Skip grep, use awk • Hacker News Board

Re: Skip grep, use awk

#61

Post by _delirium » Tue, Jul 04, 2017, 9:14 AM UTC

rflrob wrote: post #52

bluetomcat wrote: post #47

AWK is the general-purpose programmatic filter and reporting tool in the Unix pipeline. Sed, grep and cut are specializations for specific use cases whose implementations might have better performance. Perl and Python are probably too general-purpose for writing compact one-liners in a pipeline.

Even though it's more characters, I usually use perl -pe 's/search/replace/' instead of sed in pipelines because it understands /n (and other escape characters I don't remember). Because all it takes is to get burned a couple times for it to be worth sticking with what you know will work.

The Perl version is considerably faster on large files as well. Or at least, it is when compared to the version of BSD sed shipped with OSX.

Re: Skip grep, use awk

#62

Post by gbrown_ » Tue, Jul 04, 2017, 9:15 AM UTC

hiq wrote: post #51

The author mentions that using awk instead of grep -v is not a good idea, what about: awk '! /something/' Doesn't it reproduce the same behaviour as the following? grep -v 'something'

Yup, was going to post this as a comment myself.

Re: Skip grep, use awk

#63

Post by assafmo » Tue, Jul 04, 2017, 9:15 AM UTC

It's a nice idea, but I think maybe only to replace a simple egrep.

grep has tons of great option like

-F = fgrep; no regex - way faster

-v = as mentioned in the article

-o = print only matched input, not the entire line

-C = context, print lines before and after the match; can also be used partially with -A (after) and -B (before)

Re: Skip grep, use awk

#64

Post by dozzie » Tue, Jul 04, 2017, 9:16 AM UTC

thaumaturgy wrote: post #19

emmelaich wrote: post #13

My favourite bad example of using grep was from a big enterprise software vendor to kill one of their processes. It looked something like ps -ef| grep SomeDaemon | grep -v grep | grep -v perl | perl -e ' '

Serious question: what would be a better way to do this?

  do_something_with_pids $(ps -C SomeDaemon ho pid)

Though it's GNU procps.

Re: Skip grep, use awk

#65

Post by gbrown_ » Tue, Jul 04, 2017, 9:17 AM UTC

thinkMOAR wrote: post #59

is there any awk equivalent for grep "matcha\|matchb" ?

awk '/matcha|matchb/'

Re: Skip grep, use awk

#66

Post by dozzie » Tue, Jul 04, 2017, 9:18 AM UTC

rflrob wrote: post #52

bluetomcat wrote: post #47

AWK is the general-purpose programmatic filter and reporting tool in the Unix pipeline. Sed, grep and cut are specializations for specific use cases whose implementations might have better performance. Perl and Python are probably too general-purpose for writing compact one-liners in a pipeline.

Even though it's more characters, I usually use perl -pe 's/search/replace/' instead of sed in pipelines because it understands /n (and other escape characters I don't remember). Because all it takes is to get burned a couple times for it to be worth sticking with what you know will work.

"/n"? What's that? My `man perlre' doesn't say anything about this regexp flag.

Re: Skip grep, use awk

#67

Post by thinkMOAR » Tue, Jul 04, 2017, 9:24 AM UTC

gbrown_ wrote: post #65

thinkMOAR wrote: post #59

is there any awk equivalent for grep "matcha\|matchb" ?

awk '/matcha|matchb/'

excellent :) though slightly disappointed in myself, that i did not try that myself as i did imagine it when i asked...

Re: Skip grep, use awk

#68

Post by kronos29296 » Tue, Jul 04, 2017, 9:33 AM UTC

_delirium wrote: post #61

rflrob wrote: post #52

Earlier quoted context omitted.

Even though it's more characters, I usually use perl -pe 's/search/replace/' instead of sed in pipelines because it understands /n (and other escape characters I don't remember). Because all it takes is to get burned a couple times for it to be worth sticking with what you know will work.

The Perl version is considerably faster on large files as well. Or at least, it is when compared to the version of BSD sed shipped with OSX.

This is probably not the case with the sed in Linux.

Re: Skip grep, use awk

#69

Post by tribby » Tue, Jul 04, 2017, 9:56 AM UTC

dozzie wrote: post #66

rflrob wrote: post #52

Earlier quoted context omitted.

Even though it's more characters, I usually use perl -pe 's/search/replace/' instead of sed in pipelines because it understands /n (and other escape characters I don't remember). Because all it takes is to get burned a couple times for it to be worth sticking with what you know will work.

"/n"? What's that? My `man perlre' doesn't say anything about this regexp flag.

because OP is referring to escape characters I assume it is a typo for "\n"

Re: Skip grep, use awk

#70

Post by jwilk » Tue, Jul 04, 2017, 10:05 AM UTC

> awk uses modern (read “Perl”) regular expressions, by default – like grep -E

No, Perl regular expression are different than extended regular expressions that awk and "grep -E" use.