Modelspace viewport count by activex ...

[michael puckett]

Has anyone else found the results bogus or unreliable?

(vla-get-count
(vla-get-viewports
(vla-get-activedocument
(vlax-get-acad-object)
)
)
)

I have a drawing with 4 modelspace vports; *NO* viewports in paperspace ( aside from paperspace itself ), yet the above code reports 27. I've toggled the active space back and forth, re-opened the dwg. Done an audit / recover etc. No change.

Other drawings correctly report the modelspace vport count.

Is it a flaky method or am I misinterpreting the results / failing to understand the vla-get-viewports method ( collection of all <modelspace> viewports in the drawing )?

PS: I initially encountered the problem using objectdbx, ie:

(vla-get-count
(vla-get-viewports DBXDOC)
)

but the same values ( one would hope ) are realized opening the dwg in the editor and using

(vla-get-count
(vla-get-viewports
(vla-get-activedocument
(vlax-get-acad-object)
)
)
)

Thanks for any enlightening thoughts fellow lithpers.

Michael.

 

[Rudy Tovar]

Just so you know, I'm not trying to be funny, but doesn't seem everything is running as smooth as it should.

They (Autodesk) wrote a disclaimer regarding the functions they found to be flawed, leaving us the rest to find.

 

[ruul morawetz]

Has anyone else found the results bogus or unreliable?

didn't test for long but here's what i noticed:

tilemode 1, vports from 1 to 4 - function returns 1 (last state)
tm 0 - function returns 4

so the function only returns the correct count when switching tilemode (vanilla 2004 e-international)

same with saved viewport configurations.
tilemode 1 - vports saved & changed
after switching tilemode correct count of sum of all viewports (active plus saved)

and: reliable so far (10 minutes ;-).

btw has nothing to do with LAYOUTREGENCTL

 

[michael puckett]

As much as I might enjoy jumping on the "bash Autodesk" wagon, it appears that the dwg does have 27 vport objects:

I ran this:

(vlax-for viewport
(vla-get-viewports
(vla-get-activedocument
(vlax-get-acad-object)
)
)
(print
(vl-remove-if 'null
(mapcar
'(lambda (x) (if (member (car x) '(-1 0 5)) x))
(entget (vlax-vla-object->ename viewport))
)
)
)
)

and this was the result:

((-1 . <Entity name: 4091ef10>) (0 . "VPORT") (5 . "A1E2")) ((-1 . <Entity name: 4091ef18>) (0 . "VPORT") (5 . "A1E3")) ((-1 . <Entity name: 4091ef20>) (0 . "VPORT") (5 . "A1E4")) ((-1 . <Entity name: 4091ef28>) (0 . "VPORT") (5 . "A1E5")) ((-1 . <Entity name: 4091ef30>) (0 . "VPORT") (5 . "A1E6")) ((-1 . <Entity name: 4091ef38>) (0 . "VPORT") (5 . "A1E7")) ((-1 . <Entity name: 4091ef40>) (0 . "VPORT") (5 . "A1E8")) ((-1 . <Entity name: 4091ef48>) (0 . "VPORT") (5 . "A1E9")) ((-1 . <Entity name: 4091ef50>) (0 . "VPORT") (5 . "A1EA")) ((-1 . <Entity name: 4091ef58>) (0 . "VPORT") (5 . "A1EB")) ((-1 . <Entity name: 4091ef60>) (0 . "VPORT") (5 . "A1EC")) ((-1 . <Entity name: 4091ef68>) (0 . "VPORT") (5 . "A1ED")) ((-1 . <Entity name: 4091ef70>) (0 . "VPORT") (5 . "A1EE")) ((-1 . <Entity name: 4091ef78>) (0 . "VPORT") (5 . "A1EF")) ((-1 . <Entity name: 4091ef80>) (0 . "VPORT") (5 . "A1F0")) ((-1 . <Entity name: 4091ef88>) (0 . "VPORT") (5 . "A1F1")) ((-1 . <Entity name: 4091ef90>) (0 . "VPORT") (5 . "A1F2")) ((-1 . <Entity name: 4091ef98>) (0 . "VPORT") (5 . "A1F3")) ((-1 . <Entity name: 4091efa0>) (0 . "VPORT") (5 . "A1F4")) ((-1 . <Entity name: 4091efa8>) (0 . "VPORT") (5 . "A1F5")) ((-1 . <Entity name: 4091efb0>) (0 . "VPORT") (5 . "A1F6")) ((-1 . <Entity name: 4091efb8>) (0 . "VPORT") (5 . "A1F7")) ((-1 . <Entity name: 4091f000>) (0 . "VPORT") (5 . "A1F8")) ((-1 . <Entity name: 40926ef0>) (0 . "VPORT") (5 . "BA6E")) ((-1 . <Entity name: 40926ef8>) (0 . "VPORT") (5 . "BA6F")) ((-1 . <Entity name: 40926f00>) (0 . "VPORT") (5 . "BA70")) ((-1 . <Entity name: 40926f08>) (0 . "VPORT") (5 . "BA71"))

Why there are 27 in the dwg when modelspace only sports 4 (and none in paperspace) is a mystery at this point. Residual artifacts from earlier vport command sequences?

Hmmmm ...

So what's new, yeah that's not the only screwy function.

Seems to be something missing...not your code.

Has anyone else found the results bogus or unreliable?

< snippage >

 

[michael puckett]

Thanks Luis, but I'm writing an auditing program for our company using objectdbx

"Luis Esquivel" <www.draftteam.com> wrote:
why not use: (vports)

 

[Rudy Tovar]

So what you're saying is that there's 23 hidden vports or stored?

 

[Doug Broad]

Michael,
Do you want the number of paperspace viewports? If so then you don't want vla-get-viewports. Viewports are objects that contain the view settings of vports in model space. Each saved viewport configuration has a number of viewports associated with it. The active viewport configuration also (I think) has a copy of the viewports that have been saved. There is also a relationship between layouts and model space viewports but I don't know the specifics.
Does that help?

 

[michael puckett]

Hi Doug,

What I'm trying to do is get the modelspace vport count.

But - your info helped. Lots. Said drawing has 12 named viewports in modelspace ( including the *Active* Viewport, which sports 4 ).

Between all of the named viewports there are precisely 27 possible vports.

THANKS

( and thanks to Luis and Rudy who prompted me to think of other things too ).

Michael,
Do you want the number of paperspace viewports? If so then you don't want vla-get-viewports. Viewports are objects that contain the view settings of vports in model space. Each saved viewport configuration has a number of viewports associated with it. The active viewport configuration also (I think) has a copy of the viewports that have been saved. There is also a relationship between layouts and model space viewports but I don't know the specifics.
Does that help?

 

[michael puckett]

Oops, Ruul too

< snip >

( and thanks to Luis and Rudy who prompted me to think of other things too ).

< snip >

 

[michael puckett]

B y o b j e c t d b x .

But thanks

"Luis Esquivel" <www.draftteam.com> wrote:
Michael,

return the number of viewports in model:

(setvar "tilemode" 1)
(length (vports))

 

[michael puckett]

No worries mang! Appreciate you dropping by my thread; cheers

"Luis Esquivel" <www.draftteam.com> wrote:

Sorry SIR I did not read your previous... it won't happen again... I promised!... cross my heart! ;-)

 

[Doug Broad]

Glad to help.
Look at the object model for more "good" info and you might find this revealing:

(vlax-map-collection
(vla-get-viewports
(vla-get-activedocument
(vlax-get-acad-object)))
'vlax-dump-object)

 

[michael puckett]

While this may not be pretty it works ...

(setq Count 0)
(vlax-for vport (vla-get-viewports OBJDBX)
(if (eq "*ACTIVE" (strcase (vla-get-name vport)))
(setq Count (1+ Count))
)
)

I use 'vlax-dump-object quite extensively, it is most useful. Between it, the object model map and my little atoms.vlx app there are not too many secrets

Thanks again Doug.

Glad to help.
Look at the object model for more "good" info and you might find this revealing:

(vlax-map-collection
(vla-get-viewports
(vla-get-activedocument
(vlax-get-acad-object)))
'vlax-dump-object)

 

[Doug Broad]

That seems the best technique. The current configuration is a viewport and if split into other viewports, each viewport has the same name, "*ACTIVE"....

You're welcome.

 

[Mark Propst]

saved for undo?

 

[ruul morawetz]

Something else: via DBX this won't be an issue I assume, but when using either of your approaches in an open drawing how do you guarantee that you really get the current state from AutoCAD?

the 1st function [using (vla-get-count ...] and also the 2nd one [stepping through viewport collection] return a wrong result (1st: wrong number, 2nd: nil) when VPORTS is used in MS to save/delete/change the current configuration. Toggling of TILEMODE to refresh and get the correct number was necessary for me.

(vports) returns different lists, too. all data present but the numbering is different. (3 .. 4 .. 2 .. 5 ..) versus (2 .. 3 .. 4 .. 5 ..)

I haven't noticed that behaviour before....
and: as my old'n'rusty (vports) using functions still work, I don't worry too much. ;-)

 

[michael puckett]

I cannot think of any advantage using activex over vanilla in the case of determining the number of vports in modelspace *IF* the dwg is open in the AutoCAD editor. I would use (length (vports)) < in the proper space of course

myself.

I have found that in order to get reliable vport count results using activex in the AutoCAD editor < on the active document > you must toggle tilemode prior to the count < regen, regenall is not enough to force the database to update >.

Finally, this ...

(defun OBJDBX-GetVportCount ( DBXDOC / count )
; DBXDOC must be a valid ObjectDBX,
; object and a valid doc must be opened
; prior to this function call. No error
; checking is employed in this function
; in the interests of performance
(setq count 0)
(vlax-for vport (vla-get-viewports DBXDOC)
(if (eq "*ACTIVE" (strcase (vla-get-name vport)))
(setq count(1+ count))
)
)
count
)

Has proven reliable for objectdbx queries so far < 1 day of use >.

Cheers.

Something else: via DBX this won't be an issue I assume, but when using either of your approaches in an open drawing how do you guarantee that you really get the current state from AutoCAD?

the 1st function [using (vla-get-count ...] and also the 2nd one [stepping through viewport collection] return a wrong result (1st: wrong number, 2nd: nil) when VPORTS is used in MS to save/delete/change the current configuration. Toggling of TILEMODE to refresh and get the correct number was necessary for me.

(vports) returns different lists, too. all data present but the numbering is different. (3 .. 4 .. 2 .. 5 ..) versus (2 .. 3 .. 4 .. 5 ..)

I haven't noticed that behaviour before....
and: as my old'n'rusty (vports) using functions still work, I don't worry too much. ;-)

 

[Eric Schneider]

I like dbview.arx too but I don't think it works in 2004.

 

[Luis]

Eric,

I think you can email Shaan Hurley and ask him for the A2004 version.