hi there,
I've got a problem importing files where some lines end with double tab:

If the file's first line has extra columns and then other lines have missing columns, then fill=TRUE solves the problem.
However i have one file where the extra columns happen in line 82:

> test[80:84] [1] "1053982\t2014-11-30\t-1\ttimetak\t14" "1053982\t2014-11-30\t-1\tyearcompleted\t4" [3] "1053982\t24-11-2014\t-1\tAMOUNT OTHER WEBSITES\t6\t\t" "1053982\t24-11-2014\t-1\tBBC1_422\t1\t\t" [5] "1053982\t24-11-2014\t-1\tBBC1_501\t13\t\t"

In this case fill=TRUE isn't enough to get around the problem, and i get an error message saying there's text after processing all cols:

> test<-fread(filenames[9], header = FALSE, fill = TRUE, data.table=FALSE, quote='') # problem as double tab only starts in row 82
Error in fread(filenames[9], header = FALSE, fill = TRUE, data.table = FALSE,  : 
  Expecting 5 cols, but line 82 contains text after processing all cols. Try again with fill=TRUE. Another reason could be that fread's logic in distinguishing one or more fields having embedded sep='    ' and/or (unescaped) '\n' characters within unbalanced unescaped quotes has failed. If quote='' doesn't help, please file an issue to figure out if the logic could be improved.

Is there a way i could force fread to either only read the first 5 columns per line and dropping anything after that?
Failing that, can i specify the number of columns in advance to account for the lines with extra tabs, as this way fill=TRUE would work?