在R中使用regexpr，具有多个字符串肯定和否定

Question

I am grepping at a column of notes. Looking for the presence of some strings and the absence of others. The expression looks like

toMatch <- c("words", "i", "want", "to")
notToMatch <- c("not", "in", "my", "res")
insert <- paste(paste(toMatch, collapse="|"), "!", paste(notToMatch, collapse="!"), sep="")
regexpr(insert, df$notes, ignore.case=T)

It seems to me that regexpr will count

printNotes = +1 presence and -1 absence

and if that expression evaluates to printNotes > 0, it returns a value other than -1 (which in regexpr indicates not found).

Any suggested syntax for regexpr to return -1 if any of the notToMatch "!" arguments return TRUE?

Thanks much!

Answer 1

You can use grepl() to get a logical vector of where the strings have matched and then sum() that vector to see the number which are matches. You can do the same thing (roughly) with grep() and counting the length of the resultant vector but grepl() behaves a bit more consistently.

If you want to get the inverse of any match you can do !grepl("match", x) and it will show the logical inverse.

If you *specifically want it to return TRUE or ! you can do something like ifelse(grepl("m", letters), TRUE, "!") which searches the letters constant (all 26 lower-case english letters) for "m" and returns TRUE on a match and "!" on a failure to match.