Consider value piping for Pipe

[renkun-ken]

Consider supporting the following piping mechanism for Pipe(x)$y

1. Look for any y in Pipe object environment ($value, $fun)
2. If x in Pipe is S4, get slot y from x.
3. If x in Pipe is list or vector, get element named y from x.
4. Otherwise, look for function y in parent.frame()

This allows the following usage:

> Pipe(mtcars)$fun(lm(mpg~.,data=.))$summary()$fstatistic
$value : numeric 
------
   value    numdf    dendf 
13.93246 10.00000 21.00000 

[renkun-ken]

For example,

library(fUnitRoots)
Pipe(rnorm(100))$
  cumsum()$
  adfTest(lags = 5)$
  test$
  statistic

$value : numeric 
------
Dickey-Fuller 
    -2.033778 

[renkun-ken]

This implementation often leads to potential confusion.

Consider a data frame containing a column named select, if dplyr is used to select the columns, the column will be singled out and select() would be ignored.

If rule 4 has higher priority, $ is still overloaded by too much potential functionality.

[renkun-ken]

A new closure .() is added to Pipe object. For .(y),

• When y is a name then it gets getElement(x,"y") which also works for S4 object.
• When y is a call then it performs lambda piping to . or user-defined symbol.

This allows the following usage:

Pipe(mtcars)$
  .(lm(mpg ~ cyl + wt, data = .))$
  .(coefficients)

$value : numeric 
------
(Intercept)         cyl          wt 
  39.686261   -1.507795   -3.190972 

For S4 object,

library(fUnitRoots)
Pipe(rnorm(100))$
  cumsum()$
  adfTest(lags = 5)$
  .(test)$
  .(statistic)

$value : numeric 
------
Dickey-Fuller 
    -1.283631 

Then .() incorporates the functionality of fun() which may be deprecated in future version for potential name conflicts.

[renkun-ken]

Now, Pipe(x)$.(...) closure in Pipe object works in exactly the same way with x %>>% (...).