Test code:
```
import re
test='''012345678
         
012345678
'''
pattern = r'^\s+?$'
m = re.search(pattern, test, re.M)
if m:
    print(f'TEST FOUND "{m.span()}"')

    def replace(m):
        print(f'TEST REMOVE {m.span()}')
        return ''

    test = re.sub(pattern, replace, test, re.M)
    m = re.search(pattern, test, re.M)
    if m:
        print(f'TEST STILL THERE "{m.span()}"')

print('COMPILE PATTERN FIRST')
pattern_re = re.compile(pattern, re.M)
m = re.search(pattern_re, test)
if m:
    print(f'TEST FOUND "{m.span()}"')

    def replace(m):
        print(f'TEST REMOVE {m.span()}')
        return ''

    test = re.sub(pattern_re, replace, test)
    m = re.search(pattern_re, test)
    if m:
        print(f'TEST STILL THERE "{m.span()}"')
```

Actual output:

TEST FOUND "(10, 19)"
TEST STILL THERE "(10, 19)"
COMPILE PATTERN FIRST
TEST FOUND "(10, 19)"
TEST REMOVE (10, 19)

This is an inconsistency between re.search and re.sub. Either this is a bug in the code or in the documentation.

Not a bug.

Argument 4 of re.sub is the count:

    sub(pattern, repl, string, count=0, flags=0)

not the flags.