Closed GoogleCodeExporter closed 9 years ago
Will,
I'm not sure I understand. When I import your DXF file, it produces what seems
like four rows of flower patterns with the top two rows contained within a
square.
If I use the 'split sketch' command on the single sketch object, I can't find
duplicate graphical elements.
When you say the file is duplicated, are you saying that there are two copies of
each graphical element in the sketch or are you saying that there should not be
four
rows of flower patterns?
I have attached a screenshot of what I get but I don't see any duplication. Of
course, I haven't looked at the DXF file in qcad. I'm at work at the moment so
I
don't have another utility to open the DXF file with.
Thanks
David Nicholls
Original comment by David.Ni...@gmail.com
on 28 Apr 2010 at 3:26
And the file I promised...
Original comment by David.Ni...@gmail.com
on 28 Apr 2010 at 3:27
Attachments:
David,
I have attached the screen shot from within Autocad so you can see what it
should be
like. The top two rows and border should not be there!
Very strange; I deleted the border from the top image on the below part but it
is
still present in the upper image. Perhaps the dxf format has some kind of
history?
Thanks
Will
Original comment by williamr...@googlemail.com
on 28 Apr 2010 at 6:44
Attachments:
I have noticed this happening again with autocad. please see attached screens.
it
seems that there is some old geometry being saved in the file (?) from the
editing
process that heeks cad is interpreting as part of the drawing.
Have again tried saving in dxf 2007 and dxf 2004 and R12 formats and get the
same result.
Will
Original comment by williamr...@googlemail.com
on 28 Apr 2010 at 8:32
Attachments:
Will,
I have just started looking at this at home. When I open the grid2.dxf file in
QCAD, it presents the two rows of flower shapes on layer '0' but it also
defines a
'block' called 'A$C17C22E5B'.
I can select that block and place another copy in the drawing. When this is done,
the new block has a rectangle around another two rows of flowers. I suspect
that the
block's definition is being interpreted as regular graphics elements within the
DXF file.
The import routine just parses the file looking for things it recognizes. The fact
that some of them are within a block definition is the missing bit. I still
don't
know how to recognize the block definition's start and end points. If I could
figure
this out, I could add a flag to ignore graphics within the definition. The
alternative would be to create another sketch for each block definition. That
might
be more useful. You would still end up with four rows of flowers but they
would be
spread across two sketch objects. It would also mean that you could use the
block
definitions within Heeks.
I'm still going on this....
Original comment by David.Ni...@gmail.com
on 28 Apr 2010 at 8:41
Thanks! Its not really a show stopper; I am using the application of generate
toolpaths for our machine. Really useful piece of software.
If its any help I transferred the pattern into a new dxf file using cut and
paste in
autocad perhaps the block is created during that process. You would think that
autocad produces dxf files in the correct format however.
I've attached the dxf file from that gives the results in previous post which
looks
like the same issue.
Will
Original comment by williamr...@googlemail.com
on 28 Apr 2010 at 9:09
Attachments:
I'm pleased to hear that the program is being used. I sometimes think that
there are
only a handful of us using it.
Original comment by David.Ni...@gmail.com
on 28 Apr 2010 at 9:22
I have change the import to take note of the current section name, block name
and
layer name. It uses these values to group the graphics into distinct sets.
Each set
will be created in a separate sketch whose title includes the section, block and
layer names. (If the graphics is found in the ENTITIES section then the block
name
will be omitted)
It still imports all the graphics but they should be easily cleaned up now
Original comment by David.Ni...@gmail.com
on 3 Jun 2010 at 7:44
Original issue reported on code.google.com by
williamr...@googlemail.com
on 27 Apr 2010 at 8:41Attachments: