Set names for import_list() #164
Conversation
Exclude .ext from element name
Uses table's 'class' attribute otherwise blank
|
Appveyor build failed due to a suspected r-devel issue
Seems to build and check fine locally otherwise:
|
Codecov Report
@@ Coverage Diff @@
## master #164 +/- ##
=========================================
Coverage ? 85.42%
=========================================
Files ? 18
Lines ? 1036
Branches ? 0
=========================================
Hits ? 885
Misses ? 151
Partials ? 0
Continue to review full report at Codecov.
|
R/import_list.R
Outdated
| @@ -48,6 +48,7 @@ function(file, | |||
| setclass <- NULL | |||
| } | |||
| if (length(file) > 1) { | |||
| names(file) <- gsub(paste0("\\.", tools::file_ext(file[1]), "$"), "", file, ignore.case = TRUE) | |||
There was a problem hiding this comment.
I'm not sure I understand what's happening here. Can you explain?
There was a problem hiding this comment.
It's what the first comment in this PR is trying to explain but unfortunately the comment didn't attach itself to the precise line of code for clarity...
Also relates to Case 4 in the PR description
There was a problem hiding this comment.
I've since realised that assuming all files have the same extension somewhat defeats the purpose of ingesting multiple disparate datasets of differing file types at the same time.
file <- c("mtcars.csv", "iris.csvy", "flights.xlsx", "USArrests.csv")
my_list <- import_list(file, rbind = FALSE)
names(my_list)
#> [1] "mtcars" "iris.csvy" "flights.xlsx" "USArrests"If it's going to strip file extensions it probably makes more sense to lapply(file, tools::file_ext) and strip each respective one so that we end up with
exts <- paste0("\\.", lapply(file, tools::file_ext), "$")
names(file) <- sapply(seq_along(file), function(x) gsub(exts[x], "", file[x]))
names(file)
#> [1] "mtcars" "iris" "flights" "USArrests"Also, on second thoughts it also doesn't really seem appropriate to treat zip files differently so Case 3 should now be the same as Case 4
zipretains the file names without extension
Implemented in 258a50c
|
Can you add yourself to |
R/import_list.R
Outdated
| if (length(file) > 1) { | ||
| names(file) <- gsub(paste0("\\.", tools::file_ext(file[1]), "$"), "", file, ignore.case = TRUE) | ||
| names(file) <- strip_exts(file) |
There was a problem hiding this comment.
See #164 (comment)
There was a problem hiding this comment.
There's a function tools::file_path_sans_ext that might be useful (possibly in combination with basename()).
There was a problem hiding this comment.
Great call, thanks @HughParsonage.
@leeper ready for merge, I think
|
I was just looking for the functionality of maintaining sheet names for xls/xlsx files with import_list, and I found this PR. My fingers are crossed that it will be integrated soon. (If I can help with the integration, please let me know.) |
|
The PR was finished and has been ready to go pending a review. @billdenney you can install this pull request (and any other for that matter) as follows library(devtools)
install_github("leeper/rio", ref = github_pull(164)) |
|
Thanks for this! |
@leeper, some thoughts on the enhancement you suggested in #162. It may make more sense to keep the logic for cases 3 & 4 the same (see below) and/or do away with case 2 ... ?
PR to add functionality for imported lists to inherit data structure's names according to:
xlsandxlsxretain the sheet nameshtmlretains the table's "class" attribute if it exists otherwise blankzipretains the file names with extensionFixes #162