Skip to content

Merge canary into main: path security, Zod validation, AI enhancements, infra updates #180

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
sr2echa merged 2 commits into from
Mar 9, 2026
Merged

Merge canary into main: path security, Zod validation, AI enhancements, infra updates #180

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
45f6a85
Initial plan
Copilot Mar 9, 2026
6dcd2f2
Fix 4 Polynomial ReDoS vulnerabilities in backend/main.py regexes
Copilot Mar 9, 2026
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
15 changes: 9 additions & 6 deletions backend/main.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,9 +72,10 @@ def _to_seconds(value: Any) -> float | None:
except Exception:
pass
# 1h2m3s / 2m / 90s etc.
# Bounded quantifiers and longest-unit-first ordering prevent polynomial backtracking.
total = 0.0
matched = False
for m in re.finditer(r"(?P<num>\d+(?:\.\d+)?)\s*(?P<unit>h|hr|hrs|hour|hours|m|min|mins|minute|minutes|s|sec|secs|second|seconds|ms|millisecond|milliseconds)\b",
for m in re.finditer(r"(?P<num>[0-9]{1,15}(?:\.[0-9]{1,10})?)[ ]?(?P<unit>milliseconds|millisecond|minutes|minute|seconds|second|hours|hour|secs|mins|hrs|min|sec|ms|hr|m|s|h)\b",
s):
matched = True
num = float(m.group("num"))
Expand Down Expand Up @@ -109,7 +110,9 @@ def _normalize_time_fields_from_text(user_text: str, args: dict[str, Any]) -> di
text = (user_text or "").lower()

# Extract explicit FROM ... TO ... first
m = re.search(r"from\s+([^\s]+(?:\s*[a-z]+)?)\s+to\s+([^\s]+(?:\s*[a-z]+)?)", text)
# Use [0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})? so the numeric, attached-unit, and
# separated-unit character classes are disjoint — no ambiguous matching, no ReDoS.
m = re.search(r"from\s+([0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})?)\s+to\s+([0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})?)", text)
if m:
start_candidate = _to_seconds(m.group(1))
end_candidate = _to_seconds(m.group(2))
Expand All @@ -119,14 +122,14 @@ def _normalize_time_fields_from_text(user_text: str, args: dict[str, Any]) -> di
updated["end_seconds"] = end_candidate

# AT ... / START AT ... / FROM ... (single)
m2 = re.search(r"(?:at|starting\s+at|start\s+at|from)\s+([^\s]+(?:\s*[a-z]+)?)", text)
m2 = re.search(r"(?:at|starting\s+at|start\s+at|from)\s+([0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})?)", text)
if m2 and updated.get("start_seconds") is None:
start_candidate = _to_seconds(m2.group(1))
if start_candidate is not None:
updated["start_seconds"] = start_candidate

# FOR ... / SPAN FOR ...
m3 = re.search(r"(?:for|span(?:s)?\s+for)\s+([^\s]+(?:\s*[a-z]+)?)", text)
m3 = re.search(r"(?:for|span(?:s)?\s+for)\s+([0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})?)", text)
if m3 and updated.get("duration_seconds") is None:
dur_candidate = _to_seconds(m3.group(1))
if dur_candidate is not None:
Expand All @@ -135,13 +138,13 @@ def _normalize_time_fields_from_text(user_text: str, args: dict[str, Any]) -> di
# TO ... LONG / SET TO ... / MAKE (IT)? ...
# Examples: "to 12 seconds long", "set to 12s", "make it 8 sec", "12s long"
if updated.get("duration_seconds") is None:
m4 = re.search(r"(?:to\s+)?([^\s]+(?:\s*[a-z]+)?)\s+long", text)
m4 = re.search(r"(?:to\s+)?([0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})?)\s+long", text)
if m4:
dur_candidate = _to_seconds(m4.group(1))
if dur_candidate is not None:
updated["duration_seconds"] = dur_candidate
if updated.get("duration_seconds") is None:
m5 = re.search(r"(?:set\s+(?:it\s+)?to|make\s+(?:it\s+)?)\s+([^\s]+(?:\s*[a-z]+)?)", text)
m5 = re.search(r"(?:set\s+(?:it\s+)?to|make\s+(?:it\s+)?)\s+([0-9][0-9.]*[a-z]*(?:[ ][a-z]{1,12})?)", text)
if m5:
dur_candidate = _to_seconds(m5.group(1))
if dur_candidate is not None:
Expand Down