History

To clear history:

rm ~/.bash_history
./history -c

Loops

Add a .class extension to all files in the current directory:

for file in `ls -1 .`; do mv $file $file.class; done 

Renaming a range of files:

for i in {2..7}; do mv Screenshot000$i.jpg blah-$i.jpg; done

Smarter renaming:

for i in `ls -1`; do mv "$i" "${i/mar2014/2014-03}"; done

Rename a list of files:

for i in 1070 1227 1157; do echo "./pictures/IMG_$i.jpg"; done

Copying a list of files via scp:

cat ./filenames| while read p; do scp user@host:./path/$p . ; done

Text formatting

Format a text with max 80 columns width:

fmt -w 80 input.txt > output.txt

To columnate lists:

column -s ';' -t

Tar

To untar in a given directory:

tar my.tar -C dir

Zip/Unzip

When unzip fails, try 7za:

7za e thezip.7z

To unzip without the arborescence:

unzip -j thezip

Sed

Removes last character of each line:

sed 's/\(.*\)./\1/'

Remove the firsst 5 characters of each line:

sed -e 's/^.....\(.*\)$/\1/g' 

Deletes all lines starting with #

sed '/^#/ d' 

Delete first line:

sed '1d'

Like dos2unix:

sed -i 's/\r//' file

To display line number n:

sed 'NUMq;d' file

Cut

echo blah-date.eps.png |cut -d\. -f1
blah-date


Example: renaming all .eps.png to .eps:
for i in `ls -1 *.eps.png`; do mv $i `echo $i | cut -d\. -f1`.png; done


echo

Echo normally outputs a \n. To remove the trailing newline use -n:

echo -n "toto" | sha1sum


find

Print all directories:
find . -type d -exec echo {}\; -print


Find a keyword in a given directory:
find DIR | xargs grep KEYWORD


Count lines in a directory:
wc -l `find . -type f`


Remove empty files:
find -maxdepth 1 -type f -empty -print0|xargs -0 rm -f => remove empty files


Find duplicates:
find . -type f -exec sha1sum "{}" \; | sort | uniq -d -w40


Find file modified less than 10 minutes ago:
find . -mmin -10


Find file older than 365 days:
find . -mtime +365 ...


Archive files less old than one week:
find / -mtime 7 | xargs tar -zcf weekly.tgz


grep

grep -o -E "\".*\"" --color=always file.txt


This command will:

-o: print only matching lines of file.txt
-E: specifies a regular expression. In this case, we are going to match any
text between quotes. In the regular expression, the quotes must be escaped.
--color: put colour to the matching regular expression. This is much easier to read


curl

Use -F to POST information.
curl -F `username=blah` -A 'USER AGENT` http://MY-URL


awk

To get specific columns of an output line:
ls -l | awk '{print $6" "$7" " $8}'


To remove duplicate lines in a file:
awk '!_[$0]++' file


Find non empty files in a tar:
tar tfvj blah.tar.bz2 | awk '$3 > 0 {print $3 " " $6}'


To perform a sum:
... | awk '{ sum += $1 } END { print sum }'


Sort

Numeric sorting:
sort -nk 2


Comm

Compares to files line by line:

comm -23 a.sorted b.sorted > diff


ex: where -2 suppresses column 2, -3 suppresses column 3.

BC

Convert an hexadecimal number:
echo "ibase=16; 5F" | bc
95


ps

ps -o pid,ppid,comm,start,class,fuser -p 10617,10615


Converting...

Converting a text to pdf:
asciidoctor -b pdf mytext.txt