clear understanding on possesive quantifiers

Question

i have read so many questions and answers on possessive quantifiers in java in this website and java tutorials but men, am still confused!! am not understanding!! for example lets say

 my regex is .*+foo
 my input string is mdfoo

according to my understanding,

.(dot) means there are other characters before foo

*(star) means if foo should be writen in a string form if found by regex

+foo means that a string should end with foo for it to match

and in my input string,it ends with foo. but when i receive a statement that there was no match when i run my program! how is this happening and what should i do still using the (+) quantifier to receive a match ?. thanks in advance

Besides the one I already mentioned, here are more: http://stackoverflow.com/questions/1117467/can-someone-explain-possessive-quantifiers-to-me-regular-expressions and http://stackoverflow.com/questions/6485958/understanding-possessive-quantifiers-java-regex and even [more](https://www.google.com/search?q=possessive+quantifiers+stackoverflow)... — Bart Kiers, Jun 02 '14 at 14:28

sp00m · Answer 1 · 2014-06-02T14:27:05.477

0

The possessive quantifier implies that no backtracking is done.

What happens when you try to match mdfoo against .*+foo:

The first part of the pattern (i.e. .*+) matches the whole string mdfoo
But the second part of the pattern (i.e. foo) is not found after the first match
Since there is no backtracking, the attempt immediately fails

These possessive quantifiers are quite clearly explained here.

edited Jun 02 '14 at 14:27

answered Jun 02 '14 at 14:21

sp00m

44,266
23
127
230

@spOOm thanks, your answer is straight to the point. so it seems like i cant use possessive to help a user search for something? its like possessive have no great importance? – Edijae Crusar Jun 02 '14 at 16:05
@gikarasojokinene Well, the question actually is: why do you think you need a possessive quantifier? – sp00m Jun 02 '14 at 17:00
@spOOm i think i need possessive to speed up my regex and help my regex fail faster. – Edijae Crusar Jun 02 '14 at 17:15
@spOOm thats what i know about benefits of possesive. is there any great benefit you can tell me? – Edijae Crusar Jun 02 '14 at 17:31
@gikarasojokinene http://c2.com/cgi/wiki?PrematureOptimization `;)` – sp00m Jun 02 '14 at 18:11
@spOOm men, you knocked me in ma head. i got the message. – Edijae Crusar Jun 03 '14 at 14:22

score 0 · Answer 2 · answered Jun 02 '14 at 14:23

There are three types of quantifiers:

the "regular" quantifiers (*, +, ?) also called "greedy" quantifiers;
the "lazy", quantifiers (*?, +?, ??);
the "possessive" quantifiers (*+, ++, ?+).

For instance, take this input:

The answer is 42

Now, take this regex:

.*(\d+)

The question is, what will be captured by (\d+) according to which version of * you use in .*:

if *, what will be captured is 2;
if *?, what will be captured is 42;
if *+, the regex does not match.

Why:

the greedy quantifier swallows everything it can; however, it retains, on its way, the positions where it has matched; when having swallowed the full text, there is still \d+ to match; it will reluctantly backtrack until \d+ is satisfied, and \d+ is satisfied with 2;
the lazy quantifier tries and queries the following regex token: "if \d+ does not match then I swallow the next character"; when encountering 4 it lets \d+ do its job, therefore 42 is captured;
the possessive quantifier is possessive; it works like the greedy quantifier except that it does not retain any positions; therefore, when reaching the end, the regex engine asks "OK, can you give back?", .*+ says "No...", therefore no match.

clear understanding on possesive quantifiers

2 Answers2