first commint

detect_filled_cells.py

@@ -0,0 +1,100 @@
+import csv
+from openpyxl import load_workbook
+from openpyxl.styles import PatternFill
+
+
+def resolve_color_hex_from_colorobj(color):
+    """Try to resolve an openpyxl Color object to ARGB hex string.
+    Returns None if cannot resolve.
+    """
+    if color is None:
+        return None
+
+    # direct rgb (may be 'FF00B050' or '00B050')
+    rgb = getattr(color, "rgb", None)
+    if rgb:
+        s = str(rgb)
+        if len(s) == 6:
+            s = "FF" + s
+        return s.upper()
+
+    # try indexed (some workbooks may use indexed colors)
+    idx = getattr(color, "indexed", None) or getattr(color, "index", None)
+    if idx is not None:
+        try:
+            # openpyxl does not always expose a global list here; fallback to None
+            return f"INDEXED{int(idx)}"
+        except Exception:
+            pass
+
+    theme = getattr(color, "theme", None)
+    if theme is not None:
+        try:
+            return f"THEME{int(theme)}"
+        except Exception:
+            return str(theme)
+
+    return None
+
+
+def detect_colored_cells(xlsx_path, sheet_name=None, write_csv=None):
+    """Detect cells with fills (pattern fills) and return list of dicts:
+    [{'coordinate':'A1','row':1,'col':1,'value':..., 'color': 'FF00B050', 'fill_type': 'solid'}]
+
+    If write_csv is provided (path), write the results to that CSV.
+    """
+    wb = load_workbook(xlsx_path, data_only=False)
+    ws = wb[sheet_name] if sheet_name else wb.active
+
+    results = []
+    min_row = ws.min_row or 1
+    max_row = ws.max_row or 1
+    min_col = ws.min_column or 1
+    max_col = ws.max_column or 1
+
+    for row in ws.iter_rows(min_row=min_row, max_row=max_row, min_col=min_col, max_col=max_col):
+        for cell in row:
+            fill = cell.fill
+            if fill is None:
+                continue
+            # patternType for PatternFill is fill.fill_type or fill.patternType depending on openpyxl version
+            pattern = getattr(fill, "patternType", None) or getattr(fill, "fill_type", None)
+            if pattern in (None, 'none'):
+                continue
+
+            fg = getattr(fill, 'fgColor', None) or getattr(fill, 'start_color', None) or getattr(fill, 'bgColor', None)
+            color = resolve_color_hex_from_colorobj(fg)
+            if color is None or color.startswith("FF") is False:
+                continue
+
+            results.append({
+                'coordinate': cell.coordinate,
+                'row': cell.row,
+                'col': cell.column,
+                'value': cell.value,
+                'color': color,
+                'fill_type': pattern,
+            })
+
+    if write_csv:
+        keys = ['coordinate', 'row', 'col', 'value', 'color', 'fill_type']
+        with open(write_csv, 'w', newline='', encoding='utf-8') as f:
+            writer = csv.DictWriter(f, fieldnames=keys)
+            writer.writeheader()
+            for r in results:
+                writer.writerow(r)
+
+    return results
+
+
+if __name__ == '__main__':
+    import argparse
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--xlsx', '-x', required=True, help='Path to xlsx file')
+    parser.add_argument('--sheet', '-s', default=None, help='Sheet name (optional)')
+    parser.add_argument('--csv', '-c', default=None, help='Write results to CSV')
+    args = parser.parse_args()
+
+    res = detect_colored_cells(args.xlsx, sheet_name=args.sheet, write_csv=args.csv)
+    for r in res:
+        print(r)