Regular expression

[davranfor]

Hello

I need a regular expression that validate a list of numbers separated
by "-" , numbers can not be greater than 999
Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Thanks ;)

[Rik Wasmus]

On Wed, 17 Oct 2007 21:59:49 +0200, <da*******@gmail.comwrote:

Hello

I need a regular expression that validate a list of numbers separated
by "-" , numbers can not be greater than 999
Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Hmmz, showing a 'best try' would be appreciated. However, I normally can't
resist a good regex, so here it goes:

'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'
Untested BTW. I've made 78-034-89 invalid, as a number should not start
with zero, unless it's only 1 zero. 't Would be simpler if it was valid,
your choice.
--
Rik Wasmus

[davranfor]

On 17 oct, 20:20, "Rik Wasmus" <[e-mail]:

On Wed, 17 Oct 2007 21:59:49 +0200, <[e-mail]:

Hello

I need a regular expression that validate a list of numbers separated
by "-" , numbers can not be greater than 999

Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Hmmz, showing a 'best try' would be appreciated. However, I normally can't
resist a good regex, so here it goes:

'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'

Untested BTW. I've made 78-034-89 invalid, as a number should not start
with zero, unless it's only 1 zero. 't Would be simpler if it was valid,
your choice.
--
Rik Wasmus

thanks ;)))

[davranfor]

On 17 oct, 20:20, "Rik Wasmus" <[e-mail]:

On Wed, 17 Oct 2007 21:59:49 +0200, <[e-mail]:

Hello

I need a regular expression that validate a list of numbers separated
by "-" , numbers can not be greater than 999

Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Hmmz, showing a 'best try' would be appreciated. However, I normally can't
resist a good regex, so here it goes:

'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'

Untested BTW. I've made 78-034-89 invalid, as a number should not start
with zero, unless it's only 1 zero. 't Would be simpler if it was valid,
your choice.
--
Rik Wasmus

works like a charm, can you explain the simple way without check zero
at start?

thanks again

[Rik Wasmus]

On Wed, 17 Oct 2007 22:50:59 +0200, <da*******@gmail.comwrote:

On 17 oct, 20:20, "Rik Wasmus" <[e-mail]:

>On Wed, 17 Oct 2007 21:59:49 +0200, <[e-mail]:

Hello

I need a regular expression that validate a list of numbers separated
by "-" , numbers can not be greater than 999

Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Hmmz, showing a 'best try' would be appreciated. However, I normally
can't
resist a good regex, so here it goes:

'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'

Untested BTW. I've made 78-034-89 invalid, as a number should not start
with zero, unless it's only 1 zero. 't Would be simpler if it was valid,
your choice.

works like a charm, can you explain the simple way without check zero
at start?

'/^ # match start of string
[0-9]{1,3} # match 1 to 3 digits
( # start subpattern
- # literal -
[0-9]{1,3} # match 1 to 3 digits
)* # match subpattern zero or more times
$ # match has to run all the way up untill the end of the string
/x'

When starting with regexes, and certainly for the more complex ones, don't
underestimate the power of /x: comments in the regex itself, indentation
to keep things clear, it does wonders. It also keeps you from
rediscovering how the hell your regex worked 6 months after you wrote it
when you have to track a bug.....
--
Rik Wasmus

[davranfor]

On 17 oct, 21:10, "Rik Wasmus" <[e-mail]:

On Wed, 17 Oct 2007 22:50:59 +0200, <[e-mail]:

On 17 oct, 20:20, "Rik Wasmus" <[e-mail]:

On Wed, 17 Oct 2007 21:59:49 +0200, <[e-mail]:
Hello

I need a regular expression that validate a list of numbers separated
by "-" , numbers can not be greater than 999

Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Hmmz, showing a 'best try' would be appreciated. However, I normally
can't
resist a good regex, so here it goes:

'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'

Untested BTW. I've made 78-034-89 invalid, as a number should not start
with zero, unless it's only 1 zero. 't Would be simpler if it was valid,
your choice.

works like a charm, can you explain the simple way without check zero
at start?

'/^ # match start of string
[0-9]{1,3} # match 1 to 3 digits
( # start subpattern
- # literal -
[0-9]{1,3} # match 1 to 3 digits
)* # match subpattern zero or more times
$ # match has to run all the way up untill the end of the string
/x'

When starting with regexes, and certainly for the more complex ones, don't
underestimate the power of /x: comments in the regex itself, indentation
to keep things clear, it does wonders. It also keeps you from
rediscovering how the hell your regex worked 6 months after you wrote it
when you have to track a bug.....
--
Rik Wasmus

Thanks !!! simple and beatifull

[Steve]

<da*******@gmail.comwrote in message
news:11**********************@q5g2000prf.googlegro ups.com...

On 17 oct, 21:10, "Rik Wasmus" <[e-mail]:

>On Wed, 17 Oct 2007 22:50:59 +0200, <[e-mail]:

On 17 oct, 20:20, "Rik Wasmus" <[e-mail]:
On Wed, 17 Oct 2007 21:59:49 +0200, <[e-mail]:
Hello

I need a regular expression that validate a list of numbers
separated
by "-" , numbers can not be greater than 999

Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

>Hmmz, showing a 'best try' would be appreciated. However, I normally
can't
resist a good regex, so here it goes:

>'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'

>Untested BTW. I've made 78-034-89 invalid, as a number should not
start
with zero, unless it's only 1 zero. 't Would be simpler if it was
valid,
your choice.

works like a charm, can you explain the simple way without check zero
at start?

'/^ # match start of string
[0-9]{1,3} # match 1 to 3 digits
( # start subpattern
- # literal -
[0-9]{1,3} # match 1 to 3 digits
)* # match subpattern zero or more times
$ # match has to run all the way up untill the end
of the string
/x'

When starting with regexes, and certainly for the more complex ones,
don't
underestimate the power of /x: comments in the regex itself, indentation
to keep things clear, it does wonders. It also keeps you from
rediscovering how the hell your regex worked 6 months after you wrote it
when you have to track a bug.....
--
Rik Wasmus

Thanks !!! simple and beatifull

first, very good rik.

second, dave...if you're going to cross post the same question, at least
have the courtesy to go back and say, 'i found an answer, thanks for your
time.' shortly after your posting to alt.php, you had two responses...one of
which (not mine, btw), is much more simply expressed than rik's. either way,
be courteous to both of the groups you posted to...that way, we'll keep
responding to your questions rather than thinking that someone somewhere
else is responding.

[davranfor]

On 18 oct, 05:40, "Steve" <[e-mail]:

<[e-mail] in message

news:11**********************@q5g2000prf.googlegro ups.com...

On 17 oct, 21:10, "Rik Wasmus" <[e-mail]:

On Wed, 17 Oct 2007 22:50:59 +0200, <[e-mail]:
On 17 oct, 20:20, "Rik Wasmus" <[e-mail]:
On Wed, 17 Oct 2007 21:59:49 +0200, <[e-mail]:
Hello

I need a regular expression that validate a list of numbers
separated
by "-" , numbers can not be greater than 999

Valid examples
0
12-455-01
1-9
125-32-155-45-45

Invalid examples
-1
45-
1-45665456-4
12-45-
-

Hmmz, showing a 'best try' would be appreciated. However, I normally
can't
resist a good regex, so here it goes:

'/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/'

Untested BTW. I've made 78-034-89 invalid, as a number should not
start
with zero, unless it's only 1 zero. 't Would be simpler if it was
valid,
your choice.

works like a charm, can you explain the simple way without check zero
at start?

'/^ # match start of string
[0-9]{1,3} # match 1 to 3 digits
( # start subpattern
- # literal -
[0-9]{1,3} # match 1 to 3 digits
)* # match subpattern zero or more times
$ # match has to run all the way up untill the end
of the string
/x'

When starting with regexes, and certainly for the more complex ones,
don't
underestimate the power of /x: comments in the regex itself, indentation
to keep things clear, it does wonders. It also keeps you from
rediscovering how the hell your regex worked 6 months after you wrote it
when you have to track a bug.....
--
Rik Wasmus

Thanks !!! simple and beatifull

first, very good rik.

second, dave...if you're going to cross post the same question, at least
have the courtesy to go back and say, 'i found an answer, thanks for your
time.' shortly after your posting to alt.php, you had two responses...one of
which (not mine, btw), is much more simply expressed than rik's. either way,
be courteous to both of the groups you posted to...that way, we'll keep
responding to your questions rather than thinking that someone somewhere
else is responding.

ok steve, excuse my poor english: I've post in 2 groups because the
question doesn't appear in the first group (comp.lang.javascript), I
don't like cross post, but you're right

A last question Rik :))), I was looking in http://regexlib.com a way
to do that and without an answer:

Same but numbers can't be lesser than the previous number, must be
equal or greater, is this possible with regex??

0-45-90-120 -Valid
0-45-45-90 -Valid
0-90-45-120 -Invalid

thanks

[davranfor]

>you had two responses...one of

>which (not mine, btw), is much more simply expressed than rik's

Steve , not much more simply expressed

"1-9" and "\d" are identical ;)

[Captain Paralytic]

On 18 Oct, 04:40, "Steve" <[e-mail]:

second, dave...if you're going to cross post the same question, at least
have the courtesy to go back and say, 'i found an answer, thanks for your
time.' shortly after your posting to alt.php, you had two responses...one of
which (not mine, btw), is much more simply expressed than rik's. either way,
be courteous to both of the groups you posted to...that way, we'll keep
responding to your questions rather than thinking that someone somewhere
else is responding.

Actually Steve, if he had "cross-posted" ti wouldn't have been a
problem as both groups would have seen the responses.

So Dave, in future if you want to ask in 2 groups, do cross-post.

What you did in this case was multi-post.

See http://www.blakjak.demon.co.uk/mul_crss.htm

[Rik Wasmus]

On Thu, 18 Oct 2007 10:42:20 +0200, <da*******@gmail.comwrote:

Same but numbers can't be lesser than the previous number, must be
equal or greater, is this possible with regex??

0-45-90-120 -Valid
0-45-45-90 -Valid
0-90-45-120 -Invalid

Maybe it's possible. I wouldn't do it with a regex anymore though.

function _my_validate($string){
$array = explode('-',$string);
if(!is_array($array)) return false;
$previous = 0;
foreach($array as $value){
if(
intval($value)!= $value
||
$value < $previous
||
$value 999
) return false;
$previous = $value;
}
return true;
}
--
Rik Wasmus

[davranfor]

On 18 oct, 13:44, "Rik Wasmus" <[e-mail]:

On Thu, 18 Oct 2007 10:42:20 +0200, <[e-mail]:

Same but numbers can't be lesser than the previous number, must be
equal or greater, is this possible with regex??

0-45-90-120 -Valid
0-45-45-90 -Valid
0-90-45-120 -Invalid

Maybe it's possible. I wouldn't do it with a regex anymore though.

function _my_validate($string){
$array = explode('-',$string);
if(!is_array($array)) return false;
$previous = 0;
foreach($array as $value){
if(
intval($value)!= $value
||
$value < $previous
||
$value 999
) return false;
$previous = $value;
}
return true;

}

--
Rik Wasmus

ok , thanks again

[Steve]

<da*******@gmail.comwrote in message
news:11*********************@v23g2000prn.googlegro ups.com...

you had two responses...one of
which (not mine, btw), is much more simply expressed than rik's

Steve , not much more simply expressed

"1-9" and "\d" are identical ;)

uhmmmm...

/^\d{1,3}(?:-\d{1,3})*$/

is much more simply understood than

/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/

smart ass.

[Steve]

"Captain Paralytic" <pa**********@yahoo.comwrote in message
news:11**********************@e34g2000pro.googlegr oups.com...

On 18 Oct, 04:40, "Steve" <[e-mail]:

>second, dave...if you're going to cross post the same question, at least
have the courtesy to go back and say, 'i found an answer, thanks for your
time.' shortly after your posting to alt.php, you had two responses...one
of
which (not mine, btw), is much more simply expressed than rik's. either
way,
be courteous to both of the groups you posted to...that way, we'll keep
responding to your questions rather than thinking that someone somewhere
else is responding.

Actually Steve, if he had "cross-posted" ti wouldn't have been a
problem as both groups would have seen the responses.

So Dave, in future if you want to ask in 2 groups, do cross-post.

What you did in this case was multi-post.

yeah, yeah...lol

multi-post...i know. for christ's sake, i responded at 4.40 am...whatcha
expect. ;^)

[Rik Wasmus]

On Thu, 18 Oct 2007 15:15:11 +0200, Steve <no****@example.comwrote:

>
<da*******@gmail.comwrote in message
news:11*********************@v23g2000prn.googlegro ups.com...

>you had two responses...one of
which (not mine, btw), is much more simply expressed than rik's

Steve , not much more simply expressed

"1-9" and "\d" are identical ;)

uhmmmm...

/^\d{1,3}(?:-\d{1,3})*$/

is much more simply understood than

/^([1-9][0-9]?)?[0-9](-([1-9][0-9]?)?[0-9])*$/

smart ass.

Because the first hasn't got the 'starting with 0 is prohibited unless
it's only one number'-clause.
The answer without that requirement is nearly identical, and I might say
perfectly readable:
'/^ # match start of string
[0-9]{1,3} # match 1 to 3 digits
( # start subpattern
- # literal -
[0-9]{1,3} # match 1 to 3 digits
)* # match subpattern zero or more times
$ # match has to run all the way up untill the end of the string
/x'

(which in one line would result in: '/^[0-9]{1,3}(-[0-9]{1,3})*$/')

Then again I have always had an unexpainable aversion against \d, don't
know why, I just always use [0-9].
--
Rik Wasmus

[AnrDaemon]

Greetings, da*******@gmail.com.
In reply to Your message dated Wednesday, October 17, 2007, 23:59:49,

dgcI need a regular expression that validate a list of numbers separated
dgcby "-" , numbers can not be greater than 999
dgcValid examples
dgc0
dgc12-455-01
dgc1-9
dgc125-32-155-45-45

dgcInvalid examples
dgc-1
dgc45-
dgc1-45665456-4
dgc12-45-
dgc-

dgcThanks ;)

Perl one:
^\d{1,3}(\-\d{1,3})*$
--
Sincerely Yours, AnrDaemon <an*******@freemail.ru>

[Steve]

"AnrDaemon" <an*******@freemail.ruwrote in message
news:72**********************@freemail.ru...

Greetings, da*******@gmail.com.
In reply to Your message dated Wednesday, October 17, 2007, 23:59:49,

dgcI need a regular expression that validate a list of numbers separated
dgcby "-" , numbers can not be greater than 999
dgcValid examples
dgc0
dgc12-455-01
dgc1-9
dgc125-32-155-45-45

dgcInvalid examples
dgc-1
dgc45-
dgc1-45665456-4
dgc12-45-
dgc-

dgcThanks ;)

Perl one:
^\d{1,3}(\-\d{1,3})*$

wow...thanks for that! it wasn't like that was posted in two newsgroups 4
days ago or anything...in the exact form of:

/^\d{1,3}(?:-\d{1,3})*$/

i suppose i should just wait till someone else answers a post and a couple
of days later, post the exact freaking thing. the redundancy, i think, would
just ensure that the op reeeeealy gets it.

lol.

[AnrDaemon]

Greetings, Steve.
In reply to Your message dated Friday, October 19, 2007, 23:10:56,

>dgcI need a regular expression that validate a list of numbers separated
dgcby "-" , numbers can not be greater than 999
dgcValid examples
dgc0
dgc12-455-01
dgc1-9
dgc125-32-155-45-45

dgcInvalid examples
dgc-1
dgc45-
dgc1-45665456-4
dgc12-45-
dgc-

dgcThanks ;)

Perl one:
^\d{1,3}(\-\d{1,3})*$

Swow...thanks for that! it wasn't like that was posted in two newsgroups 4
Sdays ago or anything...in the exact form of:

S/^\d{1,3}(?:-\d{1,3})*$/

Only difference is post-processing of result produced by this RE.
In my form You'll get set of subpatterns, in the form You referred above
You'll get only fullstring match and no subpatterns.
"(?: ... )" construction means grouping without storing, useful for in-line
multiplication of patterns.

Si suppose i should just wait till someone else answers a post and a couple
Sof days later, post the exact freaking thing. the redundancy, i think, would
Sjust ensure that the op reeeeealy gets it.

:) It was taking me longer than 2 months trying to figure out, how RE works
exactly and can it be used in real life. But after that, I can't live without
it.
--
Sincerely Yours, AnrDaemon <an*******@freemail.ru>

[Steve]

"AnrDaemon" <an*******@freemail.ruwrote in message
news:13**********************@freemail.ru...

Greetings, Steve.
In reply to Your message dated Friday, October 19, 2007, 23:10:56,

>>dgcI need a regular expression that validate a list of numbers
separated
dgcby "-" , numbers can not be greater than 999
dgcValid examples
dgc0
dgc12-455-01
dgc1-9
dgc125-32-155-45-45

dgcInvalid examples
dgc-1
dgc45-
dgc1-45665456-4
dgc12-45-
dgc-

dgcThanks ;)

Perl one:
^\d{1,3}(\-\d{1,3})*$

Swow...thanks for that! it wasn't like that was posted in two newsgroups
4
Sdays ago or anything...in the exact form of:

S/^\d{1,3}(?:-\d{1,3})*$/

Only difference is post-processing of result produced by this RE.
In my form You'll get set of subpatterns, in the form You referred above
You'll get only fullstring match and no subpatterns.
"(?: ... )" construction means grouping without storing, useful for
in-line
multiplication of patterns.

really? i had *no* idea! you mean, you can do that sort of thing? lol.

btw, that would be a great thing to tell the op...not me. further, you
should attempt to demonstrate that you actually know what ?: does. what you
have explained is COMPLETELY false! *grouping* does NOT mean *capturing*.
BOTH patterns GROUP. i've no idea on earth what 'multiplication of patterns'
means in regex...since multiplication is not part of any regex engine of
which i know.

to be clear to the op, ?: means that whatever is inbetween the parenthesis
will not be stored as matched output. and, since there is overhead in
capturing and given that the op ONLY wants to validate, it seems feckless to
quibble about the difference here as if yours provides a better solution. it
is the SAME regex pattern...only, yours will run slower.

Si suppose i should just wait till someone else answers a post and a
couple
Sof days later, post the exact freaking thing. the redundancy, i think,
would
Sjust ensure that the op reeeeealy gets it.

:) It was taking me longer than 2 months trying to figure out, how RE
works
exactly and can it be used in real life. But after that, I can't live
without
it.

and this means what, exactly? that you think posting identical solutions to
a problem advances the learning of the op? that's just silly.