Making the headers much nicer in the tables

[lilyclements]

Should be a very small one if @Patowhiz and @rdstern agree!

Currently, if we run the tab_spanner_delim in the tables dialogs, then it automatically splits by "". This is ok, but it might cause issues if a user naturally has an "" in their variable name, or might cause issues if you want to split by a summary when it is something like "non_count_missing".

This can be fixed very easily by changing the separator in R. We just need the separator in the pivot_wider to be the same as in tab_spanner_delim. I want to suggest we do a small fix like "XXXXX", but this is assuming that no one would ever use "XXXXX" in their column headers! Is there a suggestion of what we can put in here?

E.g., we have this at the moment (except we don't call names_sep = "_" in pivot_wider as that's the default):

summary_table <- data_book$summary_table(data_name="survey", columns_to_summarise=c("size","fert","yield"), factors=c("village","variety","fertgrp"), treat_columns_as_factor=TRUE, summaries=c("summary_mean", "summary_count_non_missing"), drop = FALSE)

pivoted_table <- summary_table %>% tidyr::pivot_wider(names_from = c("village", "summary"),
                                                      values_from = value,
                                                      names_sort = TRUE,
                                                      names_sep = "_")

pivoted_table %>%
                 gt::gt() %>%
                 gt::tab_spanner_delim(delim = "_")

giving

If we set that separator to be a string, then we will not have this issue. E.g., if I generate a random 16 digit code, we can tell thet separator to be 8wu7sldpo9et7ydb

summary_table <- data_book$summary_table(data_name="survey", columns_to_summarise=c("size","fert","yield"), factors=c("village","variety","fertgrp"), treat_columns_as_factor=TRUE, summaries=c("summary_mean", "summary_count_non_missing"), drop = FALSE)

pivoted_table <- summary_table %>% tidyr::pivot_wider(names_from = c("village", "summary"),
                                                      values_from = value,
                                                      names_sort = TRUE,
                                                      names_sep = "8wu7sldpo9et7ydb")

pivoted_table %>%
                 gt::gt() %>%
                 gt::tab_spanner_delim(delim = "8wu7sldpo9et7ydb")

This is a very easy fix! We just need to:

1. Add in names_sep = "8wu7sldpo9et7ydb" in the pivot_wider
2. Change delim = "_" to delim = "8wu7sldpo9et7ydb" in the tab_spanner_delim

@Patowhiz @rdstern thoughts?