[PD] Canvas questions

[IOhannes m zmoelnig]

On 20.02.19 17:39, Laure M. Hiendl wrote:
> Dear all,
> 
> I would like to use Pd's canvas for simple, timed text projection in a
> piece, and I have a couple of questions:
> 
> 1. I'm getting the projected text from a .txt file via the [text get]
> object and [zexy/list2symbol] to then [label $1( the canvas. It's
> working very fine but when I try to open the properties menu of the
> canvas I get this error message:
> 
> (Tcl) UNHANDLED ERROR: wrong # args: should be "pdtk_iemgui_dialog
> mytoplevel mainheader dim_header wdt min_wdt wdt_label hgt min_hgt
> hgt_label rng_header min_rng min_rng_label max_rng max_rng_label
> rng_sched lin0_log1 lilo0_label lilo1_label loadbang steady num_label
> num snd rcv gui_name gn_dx gn_dy gn_f gn_fs bcol fcol lcol"
>     while executing
> "pdtk_iemgui_dialog .gfxstub100311e20 |cnv|
> ------selectable_dimensions(pix):------ 15 1 size: 0.0 0.0 empty
>       ------visible_rect..."
>     ("uplevel" body line 1)
>     invoked from within
> "uplevel #0 $docmds"
> 
> I can open the properties menu of a new canvas, but as soon as I label
> it with another long string I get the same error message. I can edit
> all the canvas properties just fine with messages, so for me it's not
> a critical issue. I was just wondering whether I should worry about
> the stability of my patch.

no need to worry.
the problem is, that the iemguis don't properly escape spaces in the
labels, and tcl gets confused.
if you need edit the properties (without messages that is), you need to
first set the label to some string without spaces.
(alternatively you could use a non-breaking-space (" ") which looks the
same as a space but isn't one and therefore doesn't confuse the tcl parser.

but of course this is a bug and ought to be fixed.

> 
> 2. To center align the projected text I'm repositioning the canvas
> label according to the character count of the string. I've added the
> char count manually for each line in the .txt file, but I was
> wondering if there's a way to get the char count of a line
> automatically…

[list fromsymbol]
|
[list length]

but that will actually count bytes rather than characters. the numbers
are the same as long as you are using ASCII; as soon as you don't the
numbers will differ. there used to be a UTF8-to-unicode patch around
somewhere¹ that would convert utf8 bytelists (as output by [list
fromsymbol] into unicode codepoint lists (one codepoint per character;
unless you are into diacritics), that would help here.
that's probably overkill for your problem at hand though...

> 
> 3. How can I get commas "," and semicolons ";" through the [text] object?
> 

you don't.

fgmasdr
IOhannes

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20190220/9d81e83e/attachment.sig>