Variable Substiution in sed executing via python

Question

Below is my sed command:

sed "s/N([0-9]*)/sum('\1')/g" old.txt > new.txt

works fine in bash terminal

But below python code doesnot execute :-

file1=old.txt
file2.new.txt
sed_cmd = 'sed "s/0x\([0-9]*\)/sum('\1')/g" %s > %s' %(file1,file2)

I get syntax error

After above I need do :-

subprocess.call([sed_cmd],shell=True)

it comes like this sed "s/0x\([0-9]*\)/sum('')/g" old.txt > new.txt — HDev007, Mar 24 '21 at 15:33
use raw strings for the sed command and then concatenate the files separately.. something like `r'''sed "s/0x\([0-9]*\)/sum('\1')/g" ''' + '%s > %s' %(file1,file2)` — Sundeep, Mar 24 '21 at 15:36

score 0 · Answer 1 · answered Mar 24 '21 at 21:38

Main issues:

Single quotes ' are used unescaped in a regular single quoted string literal (that make the syntax corrupt)
\1 inside a regular string literal defines a \x01 character, a SOH char, and not a regex backreference (that requires two char combination, \ + here, 1).

You can use

sed_cmd = r"""sed "s/0x\([0-9]*\)/sum('\1')/g" {} > {}""".format(file1,file2)

Details:

r"""...""" is a triple-quoted raw string literal inside which you may freely use unescaped ' and ", and \ chars denote literal backslashes
{} are placeholders in the format string, the first one will be filled with the contents of the first and second variables passed as arguments in parentheses (see .format(file1,file2)).

1 Answers1