fix: strip inline comments after empty unquoted values

[kamransiddiqui]

Closes theskumar#600.

Change _equal_sign in src/dotenv/parser.py from make_regex(r'(=[^%5C%5CS%5Cr%5Cn]*)') to make_regex(r'(=)') so whitespace after = is not consumed by the regex. In parse_unquoted_value, change re.sub(r'%5Cs+#.*', '', part).rstrip() to re.sub(r'%5Cs+#.*', '', part).strip() to remove leading whitespace that now remains part of the raw value. Together these prevent inline comments from being included as the value when a key has an empty unquoted value followed by whitespace and a comment, such as a= #c or a= %5Ct #c.

• Commit: fe754f9ba7e4

Iterate

gh pr checkout theskumar/python-dotenv

View this request in gitgot to access the live sandbox shell, review the agent's prompt, and request revisions:

431b5049-a65a-485f-9265-1ba7f5a34004

---

Draft PR opened by gitgot. Mark ready for review once you've verified the change.

Verification: ✅ Passed — gitgot ran the repo’s lint/typecheck/build/test in a sandbox; this PR introduces no new failures vs the base branch (any remaining red was already failing on the base).
Issue match: ✅ Addresses the issue (95% confidence) — The diff modifies src/dotenv/parser.py so the _equal_sign regex no longer consumes whitespace after =, leaving spaces/tabs before a # inside the _unquoted_value capture where the existing re.sub(r'\s+#.*', ...) can strip them; it also changes rstrip() to strip() to discard that leading whitespace, and adds tests verifying that a= #c and a= \t #c parse to an empty value.