Title: Re: LM:ssboundingbox
Post by: Lee Mac on February 28, 2021, 05:20:04 AM
The results are dependent upon the implementation of the ActiveX getboundingbox method, which is known to be inaccurate for hatches & splines (amongst other possible objects); to avoid this you would need to develop your own getboundingbox function by analysing the hatch boundary within the block definition and transforming accordingly to account for the position, scale, rotation, and orientation of the block reference.
Title: Re: LM:ssboundingbox
Post by: domenicomaria on February 28, 2021, 06:58:29 AM
... maybe, it might help, just ignore the hatch objects ... because it is probable (not sure) that the hatch is contained in plines or lines, arcs ... ...
Quote
you would need to develop your own getboundingbox function by analysing the hatch boundary within the block definition and transforming accordingly to account for the position, scale, rotation, and orientation of the block reference.
... For this, someone told me, to ask to a great Lisper (and not only) that lives in London ... ... anyway, I'll give it a try ... ... thank you, for your great work ! ... ciao
Title: Re: LM:ssboundingbox
Post by: domenicomaria on February 28, 2021, 03:29:35 PM
and in any case, avoiding to calculate the bounding box of the hatch, the correct result is obtained
this could be acceptable, supposing that the contour of any hatch has to be defined from other entities
it is not the best, but something . . .
Title: Re: LM:ssboundingbox
Post by: BIGAL on February 28, 2021, 05:27:48 PM
Just something I found a bounding box of a block of just attributes produces various results depending on how many of the attributes are filled in, so does not get all the block limits, another quirk of bounding box.