X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient: geda-user AT delorie DOT com
Message-ID: <4FFEB545.5010000@laserlinc.com>
Date: Thu, 12 Jul 2012 07:30:13 -0400
From: Joshua Lansford <Joshua DOT Lansford AT laserlinc DOT com>
User-Agent: Mozilla/5.0 (X11; Linux i686; rv:13.0) Gecko/20120614 Thunderbird/13.0.1
MIME-Version: 1.0
To: geda-user AT delorie DOT com
Subject: Re: [geda-user] refdes index locations
References: <4FFD6CF7 DOT 60707 AT laserlinc DOT com> <20120711201719 DOT A8816819FBAB AT turkos DOT aspodata DOT se>
In-Reply-To: <20120711201719.A8816819FBAB@turkos.aspodata.se>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Reply-To: geda-user AT delorie DOT com


> Very nice, but shouldn't you pick the center of the component instead
> of its lower left, or why not picking the lower left of the refdes,
> which is the thing one is loocking for when searching, that should be
> easy to find for the program ?
That is a good point.   I'm impressed you figured that out.  Did you get 
the script running or did you read though script? If you get an itch for 
it, send me a patch which corrects it.  If the script is not parsing the 
component width, it shouldn't be that hard to add it to the model.
> Instead of forcing the user to tell the program the bonding box, why
> not scan through it (recursively), let the program do the job?
I thought about that and while finding the width and height of the page 
is somewhat straightforward, scanning and obtaining the number of visual 
rows and columns the sheet is divided into could be interesting. (What 
is the "largest" single letter, and single number text field ..)  
Especially as you would also be guessing as to which symbol even is the 
title page. (who's the biggest, who has the term title in the 
filename...)  It would take less work and be more reliable, to go ahead 
and create all the entries for all the different title pages.  If 
someone were to create their own brand new sheet the guess algorithms 
for identifying and profiling the sheet might not work anyway.
> What is the interface to finding symbols, is it in libgeda or in scheme?
The python script parses the ASCII schematic files themselves. (woot 
human readable format)