如何删除特定单词的小节

Question

Aim: I am trying to get rid of subclauses in various sentences in r which contain the word normal. A subclause is defined as being separated by a starting comma and ending in either a full stop or comma. I want to get rid of the subclause.

Input sentences

I walked down the hill, which was normal, but I also walked up another hill which was dull.

I looked at him and although he looked normal, he was not normal.

I am fine, but he is not normal, and she is fine and she is normal, but I think her brother is not normal.

Desired output

I walked down the hill but I also walked up another hill which was dull

I looked at him and although he looked normal.

I am fine, and she is fine and she is normal.

Attempt

gsub(", .*normal.*?(\\.|,|$)\\R*", "", input_string, perl = T, ignore.case = T)

Current output:

I walked down the hill.
I looked at him and although he looked normal.
I am fine.

However if there are many subclauses this doesn't give the intended output mainly because it removed everything from the first comma. How do I make it match from the nearest comma to the 'normal' ?

Answer 1

Your examples and rules are not consistent (see @janos' comment). For example, you remove the last sub-clause in your last example sentence "but I think her brother is not normal" even though it doesn't end with a period.

That aside, the following should get you started:

ss <- c(
    "I walked down the hill, which was normal, but I also walked up another hill which was dull",
    "I looked at him and although he looked normal, he was not normal.",
    "I am fine, but he is not normal, and she is fine and she is normal, but I think her brother is not normal");

lapply(ss, function(x) gsub("\\,[a-zA-Z0-9_ ]+[\\,\\.]{1}", "", x));
#[[1]]
#[1] "I walked down the hill but I also walked up another hill which was dull"

#[[2]]
#[1] "I looked at him and although he looked normal"

#[[3]]
#[1] "I am fine and she is fine and she is normal, but I think her brother is not normal"