Extract characters of single word following : - r

Question

I would like to extract the name of the drug, where "Drug:", "Other:",etc precedes name of drug.
Take the first word after every ":", including characters like "-".
If there are 2 instances of ":", then "and" should join the 2 words as one string. The ourpur should be in a one column dataframe with column name Drug.
Here is my reproducible example:
my.df <- data.frame(col1 = as.character(c("Product: TLD-1433 infusion Therapy", "Biological: CG0070|Other: n-dodecyl-B-D-maltoside", "Drug: Atezolizumab",
"Drug: N-803 and BCG|Drug: N-803", "Drug: Everolimus and Intravesical Gemcitabine", "Drug: Association atezolizumab + BDB001 + RT|Drug: Association atezolizumab + BDB001+ RT
")))
The output should look something like this:
output.df <- data.frame(Drugs = c("TLD-1433", "CG0070 and n-dodecyl-B-D-matose", "Atezolizumab", "N-803 and N-803", "Everolimus and Intravesical", "Association and Association"))
This is what I've tried, which didn't work.
Attempt 1:
str_extract(my.df$col1, '(?<=:\\s)(\\w+)')
Attempt 2:
str_extract(my.df$col1, '(?<=:\\s)(\\w+)(-)(\\w+)')

Answer 1

I am not so familiar with R, but a pattern that would give you the matches from the example data could be:
(?<=:\s)\w+(?:-\w+)*(?: and \w+(?:-\w+)*)*
Then you could concatenate the matches with and in between.
The pattern matches:
(?<=:\s) Positive lookbehind, assert : and a whitespace char to the left
\w+(?:-\w+)* Match 1+ word chars, followed by optionally repeating - and 1+ word chars
(?: Non capture group
and \w+(?:-\w+)* Match and followed by 1+ word chars followed by optionally repeating - and 1+ word chars
)* Close non capture group and optionally repeat
Regex demo
To get all the matches, you can use str_match_all
str_extract_all(my.df$col1, '(?<=:\\s)\\w+(?:-\\w+)*(?: and \\w+(?:-\\w+)*)*')
For example
library(stringr)
my.df <- data.frame(col1 = as.character(c("Product: TLD-1433 infusion Therapy", "Biological: CG0070|Other: n-dodecyl-B-D-maltoside", "Drug: Atezolizumab",
"Drug: N-803 and BCG|Drug: N-803", "Drug: Everolimus and Intravesical Gemcitabine", "Drug: Association atezolizumab + BDB001 + RT|Drug: Association atezolizumab + BDB001+ RT
")))
lapply(
str_extract_all(my.df$col1, '(?<=:\\s)\\w+(?:-\\w+)*(?: and \\w+(?:-\\w+)*)*')
, paste, collapse=" and ")
Output
[[1]]
[1] "TLD-1433"
[[2]]
[1] "CG0070 and n-dodecyl-B-D-maltoside"
[[3]]
[1] "Atezolizumab"
[[4]]
[1] "N-803 and BCG and N-803"
[[5]]
[1] "Everolimus and Intravesical"
[[6]]
[1] "Association and Association"

Answer 2

Use
:\s*\b([\w-]+\b(?:\s+and\s+\b[\w-]+)*)\b
See regex proof.
EXPLANATION
--------------------------------------------------------------------------------
: ':'
--------------------------------------------------------------------------------
\s* whitespace (\n, \r, \t, \f, and " ") (0 or
more times (matching the most amount
possible))
--------------------------------------------------------------------------------
\b the boundary between a word char (\w) and
something that is not a word char
--------------------------------------------------------------------------------
( group and capture to \1:
--------------------------------------------------------------------------------
[\w-]+ any character of: word characters (a-z,
A-Z, 0-9, _), '-' (1 or more times
(matching the most amount possible))
--------------------------------------------------------------------------------
\b the boundary between a word char (\w)
and something that is not a word char
--------------------------------------------------------------------------------
(?: group, but do not capture (0 or more
times (matching the most amount
possible)):
--------------------------------------------------------------------------------
\s+ whitespace (\n, \r, \t, \f, and " ")
(1 or more times (matching the most
amount possible))
--------------------------------------------------------------------------------
and 'and'
--------------------------------------------------------------------------------
\s+ whitespace (\n, \r, \t, \f, and " ")
(1 or more times (matching the most
amount possible))
--------------------------------------------------------------------------------
\b the boundary between a word char (\w)
and something that is not a word char
--------------------------------------------------------------------------------
[\w-]+ any character of: word characters (a-
z, A-Z, 0-9, _), '-' (1 or more times
(matching the most amount possible))
--------------------------------------------------------------------------------
)* end of grouping
--------------------------------------------------------------------------------
) end of \1
--------------------------------------------------------------------------------
\b the boundary between a word char (\w) and
something that is not a word char
R code:
my.df <- data.frame(col1 = as.character(c("Product: TLD-1433 infusion Therapy", "Biological: CG0070|Other: n-dodecyl-B-D-maltoside", "Drug: Atezolizumab",
"Drug: N-803 and BCG|Drug: N-803", "Drug: Everolimus and Intravesical Gemcitabine", "Drug: Association atezolizumab + BDB001 + RT|Drug: Association atezolizumab + BDB001+ RT
")))
library(stringr)
matches <- str_match_all(my.df$col1, ":\\s*\\b([\\w-]+\\b(?:\\s+and\\s+\\b[\\w-]+)*)\\b")
Drugs <- sapply(matches, function(z) paste(z[,-1], collapse=" and "))
output.df <- data.frame(Drugs)
output.df
Results:
Drugs
1 TLD-1433
2 CG0070 and n-dodecyl-B-D-maltoside
3 Atezolizumab
4 N-803 and BCG and N-803
5 Everolimus and Intravesical
6 Association and Association