regex to allow text and punctuation marks

[shaddf]

how can I make this regex allow punction marks

"/^([a-zA-Z\s ]){4,30}$/"

At the moment it doesn't.

[Jacques1]

You'll have to be a bit more specific. What kind of puncuation marks? Period, comma, semicolon? And where exactly are they allowed? Are you OK with “.......”?

[shaddf]

You'll have to be a bit more specific. What kind of puncuation marks? Period, comma, semicolon? And where exactly are they allowed? Are you OK with “.......”?

at the end of a sentence, just like in literature i.e period,comma,semicolon,colon,hyphen and question mark

[printf]

Something like...

 

"/^[a-z\s]{4,30}+([\.!,;:-?]{1}+)?$/i",

[Jacques1]

That regex makes absolutely no sense.

 

For example, this will match

aaaa<

but not

This is an English sentence, definitely.

And what is the quantifier combination “{1}+” supposed to do?

 

Sentences look more like this:

<?php

$word      = '[a-zA-Z]+';                     // a single word
$words     = "$word( $word)*";                // a sequence of space-separated words
$sentence  = "$words([,;:] $words)*[.!?]";    // a sentence

$sentences = "/$sentence+/";

$testInput = 'This is an English sentence, definitely. And another one.';
var_dump(preg_match($sentences, $testInput));

Edited October 27, 2015 by Jacques1

[Psycho]

There are many characters that can legitimately be within and at the end of a sentence " . . . just like in literature". If this is for a real-world application that will presumably produce an error for sentences entered by users, you are going to have the very likely possibility of false negatives.

[printf]

My regex makes no sense? I think it's the other way around. The OP asked for a regex that allows for punctuation marks at the end of a sentence, yours allows for repeated punctuation marks anywhere but the beginning of the sentence;;;;;;;:,, which makes no sense to me! If you want to strictly match a English literature sentence, then something like this will work.

 

/^(?:\w+(\, |\; |\: | )){1,}(?:\w+(\.|\!|\?))$/i

Edited October 27, 2015 by printf

[Jacques1]

This is, again, nonsense. Please learn the basic syntax of regular expressions before you try to give advice.

• The \w character class includes the underscore and digits, which means you consider “_ _!” or “123 123.” valid English sentences. This obviously makes no sense.
• What's the point of “{1,}”? I guess what you're looking for is the + quantifier.
• Why are you escaping characters that don't require any escaping in the first place?  Characters like “,” or “;” or “:” can be written down verbatim, you know? They don't have any meaning in regexes.
• Why do you require more than one word? “Go!” is a valid sentence, don't you think?

[Jacques1]

Anyway, I agree with Psycho that trying to “validate” text causes more harm than good.

 

The English language is much, much more complex than anything a regex could cover. Of course we can check the basic structure, but this will exclude a large amount of perfectly valid text. Who are we to decide that a text is “invalid”, anyway? People use all kinds of nonstandard language constructs, and that doesn't mean they're all wrong.

 

Assuming that words only consist of a-zA-Z is already a misconception. What about “Raison d'être”? What about “O'Neil”?

 

So unless you have a good reason why you want to annoy your users and force them to use some primitive subset of the English language, just forget about it.