Rendering unicode proc template

[Doug____]

I have a situation where I needed to set up SGPLOT with 9.2 (TS2M3) using GTL. These code blocks (GTL) are processed prior to plot execution. The issue I'm having is that doing the plot this way causes unicode symbols in the variable labels (as macro variables) to be rendered literally when they appear in the legend. I've tried to use legend labels as macro variables but this did not work either as they do not resolve. Any suggestions?

[Reeza]

I'm not sure what you mean. You're using GTL but having trouble with unicode symbol rendering in the final template? Is unicode supported in 9.2?

[Doug____]

Yes it is but it renders it as a literal when I have to preprocess the template code and there are macro variables that create the legend labels.

In our system we create item stores when the program initially runs then another program we call sections creates the final outputs using item stores. Basically the guts of the table and its titles and footnotes are separate objects.

[Reeza]

I *think* you may be mixing macro and dynamic variables in GTL. Once you compile the template the macro variables resolve and they won't change again even if you call the template. If you set it as a Dynamic variable you can get it to resolve.

[Doug____]

Ok so in the code the variables are subgrpx1, 2, and 3 (macro variables). If I declare them as MVAR and use DYNAMIC in the same code block the following warning is issued:

WARNING: Symbol has already been declared as another type: subgrpx1.

[Reeza]

Can you post some code and/or a picture that illustrates your issue. Trying to follow without code or an image is difficult.

[Cynthia_sas]

And, another suggestion is that all the graph folks hang out in the forum for ODS GRAPHICS -- since this is specific to ODS GRAPHICS and since you are asking about SAS 9.2, the only people who may know the answer to your question quickly are probably over in that forum. Suggest you post your question there, possibly with some simplified code that uses a SASHELP dataset to illustrate your question.
cynthia