Re: [Caml-list] [1/2 OT] Indexing (and mergeable Index-algorithms)

[Oliver Bandel <oliver@first.in-berlin.de>]

On Fri, Nov 18, 2005 at 03:37:39PM +0100, Florian Weimer wrote:
> * Oliver Bandel:
> 
> >> After a quick search that only returned one dead project (Alpha, v0.0.1,
> >> last updated 2002), I have to ask: are there any OCaml bindings for
> >> BerkeleyDB?
> >
> > Isn't Dbm-module that thing?
> 
> No, it isn't.  Berkeley DB offers much richer functionality, including
> transactions and replication.
> 
> > But there only string -> string  can be used.
> > So your key as well as value has to be string-type.
> 
> You have to use some serialization/deserialization code.  The same
> issue arises if you use an SQL database.

Maybe rewriting from scratch would make sense.
Then I neither need a SQL-database nor writing
interface code.

If the Dbm is fast enbough and can handle such big
amounts of data/files efficiently enough, I would
write my own indexer then, using the mentioned
idea of inverted lists (but using a think like
inverted trees).

Ciao,
   Oliver