*ERROR* problem

I am having a problem with my *ERROR* Function. If I go into VLISP and "Check Text in Editor", if comes up saying:

[CHECKING TEXT TEST.LSP loading...]
...
; warning: too few arguments: (*ERROR*)
..............................
; Check done.

Now it works just fine if I don't compile my code into VLX. But when I do, I get errors on the command line saying it is rejecting variable setting attreq for instance.

Here is what my *ERROR* function looks like:

(VL-DOC-EXPORT '*ERROR*)
(DEFUN *ERROR* (MSG)
(SETVAR "attreq" old_attreq)
(SETVAR "attdia" old_attdia)
(SETVAR "cmdecho" old_cmdecho)
(SETVAR "clayer" old_clayer)
(PROMPT "\nResetting System Variables")
) ;_ end of defun

Can someone tell me where I am going wrong?
TIA,
Craig

 

; warning: too few arguments: (*ERROR*)

Is your error function being called without the msg argument?

 

what do you mean by that?

As Bob says, the "rejected setting" is what you will get when the error
function is triggered without having previously set values for old_attreq
etc. I'd guess that something else is causing an error, and the error
handler is blowing up.

The "too few arguments" would apply if the error function is called
incorrectly. You've defined *error* as taking one argument, msg. If you call
the function in the form of (*error* "mistake") it's legal, but if you call
(*error*) you'll get "too few arguments."

Try commenting out your error handler altogether and compiling running the
program -- see what Acad's built-in error handler reports. Yours is not
doing anything with the msg argument, meaning it's suppressing reports of
what is the real error.

 

Ok, I went through and rewrote my error based on www.afralisp.com tutorial on error trapping, and changed the variables to suit my own.

Now, it works just fine, whenever I run it as a *.lsp file. BUT, when I run it compiled (with no errors on compiling), I get this:

TEST ; error: bad argument type: stringp nil

Here is what my error trapping routine looks like:

(VL-DOC-EXPORT 'INITERR)
(DEFUN INITERR ()
(SETQ old_attreq (GETVAR "attreq")
old_attdia (GETVAR "attdia")
old_cmdecho (GETVAR "cmdecho")
old_clayer (GETVAR "clayer")
) ;_ end of SETQ
(SETQ temperr *error*)
(SETQ *error* trap)
(PRINC)
) ;_ end of defun
;;*************************************************************************
(VL-DOC-EXPORT 'TRAP)
(DEFUN TRAP (errmsg)
(IF (NOT (MEMBER errmsg '("Console Break" "Function Cancelled"))
) ;_ end of not
(PRINC (STRCAT "\nError: " errmsg))
) ;_ end of if
(SETVAR "attreq" old_attreq)
(SETVAR "attdia" old_attdia)
(SETVAR "cmdecho" old_cmdecho)
(SETVAR "clayer" old_clayer)
(PRINC "\nError Resetting Enviroment ")
(TERPRI)
(SETQ *error* temperr)
(PRINC)
) ;_ end of defun
;;*************************************************************************
(VL-DOC-EXPORT 'RESET)
(DEFUN RESET ()
(SETQ *error* temperr)
(SETVAR "attreq" old_attreq)
(SETVAR "attdia" old_attdia)
(SETVAR "cmdecho" old_cmdecho)
(SETVAR "clayer" old_clayer)
(PRINC)
) ;_ end of defun

 

Sounds like there's an error in the test routine. Why not post the whole
routine here or in customer files?

 

OT: You *do* know you can localize the *ERROR* symbol, and not jump thru the
"save-old-function-redefine-function-restore-old-function" song-and-dance,
right?

P.S. How do you spell environment? <vbg>


--
R. Robert Bell


Ok, I went through and rewrote my error based on www.afralisp.com tutorial
on error trapping, and changed the variables to suit my own.

Now, it works just fine, whenever I run it as a *.lsp file. BUT, when I run
it compiled (with no errors on compiling), I get this:

TEST ; error: bad argument type: stringp nil

Here is what my error trapping routine looks like:

(VL-DOC-EXPORT 'INITERR)
(DEFUN INITERR ()
(SETQ old_attreq (GETVAR "attreq")
old_attdia (GETVAR "attdia")
old_cmdecho (GETVAR "cmdecho")
old_clayer (GETVAR "clayer")
) ;_ end of SETQ
(SETQ temperr *error*)
(SETQ *error* trap)
(PRINC)
) ;_ end of defun
;;*************************************************************************
(VL-DOC-EXPORT 'TRAP)
(DEFUN TRAP (errmsg)
(IF (NOT (MEMBER errmsg '("Console Break" "Function Cancelled"))
) ;_ end of not
(PRINC (STRCAT "\nError: " errmsg))
) ;_ end of if
(SETVAR "attreq" old_attreq)
(SETVAR "attdia" old_attdia)
(SETVAR "cmdecho" old_cmdecho)
(SETVAR "clayer" old_clayer)
(PRINC "\nError Resetting Enviroment ")
(TERPRI)
(SETQ *error* temperr)
(PRINC)
) ;_ end of defun
;;*************************************************************************
(VL-DOC-EXPORT 'RESET)
(DEFUN RESET ()
(SETQ *error* temperr)
(SETVAR "attreq" old_attreq)
(SETVAR "attdia" old_attdia)
(SETVAR "cmdecho" old_cmdecho)
(SETVAR "clayer" old_clayer)
(PRINC)
) ;_ end of defun

 

Ok, I have commented out all of my error handling, recompiled with no errors during compiling, and I still get that error when trying to run the VLX file. If I run the FAS or LSP file, they work just fine.

What could be causing this? I have exported everything with (VL-DOC-EXPORT 'functionname). The program won't even run, so im guessing it has something to do with the initialization, but can't pin point it.

 

Here is the Program....

Way too much for me to try to analyze, sorry. Maybe someone else will help
debug. If the error routine is in doubt, try writing a short program which
does nothing but test that.

But since you get the same thing with or without the custom error handler, I
suspect an error in the larger routine. The stringp nil implies that a
function which expects a string isn't getting one. (atoi 7) will generate
this error. As to why the error appears only after compilation, sorry, no
clue, I've never seen that. Maybe it's some interaction with the dcl file --
but that's only a wild guess.

 

does anyone know what would cause when I load the program I get this:

TEST v1.0 is now loaded.
Type TEST to run program.nil

why would it have "nil" at the end? And once again, its only on the VLX, and not the LSP?????

I have at least partially figured it out in VLX format, so it loads, and gets up to picking an insertion point, but then Autocad always crashes there.

So time for 1000 more hrs of debugging.

 

There is nothing you can do about the "nil" echo from the vlx. Sorry.

--
R. Robert Bell


does anyone know what would cause when I load the program I get this:

TEST v1.0 is now loaded.
Type TEST to run program.nil

why would it have "nil" at the end? And once again, its only on the VLX,
and not the LSP?????

I have at least partially figured it out in VLX format, so it loads, and
gets up to picking an insertion point, but then Autocad always crashes
there.

So time for 1000 more hrs of debugging.

 

a (princ) at the end won't get it?
daron

 

Nope.

--
R. Robert Bell


"Daron Denton" <ask, and i'll give it to you> wrote in message
a (princ) at the end won't get it?
daron

 

The vlx runs for me without an error message.

 

Your joking right? How is that possible? OK, so whats your secret?

I am running ADT 3.3, WinXP Pro.

For support path I have it mapped to the main DIFF folder.

Did you do anything different?

 

I am on vanilla 2004, NT. I dropped the whole thing in the first folder on
my search path, and loaded and ran the vlx. I did get the nil on the loading
message, and it reported not being able to find the block on my first
attempt. Looking at the code, you're prefixing a DWG/ on the block name, and
I don't have that path. But once I had manually inserted the block in my
drawing, to make it available, the program ran without any error message.

I might recommend using findfile to locate the block, rather than hard
coding the DWG/ path, for more flexibility. Sorry I can't shed any light on
your error message.

 

ya, if you use the extract part of the winzip, it will extract the zip file with the proper hierarchy/folder structure. there will be a dwg and sld folder inside of it.

now, just to double check, I did this myself. downloaded the zip file I posted to my desktop, unzipped, mapped a support path to it.

then, I do it all three ways, I load the vlx at the command prompt, drag n' drop, and appload it. the program runs just fine. now, if I take a remove the lsp and fas files from the folder, even reload the vlx, close autocad and load the vlx, etc, it won't work. if I leave the lsp/fas in there, it works just fine. did you remove the lsp and fas files before running?

 

did you remove the lsp and fas files before running?

No I didn't. After removing the lsp and fas I got the stringp nil error on
loading the vlx. I don't understand why having those files present would
make any difference, but it did!

I don't understand what is happening on compilation. I think the stringp
error would not occur unless something is actually executing when the vlx
loaded. I've never seen anything like this.