Export renv (#55)

[schochastics]

@chainsawriot not entirely sure yet if this is robust but it seems to be a minimal working version. I keep digging to make this more stable. Something I am wondering about is the git stuff for bioconductor. If you see other vulnerabilities, do let me know

[chainsawriot]

This is the generated json for an package. I think the brackets affect renv.

    "Packages": {
        "slam": {
            "Package": [
                "slam"
            ],
            "Version": [
                "0.1-50"
            ],
            "Source": [
                "Repository"
            ],
            "Repository": [
                "CRAN"
            ]
        }

> restore()

This project has not yet been activated.
Activating this project will ensure the project library is used during restore.
Please see `?renv::activate` for more details.

Would you like to activate this project before restore? [Y/n]: y
* Project '~/dev/misc/renvtest2' loaded. [renv 0.16.0]
* The project library is out of sync with the lockfile.
* Use `renv::restore()` to install packages recorded in the lockfile.
Error in vapply(x, f, ..., FUN.VALUE = character(1)) : 
  values must be type 'character',
 but FUN(X[[1]]) result is type 'list'
Traceback (most recent calls last):
6: restore()
5: renv_restore_report_actions(diff, current, lockfile)
4: renv_pretty_print_records_pair(lhs[names(lhs) %in% names(actions)], 
       rhs[names(rhs) %in% names(actions)], "The following package(s) will be updated:")
3: renv_pretty_print_records_pair_impl(old, new)
2: map_chr(all, function(package) {
       lhs <- old[[package]]
       rhs <- new[[package]]
       case(is.null(lhs$Source) ~ rhs$Repository %||% rhs$Source, 
           is.null(rhs$Source) ~ lhs$Repository %||% lhs$Source, 
           !is.null(rhs$Repository) ~ rhs$Repository, !is.null(rhs$Source) ~ 
               rhs$Source)
   })
1: vapply(x, f, ..., FUN.VALUE = character(1))

But I was wondering what the resolve("renv") does in the function?

[chainsawriot]

pkg_list <- vector(mode = "list", length = 1)
pkg_list[[1]][["Package"]] <- "rtoot"
pkg_list[[1]][["Version"]] <- "0.1.0"
pkg_list[[1]][["Source"]] <- "Repository"
pkg_list[[1]][["Repository"]] <- "CRAN"
jsonlite::toJSON(pkg_list)
## [{"Package":["rtoot"],"Version":["0.1.0"],"Source":["Repository"],"Repository":["CRAN"]}] 
jsonlite::toJSON(pkg_list, auto_unbox = TRUE)
## [{"Package":"rtoot","Version":"0.1.0","Source":"Repository","Repository":"CRAN"}] 

[schochastics]

ah I forgot to include auto_unbox this should be fixed now.

The resolve("renv") part I am actually not sure if this is needed like this. renv is also part of the renv lockfile so we could also kind of hard code it in?

[codecov-commenter]

Codecov Report

Merging #104 (a47fef5) into v0.2 (34b2f96) will decrease coverage by 1.78%.
The diff coverage is 48.38%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

@@            Coverage Diff             @@
##             v0.2     #104      +/-   ##
==========================================
- Coverage   97.06%   95.29%   -1.78%     
==========================================
  Files           6        6              
  Lines         819      850      +31     
==========================================
+ Hits          795      810      +15     
- Misses         24       40      +16     

Impacted Files	Coverage Δ
R/installation.R	93.11% <48.38%> (-4.58%)	⬇️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[chainsawriot]

@schochastics Great! The smoke test is now working. (But it proves our point in the documentation, a lock file is not as great as a Dockerfile and rang.R. The restoration of the lock file generated by the example in the documentation doesn't work due to misconfigured rJava. And rJava is not working well with the approach renv uses. It's not an issue we can fix though.)

About the resolve("renv"), actually, it does nothing now and the smoke test still runs. And the issue is: if someone uses renv::restore() to rehydrate a lock file from someone else, renv should have been installed anyway.

[schochastics]

sounds awesome. Yeah having renv installed to rehydrate obviously makes sense. i will remove it

[schochastics]

@chainsawriot should be done now

[chainsawriot]

@schochastics Thanks a lot!