Question

Can anyone help me with translating this regex to Python re module?

sed -e "s/^[^ ]* \([^ ]*\) \([0-9]*\) \([0-9:]*\) \([0-9]*\) /\1 \2 \4 \3 /"

As I’ve understood this replaces one text to another. How can I create something like this using re module? Thanks!

e.g. Fri Mar 21 07:16:51 2008 -0600 will be converted to Mar 21 2008 07:16:51 -0600.

source = 'Fri Mar 21 07:16:51 2008 -0600'
pattern = re.compile('^[^ ]* \([^ ]*\) \([0-9]*\) \([0-9:]*\) \([0-9]*\) ')
result = re.sub('\\1 \\2 \\4 \\3 ', source)

Answer 1

Drop the backslashes on the parens in your pattern too, as in:

pattern = re.compile('^[^ ]* ([^ ]*) ([0-9]*) ([0-9:]*) ([0-9]*) ')

The way you had it, you were escaping the parens which need to be recognized as groupings so that you actually capture the four groups corresponding to the \1 … \4