Regex: Matching all words in a phrase

Question

Is this possible?

For a sentence such as hello how are you, I'd like my regular expression to return hello how are you. It only ever returns just hello and not the other words.

My Regex:

[A-Za-z]*

Any help is greatly appreciated. Thanks! If it matters, I'm using Pharo Smalltalk. I've been testing in c# too.

score 7 · Answer 1 · answered Feb 24 '16 at 01:16

7

Also in Pharo send the #substrings message:

'Hello how are you' substrings

and get the array:

#('Hello' 'how' 'are' 'you').

answered Feb 24 '16 at 01:16

Leandro Caniglia

14,495
4
29
51

score 6 · Answer 2 · answered Feb 23 '16 at 22:47

You can find a chapter about Regex in Pharo here:

https://ci.inria.fr/pharo-contribution/view/Books/job/DeepIntoPharo/lastSuccessfulBuild/artifact/tmp/PBE2.pdf

I you just want to split the string on the spaces you can just run:

Character space split: 'My String To split'

You will get an OrderedCollection with all the words.

score 5 · Accepted Answer · answered Feb 23 '16 at 17:30

5

If you only need split the sentence by spaces this can be done using string.Split() method:

var s = "hello how are you";
var words = s.Split();

If you want to use regular expressions:

var s = "hello how are you";
var regex = "\\w+";
var words = Regex.Matches(s, regex).Cast<Match>().Select(m => m.Value);

answered Feb 23 '16 at 17:30

Arturo Menchaca

15,783
1
29
53

So if I understand correctly, a regex engine can't return multiple matches itself, so it depends on the classes you can use within the programming language? – Dan Feb 23 '16 at 17:35
@KingDan: each programming language defines its regex syntax (mainly PCRE). The way you can use depends on the classes/functions that were implemented for it – Arturo Menchaca Feb 23 '16 at 17:43
Damn, so I think I've run into a Smalltalk issue then. I'm able to get this working properly in C#. Thanks for your help! – Dan Feb 23 '16 at 17:46
Nope, not a smalltalk issue. Your regex doesn't match more than hello – Stephan Eggermont Feb 25 '16 at 16:24
@StephanEggermont: my code is in C#, I just ran and gave me the 4 words. – Arturo Menchaca Feb 25 '16 at 16:28
Yours yes, the one from the question? – Stephan Eggermont Feb 25 '16 at 16:29
@StephanEggermont: I never used smalltalk, but probably there will be a function to find all matches. – Arturo Menchaca Feb 25 '16 at 16:33

score 2 · Answer 4 · answered Feb 23 '16 at 17:43

You don't need the Regex in this case at all. Simply use Split.

string str = "hello how are you";
string[] parts = str.Split(' ');

In case you really really want the Regex too much, \w+ as Regex captures any word. So in C# the Regex should look like this string regex = "\\w+" if you need at least word.

The \w stands for any word including characters as
The + quantifier stands for at least one time
The * quantifier stands for zero or more times

score 2 · Answer 5 · answered Feb 25 '16 at 19:58

The standard tries to match, which it doesn't because there are spaces

matcher := RxMatcher forString: '[A-Za-z]*'.
matcher matches: 'hello how are you'

false

If you ask for all matches it tells you there are 5, because * also matches zero characters

matcher := RxMatcher forString: '[A-Za-z]*'.
matcher matchesIn: 'hello how are you'

"an OrderedCollection('hello' 'how' 'are' 'you' '')"

And for the wanted result you might try

matcher := RxMatcher forString: '[A-Za-z]+'.
matcher matchesIn: 'hello how are you'

"an OrderedCollection('hello' 'how' 'are' 'you')"

and if you want to know how long the words are you can do

matcher := RxMatcher forString: '[A-Za-z]+'.
matcher matchesIn: 'hello how are you' collect: [ :each | each size ]

"an OrderedCollection(5 3 3 3)"

Regex: Matching all words in a phrase

5 Answers5