re.sub question (regular expressions)

[Chris Seberino]

What does this line do?...

input_ = re.sub("([a-zA-Z]+)", '"\\1"', input_)

Does it remove parentheses from words?
e.g. (foo) -> foo ???

I'd like to replace [a-zA-Z] with \w but \w makes it blow up.

In other words, re.sub("(\w+)", '"\\1"', input_) blows up.

Why?

cs

 

[Jean-Michel Pichavant]

What does this line do?...

input_ = re.sub("([a-zA-Z]+)", '"\\1"', input_)

Does it remove parentheses from words?
e.g. (foo) -> foo ???

I'd like to replace [a-zA-Z] with \w but \w makes it blow up.

In other words, re.sub("(\w+)", '"\\1"', input_) blows up.

Why?

cs

( has a special meaning of not preceded by \ (see re help)

the first line transform 'foo' into '"foo"'
your solution is doing the same.

If I'm wrong, MRAB will come into rescue anyway )

JM

 

[MRAB]

What does this line do?...

input_ = re.sub("([a-zA-Z]+)", '"\\1"', input_)

Why don't you try it?

Does it remove parentheses from words?
e.g. (foo) -> foo ???

No, it puts quotes around them.

I'd like to replace [a-zA-Z] with \w but \w makes it blow up.

In other words, re.sub("(\w+)", '"\\1"', input_) blows up.

Why?

What do you mean "blow up"? It worked for me in Python v2.6.2.

 

[Chris Seberino]

What do you mean "blow up"? It worked for me in Python v2.6.2.

My bad. False alarm. This was one of those cases where a bug in
another area appears like a bug in a different area.

Thank for the help.

cs