Find and delete lines with multiple words in a text file

Question

如果文本文件中有多个单词，如何制作bash脚本来删除行？

Answer 1

Maybe something like that:

#!/bin/bash

exec 3>output

while read -r line; do
  nwords=`echo $line | wc -w`;
  if [ ! "$nwords" -gt 1 ]; then
      echo "$line" >&3;
  fi

done < input

exec 3>&-

Then you will have all lines which don't have more than word in output file and you can delete input file than and mv output to input.

Answer 2

I used this, using sed: my txt file is file.txt:

one
one two
one two tree for
one two

sed -i.bck '/\(.\+\)\(\s\+\)\(.*\)/d' file.txt
cat file.txt
one

Answer 3

If your words are separated by a " space " character you can do it in just one line, with sed . Let " ./myfile.txt " be your file name, then:

#!/bin/bash

sed -i "/[^ ][ ]/d" ./myfile.txt

The regex replace all space characters preceded by a non space characters, that is the case when you have two words.