AutoCAD DevBlog

Displaying a DCL dialog at the last location

By Xiaodong Liang

Issue
Is it possible to display a DCL dialog at the last location it was closed at, rather than having it always open in the center of the screen ?

Solution
Yes, just feed the value of the last coordinate location of the dialog which is returned via done_dialog, to the new_dialog call. Here's some lisp code that demonstrates this, please ensure that the DCL file is on the AutoCAD search path.

(defun C:Move_the_DCL (/ dcl_id notenum1)

;; 1) loadup the Move_DCL.DCl file Dialogue Box, if it's on the path
  (setq DCLFileName "Move_DCL.dcl")
 
  (setq DCLfile (findfile DCLFileName))
  (if (null DCLfile)
     (progn
       (alert (strcat "\nDCL file:    "  DCLFileName " was Not Found !
Exiting."))
       (exit)
     )
  )
 
  (setq dcl_id (load_dialog DCLfile))


;; check to see if it's loaded, and display at last box location, if available


  (if lastpt
      (if (not (new_dialog "note" dcl_id "3" lastpt)) (progn (alert (strcat
"\nDCL file:  "  DCLFileName " was Not Found ! Exiting.")) (exit)) )
      (if (not (new_dialog "note" dcl_id "3" '(-1 -1))) (exit))
  )


;; 2) setup for return values/keys from DCL


(action_tile "cancel" "(done_dialog) (setq gperr \"\")(exit)")


(action_tile "accept"
    (strcat "(progn (setq ChkBoxReturnValue (atoi (get_tile \"note1\")))" 
    " (setq lastpt (done_dialog)))" 
    ) 
)    


;; 3) now start the dialogue box


  (start_dialog)


;; 4) unload the box


  (unload_dialog dcl_id)


;; print out the value returned


  (if ChkReturnValue
      (progn (princ "\nThe ChkBoxReturnValue is: ") (princ ChkBoxReturnValue))
  )
  (if lastpt
      (progn (princ "\nDialogue Box Location: ") (princ (car lastpt)) (princ ",
") (princ (cadr lastpt)) )
  )
(princ)
)
(princ "\nLoaded Move_DCL.lsp, type Move_the_DCL to run.")
(princ)

Here's the DCL Dialog the lisp file uses:

note : dialog {
  label = "Simple Dialog Example";
  : boxed_column {
     alignment =  centered;
     children_alignment =  centered;
    :text {
     alignment =  centered;
     is_bold = true;
     key = "label1";
     label = "This is a Centered Label";
     width = 8;
     }
    : spacer { width = 1; }
  : row {
    : spacer { width = 1; }
    :toggle {
     label = "Note Number 1";
     key = "note1";
     mnemonic = "C";
     alignment = right;
     }   
    }
    : spacer { width = 1; }
    }
  : row {        // defines the OK/Cancel button row
    : spacer { width = 1; }
    : button {    // defines the OK button

          label = "OK";
          key = "accept";
          width = 8;
          fixed_width = true;
        }
    : button {    // defines the Cancel button
        label = "Cancel";
        is_cancel = true;
        key = "cancel";
        width = 8;
        fixed_width = true;
        }
       }
}

Determine in which line of AutoLISP code an error exists

By Xiaodong Liang

In much previous versions of AutoLISP and Visual LISP (VLISP), when an AutoLISP function terminated unexpectedly in the drawing editor, the entire expression would display on the screen, which made it easy to locate the exact line which
contained the error. When this situation occurs In recent AutoCAD versions, locating the error is not as easy because only a message such as "Bad argument type: lentityp' (or stringp)" displays. The error message unfortunately does not state
in which line of code the error exists.

How errors in AutoLISP code are handled has been changed. Now, to locate the error in the code, you will need to use the Visual Lisp IDE. Although it may require that you learn about the IDE debug features, you'll find that it's a much better environment to debug AutoLISP code than in previous releases of AutoCAD. What you need to do to enable this facility is to perform the following steps:

1. Open the Visual Lisp IDE at the beginning of your AutoCAD session by typing VLISP or VLIDE at the command line.

2. Select the Debug menu in the IDE, and select 'Break On Error'.

3. Load your AutoLISP code and run it.

When an error in your code occurs, you'll be put back into the Visual LISP IDE. All you then need to do, to find the line in the code that contains the error is to select the 'Last Break' option on the Debug Toolbar, or the 'Last Break Source' from the Debug menu.

There are many more powerful tools available to debug yourAutoLISP code using the Visual Lisp IDE, To learn about them, refer to the Visual LISP Developer's Guide >> Debugging Programs for more information about the IDE and its debugging features.

How to use the AcLyLayerFilter::showEditor API?

By Philippe Leefsma

Q:

I am trying to call from my own arx application the "AcLyLayerFilter::showEditor" method in order to display the native AutoCAD layer filter editor, but it doesn't seem to work, nothing is displayed.

A:

Unfortunately displaying this native dialog from API is not possible. The purpose of this method is to be called by AutoCAD itself on a custom filter that would derive from AcLyLayerFilter. Depending on the return type that you define in your overriden showEditor method you can display your own filter editor dialog or the native one, when the user is double clicking or invoking this editor on the custom filter.

The attached sample illustrates this behavior. To test it run command "MyCommand1": it will add a new custom layer filter called "CustomFilter". When you invoke the properties editor on this filter, the overriden showEditor method will be called.

Below is the code of the command method. It shows how to insert the custom filter in the filter list:

static void asdkArxProject1_MyCommand1(void)

{

    AcLyLayerFilterManager* pFilterMgr =

        aclyGetLayerFilterManager(acdbCurDwg());

 

    if (!pFilterMgr)

    {

        acutPrintf(L"\nUnable to get Layer Filter Manager...");

        return;

    }

 

    Acad::ErrorStatus es;

 

    AcLyLayerFilter* pRoot;

    AcLyLayerFilter* pCurrent;

 

    es = pFilterMgr->getFilters(pRoot, pCurrent);

 

    //create our custom layer filter

    asdkCustomFilter* pNew = new asdkCustomFilter();

 

    //set filter properties

    es = pNew->setName(L"CustomFilter");

    es = pNew->setFilterExpression(L"NAME==\"*lay*\" AND COLOR==\"10\"");

 

    if( pRoot->allowNested() == true)

    {

        //add our filter to the root as nested

        es = pRoot->addNested(pNew);

        es = pFilterMgr->setFilters(pRoot, pNew);

    }

}

 

_ArxLayerFilter.zip

How to give back UI focus during lengthy calculations (ObjectARX)

By Philippe Leefsma

Q:

I have a large arx application which performs lengthy calculations. When the calculation is running, our clients would normally minimize the AutoCAD window down to the task bar (it can run for days). Occasionally, in order to monitor progress, the AutoCAD window is maximized but the screen within the window is not refreshed. How can I workaround this?

A:

There are a couple of ways to do what you want.

The first way is multi-threading. I will say that multi-threading isn't recommended because it can have undesirable effects, especially when dealing with the AutoCAD drawing database, and is unsupported.

The second way is to implement your own Command Message Handler - recommended. When your program is executing it takes away the Message handling control from AutoCAD. Because of this when you minimize and maximize the AutoCAD window nothing gets repainted or updated because AutoCAD doesn't get the relevant messages that are sent from the operating system. So, what we need to do is to grab these messages in our 'calculation' loop and then spend some of our calculation time posting them back to AutoCAD and waiting for AutoCAD to process them.

// get the AutoCAD CWinApp class

CWinApp *app = acedGetAcadWinApp();

 

// get the Main window

CWnd *wnd = app->GetMainWnd ();

 

// loop round while we have stuff to do

while (!acedUsrBrk ())

{

    ++count;

  

    // now we have to add our own DispatchMessage handler because our loop

    // will stop autocad's one from processing

    //

    // dispatch incoming sent messages, check the thread message queue for

    // a posted message, and retrieve the message (if any exist).

    MSG msg;  

    while (::PeekMessage (&msg, wnd->m_hWnd, 0, 0, PM_NOREMOVE))

    {

        // check to see if we can pump the messages round

        if (!app->PumpMessage())

        {

             ::PostQuitMessage(0);

             break;

        }

    }

 

    LONG lIdle = 0;

 

    // let MFC do its idle processing

    while (app->OnIdle (lIdle++));

  

    //.... calculation code...

 

}

 

Show AutoCAD color dialog

By Virupaksha Aithal

You can use “Autodesk.AutoCAD.Windows.ColorDialog” class to show the color dialog box in AutoCAD. Below code shows the procedure to show and process the input of color dialog in AutoCAD.

[CommandMethod("showColorDlg")]

public void showColorDlg()

{

    Document doc = Application.DocumentManager.MdiActiveDocument;

    Editor ed = doc.Editor;

 

    Autodesk.AutoCAD.Windows.ColorDialog dlg

                     = new Autodesk.AutoCAD.Windows.ColorDialog();

 

    if (dlg.ShowDialog() !=

                              System.Windows.Forms.DialogResult.OK)

    {

        return;

    }

 

    if (!dlg.Color.IsByAci)

    {

        if (dlg.Color.IsByLayer)

        {

            //by layer

            ed.WriteMessage("By Layer\n");

        }

        else if (dlg.Color.IsByBlock)

        {

            //by block

            ed.WriteMessage("By block\n");

        }

        else

        {

            ed.WriteMessage(dlg.Color.Red.ToString()

                + "--" + dlg.Color.Green.ToString() +

                            "--" + dlg.Color.Blue.ToString() + "\n");

        }

    }

    else

    {

        short colIndex = dlg.Color.ColorIndex;

        System.Byte byt = System.Convert.ToByte(colIndex);

 

        int rgb = Autodesk.AutoCAD.Colors.EntityColor.LookUpRgb(byt);

        long b = (rgb & 0xffL);

        long g = (rgb & 0xff00L) >> 8;

        long r = rgb >> 16; ;

 

        ed.WriteMessage(r.ToString() + "--" +

                          g.ToString() + "--" + b.ToString() + "\n");

    }

 

}

Dealing with MFC Resource conflicts between an ObjectARX MFC DLL and AutoCAD

by Fenton Webb

The CAcModuleResourceOverride class provides a nice way to switch the MFC resource pointer from AutoCAD’s resources to your own ARX DLL resources.

If you have to use AutoCAD User Interface functions in between your own resources, then you can still utilize the CAcModuleResourceOverride() class by relying on its destructor (which restores the MFC resource pointer back to AutoCAD)

Here’s what I mean…

class CMyDialog : public CDialog
{
…
};

void ftn()
{
  {
    // here we need to make sure we are using the dll’s resources

    CAcModuleResourceOverride resourceOverride;
    CMyDialog dlg;
    dlg.DoModal();

  } // switch back to AutoCAD’s resources

  // now safe to use this function
  acedGetFileD(...);

  // now back to our dialog
  {
    // here we need to make sure we are using the dll’s resources, so use:
    CAcModuleResourceOverride resourceOverride;
    CMyDialog dlg;
    dlg.DoModal();
  }

}

Linking RealDWG SDK C++ rcexelib.obj in Visual Studio Debug mode

by Fenton Webb

If you are building a RealDWG application using the C++ API then you will have to link your application with the RealDWG SDK rcexelib.obj so that your application will run.

The thing is with all Autodesk products, we give them to you compiled in Visual Studio ‘Release’ mode, not in ‘Debug’ mode therefore you have to be careful when trying to run in Debug mode.

The reason you have to be careful is because the default Visual Studio project setup for a Debug project is that it:

1. Defines _DEBUG preprocessor
2. Sets the Runtime Library as Multithreaded Debug DLL

If you try and build your RealDWG application in Debug mode with these settings, the linker will have some issues because the RealDWG SDK that you are linking to is built with the NDEBUG preprocessor symbol and Multithreaded DLL. This conflict will either give errors/warnings when building and ultimately crash your application when run.

Therefore, to get your RealDWG to build properly in debug mode, you need to make sure you have _DEBUG not set and also make sure you have Multithreaded DLL set as the Runtime Library…

Determine if point is on line bounded or unbounded

By Xiaodong Liang

We need firstly create the AcGeXXXX classes and their member functions because the abstract class AcGeEntity2d derives all the curves, arcs and lines. As such, there is an exact member function isOn(...) that can fulfill this task. Following are the codes for demo:

 

// Judge point containment in a line which is a

// bounded line segment or an unbounded line

 

//

// Name:    pointOnLine;

 

// Return value: 1 within, 0 outside, -1 error, int;

 

// Parameters ( all are input ):

// pt    -- single point to judge,

//ads_point/AcGePoint2d;

// endpt1    -- one end point of the line,

// ads_point AcGePoint2d;

//  endpt2    -- the other end point of the line,

//  ads_pointAcGePoint2d;

// bounded -- the line is bounded (not zero) or

//unbounded

 

int pointOnLine(AcGePoint2d pt,

                AcGePoint2d endpt1,

                AcGePoint2d endpt2,

                int bounded)

{

    int retCode;

 

    if( bounded )

    {

        AcGeLineSeg2d *line;

        line = new AcGeLineSeg2d(endpt1, endpt2);

 

        if(line == NULL)

            return -1;

 

        retCode = (int) line->isOn(pt);

        delete line;

    }

    else

    {

        AcGeLine2d *line;

        line = new AcGeLine2d(endpt1, endpt2);

 

        if(line == NULL)

            return -1;

 

        retCode = (int) line->isOn(pt);

        delete line;

    }

 

    return (retCode == Adesk::kTrue ? 1 : 0);

}

 

int pointOnLine( ads_point pt, ads_point endpt1,

                ads_point endpt2, int bounded)

{

    int retCode;

 

    if( bounded )

    {

        AcGeLineSeg2d *line;

        line = new AcGeLineSeg2d( asPnt2d(endpt1), asPnt2d(endpt2) );

 

        if(line == NULL)

            return -1;

 

        retCode = (int) line->isOn( asPnt2d(pt) );

    }

    else

    {

        AcGeLine2d *line;

        line = new AcGeLine2d( asPnt2d(endpt1), asPnt2d(endpt2) );

 

        if(line == NULL)

            return -1;

 

        retCode = (int) line->isOn( asPnt2d(pt) );

    }

 

    return (retCode == Adesk::kTrue ? 1 : 0);

}

Applying the current Layer Color to a new Layer using LISP

by Fenton Webb

To apply the current Layer’s colour to a newly created Layer using LISP can be done very simply like this:

; get the current layer name
(setq currentLayer (getvar "CLAYER"))
; get the ename of the current layer from the Layer table
(setq currentLayerEname (tblobjname "LAYER" currentLayer))
; extract the group code 62 (colour) from the ename using entget
(setq currentLayerColour (cdr (assoc 62 (entget currentLayerEname))))
; finally create the new layer using (command "LAYER")
(command "_.-LAYER" "_N" "TEST2" "_C" currentLayerColour "TEST2" "")

AutoCAD.NET lesson 2: User interaction - User input

By Virupaksha Aithal

The recording of the lesson 2 "User interaction - User input" of AutoCAD.NET API training Webcast that Wayne presented on 13th September is now available for download. You can download this lesson from here