Add master build, validation, and pre-push check scripts

- build.py: Master script to build all HTML pages
- validate_data.py: Data validation (required fields, URLs, images)
- pre_push_check.py: Combined validation + build check
- utils.py: Added load_spreadsheet_all_sheets function

All 76 tests pass.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <[email protected]>

Author: jeremymanning

scripts/build.py

@@ -0,0 +1,80 @@
+#!/usr/bin/env python3
+"""Master build script for the Context Lab website.
+
+Generates all HTML pages from spreadsheet data and templates.
+Run this script to regenerate all content pages.
+"""
+import sys
+from pathlib import Path
+
+from build_publications import build_publications
+from build_people import build_people
+from build_software import build_software
+
+
+def main():
+    """Build all content pages."""
+    project_root = Path(__file__).parent.parent
+
+    data_dir = project_root / 'data'
+    templates_dir = project_root / 'templates'
+
+    # Track build results
+    results = []
+
+    # Build publications.html
+    try:
+        build_publications(
+            data_dir / 'publications.xlsx',
+            templates_dir / 'publications.html',
+            project_root / 'publications.html'
+        )
+        results.append(('publications.html', 'OK'))
+    except Exception as e:
+        results.append(('publications.html', f'FAILED: {e}'))
+
+    # Build people.html
+    try:
+        build_people(
+            data_dir / 'people.xlsx',
+            templates_dir / 'people.html',
+            project_root / 'people.html'
+        )
+        results.append(('people.html', 'OK'))
+    except Exception as e:
+        results.append(('people.html', f'FAILED: {e}'))
+
+    # Build software.html
+    try:
+        build_software(
+            data_dir / 'software.xlsx',
+            templates_dir / 'software.html',
+            project_root / 'software.html'
+        )
+        results.append(('software.html', 'OK'))
+    except Exception as e:
+        results.append(('software.html', f'FAILED: {e}'))
+
+    # Print summary
+    print("\n" + "=" * 50)
+    print("Build Summary")
+    print("=" * 50)
+
+    failed = 0
+    for page, status in results:
+        print(f"  {page}: {status}")
+        if status != 'OK':
+            failed += 1
+
+    print("=" * 50)
+
+    if failed > 0:
+        print(f"Build completed with {failed} error(s)")
+        sys.exit(1)
+    else:
+        print("Build completed successfully!")
+        sys.exit(0)
+
+
+if __name__ == '__main__':
+    main()

scripts/pre_push_check.py

@@ -0,0 +1,62 @@
+#!/usr/bin/env python3
+"""Pre-push check script for the Context Lab website.
+
+Validates data files and rebuilds all HTML pages.
+This should be run before pushing to ensure consistency.
+"""
+import subprocess
+import sys
+from pathlib import Path
+
+
+def run_script(script_name: str) -> bool:
+    """Run a Python script and return True if successful."""
+    script_path = Path(__file__).parent / script_name
+    print(f"\n{'=' * 50}")
+    print(f"Running {script_name}...")
+    print('=' * 50)
+
+    result = subprocess.run(
+        [sys.executable, str(script_path)],
+        cwd=Path(__file__).parent.parent
+    )
+    return result.returncode == 0
+
+
+def main():
+    """Run all pre-push checks."""
+    print("Context Lab Website Pre-Push Check")
+    print("=" * 50)
+
+    all_passed = True
+
+    # Step 1: Validate data
+    if not run_script('validate_data.py'):
+        print("\n*** Data validation FAILED ***")
+        print("Fix validation errors before pushing.")
+        all_passed = False
+
+    # Step 2: Build all pages
+    if all_passed:
+        if not run_script('build.py'):
+            print("\n*** Build FAILED ***")
+            print("Fix build errors before pushing.")
+            all_passed = False
+
+    # Summary
+    print("\n" + "=" * 50)
+    print("Pre-Push Check Summary")
+    print("=" * 50)
+
+    if all_passed:
+        print("All checks PASSED!")
+        print("\nYou can safely push your changes.")
+        sys.exit(0)
+    else:
+        print("Some checks FAILED!")
+        print("\nPlease fix the issues before pushing.")
+        sys.exit(1)
+
+
+if __name__ == '__main__':
+    main()

scripts/utils.py

@@ -51,6 +51,48 @@ def load_spreadsheet(filepath: Path) -> List[Dict[str, Any]]:
     return rows
 
 
+def load_spreadsheet_all_sheets(filepath: Path) -> Dict[str, List[Dict[str, Any]]]:
+    """Load Excel spreadsheet with all sheets.
+
+    Args:
+        filepath: Path to the .xlsx file
+
+    Returns:
+        Dictionary with sheet names as keys, each containing list of row dicts.
+        Empty cells are converted to empty strings.
+
+    Raises:
+        FileNotFoundError: If the spreadsheet doesn't exist
+    """
+    wb = openpyxl.load_workbook(filepath, read_only=True, data_only=True)
+
+    data = {}
+    for sheet_name in wb.sheetnames:
+        sheet = wb[sheet_name]
+
+        # Get headers from first row
+        headers = [cell.value for cell in sheet[1]]
+
+        rows = []
+        for row in sheet.iter_rows(min_row=2, values_only=True):
+            # Skip completely empty rows
+            if not any(cell is not None for cell in row):
+                continue
+
+            row_dict = {}
+            for header, value in zip(headers, row):
+                if value is None:
+                    row_dict[header] = ''
+                else:
+                    row_dict[header] = value
+            rows.append(row_dict)
+
+        data[sheet_name] = rows
+
+    wb.close()
+    return data
+
+
 def inject_content(template_path: Path, output_path: Path,
                    replacements: Dict[str, str]) -> None:
     """Inject generated content into template at marker locations.