Small heaps of code

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Bash Snippets ======

This page contains a collection of useful shell / bash scripting snippets. For a general introduction see [[linux:bash|this page]].

===== Date / Time =====

==== Convert UNIX timestamps to human-readable date ====
<code bash>
date -d @1278923870 # Mon Jul 12 10:37:50 CEST 2010
</code>

==== Date loop ====
<code>
 for i in `seq -w 1 31`; do ./myscript 2007-08-$i; done
</code>
You can use inner loops for month and year of course.

===== File Juggling =====

==== Recursively count number of files in directory ====
<code bash>
find DIRECTORY -type f -print | wc -l
</code>

==== Parse a Logfile ====
<code bash>

  cat $1 | while read line; do
     sectionId=`echo $line | grep -o "[0-9]\{4\}-[0-9]\{4\}"`
     statusString=`echo $line | grep -o  "auf .*$"`

       case $statusString in
        auf\ FREI )
        STATUS=1;;
        auf\ DICHT )
        STATUS=2;;
        auf\ STOCKEND )
        STATUS=3;;
        auf\ STAU )
        STATUS=4;;
        auf\ NICHT\ VERFÜGBAR )
        STATUS=5;;
     esac

     NEWLINE="${line};${sectionId};${STATUS}"
     echo $NEWLINE >> out.txt
  done
</code>

==== Make a tar'ed archive from selected files ====
<code bash>
# starting from current directory, recursively tar and compress everything NOT ending with .csv
tar cfvz  output.tar.gz * --exclude "*.csv" 
</code>

==== Remove empty lines from text ====
<code>
 grep -v '^$' input.txt > output.txt
</code>
Or

<code>
 sed -i '/^$/d' input.txt
</code>

===== Remove newlines from text =====
<code>
 tr "\n" " " input.txt
</code>

==== Renaming Stuff ====
Use the "rename" program.

<code bash>
rename 's/\.bak$//' *.bak # removes the .bak suffix from all files matching *.bak
</code>

=== Change suffix for multiple files in a directory ===
*.hh -> *.h

<code bash>
rename 's/\.hh/.h/' *.hh
</code>

=== Change name for multiple files in a directory ===
XML taxi files from heedfled have syntax of webservices as name:

  * First step:
<code bash>
     rename -n s/.*Start=// FCD*
</code>

  * Second step:
<code bash>
      rename -n s/T.*// 2007*
</code>
NOTE: remove -n (=no-act) to run the rename

=== Dry run ===
Run your rename with -n to see which files would be renamed.


===== Finding Stuff =====

==== Find a file ====
<code bash>
  find / -name <fieename>
</code>

==== Find files with time restriction ====
Find files older than 10 minutes.
<code bash>
  find /path/to/dir -mmin +10
</code>


Find files younger than 10 minutes
<code bash>
  find /path/to/dir -mmin -10
</code>


Other options
  * -ctime 
  * -mtime
  * -atime
  * .. (see man page)

==== Delete Files Older than 15 days ====
<code bash>
  find /path/to/dir -mtime +15 -delete
</code>

==== Execute a command on every found file ====
<code bash>
  find / -name muh.txt  -exec cat '{}' \;
</code>

Multiple -exec switches can be used to execute more commands on one file: 
<code bash>
  find / -name muh.txt -exec echo '{}' \;  -exec cat '{}' \;
</code>




===== EXIV =====

=== Adjust EXIV information ===

<code bash>
# adjust date/time of exiv tag
exiv2 image.jpg              # print exiv data
exiv2 -a 0:42 image.jpg      # add 42 minutes to image timestamp
exiv2 -a -1:10 image.jpg     # subtract 1 hour 10 minutes from image timestamp
</code>

=== Rename file according to EXIV timestamp ===

<code bash>
renrot -n %Y-%m-%d_%H-%M-%S image.jpg
</code>

===== GIS =====

=== Merge shapefiles ===

<code bash>
ogr2ogr merged.shp file1.shp                                    # create a copy of file1
ogr2ogr -update -append merged.shp file2.shp -nln merge         # append file2
</code>

===== PDF =====

=== Shrink pdf files ===

<code bash>
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen -dNOPAUSE -dQUIET -dBATCH -sOutputFile=output.pdf input.pdf
</code>
Small heaps of code

User Tools

Site Tools

Page Tools
