one liner string extraction

natxo has asked for the wisdom of the Perl Monks concerning the following question:

hi, I need to get a string from a line matching the last occurrence of another string in a log file. I can get the last occurrence of in the log file like this:

perl -ne '$string=$_ if /Successfully sent/;  END{print $string}' $log
+file
[download]

The line something like this:


2017-11-02 12:08:11,120 INFO [Timer-Driven Process Thread-8] o.a.nifi.
+remote.StandardRemoteGroupPort RemoteGroupPort[name=Receiver,targets=
+https://host.sub.domain.tld:9443/nifi] Successfully sent [StandardFlo
+wFileRecord[uuid=xxxxxxxxxxxx,claim=StandardContentClaim [resourceCla
+im=StandardResourceClaim[id=xxx, container=default, section=xxx], off
+set=384244, length=163820],offset=0,name=xxxxx.json,size=163820]] (15
+9.98 KB) to https://host.sub.domain.tld:9443/nifi-api in 242 millisec
+onds at a rate of 660.28 KB/sec
[download]

Now I need to get just the timestamp at the beginning of this line. This is how I thought it should work:

perl -ne '$string=$_ if /Successfully sent/;  $string =~ /^(.*) INFO.*
+$/; END{print $1}' $logfile
[download]

But no success until now. Any tips greatly appreciated. Thanks in advance!

Comment on one liner string extraction Select or Download Code

Replies are listed 'Best First'.
Re: one liner string extraction by 1nickt (Canon) on Nov 02, 2017 at 11:37 UTC
Hi, `$1` is reset on each attempted match, so if a failed match occurs after your last match, `$1` will be unset. `perl -ne '$string=$_ if /Successfully sent/; $match = $1 if $string = +~ /^(.) INFO.$/; END{print $match}' $logfile` [download] Hope this helps! The way forward always starts with a minimal test.	[reply] [d/l] [select]
Re^2: one liner string extraction by vr (Curate) on Nov 02, 2017 at 15:33 UTC
Hm-m, while the above solution works, but by way of nitpicking and because of TIMTOWTDI: $1 is reset on each attempted match `$1` is reset on each successful match, and restored on each scope end. Scope end happens twice (rather, n + 1) in OP: 1st on every loop iteration (-n), second at program finishing but before executing END block(s). `>perl -wE "'a'=~/(a)/; 'a'=~/b/; say $1" a >perl -wE "'a'=~/(a)/; 'a'=~/a/; say $1" Use of uninitialized value $1 in say at -e line 1. >perl -wE "'a'=~/(a)/; END { say $1 }" Use of uninitialized value $1 in say at -e line 1.` [download] But we can abuse the fact that, unlike other global variables related to regular expressions, the $^R is not dynamically scoped. Further, let's match only once, and also have a little fun and save us a few keystrokes by using a secret (END will also do, of course): `>perl -ne "/^(.*) INFO.+Successfully sent(?{$1})/ }{ print $^R" file` [download]	[reply] [d/l] [select]
Re^2: one liner string extraction by natxo (Scribe) on Nov 02, 2017 at 12:00 UTC
of course! Thanks, works like a charm.	[reply]
Re: one liner string extraction by kcott (Archbishop) on Nov 03, 2017 at 13:30 UTC
G'day natxo, Would this be a simpler solution for what you're trying to achieve? `$ perl -nle 'next unless /^(.?) INFO .? SENT/; $x = $1; END { print +$x }' ts1 INFO ... SENT ts2 INFO ... RECV ts3 INFO ... SENT ts4 INFO ... RECV ts3` [download] Skip unless match, otherwise save capture. Print last capture at the end. I used `STDIN` (keyboard) for dummy logfile records: `ts1 INFO ... SENT ts2 INFO ... RECV ts3 INFO ... SENT ts4 INFO ... RECV` [download] Output last matching timestamp: `ts3` — Ken	[reply] [d/l] [select]