Batman gonna do some text processing on this guy.
#awk

I wrote some #awk to parse output from `show interface` on #Cisco IOS devices this evening as one does.

It expands the ~27 lines of dense text with multiple values on each line to be 68 discrete data points that can be worked with.

Awk: Hack the planet['s text]! (Presentation) - 2023 Update

https://youtube.com/watch?v=E5aQxIdjT0M&si=Cz2fpA_qnFh9F9eN

Looking for stats? Pete Metcalfe shows you how to quickly analyze and plot your data with just one line of Bash and tools like AWK and gnuplot
https://www.linux-magazine.com/Issues/2025/294/Bash-Stats?utm_source=mlm
#Bash #tools #AWK #gnuplot #Linux #stats #data #SQL

Hey xdg-shell-linux fedi corner

I need some eyes on this code here: https://gitlab.freedesktop.org/xdg/xdg-utils/-/merge_requests/147

Hey #entr fans: are there any examples of a `$ENTR_STATUS_SCRIPT` script out in the wild, or any explanation for what this feature was meant to do? It seems like it could be useful, and I'm a huge fan of #awk, but I can't find anyone explaining what they do with it.

I created some code to reformat the output of ip --brief to look a bit cleaner (IMO).

You can find the code in https://github.com/SebastianMeisel/mybashrc .

TIL that modifying NF in #awk has an instant effect, so if you do

$ echo a b c | awk '{while (NF){print $(NF--)}}'

the post-decrementing removes the entry before the variable gets accessed, resulting in it printing blanks, so instead you have to access the variable before doing the post-decrement:

$ echo a b c | awk '{while (NF){print $NF; NF--}}'

to print each item in reverse.

Which seems weird given what I understand about how post-decrement is *supposed* to work.

(HT: @drscriptt whose #awk sent me down this rabbit-hole of learning)

Well I did a thing.

I created a #DNS zone; .oid, on my DNS server #OIDs.

I can now easily look up OID values with dig (et al.):

% dig +short txt 2.3.7.5.5.1.6.3.1.oid.

To look up OIS 1.3.6.1.5.5.7.3.2.

I also wrote a one* line shell script to make doing the lookups easier:

\dig +short txt $(echo ${1} | awk -F. '{for (C=NF; C>1; C--){printf "%s.", $C}; printf "%s.oid.", $1}') | sed 's/"//g' #awk

So I can now run:

% oidlookup 1.3.6.1.5.5.7.3.2

and get the following output:

{iso(1) identified-organization(3) dod(6) internet(1) security(5) mechanisms(5) pkix(7) kp(3) id-kp-clientAuth(2)}

Enjoyed @kevlin on 'past and future of programming languages' yesterday. I am an outlier, I programmed in three languages outside the top 20 this week - #elixir, #lisp (emacs) and #awk (it comes bundled with most linux, and I have it on mac as well). Strangely enough, a local #llm makes the latter two more approachable. qwen3:30 is quite good at generating awk, and good enough to get almost-working #emacs lisp.

Currently testing @stefano 's #BSSG

Seems that I could, with minor mis features, switch from #Hugo to #BSSG

May be I will need some #awk, #sed and #sh

Need to take notes for a blog post.

Hey beloved #noGooMe users,

A quick reminder that you *must* have the proxy image option unchecked when using this #SearxNG instance. It can be disabled, if you previously had it enabled, browsing to /preferences, then selecting the "privacy" tab and uncheck the "image proxy" switch.

It is disabled by default so any first connection to this instance has the proper setting set up.

If you do have it enabled, you’ll quickly be blocked at the firewall level by our #awk AI agent

My reason for disabling this feature is that it generates loads of requests from the instance IP to the external engines. And this makes those block us really fast.

This means that the engines will have *your IP* logged when you search for images, only. Classical text search are still masqueraded with our instance IP.

I still can't believe that most programming systems we use today are preoccupied with numbers. AFAIK, half of (R5RS?) #Scheme standard is numbers and operations on them. Same for #C, #CommonLisp, #Java—ten different types of numbers and huge libraries for them.

Humans think in images and words. Structured text-oriented languages feel like a much better fit for everyone not corrupted by C. Yet we have little to no popular attempts in that space. Structured Regular Expressions didn't catch up; #ed1 and #awk are considered mere #regex automation tools. Modal and the term rewriting systems have their Merveilles Town, but not much beyond. sh/#bash and the like are quite successful, but aren't considered real programming languages either.

Why.