Usualy not. Relational is term for data linkage (not hierarhical, not net),
the other for data storage method.

For example dBase was (not-sql) flat-file relational database system.

Mike

 

[Tom Wilkason]

hi.
i want to know if there is any flat-file relational database system
available for python.
i require it in a project. it should have following capabilities:

1. should be small and compact on system requirments.
2. should be roubust and fast.
3. must be able to handle a bit large tables( upto 5000 rows)
4. should be available for both linux and ms windows 9x/xp/2000
platforms.
5. it will be nice if it supports sql-92.

hopping to find some real good help soon
bye.
blesess-n-luck

Have you looked at metakit ( http://www.equi4.com/metakit/python.html ), it
meets requirements 1-4 on your list above.

Tom

 

[Dennis Lee Bieber]

Peter Hansen fed this fish to the penguins on Friday 21 November 2003
10:56 am:

I thought it depended more on the definition of "relational", which I
understood to refer to operations such as "joins" between different

In classical "relational" theory, a relation is a single table -- eg,
all the data items in each row are related to each other (and, in
particular, to something that can be used as a key for that row)

/informal/ /formal/
/ terms / /terms /

table = relation
row = tuple
column = domain

On this basis, any flat file (heck, even a CSV file) could be a
"relation" as long as each row had the same number of columns, and each
column represents one type of data.


--

 

[Brian Kelley]

Have you looked at metakit ( http://www.equi4.com/metakit/python.html ), it
meets requirements 1-4 on your list above.

And best of all there is an sql binding:
http://www.mcmillan-inc.com/mksqlintro.html

There is a learning curve to using metakit at full strength, I have been
trying to update the documentation, a preview is available here:

http://jura.wi.mit.edu/people/kelley/tutorial/python.html

I have successfully used metakit in a fairly large application and
couldn't be happier with the results. The best thing about metakit is
that you can view millions of records in a nice view with ease:

See KitViewer here

http://jura.wi.mit.edu/people/kelley/

You can try out the binary and load in a csv file to play around.