I have a list of files that have columns separated by varying number of spaces. HOw can I sed or similar this so that each column is separated by a single space or tab?
I tried:
sed 's/ \+ /\t/g' file > tmp sed "s/\ /\t/g" tmp > file but R complained
line 526 did not have 11 elements Try:
sed 's/ \{1,\}/\t/g' file > tmp This takes one or more spaces and converts to the string in the second part of the expression ('\t' here).
Let say you have the file named "raw_file" that have the text below inside
COLUMN1 COLUMN2 COLUMN3 COLUMN4 I am a boy Hello word I see you I am 5 years How are you? You can use the command:
sed 's/ \{1,\}/ /g' '/root/Desktop/raw_file' |column -s ' ' -t > '/root/Desktop/my_new_file' RESULTS WHEN OUTPUT TO A NEW FILE CALLED "my_new_file"
COLUMN1 COLUMN2 COLUMN3 COLUMN4 I am a boy Hello word I see you I am 5 years RESULTS WHEN OUTPUT TO THE TERMINAL
Command used:
sed 's/ \{1,\}/ /g' '/root/Desktop/raw_file' |column -s ' ' -t 2> /dev/null Results:
COLUMN1 COLUMN2 COLUMN3 COLUMN4 I am a boy Hello word I see you I am 5 years NOTE: Each column is separated by two spaces.
awk will take care of whitespace directly and easily
awk '$1=$1' infile # convert to single space or awk '$1=$1' OFS="\t" infile # convert to single tab If you have to use SED, The character class \s will match the whitespace characters tab and space
so your sed code can be fixed as
sed 's/\s\{1,\}/ /g' infile or sed 's/\s\{1,\}/\t/g' infile if your sed support -r option,
sed -r 's/\s+/ /g' infile or sed -r 's/\s+/\t/g' infile