[Openroad-users] Clickpoint problems with SegmentShape

Antill, Jim jantill at revenue.ie
Wed Oct 17 18:29:10 EST 2007


Ross,
 
Thanks for the code. That's the sort of thing I wanted to do but just
couldn't get my head around. 
 
I'll give it a go, possibly using the ClickPoint to get the coordinates of
the click and then using your code to work out the angle the clickpoint is
from the centre of the graph and subsequently which segment it falls into.
 
Thanks,
Jim

-----Original Message-----
From: openroad-users-bounces at peerlessit.com
[mailto:openroad-users-bounces at peerlessit.com]On Behalf Of
rkummer at auspinners.com.au
Sent: 17 October 2007 01:04
To: International OpenROAD Users
Subject: Re: [Openroad-users] Clickpoint problems with SegmentShape


*************************************

This e-mail has been received by the Revenue Internet e-mail service. (IP)

*************************************
	

Hi Jim, I've had a similar problem in trying to detect where a user clicks on
a pie chart that has been generated using OpenROAD segments. The way I got
around it is by detecting the actual mouse coordinates on a click event and
then using basic trigonometry to work out which line the mouse has clicked
on. I've included the code snippet below (apologies for the lack of
comments). It requires using the standard windows API GetWindowRect and
GetCursorPos but since these are part of user32.dll you possibly already
include them as part of your II_LIBU3GL. It has been stable and has  been in
production for several years now so it definitely works. If drawing pie
graphs is the type of thing you are doing then I can provide a more detailed
example with the actual pie generation and sample calling frame. 



PROCEDURE P4_get_pie( 
    pie= COMPOSITEFIELD,p_value = FLOAT , 
    p_code = VARCHAR(32), 
    p_desc = varchar(40), 
    Li_numangles = FLOAT NOT NULL DEFAULT 500)= 
DECLARE 
    pscale = FLOAT NOT NULL, 
    Li = INTEGER NOT NULL, 
    k = INTEGER NOT NULL, 
    theta1 = FLOAT NOT NULL, 
    theta2 = FLOAT NOT NULL, 
    xpos = INTEGER NOT NULL, 
    ypos = INTEGER NOT NULL, 
    xscale = FLOAT NOT NULL , 
    yscale = FLOAT NOT NULL , 
    xleft = INTEGER NOT NULL, 
    xright = INTEGER NOT NULL, 
    ytop = INTEGER NOT NULL, 
    ybottom = INTEGER NOT NULL, 
    linelen = FLOAT NOT NULL, 
    xstart = FLOAT NOT NULL, 
    ystart = FLOAT NOT NULL, 
    Lf_angle = FLOAT NOT NULL, 
    Li_angle = INTEGER NOT NULL, 
    newlen = FLOAT NOT NULL, 
    dexp   =DynExpr, 
    Li_RtnStatus = INTEGER NOT NULL, 
    pcode = VARCHAR(32) NOT NULL, 
    pvalue = FLOAT NOT NULL, 
    pvals = uc_pie_values 

ENDDECLARE 
BEGIN 
    linelen = 1.1 * FLOAT8(Pie.AbsYBOTTOM - Pie.ABsYTop) /2.0;   
    Xstart = Pie.Absxright - (Pie.Absxleft + Pie.Absxright)/2.0; 
    Ystart = Pie.Absybottom - (Pie.Absytop + Pie.AbsYBottom)/2.0; 
    
    xscale =
FLOAT8(CurSession.ScreenWIdth)/FLOAT8(CurSession.PixelScreenWidth); 
    yscale =
FLOAT8(CurSession.ScreenHeight)/FLOAT8(CurSession.PixelScreenHeight); 
    Callproc
GetWindowRect(SUBFORM(Pie).WidgetId,BYREF(xleft),BYREF(ytop),BYREF(xright),BY
REF(ybottom)); 
    CallProc GetCursorPos(BYREF(xpos),BYREF(ypos)); 
    xpos = xscale*FLOAT8(xpos-xleft)  -  xstart; 
    ypos = yscale*FLOAT8(ypos-ytop)  -  ystart; 
    newlen = sqrt(FLOAT8(xpos*xpos) + FLOAT8(ypos*ypos) ); 
    IF newlen > linelen THEN 
        RETURN -1; 
    ENDIF; 
    Lf_angle =  atan(FLOAT8(ypos) / FLOAT8(xpos)) ; 
    IF xpos >=0 AND ypos <=0 THEN 
        lf_angle = -1*lf_angle; 
    ELSEIF xpos <= 0 AND ypos <= 0 THEN 
        lf_angle = 3.1415 - lf_angle ; 
    ELSEIF xpos <= 0 AND ypos >= 0 THEN 
        lf_angle = -1*lf_angle + 3.1415; 
    ELSEIF xpos >= 0 AND ypos >= 0 THEN 
        lf_angle = 6.283 - lf_angle ; 
    ENDIF; 

    lf_angle = Li_numangles * lf_angle /6.283; 
    Li_angle =lf_angle; 
    FOR Li = 1 to pie.Childfields.LastROw DO 
        IF pie.childfields[Li].name = SQUEEZE('line' + varchar(Li_angle))
THEN 
            pvals =
uc_pie_values(pie.childfields[Li].ClientData).Duplicate(); 
            ENDLOOP; 
        ENDIF; 
    ENDFOR; 

    p_code =  pvals.P_code; 
    p_desc =  pvals.P_desc; 
    p_value = pvals.P_value; 

    IF p_code = 'Other' THEN 
        MESSAGE 'You can not drill down into the "Other" category.'; 
        RETURN C_RTNFAIL; 
    ELSE 
        RETURN C_RTNSUCCESS; 
    ENDIF; 
} 


Regards, Ross Kummer 
rkummer at auspinners.com.au 




"Antill, Jim" <jantill at revenue.ie> 
Sent by: openroad-users-bounces at peerlessit.com 


16/10/2007 08:17 PM 


Please respond to
International OpenROAD Users <openroad-users at peerlessit.com>



To
<openroad-users at peerlessit.com> 

cc

Subject
[Openroad-users] Clickpoint problems with SegmentShape

	





Hi,

I'm having quite a bit of difficulty accurately detecting a childclickpoint
event on a segment shape within a composite and was wondering if anyone else
had found a solution to similar problems. I'm using OR4.0 .

The problem seems to be that when a childclickpoint event is detected the
formfield itself doesn't pick the event up. Rather the coordinates of the
clickpoint are taken and the array of fields within the composite gone
through to find the first field that contains the clicked on position. That
works pretty well for rectangular shapes (as most OpenROAD fields are).
However, in the case of segment shapes it seems that a box is formed from the
start and end points and points within this are checked rather than points
that fall on the line itself. So if you have segments that are close to each
other, but not overlapping, the wrong segment is invariably picked up. I have
some example code here that demonstrates the problem. If anyone is interested
I'll post it up.

More worryingly this can result in code being fired in error as any
clickpoint code behind the segmentshape that OpenROAD mistakenly thinks was
clicked is then executed.

As you can imagine this leads to severe problems when doing things such as
Pie Charts which isn't an unusual requirement. I have tried to use MSGraph in
the past but had a number of issues with it's behaviour with OpenROAD (mainly
due to sizing). This might be improved with OR2006 I guess.

So, my questions are: Does anyone know whether this behaviour will be changed
in OR2006, will we be given better drawing tools in the near future e.g draw
a filled arc in a circle etc. and has anyone a favourite alternative to
graphing that allows interaction between the graph and OpenROAD?

Thanks for any advice.

Regards,
Jim

************************

This message has been delivered to the Internet by the Revenue Internet
e-mail service (OP)

*************************

________________________________________________________________
OpenROAD-Users mailing list

You can maintain your subscription here:
http://www.peerlessit.com/mailman/listinfo/openroad-users

To unsubscribe click on this link
mailto:openroad-users-unsubscribe at peerlessit.com&subject=unsubscribe

To subscribe click on this link
mailto:openroad-users-subscribe at peerlessit.com&subject=subscribe 

______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email 
______________________________________________________________________





************************

This message has been delivered to the Internet by the Revenue Internet e-mail service (OP)

*************************
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.peerlessit.com/pipermail/openroad-users/attachments/20071017/fd3f4210/attachment.html 


More information about the Openroad-users mailing list