正则表达式与？ 和{}量词

Question

I'm trying to create a regex to validate numbers that checks the year for leap year. here's a part of the code. for some reason this code would let number 4 8 24 28 as a valid regex.

(0{2}?)
([2468][480] | [13579][26])



pattern = re.compile (r"""

    ((0{2}?)([2468][480] | [13579][26]))

    """, re.X)

when I left out

(0{2}?)

24 12 and everything works..

I'm using verbose so spacing shouldn't matter..

Invalid

12
24
28
16

EDIT :: Actually all is invalid now..

i don't understand why 24 is invalid and 28 is invalid this doesn't make sense at all. I appreciate your guidance.

Answer 1

When you write (0{2}?) , that means “match two 0 s here, but match as few as possible”. Non-greediness doesn't really make sense for an {n} quantifier (it does for {n,} , and {m,n} ) – did you mean (0{2})? ?

Oh, and do keep in mind that years divisible by 400 are leap years.

Answer 2

Using the re.DEBUG flag to show debug info about the expression, we get

>>> pattern = re.compile(r'0{2}?', re.DEBUG)
min_repeat 2 2
  literal 48

The min_repeat shows that 0{2}? isn't being interpreted as ? applied to 0{2} . It's being interpreted as a lazy quantifier, attempting to match 0 any number of times from 2 to 2, but as few as possible. This doesn't quite seem consistent with the documentation; the docs only show the {m,n}? form.