Initial commit

grouping.py

@@ -0,0 +1,248 @@
+import os
+import json
+import re
+import sys
+import shutil
+from collections import defaultdict
+from PIL import Image, ImageDraw, ImageFont
+from pdf2image import convert_from_path, pdfinfo_from_path
+
+# Configuration
+DPI = 200  # Good balance for readability and size
+A4_HEIGHT_INCHES = 11.69
+FULL_PAGE_PX = int(A4_HEIGHT_INCHES * DPI)
+MAX_GROUP_HEIGHT = 2.5 * FULL_PAGE_PX
+MAX_GROUP_COUNT = 15
+SEPARATOR_HEIGHT = 20
+LABEL_HEIGHT = 50
+MAX_FILE_SIZE_BYTES = 2.5 * 1024 * 1024  # 2MB
+
+
+# def get_pdf_height(path):
+#     """Returns height in pixels at defined DPI without rendering."""
+#     try:
+#         info = pdfinfo_from_path(path)
+#         # info["Page size"] is usually "width height pts"
+#         # 1 pt = 1/72 inch
+#         # We assume single page PDFs as per prompt implication, or take the first page
+#         pts_height = float(info['Page size'].split(' ')[2]) if 'Page size' in info else 0
+#         return int((pts_height / 72.0) * DPI)
+#     except Exception as e:
+#         print(f"Error reading {path}: {e}")
+#         return 0
+
+def get_pdf_height(path):
+    """Returns total height of all pages in pixels at defined DPI."""
+    try:
+        info = pdfinfo_from_path(path)
+        # Get page count (default to 1)
+        num_pages = int(info["Pages"]) if "Pages" in info else 1
+
+        # 1 pt = 1/72 inch
+        pts_height = float(info['Page size'].split(' ')[2]) if 'Page size' in info else 0
+
+        # Height of one page in pixels
+        single_page_px = int((pts_height / 72.0) * DPI)
+
+        # Return total height
+        return single_page_px * num_pages
+    except Exception as e:
+        print(f"Error reading {path}: {e}")
+        return 0
+
+def collect_files(root_dir):
+    """
+    Scans Dir/Copiedd/identifier.pdf
+    Returns dict: {identifier: [(dd, path, height), ...]}
+    """
+    data = defaultdict(list)
+
+    # Regex to match 'Copie' followed by 2 digits
+    folder_pattern = re.compile(r'Copie(\d{2})')
+
+    for root, dirs, files in os.walk(root_dir):
+        folder_name = os.path.basename(root)
+        match = folder_pattern.match(folder_name)
+
+        if match:
+            dd = match.group(1)
+            for file in files:
+                if file.lower().endswith('.pdf'):
+                    identifier = os.path.splitext(file)[0]
+                    full_path = os.path.join(root, file)
+
+                    # Calculate height (c)
+                    height = get_pdf_height(full_path)
+
+                    # Store triple (a, b, c)
+                    data[identifier].append((dd, full_path, height))
+    return data
+
+def group_files(file_list):
+    """Groups files based on constraints."""
+    sorted_files = sorted(file_list, key=lambda x: x[0])
+
+    groups = []
+    current_group = []
+    current_height = 0
+
+    for item in sorted_files:
+        dd, path, height = item
+
+        # Calculate added height (image + separator + approx text space)
+        # We add separator height only if it's not the first image
+        added_overhead = SEPARATOR_HEIGHT + 30 if current_group else 0
+
+        # Check conditions
+        if (len(current_group) >= MAX_GROUP_COUNT or
+            (current_height + height + added_overhead) > MAX_GROUP_HEIGHT):
+
+            # Push current group and start new
+            if current_group:
+                groups.append(current_group)
+            current_group = []
+            current_height = 0
+            added_overhead = 0 # Reset for first file of new group
+
+        current_group.append(item)
+        current_height += height + added_overhead
+
+    if current_group:
+        groups.append(current_group)
+
+    return groups
+
+def stitch_pdf_pages(images_list):
+    """Vertically concatenates a list of PIL images with no separator."""
+    if not images_list:
+        return None
+    if len(images_list) == 1:
+        return images_list[0]
+
+    max_width = max(img.width for img in images_list)
+    total_height = sum(img.height for img in images_list)
+
+    combined = Image.new('RGB', (max_width, total_height), 'white')
+
+    y_offset = 0
+    for img in images_list:
+        combined.paste(img, (0, y_offset))
+        y_offset += img.height
+
+    return combined
+
+def create_jpg(identifier, group_index, group, root_dir):
+    images = []
+    metadata = []  # To store (id, h_min, h_max)
+
+    # Render PDFs to images
+    for dd, path, _ in group:
+        try:
+            # Convert pdf to image
+            imgs = convert_from_path(path, dpi=DPI)
+            # if imgs:
+                # images.append((dd, imgs[0])) # Assume 1 page per pdf !! ??
+            if imgs:
+                # Concatenate multi-page PDFs into one single image object
+                combined_img = stitch_pdf_pages(imgs)
+                if combined_img:
+                    images.append((dd, combined_img))
+        except Exception as e:
+            print(f"Failed to convert {path}: {e}")
+
+    if not images:
+        return
+
+    # Calculate total canvas size
+    total_width = max(img.width for _, img in images)
+    total_height = sum(img.height for _, img in images) + ((len(images) - 1) * SEPARATOR_HEIGHT)
+
+    # Add space for text (approx 40px per label)
+    total_height += len(images) * LABEL_HEIGHT
+
+    canvas = Image.new('RGB', (total_width, total_height), 'white')
+    draw = ImageDraw.Draw(canvas)
+
+    # Try loading a font, fallback to default
+    try:
+        font = ImageFont.truetype("DejaVuSans.ttf", 40)
+    except IOError:
+        print("font not found")
+        font = ImageFont.load_default()
+
+    y_offset = 0
+
+    for i, (dd, img) in enumerate(images):
+        # Draw separator if not first image
+        if i > 0:
+            draw.rectangle([0, y_offset, total_width, y_offset + SEPARATOR_HEIGHT], fill='black')
+            y_offset += SEPARATOR_HEIGHT
+
+        # Draw Text (dd)
+        text = f"ID: {dd}"
+        draw.text((10, y_offset + 5), text, fill='black', font=font)
+        y_offset += LABEL_HEIGHT # Space for text
+
+        # Record Image Coordinates
+        h_min = y_offset
+        h_max = y_offset + img.height
+        metadata.append((dd, h_min, h_max))
+
+        # Draw Image
+        x_pos = 0
+        canvas.paste(img, (x_pos, y_offset))
+        y_offset += img.height
+
+    target_folder = os.path.join(root_dir, identifier)
+    os.makedirs(target_folder, exist_ok=True)
+
+    # Save JSON metadata
+    json_filename = f"Group_{group_index+1}.json"
+    json_path = os.path.join(target_folder, json_filename)
+    with open(json_path, 'w') as f:
+        json.dump(metadata, f)
+
+    # Save with size constraints
+    output_filename = f"Group_{group_index+1}.jpg"
+    output_path = os.path.join(target_folder, output_filename)
+
+    quality = 90
+    while quality > 10:
+        canvas.save(output_path, "JPEG", quality=quality, optimize=True)
+        if os.path.getsize(output_path) <= MAX_FILE_SIZE_BYTES:
+            if quality < 90:
+                print("quality : ", quality)
+            break
+        quality -= 5
+
+    print(f"Saved {output_path} ({os.path.getsize(output_path)/1024/1024:.2f} MB)")
+
+def main():
+    if len(sys.argv) < 2:
+        print("Usage: python app.py <Path_to_Dir>")
+        sys.exit(1)
+
+    root_dir = sys.argv[1]
+
+    print("Scanning files...")
+    data = collect_files(root_dir)
+
+    print(f"Found {len(data)} identifiers. Processing...")
+
+    for identifier, files_info in data.items():
+        # Clear output directory if it exists
+        target_folder = os.path.join(root_dir, identifier)
+        if os.path.exists(target_folder):
+            shutil.rmtree(target_folder)
+        os.makedirs(target_folder, exist_ok=True)
+
+        # files_info is list of (dd, path, height)
+        file_groups = group_files(files_info)
+
+        for idx, group in enumerate(file_groups):
+            create_jpg(identifier, idx, group, root_dir)
+
+    print("Done.")
+
+if __name__ == "__main__":
+    main()