用re解析python文件

Question

I have a python file as

test.py 

import os
class test():

    def __init__(self):
        pass

    def add(num1, num2):
        return num1+num2

I am reading this file in a string as :

with open('test.py', 'r') as myfile:
    data=myfile.read()

print data

Now, my data contains the string with all lines and new lines. I need to find lines with start of class and def.

for example:

I need the output to be printed as :

class test():
def __init__(self):
def add(num1, num2):

How can I process this using regular expressions?

Answer 1

So if you need to find all def and class lines it is much easier to avoid regex.

You read the whole content of the file here

with open('test.py', 'r') as myfile:
    data=myfile.read()

print data

Why don't you just find the answer right there?

with open('test.py', 'r') as myfile:
    for line in myfile:
        stripped = line.strip()  # get rid of spaces left and right
        if stripped.startswith('def') or stripped.startswith('class'):
             print(line)

To work with a whole string as you requested:

import re
with open('test.py', 'r') as myfile:
    data = myfile.read()

print(data)

print(re.findall("class.+\n|def.+\n",data))

As you can see from the comments this will match ''definied as bla bla' as well. So it is better to use

print(re.findall("class .+\n|def .+\n",data))

Answer 2

If you want to follow a regex approach, use

re.findall(r'(?m)^[ \t]*((?:class|def)[ \t].*)', data)

or

re.findall(r'^[ \t]*((?:class|def)[ \t].*)', data, flags=re.M)

See regex demo

The point is that you should use ^ as the beginning of the line anchor (hence, (?m) at the start or re.M flag are necessary), then you match horizontal whitespaces (with [ \\t] ), then either class or def (with (?:class|def) ), and then again a space or tab and then 0+ chars other than a newline ( .* ).

If you plan to also handle Unicode whitespace, you need to replace [ \\t] with [^\\S\\r\\n\\f\\v] (and use the re.UNICODE flag).

Python demo :

import re
p = re.compile(r'^[ \t]*((?:class|def)[ \t].*)', re.MULTILINE)
s = "test.py \n\nimport os\nclass test():\n\n    def __init__(self):\n        pass\n\n    def add(num1, num2):\n        return num1+num2"
print(p.findall(s))
# => ['class test():', 'def __init__(self):', 'def add(num1, num2):']

Answer 3

with open('test.py', 'r') as myfile:
    data=myfile.read().split('\n')
    for line in data:
        if re.search("(\s+)?class ", line) or re.search("^\s+def ", line):
            print line