Can I use python “re.sub” instead of “sed”?

Question

I would like to convert a command from one line to multiple:

Example Input:

myprog  -library lib -topcell top -view layout

Output:

myprog \
          -library  lib\
          -topCell  top\
          -view  layout\

I can do that using "sed" as following:

echo $cmd | sed 's/\s-[a-zA-Z0-9]*\s/\\\n\t & /g'

But I can't replicate that by python using re.sub. What I noticed is that re.sub doesn't accept a regular expression as a second argument as following:

>>> re.sub(r'-[a-zA-Z0-9]*\s',r'[a-zA-Z0-9]*\s',cmd)

‘myprog [a-zA-Z0-9]*\\slib [a-zA-Z0-9]*\\stop [a-zA-Z0-9]*\\slayout'

Do you have a solution?

Answer 1

x="myprog  -library lib -topcell top -view layout"
print re.sub(r"(?=-)",r"\\\n\t",x)

Try this.

Answer 2

In python it would be like,

>>> s = "myprog  -library lib -topcell top -view layout"
>>> print re.sub(r'(-[a-zA-Z0-9]*\s)',r'\\\n\t\1', s)
myprog  \
    -library lib \
    -topcell top \
    -view layout
>>> print re.sub(r'(-[a-zA-Z0-9]*\s)',r'\\\n\t \1 ', s)
myprog  \
     -library  lib \
     -topcell  top \
     -view  layout

In sed & in the replacement part prints the matched characters. In python, i just used capturing groups.