search

quanteda / readtext

an R package for reading text files
https://readtext.quanteda.io
120 stars 28 forks source link

Can't read "European" csv #170

Open chainsawriot opened 1 year ago

chainsawriot commented 1 year ago

The so-called "European" csv is delimited by semi-colon, because comma is used for decimal.

The underlying fread can actually handle this. Despite the fact that the documentation of readtext says "..." will be passed to fread, it can't for this case.

content <- "number;country;platform;tandem;csystem_dummy;msystem_rank;csystem_rank;arena_rank;issdriven;prefdriven;charcount;ic_ordinal;ic_categorized;post;main_language;languages;is_reliable;hashtag_count;is_redacted;WC
1;GER;FB II;JJ/SG;1;2;3;2;0;1;16;1;1;Merkel sei Dank!;un;(('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0));False;0;False;3
2;GER;FB II;JJ/SG;1;2;3;2;0;1;125;2;2;Es war ya auch der Islam der 6 Millionen Jüdische Menschen systematisch etmordet hat. Das hat ya nichts mit euch zu tun sorry;de;(('GERMAN', 'de', 99, 867.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0));True;0;False;23
3;GER;TW;JJ/JW;1;2;3;1;0;1;41;0;0;Mijn maag keert ! #zaventem #grenzendicht;nl;(('DUTCH', 'nl', 97, 404.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0));True;2;False;5"

temp_csv <- tempfile(fileext = ".csv")
writeLines(content, temp_csv)
data.table::fread(temp_csv)
#>    number country platform tandem csystem_dummy msystem_rank csystem_rank
#> 1:      1     GER    FB II  JJ/SG             1            2            3
#> 2:      2     GER    FB II  JJ/SG             1            2            3
#> 3:      3     GER       TW  JJ/JW             1            2            3
#>    arena_rank issdriven prefdriven charcount ic_ordinal ic_categorized
#> 1:          2         0          1        16          1              1
#> 2:          2         0          1       125          2              2
#> 3:          1         0          1        41          0              0
#>                                                                                                                             post
#> 1:                                                                                                              Merkel sei Dank!
#> 2: Es war ya auch der Islam der 6 Millionen Jüdische Menschen systematisch etmordet hat. Das hat ya nichts mit euch zu tun sorry
#> 3:                                                                                     Mijn maag keert ! #zaventem #grenzendicht
#>    main_language
#> 1:            un
#> 2:            de
#> 3:            nl
#>                                                                              languages
#> 1:   (('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0))
#> 2: (('GERMAN', 'de', 99, 867.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0))
#> 3:  (('DUTCH', 'nl', 97, 404.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0))
#>    is_reliable hashtag_count is_redacted WC
#> 1:       FALSE             0       FALSE  3
#> 2:        TRUE             0       FALSE 23
#> 3:        TRUE             2       FALSE  5
data.table::fread(temp_csv, sep = ";")
#>    number country platform tandem csystem_dummy msystem_rank csystem_rank
#> 1:      1     GER    FB II  JJ/SG             1            2            3
#> 2:      2     GER    FB II  JJ/SG             1            2            3
#> 3:      3     GER       TW  JJ/JW             1            2            3
#>    arena_rank issdriven prefdriven charcount ic_ordinal ic_categorized
#> 1:          2         0          1        16          1              1
#> 2:          2         0          1       125          2              2
#> 3:          1         0          1        41          0              0
#>                                                                                                                             post
#> 1:                                                                                                              Merkel sei Dank!
#> 2: Es war ya auch der Islam der 6 Millionen Jüdische Menschen systematisch etmordet hat. Das hat ya nichts mit euch zu tun sorry
#> 3:                                                                                     Mijn maag keert ! #zaventem #grenzendicht
#>    main_language
#> 1:            un
#> 2:            de
#> 3:            nl
#>                                                                              languages
#> 1:   (('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0))
#> 2: (('GERMAN', 'de', 99, 867.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0))
#> 3:  (('DUTCH', 'nl', 97, 404.0), ('Unknown', 'un', 0, 0.0), ('Unknown', 'un', 0, 0.0))
#>    is_reliable hashtag_count is_redacted WC
#> 1:       FALSE             0       FALSE  3
#> 2:        TRUE             0       FALSE 23
#> 3:        TRUE             2       FALSE  5
readtext::readtext(temp_csv, text_field = "post")
#> Warning in data.table::fread(input = path, data.table = FALSE, stringsAsFactors
#> = FALSE, : Detected 1 column names but the data has 12 columns (i.e. invalid
#> file). Added 11 extra default column names at the end.
#> Error in sort_fields(result, path, text_field): There is no field called post in file /tmp/RtmpoR2R2H/file4986b618527ca.csv.
readtext::readtext(temp_csv, text_field = "post", sep = ";")
#> Error in data.table::fread(input = path, data.table = FALSE, stringsAsFactors = FALSE, : formal argument "sep" matched by multiple actual arguments

sessionInfo()
#> R version 4.2.2 Patched (2022-11-10 r83330)
#> Platform: x86_64-pc-linux-gnu (64-bit)
#> Running under: Ubuntu 20.04.5 LTS
#> 
#> Matrix products: default
#> BLAS:   /usr/lib/x86_64-linux-gnu/blas/libblas.so.3.9.0
#> LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.9.0
#> 
#> locale:
#>  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
#>  [3] LC_TIME=de_DE.UTF-8        LC_COLLATE=en_US.UTF-8    
#>  [5] LC_MONETARY=de_DE.UTF-8    LC_MESSAGES=en_US.UTF-8   
#>  [7] LC_PAPER=de_DE.UTF-8       LC_NAME=C                 
#>  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
#> [11] LC_MEASUREMENT=de_DE.UTF-8 LC_IDENTIFICATION=C       
#> 
#> attached base packages:
#> [1] stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> loaded via a namespace (and not attached):
#>  [1] knitr_1.39        magrittr_2.0.3    R.cache_0.16.0    R6_2.5.1         
#>  [5] rlang_1.0.6       fastmap_1.1.0     fansi_1.0.3       httr_1.4.4       
#>  [9] stringr_1.4.1     styler_1.7.0      highr_0.9         tools_4.2.2      
#> [13] data.table_1.14.2 xfun_0.31         R.oo_1.25.0       utf8_1.2.2       
#> [17] cli_3.4.1         withr_2.5.0       htmltools_0.5.3   yaml_2.3.6       
#> [21] digest_0.6.30     tibble_3.1.8      lifecycle_1.0.3   purrr_0.3.5      
#> [25] vctrs_0.5.1       R.utils_2.12.0    fs_1.5.2          glue_1.6.2       
#> [29] evaluate_0.15     rmarkdown_2.14    reprex_2.0.1      stringi_1.7.8    
#> [33] compiler_4.2.2    pillar_1.8.1      R.methodsS3_1.8.2 readtext_0.81    
#> [37] pkgconfig_2.0.3

Created on 2022-11-30 by the reprex package (v2.0.1)

chainsawriot commented 1 year ago

And actually this works.

readtext:::get_csv(temp_csv, text_field = "post", sep = ";", 
encoding = getOption("encoding"), docid_field = NULL)

https://github.com/quanteda/readtext/blob/d480d9c55f2d4192519cb4a67309075e66023638/R/readtext.R#L267

But this line hard-codes sep to comma for csv.