initial webapp draft

cost_calculator.py

@@ -0,0 +1,87 @@
+# cost_calculator.py
+import os
+from PyPDF2 import PdfReader
+from pdf2image import convert_from_path
+import numpy as np
+
+UPLOAD_FOLDER = "data/uploads"
+ALLOWED_EXTENSIONS = {"pdf"}
+
+
+def allowed_file(filename: str) -> bool:
+    return "." in filename and filename.rsplit(".", 1)[1].lower() in ALLOWED_EXTENSIONS
+
+
+def points_to_meters(points: float) -> float:
+    return (points / 72.0) * 0.0254
+
+
+def get_rate_black() -> float:
+    return float(os.environ.get("RATE_PER_M2_BLACK", "4.0"))
+
+
+def get_rate_color() -> float:
+    return float(os.environ.get("RATE_PER_M2_COLOR", "5.0"))
+
+
+def analyze_pdf(path, compute_ink=True):
+    reader = PdfReader(path)
+    pages_info = []
+    total_area_black = total_area_color = 0.0
+    total_cost_black = total_cost_color = 0.0
+
+    for i, page in enumerate(reader.pages):
+        box = page.mediabox
+        width_m = points_to_meters(float(box.width))
+        height_m = points_to_meters(float(box.height))
+        area = width_m * height_m
+
+        ink_pct = None
+        is_color = False
+
+        if compute_ink:
+            try:
+                images = convert_from_path(path, first_page=i + 1, last_page=i + 1, dpi=150)
+                img = images[0].convert("RGB")
+                arr = np.array(img)
+
+                # ink pixels: any channel < 240
+                ink_mask = np.any(arr < 240, axis=2)
+                ink_pct = float(np.count_nonzero(ink_mask)) / (arr.shape[0] * arr.shape[1]) * 100.0
+
+                # simple color detection: if RGB channels differ significantly
+                avg_rgb = arr.mean(axis=(0, 1))
+                if np.ptp(avg_rgb) > 30:
+                    is_color = True
+            except Exception as e:
+                print(f"Page {i+1} ink/color calc failed: {e}")
+                ink_pct = None
+
+        if is_color:
+            rate = get_rate_color()
+            total_area_color += area
+            total_cost_color += area * rate
+        else:
+            rate = get_rate_black()
+            total_area_black += area
+            total_cost_black += area * rate
+
+        pages_info.append({
+            "page": i + 1,
+            "width_m": width_m,
+            "height_m": height_m,
+            "area_m2": area,
+            "ink_pct": ink_pct,
+            "is_color": is_color,
+            "cost": round(area * rate, 2)
+        })
+
+    return {
+        "filename": os.path.basename(path),
+        "pages": pages_info,
+        "total_area_black": total_area_black,
+        "total_area_color": total_area_color,
+        "total_cost_black": round(total_cost_black, 2),
+        "total_cost_color": round(total_cost_color, 2),
+        "grand_total": round(total_cost_black + total_cost_color, 2)
+    }