Simple text and numbers?

deep

Hello,

How do i make sure that the characters entered in any of my textfields are ONLY letters or numbers?

How does ereg and those functions work?

Thanks.

bodzan

ereg("[a-zA-Z0-9]",$field);

but use preg_match() because it's faster

preg_match("[a-zA-Z0-9]",$field);

And try this first: it's way better than posting this kind of simple posts. You might even actually learn something...

deep

Thanks,

I havnt been able to find a good explanation of [a-zA-Z0-9] thats why im asking. Every where i look i see
[u#%"&&"66#&&%%] (baad example) :rolleyes:

bodzan

Go here

deep

I havnt read the whole tutorial on regex yet but writing

preg_match("[a-zA-Z0-9]",$field);

will return true eaven if any of the letters a-z A-Z or the numbers 0-9 are found in the string mixed with any other scaracters...

How do I write to get it to only return true if the string is containing characters from a-z or numbers 0-9, no other characters are allowed.

Thanks.

tsinka

Hi,

try

preg_match("^[a-zA-Z0-9]+$",$field);

Thomas

deep

preg_match("^[a-zA-Z0-9]+$",$field);

Thomas [/B]

Error: Warning: No ending delimiter '^' found ?

tsinka

preg_match("§^{[a-zA-Z0-9]+$§",$field);}

Klaus_Paiva

That error ocurred because every perl-compatible regex must start and end with the same delimiter, like was modified in the post above.

deep

Originally posted by tsinka
preg_match("§^{[a-zA-Z0-9]+$§",$field);}

that gives:

Notice: Undefined variable: §
Warning: No ending delimiter '§' found :queasy:

deep

okok, i replaced § with / and it works now!

Last question. Swedish alphabet uses å ä ö at the end of the alphabet. How do I make these availible just like a-z ?

Thanks!

deep

ah! okok just fixed it! this works:

/^{[a-öA-Ö0-9]+$/}

EDIT:

No wait! Just noticed that the regex will allow characters like _ £ and a few more... thats not right!?

SECOND EDIT:

ok, fixed it with this: /^{[a-zA-Z0-9+å+ä+ö]+$/}

hope it works. 🙂

laserlight

Actually, for this special case you can avoid regex by using [man]ctype_alnum/man, and setting the appropriate locale with [man]setlocale/man if necessary.