Columns are read even if not mentioned in col_types in read_tsv

[anders-sjogren]

From ?read_tsv:

col_types
[...]
If a list, it must contain one "collector" for each column. If you only want to read a subset of the columns, you can use a named list (where the names give the column names). If a column is not mentioned by name, it will not be included in the output. [...]

Bug reproduction steps

read_tsv("1\t2\n3\t4\n",col_names=c("X","Y"),col_types=list("X"=col_integer()))

Expected result

  X
1 1
2 3

Observed result

  X Y
1 1 2
2 3 4

Sidenote: read_tsv("1\t2\n3\t4\n",col_names=c("X","Y"),col_types=c("i_")) yields the correct result.

[anders-sjogren]

A workaround is to define

makeColTypesString<-function(colTypes,allColNames) {
  getColChar<-function(collector){
    if(is(collector,"collector_double")) return("d");
    if(is(collector,"collector_integer")) return("i");
    if(is(collector,"collector_character")) return("c");
    if(is(collector,"collector_logical")) return("l");
  }
  chars = sapply(allColNames, function(x){
    if(x %in% names(colTypes)) getColChar(colTypes[[x]]) else "_";
  })
  paste(chars,collapse="")
}

and to use it as

cnames=c("X","Y")
ctypes=list("X"=col_integer())
read_tsv("1\t2\n3\t4\n",col_names=cnames,col_types=makeColTypesString(ctypes,cnames))

[hadley]

Oops, yes, that's a documentation error. See also #72

[anders-sjogren]

Ok. IMHO: if it'd work as documented, I think it'd be neat.

[hadley]

That's what I had originally planned, but it's a huge pain if you just want to adjust the type of one column. That seems to be more common in practice - but I think it should be possible to have a solution where you just replace list() with something like only() and then you only get the specified cols

[anders-sjogren]

Is there some ignore collector, that one could use to express that one would want to ignore that column (cf. "Ignore Token Processing" on http://www.codeproject.com/Articles/23198/C-String-Toolkit-StrTk-Tokenizer)

[anders-sjogren]

but I think it should be possible to have a solution where you just replace list() with something like only() and then you only get those cols.

That seems like a good idea.