Turn singular row with interval into multiple rows which equal the interval

Question

Suppose you have something like this:

Col1 Col2
a    odd from 1 to 9
b    even from 2 to 14
c    even from 30 to 50
...

I would like to expand the rows by separating the interval into individual row, so:

Col1 Col2
a    1
a    3
a    5
...
b    2
b    4
b    6
...
c    30
c    32
c    34
...

Note that when it says "even from", the lower and upper bounds are also even, and the same goes for odd numbers.

Answer 1

separate Col2 into individual columns and then for each row create the sequence:

library(dplyr)
library(tidyr)
DF %>% 
   separate(Col2, into = c("parity", "X1", "from", "X2", "to")) %>% 
   group_by(Col1) %>% 
   do(data.frame(Col2 = seq(.$from, .$to, 2))) %>%
   ungroup

Note 1

The input DF in reproducible form is assumed to be:

DF <- structure(list(Col1 = c("a", "b", "c"), Col2 = c("odd from 1 to 9", 
"even from 2 to 14", "even from 30 to 50")), .Names = c("Col1", 
"Col2"), row.names = c(NA, -3L), class = "data.frame")

Note 2

The next version of tidyr supports NA in the into vector to denote fields to ignore so the separate statement above could be written:

   separate(Col2, into = c("parity", NA, "from", NA, "to")) %>% 

Answer 2

Here is an option using base R. We extract the numeric elements in 'Col2' using gregexpr/regmatches into a list, then the get the sequence of elements by 2 with seq and stack it to data.frame

res <- stack(setNames(lapply(regmatches(DF$Col2, gregexpr("\\d+", DF$Col2)), function(x)
     seq(as.numeric(x[1]), as.numeric(x[2]), by = 2)), DF$Col1))[2:1]
colnames(res) <- colnames(DF)
head(res)
#  Col1 Col2
#1    a    1
#2    a    3
#3    a    5
#4    a    7
#5    a    9
#6    b    2