Skip to contents

Takes an input string and parses it into a sequence of regular expressions.

Usage

REparser(
  res,
  parse.strict = TRUE,
  parse.exhaust = TRUE,
  parse.lead = FALSE,
  parse.rest = FALSE,
  toEnv = FALSE
)

REparse(
  str,
  res,
  parse.strict = TRUE,
  parse.exhaust = TRUE,
  parse.lead = FALSE,
  parse.rest = FALSE,
  reverse = FALSE,
  sep = NULL,
  toEnv = FALSE
)

Details

If exhaustive is TRUE, the string must be exhaustively broken up by the matching regular expressions. Superfluous (non-match) characters at the begginning, end, or in bettween matches, will result in all NA being returned.

-------------------------------------------> NEEDS DOCUMENTATION <-------------------------------------------