lifecycle
Guidance for managing R package lifecycle according to tidyverse principles using the lifecycle package. Use when: (1) Setting up lifecycle infrastructure in a package, (2) Deprecating functions or arguments, (3) Renaming functions or arguments, (4) Superseding functions, (5) Mar
Install
npx skills add https://github.com/posit-dev/skills/tree/main/r-lib/lifecycle
claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install posit-dev-skills@llmmart
git clone https://github.com/posit-dev/skills.git
The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole posit-dev/skills collection as a plugin from our marketplace. Git is the plain clone.
Skill manifest
R Package Lifecycle Management
Manage function and argument lifecycle using tidyverse conventions and the lifecycle package.
Setup
Check if lifecycle is configured by looking for lifecycle-*.svg files in man/figures/.
If not configured, run:
usethis::use_lifecycle()
This:
- Adds lifecycle to
Importsin DESCRIPTION - Adds
@importFrom lifecycle deprecatedto the package documentation file - Copies badge SVGs to
man/figures/
Lifecycle Badges
Insert badges in roxygen2 documentation:
#' @description
#' `r lifecycle::badge("experimental")`
#' `r lifecycle::badge("deprecated")`
#' `r lifecycle::badge("superseded")`
For arguments:
#' @param old_arg `r lifecycle::badge("deprecated")` Use `new_arg` instead.
Only badge functions/arguments whose stage differs from the package's overall stage.
Deprecating a Function
- Add badge and explanation to
@description:
#' Do something
#'
#' @description
#' `r lifecycle::badge("deprecated")`
#'
#' `old_fun()` was deprecated in mypkg 1.0.0. Use [new_fun()] instead.
#' @keywords internal
- Add
deprecate_warn()as first line of function body:
old_fun <- function(x) {
lifecycle::deprecate_warn("1.0.0", "old_fun()", "new_fun()")
new_fun(x)
}
- Show migration in examples:
#' @examples
#' old_fun(x)
#' # ->
#' new_fun(x)
Deprecation Functions
| Function | When to Use |
|---|---|
deprecate_soft() |
First stage; warns only direct users and during tests |
deprecate_warn() |
Standard deprecation; warns once per 8 hours |
deprecate_stop() |
Final stage before removal; errors with helpful message |
Deprecation workflow for major releases:
- Search
deprecate_stop()- consider removing function entirely - Replace
deprecate_warn()withdeprecate_stop() - Replace
deprecate_soft()withdeprecate_warn()
Renaming a Function
Move implementation to new name, call from old name with deprecation:
#' @description
#' `r lifecycle::badge("deprecated")`
#'
#' `add_two()` was renamed to `number_add()` for API consistency.
#' @keywords internal
#' @export
add_two <- function(x, y) {
lifecycle::deprecate_warn("1.0.0", "add_two()", "number_add()")
number_add(x, y)
}
#' Add two numbers
#' @export
number_add <- function(x, y) {
x + y
}
Deprecating an Argument
Use deprecated() as default value with is_present() check:
#' @param path `r lifecycle::badge("deprecated")` Use `file` instead.
write_file <- function(x, file, path = deprecated()) {
if (lifecycle::is_present(path)) {
lifecycle::deprecate_warn("1.4.0", "write_file(path)", "write_file(file)")
file <- path
}
# ... rest of function
}
Renaming an Argument
add_two <- function(x, y, na_rm = TRUE, na.rm = deprecated()) {
if (lifecycle::is_present(na.rm)) {
lifecycle::deprecate_warn("1.0.0", "add_two(na.rm)", "add_two(na_rm)")
na_rm <- na.rm
}
sum(x, y, na.rm = na_rm)
}
Superseding a Function
For functions with better alternatives that shouldn't be removed:
#' Gather columns into key-value pairs
#'
#' @description
#' `r lifecycle::badge("superseded")`
#'
#' Development on `gather()` is complete. For new code, use [pivot_longer()].
#'
#' `df %>% gather("key", "value", x, y, z)` is equivalent to
#' `df %>% pivot_longer(c(x, y, z), names_to = "key", values_to = "value")`.
No warning needed - just document the preferred alternative.
Marking as Experimental
#' @description
#' `r lifecycle::badge("experimental")`
cool_function <- function() {
lifecycle::signal_stage("experimental", "cool_function()")
# ...
}
Testing Deprecations
Test that deprecated functions work and warn appropriately:
test_that("old_fun is deprecated", {
expect_snapshot({
x <- old_fun(1)
expect_equal(x, expected_value)
})
})
Suppress warnings in existing tests:
test_that("old_fun returns correct value", {
withr::local_options(lifecycle_verbosity = "quiet")
expect_equal(old_fun(1), expected_value)
})
Deprecation Helpers
For deprecations affecting many functions (e.g., removing a common argument), create an internal helper:
warn_for_verbose <- function(
verbose = TRUE,
env = rlang::caller_env(),
user_env = rlang::caller_env(2)
) {
if (!lifecycle::is_present(verbose) || isTRUE(verbose)) {
return(invisible())
}
lifecycle::deprecate_warn(
when = "2.0.0",
what = I("The `verbose` argument"),
details = c(
"Set `options(mypkg_quiet = TRUE)` to suppress messages.",
"The `verbose` argument will be removed in a future release."
),
user_env = user_env
)
invisible()
}
Then use in affected functions:
my_function <- function(..., verbose = deprecated()) {
warn_for_verbose(verbose)
# ...
}
Custom Deprecation Messages
For non-standard deprecations, use I() to wrap custom text:
lifecycle::deprecate_warn(
when = "1.0.0",
what = I('Setting option "pkg.opt" to "foo"'),
with = I('"pkg.new_opt"')
)
The what fragment must work with "was deprecated in..." appended.
Reference
See references/lifecycle-stages.md for detailed stage definitions and transitions.
Files (skills)
-
references
-
lifecycle-stages.md 2.2 KB
# Lifecycle Stages Reference ## Table of Contents - [Stage Diagram](#stage-diagram) - [Stable](#stable) - [Experimental](#experimental) - [Deprecated](#deprecated) - [Superseded](#superseded) - [Retired Stages](#retired-stages) ## Stage Diagram ``` experimental --> stable --> deprecated | +--> superseded ``` Lifecycle stages apply to packages, functions, arguments, or specific argument values. ## Stable The default stage. Functions are stable when the author is satisfied with the interface. **Promises:** - Breaking changes avoided where possible - Breaking changes happen gradually through deprecation **Breaking changes** reduce working code: removing functions/arguments, narrowing valid inputs, or changing output types. **Not breaking changes:** Bug fixes that happen to break code relying on buggy behavior. ## Experimental For functions released without long-term stability promises. Author reserves the right to make breaking changes without deprecation cycle. **Guidelines:** - Packages with version < 1.0.0 are at least somewhat experimental - Non-CRAN packages require active relationship with maintainers ## Deprecated Function has a better alternative and is scheduled for removal. **Deprecation stages (in order):** 1. **Soft deprecated** (`deprecate_soft()`) - Warns only: - Users calling from global environment - Developers running tests - Does NOT warn when called indirectly by another package 2. **Deprecated** (`deprecate_warn()`) - Warns all users once per 8 hours 3. **Defunct** (`deprecate_stop()`) - Errors with helpful message about replacement ## Superseded Function has a known better alternative but won't be removed. Receives only critical bug fixes. **Key difference from deprecated:** - No warning emitted - Documentation explains recommended alternative - Function preserved indefinitely (safer than stable in some ways) **Examples:** `tidyr::gather()` and `tidyr::spread()` superseded by `pivot_longer()` and `pivot_wider()`. ## Retired Stages No longer recommended: - **Questioning** - Author uncertain about function; not actionable for users - **Maturing** - Between experimental and stable; unclear actionable information
-
-
SKILL.md 5.8 KB
--- name: lifecycle description: "Guidance for managing R package lifecycle according to tidyverse principles using the lifecycle package. Use when: (1) Setting up lifecycle infrastructure in a package, (2) Deprecating functions or arguments, (3) Renaming functions or arguments, (4) Superseding functions, (5) Marking functions as experimental, (6) Understanding lifecycle stages (stable, experimental, deprecated, superseded), or (7) Writing deprecation helpers for complex scenarios." metadata: author: Garrick Aden-Buie (@gadenbuie) version: "1.1" license: MIT --- # R Package Lifecycle Management Manage function and argument lifecycle using tidyverse conventions and the lifecycle package. ## Setup Check if lifecycle is configured by looking for `lifecycle-*.svg` files in `man/figures/`. If not configured, run: ```r usethis::use_lifecycle() ``` This: - Adds lifecycle to `Imports` in DESCRIPTION - Adds `@importFrom lifecycle deprecated` to the package documentation file - Copies badge SVGs to `man/figures/` ## Lifecycle Badges Insert badges in roxygen2 documentation: ```r #' @description #' `r lifecycle::badge("experimental")` #' `r lifecycle::badge("deprecated")` #' `r lifecycle::badge("superseded")` ``` For arguments: ```r #' @param old_arg `r lifecycle::badge("deprecated")` Use `new_arg` instead. ``` Only badge functions/arguments whose stage differs from the package's overall stage. ## Deprecating a Function 1. Add badge and explanation to `@description`: ```r #' Do something #' #' @description #' `r lifecycle::badge("deprecated")` #' #' `old_fun()` was deprecated in mypkg 1.0.0. Use [new_fun()] instead. #' @keywords internal ``` 2. Add `deprecate_warn()` as first line of function body: ```r old_fun <- function(x) { lifecycle::deprecate_warn("1.0.0", "old_fun()", "new_fun()") new_fun(x) } ``` 3. Show migration in examples: ```r #' @examples #' old_fun(x) #' # -> #' new_fun(x) ``` ## Deprecation Functions | Function | When to Use | | ------------------ | ------------------------------------------------------- | | `deprecate_soft()` | First stage; warns only direct users and during tests | | `deprecate_warn()` | Standard deprecation; warns once per 8 hours | | `deprecate_stop()` | Final stage before removal; errors with helpful message | **Deprecation workflow for major releases:** 1. Search `deprecate_stop()` - consider removing function entirely 2. Replace `deprecate_warn()` with `deprecate_stop()` 3. Replace `deprecate_soft()` with `deprecate_warn()` ## Renaming a Function Move implementation to new name, call from old name with deprecation: ```r #' @description #' `r lifecycle::badge("deprecated")` #' #' `add_two()` was renamed to `number_add()` for API consistency. #' @keywords internal #' @export add_two <- function(x, y) { lifecycle::deprecate_warn("1.0.0", "add_two()", "number_add()") number_add(x, y) } #' Add two numbers #' @export number_add <- function(x, y) { x + y } ``` ## Deprecating an Argument Use `deprecated()` as default value with `is_present()` check: ```r #' @param path `r lifecycle::badge("deprecated")` Use `file` instead. write_file <- function(x, file, path = deprecated()) { if (lifecycle::is_present(path)) { lifecycle::deprecate_warn("1.4.0", "write_file(path)", "write_file(file)") file <- path } # ... rest of function } ``` ## Renaming an Argument ```r add_two <- function(x, y, na_rm = TRUE, na.rm = deprecated()) { if (lifecycle::is_present(na.rm)) { lifecycle::deprecate_warn("1.0.0", "add_two(na.rm)", "add_two(na_rm)") na_rm <- na.rm } sum(x, y, na.rm = na_rm) } ``` ## Superseding a Function For functions with better alternatives that shouldn't be removed: ```r #' Gather columns into key-value pairs #' #' @description #' `r lifecycle::badge("superseded")` #' #' Development on `gather()` is complete. For new code, use [pivot_longer()]. #' #' `df %>% gather("key", "value", x, y, z)` is equivalent to #' `df %>% pivot_longer(c(x, y, z), names_to = "key", values_to = "value")`. ``` No warning needed - just document the preferred alternative. ## Marking as Experimental ```r #' @description #' `r lifecycle::badge("experimental")` cool_function <- function() { lifecycle::signal_stage("experimental", "cool_function()") # ... } ``` ## Testing Deprecations Test that deprecated functions work and warn appropriately: ```r test_that("old_fun is deprecated", { expect_snapshot({ x <- old_fun(1) expect_equal(x, expected_value) }) }) ``` Suppress warnings in existing tests: ```r test_that("old_fun returns correct value", { withr::local_options(lifecycle_verbosity = "quiet") expect_equal(old_fun(1), expected_value) }) ``` ## Deprecation Helpers For deprecations affecting many functions (e.g., removing a common argument), create an internal helper: ```r warn_for_verbose <- function( verbose = TRUE, env = rlang::caller_env(), user_env = rlang::caller_env(2) ) { if (!lifecycle::is_present(verbose) || isTRUE(verbose)) { return(invisible()) } lifecycle::deprecate_warn( when = "2.0.0", what = I("The `verbose` argument"), details = c( "Set `options(mypkg_quiet = TRUE)` to suppress messages.", "The `verbose` argument will be removed in a future release." ), user_env = user_env ) invisible() } ``` Then use in affected functions: ```r my_function <- function(..., verbose = deprecated()) { warn_for_verbose(verbose) # ... } ``` ## Custom Deprecation Messages For non-standard deprecations, use `I()` to wrap custom text: ```r lifecycle::deprecate_warn( when = "1.0.0", what = I('Setting option "pkg.opt" to "foo"'), with = I('"pkg.new_opt"') ) ``` The `what` fragment must work with "was deprecated in..." appended. ## Reference See `references/lifecycle-stages.md` for detailed stage definitions and transitions.
Comments (0)
Sign in to join the conversation.
Reviews (0)
No reviews yet.
No comments yet.