feat(scripts): rewrite parser as modular Python CLI

Replace monolithic scraping scripts with sportstime_parser package: - Multi-source scrapers with automatic fallback for 7 sports - Canonical ID generation for games, teams, and stadiums - Fuzzy matching with configurable thresholds for name resolution - CloudKit Web Services uploader with JWT auth, diff-based updates - Resumable uploads with checkpoint state persistence - Validation reports with manual review items and suggested matches - Comprehensive test suite (249 tests) CLI: sportstime-parser scrape|validate|upload|status|retry|clear Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2026-01-10 21:06:12 -06:00
parent 284a10d9e1
commit eeaf900e5a
109 changed files with 18415 additions and 266211 deletions
--- a/Scripts/sportstime_parser/tests/test_timezone.py
+++ b/Scripts/sportstime_parser/tests/test_timezone.py
@@ -0,0 +1,187 @@
+"""Tests for timezone conversion utilities."""
+
+import pytest
+from datetime import datetime, date
+from zoneinfo import ZoneInfo
+
+from sportstime_parser.normalizers.timezone import (
+    detect_timezone_from_string,
+    detect_timezone_from_location,
+    parse_datetime,
+    convert_to_utc,
+    get_stadium_timezone,
+    TimezoneResult,
+)
+
+
+class TestDetectTimezoneFromString:
+    """Tests for detect_timezone_from_string function."""
+
+    def test_eastern_time(self):
+        """Test Eastern Time detection."""
+        assert detect_timezone_from_string("7:00 PM ET") == "America/New_York"
+        assert detect_timezone_from_string("7:00 PM EST") == "America/New_York"
+        assert detect_timezone_from_string("7:00 PM EDT") == "America/New_York"
+
+    def test_central_time(self):
+        """Test Central Time detection."""
+        assert detect_timezone_from_string("8:00 PM CT") == "America/Chicago"
+        assert detect_timezone_from_string("8:00 PM CST") == "America/Chicago"
+        assert detect_timezone_from_string("8:00 PM CDT") == "America/Chicago"
+
+    def test_mountain_time(self):
+        """Test Mountain Time detection."""
+        assert detect_timezone_from_string("7:00 PM MT") == "America/Denver"
+        assert detect_timezone_from_string("7:00 PM MST") == "America/Denver"
+
+    def test_pacific_time(self):
+        """Test Pacific Time detection."""
+        assert detect_timezone_from_string("7:00 PM PT") == "America/Los_Angeles"
+        assert detect_timezone_from_string("7:00 PM PST") == "America/Los_Angeles"
+        assert detect_timezone_from_string("7:00 PM PDT") == "America/Los_Angeles"
+
+    def test_no_timezone(self):
+        """Test string with no timezone."""
+        assert detect_timezone_from_string("7:00 PM") is None
+        assert detect_timezone_from_string("19:00") is None
+
+    def test_case_insensitive(self):
+        """Test case insensitive matching."""
+        assert detect_timezone_from_string("7:00 PM et") == "America/New_York"
+        assert detect_timezone_from_string("7:00 PM Et") == "America/New_York"
+
+
+class TestDetectTimezoneFromLocation:
+    """Tests for detect_timezone_from_location function."""
+
+    def test_eastern_states(self):
+        """Test Eastern timezone states."""
+        assert detect_timezone_from_location(state="NY") == "America/New_York"
+        assert detect_timezone_from_location(state="MA") == "America/New_York"
+        assert detect_timezone_from_location(state="FL") == "America/New_York"
+
+    def test_central_states(self):
+        """Test Central timezone states."""
+        assert detect_timezone_from_location(state="TX") == "America/Chicago"
+        assert detect_timezone_from_location(state="IL") == "America/Chicago"
+
+    def test_mountain_states(self):
+        """Test Mountain timezone states."""
+        assert detect_timezone_from_location(state="CO") == "America/Denver"
+        assert detect_timezone_from_location(state="AZ") == "America/Phoenix"
+
+    def test_pacific_states(self):
+        """Test Pacific timezone states."""
+        assert detect_timezone_from_location(state="CA") == "America/Los_Angeles"
+        assert detect_timezone_from_location(state="WA") == "America/Los_Angeles"
+
+    def test_canadian_provinces(self):
+        """Test Canadian provinces."""
+        assert detect_timezone_from_location(state="ON") == "America/Toronto"
+        assert detect_timezone_from_location(state="BC") == "America/Vancouver"
+        assert detect_timezone_from_location(state="AB") == "America/Edmonton"
+
+    def test_case_insensitive(self):
+        """Test case insensitive matching."""
+        assert detect_timezone_from_location(state="ny") == "America/New_York"
+        assert detect_timezone_from_location(state="Ny") == "America/New_York"
+
+    def test_unknown_state(self):
+        """Test unknown state returns None."""
+        assert detect_timezone_from_location(state="XX") is None
+        assert detect_timezone_from_location(state=None) is None
+
+
+class TestParseDatetime:
+    """Tests for parse_datetime function."""
+
+    def test_basic_date_time(self):
+        """Test basic date and time parsing."""
+        result = parse_datetime("2025-12-25", "7:00 PM ET")
+        assert result.datetime_utc.year == 2025
+        assert result.datetime_utc.month == 12
+        assert result.datetime_utc.day == 26  # UTC is +5 hours ahead
+        assert result.source_timezone == "America/New_York"
+        assert result.confidence == "high"
+
+    def test_date_only(self):
+        """Test date only parsing."""
+        result = parse_datetime("2025-10-21")
+        assert result.datetime_utc.year == 2025
+        assert result.datetime_utc.month == 10
+        assert result.datetime_utc.day == 21
+
+    def test_timezone_hint(self):
+        """Test timezone hint is used when no timezone in string."""
+        result = parse_datetime(
+            "2025-10-21",
+            "7:00 PM",
+            timezone_hint="America/Chicago",
+        )
+        assert result.source_timezone == "America/Chicago"
+        assert result.confidence == "medium"
+
+    def test_location_inference(self):
+        """Test timezone inference from location."""
+        result = parse_datetime(
+            "2025-10-21",
+            "7:00 PM",
+            location_state="CA",
+        )
+        assert result.source_timezone == "America/Los_Angeles"
+        assert result.confidence == "medium"
+
+    def test_default_to_eastern(self):
+        """Test defaults to Eastern when no timezone info."""
+        result = parse_datetime("2025-10-21", "7:00 PM")
+        assert result.source_timezone == "America/New_York"
+        assert result.confidence == "low"
+        assert result.warning is not None
+
+    def test_invalid_date(self):
+        """Test handling of invalid date."""
+        result = parse_datetime("not a date")
+        assert result.confidence == "low"
+        assert result.warning is not None
+
+
+class TestConvertToUtc:
+    """Tests for convert_to_utc function."""
+
+    def test_convert_naive_datetime(self):
+        """Test converting naive datetime to UTC."""
+        dt = datetime(2025, 12, 25, 19, 0)  # 7:00 PM
+        utc = convert_to_utc(dt, "America/New_York")
+
+        # In December, Eastern Time is UTC-5
+        assert utc.hour == 0  # Next day 00:00 UTC
+        assert utc.day == 26
+
+    def test_convert_aware_datetime(self):
+        """Test converting timezone-aware datetime."""
+        tz = ZoneInfo("America/Los_Angeles")
+        dt = datetime(2025, 7, 4, 19, 0, tzinfo=tz)  # 7:00 PM PT
+        utc = convert_to_utc(dt, "America/Los_Angeles")
+
+        # In July, Pacific Time is UTC-7
+        assert utc.hour == 2  # 02:00 UTC next day
+        assert utc.day == 5
+
+
+class TestGetStadiumTimezone:
+    """Tests for get_stadium_timezone function."""
+
+    def test_explicit_timezone(self):
+        """Test explicit timezone override."""
+        tz = get_stadium_timezone("AZ", stadium_timezone="America/Phoenix")
+        assert tz == "America/Phoenix"
+
+    def test_state_inference(self):
+        """Test timezone from state."""
+        tz = get_stadium_timezone("NY")
+        assert tz == "America/New_York"
+
+    def test_default_eastern(self):
+        """Test default to Eastern for unknown state."""
+        tz = get_stadium_timezone("XX")
+        assert tz == "America/New_York"