How to delete a line with re?

[Peng Yu]

Hi,

I want to delete the line with abc in the following program. But the
following program does not do what I want. Can somebody let me know
how to do it?

Thanks,
Peng

#!/usr/bin/python

import re

file="""abcd
efg
hijk
lmn
"""

regex = re.compile("^abcd$", re.MULTILINE)

print file,
print "$"
print regex.sub('', file),

[Peter Otten]

Peng Yu wrote:

I want to delete the line with abc in the following program. But the
following program does not do what I want. Can somebody let me know
how to do it?

file="""abcd
efg
hijk
lmn
"""

regex = re.compile("^abcd$", re.MULTILINE)

print regex.sub('', file),

What /do/ you want? If you want to remove the trailing newline

regex = re.compile("^abcd$\n?", re.MULTILINE)

might work.

Peter

[John Machin]

On Aug 18, 4:22*pm, Peter Otten <[e-mail]:

Peng Yu wrote:

I want to delete the line with abc in the following program. But the
following program does not do what I want. Can somebody let me know
how to do it?
file="""abcd
efg
hijk
lmn
"""

regex = re.compile("^abcd$", re.MULTILINE)
print regex.sub('', file),

What /do/ you want? If you want to remove the trailing newline

regex = re.compile("^abcd$\n?", re.MULTILINE)

might work.

Not only might work, but does work, including covering the corner
cases where the abcd line is immediately followed by (1) an empty line
(2) no newline then end-of-file. It is also more elegant [yes, even
regular expressions can be elegant] than what I came up with (see
below).

Hint for the OP: repr() is your friend.

>>import re
src="""abcd

... efg
... abcd
...
... hijk
... abcd"""

>>expected = """efg

...
... hijk
... """

>>print repr(src)

'abcd\nefg\nabcd\n\nhijk\nabcd'

>>print repr(expected)

'efg\n\nhijk\n'

>>for pattern in ["^abcd$\n?", r"^abcd(\n|\Z)"]:

... regex = re.compile(pattern, re.MULTILINE)
... actual = regex.sub('', src)
... print repr(actual)
... print actual == expected
...
'efg\n\nhijk\n'
True
'efg\n\nhijk\n'
True

>>>

Cheers,
John