Potential issue with new ggplot2 version?

[teunbrand]

Hi there,

We've been preparing a new major release for ggplot2 and a reverse dependency check marked an issue with the flextable package.

The following information was received from the check:

## Newly broken

*   checking examples ... ERROR
    ```
    Running examples in ‘flextable-Ex.R’ failed
    The error most likely occurred in:
    
    > ### Name: gg_chunk
    > ### Title: 'ggplots' chunk wrapper
    > ### Aliases: gg_chunk
    > 
    > ### ** Examples
    > 
    > library(data.table)
    ...
    Loading required package: ggplot2
    a flextable object.
    col_keys: `Species`, `gg` 
    header has 1 row(s) 
    body has 3 row(s) 
    original dataset sample: 
    Error in format.default(unlist(xx), ...) : 
      c("Found no format() method for class \"ggplot2::ggplot\"", "Found no format() method for class \"ggplot\"", "Found no format() method for class \"ggplot2::gg\"", "Found no format() method for class \"S7_object\"", "Found no format() method for class \"gg\"")
    Calls:  ... format -> format.default -> lapply -> FUN -> format.default
    Execution halted
    ```

However, when I input that example in the R console, it seems to work fine. The error does not show up when I run devtools::run_tests() and only appears to pop up during R CMD Check. I also couldn't find the offending call in flextable's code ( format.default(unlist(xx), ...)). I'm a bit at a loss about what should be done about this, but I think the very minimum is to report here that there might possibly be an incompatibility.

You can test your code with the development version of ggplot2 by installing it as follows:

# install.packages("pak")
pak::pak("tidyverse/ggplot2")

We aim to release the new ggplot2 version in about 2 weeks, and hope you can submit a fix to CRAN around that time. Hopefully this will inform you in a timely manner.

Best wishes,
Teun

[davidgohel]

thanks @teunbrand I will have a look

first priority is to find free days to adapt ggiraph to changes in ggplot2 and then I'll see if I can spend time on flextable, if not, I will drop the feature :)

[teunbrand]

That sounds like a sensible choice to me :)
Does it need much adapting after davidgohel/ggiraph#319 still?
I was looking into a flextable solution and helping out as well but couldn't trace the source of the error correctly.

[davidgohel]

Ah!

I think the correct minimal example for this case (only talking about flextable issue you met while revchecking) would be

library(ggplot2)
z <- iris
z$gg <- list(ggplot())
print(z) # this need to be eval. in the R console, no Rmd, no quarto, etc

[teunbrand]

Ah brilliant! This should be fixed on ggplot2's side then. Thanks for your help in figuring this out!