Changelog

`hdd` version 0.1.2

fix major bug in [.hdd preventing this function to be used in nested environments

changing the default for the argument replace: it becomes TRUE in all functions using it (previously it was FALSE)
argument newfile becomes destDir, retro compatibility is ensured
in argument destDir, the parent and grand-parent directories are now automatically created if they are missing
add progress bar to txt2hdd and hdd_slice

txt2hdd: argument path now accepts vectors and even patterns. All files matching the pattern will be imported to the hdd format.
txt2hdd: gains the arguments encoding and locale.
txt2hdd: messages are more informative, in particular import time is reported.

Fixed a small bug in the function guess_col_types when the data contained dates.

This package is an effort to provide a simple way to deal with out of memory data sets within R. The tasks does is to import and perfom simple manipulations: like subsetting, creating new variables, etc.
As in other data base management systems (DBMS), the data is split into several chunks that fit in memory, and operations are performed “chunk-wise”. These operations are hidden to the user.
Why creating this package? First, it avoids having to switch to alternative software like DBMS to deal with out of memory data sets (you can also use DBMS in R but it is cumbersome). Second, it allows the user to use the full power of R to create complex new variables. Finally, and maybe most importantly, the syntax is very simple and requires little effort from the user.