Commands by dopeman (11)

  • The command (above) will remove any duplicate rows based on the FIRST column of data in an un-sorted file. The '$1' represents a positional parameter. You can change both instances of '$1' in the command to remove duplicates based on a different column, for instance, the third: awk '{ if ($3 in stored_lines) x=1; else print; stored_lines[$3]=1 }' infile.txt > outfile.txt Or you can change it to '$0' to base the removal on the whole row: awk '{ if ($0 in stored_lines) x=1; else print; stored_lines[$0]=1 }' infile.txt > outfile.txt ** Note: I wouldn't use this on a MASSIVE file, unless you're RAM-rich ;) **


    4
    awk '{ if ($1 in stored_lines) x=1; else print; stored_lines[$1]=1 }' infile.txt > outfile.txt
    dopeman · 2010-12-15 17:08:47 0
  • buf myfile.txt This is useful when you are making small but frequent changes to a file. It keeps things organised and clear for another administrator to see what changed and at what time. An overview of changes can be deduced using a simple: ls -ltr


    1
    buf () { filename=$1; filetime=$(date +%Y%m%d_%H%M%S); cp ${filename} ${filename}_${filetime}; }
    dopeman · 2010-12-14 13:19:52 3
  • Essentially the same as funky's alias, but will not traverse filesystems and has nicer formatting. Show Sample Output


    -1
    alias dush="du -xsm * | sort -n | awk '{ printf(\"%4s MB ./\",\$1) ; for (i=1;i<=NF;i++) { if (i>1) printf(\"%s \",\$i) } ; printf(\"\n\") }' | tail"
    dopeman · 2010-07-15 10:38:27 1
  • I have come across a situation in the past where someone has unlinked a file by running an 'rm' command against it while it was still being written to by a running process. The problem manifested itself when a 'df' command showed a filesystem at 100%, but this did not match the total value of a 'du -sk *'. When this happens, the process continues to write to the file but you can no longer see the file on the filesystem. Stopping and starting the process will, more often than not, get rid of the unlinked file, however this is not always possible on a live server. When you are in this situation you can use the 'lsof' command above to get the PID of the process that owns the file (in the sample output this is 23521). Run the following command to see a sym-link to the file (marked as deleted): cd /proc/23521/fd && ls -l Truncate the sym-link to regain your disk space: > /proc/23521/fd/3 I should point out that this is pretty brutal and *could* potentially destabilise your system depending on what process the file belongs to that you are truncating. Show Sample Output


    16
    lsof +L1
    dopeman · 2010-07-14 17:21:01 2
  • This is a handy way to circumvent the "Maximum line length of 2048 exceeded" grep error. Once you have run the above command (or put it in your .bashrc), files can be searched using: lgrep search-string /file/to/search


    1
    lgrep() { string=$1; file=$2; awk -v String=${string} '$0 ~ String' ${file}; }
    dopeman · 2010-01-19 09:42:19 0
  • This does the same thing as many of the 'grep' based alternatives but allows a more finite control over the output. For example if you only wanted the process ID you could change the command: ps -ef | awk '/mingetty/ && !/awk/ {print $2}' If you wanted to kill the returned PID's: ps -ef | awk '/mingetty/ && !/awk/ {print $2}' | xargs -i kill {} Show Sample Output


    1
    ps -ef | awk '/process-name/ && !/awk/ {print}'
    dopeman · 2009-08-19 11:22:09 0
  • This command will copy files and directories from a remote machine to the local one. Ensure you are in the local directory you want to populate with the remote files before running the command. To copy a directory and it's contents, you could: ssh user@host "(cd /path/to/a/directory ; tar cvf - ./targetdir)" | tar xvf - This is especially useful on *nix'es that don't have 'scp' installed by default.


    1
    ssh user@host "(cd /path/to/remote/top/dir ; tar cvf - ./*)" | tar xvf -
    dopeman · 2009-03-31 13:08:45 3
  • This command will show the 20 processes using the most CPU time (hungriest at the bottom). You can see the 20 most memory intensive processes (hungriest at the bottom) by running: ps aux | sort +3n | tail -20 Or, run both: echo "CPU:" && ps aux | sort +2n | tail -20 && echo "Memory:" && ps aux | sort +3n | tail -20


    3
    ps aux | sort +2n | tail -20
    dopeman · 2009-03-31 12:03:34 1
  • This command will tell you the 20 biggest directories starting from your working directory and skips directories on other filesystems. Useful for resolving disk space issues.


    7
    du -xk | sort -n | tail -20
    dopeman · 2009-03-30 11:37:43 3
  • This command will replace all instances of 'foo' with 'bar' in all files in the current working directory and any sub-directories.


    -1
    perl -pi -e 's/foo/bar/g' $(grep -rl foo ./*)
    dopeman · 2009-03-27 17:21:35 2
  • This command will replace all instances of 'foo' with 'bar' in all files in the current working directory.


    -1
    perl -pi -e 's/foo/bar/g' $(grep -l foo ./*)
    dopeman · 2009-03-27 17:18:08 1

What's this?

commandlinefu.com is the place to record those command-line gems that you return to again and again. That way others can gain from your CLI wisdom and you from theirs too. All commands can be commented on, discussed and voted up or down.

Share Your Commands


Check These Out

List the popular module namespaces on CPAN
Grabs the complete module list from CPAN, pulls the first column, ditches html lines, counts, ditches small namespaces.

Create an animated gif from a Youtube video
requires "youtube-dl" -- sure you can do this with wget and some more obscurity but why waste your time when this great tool is available? the guts consist of mplayer converting a video to a gif -- study this command and read the man page for more information $ mplayer video.flv -ss 00:23 -endpos 6 -vo gif89a:fps=5:output=output.gif -vf scale=400:300 -nosound generates a 6 second gif starting at 23 seconds of play time at 5 fps and a scale of 400x300 start time (-ss)/end time (-endpos) formats: 00:00:00.000 end time should be relative to start time, not absolute. i.e. -endpos 5 == seconds after 0:42 = 0:47 end point play with fps and scale for lower gif sizes the subshell is a solution for the -b flag on youtube-dl which downloads the best quality video, sometimes, which can be various video formats $(ls ${url##*=}*| tail -n1)

The top ten commands you use

Get list of servers with a specific port open
Change the -p argument for the port number. See "man nmap" for different ways to specify address ranges.

reverse order of file
awk reverse order of lines (emulates "tac")

generate random password (works on Mac OS X)
Prepending env LC_CTYPE=C fixes a problem with bad bytes in /dev/urandom on Mac OS X

Nicely display permissions in octal format with filename
Nicely display permissions in octal format and filename for a given directory

Mouse Tracking
Will track your mouse and save it to a file. You can use gnuplot to graph it: $ gnuplot -persist

Converts uppercase chars in a string to lowercase

Make vim open in tabs by default (save to .profile)
I always add this to my .profile rc so I can do things like: "vim *.c" and the files are opened in tabs.


Stay in the loop…

Follow the Tweets.

Every new command is wrapped in a tweet and posted to Twitter. Following the stream is a great way of staying abreast of the latest commands. For the more discerning, there are Twitter accounts for commands that get a minimum of 3 and 10 votes - that way only the great commands get tweeted.

» http://twitter.com/commandlinefu
» http://twitter.com/commandlinefu3
» http://twitter.com/commandlinefu10

Subscribe to the feeds.

Use your favourite RSS aggregator to stay in touch with the latest commands. There are feeds mirroring the 3 Twitter streams as well as for virtually every other subset (users, tags, functions,…):

Subscribe to the feed for: