Help needed with a Regular Expression

[MJ]

I'm having problems coming up with a regular expression that works for what
I want. I need to extract a few numbers from from html. Here's a snippet:

<tr><td align="right">C onsolidated Metro Area (CMSA)</td><td
align="left"><s trong> SEATTLE-TACOMA-BREMERTON, WA
( 7602 )</td></tr>
<tr><td align="right">P rimary Metro Area (PMSA)</td><td
align="left"><s trong>SEATTLE-BELLEVUE-EVERETT, WA
( 7600 )</td></tr>

What I am trying to do is extract the first 4 digit number that occurs after
the letters (CMSA) and the first 4 digit number that occurs after (PMSA).

In Dreamweaver, I am able to use the following reg ex to search and it works
fine:

\(CMSA\).*\d{4}

When I try to use the same reg ex in the code on a page it fails to make a
match:

$getCMSA = eregi("\(CMSA\) .*\(\s*\d{4}",$ result,$CMSA_va lue);

I want the reg ex to be as flexible as can be with the code (the least
amount of literal character searches) it is searching so if the site the
values are being pulled from ever changes their code slightly, there will be
a higher chance of the reg ex still working.

Can anybody help me out here? It would be GREATLY appreciated. Thanks!

[Doug Hutcheson]

Re: Help needed with a Regular Expression

A regexp along these lines does what you want - conversion to eregi or other
syntax as required I leave as an exercise for the reader:
(?:[C|P]MSA.*?)(\d{4})
Cheers,
Doug

--
Remove the blots from my address to reply
"MJ" <no_spam@thank. you> wrote in message
news:JYxvc.157$ Q03.171859@news .uswest.net...[color=blue]
> I'm having problems coming up with a regular expression that works for[/color]
what[color=blue]
> I want. I need to extract a few numbers from from html. Here's a[/color]
snippet:[color=blue]
>
> <tr><td align="right">C onsolidated Metro Area (CMSA)</td><td
> align="left"><s trong> SEATTLE-TACOMA-BREMERTON, WA
> ( 7602 )</td></tr>
> <tr><td align="right">P rimary Metro Area (PMSA)</td><td
> align="left"><s trong>SEATTLE-BELLEVUE-EVERETT, WA
> ( 7600 )</td></tr>
>
> What I am trying to do is extract the first 4 digit number that occurs[/color]
after[color=blue]
> the letters (CMSA) and the first 4 digit number that occurs after (PMSA).
>
> In Dreamweaver, I am able to use the following reg ex to search and it[/color]
works[color=blue]
> fine:
>
> \(CMSA\).*\d{4}
>
> When I try to use the same reg ex in the code on a page it fails to make a
> match:
>
> $getCMSA = eregi("\(CMSA\) .*\(\s*\d{4}",$ result,$CMSA_va lue);
>
> I want the reg ex to be as flexible as can be with the code (the least
> amount of literal character searches) it is searching so if the site the
> values are being pulled from ever changes their code slightly, there will[/color]
be[color=blue]
> a higher chance of the reg ex still working.
>
> Can anybody help me out here? It would be GREATLY appreciated. Thanks!
>
>[/color]

[Doug Hutcheson]

Re: Help needed with a Regular Expression

> "MJ" <no_spam@thank. you> wrote in message[color=blue]
> news:JYxvc.157$ Q03.171859@news .uswest.net...[color=green]
> > I'm having problems coming up with a regular expression that works for[/color]
> what[color=green]
> > I want. I need to extract a few numbers from from html. Here's a[/color]
> snippet:[color=green]
> >
> > <tr><td align="right">C onsolidated Metro Area (CMSA)</td><td
> > align="left"><s trong> SEATTLE-TACOMA-BREMERTON, WA
> > ( 7602 )</td></tr>
> > <tr><td align="right">P rimary Metro Area (PMSA)</td><td
> > align="left"><s trong>SEATTLE-BELLEVUE-EVERETT, WA
> > ( 7600 )</td></tr>
> >
> > What I am trying to do is extract the first 4 digit number that occurs[/color]
> after[color=green]
> > the letters (CMSA) and the first 4 digit number that occurs after[/color][/color]
(PMSA).[color=blue][color=green]
> >
> > In Dreamweaver, I am able to use the following reg ex to search and it[/color]
> works[color=green]
> > fine:
> >
> > \(CMSA\).*\d{4}
> >
> > When I try to use the same reg ex in the code on a page it fails to make[/color][/color]
a[color=blue][color=green]
> > match:
> >
> > $getCMSA = eregi("\(CMSA\) .*\(\s*\d{4}",$ result,$CMSA_va lue);
> >
> > I want the reg ex to be as flexible as can be with the code (the least
> > amount of literal character searches) it is searching so if the site the
> > values are being pulled from ever changes their code slightly, there[/color][/color]
will[color=blue]
> be[color=green]
> > a higher chance of the reg ex still working.
> >
> > Can anybody help me out here? It would be GREATLY appreciated. Thanks!
> >
> >[/color]
>
>[/color]
"Doug Hutcheson" <doug.blot.hutc heson@nrm.blot. qld.blot.gov.bl ot.au> wrote
in message news:5Jyvc.21$E P1.1485@news.op tus.net.au...[color=blue]
> A regexp along these lines does what you want - conversion to eregi or[/color]
other[color=blue]
> syntax as required I leave as an exercise for the reader:
> (?:[C|P]MSA.*?)(\d{4})
> Cheers,
> Doug
>
> --
> Remove the blots from my address to reply[/color]


Damn!
Using brain-dead M$ tools for posting and I forgot to cut 'n paste top post
to bottom post. Hereby fixed!
Apologies, all
"8-\

--
Remove the blots from my address to reply

[John Dunlop]

Re: Help needed with a Regular Expression

Doug Hutcheson wrote:
[color=blue]
> [MJ wrote:]
>[color=green]
> > What I am trying to do is extract the first 4 digit number that occurs
> > after the letters (CMSA) and the first 4 digit number that occurs after
> > (PMSA).[/color][/color]

[ ... ]
[color=blue]
> A regexp along these lines does what you want - conversion to eregi or other
> syntax as required I leave as an exercise for the reader:
> (?:[C|P]MSA.*?)(\d{4})[/color]

Yes, you'd need a non-greedy match. But I reckon the first problem
was the dot metacharacter not matching newlines. By default, dot
metacharacters don't match newlines in PCREs *or* POSIX regular
expressions. If the numbers and letters weren't on the same line,
there could be no match.

Also, to match the first four-digit number, the characters after a
four-digit sequence must be checked to make sure they aren't digits.

So, assuming parentheses surround the letters, here's one PCRE
possibility (untested):

`\([CP]MSA\).*?\d{4}(? !\d)`s

--
Jock

[OhBoy!]

Re: Help needed with a Regular Expression

That did the trick. Thank you VERY much!


"Doug Hutcheson" <doug.blot.hutc heson@nrm.blot. qld.blot.gov.bl ot.au> wrote
in message news:5Jyvc.21$E P1.1485@news.op tus.net.au...[color=blue]
> A regexp along these lines does what you want - conversion to eregi or[/color]
other[color=blue]
> syntax as required I leave as an exercise for the reader:
> (?:[C|P]MSA.*?)(\d{4})
> Cheers,
> Doug
>[/color]

[Doug Hutcheson]

Re: Help needed with a Regular Expression

"John Dunlop" <usenet+2004@jo hn.dunlop.name> wrote in message
news:MPG.1b2968 466f917f4f98973 4@News.Individu al.NET...[color=blue]
> Doug Hutcheson wrote:
>[color=green]
> > [MJ wrote:]
> >[color=darkred]
> > > What I am trying to do is extract the first 4 digit number that occurs
> > > after the letters (CMSA) and the first 4 digit number that occurs[/color][/color][/color]
after[color=blue][color=green][color=darkred]
> > > (PMSA).[/color][/color]
>
> [ ... ]
>[color=green]
> > A regexp along these lines does what you want - conversion to eregi or[/color][/color]
other[color=blue][color=green]
> > syntax as required I leave as an exercise for the reader:
> > (?:[C|P]MSA.*?)(\d{4})[/color]
>
> Yes, you'd need a non-greedy match. But I reckon the first problem
> was the dot metacharacter not matching newlines. By default, dot
> metacharacters don't match newlines in PCREs *or* POSIX regular
> expressions. If the numbers and letters weren't on the same line,
> there could be no match.
>
> Also, to match the first four-digit number, the characters after a
> four-digit sequence must be checked to make sure they aren't digits.
>
> So, assuming parentheses surround the letters, here's one PCRE
> possibility (untested):
>
> `\([CP]MSA\).*?\d{4}(? !\d)`s
>
> --
> Jock[/color]

Jock,
Two good observations - thanks.
This might solve those issues:
(?:[C|P]MSA)[\S|\s]*?(\d{4}\D)
Cheers,
Doug

--
Remove the blots from my address to reply

[Doug Hutcheson]

Re: Help needed with a Regular Expression

Dang it!
That picks up the following non-digit - I'm an idjit!
Try this:
(?:[C|P]MSA)[\S|\s]*?(\d{4})[\D]*
Sigh...oldtimer s disease...
"8-\
Doug

--
Remove the blots from my address to reply
"Doug Hutcheson" <doug.blot.hutc heson@nrm.blot. qld.blot.gov.bl ot.au> wrote
in message news:H1Ovc.34$E P1.2344@news.op tus.net.au...[color=blue]
> "John Dunlop" <usenet+2004@jo hn.dunlop.name> wrote in message
> news:MPG.1b2968 466f917f4f98973 4@News.Individu al.NET...[color=green]
> > Doug Hutcheson wrote:
> >[color=darkred]
> > > [MJ wrote:]
> > >
> > > > What I am trying to do is extract the first 4 digit number that[/color][/color][/color]
occurs[color=blue][color=green][color=darkred]
> > > > after the letters (CMSA) and the first 4 digit number that occurs[/color][/color]
> after[color=green][color=darkred]
> > > > (PMSA).[/color]
> >
> > [ ... ]
> >[color=darkred]
> > > A regexp along these lines does what you want - conversion to eregi or[/color][/color]
> other[color=green][color=darkred]
> > > syntax as required I leave as an exercise for the reader:
> > > (?:[C|P]MSA.*?)(\d{4})[/color]
> >
> > Yes, you'd need a non-greedy match. But I reckon the first problem
> > was the dot metacharacter not matching newlines. By default, dot
> > metacharacters don't match newlines in PCREs *or* POSIX regular
> > expressions. If the numbers and letters weren't on the same line,
> > there could be no match.
> >
> > Also, to match the first four-digit number, the characters after a
> > four-digit sequence must be checked to make sure they aren't digits.
> >
> > So, assuming parentheses surround the letters, here's one PCRE
> > possibility (untested):
> >
> > `\([CP]MSA\).*?\d{4}(? !\d)`s
> >
> > --
> > Jock[/color]
>
> Jock,
> Two good observations - thanks.
> This might solve those issues:
> (?:[C|P]MSA)[\S|\s]*?(\d{4}\D)
> Cheers,
> Doug
>
> --
> Remove the blots from my address to reply
>
>[/color]

[Doug Hutcheson]

Re: Help needed with a Regular Expression

"Doug Hutcheson" <doug.blot.hutc heson@nrm.blot. qld.blot.gov.bl ot.au> wrote
in message news:0dOvc.36$E P1.967@news.opt us.net.au...[color=blue]
> "Doug Hutcheson" <doug.blot.hutc heson@nrm.blot. qld.blot.gov.bl ot.au> wrote
> in message news:H1Ovc.34$E P1.2344@news.op tus.net.au...[color=green]
> > "John Dunlop" <usenet+2004@jo hn.dunlop.name> wrote in message
> > news:MPG.1b2968 466f917f4f98973 4@News.Individu al.NET...[color=darkred]
> > > Doug Hutcheson wrote:
> > >
> > > > [MJ wrote:]
> > > >
> > > > > What I am trying to do is extract the first 4 digit number that[/color][/color]
> occurs[color=green][color=darkred]
> > > > > after the letters (CMSA) and the first 4 digit number that occurs[/color]
> > after[color=darkred]
> > > > > (PMSA).
> > >
> > > [ ... ]
> > >
> > > > A regexp along these lines does what you want - conversion to eregi[/color][/color][/color]
or[color=blue][color=green]
> > other[color=darkred]
> > > > syntax as required I leave as an exercise for the reader:
> > > > (?:[C|P]MSA.*?)(\d{4})
> > >
> > > Yes, you'd need a non-greedy match. But I reckon the first problem
> > > was the dot metacharacter not matching newlines. By default, dot
> > > metacharacters don't match newlines in PCREs *or* POSIX regular
> > > expressions. If the numbers and letters weren't on the same line,
> > > there could be no match.
> > >
> > > Also, to match the first four-digit number, the characters after a
> > > four-digit sequence must be checked to make sure they aren't digits.
> > >
> > > So, assuming parentheses surround the letters, here's one PCRE
> > > possibility (untested):
> > >
> > > `\([CP]MSA\).*?\d{4}(? !\d)`s
> > >
> > > --
> > > Jock[/color]
> >
> > Jock,
> > Two good observations - thanks.
> > This might solve those issues:
> > (?:[C|P]MSA)[\S|\s]*?(\d{4}\D)
> > Cheers,
> > Doug
> >
> > --
> > Remove the blots from my address to reply
> >
> >[/color]
>
> Dang it!
> That picks up the following non-digit - I'm an idjit!
> Try this:
> (?:[C|P]MSA)[\S|\s]*?(\d{4})[\D]*
> Sigh...oldtimer s disease...
> "8-\
> Doug
>
> --
> Remove the blots from my address to reply
>[/color]

Double dang it!
M$ Outlook Express does not allow me to bottom post by default.
Forgot to copy and paste from top to bottom in my last message - sorry,
folks.
Fixed here.
<beats head against keyboard: "Why won't they let me use a decent
newsreader?" />
Doug

--
Remove the blots from my address to reply