Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/moodymudskipper/now

Remove Exported Functions Depending On Lifecycle
https://github.com/moodymudskipper/now

Last synced: 16 days ago
JSON representation

Remove Exported Functions Depending On Lifecycle

Awesome Lists containing this project

README 

        
---

output: github_document

---



```{r, include = FALSE}

knitr::opts_chunk$set(

  collapse = TRUE,

  comment = "#>",

  fig.path = "man/figures/README-",

  out.width = "100%"

)

options(tidyverse.quiet = TRUE)

```



# now



Experimental!



Forget about lifecycles, live in the {now}. 



`now::clean_up()` removes functions of given life cycles from the namespace's exports so they're not attached with `library()` and not available or suggested with `::`.



Why would we want that ?



* You might not want to use deprecated functions, to use best current practice

* You might not want to use experimental functions, because you need to write stable code

* You might not want to run into defunct functions, that exist only to fail

* You might want to explore a package without being distracted by outdated or unstable

  features.



In short:



* `clean_up(pkgs)` is `clean_up(pkgs, remove = getOption("now.lifecycles", c("superseded", "deprecated", "defunct")))`.

* Additional possible options are `"experimental"`, `"questioning"`, and `"reexports"`, 

  the latter is not a life cycle, but allows you to drop reexports so you're forced

  to use the source package when programming.

* For instance use`clean_up(pkg, lifecycle = "defunct")` to cleanup only defunct functions, and not break anything.

* Use`clean_up_tidyverse()` to do the above on all tidyverse packages at once.

* `clean_attach()` and `clean_attach_tidyverse()` are wrappers that call library



> **It's not meant to be used in production, because it will break some dependencies that rely on the

features you've removed**



Obviously life cycles are important and there are reasons why they exist, they

just sometimes get in the way.



It works by looking at the doc and the code of a function, but we might miss

some "life-cycled" functions if the doc doesn't clearly use life cycle tags and the code

doesn't call a {lifecycle} function directly.



## Installation



Install with



``` r

remotes::install_github("moodymudskipper/now")

```



## Example



```{r, error = TRUE}

library(tidyverse)

before <- sapply(tidyverse_packages(), \(pkg) length(getNamespaceExports(pkg)))

do

location



now::clean_up_tidyverse()

after <- sapply(tidyverse_packages(), \(pkg) length(getNamespaceExports(pkg)))

do

location

dplyr::do



# expect some failures! Many packages use superseded functions in particular

library(dm)

```



To avoid the latter, either :



* call `library(dm)` before cleaning up (these dependencies would

be copied in dm's imports env so `dm::separate()` would work but `tidyr::separate()` wouldn't)

* call `clean_up_tidyverse(force_keep = c("separate", "transmute"))` above to keep the necessary dependencies around.

* call `clean_up("dm")` or `clean_attach("dm")` before `clean_up_tidyverse()` and

don't be bothered with {dm}'s re-exports



Let's take a look at how much we can expect to trim off



```{r, error = TRUE}

# removing only superseded/deprecated/defunct

left_join(enframe(before, value = "before"), enframe(after, value = "after"), by = "name") |> 

  filter(after != before) |>

  mutate(keep_ratio = sprintf("%2.0f%%", 100*after/before)) |>

  arrange(keep_ratio) |>

  knitr::kable()



# removing additionally all experimental and questioning features, along with reexports

now::clean_up_tidyverse(c("experimental", "superseded", "deprecated", "defunct", "questioning", "reexports"))

after <- sapply(tidyverse_packages(), \(pkg) length(getNamespaceExports(pkg)))

left_join(enframe(before, value = "before"), enframe(after, value = "after"), by = "name") |> 

  filter(after != before) |>

  mutate(keep_ratio = sprintf("%2.0f%%", 100*after/before)) |>

  arrange(keep_ratio) |>

  knitr::kable()

```