regexp in Python (from Perl)

[Pat]

I have a regexp in Perl that converts the last digit of an ip address to
'9'. This is a very particular case so I don't want to go off on a
tangent of IP octets.

( my $s = $str ) =~ s/((\d+\.){3})\d+/${1}9/ ;

While I can do this in Python which accomplishes the same thing:

ip = ip[ :-1 ]
ip =+ '9'

I'm more interested, for my own edification in non-trivial cases, in how
one would convert the Perl RE to a Python RE that use groups. I am
somewhat familiar using the group method from the re package but I
wanted to know if there was a one-line solution.

Thank you.

[MRAB]

On Oct 19, 5:47*pm, Bruno Desthuilliers
<bdesth.quelque [e-mail] quepart.frwrote :

Pat a écrit :

I have a regexp in Perl that converts the last digit of an ip address to
*'9'. *This is a very particular case so I don't want to go off on a
tangent of IP octets.

*( my $s = $str ) =~ s/((\d+\.){3})\d+/${1}9/ ;

While I can do this in Python which accomplishes the same thing:

ip = ip[ :-1 ]
ip =+ '9'

or:

ip = ip[:-1]+"9"

I'm more interested, for my own edification in non-trivial cases, in how
one would convert the Perl RE to a Python RE that use groups. *I am
somewhat familiar using the group method from the re package but I
wanted to know if there was a one-line solution.

Is that what you want ?

*>>re.sub(r'^(( (\d+)\.){3})\d+ $', "\g<1>9", "192.168.1. 1")
'192.168.1.9'

>re.sub(r'^(((\ d+)\.){3})\d+$' , "\g<1>9", "192.168.1.100" )

'192.168.1.9'

The regular expression changes the last sequence of digits to
"9" ("192.168.1.100 " ="192.168.1. 9") but the other code replaces the
last digit ("192.168.1.100 " ="192.168.1.109 ").

[bearophileHUGS]

MRAB:

The regular expression changes the last sequence of digits to
"9" ("192.168.1.100 " ="192.168.1. 9") but the other code replaces the
last digit ("192.168.1.100 " ="192.168.1.109 ").

Uhmm, this is a possible alternative:

>>s = " 192.168.1.100 "
".".join(s.st rip().split("." )[:3]) + ".9"

'192.168.1.9'

Bye,
bearophile

[Bruno Desthuilliers]

MRAB a écrit :

On Oct 19, 5:47 pm, Bruno Desthuilliers
<bdesth.quelque [e-mail] quepart.frwrote :

>Pat a écrit :

(snip)

>>ip = ip[ :-1 ]
ip =+ '9'

or:

ip = ip[:-1]+"9"

(snip)

> >>re.sub(r'^((( \d+)\.){3})\d+$ ', "\g<1>9", "192.168.1. 1")

'192.168.1.9 '

>>>>re.sub(r'^( ((\d+)\.){3})\d +$', "\g<1>9", "192.168.1.100" )

'192.168.1.9 '

The regular expression changes the last sequence of digits to
"9" ("192.168.1.100 " ="192.168.1. 9") but the other code replaces the
last digit ("192.168.1.100 " ="192.168.1.109 ").

Mmm - yes, true.

ip = ".".join(ip.spl it('.')[0:3] + ['9'])

[Pat]

Bruno Desthuilliers wrote:

Pat a écrit :

>I have a regexp in Perl that converts the last digit of an ip address
to '9'. This is a very particular case so I don't want to go off on
a tangent of IP octets.

( my $s = $str ) =~ s/((\d+\.){3})\d+/${1}9/ ;

While I can do this in Python which accomplishes the same thing:

ip = ip[ :-1 ]
ip =+ '9'

or:

ip = ip[:-1]+"9"

Yes! That's exactly what I was looking for.

>

>I'm more interested, for my own edification in non-trivial cases, in
how one would convert the Perl RE to a Python RE that use groups. I
am somewhat familiar using the group method from the re package but I
wanted to know if there was a one-line solution.

Is that what you want ?

>>re.sub(r'^((( \d+)\.){3})\d+$ ', "\g<1>9", "192.168.1. 1")

'192.168.1.9'

>>>re.sub(r'^(( (\d+)\.){3})\d+ $', "\g<1>9", "192.168.1.100" )

'192.168.1.9'

Ah-hah! That's how one uses groups. It's beautiful. I couldn't find
that in my books. Thank you very, very much!

At first, I thought that using RE's in Python was going to be more
difficult than Perl. A lot of my Perl code makes heavy use of RE
searching and substitution.

I will never, ever write another line of Perl code as long as I live.

[Bruno Desthuilliers]

Pat a écrit :

Bruno Desthuilliers wrote:

>Pat a écrit :

>>I have a regexp in Perl that converts the last digit of an ip address
to '9'. This is a very particular case so I don't want to go off on
a tangent of IP octets.

( my $s = $str ) =~ s/((\d+\.){3})\d+/${1}9/ ;

While I can do this in Python which accomplishes the same thing:

ip = ip[ :-1 ]
ip =+ '9'

or:

ip = ip[:-1]+"9"

Yes! That's exactly what I was looking for.

Perhaps not - cf MRAB's post earlier in this thread and bearophile's answer.

While this was a straightforward one-liner version of your own code, it
doesn't behave like the re version : this one only replace the last
_character_, while the re version replaces the last _group_. If you want
the re version's behaviour, use this instead:

ip = ".".join(ip.spl it('.')[0:3] + ['9'])

>>

>>I'm more interested, for my own edification in non-trivial cases, in
how one would convert the Perl RE to a Python RE that use groups. I
am somewhat familiar using the group method from the re package but I
wanted to know if there was a one-line solution.

Is that what you want ?

> >>re.sub(r'^((( \d+)\.){3})\d+$ ', "\g<1>9", "192.168.1. 1")

'192.168.1.9 '

>>>>re.sub(r'^( ((\d+)\.){3})\d +$', "\g<1>9", "192.168.1.100" )

'192.168.1.9 '

Ah-hah! That's how one uses groups. It's beautiful. I couldn't find
that in my books.

Did you look at the FineManual(tm) ?-)

>
At first, I thought that using RE's in Python was going to be more
difficult than Perl. A lot of my Perl code makes heavy use of RE
searching and substitution.

Well... Python's re module makes for a bit more verbose code, and I'm
not sure all of the perl's regexps features are implemented. But that's
still quite enough to shoot yourself in the foot IMHO !-)

Now while regexps are a must-have in a programmer's toolbox, you can
already do quite a few things just using slicing and string methods. I
highly recommend you read the relevant doc.

I will never, ever write another line of Perl code as long as I live.

Hmmm... Never say "never again" ?-)

[Pat]

Bruno Desthuilliers wrote:

MRAB a écrit :

>On Oct 19, 5:47 pm, Bruno Desthuilliers
<bdesth.quelqu [e-mail] lquepart.frwrot e:

>>Pat a écrit :

(snip)

>>>ip = ip[ :-1 ]
ip =+ '9'
or:

ip = ip[:-1]+"9"

(snip)

>> >>re.sub(r'^((( \d+)\.){3})\d+$ ', "\g<1>9", "192.168.1. 1")
'192.168.1. 9'

>re.sub(r'^ (((\d+)\.){3})\ d+$', "\g<1>9", "192.168.1.100" )
'192.168.1. 9'

The regular expression changes the last sequence of digits to
"9" ("192.168.1.100 " ="192.168.1. 9") but the other code replaces the
last digit ("192.168.1.100 " ="192.168.1.109 ").

Mmm - yes, true.

ip = ".".join(ip.spl it('.')[0:3] + ['9'])

As I first stated, in my very particular case, I knew that the last
octet was always going to be a single digit.

But I did learn a lot from everyone else's posts for the more generic
cases. thx!