BERT-R中如何隐藏部分函数？如何控制functions.R函数向Excel的暴露？

Great questions—controlling function exposure in BERT-R is totally doable, and it’s a common need to keep internal helper functions hidden from Excel users. Let’s walk through how to handle both scenarios:

1. Hiding Specific Functions in BERT-R #

By default, BERT-R automatically scans your functions.R file and registers every top-level R function (those in the global environment) with Excel. To hide functions, you need to override this automatic behavior and explicitly choose which functions to expose. Here are a few reliable approaches:

Option 1: Manually Register Only the Functions You Want #

Use the bert::bert_register_function() method to explicitly register the functions you want Excel to see. Keep internal helper functions out of the global environment (either wrapped in a local scope or stored in a separate file) so they don’t get picked up by BERT’s auto-scan.

Example functions.R:

# Wrap internal helper functions in a local scope to keep them hidden
local({
  hidden_calculator <- function(input) {
    return(input * 1.5)
  }
  
  # Define the public function that Excel will use
  assign("calculate_margin", function(revenue) {
    return(hidden_calculator(revenue) - 100)
  }, envir = .GlobalEnv)
})

# Explicitly register the public function with Excel
bert::bert_register_function(calculate_margin)

Option 2: Split Code Across Files #

Move internal functions to a separate file (like utils.R) and source it in functions.R, then only register your public functions. This keeps your code organized and ensures hidden functions never enter the global environment.

• utils.R (hidden from Excel):

hidden_data_cleaner <- function(raw_data) {
  return(na.omit(raw_data))
}

• functions.R (exposed to Excel):

# Import internal functions without exposing them
source("utils.R")

# Define public function
clean_excel_data <- function(data_range) {
  return(hidden_data_cleaner(data_range))
}

# Register the public function
bert::bert_register_function(clean_excel_data)

2. Hiding Existing Functions in functions.R #

Absolutely—you can hide any existing functions in functions.R and pick only the ones to expose. The key is to prevent the functions you want hidden from being in the global environment (where BERT scans for functions to register).

If you already have functions in functions.R that you want to hide, simply refactor them into a local scope or move them to a separate helper file, then use bert::bert_register_function() to explicitly register the functions you want Excel to access.

For example, if you had this original functions.R:

# Function I want to hide
old_hidden_function <- function(x) {
  return(x + 5)
}

# Function I want to expose
old_public_function <- function(y) {
  return(old_hidden_function(y) * 2)
}

Refactor it to:

# Hide the internal function in a local scope
local({
  old_hidden_function <- function(x) {
    return(x + 5)
  }
  
  assign("old_public_function", function(y) {
    return(old_hidden_function(y) * 2)
  }, envir = .GlobalEnv)
})

# Only register the public function
bert::bert_register_function(old_public_function)

Now only old_public_function will show up in Excel—old_hidden_function stays hidden.

内容的提问来源于stack exchange，提问作者user9630195