View Category
Regex
Matching

Check if a string matches a regular expression

Display "ok" if "Hello" matches /[A-Z][a-z]+/
python
found = re.match(r'[A-Z][a-z]+', 'Hello')
if found:
print 'ok'
clojure
(if (re-matches #"[A-Z][a-z]+" "Hello")
(println "ok"))
erlang
String = "Hello", Regexp = "[A-Z][a-z]+",
is_match(String, Regexp) andalso (begin io:format("ok~n"), true end).
case re:run("Hello", "[A-Z][a-z]+") of {match, _} -> ok end.

Check if a string matches with groups

Display "two" if "one two three" matches /one (.*) three/
python
match = re.match(r'one (.*) three', 'one two three')
if match:
print match.group(1)
clojure
(if-let [groups (re-matches #"one (.*) three" "one two three")]
(println (second groups)))
erlang
case re:run("one two three", "one (.*) three", [{capture, [1], list}]) of {match, Res} -> hd(Res) end.
Searching

Check if a string contains a match to a regular expression

Display "ok" if "abc 123 @#$" matches /\d+/
python
found = re.search(r'\d+', 'abc 123 @#$')
if found:
print 'ok'
clojure
(if (re-find #"\d+" "abc 123 @#$")
(println "ok"))
erlang
% Erlang uses 'egrep'-compatible regular expressions, so shortcuts like '\d' not supported
String = "abc 123 @#$", Regexp = "[0-9]+",
is_match(String, Regexp) andalso (begin io:format("ok~n"), true end).
case re:run("abc 123 @#$", "\\d+") of {match, _} -> ok end.

Loop through a string matching a regex and performing an action for each match

Create a list [fish1,cow3,boat4] when matching "(fish):1 sausage (cow):3 tree (boat):4" with regex /\((\w+)\):(\d+)/
python
map(''.join, re.findall(r"\((\w+)\):(\d+)", "(fish):1 sausage (cow):3 tree (boat):4"))
--------------------------------------------------------------------------
(''.join(m.groups()) for m in re.finditer(r"\((\w+)\):(\d+)", "(fish):1 sausage (cow):3 tree (boat):4"))
clojure
(let [matcher (re-matcher #"\((\w+)\):(\d+)" "(fish):1 sausage (cow):3 tree (boat):4")]
(loop [match (re-find matcher)
lst []]
(if match
(recur (re-find matcher) (conj lst (str (second match) (nth match 2))))
lst)))
erlang
solve(S) ->
R = "\\((\\w+?)\\):(\\d+)",
{match, M} = re:run(S,R, [global, {capture, all_but_first, list}]),
[ A++N || [A, N] <- M].
Replacing

Replace the first regex match in a string with a static string

Transform "Red Green Blue" into "R*d Green Blue" by replacing /e/ with "*"
python
print re.sub(r'e', '*', 'Red Green Blue', 1)
clojure
(.replaceFirst (re-matcher #"e" "Red Green Blue") "*")
erlang
{ok, Replaced, _} = regexp:sub("Red Green Blue", "e", "*"),
re:replace("Red Green Blue", "e", "*", [{return, list}]).

Replace all regex matches in a string with a static string

Transform "She sells sea shells" into "She X X shells" by replacing /se\w+/ with "X"
python
transformed = re.sub(r'se\w+', 'X', 'She sells sea shells')
clojure
(.replaceAll (re-matcher #"se\w+" "She sells sea shells") "X")
erlang
% Erlang uses 'egrep'-compatible regular expressions, so shortcuts like '\w' not supported
{ok, Replaced, _} = regexp:gsub("She sells sea shells", "se[A-Za-z0-9_]+", "X"),
re:replace("She sells sea shells", "se\\w+", "X", [global, {return, list}]).

Replace all regex matches in a string with a dynamic string

Transform "The {Quick} Brown {Fox}" into "The kciuQ Brown xoF" by reversing words in braces using the regex /\{(\w+)\}/.
python
transformed = re.sub(r'\{(\w+)\}',
lambda match: match.group(1)[::-1],
'The {Quick} Brown {Fox}')
clojure
(def *string* "The {Quick} Brown {Fox}")
(def *regex* (re-pattern #"\{(\w+)\}"))

(println
(loop [result ""
src *string*
replace-strs (re-seq *regex* *string*)]
(if (empty? src)
result
(let [[match replacement] (first replace-strs)]
(if (= (first src) (first match))
; At the beginning of a sequence that should be replaced.
; Do replacement of a single match
(recur (str result (apply str (reverse replacement)))
(drop (count match) src)
(rest replace-strs))
; else, just copy one char from the source to the result
(recur (str result (first src))
(rest src)
replace-strs))))))
(clojure.string/replace "The {Quick} Brown {Fox}"
#"\{(\w+)\}"
(fn [[_ word]] (apply str (reverse word))))
erlang
% Erlang regular expressions lack both group capture and backreferences, thus this problem is not directly
% solvable. Presented solution is close, but not on-spec

String = "The {Quick} Brown {Fox}",
{match, FieldList} = regexp:matches(String, "\{([A-Za-z0-9_]+)\}"),

NewString = lists:foldl(fun ({Start, Length}, S) -> replstr(S, lists:reverse(string:substr(S, Start, Length)), Start) end, String, FieldList),