"Dan" schreef:
if (substr($line,0,5) eq "From:")
You don't even need to know that 'From:' is 5 characters, if you use
if ( 0 == index $line, 'From:' )
i wrote a program some months back which utilised a compelx regex sub
$onchan{lc($data[0])} =~ s/(,|^)\Q$data[1]\E(?=,|$)//;
which substitutes the exact match for $data[1] in a long string which
is csv, and replace it with nothing. i'm trying to use the same
routine, or the same method, to get the 'dan' out of
"dan" dan[@]domain.com
but a) the regex confuses me enough to not know how to get that out
of there (i've replaced the ,'s with "'s, but that's obviously not
enough), and b) i don't know how to give the regex the full line of
text, and assign the extracted value into a variable for use later on.
Please give better examples of input and expected output. Use
"example.com" in examples.
This contains a regexp that will remove the non-@ characters from the
start of a string:
echo 'dan (AT) example (DOT) com' | perl -pe 's/^[^@]+//'
See also: perldoc -q strip

i been doing perl long enough that you'd think i should know this, but one
thing i've never ever ever managed to get my head around is how regex works.

i'm using net::pop3 (mail::pop3client doesn't work!), and i'm trying to
extract certain data from the pop3 stream (from, subject, and some of the
body eventually). but the regex behind matching the line required is just
baffling me. i could cheat and say
if (substr($line,0,5) eq "From:") { }, but i want to get past the phase of
doing things the such a long way round, and learn more about regex.

i wrote a program some months back which utilised a compelx regex sub
$onchan{lc($data[0])} =~ s/(,|^)\Q$data[1]\E(?=,|$)//;
which substitutes the exact match for $data[1] in a long string which is
csv, and replace it with nothing. i'm trying to use the same routine, or the
same method, to get the 'dan' out of
"dan" dan (AT) domain (DOT) com
but a) the regex confuses me enough to not know how to get that out of there
(i've replaced the ,'s with "'s, but that's obviously not enough), and b) i
don't know how to give the regex the full line of text, and assign the
extracted value into a variable for use later on.

i'd really like to get my head round this. if anyone's got any good guides,
pointers, or places i can go to read and help me learn more, i'd much
appreciate it.

many thanks.

dan

hi

thanks for the pointers. i think the news thing, or something, made the
question a little unclear, as it stripped some characters from the input.
the input is the default way email headers are sent with the "from" and "to"
portions,
"name" <email (AT) example (DOT) com>
the question was pertaining to get 'name' out of "name" at the start, and
the email address out of the <<>brackets. however i managed to figure that
one out by playing.
the other question i have involves searching a string. i'm a little
unfamiliar with this process as well. i'm basically writing a script to
automate the sending out of emails, and it checks the pop3 box every so
often and sends out emails depending on what's sat in there, then deletes
them. in this case it's paypal emails. messages paypal send to say i've
received a payment. this needs to be expanded further than just picking
particular words out of the entire subject line as this may not always be
the case. say for example, the subject said:
item no.00001313131 - notification of an instant payment received from
ebay_user (ebayuser (AT) example (DOT) com)
i can if ($subject =~ /$item_number/) to pick out the item number from that,
but i need to search for the email address and the username in the subject.
i suppose the easiest way to do this would be to search the string for
"(wildcard-here)" and take one off the index of that to find the username in
question (as this is always going to change). is there any easy way to do
this with regex or otherwise?

i hope this makes it a little clearer!

thanks, dan

""Dr.Ruud"" <rvtol+news (AT) isolution (DOT) nlwrote in message
@lists.develooper.com
"Dan" schreef:
>
>if (substr($line,0,5) eq "From:")
>
You don't even need to know that 'From:' is 5 characters, if you use

if ( 0 == index $line, 'From:' )
>
>
>i wrote a program some months back which utilised a compelx regex sub
>>
>$onchan{lc($data[0])} =~ s/(,|^)\Q$data[1]\E(?=,|$)//;
>>
>which substitutes the exact match for $data[1] in a long string which
>is csv, and replace it with nothing. i'm trying to use the same
>routine, or the same method, to get the 'dan' out of
>"dan" dan[@]domain.com
>but a) the regex confuses me enough to not know how to get that out
>of there (i've replaced the ,'s with "'s, but that's obviously not
>enough), and b) i don't know how to give the regex the full line of
>text, and assign the extracted value into a variable for use later on.
>
Please give better examples of input and expected output. Use
"example.com" in examples.
--
This contains a regexp that will remove the non-@ characters from the
start of a string:

echo 'dan (AT) example (DOT) com' | perl -pe 's/^[^@]+//'

See also: perldoc -q strip

5/7/06, Dan <dan (AT) danneh (DOT) orgwrote:
i been doing perl long enough that you'd think i should know this, but one
thing i've never ever ever managed to get my head around is how regex works.

i'm using net::pop3 (mail::pop3client doesn't work!), and i'm trying to
extract certain data from the pop3 stream (from, subject, and some of the
body eventually). but the regex behind matching the line required is just
baffling me. i could cheat and say
if (substr($line,0,5) eq "From:") { }, but i want to get past the phase of
doing things the such a long way round, and learn more about regex.

i wrote a program some months back which utilised a compelx regex sub
$onchan{lc($data[0])} =~ s/(,|^)\Q$data[1]\E(?=,|$)//;
which substitutes the exact match for $data[1] in a long string which is
csv, and replace it with nothing. i'm trying to use the same routine, or the
same method, to get the 'dan' out of
"dan" dan (AT) domain (DOT) com
but a) the regex confuses me enough to not know how to get that out of there
(i've replaced the ,'s with "'s, but that's obviously not enough), and b) i
don't know how to give the regex the full line of text, and assign the
extracted value into a variable for use later on.

i'd really like to get my head round this. if anyone's got any good guides,
pointers, or places i can go to read and help me learn more, i'd much
appreciate it.

many thanks.

dan

Not to clear on what this means:
>to get the 'dan' out of "dan" dan (AT) domain (DOT) com

But here's a stab

my $string = 'dan (AT) domain (DOT) com';
$string =~ s/^(.*?)@//;
This would actualy return "@domain.com";

If you want the username, then:
my ($username) = $string =~ m/^(.*?)@/;

if (substr($line,0,5) eq "From:") { }, but i want to get past the phase of
if ($line =~ m/^From:/i) { do something with "From" line; }

Dan wrote:
i been doing perl long enough that you'd think i should know this, but one
thing i've never ever ever managed to get my head around is how regex works.

i'm using net::pop3 (mail::pop3client doesn't work!), and i'm trying to
extract certain data from the pop3 stream (from, subject, and some of the
body eventually). but the regex behind matching the line required is just
baffling me. i could cheat and say
if (substr($line,0,5) eq "From:") { }, but i want to get past the phase of
doing things the such a long way round, and learn more about regex.

i wrote a program some months back which utilised a compelx regex sub
$onchan{lc($data[0])} =~ s/(,|^)\Q$data[1]\E(?=,|$)//;
which substitutes the exact match for $data[1] in a long string which is
csv, and replace it with nothing. i'm trying to use the same routine, or the
same method, to get the 'dan' out of
"dan" dan (AT) domain (DOT) com
but a) the regex confuses me enough to not know how to get that out of there
(i've replaced the ,'s with "'s, but that's obviously not enough), and b) i
don't know how to give the regex the full line of text, and assign the
extracted value into a variable for use later on.

i'd really like to get my head round this. if anyone's got any good guides,
pointers, or places i can go to read and help me learn more, i'd much
appreciate it.

many thanks.

dan

The perldocs are an excellent source for learning regex:
perlrequick
perlretut
perlre
perlreref

I've also found a nifty little program that will help you test out your
regex to see if it works:
http://weitz.de/regex-coach/

SkyBlueshoes

"Dan" schreef:

Please don't top-post.

"name" <email (AT) example (DOT) com>
the question was pertaining to get 'name' out of "name" at the start,
and the email address out of the <<>brackets. however i managed to
figure that one out by playing.

Don't play but use Regexp::Common::Email::Address

say for example, the subject said:
item no.00001313131 - notification of an instant payment received
from ebay_user (ebayuser (AT) example (DOT) com)

Use something like

if ( /^Subject:\sitem no[.](\d+)\s-\s
notification\sof\san\sinstant\s
payment\sreceived\sfrom\s
(.*?)\s
[(](.+?[@].+?)[)]$/x )
{
print "item:$1, name:$2, address:$3\n";
}

(untested)

If the length of some of the \s (whitespace) can be >1, make those \s+.