Skip to content

emrg: fix flaky session ID uniqueness test — 16-bit → 32-bit entropy #539

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
argszero merged 1 commit into from
Aug 7, 2026
Merged

emrg: fix flaky session ID uniqueness test — 16-bit → 32-bit entropy #539

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 10 additions & 4 deletions emrg/session.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,17 +31,23 @@


def generate_session_id(cwd: Path) -> str:
"""Generate a human-friendly session ID: s_YYMMDD_HHMM_xxxx."""
"""Generate a human-friendly session ID: s_YYMMDD_HHMM_xxxxxxxx.

Suffix uses 4 random bytes (32 bits of entropy) — a 2-byte suffix
(16 bits, 65536 possibilities) made the 10-sample uniqueness test
flaky via the birthday paradox (~0.07% collision per run; observed
on CI master push run 31145421676).
"""
now = datetime.now()
prefix = f"s_{now.strftime('%y%m%d_%H%M')}_"
sessions_dir = cwd / ".emrg" / "sessions"
for _ in range(100):
suffix = secrets.token_hex(2)[:4]
suffix = secrets.token_hex(4)
sid = prefix + suffix
if not (sessions_dir / sid).exists():
return sid
# Fallback: use longer suffix
suffix = secrets.token_hex(4)[:8]
# Fallback: same 4-byte entropy (loop exhaustion is astronomically rare)
suffix = secrets.token_hex(4)
return prefix + suffix


Expand Down
Loading