Package 'rex'

Or

Description

The special binary function %or% can be used to specify a set of optional matches.

Usage

x %or% y

or(...)

Arguments

x	A string.
y	A string.
...	shortcuts, R variables, text, or other rex functions.

See Also

Other rex: capture(), character_class(), counts, group(), lookarounds, not(), rex(), shortcuts, wildcards

---

Regular Expression

Description

Specify an explicit regular expression. This expression must already be escaped.

Usage

## S3 method for class 'regex'
as.character(x, ...)

## S3 method for class 'regex'
print(x, ...)

regex(x, ...)

Arguments

x	Object
...	further arguments

Methods (by generic)

• as.character: coerce regex object to a character

• print: Print regex object

See Also

as.regex to coerce to a regex object.

---

Coerce objects to a regex.

Description

Coerce objects to a regex.

Usage

as.regex(x, ...)

## Default S3 method:
as.regex(x, ...)

Arguments

x	Object to coerce to regex.
...	further arguments passed to methods.

Methods (by class)

• default: Simply escape the Object.

---

Create a capture group

Description

Used to save the matched value within the group for use later in the regular expression or to extract the values captured. Both named and unnamed groups can later be referenced using capture_group.

Usage

capture(..., name = NULL)

capture_group(name)

Arguments

...	shortcuts, R variables, text, or other rex functions.
name	of the group. Unnamed capture groups are numbers starting at 1 in the order they appear in the regular expression. If two groups have the same name, the leftmost group is the used in any reference.

See Also

group for grouping without capturing. Perl 5 Capture Groups https://perldoc.perl.org/perlre#Capture-groups

Other rex: %or%(), character_class(), counts, group(), lookarounds, not(), rex(), shortcuts, wildcards

Examples

# Match paired quotation marks
re <- rex(
  # first quotation mark
  capture(quotes),

  # match all non-matching quotation marks
  zero_or_more(except(capture_group(1))),

  # end quotation mark (matches first)
  capture_group(1)
)

#named capture - don't match apples to oranges
re <- rex(
  capture(name = "fruit", or("apple", "orange")),
  "=",
  capture_group("fruit")
)

---

Create character classes

Description

There are multiple ways you can define a character class.

Usage

character_class(x)

one_of(...)

any_of(..., type = c("greedy", "lazy", "possessive"))

some_of(..., type = c("greedy", "lazy", "possessive"))

none_of(...)

except_any_of(..., type = c("greedy", "lazy", "possessive"))

except_some_of(..., type = c("greedy", "lazy", "possessive"))

range(start, end)

`:`(start, end)

exclude_range(start, end)

Arguments

x	text to include in the character class (must be escaped manually)
...	shortcuts, R variables, text, or other rex functions.
type	the type of match to perform. There are three match types greedy: match the longest string. This is the default matching type. lazy: match the shortest string. This matches the shortest string from the same anchor point, not necessarily the shortest global string. possessive: match and don't allow backtracking
start	beginning of character class
end	end of character class

Functions

• character_class: explicitly define a character class

• one_of: matches one of the specified characters.

• any_of: matches zero or more of the specified characters.

• some_of: matches one or more of the specified characters.

• none_of: matches anything but one of the specified characters.

• except_any_of: matches zero or more of anything but the specified characters.

• except_some_of: matches one or more of anything but the specified characters.

• range: matches one of any of the characters in the range.

• :: matches one of any of the characters in the range.

• exclude_range: matches one of any of the characters except those in the range.

See Also

Other rex: %or%(), capture(), counts, group(), lookarounds, not(), rex(), shortcuts, wildcards

Examples

# grey = gray
re <- rex("gr", one_of("a", "e"), "y")
grepl(re, c("grey", "gray")) # TRUE TRUE

# Match non-vowels
re <- rex(none_of("a", "e", "i", "o", "u"))
# They can also be in the same string
re <- rex(none_of("aeiou"))
grepl(re, c("k", "l", "e")) # TRUE TRUE FALSE

# Match range
re <- rex(range("a", "e"))
grepl(re, c("b", "d", "f")) # TRUE TRUE FALSE

# Explicit creation
re <- rex(character_class("abcd\\["))
grepl(re, c("a", "d", "[", "]")) # TRUE TRUE TRUE FALSE

---

Character class escapes

Description

Character class escapes

Usage

character_class_escape(x)

## S3 method for class 'regex'
character_class_escape(x)

## S3 method for class 'character_class'
character_class_escape(x)

## S3 method for class 'character'
character_class_escape(x)

## S3 method for class 'list'
character_class_escape(x)

## Default S3 method:
character_class_escape(x)

Arguments

x	Object to escape.

Methods (by class)

• regex: objects are passed through unchanged.

• character_class: objects are passed through unchanged.

• character: objects properly escaped for character classes.

• list: call character_class_escape on all elements of the list.

• default: coerce to character and character_class_escape.

---

Counts

Description

Functions to restrict a regex to a specific number

Usage

n_times(x, n, type = c("greedy", "lazy", "possessive"))

between(x, low, high, type = c("greedy", "lazy", "possessive"))

at_least(x, n, type = c("greedy", "lazy", "possessive"))

at_most(x, n, type = c("greedy", "lazy", "possessive"))

Arguments

x	A regex pattern.
n	An integer number
type	the type of match to perform. There are three match types greedy: match the longest string. This is the default matching type. lazy: match the shortest string. This matches the shortest string from the same anchor point, not necessarily the shortest global string. possessive: match and don't allow backtracking
low	An integer number for the lower limit.
high	An integer number for the upper limit.

Functions

• n_times: x must occur exactly n times.

• between: x must occur between low and high times.

• at_least: x must occur at least n times.

• at_most: x must occur at most n times.

See Also

Other rex: %or%(), capture(), character_class(), group(), lookarounds, not(), rex(), shortcuts, wildcards

---

Escape characters for a regex

Description

Escape characters for a regex

Usage

escape(x)

## S3 method for class 'regex'
escape(x)

## S3 method for class 'character_class'
escape(x)

## S3 method for class 'character'
escape(x)

## Default S3 method:
escape(x)

## S3 method for class 'list'
escape(x)

Arguments

x	Object to escape.

Methods (by class)

• regex: Objects are simply passed through unchanged.

• character_class: Objects are surrounded by braces.

• character: Objects are properly escaped for regular expressions.

• default: default escape coerces to character and escapes.

• list: simply call escape on all elements of the list.

---

Create a grouped expression

Description

This is similar to capture except that it does not store the value of the group. Best used when you want to combine several parts together and do not reference or extract the grouped value later.

Usage

group(...)

Arguments

...	shortcuts, R variables, text, or other rex functions.

See Also

capture for grouping with capturing. Perl 5 Extended Patterns https://perldoc.perl.org/perlre#Extended-Patterns

Other rex: %or%(), capture(), character_class(), counts, lookarounds, not(), rex(), shortcuts, wildcards

---

Lookarounds

Description

Lookarounds

Usage

x %if_next_is% y

x %if_next_isnt% y

x %if_prev_is% y

x %if_prev_isnt% y

Arguments

x	A regex pattern.
y	A regex pattern.

Details

These functions provide an interface to perl lookarounds.

Special binary functions are used to infer an ordering, since often you might wish to match a word / set of characters conditional on the start and end of that word.

• %if_next_is%: TRUE if x follows y

• %if_next_isnt%: TRUE if x does not follow y

• %if_prev_is%: TRUE if y comes before x

• %if_prev_isnt%: TRUE if y does not come before x

See Also

Perl 5 Documentation https://perldoc.perl.org/perlre#Extended-Patterns

Other rex: %or%(), capture(), character_class(), counts, group(), not(), rex(), shortcuts, wildcards

Examples

stopifnot(grepl(rex("crab" %if_next_is% "apple"), "crabapple", perl = TRUE))
stopifnot(grepl(rex("crab" %if_prev_is% "apple"), "applecrab", perl = TRUE))
stopifnot(grepl(rex(range("a", "e") %if_next_isnt% range("f", "g")),
  "ah", perl = TRUE))
stopifnot(grepl(rex(range("a", "e") %if_next_is% range("f", "i")),
  "ah", perl = TRUE))

---

Do not match

Description

Do not match

Usage

not(..., type = c("greedy", "lazy", "possessive"))

Arguments

...	shortcuts, R variables, text, or other rex functions.
type	the type of match to perform. There are three match types greedy: match the longest string. This is the default matching type. lazy: match the shortest string. This matches the shortest string from the same anchor point, not necessarily the shortest global string. possessive: match and don't allow backtracking

See Also

Other rex: %or%(), capture(), character_class(), counts, group(), lookarounds, rex(), shortcuts, wildcards

---

Match function

Description

Match function

Usage

re_matches(
  data,
  pattern,
  global = FALSE,
  options = NULL,
  locations = FALSE,
  ...
)

Arguments

data	character vector to match against
pattern	regular expression to use for matching
global	use global matching
options	regular expression options
locations	rather than returning the values of the matched (or captured) string, return a data.frame of the match locations in the string.
...	options passed to regexpr or gregexpr

Value

if no captures, returns a logical vector the same length as the input character vector specifying if the relevant value matched or not. If there are captures in the regular expression, returns a data.frame with a column for each capture group. If global is TRUE, returns a list of data.frames.

See Also

regexp Section "Perl-like Regular Expressions" for a discussion of the supported options

Examples

string <- c("this is a", "test string")
re_matches(string, rex("test")) # FALSE FALSE

# named capture
re_matches(string, rex(capture(alphas, name = "first_word"), space,
  capture(alphas, name = "second_word")))
#   first_word second_word
# 1       this          is
# 2       test      string

# capture returns NA when it fails to match
re_matches(string, rex(capture("test")))
#      1
# 1 test
# 2 <NA>

---

Substitute regular expressions in a string with another string.

Description

Substitute regular expressions in a string with another string.

Usage

re_substitutes(data, pattern, replacement, global = FALSE, options = NULL, ...)

Arguments

data	character vector to substitute
pattern	regular expression to match
replacement	replacement text to use
global	substitute all occurrences
options	option flags
...	options passed to sub or gsub

See Also

regexp Section "Perl-like Regular Expressions" for a discussion of the supported options

Examples

string <- c("this is a Test", "string")
re_substitutes(string, "test", "not a test", options = "insensitive")
re_substitutes(string, "i", "x", global = TRUE)
re_substitutes(string, "(test)", "not a \\1", options = "insensitive")

---

Register the Rex shortcuts

Description

If you are using rex in another package you need to call this function to register all of the rex shortcuts so that spurious NOTEs about global variables being generated during R CMD check.

Usage

register_shortcuts(pkg_name)

Arguments

pkg_name

the package to register the shortcuts in

---

Generate a regular expression.

Description

Generate a regular expression.

Usage

rex(..., env = parent.frame())

Arguments

...	shortcuts, R variables, text, or other rex functions.
env	environment to evaluate the rex expression in.

See Also

Other rex: %or%(), capture(), character_class(), counts, group(), lookarounds, not(), shortcuts, wildcards

---

Toggles rex mode.

Description

While within rex mode, functions used within the rex function are attached, so one can get e.g. auto-completion within editors.

Usage

rex_mode()

---

Shortcuts

Description

Commonly used character classes and regular expressions. These shortcuts are substituted inside rex calls.

Usage

shortcuts

Format

An object of class shortcut of length 116.

Details

names(shortcuts) will give you the full list of available shortcuts.

See Also

Other rex: %or%(), capture(), character_class(), counts, group(), lookarounds, not(), rex(), wildcards

---

Single shortcuts

Description

Each of these shortcuts has both a plural (-s) and inverse (non_) form.

Usage

single_shortcuts

Format

An object of class shortcut of length 18.

---

Wildcards

Description

Wildcards

Usage

zero_or_more(..., type = c("greedy", "lazy", "possessive"))

one_or_more(..., type = c("greedy", "lazy", "possessive"))

maybe(..., type = c("greedy", "lazy", "possessive"))

Arguments

...	shortcuts, R variables, text, or other rex functions.
type	the type of match to perform. There are three match types greedy: match the longest string. This is the default matching type. lazy: match the shortest string. This matches the shortest string from the same anchor point, not necessarily the shortest global string. possessive: match and don't allow backtracking

Functions

• zero_or_more: match ... zero or more times.

• one_or_more: match ... one or more times.

• maybe: match ... zero or one times.

See Also

Other rex: %or%(), capture(), character_class(), counts, group(), lookarounds, not(), rex(), shortcuts

---