If you run those strings against re instead you’ll see something different. E.g. via re.match:
0>>> import re
1>>> re.match(".", "K") # standard string, no escape, matches any character
2<re.Match object; span=(0, 1), match='K'>
3>>> re.match("\.", "K") # standard string, with escape, does not match any character (hence no result)
4>>> re.match("\.", ".") # standard string, with escape, does match "."
5<re.Match object; span=(0, 1), match='.'>
6>>> re.match(r"\.", "K") # raw string, with escape, does not match any character
7>>> re.match(r"\.", ".") # raw string, with escape, does match "."
8<re.Match object; span=(0, 1), match='.'>
9>>> re.match("\\.", "K") # standard string, double escape, does not match any character
10>>> re.match("\\.", ".") # standard string, double escape, does match "."
11<re.Match object; span=(0, 1), match='.'>
12>>> re.match(r"\\.", "K") # raw string, double escape, does not match any character
13>>> re.match(r"\\.", ".") # raw string, double escape, does not match "."
You can see above that absent the \, “.” is interpreted as any character, while “\.”, r"\." and “\\.” are both interpreted as an escaped period. Python is planning to error on “\.” in the future.
https://bugs.python.org/issue27364
