Capturing every repetition of group

Question

I need to check whether a string matches the following logical pattern:

string, string, string

I NEED TO USE A REGEXP and I cannot use functions.

For example the following string should match:

hello,hola, ciao,bye -> [hello,hola,ciao,bye]

The following shouldn't match:

hello -> []

I'm using this regexp right now but it's working fine only for the first string. I'm trying to exclude the single string without a comma in the second example.

([^,]+)

The following regexp doesn't work neither: isn't it supposed to match every group and then the end of the line BUT before the end of the line a comma followed by a string must be present?

([^,]+)(?<=,)$

Any ideas?

No functions allowed, the language shouldn't matter! However javascript :) — Damiano Barbati, Dec 14 '14 at 17:06
@DamianoBarbati: Well the capturing facility is not something *inherent* to regexes. It's just because many programming languages have made large regex libraries aspects like capturing are well supported. But the framework thus matters. — Willem Van Onsem, Dec 14 '14 at 17:07
It's false, the language matter because languages haven't always the same regex engine. — Casimir et Hippolyte, Dec 14 '14 at 17:07

score 0 · Answer 1 · answered Dec 14 '14 at 17:09

0

You have to match on ([^,]+),?, and iterate over each match. Regex doesn't have the concept of "arrays".

Alternatively, you can "split" on , and be done.

answered Dec 14 '14 at 17:09

Daniel

4,481
14
34

Edit: I don't quite need the array, but just matching correctly a sequence of "string,string,string" and not a simple "string" – Damiano Barbati Dec 14 '14 at 17:11
@DamianoBarbati What about `if (str.indexOf(',') > -1) {`?? – h2ooooooo Dec 14 '14 at 17:11
I see, do you need it to be "space" free? You might not be able to have a one-liner to do that. You could first see if it "has a comma", and then just strip spaces. – Daniel Dec 14 '14 at 17:12
what do you mean with space free? – Damiano Barbati Dec 14 '14 at 17:15
I mean turning `hola, ciao` into `hola,ciao` (without whitespace) – Daniel Dec 14 '14 at 17:19
nope that's not needed, I don't need the groups but just correct matching even if groups are not trimmed. – Damiano Barbati Dec 14 '14 at 17:21
I updated the question anyway. – Damiano Barbati Dec 14 '14 at 17:23

Gilles Quénot · Answer 2 · 2014-12-14T17:15:43.780

0

Try doing this with a positive look ahead :

/\w+(?=,)/g

Example with perl:

$ echo 'hello,hola, ciao,bye' | 
    perl -lne 'my @a = /\w+(?=,)/g; print join "\n", @a'
hello
hola
ciao

Or using split :

$ echo 'hello,hola, ciao,bye' |
    perl -lne 'my @a = split /[,\s]+/g; print join "\n", @a'
hello
hola
ciao
bye

edited Dec 14 '14 at 17:15

answered Dec 14 '14 at 17:09

Gilles Quénot

173,512
41
224
223

nice one, i'm on that path too but that doesn't match last repetition – Damiano Barbati Dec 14 '14 at 17:10
I don't understand: which one is supposed to be the correct regexp? Thanks anyway! – Damiano Barbati Dec 14 '14 at 17:15

score 0 · Accepted Answer · answered Dec 14 '14 at 17:24

0

I think what you are looking for is:

^[^,]+(?:,[^,]+)+$

answered Dec 14 '14 at 17:24

Casimir et Hippolyte

88,009
5
94
125

yes! It works! May I ask you to explain a little? ^[^,]+ is to match the first string, then a non capturing group which matches a comma and again a string without comma repeated until the end of the line. right? – Damiano Barbati Dec 14 '14 at 17:27
@DamianoBarbati: yes, the description is good. – Casimir et Hippolyte Dec 14 '14 at 17:29

Capturing every repetition of group

3 Answers3