assorted sed commands 

Remove spaces from filenames 

removing spaces from a filename:
mv -v "$FILE" `echo $FILE | tr ' ' '_' | tr -d '[{}(),\!]' | tr -d "\'" | tr '[A-Z]' '[a-z]' | sed 's/_-_/_/g'`
mv -v "$FILE" `echo $FILE | tr ' ' '_' | tr -d '[{}(),\!]' | tr -d "\'" |  sed 's/_-_/_/g'`

add year on files less than a year old:
ls -l | grep -v ^d  | sort | awk '{print "mv "$9" "$8"_"$6"_"$7"_"$9}' | sed -e s'$..:..$2014$g' \
~                                                                                                         

Remove Blank Lines from file
     cat file-blanklines-2-be-deleted | sed -e '/^$/d' | tee newfile-no-blanklines

 some "sed-like" tricks from within vi: 
 insert # at beginning of lines 

:%s/.*/# &/g   - inserts a # at the beginning of every line

:15,$s/.*/# &/g   - inserts a # at the beginning of line 15, until the end of the file

:15,250s/.*/# &/g   - inserts a # from line 15 through 250.

 turn one or more blank spaces into one space
	:%s/  */ /g

 remove blank lines, or lines with only spaces or tabs
	:g/^[ t]*$/d  --> WHERE t=tab

 redo entire line adding parentheses: 
  at command line: 	sed s/.*/( & )/  file > new.file
	 	or
         from within vi:	:%s/.*/( & )/g 

 create a list of commands from a directory listing: 

 ls > list
	cat list:
		list
		sed.html
		file.txt
 sed 's/.*/mv & &.save/' list > list2mv  
	cat list2mv:
		mv list list.save
		mv sed.html sed.html.save
		mv file.txt file.txt.save
 rm list
 chmod 750 list2mv
 ./list2mv (to move files to *.save version, could use cp to make backup copies)

 delete blank spaces: 
     FROM within vi:
         :%s/^  *\(.*\)/\1/g    OR
   	    :%s$^  *\(.*\)$\1$g

 delete blank lines: 
            sed  /^$/d  file > newfile
     FROM within vi:
               :g/^$/d  or :%s/^$/d

 some "sed-like" tricks using  perl at the command line: 

	THIS WILL NOT WORK:   perl -pi -e 's/.*/# &/g'   - for two reasons:
		1) need to place the variables in parens e.g.  (.*)
		2) the ampersand will not work, need to use $1 instead.

THIS WORKS:

perl -pi -e 's/(.*)/# $1/' filename - inserts a # at the beginning of every line

---------------------------------------------------------------------------------------------------
  note: to do specific lines in perl gets a bit ugly...  (as far as I know at this point)

perl -pi'*.bak' -e 'if ($.==3){s/$_/Replacement text/;}' file.txt
        3 is the line affected
    (this might work at the command line, ymmv - tested on HP)

perl -pi'*.bak' -e 'if ($.==3..20){s/$_/Replacement text/;}' file.txt
 	3 through 20 would be affected

---------------------------------------------------------------------------------------------------
To help reduce the harvesting of emails, one can use perl to replace the at sign with -at-:

find . *.txt -exec perl -pi -e 's/@/-at-/g' {} \;

---------------------------------------------------------------------------------------------------