I know it's possible to match a word and then reverse the matches using other tools (e.g. grep -v). However, is it possible to match lines that do not contain a specific word, e.g. hede, using a regular expression?
Input:
hoho
hihi
haha
hede
Code:
grep "<Regex for 'doesn't contain hede'>" input
Desired output:
hoho
hihi
haha
The notion that regex doesn't support inverse matching is not entirely true. You can mimic this behavior by using negative look-arounds:
^((?!hede).)*$
The regex above will match any string, or line without a line break, not containing the (sub)string 'hede'. As mentioned, this is not something regex is "good" at (or should do), but still, it is possible.
And if you need to match line break chars as well, use the DOT-ALL modifier (the trailing s in the following pattern):
/^((?!hede).)*$/s
or use it inline:
/(?s)^((?!hede).)*$/
(where the /.../ are the regex delimiters, i.e., not part of the pattern)
If the DOT-ALL modifier is not available, you can mimic the same behavior with the character class [\s\S]:
/^((?!hede)[\s\S])*$/
Explanation
A string is just a list of n characters. Before, and after each character, there's an empty string. So a list of n characters will have n+1 empty strings. Consider the string "ABhedeCD":
┌──┬───┬──┬───┬──┬───┬──┬───┬──┬───┬──┬───┬──┬───┬──┬───┬──┐
S = │e1│ A │e2│ B │e3│ h │e4│ e │e5│ d │e6│ e │e7│ C │e8│ D │e9│
└──┴───┴──┴───┴──┴───┴──┴───┴──┴───┴──┴───┴──┴───┴──┴───┴──┘
index 0 1 2 3 4 5 6 7
where the e's are the empty strings. The regex (?!hede). looks ahead to see if there's no substring "hede" to be seen, and if that is the case (so something else is seen), then the . (dot) will match any character except a line break. Look-arounds are also called zero-width-assertions because they don't consume any characters. They only assert/validate something.
So, in my example, every empty string is first validated to see if there's no "hede" up ahead, before a character is consumed by the . (dot). The regex (?!hede). will do that only once, so it is wrapped in a group, and repeated zero or more times: ((?!hede).)*. Finally, the start- and end-of-input are anchored to make sure the entire input is consumed: ^((?!hede).)*$
As you can see, the input "ABhedeCD" will fail because on e3, the regex (?!hede) fails (there is "hede" up ahead!).
Note that the solution to does not start with “hede”:
^(?!hede).*$
is generally much more efficient than the solution to does not contain “hede”:
^((?!hede).)*$
The former checks for “hede” only at the input string’s first position, rather than at every position.
If you're just using it for grep, you can use grep -v hede to get all lines which do not contain hede.
ETA Oh, rereading the question, grep -v is probably what you meant by "tools options".
Answer:
^((?!hede).)*$
Explanation:
^the beginning of the string,
( group and capture to \1 (0 or more times (matching the most amount possible)),
(?! look ahead to see if there is not,
hede your string,
) end of look-ahead,
. any character except \n,
)* end of \1 (Note: because you are using a quantifier on this capture, only the LAST repetition of the captured pattern will be stored in \1)
$ before an optional \n, and the end of the string
The given answers are perfectly fine, just an academic point:
Regular Expressions in the meaning of theoretical computer sciences ARE NOT ABLE do it like this. For them it had to look something like this:
^([^h].*$)|(h([^e].*$|$))|(he([^h].*$|$))|(heh([^e].*$|$))|(hehe.+$)
This only does a FULL match. Doing it for sub-matches would even be more awkward.
If you want the regex test to only fail if the entire string matches, the following will work:
^(?!hede$).*
e.g. -- If you want to allow all values except "foo" (i.e. "foofoo", "barfoo", and "foobar" will pass, but "foo" will fail), use: ^(?!foo$).*
Of course, if you're checking for exact equality, a better general solution in this case is to check for string equality, i.e.
myStr !== 'foo'
You could even put the negation outside the test if you need any regex features (here, case insensitivity and range matching):
!/^[a-f]oo$/i.test(myStr)
The regex solution at the top of this answer may be helpful, however, in situations where a positive regex test is required (perhaps by an API).
FWIW, since regular languages (aka rational languages) are closed under complementation, it's always possible to find a regular expression (aka rational expression) that negates another expression. But not many tools implement this.
Vcsn supports this operator (which it denotes {c}, postfix).
You first define the type of your expressions: labels are letter (lal_char) to pick from a to z for instance (defining the alphabet when working with complementation is, of course, very important), and the "value" computed for each word is just a Boolean: true the word is accepted, false, rejected.
In Python:
In [5]: import vcsn
c = vcsn.context('lal_char(a-z), b')
c
Out[5]: {a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z} → 𝔹
then you enter your expression:
In [6]: e = c.expression('(hede){c}'); e
Out[6]: (hede)^c
convert this expression to an automaton:
In [7]: a = e.automaton(); a
finally, convert this automaton back to a simple expression.
In [8]: print(a.expression())
\e+h(\e+e(\e+d))+([^h]+h([^e]+e([^d]+d([^e]+e[^]))))[^]*
where + is usually denoted |, \e denotes the empty word, and [^] is usually written . (any character). So, with a bit of rewriting ()|h(ed?)?|([^h]|h([^e]|e([^d]|d([^e]|e.)))).*.
You can see this example here, and try Vcsn online there.
Here's a good explanation of why it's not easy to negate an arbitrary regex. I have to agree with the other answers, though: if this is anything other than a hypothetical question, then a regex is not the right choice here.
With negative lookahead, regular expression can match something not contains specific pattern. This is answered and explained by Bart Kiers. Great explanation!
However, with Bart Kiers' answer, the lookahead part will test 1 to 4 characters ahead while matching any single character. We can avoid this and let the lookahead part check out the whole text, ensure there is no 'hede', and then the normal part (.*) can eat the whole text all at one time.
Here is the improved regex:
/^(?!.*?hede).*$/
Note the (*?) lazy quantifier in the negative lookahead part is optional, you can use (*) greedy quantifier instead, depending on your data: if 'hede' does present and in the beginning half of the text, the lazy quantifier can be faster; otherwise, the greedy quantifier be faster. However if 'hede' does not present, both would be equal slow.
Here is the demo code.
For more information about lookahead, please check out the great article: Mastering Lookahead and Lookbehind.
Also, please check out RegexGen.js, a JavaScript Regular Expression Generator that helps to construct complex regular expressions. With RegexGen.js, you can construct the regex in a more readable way:
var _ = regexGen;
var regex = _(
_.startOfLine(),
_.anything().notContains( // match anything that not contains:
_.anything().lazy(), 'hede' // zero or more chars that followed by 'hede',
// i.e., anything contains 'hede'
),
_.endOfLine()
);
Benchmarks
I decided to evaluate some of the presented Options and compare their performance, as well as use some new Features.
Benchmarking on .NET Regex Engine: http://regexhero.net/tester/
Benchmark Text:
The first 7 lines should not match, since they contain the searched Expression, while the lower 7 lines should match!
Regex Hero is a real-time online Silverlight Regular Expression Tester.
XRegex Hero is a real-time online Silverlight Regular Expression Tester.
Regex HeroRegex HeroRegex HeroRegex HeroRegex Hero is a real-time online Silverlight Regular Expression Tester.
Regex Her Regex Her Regex Her Regex Her Regex Her Regex Her Regex Hero is a real-time online Silverlight Regular Expression Tester.
Regex Her is a real-time online Silverlight Regular Expression Tester.Regex Hero
egex Hero egex Hero egex Hero egex Hero egex Hero egex Hero Regex Hero is a real-time online Silverlight Regular Expression Tester.
RRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRegex Hero is a real-time online Silverlight Regular Expression Tester.
Regex Her
egex Hero
egex Hero is a real-time online Silverlight Regular Expression Tester.
Regex Her is a real-time online Silverlight Regular Expression Tester.
Regex Her Regex Her Regex Her Regex Her Regex Her Regex Her is a real-time online Silverlight Regular Expression Tester.
Nobody is a real-time online Silverlight Regular Expression Tester.
Regex Her o egex Hero Regex Hero Reg ex Hero is a real-time online Silverlight Regular Expression Tester.
Results:
Results are Iterations per second as the median of 3 runs - Bigger Number = Better
01: ^((?!Regex Hero).)*$ 3.914 // Accepted Answer
02: ^(?:(?!Regex Hero).)*$ 5.034 // With Non-Capturing group
03: ^(?!.*?Regex Hero).* 7.356 // Lookahead at the beginning, if not found match everything
04: ^(?>[^R]+|R(?!egex Hero))*$ 6.137 // Lookahead only on the right first letter
05: ^(?>(?:.*?Regex Hero)?)^.*$ 7.426 // Match the word and check if you're still at linestart
06: ^(?(?=.*?Regex Hero)(?#fail)|.*)$ 7.371 // Logic Branch: Find Regex Hero? match nothing, else anything
P1: ^(?(?=.*?Regex Hero)(*FAIL)|(*ACCEPT)) ????? // Logic Branch in Perl - Quick FAIL
P2: .*?Regex Hero(*COMMIT)(*FAIL)|(*ACCEPT) ????? // Direct COMMIT & FAIL in Perl
Since .NET doesn't support action Verbs (*FAIL, etc.) I couldn't test the solutions P1 and P2.
Summary:
The overall most readable and performance-wise fastest solution seems to be 03 with a simple negative lookahead. This is also the fastest solution for JavaScript, since JS does not support the more advanced Regex Features for the other solutions.
Not regex, but I've found it logical and useful to use serial greps with pipe to eliminate noise.
eg. search an apache config file without all the comments-
grep -v '\#' /opt/lampp/etc/httpd.conf # this gives all the non-comment lines
and
grep -v '\#' /opt/lampp/etc/httpd.conf | grep -i dir
The logic of serial grep's is (not a comment) and (matches dir)
Since no one else has given a direct answer to the question that was asked, I'll do it.
The answer is that with POSIX grep, it's impossible to literally satisfy this request:
grep "<Regex for 'doesn't contain hede'>" input
The reason is that with no flags, POSIX grep is only required to work with Basic Regular Expressions (BREs), which are simply not powerful enough for accomplishing that task, because of lack of alternation in subexpressions. The only kind of alternation it supports involves providing multiple regular expressions separated by newlines, and that doesn't cover all regular languages, e.g. there's no finite collection of BREs that matches the same regular language as the extended regular expression (ERE) ^(ab|cd)*$.
However, GNU grep implements extensions that allow it. In particular, \| is the alternation operator in GNU's implementation of BREs. If your regular expression engine supports alternation, parentheses and the Kleene star, and is able to anchor to the beginning and end of the string, that's all you need for this approach. Note however that negative sets [^ ... ] are very convenient in addition to those, because otherwise, you need to replace them with an expression of the form (a|b|c| ... ) that lists every character that is not in the set, which is extremely tedious and overly long, even more so if the whole character set is Unicode.
Thanks to formal language theory, we get to see how such an expression looks like. With GNU grep, the answer would be something like:
grep "^\([^h]\|h\(h\|eh\|edh\)*\([^eh]\|e[^dh]\|ed[^eh]\)\)*\(\|h\(h\|eh\|edh\)*\(\|e\|ed\)\)$" input
(found with Grail and some further optimizations made by hand).
You can also use a tool that implements EREs, like egrep, to get rid of the backslashes, or equivalently, pass the -E flag to POSIX grep (although I was under the impression that the question required avoiding any flags to grep whatsoever):
egrep "^([^h]|h(h|eh|edh)*([^eh]|e[^dh]|ed[^eh]))*(|h(h|eh|edh)*(|e|ed))$" input
Here's a script to test it (note it generates a file testinput.txt in the current directory). Several of the expressions presented in other answers fail this test.
#!/bin/bash
REGEX="^\([^h]\|h\(h\|eh\|edh\)*\([^eh]\|e[^dh]\|ed[^eh]\)\)*\(\|h\(h\|eh\|edh\)*\(\|e\|ed\)\)$"
# First four lines as in OP's testcase.
cat > testinput.txt <<EOF
hoho
hihi
haha
hede
h
he
ah
head
ahead
ahed
aheda
ahede
hhede
hehede
hedhede
hehehehehehedehehe
hedecidedthat
EOF
diff -s -u <(grep -v hede testinput.txt) <(grep "$REGEX" testinput.txt)
In my system it prints:
Files /dev/fd/63 and /dev/fd/62 are identical
as expected.
For those interested in the details, the technique employed is to convert the regular expression that matches the word into a finite automaton, then invert the automaton by changing every acceptance state to non-acceptance and vice versa, and then converting the resulting FA back to a regular expression.
As everyone has noted, if your regular expression engine supports negative lookahead, the regular expression is much simpler. For example, with GNU grep:
grep -P '^((?!hede).)*$' input
However, this approach has the disadvantage that it requires a backtracking regular expression engine. This makes it unsuitable in installations that are using secure regular expression engines like RE2, which is one reason to prefer the generated approach in some circumstances.
Using Kendall Hopkins' excellent FormalTheory library, written in PHP, which provides a functionality similar to Grail, and a simplifier written by myself, I've been able to write an online generator of negative regular expressions given an input phrase (only alphanumeric and space characters currently supported, and the length is limited): http://www.formauri.es/personal/pgimeno/misc/non-match-regex/
For hede it outputs:
^([^h]|h(h|e(h|dh))*([^eh]|e([^dh]|d[^eh])))*(h(h|e(h|dh))*(ed?)?)?$
which is equivalent to the above.
with this, you avoid to test a lookahead on each positions:
/^(?:[^h]+|h++(?!ede))*+$/
equivalent to (for .net):
^(?>(?:[^h]+|h+(?!ede))*)$
Old answer:
/^(?>[^h]+|h+(?!ede))*$/
Aforementioned (?:(?!hede).)* is great because it can be anchored.
^(?:(?!hede).)*$ # A line without hede
foo(?:(?!hede).)*bar # foo followed by bar, without hede between them
But the following would suffice in this case:
^(?!.*hede) # A line without hede
This simplification is ready to have "AND" clauses added:
^(?!.*hede)(?=.*foo)(?=.*bar) # A line with foo and bar, but without hede
^(?!.*hede)(?=.*foo).*bar # Same
An, in my opinon, more readable variant of the top answer:
^(?!.*hede)
Basically, "match at the beginning of the line if and only if it does not have 'hede' in it" - so the requirement translated almost directly into regex.
Of course, it's possible to have multiple failure requirements:
^(?!.*(hede|hodo|hada))
Details: The ^ anchor ensures the regex engine doesn't retry the match at every location in the string, which would match every string.
The ^ anchor in the beginning is meant to represent the beginning of the line. The grep tool matches each line one at a time, in contexts where you're working with a multiline string, you can use the "m" flag:
/^(?!.*hede)/m # JavaScript syntax
or
(?m)^(?!.*hede) # Inline flag
Here's how I'd do it:
^[^h]*(h(?!ede)[^h]*)*$
Accurate and more efficient than the other answers. It implements Friedl's "unrolling-the-loop" efficiency technique and requires much less backtracking.
Another option is that to add a positive look-ahead and check if hede is anywhere in the input line, then we would negate that, with an expression similar to:
^(?!(?=.*\bhede\b)).*$
with word boundaries.
The expression is explained on the top right panel of regex101.com, if you wish to explore/simplify/modify it, and in this link, you can watch how it would match against some sample inputs, if you like.
RegEx Circuit
jex.im visualizes regular expressions:
If you want to match a character to negate a word similar to negate character class:
For example, a string:
<?
$str="aaa bbb4 aaa bbb7";
?>
Do not use:
<?
preg_match('/aaa[^bbb]+?bbb7/s', $str, $matches);
?>
Use:
<?
preg_match('/aaa(?:(?!bbb).)+?bbb7/s', $str, $matches);
?>
Notice "(?!bbb)." is neither lookbehind nor lookahead, it's lookcurrent, for example:
"(?=abc)abcde", "(?!abc)abcde"
The OP did not specify or Tag the post to indicate the context (programming language, editor, tool) the Regex will be used within.
For me, I sometimes need to do this while editing a file using Textpad.
Textpad supports some Regex, but does not support lookahead or lookbehind, so it takes a few steps.
If I am looking to retain all lines that Do NOT contain the string hede, I would do it like this:
1. Search/replace the entire file to add a unique "Tag" to the beginning of each line containing any text.
Search string:^(.)
Replace string:<##-unique-##>\1
Replace-all
2. Delete all lines that contain the string hede (replacement string is empty):
Search string:<##-unique-##>.*hede.*\n
Replace string:<nothing>
Replace-all
3. At this point, all remaining lines Do NOT contain the string hede. Remove the unique "Tag" from all lines (replacement string is empty):
Search string:<##-unique-##>
Replace string:<nothing>
Replace-all
Now you have the original text with all lines containing the string hede removed.
If I am looking to Do Something Else to only lines that Do NOT contain the string hede, I would do it like this:
1. Search/replace the entire file to add a unique "Tag" to the beginning of each line containing any text.
Search string:^(.)
Replace string:<##-unique-##>\1
Replace-all
2. For all lines that contain the string hede, remove the unique "Tag":
Search string:<##-unique-##>(.*hede)
Replace string:\1
Replace-all
3. At this point, all lines that begin with the unique "Tag", Do NOT contain the string hede. I can now do my Something Else to only those lines.
4. When I am done, I remove the unique "Tag" from all lines (replacement string is empty):
Search string:<##-unique-##>
Replace string:<nothing>
Replace-all
Since the introduction of ruby-2.4.1, we can use the new Absent Operator in Ruby’s Regular Expressions
from the official doc
(?~abc) matches: "", "ab", "aab", "cccc", etc.
It doesn't match: "abc", "aabc", "ccccabc", etc.
Thus, in your case ^(?~hede)$ does the job for you
2.4.1 :016 > ["hoho", "hihi", "haha", "hede"].select{|s| /^(?~hede)$/.match(s)}
=> ["hoho", "hihi", "haha"]
Through PCRE verb (*SKIP)(*F)
^hede$(*SKIP)(*F)|^.*$
This would completely skips the line which contains the exact string hede and matches all the remaining lines.
DEMO
Execution of the parts:
Let us consider the above regex by splitting it into two parts.
Part before the | symbol. Part shouldn't be matched.
^hede$(*SKIP)(*F)
Part after the | symbol. Part should be matched.
^.*$
PART 1
Regex engine will start its execution from the first part.
^hede$(*SKIP)(*F)
Explanation:
^ Asserts that we are at the start.
hede Matches the string hede
$ Asserts that we are at the line end.
So the line which contains the string hede would be matched. Once the regex engine sees the following (*SKIP)(*F) (Note: You could write (*F) as (*FAIL)) verb, it skips and make the match to fail. | called alteration or logical OR operator added next to the PCRE verb which inturn matches all the boundaries exists between each and every character on all the lines except the line contains the exact string hede. See the demo here. That is, it tries to match the characters from the remaining string. Now the regex in the second part would be executed.
PART 2
^.*$
Explanation:
^ Asserts that we are at the start. ie, it matches all the line starts except the one in the hede line. See the demo here.
.* In the Multiline mode, . would match any character except newline or carriage return characters. And * would repeat the previous character zero or more times. So .* would match the whole line. See the demo here.
Hey why you added .* instead of .+ ?
Because .* would match a blank line but .+ won't match a blank. We want to match all the lines except hede , there may be a possibility of blank lines also in the input . so you must use .* instead of .+ . .+ would repeat the previous character one or more times. See .* matches a blank line here.
$ End of the line anchor is not necessary here.
The TXR Language supports regex negation.
$ txr -c '#(repeat)
#{nothede /~hede/}
#(do (put-line nothede))
#(end)' Input
A more complicated example: match all lines that start with a and end with z, but do not contain the substring hede:
$ txr -c '#(repeat)
#{nothede /a.*z&~.*hede.*/}
#(do (put-line nothede))
#(end)' -
az <- echoed
az
abcz <- echoed
abcz
abhederz <- not echoed; contains hede
ahedez <- not echoed; contains hede
ace <- not echoed; does not end in z
ahedz <- echoed
ahedz
Regex negation is not particularly useful on its own but when you also have intersection, things get interesting, since you have a full set of boolean set operations: you can express "the set which matches this, except for things which match that".
It may be more maintainable to two regexes in your code, one to do the first match, and then if it matches run the second regex to check for outlier cases you wish to block for example ^.*(hede).* then have appropriate logic in your code.
OK, I admit this is not really an answer to the posted question posted and it may also use slightly more processing than a single regex. But for developers who came here looking for a fast emergency fix for an outlier case then this solution should not be overlooked.
The below function will help you get your desired output
<?PHP
function removePrepositions($text){
$propositions=array('/\bfor\b/i','/\bthe\b/i');
if( count($propositions) > 0 ) {
foreach($propositions as $exceptionPhrase) {
$text = preg_replace($exceptionPhrase, '', trim($text));
}
$retval = trim($text);
}
return $retval;
}
?>
I wanted to add another example for if you are trying to match an entire line that contains string X, but does not also contain string Y.
For example, let's say we want to check if our URL / string contains "tasty-treats", so long as it does not also contain "chocolate" anywhere.
This regex pattern would work (works in JavaScript too)
^(?=.*?tasty-treats)((?!chocolate).)*$
(global, multiline flags in example)
Interactive Example: https://regexr.com/53gv4
Matches
(These urls contain "tasty-treats" and also do not contain "chocolate")
example.com/tasty-treats/strawberry-ice-cream
example.com/desserts/tasty-treats/banana-pudding
example.com/tasty-treats-overview
Does Not Match
(These urls contain "chocolate" somewhere - so they won't match even though they contain "tasty-treats")
example.com/tasty-treats/chocolate-cake
example.com/home-cooking/oven-roasted-chicken
example.com/tasty-treats/banana-chocolate-fudge
example.com/desserts/chocolate/tasty-treats
example.com/chocolate/tasty-treats/desserts
As long as you are dealing with lines, simply mark the negative matches and target the rest.
In fact, I use this trick with sed because ^((?!hede).)*$ looks not supported by it.
For the desired output
Mark the negative match: (e.g. lines with hede), using a character not included in the whole text at all. An emoji could probably be a good choice for this purpose.
s/(.*hede)/🔒\1/g
Target the rest (the unmarked strings: e.g. lines without hede). Suppose you want to keep only the target and delete the rest (as you want):
s/^🔒.*//g
For a better understanding
Suppose you want to delete the target:
Mark the negative match: (e.g. lines with hede), using a character not included in the whole text at all. An emoji could probably be a good choice for this purpose.
s/(.*hede)/🔒\1/g
Target the rest (the unmarked strings: e.g. lines without hede). Suppose you want to delete the target:
s/^[^🔒].*//g
Remove the mark:
s/🔒//g
^((?!hede).)*$ is an elegant solution, except since it consumes characters you won't be able to combine it with other criteria. For instance, say you wanted to check for the non-presence of "hede" and the presence of "haha." This solution would work because it won't consume characters:
^(?!.*\bhede\b)(?=.*\bhaha\b)
How to use PCRE's backtracking control verbs to match a line not containing a word
Here's a method that I haven't seen used before:
/.*hede(*COMMIT)^|/
How it works
First, it tries to find "hede" somewhere in the line. If successful, at this point, (*COMMIT) tells the engine to, not only not backtrack in the event of a failure, but also not to attempt any further matching in that case. Then, we try to match something that cannot possibly match (in this case, ^).
If a line does not contain "hede" then the second alternative, an empty subpattern, successfully matches the subject string.
This method is no more efficient than a negative lookahead, but I figured I'd just throw it on here in case someone finds it nifty and finds a use for it for other, more interesting applications.
Simplest thing that I could find would be
[^(hede)]
Tested at https://regex101.com/
You can also add unit-test cases on that site
A simpler solution is to use the not operator !
Your if statement will need to match "contains" and not match "excludes".
var contains = /abc/;
var excludes =/hede/;
if(string.match(contains) && !(string.match(excludes))){ //proceed...
I believe the designers of RegEx anticipated the use of not operators.
What are these two terms in an understandable way?
Greedy will consume as much as possible. From http://www.regular-expressions.info/repeat.html we see the example of trying to match HTML tags with <.+>. Suppose you have the following:
<em>Hello World</em>
You may think that <.+> (. means any non newline character and + means one or more) would only match the <em> and the </em>, when in reality it will be very greedy, and go from the first < to the last >. This means it will match <em>Hello World</em> instead of what you wanted.
Making it lazy (<.+?>) will prevent this. By adding the ? after the +, we tell it to repeat as few times as possible, so the first > it comes across, is where we want to stop the matching.
I'd encourage you to download RegExr, a great tool that will help you explore Regular Expressions - I use it all the time.
'Greedy' means match longest possible string.
'Lazy' means match shortest possible string.
For example, the greedy h.+l matches 'hell' in 'hello' but the lazy h.+?l matches 'hel'.
Greedy quantifier
Lazy quantifier
Description
*
*?
Star Quantifier: 0 or more
+
+?
Plus Quantifier: 1 or more
?
??
Optional Quantifier: 0 or 1
{n}
{n}?
Quantifier: exactly n
{n,}
{n,}?
Quantifier: n or more
{n,m}
{n,m}?
Quantifier: between n and m
Add a ? to a quantifier to make it ungreedy i.e lazy.
Example:
test string : stackoverflow
greedy reg expression : s.*o output: stackoverflow
lazy reg expression : s.*?o output: stackoverflow
Greedy means your expression will match as large a group as possible, lazy means it will match the smallest group possible. For this string:
abcdefghijklmc
and this expression:
a.*c
A greedy match will match the whole string, and a lazy match will match just the first abc.
As far as I know, most regex engine is greedy by default. Add a question mark at the end of quantifier will enable lazy match.
As #Andre S mentioned in comment.
Greedy: Keep searching until condition is not satisfied.
Lazy: Stop searching once condition is satisfied.
Refer to the example below for what is greedy and what is lazy.
import java.util.regex.Matcher;
import java.util.regex.Pattern;
public class Test {
public static void main(String args[]){
String money = "100000000999";
String greedyRegex = "100(0*)";
Pattern pattern = Pattern.compile(greedyRegex);
Matcher matcher = pattern.matcher(money);
while(matcher.find()){
System.out.println("I'm greedy and I want " + matcher.group() + " dollars. This is the most I can get.");
}
String lazyRegex = "100(0*?)";
pattern = Pattern.compile(lazyRegex);
matcher = pattern.matcher(money);
while(matcher.find()){
System.out.println("I'm too lazy to get so much money, only " + matcher.group() + " dollars is enough for me");
}
}
}
The result is:
I'm greedy and I want 100000000 dollars. This is the most I can get.
I'm too lazy to get so much money, only 100 dollars is enough for me
Taken From www.regular-expressions.info
Greediness: Greedy quantifiers first tries to repeat the token as many times
as possible, and gradually gives up matches as the engine backtracks to find
an overall match.
Laziness: Lazy quantifier first repeats the token as few times as required, and
gradually expands the match as the engine backtracks through the regex to
find an overall match.
From Regular expression
The standard quantifiers in regular
expressions are greedy, meaning they
match as much as they can, only giving
back as necessary to match the
remainder of the regex.
By using a lazy quantifier, the
expression tries the minimal match
first.
Greedy matching. The default behavior of regular expressions is to be greedy. That means it tries to extract as much as possible until it conforms to a pattern even when a smaller part would have been syntactically sufficient.
Example:
import re
text = "<body>Regex Greedy Matching Example </body>"
re.findall('<.*>', text)
#> ['<body>Regex Greedy Matching Example </body>']
Instead of matching till the first occurrence of ‘>’, it extracted the whole string. This is the default greedy or ‘take it all’ behavior of regex.
Lazy matching, on the other hand, ‘takes as little as possible’. This can be effected by adding a ? at the end of the pattern.
Example:
re.findall('<.*?>', text)
#> ['<body>', '</body>']
If you want only the first match to be retrieved, use the search method instead.
re.search('<.*?>', text).group()
#> '<body>'
Source: Python Regex Examples
Greedy Quantifiers are like the IRS
They’ll take as much as they can. e.g. matches with this regex: .*
$50,000
Bye-bye bank balance.
See here for an example: Greedy-example
Non-greedy quantifiers - they take as little as they can
Ask for a tax refund: the IRS sudden becomes non-greedy - and return as little as possible: i.e. they use this quantifier:
(.{2,5}?)([0-9]*) against this input: $50,000
The first group is non-needy and only matches $5 – so I get a $5 refund against the $50,000 input.
See here: Non-greedy-example.
Why do we need greedy vs non-greedy?
It becomes important if you are trying to match certain parts of an expression. Sometimes you don't want to match everything - as little as possible. Sometimes you want to match as much as possible. Nothing more to it.
You can play around with the examples in the links posted above.
(Analogy used to help you remember).
Greedy means it will consume your pattern until there are none of them left and it can look no further.
Lazy will stop as soon as it will encounter the first pattern you requested.
One common example that I often encounter is \s*-\s*? of a regex ([0-9]{2}\s*-\s*?[0-9]{7})
The first \s* is classified as greedy because of * and will look as many white spaces as possible after the digits are encountered and then look for a dash character "-". Where as the second \s*? is lazy because of the present of *? which means that it will look the first white space character and stop right there.
Best shown by example. String. 192.168.1.1 and a greedy regex \b.+\b
You might think this would give you the 1st octet but is actually matches against the whole string. Why? Because the.+ is greedy and a greedy match matches every character in 192.168.1.1 until it reaches the end of the string. This is the important bit! Now it starts to backtrack one character at a time until it finds a match for the 3rd token (\b).
If the string a 4GB text file and 192.168.1.1 was at the start you could easily see how this backtracking would cause an issue.
To make a regex non greedy (lazy) put a question mark after your greedy search e.g
*?
??
+?
What happens now is token 2 (+?) finds a match, regex moves along a character and then tries the next token (\b) rather than token 2 (+?). So it creeps along gingerly.
To give extra clarification on Laziness, here is one example which is maybe not intuitive on first look but explains idea of "gradually expands the match" from Suganthan Madhavan Pillai answer.
input -> some.email#domain.com#
regex -> ^.*?#$
Regex for this input will have a match. At first glance somebody could say LAZY match(".*?#") will stop at first # after which it will check that input string ends("$"). Following this logic someone would conclude there is no match because input string doesn't end after first #.
But as you can see this is not the case, regex will go forward even though we are using non-greedy(lazy mode) search until it hits second # and have a MINIMAL match.
try to understand the following behavior:
var input = "0014.2";
Regex r1 = new Regex("\\d+.{0,1}\\d+");
Regex r2 = new Regex("\\d*.{0,1}\\d*");
Console.WriteLine(r1.Match(input).Value); // "0014.2"
Console.WriteLine(r2.Match(input).Value); // "0014.2"
input = " 0014.2";
Console.WriteLine(r1.Match(input).Value); // "0014.2"
Console.WriteLine(r2.Match(input).Value); // " 0014"
input = " 0014.2";
Console.WriteLine(r1.Match(input).Value); // "0014.2"
Console.WriteLine(r2.Match(input).Value); // ""
What are these two terms in an understandable way?
Greedy will consume as much as possible. From http://www.regular-expressions.info/repeat.html we see the example of trying to match HTML tags with <.+>. Suppose you have the following:
<em>Hello World</em>
You may think that <.+> (. means any non newline character and + means one or more) would only match the <em> and the </em>, when in reality it will be very greedy, and go from the first < to the last >. This means it will match <em>Hello World</em> instead of what you wanted.
Making it lazy (<.+?>) will prevent this. By adding the ? after the +, we tell it to repeat as few times as possible, so the first > it comes across, is where we want to stop the matching.
I'd encourage you to download RegExr, a great tool that will help you explore Regular Expressions - I use it all the time.
'Greedy' means match longest possible string.
'Lazy' means match shortest possible string.
For example, the greedy h.+l matches 'hell' in 'hello' but the lazy h.+?l matches 'hel'.
Greedy quantifier
Lazy quantifier
Description
*
*?
Star Quantifier: 0 or more
+
+?
Plus Quantifier: 1 or more
?
??
Optional Quantifier: 0 or 1
{n}
{n}?
Quantifier: exactly n
{n,}
{n,}?
Quantifier: n or more
{n,m}
{n,m}?
Quantifier: between n and m
Add a ? to a quantifier to make it ungreedy i.e lazy.
Example:
test string : stackoverflow
greedy reg expression : s.*o output: stackoverflow
lazy reg expression : s.*?o output: stackoverflow
Greedy means your expression will match as large a group as possible, lazy means it will match the smallest group possible. For this string:
abcdefghijklmc
and this expression:
a.*c
A greedy match will match the whole string, and a lazy match will match just the first abc.
As far as I know, most regex engine is greedy by default. Add a question mark at the end of quantifier will enable lazy match.
As #Andre S mentioned in comment.
Greedy: Keep searching until condition is not satisfied.
Lazy: Stop searching once condition is satisfied.
Refer to the example below for what is greedy and what is lazy.
import java.util.regex.Matcher;
import java.util.regex.Pattern;
public class Test {
public static void main(String args[]){
String money = "100000000999";
String greedyRegex = "100(0*)";
Pattern pattern = Pattern.compile(greedyRegex);
Matcher matcher = pattern.matcher(money);
while(matcher.find()){
System.out.println("I'm greedy and I want " + matcher.group() + " dollars. This is the most I can get.");
}
String lazyRegex = "100(0*?)";
pattern = Pattern.compile(lazyRegex);
matcher = pattern.matcher(money);
while(matcher.find()){
System.out.println("I'm too lazy to get so much money, only " + matcher.group() + " dollars is enough for me");
}
}
}
The result is:
I'm greedy and I want 100000000 dollars. This is the most I can get.
I'm too lazy to get so much money, only 100 dollars is enough for me
Taken From www.regular-expressions.info
Greediness: Greedy quantifiers first tries to repeat the token as many times
as possible, and gradually gives up matches as the engine backtracks to find
an overall match.
Laziness: Lazy quantifier first repeats the token as few times as required, and
gradually expands the match as the engine backtracks through the regex to
find an overall match.
From Regular expression
The standard quantifiers in regular
expressions are greedy, meaning they
match as much as they can, only giving
back as necessary to match the
remainder of the regex.
By using a lazy quantifier, the
expression tries the minimal match
first.
Greedy matching. The default behavior of regular expressions is to be greedy. That means it tries to extract as much as possible until it conforms to a pattern even when a smaller part would have been syntactically sufficient.
Example:
import re
text = "<body>Regex Greedy Matching Example </body>"
re.findall('<.*>', text)
#> ['<body>Regex Greedy Matching Example </body>']
Instead of matching till the first occurrence of ‘>’, it extracted the whole string. This is the default greedy or ‘take it all’ behavior of regex.
Lazy matching, on the other hand, ‘takes as little as possible’. This can be effected by adding a ? at the end of the pattern.
Example:
re.findall('<.*?>', text)
#> ['<body>', '</body>']
If you want only the first match to be retrieved, use the search method instead.
re.search('<.*?>', text).group()
#> '<body>'
Source: Python Regex Examples
Greedy Quantifiers are like the IRS
They’ll take as much as they can. e.g. matches with this regex: .*
$50,000
Bye-bye bank balance.
See here for an example: Greedy-example
Non-greedy quantifiers - they take as little as they can
Ask for a tax refund: the IRS sudden becomes non-greedy - and return as little as possible: i.e. they use this quantifier:
(.{2,5}?)([0-9]*) against this input: $50,000
The first group is non-needy and only matches $5 – so I get a $5 refund against the $50,000 input.
See here: Non-greedy-example.
Why do we need greedy vs non-greedy?
It becomes important if you are trying to match certain parts of an expression. Sometimes you don't want to match everything - as little as possible. Sometimes you want to match as much as possible. Nothing more to it.
You can play around with the examples in the links posted above.
(Analogy used to help you remember).
Greedy means it will consume your pattern until there are none of them left and it can look no further.
Lazy will stop as soon as it will encounter the first pattern you requested.
One common example that I often encounter is \s*-\s*? of a regex ([0-9]{2}\s*-\s*?[0-9]{7})
The first \s* is classified as greedy because of * and will look as many white spaces as possible after the digits are encountered and then look for a dash character "-". Where as the second \s*? is lazy because of the present of *? which means that it will look the first white space character and stop right there.
Best shown by example. String. 192.168.1.1 and a greedy regex \b.+\b
You might think this would give you the 1st octet but is actually matches against the whole string. Why? Because the.+ is greedy and a greedy match matches every character in 192.168.1.1 until it reaches the end of the string. This is the important bit! Now it starts to backtrack one character at a time until it finds a match for the 3rd token (\b).
If the string a 4GB text file and 192.168.1.1 was at the start you could easily see how this backtracking would cause an issue.
To make a regex non greedy (lazy) put a question mark after your greedy search e.g
*?
??
+?
What happens now is token 2 (+?) finds a match, regex moves along a character and then tries the next token (\b) rather than token 2 (+?). So it creeps along gingerly.
To give extra clarification on Laziness, here is one example which is maybe not intuitive on first look but explains idea of "gradually expands the match" from Suganthan Madhavan Pillai answer.
input -> some.email#domain.com#
regex -> ^.*?#$
Regex for this input will have a match. At first glance somebody could say LAZY match(".*?#") will stop at first # after which it will check that input string ends("$"). Following this logic someone would conclude there is no match because input string doesn't end after first #.
But as you can see this is not the case, regex will go forward even though we are using non-greedy(lazy mode) search until it hits second # and have a MINIMAL match.
try to understand the following behavior:
var input = "0014.2";
Regex r1 = new Regex("\\d+.{0,1}\\d+");
Regex r2 = new Regex("\\d*.{0,1}\\d*");
Console.WriteLine(r1.Match(input).Value); // "0014.2"
Console.WriteLine(r2.Match(input).Value); // "0014.2"
input = " 0014.2";
Console.WriteLine(r1.Match(input).Value); // "0014.2"
Console.WriteLine(r2.Match(input).Value); // " 0014"
input = " 0014.2";
Console.WriteLine(r1.Match(input).Value); // "0014.2"
Console.WriteLine(r2.Match(input).Value); // ""
Java is hanging with 100% CPU usage when I use the below string as input for a regular expression.
RegEx Used:
Here is the regular expression used for the description field in my application.
^([A-Za-z0-9\\-\\_\\.\\&\\,]+[\\s]*)+
String used for testing:
SaaS Service VLAN from Provider_One
2nd attempt with Didier SPT because the first one he gave me was wrong :-(
It works properly when I split the same string in different combinations. Like "SaaS Service VLAN from Provider_One", "first one he gave me was wrong :-(", etc. Java is hanging only for the above given string.
I also tried optimizing the regex as below.
^([\\w\\-\\.\\&\\,]+[\\s]*)+
Even with this is not working.
Another classic case of catastrophic backtracking.
You have nested quantifiers that cause a gigantic number of permutations to be checked when the regex arrives at the : in your input string which is not part of your character class (assuming you're using the .matches() method).
Let's simplify the problem to this regex:
^([^:]+)+$
and this string:
1234:
The regex engine needs to check
1234 # no repetition of the capturing group
123 4 # first repetition of the group: 123; second repetition: 4
12 34 # etc.
12 3 4
1 234
1 23 4
1 2 34
1 2 3 4
...and that's just for four characters. On your sample string, RegexBuddy aborts after 1 million attempts. Java will happily keep on chugging... before finally admitting that none of these combinations allows the following : to match.
How can you solve this?
You can forbid the regex from backtracking by using possessive quantifiers:
^([A-Za-z0-9_.&,-]++\\s*+)+
will allow the regex to fail faster. Incidentally, I removed all those unnecessary backslashes.
Edit:
A few measurements:
On the string "was wrong :-)", it takes RegexBuddy 862 steps to figure out a non-match.
For "me was wrong :-)", it's 1,742 steps.
For "gave me was wrong :-)", 14,014 steps.
For "he gave me was wrong :-)", 28,046 steps.
For "one he gave me was wrong :-)", 112,222 steps.
For "first one he gave me was wrong :-)", >1,000,000 steps.
First, you need to realize that your regexes CANNOT match the supplied input string. The strings contain a number of characters ('<' '>' '/' ':' and ')') that are not "word" characters.
So why is it taking so long?
Basically "catastrophic backtracking". More specifically, the repeating structures of your regex give an exponential number of alternatives for the regex backtracking algorithm to try!
Here's what your regex says:
One or more word characters
Followed by zero or more space characters
Repeat the previous 2 patterns as many times as you like.
The problem is with the "zero or more space characters" part. The first time, the matcher will match everything up to the first unexpected character (i.e. the '<'). Then it will back off a bit and try again with a different alternative ... that involves "zero spaces" before the last letter, then when that fails, it will move the "zero spaces" back one position.
The problem is that for String with N non-space characters, there as N different places that "zero spaces" can be matched, and that makes 2^N different combinations. That rapidly turns into a HUGE number as N grows, and the end result is hard to distinguish from an infinite loop.
Why are you matching whitespace separately from the other characters? And why are you anchoring the match at the beginning, but not at the end? If you want to make sure the string doesn't start or end with whitespace, you should do something like this:
^[A-Za-z0-9_.&,-]+(?:\s+[A-Za-z0-9_.&,-]+)*$
Now there's only one "path" the regex engine can take through the string. If it runs out of characters that match [A-Za-z0-9_.&,-] before reaching the end, and the next character doesn't match \s, it fails immediately. If it reaches the end while still matching whitespace characters, it fails because it's required to match at least one non-whitespace character after each run of whitespace.
If you want to make sure there's exactly one whitespace character separating the runs of non-whitespace, just remove the quantifier from \s+:
^[A-Za-z0-9_.&,-]+(?:\s[A-Za-z0-9_.&,-]+)*$
If you don't care where the whitespace is in relation to the non-whitespace, just match them all with the same character class:
^[A-Za-z0-9_.&,\s-]+$
I'm assuming you know that your regex won't match the given input because of the : and ( in the smiley, and you just want to know why it takes so long to fail.
And of course, since you're creating the regex in the form of a Java string literal, you would write:
"^[A-Za-z0-9_.&,-]+(?:\\s+[A-Za-z0-9_.&,-]+)*$"
or
"^[A-Za-z0-9_.&,-]+(?:\\s[A-Za-z0-9_.&,-]+)*$"
or
"^[A-Za-z0-9_.&,\\s-]+$"
(I know you had double backslashes in the original question, but that was probably just to get them to display properly, since you weren't using SO's excellent code formatting feature.)