Export csv with ISO-8859-1 encoding instead of UTF-8

user_5816847 0
Viewed 10398

I struggle with encoding in csv exports. I'm from the Netherlands and we use quite some trema's (e.g. ë, ï) and accents (e.g. é, ó) etc. This causes troubles when exporting to csv and open file in excel.

On macOS Mojave.

I've tried multiple encoding functions like the following.

library(stringr)
library(readr)

test <- c("Argentinië", "België", "Haïti")

test %>%
  stringi::stri_conv(., "UTF-8", "ISO-8859-1") %>%
  write.csv2("~/Downloads/test.csv")

But still, this causes weird characters:

enter image description here

stringr r character-encoding
3 Answers

How do I take care of people importing the right encoding in their Excel..?

Don’t convert to iso-8859-1 but export with readr::write_excel_csv2(). It writes the file as UTF-8, but with byte order mark (BOM), which Excel understands).

library(readr)
test <- c("Argentinië", "België", "Haïti")

I need to convert test to UTF-8, because I am on Windows.

test <- enc2utf8(test)

On MacOS test should be in UTF-8 already, as that is the native encoding.

Encoding(test)
#> [1] "UTF-8" "UTF-8" "UTF-8"

Save as an Excel csv with readr::write_excel_csv2()

write_excel_csv2(data.frame(x = test), "test.csv")

Alternatively save as xlsx with writexl::write_xlsx()

writexl::write_xlsx(data.frame(x = test), "test.xlsx")

test.csv in Excel

user_21149320

test <- c("Argentinië", "België", "Haïti")

con <- file('~/test.csv', encoding = "ISO-8859-1")

write.csv(x = test, file = con)

A call to csvlook test.csv -e "ISO-8859-1" in my bash returns | a | x | | - | ---------- | | 1 | Argentinië | | 2 | België | | 3 | Haïti |

user_40231360

do the simple thing

Just open the file in note pad and save as UTF -8 in another name, now open the saved notepad file in excel and it will ask you import, do delimiter based on your report and use , also as delimiter for columns separation and finish import. you will get your clean file

user_118708820
Related