[Plugin] Hatchfaces (v1.8 beta) UPDATED 15-Dec-2012

jolran

Yes that make sence. I had already put a similar code from your pm in my editor where I had used ents.

That was wrong then from my part. Makes thing clearer. Thanks!

jolran

Strange. Should work, but no innerface is created.
Hatch gets created over whole face. I tried different combinations with
these new lines added, but get errors then..

gp=ents.add_group()
      gp.name=name
      gents=gp.entities
      face=gents.add_face(face.outer_loop.vertices) # adds a faceclone from outer vertices loop
      faces2go=[]  # was faces2gp? Typo? array for innerloups faces
      face.loops.each{|loop|gents.add_face(loop.vertices)}   # creates inner loop 
      gents.each{|face|
      next if face.class!=Sketchup;;Face     
      face.edges.each{|e|
      if not e.faces[1]
      break
      end
    faces2go << face
  }
}
      gents.erase_entities(faces2go)                #erase the faces in the holes

TIG

Try first adding the faces for the 'inner_loops' - each using loop.vertices in turn [i.e. face.loops-[face.outer_loop]], then add the face from face.outer_loop.vertices...
This might work... To see what you are getting stop the code temporarily before the inner faces are erased and before any hatching is done - add a line 'return nil' in the code so that it stops... You should get the faces reproduced in the group. Then try with the inner face erase bit to see if they go [sans hatching...]

sergey2402

Perhaps the thickness of the hatch lines should not be equal to the thickness of the edges? Like dimension lines.

jolran

Not noow Kato!!

I don't understand Sergey. Do you want line weight? That is only possible
through styles in Sketchup. But it's a global feature. I am afraid I can't do anything about that. Maybe if thin rectangles are grouped with materials
like TIG,s dashed lines. But that is further on in developement.

Look at my last minitut for an example how you can use differnt styles, with edge profiles for ex.

Thanks for input Sergey! MUST CODE ON!!

And yes TIG. I'm trying!

jolran

This code gives me a face with holes. If I stop after this code.
But cannot Hatch faces with holes in full script.
Get an error
Normal (square faces) works fine in full script with the changed code, strange enough.

gp.layer=hatch_layer    # group to Hatching layer
      gp.name=name
      gents=gp.entities
      faces2go=[]
      face.loops.each{|loop|gents.add_face(loop.vertices)}
      
      gents.each{|face| 
      next if face.class!=Sketchup;;Face     
      face.edges.each{|e|
      if not e.faces[1]
      break
      end
      faces2go << face
  }
}
 
 gents.erase_entities(faces2go)                #erase the faces in the holes

TIG

Why not?
How are you hatching?
Stop at each stage with a temporary 'return nil' and see what you get or error messages etc...

linea

Tig this is brilliant, wanted this for so long.

sergey2402

Whether not too much I want?

kyyu

jolran, a few suggests for you to consider.

Seperate out the code that draws the line pattern, into it's own method. That will make your main program simpler to understand. It will look more like an outline. Check out my later example code snippet. I just premade the line pattern, by hand, which would represent simply calling a method to make it. Also, you can reuse the draw_lines method, say if you want to draw a 2nd set in the criss cross direction.
Might want to consider a rethink. The program will become more complicated, if you just keep adding stuff on top of exsiting. Here's a possible new way, shown below. And the way I test stuff quickly, is to use the "Ruby Web Console" plugin by Jim Foltz. I often check parts of my program, like this, as snippets of code; before adding it to the plugin. You simply undo <ctrl+z>, change the line(s) of code and execute the code again, by hitting the "eval button".

#This snippet puts the face,lines & hatch into 3 seperate groups.  Face is simply grouped, not copied.
#The "lines" group is premade and named "lines", so the script can find it.
#You need to select the face, before running script
#No need to delete seperate edges, just delete the lines group when done.
model = Sketchup.active_model
ent = model.entities
sel = model.selection
f = sel[0]
face = ent.add_group(f)
lines = (ent.to_a.select {|e| e.is_a?(Sketchup;;Group) and e.name=="lines"})[0]
puts lines
hatch = ent.add_group
hent=hatch.entities; lent=lines.entities; fent=face.entities
htr=hatch.transformation; ltr=lines.transformation; ftr=face.transformation
fent.intersect_with(true, ftr, hent, htr, false, lines)
lines.erase!
face.explode

Snippet of code and model set up to run
![After pressing "eval" button](/uploads/imported_attachments/Hdko_test_hatch_rwc_after.png "After pressing "eval" button")
test_hatch_RubyWebConsole.skp

Jim

This is very useful, jolran. Thanks for making it happen.

kaas

wonderful!! This kind of plugin (hopefully someday with some extra hatch options) is really what I was missing in SU.

Many thanks to those involved!

jolran

Sergey, thanks for your suggestions I will probably go for something like the picture, first time around.
Unless I'm wrong you have to use webdialog for checkboxes, and I'll put that on hold for now.
More important issues to deal with first.

Hi Kyyu. Many thanks for your input and experimenting. Really appreciate it
I got a strong impression from reading this forum, that making groupes from selection is a bad choice?
Thats why I neglected my original idea and went for TIG's "face.clone iterating stuff". Don't you get splats doing it like that? Have you tried this inside a group? I got constant splats doing almost exactly as you did.

However grouping the edges before doing an intersect_with might be a good idea. That might actually be an easier method than iteration the edges for erasing? Depends on how "splat-prone" the method is.

About separating to 3 groups. Don't know if that is necessary? Doesent it suffice to add the hatches to a new layer? (Which is already implemented in my current code). This grouping In my opinion does no good for the workflow. Need more feedback about that, please

@unknownuser said:

Seperate out the code that draws the line pattern, into it's own method

Yeah, I know code's a bit messy, but doing a second hatching that is probably a must anyway.
I'm using AS ruby code editor. It's the same as Jim's no?

Thank you Jim and Kaas! And TIG's been helping me a LOT, that goes without saying.

jolran

YES!!!! Faces with holes are hatching now. Will update tomorrow, hopefully.

Stupid newbie-mistake. I did not create "@face" object. So for ex bb=face.bounds
did not have anything to refer to since face was in Faces2go's scope. At least that is what I think happend.

Will test more before I celebrate too much...

Sorry TIG, I missed your post from yesterday. I wasent using return nil. I used =begin and =end. That works equally good?

jolran

Testing.

Bob James

Looking Good

kyyu

Sorry, disregard, was trying to edit the previous post.

sergey2402

@jolran said:

YES!!!! Faces with holes are hatching now. Will update tomorrow, hopefully.

So when will come tomorrow? I am looking forward!
I am sure - in the new version of SketchUp this plugin should become a standard feature.

kyyu

jolran, looks like you got it working well. So no need to change anything. I will answer the questions, anyways.

@unknownuser said:

Hi Kyyu. Many thanks for your input and experimenting. Really appreciate it
I got a strong impression from reading this forum, that making groupes from selection is a bad choice?
Thats why I neglected my original idea and went for TIG's "face.clone iterating stuff". Don't you get splats doing it like that? Have you tried this inside a group? I got constant splats doing almost exactly as you did.

However grouping the edges before doing an intersect_with might be a good idea. That might actually be an easier method than iteration the edges for erasing? Depends on how "splat-prone" the method is.

Yes, it's to be avoided. But you can do it, under certain circumstances. Grouping loose geometry is ok, especially a single edge or face. In my example, I grouped a face at the very beginning. And it do it, just once. I've written plugins, like this, before. You can't do it what an entitiy, already inside a group. I'm assuming it has to be in the current active_entities.

If, I am just testing a quick snippet of code, I certainly will just group the face instead of writing a bunch of lines to clone it.

You only need to consider doing all this, if you didn't make the entitiy. If you want your lines in a group, then you would make an empty group 1st, and then create the lines directly in the group (gents.add_line instead of ents.add_line).

@unknownuser said:

About separating to 3 groups. Don't know if that is necessary? Doesent it suffice to add the hatches to a new layer? (Which is already implemented in my current code). This grouping In my opinion does no good for the workflow. Need more feedback about that, please

It all depends on how you are doing stuff. My thoughts on 3 groups is as follows: Group1 is your cloned face group. If you decide to group your line pattern, then that's group2. When you use intersec_with, you can choose where the new lines go. You can't put them in group1 or group2. They would merge and you have to clean them up. So you put them in group3. Then you only need to delete group1 & group2. No need to find and delete any edges. That's what my example does. Except, I just explode group1, becasue I don't make a clone.

@unknownuser said:

I'm using AS ruby code editor. It's the same as Jim's no?

Should be the same. Just suggesting an editor that can execute the code, if you aren't using one. I use Notepad++ for that and reload using Jim's Ruby Toolbar. But Jim's Ruby Web Console does the undo automatically. And very easy to test small bits of code. I still use his older version.

-Kwok

sergey2402

Looking to the future, I regret to foresee, this plugin will be not freeware.