Post your Command line tricks

[wuxmedia]

^^ shouldn't that be 'pstree' ?

[wuxmedia]

bored, how many files on my system?

Code: Select all

find / | sort | uniq | wc -l

[machinebacon]

Search for bloat (lol):

Code: Select all

find / -size +20000k

Search for the pattern "grill" within files inside of /usr/local/

Code: Select all

grep -nr "grill" /usr/local/

[johnraff]

^^ shouldn't that be 'pstree' ?

YES!!

Sorry bout that - a beer too many?
Thanks for catching it.

[wuxmedia]

heh, I thought it was a new one, ran it, nope - apt-cache searched it, nope... Ahhh ok...

[dkeg]

^a^b
take previous commnad and replace a with b.

[kiiroitori]

They are not really tricks but I find this useful and it's the best I can do, being a regular noob kind-a-guy (with a beard though).

This came after catfish crashed on me and I was like "what the hell I just want to find a damn file stupid GUI software crashing on me damn how hard can this be man it should be simple damnit" (note that when I swear I don't use punctuation).

So here is some find goodness:

Code: Select all

find . -name *ass.jpg -newermt 2013-10-01 ! -newermt 2013-11-01

will find all files that end with "ass.jpg" (e.g. biomass.jpg, compass.jpg, etc) created in October of this year.

Feel like wiping all asses pictures from your drive? Be careful and use this at your own risks (keep in mind that I am still a noob despite my beard):

Code: Select all

find . -name \*ass.jpg -type f -delete

"type -f" shouldn't be necessary but will prevent from deleting directories, it feels safer to me.

[machinebacon]

^ or you can pipe the output into a file, review it, and then cat the file into rm :)

[kiiroitori]

^ Excellent Mr Bacon, piping is good!

[machinebacon]

so something like

Code: Select all

find . -name \*ass.jpg > asses.txt && nano asses.txt && xargs -a asses.txt -d'\n' rm

and all the ass is gone. Probably there's a more elegant way. Or, I'm sure there is.

[ivanovnegro]

@Kiiroitiri: See, this is why I do not understand all the search features that come with some file browsers or in KDE.

[machinebacon]

To be honest, only Catfish is a really decent front end for complex searches (you can choose the backend and limit the results), but I find myself using find and grep actually every day, because it 'goes loosely from the hand' :D

[pidsley]

Let's not forget "locate". Fewer options than find, but much faster when you want to search the whole system.

[rhowaldt]

pidsley just made a valid point.
also:

(note that when I swear I don't use punctuation).

lol.

[DebianJoe]

"Hey brah, I just downloaded this totally sweet color theme, but someone put a '*color' at the beginning of all of the lines instead of the "Urxvt.color" that I normally use to make it work. I don't like editing files manually because I'm lazy." Basic sed to the rescue!

Code: Select all

sed -i.bak 's/^\*/URxvt./' .Xresources

Plus, .Xresources.bak is left if you want to renege on the decision.

[johnraff]

@joe you sure that worked? I've just tried it on a random list file (not Xresources) to no effect. However, this put a 'URxvt' at the front of every line (no discrimination):

Code: Select all

sed -i.bak 's/^/URxvt./' list

Thanks for that -i.bak tweak though!

@dkeg that ^a^b trick is amazing!

[dkeg]

small but mighty :D

[DebianJoe]

@ Johnraff

....I accidentally left a ' at the end of the command. I'll fix it. Thanks for the catch. Remove extra single quote, profit.

The regex needs to be tied to the "*" which should be escaped first.

Edit: Sed is my favorite tool on systems that don't have emacs (where I'd normally just open and do a regex search and replace). Let's break this one down for fun, because you're a Bash guy, this is handy to know.

sed -i.bak (defines that we're going to use sed with the -i which edits the original file while making a .bak of it.)
's/ (search and replace, the / is a delimiter, expression starts at the single quote)
^\*/ (like a bash regex, the ^ anchors us to the start of the line, and we escape the * so that it's literal, and not a glob. Then we add another delimiter to let sed know we're done with what we're looking for.)
URxvt./' (the string to replace the occurance with. If we find a * at the beginning of the line, we sub this in for it. One more delimiter. Then we close the command with the single quote)
.Xresources (the file that we wish to perform this function on.)

The extra quote at the end meant that it would have only performed this operation on a file named .Xresources' which you probably don't have either. Sorry for the mistake.

[johnraff]

^ah got it. Thanks. It looks as if there might not have been anything wrong with your original code anyway. I didn't see the * (blame it on the font+my eyes) in front of color in " ...but someone put a '*color' at the beginning of..." so of course as my test file had no asterisks it didn't do anything...

[DebianJoe]

New useful tool, I have it as a function in my .zshrc, but should work in bash as well.

Code: Select all

internetquery() {
    ping -q -w 1 -c 1 $(ip r | grep default | cut -d ' ' -f 3) \
               > /dev/null && echo "Internet up" || echo "No Internet"
}

Let's me know if it's my connection or not by pinging default gateway and then telling me if internet is up, or not.