forums.silverfrost.com

[Smib]

Does anybody have sample code to fill a polygon with a hatch pattern? I don't seem to be making any progress with createhatchbrush

Cheers

Brian

[PaulLaidler]

draw_filled_polygon@ only allows you to specify the fill as a solid colour.

I guess that you may be able to create your own brush hBrush using CreateHatchBrush but you will have to select it into the relevant device context before calling draw_filled_polygon@ etc.

You can get the hDC from clearwin_info@("GRAPHICS_HDC") and then select the created hBrush with a call to the Windows API SelectObject.

oldBrush = SelectObject(hDC, hBrush)

[LitusSaxonicum]

In I.O. Angell & G. Griffith "High-resolution computer graphics using Fortran 77" published by Macmillan, ISBN 0-333-40398-3 (0-333-430399-1 for the paperback) (1987) there is a routine for doing just that on pages 93 & 94. (You are a bit too far away from SW London for me to offer you the loan of my copy. I could scan the pages and e mail them if you send me an e mail address via "private message").

However, when I typed up the code and ran it my recollection is that it missed out one of the hatch lines through a vertex. I never got to the bottom of it, and gave up ... (it might have simply been a precision problem).

Eddie

[IanLambley]

Eddie,

You are right, lines which pass through a vertex go wrong. I seem to remember having a fight with that routine. I have attached where I got to in 1994, but it does seem to be a precision problem where one of the hatching lines is very close to the end of a boundary line.

This code just checks a series of lines against a boundary.

[IanLambley]

Part 2

[IanLambley]

Part 3

[LitusSaxonicum]

Thanks Ian,

That saved me a job.

Eddie

[Smib]

Gents Thanks for the very prompt replies

Paul I have not had a chace to try this yet;

Eddie Thanks for the offer; and

Ian thanks for the code

Cheers

Brian

[JohnCampbell]

Ian,

I could not resist converting your SORT to fortran 90, using a SELECT CASE (ITY) structure. I do think that fortran 90 is an improvement in style and probably with little overhead.

I also noted your code sets ITY to 1, however ITY is not supplied as a variable. This could be one of those nasty bugs that is difficult to find.

[Smib]

Paul

Worked perfect;ly when i used the polygon API and after I twiged that I had to define my coordinates in a vector of alternating x and y values

Thanks again to all

[JohnHorspool]

Brian,

Could you post the code please, for all to see?

John

[LitusSaxonicum]

May I add my voice to John H's, asking to see a code fragment please?

Apart from the line through a vertex problem, I think that the ultimate reason I abandoned the quest to program hatching was the incredible price drop in colour printers, and my abandoning plotters of the type that used to pick up a pen and rush round the page with it.

To John C: As Angell and Griffiths were writing in Fortran 77, they probably eschewed Select Case deliberately. Think yourself lucky the code isn't all in arithmetic IFs, ASSIGNed GOTOs and Holleriths, with no comments and worse ....

Eddie

[Smib]

Apologies

Here is code:

integer hatchbrush,oldbrush,npts
integer*4 ijk(50)

c put xy coordinates into a vector array ijk
c alternating x and y

ijk(1)=x(1)
ijk(2)=y(1)
.
.
.
.
.
ijk(2n-1)=x(n)
ijk(2n)=x(n)

npts=n

ij=clearwin_info@('GRAPHICS_HDC')
hatchBrush = CreateHatchBrush(HS_DIAGCROSS,RGB@(255,128,0))
oldBrush = SelectObject (ij, hatchBrush)
call polygon(ij,ijk,npts)


Cheers

Brian