lout-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Inputting data for web-based forms generator


From: kersten
Subject: Re: Inputting data for web-based forms generator
Date: Mon, 13 Dec 1999 10:51:32 +0100 (CET)

On 13 Dec, Jamieson Becker wrote:
...
> What is the best way to "fill in the blanks" of a pre-designed lout form
> from a CGI script? (ie,
> a simple bash or python script)? Do I need to generate a text file on
> the fly with appropriate
> symbol declarations and sysinclude it so that those declarations are
> used when calling lout or is
> there an easier way?
> 
...
> My ultimate goal is to generate a PDF from a lout "template" source file
> that has a few variables that are generated on the fly and hand it back
> to the client.

Hello Jamieson,

I'm useing lout based forms on low-end computers (PC-XT, PDA) without
using lout itself. It works in a two stage process: first you make a
lout form with special tags in it and generate what I've called a tagged
PostScript file, then you use a simple script to replace the tags by
data and thus generate the printable form. For this second stage even
'sed' will do, given the right environment (Web backend?)

In this way, you generate the filled form with minimum fuss and minimum
system resources. The catch is that my approach can only use fixed width
fonts in the variable data and that the form has to be *really* static
(no floating elements). Aside from these restrictions, it works great
and very fast.

I intend to use proportional fonts in the (not so near) future, but the
simplicity of the filling application is gone by then, as it has to
interpret PDA files. It should still be much faster than using lout all
the way though.

I've tried using 'tagged PDF forms', but the restrictions on graphics
lout enforces appeared to be unacceptable to me. You can always use
'ps2pdf' from 'gs' if you're stuck with it. It should still be faster
than using lout itself and it does a far better job anyway (sorry Jeff).

I don't have the python scripts and lout templates by hand here, but if
you're interested, I can mail them to you when I do.

Just my 2c,
-- 
Pieter J. Kersten; Erasmus Universiteit Rotterdam, dept. DI&A/S&N,
WH2-20, B. Oudlaan 50, 3062 PA Rotterdam, Netherlands;
Voice: +31-10-4082642; Fax: +31-10-4089064


reply via email to

[Prev in Thread] Current Thread [Next in Thread]