Even if files are not saved in UTF8, you can can extract information on character encoding from the file names and import the texts correctly.
require(quanteda)
require(readtext)
temp_dir contains the example files in various character encodings.
temp_dir <- tempdir()
unzip(system.file("extdata", "data_files_encodedtexts.zip", package = "readtext"), exdir = temp_dir)
list.files() returns names of all the text files (.txt) in the directory
filename <- list.files(temp_dir, "^(Indian|UDHR_).*\\.txt$")
head(filename)
## [1] "IndianTreaty_English_UTF-16LE.txt"
## [2] "IndianTreaty_English_UTF-8-BOM.txt"
## [3] "UDHR_Arabic_ISO-8859-6.txt"
## [4] "UDHR_Arabic_UTF-8.txt"
## [5] "UDHR_Arabic_WINDOWS-1256.txt"
## [6] "UDHR_Chinese_GB2312.txt"
You can extract character encoding information from the file names using R’s basic commands.
filename <- gsub(".txt$", "", filename)
encoding <- sapply(strsplit(filename, "_"), "[", 3)
head(encoding)
## [1] "UTF-16LE" "UTF-8-BOM" "ISO-8859-6" "UTF-8"
## [5] "WINDOWS-1256" "GB2312"
There is a character encoding not supported by R.
setdiff(encoding, iconvlist())
## [1] "UTF-8-BOM"
You then pass encoding to readtext() to convert various character encodings into UTF8.
data_dir <- system.file("extdata/", package = "readtext")
txt_data <- readtext(paste0(data_dir, "/data_files_encodedtexts.zip"),
encoding = encoding,
docvarsfrom = "filenames",
docvarnames = c("document", "language", "input_encoding"))
print(txt_data, n = 50)
## readtext object consisting of 36 documents and 3 docvars.
## # data.frame [36 × 5]
## doc_id text document language input_encoding
## <chr> <chr> <chr> <chr> <chr>
## 1 IndianTreaty_Engli… "\"WHEREAS, t\"… IndianTre… English UTF-16LE
## 2 IndianTreaty_Engli… "\"ARTICLE 1.\"… IndianTre… English UTF-8-BOM
## 3 UDHR_Arabic_ISO-88… "\"الديباجة\nل\… UDHR Arabic ISO-8859-6
## 4 UDHR_Arabic_UTF-8.… "\"الديباجة\nل\… UDHR Arabic UTF-8
## 5 UDHR_Arabic_WINDOW… "\"الديباجة\nل\… UDHR Arabic WINDOWS-1256
## 6 UDHR_Chinese_GB231… "\"世界人权宣言\n联合国\… UDHR Chinese GB2312
## 7 UDHR_Chinese_GBK.t… "\"世界人权宣言\n联合国\… UDHR Chinese GBK
## 8 UDHR_Chinese_UTF-8… "\"世界人权宣言\n联合国\… UDHR Chinese UTF-8
## 9 UDHR_English_UTF-1… "\"Universal \"… UDHR English UTF-16BE
## 10 UDHR_English_UTF-1… "\"Universal \"… UDHR English UTF-16LE
## 11 UDHR_English_UTF-8… "\"Universal \"… UDHR English UTF-8
## 12 UDHR_English_WINDO… "\"Universal \"… UDHR English WINDOWS-1252
## 13 UDHR_French_ISO-88… "\"Déclaratio\"… UDHR French ISO-8859-1
## 14 UDHR_French_UTF-8.… "\"Déclaratio\"… UDHR French UTF-8
## 15 UDHR_French_WINDOW… "\"Déclaratio\"… UDHR French WINDOWS-1252
## 16 UDHR_German_ISO-88… "\"Die Allgem\"… UDHR German ISO-8859-1
## 17 UDHR_German_UTF-8.… "\"Die Allgem\"… UDHR German UTF-8
## 18 UDHR_German_WINDOW… "\"Die Allgem\"… UDHR German WINDOWS-1252
## 19 UDHR_Greek_CP1253.… "\"ΟΙΚΟΥΜΕΝΙΚ\"… UDHR Greek CP1253
## 20 UDHR_Greek_ISO-885… "\"ΟΙΚΟΥΜΕΝΙΚ\"… UDHR Greek ISO-8859-7
## 21 UDHR_Greek_UTF-8.t… "\"ΟΙΚΟΥΜΕΝΙΚ\"… UDHR Greek UTF-8
## 22 UDHR_Hindi_UTF-8.t… "\"मानव अधिका\"… UDHR Hindi UTF-8
## 23 UDHR_Icelandic_ISO… "\"Mannréttin\"… UDHR Iceland… ISO-8859-1
## 24 UDHR_Icelandic_UTF… "\"Mannréttin\"… UDHR Iceland… UTF-8
## 25 UDHR_Icelandic_WIN… "\"Mannréttin\"… UDHR Iceland… WINDOWS-1252
## 26 UDHR_Japanese_CP93… "\"『世界人権宣言』\n \… UDHR Japanese CP932
## 27 UDHR_Japanese_ISO-… "\"『世界人権宣言』\n \… UDHR Japanese ISO-2022-JP
## 28 UDHR_Japanese_UTF-… "\"『世界人権宣言』\n \… UDHR Japanese UTF-8
## 29 UDHR_Japanese_WIND… "\"『世界人権宣言』\n \… UDHR Japanese WINDOWS-936
## 30 UDHR_Korean_ISO-20… "\"세 계 인 권 선 \"… UDHR Korean ISO-2022-KR
## 31 UDHR_Korean_UTF-8.… "\"세 계 인 권 선 \"… UDHR Korean UTF-8
## 32 UDHR_Russian_ISO-8… "\"Всеобщая д\"… UDHR Russian ISO-8859-5
## 33 UDHR_Russian_KOI8-… "\"Всеобщая д\"… UDHR Russian KOI8-R
## 34 UDHR_Russian_UTF-8… "\"Всеобщая д\"… UDHR Russian UTF-8
## 35 UDHR_Russian_WINDO… "\"Всеобщая д\"… UDHR Russian WINDOWS-1251
## 36 UDHR_Thai_UTF-8.txt "\"ปฏิญญาสากล\"… UDHR Thai UTF-8