如何在python正则表达式中重复模式？

Question

I'm doing a python regex and have a working expression:

\n(?P<curve>\w+)(?:.+)(?P<unit>\.\S*)(?:\s+.\s+)(?P<desc>:.+)|\n(?P<curve2>\w+)(?:.+)(?P<unit2>\.\S*)|\n(?P<curve3>\w+)

I would like to know I could repeat the pattern from the first if, the reason is that I would like to not group in many "curve" or "unit" for each case.

My test data is as follows:

#-------------
MD              
BMK_STA            .Mpsi                                   : Modulus
FANG        .                                   : Friction Angle
PR             .unitless                               :  
RHO           .g/cm3                                  

The idea is to have MD and RHO also in "curve" group.

Answer 1

I am not entirely sure what you mean, but the following may help:

If you want to find every match for a pattern, you can use re.findall(pattern, string)

It returns a list of the matches..

re module docs

Answer 2

There is no special syntax to avoid that kind of repetition in regexes, so in the general case you can't avoid a certain amount of repetition. However in your specific case you should be able to solve your problem using optional groups:

\n(?P<curve>\w+)((?:.+)(?P<unit>\.\S*)((?:\s+.\s+)(?P<desc>:.+))?)?

Which is probably better written in verbose mode as:

\n(?P<curve>\w+)
(
    .+
    (?P<unit>\.\S*)
    (
        \s+.\s+
        (?P<desc>:.+)
    )?
)?

to make the group nesting easier to read. I've also remove the ?: groups since in this case they are useless.

Answer 3

Assuming your regex is correct. Use the finditer() method for this purpose to iterate all the matches.

Example:

for m in re.finditer(r'REGEX_GOES_HERE', text):
    print m.group('curve')
    print m.group("unit")

In this way you picked all the matches, as well as their named groups are intact as you wanted!