Possibly wrong regex patterns?

[RadhiFadlillah]

Hi @adbar. Thanks for this awesome library. While working on my port I noticed a strange issue with several regexes that don't give me the expected output.

First is COPYRIGHT_PATTERN which currently defined like this:

COPYRIGHT_PATTERN = re.compile(
    rf"(?:©|\©|Copyright|\(c\))\D*(?:{YEAR_RE}-)?({YEAR_RE})\D"
)

Given the following string:

<p>&copy; Copyright 1999-2020 Asia Pacific Star. All rights reserved.</p>

I expected the COPYRIGHT_PATTERN to capture 2020, however right now it's capture the 1999 as can be seen here. So what do you think? Is that the expected behavior?

---

Next is TIMESTAMP_PATTERN which currently defined like this:

TIMESTAMP_PATTERN = re.compile(
    rf"({YEAR_RE}-{MONTH_RE}-{DAY_RE}).[0-9]{{2}}:[0-9]{{2}}:[0-9]{{2}}"
)

Given the following string:

1991-03-21T00:00:00
1995-07-23T00:00:00
2020-05-10T18:59:01
2021-02-15T11:29:21
2023-01-10T22:18:38

TIMESTAMP_PATTERN will only capture the last three while ignoring the first two as can be seen here. I believe it was a mistake since 1991-03-21T00:00:00 and 1995-07-23T00:00:00 are valid timestamps as well.

Thanks!

[RadhiFadlillah]

For COPYRIGHT_PATTERN I think it can be modified by putting the dash outside the optional non-capturing group as can be seen here:

COPYRIGHT_PATTERN = re.compile(
-    rf"(?:©|\©|Copyright|\(c\))\D*(?:{YEAR_RE}-)?({YEAR_RE})\D"
+    rf"(?:©|\©|Copyright|\(c\))\D*(?:{YEAR_RE})?-?({YEAR_RE})\D"
)

And for TIMESTAMP_PATTERN I think it can be fixed by surrounding YEAR_RE, MONTH_RE and DAY_RE with non-capturing group as can be seen here:

TIMESTAMP_PATTERN = re.compile(
-    rf"({YEAR_RE}-{MONTH_RE}-{DAY_RE}).[0-9]{{2}}:[0-9]{{2}}:[0-9]{{2}}"
+    rf"((?:{YEAR_RE})-(?:{MONTH_RE})-(?:{DAY_RE})).[0-9]{{2}}:[0-9]{{2}}:[0-9]{{2}}"
)