help - sed - insert space between any string of form XxxXxx, without replacing the pattern

If the string is of the pattern XxxXyzAbc...

The expected out put from sed has to be Xxx Xyz Abc ...

eg: if the string is QcfEfQfs, then the expec开发者_如何学Goted output is Qcf Ef Efs.

If i try to substitute the pattern [a-z][A-Z] with space then the sed will replace the character or pattern with space, like Qc f fs.

Is there any way to insert space in between without replacing the pattern ?.

Kindly help. Thanks.

Use match groups.

$ sed 's/\([a-z]\)\([A-Z]\)/\1 \2/g' <<< 'XxxXyzAbc'
Xxx Xyz Abc

A more accurate pattern match - one with fewer spurious matches on strings that are not in the XxxYyyZzz format - is:

sed 's/\([A-Z][a-z][a-z]*\)\([A-Z][a-z][a-z]*\)\([A-Z][a-z][a-z]*\)/\1 \2 \3/'

Consider the inputs:

XxxYyyZzz
xxxYyyZzz
X xY yZ z

One other answer I see will insert spaces in all the strings - when it should only modify the first string.

Updated to allow for the QcfEfEfs example - it isn't clear from the question how many lower case letters in a row are allowed. It could be 1 or 2; or it could be 1 or more - using a * allows for 1 or more; otherwise, use \{1,2\} for 1 or 2.

This might work for you (GNU sed):

sed 's/\B[[:upper:]]/ &/g' file

Looks for a non-word boundary followed by an uppercase character and inserts a space throughout the line.