grepRaw
Pattern Matching for Raw Vectors
grepRaw searches for substring pattern matches within a
raw vector x.
- Keywords
- utilities
Usage
grepRaw(pattern, x, offset = 1L, ignore.case = FALSE, value = FALSE, fixed = FALSE, all = FALSE, invert = FALSE)Arguments
- pattern
- raw vector containing a regular expression
(or fixed pattern for
fixed = TRUE) to be matched in the given raw vector. Coerced bycharToRawto a character string if possible. - x
- a raw vector where matches are sought, or an object which can
be coerced by
charToRawto a raw vector. Long vectors are not supported. - ignore.case
- if
FALSE, the pattern matching is case sensitive and ifTRUE, case is ignored during matching. - offset
- An integer specifying the offset from
which the search should start. Must be positive. The beginning of
line is defined to be at that offset so
"^"will match there. - value
- logical. Determines the return value: see ‘Value’.
- fixed
- logical. If
TRUE,patternis a pattern to be matched as is. - all
- logical. If
TRUEall matches are returned, otherwise just the first one. - invert
- logical. If
TRUEreturn indices or values for elements that do not match. Ignored (with a warning) unlessvalue = TRUE.
Details
Unlike grep, seeks matching patterns within the raw
vector x . This has implications especially in the all =
TRUE case, e.g., patterns matching empty strings are inherently
infinite and thus may lead to unexpected results.
The argument invert is interpreted as asking to return the
complement of the match, which is only meaningful for value =
TRUE. Argument offset determines the start of the search, not
of the complement. Note that invert = TRUE with all =
TRUE will split x into pieces delimited by the pattern
including leading and trailing empty strings (consequently the use of
regular expressions with "^" or "$" in that case may
lead to less intuitive results).
Some combinations of arguments such as fixed = TRUE with
value = TRUE are supported but are less meaningful.
Value
grepRaw(value = FALSE) returns an integer vector of the offsets
at which matches have occurred. If all = FALSE then it will be
either of length zero (no match) or length one (first matching
position).grepRaw(value = TRUE, all = FALSE) returns a raw vector which
is either empty (no match) or the matched part of x.grepRaw(value = TRUE, all = TRUE) returns a (potentially
empty) list of raw vectors corresponding to the matched parts.
Source
The TRE library of Ville Laurikari (http://laurikari.net/tre/)
is used except for fixed = TRUE.
See Also
regular expression (aka regexp) for the details
of the pattern specification.
grep for matching character vectors.