It would be nice if you modellers out there would get together and agree on a standard for material naming (or zoning).

I was trying to write a script that would automatically update EZSkin definition files just by loading a new character and pressing a button. (it checks that the definition does not already exist)

I was doing rather well with it until I came across some of Snarly's EZSkin definitions and found that, although some models have the same names in their materials, they belong in different categories.

for example, cornea, - great, easy to see where that goes right? WRONG. some characters have a cornea, and it is ignored, some have one in the eyetrans catagory, still others are in the invisible category, a nightmare to try to code for.

instead of using

if mat in material:

you have to use something like

if mat in material and figure != 'thisscrewedupmaterialzonesfigure' and figure != 'thatscrewedupmaterialzonesfigure':

etc. etc.

Understand what you mean, it would be easier in the long run.  But modelers have a different mindset about structuring mat zones for different models, in their own way, they believe texturing would be simpler and easier.  Creature models don't necessarily follow any convention, yet they benefit just the same when applying EZSkin.  

Now figure the nightmare you can face when attempting to script Genesis 2 after a UV change, you have all the mat zones duplicated in G2 and generation 4 nomenclature and trying to determine which set actually appears in Poser.

I think I know what you want to achieve, and I think it won't be possible:

With all the different figures, different material zones, and most of all: node setups, it is almost impossible to write a "one script covers it all".

A simple example:
I made a figure which has several "options" attached to the diffuse map, like makeup, blushing/sweating, lip's moisture, and whatever.
This is all in a shader tree which also includes SSS, specularity and whatever, and it would be a nightmare to edit manually.
But "I was smart and added a lot of blender nodes" to easyly swich between the options :/

Good so far.

However, if you write a script that shall hunt down the specific node which you want to change, your script would have to have some kind of AI - in order to determine what each node actually does, and which node is "THE ONE" which must be modified.
It also must account for the settings of the blender nodes, i.e. which part of the shader tree is current.

Now back to the initial problem of identical material naming:

I model a bird with "wings" and "legs/feet/claws"
Next I model a whale, which has "flukes" and "fins"
because I have nothing better to do, I spend the rest of the week modeling a centipede: no "arms" at all, but lLeg1 + rLeg1 up to lLeg100 to rLeg100.

Any idea about a common naming convention for these three?
And I don't even talk about tails here, though some animals and male humans could eventually share some here :D

Summary:

I believe that such a script can be coded, and it would probably be of great help for people who frequently work with a particular figure.
However, I see no other way than that this script must be written for this particular figure, and all it's particularities.
Anything different in other figures would break the script, because it's nearly impossible to cover all eventualities (as I described above)
So a "personalized" script would be necessary for every original figure, and all it's particularities, and 100 legs or wings...

Peter

The nodes do not affect this script, the material zones however, do. SnarlyGribbly's EZMat and EZSkin scripts do the hard work (with the nodes) and rely on a definitions file to tell them what category to place the materials into. :)