CODEWITHSUNDEEP
javaregex
Dragan
Dragan

Reputation: 3743

Regex - starts with OPERATION and must be followed by either an integer or double

SO I have the follow input string

OPERATION: 12, 12.32, 54.3332

OPERATION can be any of MIN, MAX, SUM. The regex should accept strings that only start with any of those three words follow by colon and the followed by either integer or double.

I have been googling and fidling around for hours now before I decided to turn to you guys.

Thanks for your time!

EDIT:

So far I have ended with this regex

^[MAX:SUM:AVERAGE:MIN:(\\d+(\\.\\d+)?), ]+$

It matches "MAX: " as correct as well as "MAX:" and "MA: . It also matches the strings of the following format : "MAX: 12, 12.3323......"

Upvotes: 0

Views: 495

Answers (3)

Bernhard Barker
Bernhard Barker

Reputation: 55609

You misunderstand the meaning of []. These refer to single character alternatives.

So [MAX:SUM:AVERAGE:MIN:(\\d+(\\.\\d+)?), ] is either M or A or X or .... So something like MMMMM should also match.

You may want something more like this:

^(MAX|SUM|AVERAGE|MIN): (\\d+(\\.\\d+)?(, (?=.)|$))+$

Explanation:

(MAX|SUM|AVERAGE|MIN) is either MAX, SUM, AVERAGE or MIN.
": " refers to the actual characters : and space.
\\d+(\\.\\d+)? is what you had before.
", " refers to the actual characters , and space.
(?=.) is look-ahead. Checking that the following characters (?=) matches any single character (.) (thus not end-of-string) so there isn't a ", " required at the end.
So , (?=.)|$ is either ", " not at the end or nothing at the end.

Alternative: Not using look-ahead

^(MAX|SUM|AVERAGE|MIN): (\\d+(\\.\\d+)?(, \\d+(\\.\\d+)?)*)$

Test for both.

Reference.

Upvotes: 2

Micha Wiedenmann
Micha Wiedenmann

Reputation: 20843

(?:MIN|MAX|SUM):[ ]+\d+(?:[.]\d+)?(?:,[ ]+\d+(?:[.]\d+)?)*

Explanation

(?:MIN|MAX|SUM) - either one of the operations, non-capturing
:             - a literal :
[ ]+          - 1 or more space characters
\d+           - 1 or more digits
(?:[.]\d+)?   - optionally a literal . followed by 1 or more digits, non-capturing
(?:,...)*     - a literal comma and another number, 0 or more times, non-capturing

[.] could also be written as \. but putting magic characters in a character class instead is a best practice, same goes for using [ ] instead of a space.

References

Hope that helps.

Upvotes: 2

Giorgio Desideri
Giorgio Desideri

Reputation: 179

java.util.regex.Pattern accept regular expression formed by group of Greedy format expressions. You can create group to divide your input string.

OPERATION: NUM, NUM, DECIMAL, DECIMAL

Your regular expression is:

([MIN,MAX,SUM]+):([0-9]{4}.[0-9]{2}),(...),(...),(...)

and use:

Matcher.group()

to iterate upon found groups.

Note: I supposed that your number format is "dddd.dd"

Upvotes: 1

Related Questions