locate_groups: Locate multiple header groups
In ianmoran11/locatr: Easier Tidying of Data from Spreadsheets

This function adds direction annations to tidyxl data frame, specifying annotions for several headers at once. The resulting data frame will contain the following columns: .value, .direction, and .header_label. These columns indicate the value, direction and name of the header columns that will be produced in the tidy data frame produced by migrate.

locate_groups(sheet = NULL, direction = "N",
  .groupings = groupings(fmt_alignment_indent),
  .hook_if = hook_if(any(FALSE)), .hook_if_rev = hook_if(any(FALSE)),
  default_col_header_direction = "N",
  default_row_header_direction = "W", header_fill = "local_format_id")

`sheet`	data frame created by xlsx_cells
`direction`	a string indicating which type of headers are to be labelled. Options include compass direction or up/down/left/right.
`.groupings`	expressions representing how header cells are differentiated. Most naturally works with fmt_* functions. These expressions must be provided within `groupings`.
`.hook_if`	expression determining whether directions are hooked. That is, whether a header cell to the left of data cells should be annotated as "W" or "WNW". This argument contains an expression within the function `hook`. This expression must evaluate to a single value for each header group. For example, see below that any(fmt_alignment_indent == 0) is used rather that fmt_alignment_indent == 0.
`.hook_if_rev`	expression determining whether directions are hooked. That is, whether a header cell to the left of data cells should be annotated as "W" or "WNW". This argument contains an expression within the function `hook`. This expression must evaluate to a single value for each header group. For example, see below that any(fmt_alignment_indent == 0) is used rather that fmt_alignment_indent == 0.
`default_col_header_direction`	Indicates which direction is given to col headers by default. Only need if "NNW" is required, rather than "N".
`default_row_header_direction`	Indicates which direction is given to row headers by default. Only need if "WNW" is required, rather than "W".
`header_fill`	deals with merged cells. Fills in neighbouring cells if they have the same "local_format_id", "style" or are within "borders".

## Not run: 

library(tidyverse)

# Read in tidyxl data frame
xl_df <-
  locatr_example("worked-examples.xlsx") %>%
  xlsx_cells_fmt(sheets = "pivot-hierarchy") %>%
  append_fmt(fmt_alignment_indent)

# Add location annotations
xl_df <-
  xl_df %>%
  locate_data(data_type == "numeric") %>%
  locate_groups(
    direction = "W",
    .groupings = groupings(fmt_alignment_indent),
    .hook_if = hook_if(any(fmt_alignment_indent == 0))
  ) %>%
  locate(direction = "N", name = student)

# Use `migrate` to reshape the data frame such that each data cells has its own row and each
 header variable has its own column.
xl_df %>% migrate()

## End(Not run)