Add Python web scraper for NJC travel rates with currency extraction

- Implemented Python scraper using BeautifulSoup and pandas to automatically collect travel rates from official NJC website
- Added currency extraction from table titles (supports EUR, USD, AUD, CAD, ARS, etc.)
- Added country extraction from table titles for international rates
- Flatten pandas MultiIndex columns for cleaner data structure
- Default to CAD for domestic Canadian sources (accommodations and domestic tables)
- Created SQLite database schema (raw_tables, rate_entries, exchange_rates, accommodations)
- Successfully scraped 92 tables with 17,205 rate entries covering 25 international cities
- Added migration script to convert scraped data to Node.js database format
- Updated .gitignore for Python files (.venv/, __pycache__, *.pyc, *.sqlite3)
- Fixed city validation and currency conversion in main app
- Added comprehensive debug and verification scripts

This replaces manual JSON maintenance with automated data collection from official government source.

scripts/inspect_table_structure.py

@@ -0,0 +1,31 @@
+"""Inspect the actual table structure from NJC"""
+import sys
+sys.path.insert(0, 'src')
+
+from gov_travel.scrapers import SourceConfig, scrape_tables_from_source
+import json
+
+# Create a test source config
+source = SourceConfig(name="international", url="https://www.njc-cnm.gc.ca/directive/app_d.php?lang=en")
+
+# Get just the first table
+print("Fetching tables...")
+tables = scrape_tables_from_source(source)
+
+first_table = tables[0]
+print(f"\nTable {first_table['table_index']}")
+print(f"Title: {first_table['title']}")
+print(f"\nFirst data row:")
+print(json.dumps(first_table['data'][0], indent=2))
+
+print(f"\nSecond data row:")
+print(json.dumps(first_table['data'][1], indent=2))
+
+# Now try Argentina
+for table in tables:
+    if table['title'] and 'Argentina' in table['title']:
+        print(f"\n\n=== Argentina Table ===")
+        print(f"Title: {table['title']}")
+        print(f"\nFirst row:")
+        print(json.dumps(table['data'][0], indent=2))
+        break