| grep {base} | R Documentation |
grep, grepl, regexpr and gregexpr search
for matches to argument pattern within each element of a
character vector: they differ in the format of and amount of detail in
the results.
sub and gsub perform replacement of the first and all
matches respectively.
grep(pattern, x, ignore.case = FALSE, perl = FALSE, value = FALSE,
fixed = FALSE, useBytes = FALSE, invert = FALSE)
grepl(pattern, x, ignore.case = FALSE, perl = FALSE,
fixed = FALSE, useBytes = FALSE)
sub(pattern, replacement, x, ignore.case = FALSE, perl = FALSE,
fixed = FALSE, useBytes = FALSE)
gsub(pattern, replacement, x, ignore.case = FALSE, perl = FALSE,
fixed = FALSE, useBytes = FALSE)
regexpr(pattern, text, ignore.case = FALSE, perl = FALSE,
fixed = FALSE, useBytes = FALSE)
gregexpr(pattern, text, ignore.case = FALSE, perl = FALSE,
fixed = FALSE, useBytes = FALSE)
regexec(pattern, text, ignore.case = FALSE,
fixed = FALSE, useBytes = FALSE)
pattern |
character string containing a regular expression
(or character string for |
x, text |
a character vector where matches are sought, or an
object which can be coerced by |
ignore.case |
if |
perl |
logical. Should perl-compatible regexps be used? |
value |
if |
fixed |
logical. If |
useBytes |
logical. If |
invert |
logical. If |
replacement |
a replacement for matched pattern in |
Arguments which should be character strings or character vectors are coerced to character if possible.
Each of these functions (apart from regexec, which currently
does not support Perl-style regular expressions) operates in one of
three modes:
fixed = TRUE: use exact matching.
perl = TRUE: use Perl-style regular expressions.
fixed = FALSE, perl = FALSE: use POSIX 1003.2
extended regular expressions.
See the help pages on regular expression for details of the different types of regular expressions.
The two *sub functions differ only in that sub replaces
only the first occurrence of a pattern whereas gsub
replaces all occurrences. If replacement contains
backreferences which are not defined in pattern the result is
undefined (but most often the backreference is taken to be "").
For regexpr, gregexpr and regexec it is an error
for pattern to be NA, otherwise NA is permitted
and gives an NA match.
The main effect of useBytes is to avoid errors/warnings about
invalid inputs and spurious matches in multibyte locales, but for
regexpr it changes the interpretation of the output.
It inhibits the conversion of inputs with marked encodings, and is
forced if any input is found which is marked as "bytes".
Caseless matching does not make much sense for bytes in a multibyte
locale, and you should expect it only to work for ASCII characters if
useBytes = TRUE.
As from R 2.14.0, regexpr and gregexpr with perl
= TRUE allow Python-style named captures.
grep(value = FALSE) returns an integer vector of the indices
of the elements of x that yielded a match (or not, for
invert = TRUE.
grep(value = TRUE) returns a character vector containing the
selected elements of x (after coercion, preserving names but no
other attributes).
grepl returns a logical vector (match or not for each element of
x).
For sub and gsub return a character vector of the same
length and with the same attributes as x (after possible
coercion to character). Elements of character vectors x which
are not substituted will be returned unchanged (including any declared
encoding). If useBytes = FALSE a non-ASCII substituted result
will often be in UTF-8 with a marked encoding (e.g. if there is a
UTF-8 input, and in a multibyte locale unless fixed = TRUE).
Such strings can be re-encoded by enc2native.
regexpr returns an integer vector of the same length as
text giving the starting position of the first match or
-1 if there is none, with attribute "match.length", an
integer vector giving the length of the matched text (or -1 for
no match). The match positions and lengths are in characters unless
useBytes = TRUE is used, when they are in bytes. If named
capture is used there are further attributes "capture.start",
"capture.length" and "capture.names".
gregexpr returns a list of the same length as text each
element of which is of the same form as the return value for
regexpr, except that the starting positions of every (disjoint)
match are given.
regexec returns a list of the same length as text each
element of which is either -1 if there is no match, or a
sequence of integers with the starting positions of the match and all
substrings corresponding to parenthesized subexpressions of
pattern, with attribute "match.length" an integer vector
giving the lengths of the matches (or -1 for no match).
POSIX 1003.2 mode of gsub and gregexpr does not
work correctly with repeated word-boundaries (e.g. pattern =
"\b"). Use perl = TRUE for such matches (but that may not
work as expected with non-ASCII inputs, as the meaning of
‘word’ is system-dependent).
If you are doing a lot of regular expression matching, including on
very long strings, you will want to consider the options used.
Generally PCRE will be faster than the default regular expression
engine, and fixed = TRUE faster still (especially when each
pattern is matched only a few times).
If you are working in a single-byte locale and have marked UTF-8 strings that are representable in that locale, convert them first as just one UTF-8 string will force all the matching to be done in Unicode, which attracts a penalty of around 3x for the default POSIX 1003.2 mode.
If you can make use of useBytes = TRUE, the strings will not be
checked before matching, and the actual matching will be faster.
Often byte-based matching suffices in a UTF-8 locale since byte
patterns of one character never match part of another.
Prior to R 2.11.0 there was an argument extended which could
be used to select ‘basic’ regular expressions: this was often
used when fixed = TRUE would be preferable. In the actual
implementation (as distinct from the POSIX standard) the only
difference was that ?, +, {, |, (,
and ) were not interpreted as metacharacters.
The C code for POSIX-style regular expression matching has changed
over the years. As from R 2.10.0 the TRE library of Ville Laurikari
(http://laurikari.net/tre/) is used. From 2005 to R 2.9.2,
code based on glibc was used (and before that, code from GNU
grep). The POSIX standard does give some room for
interpretation, especially in the handling of invalid regular
expressions and the collation of character ranges, so the results will
have changed slightly.
For Perl-style matching PCRE (http://www.pcre.org) is used.
Becker, R. A., Chambers, J. M. and Wilks, A. R. (1988)
The New S Language.
Wadsworth & Brooks/Cole (grep)
regular expression (aka regexp) for the details
of the pattern specification.
regmatches for extracting matched substrings based on
the results of regexpr, gregexpr and regexec.
glob2rx to turn wildcard matches into regular expressions.
agrep for approximate matching.
charmatch, pmatch for partial matching,
match for matching to whole strings.
tolower, toupper and chartr
for character translations.
apropos uses regexps and has more examples.
grepRaw for matching raw vectors.
grep("[a-z]", letters)
txt <- c("arm","foot","lefroo", "bafoobar")
if(length(i <- grep("foo",txt)))
cat("'foo' appears at least once in\n\t",txt,"\n")
i # 2 and 4
txt[i]
## Double all 'a' or 'b's; "\" must be escaped, i.e., 'doubled'
gsub("([ab])", "\\1_\\1_", "abc and ABC")
txt <- c("The", "licenses", "for", "most", "software", "are",
"designed", "to", "take", "away", "your", "freedom",
"to", "share", "and", "change", "it.",
"", "By", "contrast,", "the", "GNU", "General", "Public", "License",
"is", "intended", "to", "guarantee", "your", "freedom", "to",
"share", "and", "change", "free", "software", "--",
"to", "make", "sure", "the", "software", "is",
"free", "for", "all", "its", "users")
( i <- grep("[gu]", txt) ) # indices
stopifnot( txt[i] == grep("[gu]", txt, value = TRUE) )
## Note that in locales such as en_US this includes B as the
## collation order is aAbBcCdEe ...
(ot <- sub("[b-e]",".", txt))
txt[ot != gsub("[b-e]",".", txt)]#- gsub does "global" substitution
txt[gsub("g","#", txt) !=
gsub("g","#", txt, ignore.case = TRUE)] # the "G" words
regexpr("en", txt)
gregexpr("e", txt)
## Using grepl() for filtering
## Find functions with argument names matching "warn":
findArgs <- function(env, pattern) {
nms <- ls(envir = as.environment(env))
nms <- nms[is.na(match(nms, c("F","T")))] # <-- work around "checking hack"
aa <- sapply(nms, function(.) { o <- get(.)
if(is.function(o)) names(formals(o)) })
iw <- sapply(aa, function(a) any(grepl(pattern, a, ignore.case=TRUE)))
aa[iw]
}
findArgs("package:base", "warn")
## trim trailing white space
str <- 'Now is the time '
sub(' +$', '', str) ## spaces only
sub('[[:space:]]+$', '', str) ## white space, POSIX-style
sub('\\s+$', '', str, perl = TRUE) ## Perl-style white space
## capitalizing
txt <- "a test of capitalizing"
gsub("(\\w)(\\w*)", "\\U\\1\\L\\2", txt, perl=TRUE)
gsub("\\b(\\w)", "\\U\\1", txt, perl=TRUE)
txt2 <- "useRs may fly into JFK or laGuardia"
gsub("(\\w)(\\w*)(\\w)", "\\U\\1\\E\\2\\U\\3", txt2, perl=TRUE)
sub("(\\w)(\\w*)(\\w)", "\\U\\1\\E\\2\\U\\3", txt2, perl=TRUE)
## named capture
notables <- c(" Ben Franklin and Jefferson Davis",
"\tMillard Fillmore")
# name groups 'first' and 'last'
name.rex <- "(?<first>[[:upper:]][[:lower:]]+) (?<last>[[:upper:]][[:lower:]]+)"
(parsed <- regexpr(name.rex, notables, perl = TRUE))
gregexpr(name.rex, notables, perl = TRUE)[[2]]
parse.one <- function(res, result) {
m <- do.call(rbind, lapply(seq_along(res), function(i) {
if(result[i] == -1) return("")
st <- attr(result, "capture.start")[i, ]
substring(res[i], st, st + attr(result, "capture.length")[i, ] - 1)
}))
colnames(m) <- attr(result, "capture.names")
m
}
parse.one(notables, parsed)
## Decompose a URL into its components.
## Example by LT (http://www.cs.uiowa.edu/~luke/R/regexp.html).
x <- "http://stat.umn.edu:80/xyz"
m <- regexec("^(([^:]+)://)?([^:/]+)(:([0-9]+))?(/.*)", x)
m
regmatches(x, m)
## Element 3 is the protocol, 4 is the host, 6 is the port, and 7
## is the path. We can use this to make a function for extracting the
## parts of a URL:
URL_parts <- function(x) {
m <- regexec("^(([^:]+)://)?([^:/]+)(:([0-9]+))?(/.*)", x)
parts <- do.call(rbind,
lapply(regmatches(x, m), `[`, c(3L, 4L, 6L, 7L)))
colnames(parts) <- c("protocol","host","port","path")
parts
}
URL_parts(x)