read.csv warning 'EOF within quoted string' prevents complete reading of file

I have a CSV file (24.1 MB) that I cannot fully read into my R session. When I open the file in a spreadsheet program I can see 112,544 rows. When I read it into R with read.csv I only get 56,952 rows and this warning:

cit <- read.csv("citations.CSV", row.names = NULL, 
                comment.char = "", header = TRUE, 
                stringsAsFactors = FALSE,  
                colClasses= "character", encoding= "utf-8")

Warning message:
In scan(file, what, nmax, sep, dec, quote, skip, nlines, na.strings,  :
  EOF within quoted string

I can read the whole file into R with readLines:

rl <- readLines(file("citations.CSV", encoding = "utf-8"))
length(rl)
[1] 112545

But I can't get this back into R as a table (via read.csv):

write.table(rl, "rl.txt", quote = FALSE, row.names = FALSE)
rl_in <- read.csv("rl.txt", skip = 1, row.names = NULL)

Warning message:
In scan(file, what, nmax, sep, dec, quote, skip, nlines, na.strings,  :
  EOF within quoted string

How can I solve or workaround this EOF message (which seems to be more of an error than a warning) to get the entire file into my R session?

I have similar problems with other methods of reading CSV files:

require(sqldf)
cit_sql <- read.csv.sql("citations.CSV", sql = "select * from file")
require(data.table)
cit_dt <- fread("citations.CSV")
require(ff)
cit_ff <- read.csv.ffdf(file="citations.CSV")

Here's my sessionInfo()

R version 3.0.1 (2013-05-16)
Platform: x86_64-w64-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=English_United States.1252  LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

attached base packages:
[1] tools     tcltk     stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] ff_2.2-11             bit_1.1-10            data.table_1.8.8      sqldf_0.4-6.4        
 [5] RSQLite.extfuns_0.0.1 RSQLite_0.11.4        chron_2.3-43          gsubfn_0.6-5         
 [9] proto_0.3-10          DBI_0.2-7   

You need to disable quoting.

cit <- read.csv("citations.CSV", quote = "", 
                 row.names = NULL, 
                 stringsAsFactors = FALSE)

str(cit)
## 'data.frame':    112543 obs. of  13 variables:
##  $ row.names    : chr  "10.2307/675394" "10.2307/30007362" "10.2307/4254931" "10.2307/20537934" ...
##  $ id           : chr  "10.2307/675394\t" "10.2307/30007362\t" "10.2307/4254931\t" "10.2307/20537934\t" ...
##  $ doi          : chr  "Archaeological Inference and Inductive Confirmation\t" "Sound and Sense in Cath Almaine\t" "Oak Galls Preserved by the Eruption of Mount Vesuvius in A.D. 79_ and Their Probable Use\t" "The Arts Four Thousand Years Ago\t" ...
##  $ title        : chr  "Bruce D. Smith\t" "Tomás Ó Cathasaigh\t" "Hiram G. Larew\t" "\t" ...
##  $ author       : chr  "American Anthropologist\t" "Ériu\t" "Economic Botany\t" "The Illustrated Magazine of Art\t" ...
##  $ journaltitle : chr  "79\t" "54\t" "41\t" "1\t" ...
##  $ volume       : chr  "3\t" "\t" "1\t" "3\t" ...
##  $ issue        : chr  "1977-09-01T00:00:00Z\t" "2004-01-01T00:00:00Z\t" "1987-01-01T00:00:00Z\t" "1853-01-01T00:00:00Z\t" ...
##  $ pubdate      : chr  "pp. 598-617\t" "pp. 41-47\t" "pp. 33-40\t" "pp. 171-172\t" ...
##  $ pagerange    : chr  "American Anthropological Association\tWiley\t" "Royal Irish Academy\t" "New York Botanical Garden Press\tSpringer\t" "\t" ...
##  $ publisher    : chr  "fla\t" "fla\t" "fla\t" "fla\t" ...
##  $ type         : logi  NA NA NA NA NA NA ...
##  $ reviewed.work: logi  NA NA NA NA NA NA ...

I think is because of this kind of lines (check "Thorn" and "Minus")

 readLines("citations.CSV")[82]
[1] "10.2307/3642839,10.2307/3642839\t,\"Thorn\" and \"Minus\" in Hieroglyphic Luvian Orthography\t,H. Craig Melchert\t,Anatolian Studies\t,38\t,\t,1988-01-01T00:00:00Z\t,pp. 29-42\t,British Institute at Ankara\t,fla\t,\t,"

I'm a new-ish R user and thought I'd post this in case it helps anyone else. I was trying to read in data from a text file (separated with commas) that included a few Spanish characters and it took me forever to figure it out. I knew I needed to use UTF-8 encoding, set the header arg to TRUE, and that I need to set the sep arguemnt to ",", but then I still got hang ups. After reading this post I tried setting the fill arg to TRUE, but then got the same "EOF within quoted string" which I was able to fix in the same manner as above. My successful read.table looks like this:

target <- read.table("target2.txt", fill=TRUE, header=TRUE, quote="", sep=",", encoding="UTF-8")

The result has Spanish language characters and same dims I had originally, so I'm calling it a success! Thanks all!


In the R help section, as pointed out above, just disabling quoting altogether, by simply adding:

    quote = "" 

to the read.csv() worked for me.

The error, "EOF within quoted string", occurred with:

    > iproscan.53A.neg     = read.csv("interproscan.53A.neg.n.csv",
    +                        colClasses=c(pb.id      = "character",
    +                                     genLoc     = "character",
    +                                     icode      = "character",
    +                                     length     = "character",
    +                                     proteinDB  = "character",
    +                                     protein.id = "character",
    +                                     prot.desc  = "character",
    +                                     start      = "character",
    +                                     end        = "character",
    +                                     evalue     = "character",
    +                                     tchar      = "character",
    +                                     date       = "character",
    +                                     ipro.id    = "character",
    +                                     prot.name  = "character",
    +                                     go.cat     = "character",
    +                                     reactome.id= "character"),
    +                                     as.is=T,header=F)
    Warning message:
    In scan(file, what, nmax, sep, dec, quote, skip, nlines, na.strings,  :
      EOF within quoted string
    > dim(iproscan.53A.neg)
    [1] 69383    16

And the file read in was missing 6,619 lines. But by disabling quoting

    > iproscan.53A.neg     = read.csv("interproscan.53A.neg.n.csv",
    +                        colClasses=c(pb.id      = "character",
    +                                     genLoc     = "character",
    +                                     icode      = "character",
    +                                     length     = "character",
    +                                     proteinDB  = "character",
    +                                     protein.id = "character",
    +                                     prot.desc  = "character",
    +                                     start      = "character",
    +                                     end        = "character",
    +                                     evalue     = "character",
    +                                     tchar      = "character",
    +                                     date       = "character",
    +                                     ipro.id    = "character",
    +                                     prot.name  = "character",
    +                                     go.cat     = "character",
    +                                     reactome.id= "character"),
    +                                     as.is=T,header=F,**quote=""**)    
    > 
    > dim(iproscan.53A.neg)
    [1] 76002    16

Worked without error and all lines were successfully read in.


I also ran into this problem, and was able to work around a similar EOF error using:

read.table("....csv", sep=",", ...)

Notice that the separator parameter is defined within the more general read.table().