在bash shell中删除字符串中逗号的空格

Question

I would like to replace blank spaces/white spaces in a string with commas.

STR1=This is a string 

to

STR1=This,is,a,string

Answer 1

Without using external tools:

echo ${STR1// /,}

Demo:

$ STR1="This is a string"
$ echo ${STR1// /,}
This,is,a,string

See bash: Manipulating strings .

Answer 2

Just use sed:

echo $STR1 | sed 's/ /,/g'

or pure BASH way::

echo ${STR1// /,}

Answer 3

kent$  echo "STR1=This is a           string"|awk -v OFS="," '$1=$1'
STR1=This,is,a,string

Note:

if there are continued blanks, they would be replaced with a single comma. as example above shows.

Answer 4

How about

STR1="This is a string"
StrFix="$( echo "$STR1" | sed 's/[[:space:]]/,/g')"

echo "$StrFix"

**output**
This,is,a,string

If you have multiple adjacent spaces in your string and what to reduce them to just 1 comma, then change the sed to

STR1="This is    a    string"
StrFix="$( echo "$STR1" | sed 's/[[:space:]][[:space:]]*/,/g')"

 echo "$StrFix"

**output**
This,is,a,string

I'm using a non-standard sed, and so have used ``[[:space:]][[:space:]]* to indicate one or more "white-space" characters (including tabs, VT, maybe a few others). In a modern sed, I would expect to indicate one or more "white-space" characters (including tabs, VT, maybe a few others). In a modern sed, I would expect [[:space:]]+` to work as well.

Answer 5

This might work for you:

echo 'STR1=This is a string' | sed 'y/ /,/'
STR1=This,is,a,string

or:

echo 'STR1=This is a string' | tr ' ' ','  
STR1=This,is,a,string

Answer 6

STR1=`echo $STR1 | sed 's/ /,/g'`