Re: electronic-izing unicode texts - Mailing list pgsql-general

From Richard Huxton
Subject Re: electronic-izing unicode texts
Date
Msg-id 42675341.7030107@archonet.com
Whole thread Raw
In response to electronic-izing unicode texts  ("A. Cropi" <cropister@gmail.com>)
List pgsql-general
A. Cropi wrote:
> my objectives: (1) allow users to make query through the web (i guess
> i will do this via PHP interacting with the postgresql)
>
> my questions are: (1) is it reasonable to put the bookcontent into the
> CONTENT column? (2) the content of the book can be very long (some of
> them have nearly 1 milloin words), so, what kind of considerations
> should i be making? (3) how should i design something like this? there
> must be someone outthere that has done somethign similar to this.. if
> so, please share your experiences.

You might be better off with a web-indexing package.
   http://freshmeat.net/search/?q=web+indexing§ion=projects

Since you're not structuring the content of the book, most of the
advantages of a RDBMS don't apply. If you're going to treat it as text,
just use one of the text indexing systems above.

I would convert each book into one or more web-pages (perhaps one page
per section/chapter) and then use htdig or swish.

--
   Richard Huxton
   Archonet Ltd

pgsql-general by date:

Previous
From: Alexandre
Date:
Subject: Re: Simplified (a-la [G|N]DBM) DB access
Next
From: Daniel Schuchardt
Date:
Subject: Re: Windows install/uninstall as a "service"