Mail Archives: geda-user/2013/11/07/15:45:03
X-Authentication-Warning: | delorie.com: mail set sender to geda-user-bounces using -f
|
X-Recipient: | geda-user AT delorie DOT com
|
DKIM-Signature: | v=1; a=rsa-sha256; c=relaxed/relaxed;
|
| d=gmail.com; s=20120113;
|
| h=mime-version:from:date:message-id:subject:to:content-type;
|
| bh=HS1JKZcVsJ77Pev9zSPtW/xThT+CMPVJGHdy/9ktXRg=;
|
| b=eWYfRDVNV4wJILp9YycglxkVF3Hxdga2JdJr32Hq8aLdcFoWAiIex1LlHvKOcHjYdU
|
| h4WSoJ7HMd/fzOnrLZOc6yyL6VGjRu4w+TL3uqgHA/d02AF0ptBVycJYDqpO1cEp43Va
|
| SilUxn6YvaO0TPCYt8lY5b/t1bvLUYPPUbg54LiMZZoyE8aNZp7WqE0ewYOx8WwN/Cof
|
| RRvxNyNequP9FKzTijX5ZXkKqmOxVhkN5FYJwPLTCBkARyg0WU3QDVLPgTJTZl+YOpiZ
|
| km0v+Z+0J+RNiNyq6nKxo8iH3iywvT6QpHRPY2bXJiOx46UTcxCKBR+3jHBj8bNLtZ3N
|
| UcBQ==
|
X-Received: | by 10.112.144.105 with SMTP id sl9mr1293672lbb.37.1383857021091;
|
| Thu, 07 Nov 2013 12:43:41 -0800 (PST)
|
MIME-Version: | 1.0
|
From: | Martin Held <diver DOT martin AT gmail DOT com>
|
Date: | Thu, 7 Nov 2013 12:43:21 -0800
|
Message-ID: | <CAGojTMExP001M3CG5Xd8LtpA8aq+VzMGmBixU_Z9Cjqccmc-dg@mail.gmail.com>
|
Subject: | [geda-user] Perl hack for bulk updates of refdes'
|
To: | geda-user <geda-user AT delorie DOT com>
|
Reply-To: | geda-user AT delorie DOT com
|
--047d7b3a82f429842804ea9c5237
Content-Type: text/plain; charset=ISO-8859-1
DJ suggested I send this hack to the mailing list directly. This is an
ugly perl hack, but then again, most perl hacks are. I used it again today
and remembered I forgot to mail this one in.
I'm not a normal geda-user subscriber, so your best bet is to reply
directly to me.
I've found myself doing some big designs recently, wherein I add lots of
components, redo stuff, do a layout, make some schematic changes, split off
schematic files, split off board, recombine boards, make more changes, more
pcb changes, etc etc. In the end, the reference designators get all sorts
of out-of-whack. C200 near R1, stupid little OCD annoying stuff like that.
The problem is that if you re-annotate your schematic (refdes, current
page, overwrite all existing) when you have a PCB layout that's complete,
there's no current 'good' way to forward annotate those changes without
gnetlist (in my system) deleting half the components, only to make you add
them all back in again, redo the refdes locations, yadda yadda.
So here's a perl script to do what I needed. Do this where you have a
schematic and a corresponding layout. You can probably break things by
adding/removing components during this process.
The first step is to grep refdes file.sch > pre.txt. This creates a file
of all reference designators.
Then go into gschem with your schematic, and attributes -> autonumber text
(or shortcut 'tu'), and wipe out all your reference designators with new
ones. Save and exit (backing up along the way might be a good idea too.. I
commit to SVN each time before/after I do this). You can also do any
manual changes, e.g. if you don't like the top to bottom, diagonal/etc
options in the autonumber menu.
Then grep refdes file.sch > post.txt
Now run the perl script. It'll create a hash out of pre and post.txt, and
then scan through the pcb file and change the reference designators as
necessary. Naturally, if you add or remove components, you can change the
order of pre/post.txt... that will screw things up in the hash table.
./update.pl pre.txt post.txt file.pcb output.pcb
You may also wish to re-run gnetlist, and then when you open the output.pcb
file, reload the new netlist and ensure you didn't get any munged
connectivity.
If all goes well, you have reference designators that make more sense.
License: None, I don't care if you sell it or something, I use it for
myself, it works for my needs. You can even take credit for all I care.
Cheers,
Martin
AE7IK
-------------------------------------------------------------------------------
If there aren't any questions, then what is there to learn?
--047d7b3a82f429842804ea9c5237
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
<div dir=3D"ltr"><div>DJ suggested I send this hack to the mailing list dir=
ectly.=A0 This=20
is an ugly perl hack, but then again, most perl hacks are.=A0 I used it=20
again today and remembered I forgot to mail this one in.<br><br></div>I'=
;m not a normal geda-user subscriber, so your best bet is to reply directly=
to me.<br><div>
<br>I've found myself doing some big designs recently, wherein I add=20
lots of components, redo stuff, do a layout, make some schematic=20
changes, split off schematic files, split off board, recombine boards,=20
make more changes, more pcb changes, etc etc.=A0 In the end, the reference
designators get all sorts of out-of-whack.=A0 C200 near R1, stupid little
OCD annoying stuff like that.<br>
<div><br>The problem is that if you re-annotate your schematic=20
(refdes, current page, overwrite all existing) when you have a PCB=20
layout that's complete, there's no current 'good' way to fo=
rward=20
annotate those changes without gnetlist (in my system) deleting half the
components, only to make you add them all back in again, redo the=20
refdes locations, yadda yadda.<br>
<br>So here's a perl script to do what I needed.=A0 Do this where you h=
ave
a schematic and a corresponding layout.=A0 You can probably break things=
=20
by adding/removing components during this process.<br><br></div><div>The fi=
rst step is to grep refdes file.sch > pre.txt.=A0 This creates a file of=
all reference designators.<br>
</div><div>Then go into gschem with your schematic, and=20
attributes -> autonumber text (or shortcut 'tu'), and wipe out a=
ll=20
your reference designators with new ones.=A0 Save and exit (backing up=20
along the way might be a good idea too.. I commit to SVN each time=20
before/after I do this).=A0 You can also do any manual changes, e.g. if=20
you don't like the top to bottom, diagonal/etc options in the autonumbe=
r
menu.<br>
<br></div><div>Then grep refdes file.sch > post.txt<br><br></div><div>No=
w
run the perl script.=A0 It'll create a hash out of pre and post.txt, a=
nd=20
then scan through the pcb file and change the reference designators as=20
necessary.=A0 Naturally, if you add or remove components, you can change=20
the order of pre/post.txt... that will screw things up in the hash=20
table.<br>
<br></div><div>./<a href=3D"http://update.pl" target=3D"_blank">update.pl</=
a> pre.txt post.txt file.pcb output.pcb<br><br></div><div>You
may also wish to re-run gnetlist, and then when you open the output.pcb
file, reload the new netlist and ensure you didn't get any munged=20
connectivity.<br>
<br>If all goes well, you have reference designators that make more sense.<=
br><br></div><div>License:=A0
None, I don't care if you sell it or something, I use it for myself, i=
t
works for my needs.=A0 You can even take credit for all I care.<br>
<br></div><div>Cheers,<br></div><div>Martin<br></div><div>AE7IK<br></div><b=
r>-------------------------------------------------------------------------=
------<br>If there aren't any questions, then what is there to learn?
</div></div>
--047d7b3a82f429842804ea9c5237--
- Raw text -