Update vignette w non-ASCII chars

[majazaloznik]

Add subsection to the "Things to be careful about" section explaining issues with non-ASCII characters in the db responses. See #165

[jonkeane]

Thanks for this! I should have looked a bit more closely on the issue (I was distracted by the {bit64} problem, I think!)

I'm curious if maybe this is an encoding issue specific to your database or runners. Do you happen to have an example of a github actions run that failed remotely when it worked locally? We might even be able to resolve this from happening without needing to manually change up fixtures

I just tried the following in a console and these seemed to go fine (I'm on MacOS, and I believe unicode is being used throughout the saving and then reading process).

> dput("😀", file = "foo.R", control = c("all", "hexNumeric"))
> source("foo.R")$value
[1] "😀"
> dput("é", file = "foo.R", control = c("all", "hexNumeric"))
> source("foo.R")$value
[1] "é"
> dput("č", file = "foo.R", control = c("all", "hexNumeric"))
> source("foo.R")$value
[1] "č"
> dput(data.frame(col=c("č", "é", "😀")), file = "foo.R", control = c("all", "hexNumeric"))
> source("foo.R")$value
  col
1   č
2   é
3  😀

[majazaloznik]

Here's the offending actions: the one before last (actually at least two if not three) were passing locally but not remotely (sorry for the very descriptive commit messages 😄 ). The last one then passed on both.
It's not the mocks directly that are causing the issue, but it's a second order consequence, when they are used as inputs into other functions.. I think.

The substantive change in the last commit was the manual fix of that one mock file. I also added some print() statements for debugging and here's the output of one of them locally:

# A tibble: 6 x 3
  tab_dim_id level_value level_text                                             
       <int> <chr>       <chr>                                                  
1          2 V           Teko�e cene (mio EUR)                                  
2          2 Y           Stalne cene predhodnega leta (mio EUR)                 
3          2 L           Stalne cene, referen�no leto 2010 (mio EUR)            
4          2 G4          Stopnja rasti obsega glede na enako �etrtletje predhod~
5          2 G1          Stopnja rasti obsega glede na predhodno �etrtletje (%) 
6          2 GO4         Prispevek k rasti obsega BDP glede na enako �etrtletje~

And remotely:

# A tibble: 6 x 3
  tab_dim_id level_value level_text                                             
       <int> <chr>       <chr>                                                  
1          2 V           Tekoče cene (mio EUR)                                  
2          2 Y           Stalne cene predhodnega leta (mio EUR)                 
3          2 L           Stalne cene, referenčno leto 2010 (mio EUR)            
4          2 G4          Stopnja rasti obsega glede na enako četrtletje predhod~
5          2 G1          Stopnja rasti obsega glede na predhodno četrtletje (%) 
6          2 GO4         Prispevek k rasti obsega BDP glede na enako četrtletje~

Locally I'm on a Windows machine and I get this btw:

> dput(data.frame(col=c("č", "é", "😀")), file = "foo.R", control = c("all", "hexNumeric"))
> source("foo.R")$value
           col
1            č
2            é
3 <U+0001F600>

Thanks for taking the time to look into this!