Re: gEDA-dev: tragesym patch

[Werner Hoch <werner_dot_ho-at-NOSPAMgmx_dot_de>]

Hi Endre,

On Donnerstag, 13. November 2008, Endre Bak wrote:
> 1. If the size calculation of the symbol width is left to tragesym
> (sym_width=0) there is a possibility to configure increments for
> symbol width and height, by this feature it is easy to eliminate
> overlapping pin labels at the symbol corners.

That's a nice feature.

> 2. If "package_order=yes" option is passed, the pins are placed into
> the symbol with counterclockwise order sorted by the pin numbers. It
> is good if one wants to make a symbol where the layout is similar to
> the package layout. (Of course it is not true for BGA and such
> esoteric packages).

Not sure about that one because of the following reasons:

1. It is easy to write the input file already in the required ordering.

2. The ordering will fail if you have a PLCC package, where the top pin
   row isn't ordered in increasing order.

3. Personally I'm not a fan of package order style of symbols.

I also saw that your sort function is quite complicated.
Here's the code I'd use for it: (untested)
--------
def compare_func(pin_a, pin_b):
    if pin_a.pos != pin_b.pos:
        return cmp(pin_a.pos, pin_b.pos)

    if pin_a.pos in ['r','t']:
        return cmp(-int(pin_a.nr, -int(pin_b.nr))
    else:
	return cmp(int(pin_a.nr, int(pin_b.nr))

pins.sort(compare_func)
--------

Maybe it's even possible to define different __cmp__ functions for the 
pin object. Thus the compare function would be part of the pin class.

> The patch is based on the latest development version which I've
> found: http://geda.seul.org/devel/v1.5/1.5.0/geda-utils-1.5.0.tar.gz

applied to my local git repo. Not yet sure how I will proceed.

Regards
Werner
PS: Please use the unified diff format to create patches (diff -u)


_______________________________________________
geda-dev mailing list
geda-dev@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev