{"owner":"ageerle","repo":"ruoyi-ai","hasSkills":true,"totalSkillsCount":5,"totalTokensCount":17049,"categories":["anthropic-skill"],"hasMcp":false,"mcpConfig":null,"found":["ruoyi-admin/src/main/resources/skills/pdf/forms.md","ruoyi-admin/src/main/resources/skills/pdf/reference.md","ruoyi-admin/src/main/resources/skills/docx/SKILL.md","ruoyi-admin/src/main/resources/skills/pdf/SKILL.md","ruoyi-admin/src/main/resources/skills/xlsx/SKILL.md"],"skills":{"ruoyi-admin/src/main/resources/skills/pdf/forms.md":"**CRITICAL: You MUST complete these steps in order. Do not skip ahead to writing code.**\n\nIf you need to fill out a PDF form, first check to see if the PDF has fillable form fields. Run this script from this file's directory:\n `python scripts/check_fillable_fields <file.pdf>`, and depending on the result go to either the \"Fillable fields\" or \"Non-fillable fields\" and follow those instructions.\n\n# Fillable fields\nIf the PDF has fillable form fields:\n- Run this script from this file's directory: `python scripts/extract_form_field_info.py <input.pdf> <field_info.json>`. It will create a JSON file with a list of fields in this format:\n```\n[\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"rect\": ([left, bottom, right, top] bounding box in PDF coordinates, y=0 is the bottom of the page),\n    \"type\": (\"text\", \"checkbox\", \"radio_group\", or \"choice\"),\n  },\n  // Checkboxes have \"checked_value\" and \"unchecked_value\" properties:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"checkbox\",\n    \"checked_value\": (Set the field to this value to check the checkbox),\n    \"unchecked_value\": (Set the field to this value to uncheck the checkbox),\n  },\n  // Radio groups have a \"radio_options\" list with the possible choices.\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"radio_group\",\n    \"radio_options\": [\n      {\n        \"value\": (set the field to this value to select this radio option),\n        \"rect\": (bounding box for the radio button for this option)\n      },\n      // Other radio options\n    ]\n  },\n  // Multiple choice fields have a \"choice_options\" list with the possible choices:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"choice\",\n    \"choice_options\": [\n      {\n        \"value\": (set the field to this value to select this option),\n        \"text\": (display text of the option)\n      },\n      // Other choice options\n    ],\n  }\n]\n```\n- Convert the PDF to PNGs (one image for each page) with this script (run from this file's directory):\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThen analyze the images to determine the purpose of each form field (make sure to convert the bounding box PDF coordinates to image coordinates).\n- Create a `field_values.json` file in this format with the values to be entered for each field:\n```\n[\n  {\n    \"field_id\": \"last_name\", // Must match the field_id from `extract_form_field_info.py`\n    \"description\": \"The user's last name\",\n    \"page\": 1, // Must match the \"page\" value in field_info.json\n    \"value\": \"Simpson\"\n  },\n  {\n    \"field_id\": \"Checkbox12\",\n    \"description\": \"Checkbox to be checked if the user is 18 or over\",\n    \"page\": 1,\n    \"value\": \"/On\" // If this is a checkbox, use its \"checked_value\" value to check it. If it's a radio button group, use one of the \"value\" values in \"radio_options\".\n  },\n  // more fields\n]\n```\n- Run the `fill_fillable_fields.py` script from this file's directory to create a filled-in PDF:\n`python scripts/fill_fillable_fields.py <input pdf> <field_values.json> <output pdf>`\nThis script will verify that the field IDs and values you provide are valid; if it prints error messages, correct the appropriate fields and try again.\n\n# Non-fillable fields\nIf the PDF doesn't have fillable form fields, you'll need to visually determine where the data should be added and create text annotations. Follow the below steps *exactly*. You MUST perform all of these steps to ensure that the the form is accurately completed. Details for each step are below.\n- Convert the PDF to PNG images and determine field bounding boxes.\n- Create a JSON file with field information and validation images showing the bounding boxes.\n- Validate the the bounding boxes.\n- Use the bounding boxes to fill in the form.\n\n## Step 1: Visual Analysis (REQUIRED)\n- Convert the PDF to PNG images. Run this script from this file's directory:\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThe script will create a PNG image for each page in the PDF.\n- Carefully examine each PNG image and identify all form fields and areas where the user should enter data. For each form field where the user should enter text, determine bounding boxes for both the form field label, and the area where the user should enter text. The label and entry bounding boxes MUST NOT INTERSECT; the text entry box should only include the area where data should be entered. Usually this area will be immediately to the side, above, or below its label. Entry bounding boxes must be tall and wide enough to contain their text.\n\nThese are some examples of form structures that you might see:\n\n*Label inside box*\n```\n┌────────────────────────┐\n│ Name:                  │\n└────────────────────────┘\n```\nThe input area should be to the right of the \"Name\" label and extend to the edge of the box.\n\n*Label before line*\n```\nEmail: _______________________\n```\nThe input area should be above the line and include its entire width.\n\n*Label under line*\n```\n_________________________\nName\n```\nThe input area should be above the line and include the entire width of the line. This is common for signature and date fields.\n\n*Label above line*\n```\nPlease enter any special requests:\n________________________________________________\n```\nThe input area should extend from the bottom of the label to the line, and should include the entire width of the line.\n\n*Checkboxes*\n```\nAre you a US citizen? Yes □  No □\n```\nFor checkboxes:\n- Look for small square boxes (□) - these are the actual checkboxes to target. They may be to the left or right of their labels.\n- Distinguish between label text (\"Yes\", \"No\") and the clickable checkbox squares.\n- The entry bounding box should cover ONLY the small square, not the text label.\n\n### Step 2: Create fields.json and validation images (REQUIRED)\n- Create a file named `fields.json` with information for the form fields and bounding boxes in this format:\n```\n{\n  \"pages\": [\n    {\n      \"page_number\": 1,\n      \"image_width\": (first page image width in pixels),\n      \"image_height\": (first page image height in pixels),\n    },\n    {\n      \"page_number\": 2,\n      \"image_width\": (second page image width in pixels),\n      \"image_height\": (second page image height in pixels),\n    }\n    // additional pages\n  ],\n  \"form_fields\": [\n    // Example for a text field.\n    {\n      \"page_number\": 1,\n      \"description\": \"The user's last name should be entered here\",\n      // Bounding boxes are [left, top, right, bottom]. The bounding boxes for the label and text entry should not overlap.\n      \"field_label\": \"Last name\",\n      \"label_bounding_box\": [30, 125, 95, 142],\n      \"entry_bounding_box\": [100, 125, 280, 142],\n      \"entry_text\": {\n        \"text\": \"Johnson\", // This text will be added as an annotation at the entry_bounding_box location\n        \"font_size\": 14, // optional, defaults to 14\n        \"font_color\": \"000000\", // optional, RRGGBB format, defaults to 000000 (black)\n      }\n    },\n    // Example for a checkbox. TARGET THE SQUARE for the entry bounding box, NOT THE TEXT\n    {\n      \"page_number\": 2,\n      \"description\": \"Checkbox that should be checked if the user is over 18\",\n      \"entry_bounding_box\": [140, 525, 155, 540],  // Small box over checkbox square\n      \"field_label\": \"Yes\",\n      \"label_bounding_box\": [100, 525, 132, 540],  // Box containing \"Yes\" text\n      // Use \"X\" to check a checkbox.\n      \"entry_text\": {\n        \"text\": \"X\",\n      }\n    }\n    // additional form field entries\n  ]\n}\n```\n\nCreate validation images by running this script from this file's directory for each page:\n`python scripts/create_validation_image.py <page_number> <path_to_fields.json> <input_image_path> <output_image_path>\n\nThe validation images will have red rectangles where text should be entered, and blue rectangles covering label text.\n\n### Step 3: Validate Bounding Boxes (REQUIRED)\n#### Automated intersection check\n- Verify that none of bounding boxes intersect and that the entry bounding boxes are tall enough by checking the fields.json file with the `check_bounding_boxes.py` script (run from this file's directory):\n`python scripts/check_bounding_boxes.py <JSON file>`\n\nIf there are errors, reanalyze the relevant fields, adjust the bounding boxes, and iterate until there are no remaining errors. Remember: label (blue) bounding boxes should contain text labels, entry (red) boxes should not.\n\n#### Manual image inspection\n**CRITICAL: Do not proceed without visually inspecting validation images**\n- Red rectangles must ONLY cover input areas\n- Red rectangles MUST NOT contain any text\n- Blue rectangles should contain label text\n- For checkboxes:\n  - Red rectangle MUST be centered on the checkbox square\n  - Blue rectangle should cover the text label for the checkbox\n\n- If any rectangles look wrong, fix fields.json, regenerate the validation images, and verify again. Repeat this process until the bounding boxes are fully accurate.\n\n\n### Step 4: Add annotations to the PDF\nRun this script from this file's directory to create a filled-out PDF using the information in fields.json:\n`python scripts/fill_pdf_form_with_annotations.py <input_pdf_path> <path_to_fields.json> <output_pdf_path>\n","ruoyi-admin/src/main/resources/skills/pdf/reference.md":"# PDF Processing Advanced Reference\n\nThis document contains advanced PDF processing features, detailed examples, and additional libraries not covered in the main skill instructions.\n\n## pypdfium2 Library (Apache/BSD License)\n\n### Overview\npypdfium2 is a Python binding for PDFium (Chromium's PDF library). It's excellent for fast PDF rendering, image generation, and serves as a PyMuPDF replacement.\n\n### Render PDF to Images\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\n\n# Load PDF\npdf = pdfium.PdfDocument(\"document.pdf\")\n\n# Render page to image\npage = pdf[0]  # First page\nbitmap = page.render(\n    scale=2.0,  # Higher resolution\n    rotation=0  # No rotation\n)\n\n# Convert to PIL Image\nimg = bitmap.to_pil()\nimg.save(\"page_1.png\", \"PNG\")\n\n# Process multiple pages\nfor i, page in enumerate(pdf):\n    bitmap = page.render(scale=1.5)\n    img = bitmap.to_pil()\n    img.save(f\"page_{i+1}.jpg\", \"JPEG\", quality=90)\n```\n\n### Extract Text with pypdfium2\n```python\nimport pypdfium2 as pdfium\n\npdf = pdfium.PdfDocument(\"document.pdf\")\nfor i, page in enumerate(pdf):\n    text = page.get_text()\n    print(f\"Page {i+1} text length: {len(text)} chars\")\n```\n\n## JavaScript Libraries\n\n### pdf-lib (MIT License)\n\npdf-lib is a powerful JavaScript library for creating and modifying PDF documents in any JavaScript environment.\n\n#### Load and Manipulate Existing PDF\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function manipulatePDF() {\n    // Load existing PDF\n    const existingPdfBytes = fs.readFileSync('input.pdf');\n    const pdfDoc = await PDFDocument.load(existingPdfBytes);\n\n    // Get page count\n    const pageCount = pdfDoc.getPageCount();\n    console.log(`Document has ${pageCount} pages`);\n\n    // Add new page\n    const newPage = pdfDoc.addPage([600, 400]);\n    newPage.drawText('Added by pdf-lib', {\n        x: 100,\n        y: 300,\n        size: 16\n    });\n\n    // Save modified PDF\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('modified.pdf', pdfBytes);\n}\n```\n\n#### Create Complex PDFs from Scratch\n```javascript\nimport { PDFDocument, rgb, StandardFonts } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function createPDF() {\n    const pdfDoc = await PDFDocument.create();\n\n    // Add fonts\n    const helveticaFont = await pdfDoc.embedFont(StandardFonts.Helvetica);\n    const helveticaBold = await pdfDoc.embedFont(StandardFonts.HelveticaBold);\n\n    // Add page\n    const page = pdfDoc.addPage([595, 842]); // A4 size\n    const { width, height } = page.getSize();\n\n    // Add text with styling\n    page.drawText('Invoice #12345', {\n        x: 50,\n        y: height - 50,\n        size: 18,\n        font: helveticaBold,\n        color: rgb(0.2, 0.2, 0.8)\n    });\n\n    // Add rectangle (header background)\n    page.drawRectangle({\n        x: 40,\n        y: height - 100,\n        width: width - 80,\n        height: 30,\n        color: rgb(0.9, 0.9, 0.9)\n    });\n\n    // Add table-like content\n    const items = [\n        ['Item', 'Qty', 'Price', 'Total'],\n        ['Widget', '2', '$50', '$100'],\n        ['Gadget', '1', '$75', '$75']\n    ];\n\n    let yPos = height - 150;\n    items.forEach(row => {\n        let xPos = 50;\n        row.forEach(cell => {\n            page.drawText(cell, {\n                x: xPos,\n                y: yPos,\n                size: 12,\n                font: helveticaFont\n            });\n            xPos += 120;\n        });\n        yPos -= 25;\n    });\n\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('created.pdf', pdfBytes);\n}\n```\n\n#### Advanced Merge and Split Operations\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function mergePDFs() {\n    // Create new document\n    const mergedPdf = await PDFDocument.create();\n\n    // Load source PDFs\n    const pdf1Bytes = fs.readFileSync('doc1.pdf');\n    const pdf2Bytes = fs.readFileSync('doc2.pdf');\n\n    const pdf1 = await PDFDocument.load(pdf1Bytes);\n    const pdf2 = await PDFDocument.load(pdf2Bytes);\n\n    // Copy pages from first PDF\n    const pdf1Pages = await mergedPdf.copyPages(pdf1, pdf1.getPageIndices());\n    pdf1Pages.forEach(page => mergedPdf.addPage(page));\n\n    // Copy specific pages from second PDF (pages 0, 2, 4)\n    const pdf2Pages = await mergedPdf.copyPages(pdf2, [0, 2, 4]);\n    pdf2Pages.forEach(page => mergedPdf.addPage(page));\n\n    const mergedPdfBytes = await mergedPdf.save();\n    fs.writeFileSync('merged.pdf', mergedPdfBytes);\n}\n```\n\n### pdfjs-dist (Apache License)\n\nPDF.js is Mozilla's JavaScript library for rendering PDFs in the browser.\n\n#### Basic PDF Loading and Rendering\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\n// Configure worker (important for performance)\npdfjsLib.GlobalWorkerOptions.workerSrc = './pdf.worker.js';\n\nasync function renderPDF() {\n    // Load PDF\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    console.log(`Loaded PDF with ${pdf.numPages} pages`);\n\n    // Get first page\n    const page = await pdf.getPage(1);\n    const viewport = page.getViewport({ scale: 1.5 });\n\n    // Render to canvas\n    const canvas = document.createElement('canvas');\n    const context = canvas.getContext('2d');\n    canvas.height = viewport.height;\n    canvas.width = viewport.width;\n\n    const renderContext = {\n        canvasContext: context,\n        viewport: viewport\n    };\n\n    await page.render(renderContext).promise;\n    document.body.appendChild(canvas);\n}\n```\n\n#### Extract Text with Coordinates\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractText() {\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    let fullText = '';\n\n    // Extract text from all pages\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const textContent = await page.getTextContent();\n\n        const pageText = textContent.items\n            .map(item => item.str)\n            .join(' ');\n\n        fullText += `\\n--- Page ${i} ---\\n${pageText}`;\n\n        // Get text with coordinates for advanced processing\n        const textWithCoords = textContent.items.map(item => ({\n            text: item.str,\n            x: item.transform[4],\n            y: item.transform[5],\n            width: item.width,\n            height: item.height\n        }));\n    }\n\n    console.log(fullText);\n    return fullText;\n}\n```\n\n#### Extract Annotations and Forms\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractAnnotations() {\n    const loadingTask = pdfjsLib.getDocument('annotated.pdf');\n    const pdf = await loadingTask.promise;\n\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const annotations = await page.getAnnotations();\n\n        annotations.forEach(annotation => {\n            console.log(`Annotation type: ${annotation.subtype}`);\n            console.log(`Content: ${annotation.contents}`);\n            console.log(`Coordinates: ${JSON.stringify(annotation.rect)}`);\n        });\n    }\n}\n```\n\n## Advanced Command-Line Operations\n\n### poppler-utils Advanced Features\n\n#### Extract Text with Bounding Box Coordinates\n```bash\n# Extract text with bounding box coordinates (essential for structured data)\npdftotext -bbox-layout document.pdf output.xml\n\n# The XML output contains precise coordinates for each text element\n```\n\n#### Advanced Image Conversion\n```bash\n# Convert to PNG images with specific resolution\npdftoppm -png -r 300 document.pdf output_prefix\n\n# Convert specific page range with high resolution\npdftoppm -png -r 600 -f 1 -l 3 document.pdf high_res_pages\n\n# Convert to JPEG with quality setting\npdftoppm -jpeg -jpegopt quality=85 -r 200 document.pdf jpeg_output\n```\n\n#### Extract Embedded Images\n```bash\n# Extract all embedded images with metadata\npdfimages -j -p document.pdf page_images\n\n# List image info without extracting\npdfimages -list document.pdf\n\n# Extract images in their original format\npdfimages -all document.pdf images/img\n```\n\n### qpdf Advanced Features\n\n#### Complex Page Manipulation\n```bash\n# Split PDF into groups of pages\nqpdf --split-pages=3 input.pdf output_group_%02d.pdf\n\n# Extract specific pages with complex ranges\nqpdf input.pdf --pages input.pdf 1,3-5,8,10-end -- extracted.pdf\n\n# Merge specific pages from multiple PDFs\nqpdf --empty --pages doc1.pdf 1-3 doc2.pdf 5-7 doc3.pdf 2,4 -- combined.pdf\n```\n\n#### PDF Optimization and Repair\n```bash\n# Optimize PDF for web (linearize for streaming)\nqpdf --linearize input.pdf optimized.pdf\n\n# Remove unused objects and compress\nqpdf --optimize-level=all input.pdf compressed.pdf\n\n# Attempt to repair corrupted PDF structure\nqpdf --check input.pdf\nqpdf --fix-qdf damaged.pdf repaired.pdf\n\n# Show detailed PDF structure for debugging\nqpdf --show-all-pages input.pdf > structure.txt\n```\n\n#### Advanced Encryption\n```bash\n# Add password protection with specific permissions\nqpdf --encrypt user_pass owner_pass 256 --print=none --modify=none -- input.pdf encrypted.pdf\n\n# Check encryption status\nqpdf --show-encryption encrypted.pdf\n\n# Remove password protection (requires password)\nqpdf --password=secret123 --decrypt encrypted.pdf decrypted.pdf\n```\n\n## Advanced Python Techniques\n\n### pdfplumber Advanced Features\n\n#### Extract Text with Precise Coordinates\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract all text with coordinates\n    chars = page.chars\n    for char in chars[:10]:  # First 10 characters\n        print(f\"Char: '{char['text']}' at x:{char['x0']:.1f} y:{char['y0']:.1f}\")\n    \n    # Extract text by bounding box (left, top, right, bottom)\n    bbox_text = page.within_bbox((100, 100, 400, 200)).extract_text()\n```\n\n#### Advanced Table Extraction with Custom Settings\n```python\nimport pdfplumber\nimport pandas as pd\n\nwith pdfplumber.open(\"complex_table.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract tables with custom settings for complex layouts\n    table_settings = {\n        \"vertical_strategy\": \"lines\",\n        \"horizontal_strategy\": \"lines\",\n        \"snap_tolerance\": 3,\n        \"intersection_tolerance\": 15\n    }\n    tables = page.extract_tables(table_settings)\n    \n    # Visual debugging for table extraction\n    img = page.to_image(resolution=150)\n    img.save(\"debug_layout.png\")\n```\n\n### reportlab Advanced Features\n\n#### Create Professional Reports with Tables\n```python\nfrom reportlab.platypus import SimpleDocTemplate, Table, TableStyle, Paragraph\nfrom reportlab.lib.styles import getSampleStyleSheet\nfrom reportlab.lib import colors\n\n# Sample data\ndata = [\n    ['Product', 'Q1', 'Q2', 'Q3', 'Q4'],\n    ['Widgets', '120', '135', '142', '158'],\n    ['Gadgets', '85', '92', '98', '105']\n]\n\n# Create PDF with table\ndoc = SimpleDocTemplate(\"report.pdf\")\nelements = []\n\n# Add title\nstyles = getSampleStyleSheet()\ntitle = Paragraph(\"Quarterly Sales Report\", styles['Title'])\nelements.append(title)\n\n# Add table with advanced styling\ntable = Table(data)\ntable.setStyle(TableStyle([\n    ('BACKGROUND', (0, 0), (-1, 0), colors.grey),\n    ('TEXTCOLOR', (0, 0), (-1, 0), colors.whitesmoke),\n    ('ALIGN', (0, 0), (-1, -1), 'CENTER'),\n    ('FONTNAME', (0, 0), (-1, 0), 'Helvetica-Bold'),\n    ('FONTSIZE', (0, 0), (-1, 0), 14),\n    ('BOTTOMPADDING', (0, 0), (-1, 0), 12),\n    ('BACKGROUND', (0, 1), (-1, -1), colors.beige),\n    ('GRID', (0, 0), (-1, -1), 1, colors.black)\n]))\nelements.append(table)\n\ndoc.build(elements)\n```\n\n## Complex Workflows\n\n### Extract Figures/Images from PDF\n\n#### Method 1: Using pdfimages (fastest)\n```bash\n# Extract all images with original quality\npdfimages -all document.pdf images/img\n```\n\n#### Method 2: Using pypdfium2 + Image Processing\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\nimport numpy as np\n\ndef extract_figures(pdf_path, output_dir):\n    pdf = pdfium.PdfDocument(pdf_path)\n    \n    for page_num, page in enumerate(pdf):\n        # Render high-resolution page\n        bitmap = page.render(scale=3.0)\n        img = bitmap.to_pil()\n        \n        # Convert to numpy for processing\n        img_array = np.array(img)\n        \n        # Simple figure detection (non-white regions)\n        mask = np.any(img_array != [255, 255, 255], axis=2)\n        \n        # Find contours and extract bounding boxes\n        # (This is simplified - real implementation would need more sophisticated detection)\n        \n        # Save detected figures\n        # ... implementation depends on specific needs\n```\n\n### Batch PDF Processing with Error Handling\n```python\nimport os\nimport glob\nfrom pypdf import PdfReader, PdfWriter\nimport logging\n\nlogging.basicConfig(level=logging.INFO)\nlogger = logging.getLogger(__name__)\n\ndef batch_process_pdfs(input_dir, operation='merge'):\n    pdf_files = glob.glob(os.path.join(input_dir, \"*.pdf\"))\n    \n    if operation == 'merge':\n        writer = PdfWriter()\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                for page in reader.pages:\n                    writer.add_page(page)\n                logger.info(f\"Processed: {pdf_file}\")\n            except Exception as e:\n                logger.error(f\"Failed to process {pdf_file}: {e}\")\n                continue\n        \n        with open(\"batch_merged.pdf\", \"wb\") as output:\n            writer.write(output)\n    \n    elif operation == 'extract_text':\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                text = \"\"\n                for page in reader.pages:\n                    text += page.extract_text()\n                \n                output_file = pdf_file.replace('.pdf', '.txt')\n                with open(output_file, 'w', encoding='utf-8') as f:\n                    f.write(text)\n                logger.info(f\"Extracted text from: {pdf_file}\")\n                \n            except Exception as e:\n                logger.error(f\"Failed to extract text from {pdf_file}: {e}\")\n                continue\n```\n\n### Advanced PDF Cropping\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\n# Crop page (left, bottom, right, top in points)\npage = reader.pages[0]\npage.mediabox.left = 50\npage.mediabox.bottom = 50\npage.mediabox.right = 550\npage.mediabox.top = 750\n\nwriter.add_page(page)\nwith open(\"cropped.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Performance Optimization Tips\n\n### 1. For Large PDFs\n- Use streaming approaches instead of loading entire PDF in memory\n- Use `qpdf --split-pages` for splitting large files\n- Process pages individually with pypdfium2\n\n### 2. For Text Extraction\n- `pdftotext -bbox-layout` is fastest for plain text extraction\n- Use pdfplumber for structured data and tables\n- Avoid `pypdf.extract_text()` for very large documents\n\n### 3. For Image Extraction\n- `pdfimages` is much faster than rendering pages\n- Use low resolution for previews, high resolution for final output\n\n### 4. For Form Filling\n- pdf-lib maintains form structure better than most alternatives\n- Pre-validate form fields before processing\n\n### 5. Memory Management\n```python\n# Process PDFs in chunks\ndef process_large_pdf(pdf_path, chunk_size=10):\n    reader = PdfReader(pdf_path)\n    total_pages = len(reader.pages)\n    \n    for start_idx in range(0, total_pages, chunk_size):\n        end_idx = min(start_idx + chunk_size, total_pages)\n        writer = PdfWriter()\n        \n        for i in range(start_idx, end_idx):\n            writer.add_page(reader.pages[i])\n        \n        # Process chunk\n        with open(f\"chunk_{start_idx//chunk_size}.pdf\", \"wb\") as output:\n            writer.write(output)\n```\n\n## Troubleshooting Common Issues\n\n### Encrypted PDFs\n```python\n# Handle password-protected PDFs\nfrom pypdf import PdfReader\n\ntry:\n    reader = PdfReader(\"encrypted.pdf\")\n    if reader.is_encrypted:\n        reader.decrypt(\"password\")\nexcept Exception as e:\n    print(f\"Failed to decrypt: {e}\")\n```\n\n### Corrupted PDFs\n```bash\n# Use qpdf to repair\nqpdf --check corrupted.pdf\nqpdf --replace-input corrupted.pdf\n```\n\n### Text Extraction Issues\n```python\n# Fallback to OCR for scanned PDFs\nimport pytesseract\nfrom pdf2image import convert_from_path\n\ndef extract_text_with_ocr(pdf_path):\n    images = convert_from_path(pdf_path)\n    text = \"\"\n    for i, image in enumerate(images):\n        text += pytesseract.image_to_string(image)\n    return text\n```\n\n## License Information\n\n- **pypdf**: BSD License\n- **pdfplumber**: MIT License\n- **pypdfium2**: Apache/BSD License\n- **reportlab**: BSD License\n- **poppler-utils**: GPL-2 License\n- **qpdf**: Apache License\n- **pdf-lib**: MIT License\n- **pdfjs-dist**: Apache License","ruoyi-admin/src/main/resources/skills/docx/SKILL.md":"---\nname: docx\ndescription: \"Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# DOCX creation, editing, and analysis\n\n## Overview\n\nA .docx file is a ZIP archive containing XML files.\n\n## Quick Reference\n\n| Task | Approach |\n|------|----------|\n| Read/analyze content | `pandoc` or unpack for raw XML |\n| Create new document | Use `docx-js` - see Creating New Documents below |\n| Edit existing document | Unpack → edit XML → repack - see Editing Existing Documents below |\n\n### Converting .doc to .docx\n\nLegacy `.doc` files must be converted before editing:\n\n```bash\npython scripts/office/soffice.py --headless --convert-to docx document.doc\n```\n\n### Reading Content\n\n```bash\n# Text extraction with tracked changes\npandoc --track-changes=all document.docx -o output.md\n\n# Raw XML access\npython scripts/office/unpack.py document.docx unpacked/\n```\n\n### Converting to Images\n\n```bash\npython scripts/office/soffice.py --headless --convert-to pdf document.docx\npdftoppm -jpeg -r 150 document.pdf page\n```\n\n### Accepting Tracked Changes\n\nTo produce a clean document with all tracked changes accepted (requires LibreOffice):\n\n```bash\npython scripts/accept_changes.py input.docx output.docx\n```\n\n---\n\n## Creating New Documents\n\nGenerate .docx files with JavaScript, then validate. Install: `npm install -g docx`\n\n### Setup\n```javascript\nconst { Document, Packer, Paragraph, TextRun, Table, TableRow, TableCell, ImageRun,\n        Header, Footer, AlignmentType, PageOrientation, LevelFormat, ExternalHyperlink,\n        InternalHyperlink, Bookmark, FootnoteReferenceRun, PositionalTab,\n        PositionalTabAlignment, PositionalTabRelativeTo, PositionalTabLeader,\n        TabStopType, TabStopPosition, Column, SectionType,\n        TableOfContents, HeadingLevel, BorderStyle, WidthType, ShadingType,\n        VerticalAlign, PageNumber, PageBreak } = require('docx');\n\nconst doc = new Document({ sections: [{ children: [/* content */] }] });\nPacker.toBuffer(doc).then(buffer => fs.writeFileSync(\"doc.docx\", buffer));\n```\n\n### Validation\nAfter creating the file, validate it. If validation fails, unpack, fix the XML, and repack.\n```bash\npython scripts/office/validate.py doc.docx\n```\n\n### Page Size\n\n```javascript\n// CRITICAL: docx-js defaults to A4, not US Letter\n// Always set page size explicitly for consistent results\nsections: [{\n  properties: {\n    page: {\n      size: {\n        width: 12240,   // 8.5 inches in DXA\n        height: 15840   // 11 inches in DXA\n      },\n      margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } // 1 inch margins\n    }\n  },\n  children: [/* content */]\n}]\n```\n\n**Common page sizes (DXA units, 1440 DXA = 1 inch):**\n\n| Paper | Width | Height | Content Width (1\" margins) |\n|-------|-------|--------|---------------------------|\n| US Letter | 12,240 | 15,840 | 9,360 |\n| A4 (default) | 11,906 | 16,838 | 9,026 |\n\n**Landscape orientation:** docx-js swaps width/height internally, so pass portrait dimensions and let it handle the swap:\n```javascript\nsize: {\n  width: 12240,   // Pass SHORT edge as width\n  height: 15840,  // Pass LONG edge as height\n  orientation: PageOrientation.LANDSCAPE  // docx-js swaps them in the XML\n},\n// Content width = 15840 - left margin - right margin (uses the long edge)\n```\n\n### Styles (Override Built-in Headings)\n\nUse Arial as the default font (universally supported). Keep titles black for readability.\n\n```javascript\nconst doc = new Document({\n  styles: {\n    default: { document: { run: { font: \"Arial\", size: 24 } } }, // 12pt default\n    paragraphStyles: [\n      // IMPORTANT: Use exact IDs to override built-in styles\n      { id: \"Heading1\", name: \"Heading 1\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 32, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 240, after: 240 }, outlineLevel: 0 } }, // outlineLevel required for TOC\n      { id: \"Heading2\", name: \"Heading 2\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 28, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 180, after: 180 }, outlineLevel: 1 } },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ heading: HeadingLevel.HEADING_1, children: [new TextRun(\"Title\")] }),\n    ]\n  }]\n});\n```\n\n### Lists (NEVER use unicode bullets)\n\n```javascript\n// ❌ WRONG - never manually insert bullet characters\nnew Paragraph({ children: [new TextRun(\"• Item\")] })  // BAD\nnew Paragraph({ children: [new TextRun(\"\\u2022 Item\")] })  // BAD\n\n// ✅ CORRECT - use numbering config with LevelFormat.BULLET\nconst doc = new Document({\n  numbering: {\n    config: [\n      { reference: \"bullets\",\n        levels: [{ level: 0, format: LevelFormat.BULLET, text: \"•\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n      { reference: \"numbers\",\n        levels: [{ level: 0, format: LevelFormat.DECIMAL, text: \"%1.\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ numbering: { reference: \"bullets\", level: 0 },\n        children: [new TextRun(\"Bullet item\")] }),\n      new Paragraph({ numbering: { reference: \"numbers\", level: 0 },\n        children: [new TextRun(\"Numbered item\")] }),\n    ]\n  }]\n});\n\n// ⚠️ Each reference creates INDEPENDENT numbering\n// Same reference = continues (1,2,3 then 4,5,6)\n// Different reference = restarts (1,2,3 then 1,2,3)\n```\n\n### Tables\n\n**CRITICAL: Tables need dual widths** - set both `columnWidths` on the table AND `width` on each cell. Without both, tables render incorrectly on some platforms.\n\n```javascript\n// CRITICAL: Always set table width for consistent rendering\n// CRITICAL: Use ShadingType.CLEAR (not SOLID) to prevent black backgrounds\nconst border = { style: BorderStyle.SINGLE, size: 1, color: \"CCCCCC\" };\nconst borders = { top: border, bottom: border, left: border, right: border };\n\nnew Table({\n  width: { size: 9360, type: WidthType.DXA }, // Always use DXA (percentages break in Google Docs)\n  columnWidths: [4680, 4680], // Must sum to table width (DXA: 1440 = 1 inch)\n  rows: [\n    new TableRow({\n      children: [\n        new TableCell({\n          borders,\n          width: { size: 4680, type: WidthType.DXA }, // Also set on each cell\n          shading: { fill: \"D5E8F0\", type: ShadingType.CLEAR }, // CLEAR not SOLID\n          margins: { top: 80, bottom: 80, left: 120, right: 120 }, // Cell padding (internal, not added to width)\n          children: [new Paragraph({ children: [new TextRun(\"Cell\")] })]\n        })\n      ]\n    })\n  ]\n})\n```\n\n**Table width calculation:**\n\nAlways use `WidthType.DXA` — `WidthType.PERCENTAGE` breaks in Google Docs.\n\n```javascript\n// Table width = sum of columnWidths = content width\n// US Letter with 1\" margins: 12240 - 2880 = 9360 DXA\nwidth: { size: 9360, type: WidthType.DXA },\ncolumnWidths: [7000, 2360]  // Must sum to table width\n```\n\n**Width rules:**\n- **Always use `WidthType.DXA`** — never `WidthType.PERCENTAGE` (incompatible with Google Docs)\n- Table width must equal the sum of `columnWidths`\n- Cell `width` must match corresponding `columnWidth`\n- Cell `margins` are internal padding - they reduce content area, not add to cell width\n- For full-width tables: use content width (page width minus left and right margins)\n\n### Images\n\n```javascript\n// CRITICAL: type parameter is REQUIRED\nnew Paragraph({\n  children: [new ImageRun({\n    type: \"png\", // Required: png, jpg, jpeg, gif, bmp, svg\n    data: fs.readFileSync(\"image.png\"),\n    transformation: { width: 200, height: 150 },\n    altText: { title: \"Title\", description: \"Desc\", name: \"Name\" } // All three required\n  })]\n})\n```\n\n### Page Breaks\n\n```javascript\n// CRITICAL: PageBreak must be inside a Paragraph\nnew Paragraph({ children: [new PageBreak()] })\n\n// Or use pageBreakBefore\nnew Paragraph({ pageBreakBefore: true, children: [new TextRun(\"New page\")] })\n```\n\n### Hyperlinks\n\n```javascript\n// External link\nnew Paragraph({\n  children: [new ExternalHyperlink({\n    children: [new TextRun({ text: \"Click here\", style: \"Hyperlink\" })],\n    link: \"https://example.com\",\n  })]\n})\n\n// Internal link (bookmark + reference)\n// 1. Create bookmark at destination\nnew Paragraph({ heading: HeadingLevel.HEADING_1, children: [\n  new Bookmark({ id: \"chapter1\", children: [new TextRun(\"Chapter 1\")] }),\n]})\n// 2. Link to it\nnew Paragraph({ children: [new InternalHyperlink({\n  children: [new TextRun({ text: \"See Chapter 1\", style: \"Hyperlink\" })],\n  anchor: \"chapter1\",\n})]})\n```\n\n### Footnotes\n\n```javascript\nconst doc = new Document({\n  footnotes: {\n    1: { children: [new Paragraph(\"Source: Annual Report 2024\")] },\n    2: { children: [new Paragraph(\"See appendix for methodology\")] },\n  },\n  sections: [{\n    children: [new Paragraph({\n      children: [\n        new TextRun(\"Revenue grew 15%\"),\n        new FootnoteReferenceRun(1),\n        new TextRun(\" using adjusted metrics\"),\n        new FootnoteReferenceRun(2),\n      ],\n    })]\n  }]\n});\n```\n\n### Tab Stops\n\n```javascript\n// Right-align text on same line (e.g., date opposite a title)\nnew Paragraph({\n  children: [\n    new TextRun(\"Company Name\"),\n    new TextRun(\"\\tJanuary 2025\"),\n  ],\n  tabStops: [{ type: TabStopType.RIGHT, position: TabStopPosition.MAX }],\n})\n\n// Dot leader (e.g., TOC-style)\nnew Paragraph({\n  children: [\n    new TextRun(\"Introduction\"),\n    new TextRun({ children: [\n      new PositionalTab({\n        alignment: PositionalTabAlignment.RIGHT,\n        relativeTo: PositionalTabRelativeTo.MARGIN,\n        leader: PositionalTabLeader.DOT,\n      }),\n      \"3\",\n    ]}),\n  ],\n})\n```\n\n### Multi-Column Layouts\n\n```javascript\n// Equal-width columns\nsections: [{\n  properties: {\n    column: {\n      count: 2,          // number of columns\n      space: 720,        // gap between columns in DXA (720 = 0.5 inch)\n      equalWidth: true,\n      separate: true,    // vertical line between columns\n    },\n  },\n  children: [/* content flows naturally across columns */]\n}]\n\n// Custom-width columns (equalWidth must be false)\nsections: [{\n  properties: {\n    column: {\n      equalWidth: false,\n      children: [\n        new Column({ width: 5400, space: 720 }),\n        new Column({ width: 3240 }),\n      ],\n    },\n  },\n  children: [/* content */]\n}]\n```\n\nForce a column break with a new section using `type: SectionType.NEXT_COLUMN`.\n\n### Table of Contents\n\n```javascript\n// CRITICAL: Headings must use HeadingLevel ONLY - no custom styles\nnew TableOfContents(\"Table of Contents\", { hyperlink: true, headingStyleRange: \"1-3\" })\n```\n\n### Headers/Footers\n\n```javascript\nsections: [{\n  properties: {\n    page: { margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } } // 1440 = 1 inch\n  },\n  headers: {\n    default: new Header({ children: [new Paragraph({ children: [new TextRun(\"Header\")] })] })\n  },\n  footers: {\n    default: new Footer({ children: [new Paragraph({\n      children: [new TextRun(\"Page \"), new TextRun({ children: [PageNumber.CURRENT] })]\n    })] })\n  },\n  children: [/* content */]\n}]\n```\n\n### Critical Rules for docx-js\n\n- **Set page size explicitly** - docx-js defaults to A4; use US Letter (12240 x 15840 DXA) for US documents\n- **Landscape: pass portrait dimensions** - docx-js swaps width/height internally; pass short edge as `width`, long edge as `height`, and set `orientation: PageOrientation.LANDSCAPE`\n- **Never use `\\n`** - use separate Paragraph elements\n- **Never use unicode bullets** - use `LevelFormat.BULLET` with numbering config\n- **PageBreak must be in Paragraph** - standalone creates invalid XML\n- **ImageRun requires `type`** - always specify png/jpg/etc\n- **Always set table `width` with DXA** - never use `WidthType.PERCENTAGE` (breaks in Google Docs)\n- **Tables need dual widths** - `columnWidths` array AND cell `width`, both must match\n- **Table width = sum of columnWidths** - for DXA, ensure they add up exactly\n- **Always add cell margins** - use `margins: { top: 80, bottom: 80, left: 120, right: 120 }` for readable padding\n- **Use `ShadingType.CLEAR`** - never SOLID for table shading\n- **Never use tables as dividers/rules** - cells have minimum height and render as empty boxes (including in headers/footers); use `border: { bottom: { style: BorderStyle.SINGLE, size: 6, color: \"2E75B6\", space: 1 } }` on a Paragraph instead. For two-column footers, use tab stops (see Tab Stops section), not tables\n- **TOC requires HeadingLevel only** - no custom styles on heading paragraphs\n- **Override built-in styles** - use exact IDs: \"Heading1\", \"Heading2\", etc.\n- **Include `outlineLevel`** - required for TOC (0 for H1, 1 for H2, etc.)\n\n---\n\n## Editing Existing Documents\n\n**Follow all 3 steps in order.**\n\n### Step 1: Unpack\n```bash\npython scripts/office/unpack.py document.docx unpacked/\n```\nExtracts XML, pretty-prints, merges adjacent runs, and converts smart quotes to XML entities (`&#x201C;` etc.) so they survive editing. Use `--merge-runs false` to skip run merging.\n\n### Step 2: Edit XML\n\nEdit files in `unpacked/word/`. See XML Reference below for patterns.\n\n**Use \"Claude\" as the author** for tracked changes and comments, unless the user explicitly requests use of a different name.\n\n**Use the Edit tool directly for string replacement. Do not write Python scripts.** Scripts introduce unnecessary complexity. The Edit tool shows exactly what is being replaced.\n\n**CRITICAL: Use smart quotes for new content.** When adding text with apostrophes or quotes, use XML entities to produce smart quotes:\n```xml\n<!-- Use these entities for professional typography -->\n<w:t>Here&#x2019;s a quote: &#x201C;Hello&#x201D;</w:t>\n```\n| Entity | Character |\n|--------|-----------|\n| `&#x2018;` | ‘ (left single) |\n| `&#x2019;` | ’ (right single / apostrophe) |\n| `&#x201C;` | “ (left double) |\n| `&#x201D;` | ” (right double) |\n\n**Adding comments:** Use `comment.py` to handle boilerplate across multiple XML files (text must be pre-escaped XML):\n```bash\npython scripts/comment.py unpacked/ 0 \"Comment text with &amp; and &#x2019;\"\npython scripts/comment.py unpacked/ 1 \"Reply text\" --parent 0  # reply to comment 0\npython scripts/comment.py unpacked/ 0 \"Text\" --author \"Custom Author\"  # custom author name\n```\nThen add markers to document.xml (see Comments in XML Reference).\n\n### Step 3: Pack\n```bash\npython scripts/office/pack.py unpacked/ output.docx --original document.docx\n```\nValidates with auto-repair, condenses XML, and creates DOCX. Use `--validate false` to skip.\n\n**Auto-repair will fix:**\n- `durableId` >= 0x7FFFFFFF (regenerates valid ID)\n- Missing `xml:space=\"preserve\"` on `<w:t>` with whitespace\n\n**Auto-repair won't fix:**\n- Malformed XML, invalid element nesting, missing relationships, schema violations\n\n### Common Pitfalls\n\n- **Replace entire `<w:r>` elements**: When adding tracked changes, replace the whole `<w:r>...</w:r>` block with `<w:del>...<w:ins>...` as siblings. Don't inject tracked change tags inside a run.\n- **Preserve `<w:rPr>` formatting**: Copy the original run's `<w:rPr>` block into your tracked change runs to maintain bold, font size, etc.\n\n---\n\n## XML Reference\n\n### Schema Compliance\n\n- **Element order in `<w:pPr>`**: `<w:pStyle>`, `<w:numPr>`, `<w:spacing>`, `<w:ind>`, `<w:jc>`, `<w:rPr>` last\n- **Whitespace**: Add `xml:space=\"preserve\"` to `<w:t>` with leading/trailing spaces\n- **RSIDs**: Must be 8-digit hex (e.g., `00AB1234`)\n\n### Tracked Changes\n\n**Insertion:**\n```xml\n<w:ins w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:t>inserted text</w:t></w:r>\n</w:ins>\n```\n\n**Deletion:**\n```xml\n<w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n```\n\n**Inside `<w:del>`**: Use `<w:delText>` instead of `<w:t>`, and `<w:delInstrText>` instead of `<w:instrText>`.\n\n**Minimal edits** - only mark what changes:\n```xml\n<!-- Change \"30 days\" to \"60 days\" -->\n<w:r><w:t>The term is </w:t></w:r>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:delText>30</w:delText></w:r>\n</w:del>\n<w:ins w:id=\"2\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:t>60</w:t></w:r>\n</w:ins>\n<w:r><w:t> days.</w:t></w:r>\n```\n\n**Deleting entire paragraphs/list items** - when removing ALL content from a paragraph, also mark the paragraph mark as deleted so it merges with the next paragraph. Add `<w:del/>` inside `<w:pPr><w:rPr>`:\n```xml\n<w:p>\n  <w:pPr>\n    <w:numPr>...</w:numPr>  <!-- list numbering if present -->\n    <w:rPr>\n      <w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\"/>\n    </w:rPr>\n  </w:pPr>\n  <w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n    <w:r><w:delText>Entire paragraph content being deleted...</w:delText></w:r>\n  </w:del>\n</w:p>\n```\nWithout the `<w:del/>` in `<w:pPr><w:rPr>`, accepting changes leaves an empty paragraph/list item.\n\n**Rejecting another author's insertion** - nest deletion inside their insertion:\n```xml\n<w:ins w:author=\"Jane\" w:id=\"5\">\n  <w:del w:author=\"Claude\" w:id=\"10\">\n    <w:r><w:delText>their inserted text</w:delText></w:r>\n  </w:del>\n</w:ins>\n```\n\n**Restoring another author's deletion** - add insertion after (don't modify their deletion):\n```xml\n<w:del w:author=\"Jane\" w:id=\"5\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n<w:ins w:author=\"Claude\" w:id=\"10\">\n  <w:r><w:t>deleted text</w:t></w:r>\n</w:ins>\n```\n\n### Comments\n\nAfter running `comment.py` (see Step 2), add markers to document.xml. For replies, use `--parent` flag and nest markers inside the parent's.\n\n**CRITICAL: `<w:commentRangeStart>` and `<w:commentRangeEnd>` are siblings of `<w:r>`, never inside `<w:r>`.**\n\n```xml\n<!-- Comment markers are direct children of w:p, never inside w:r -->\n<w:commentRangeStart w:id=\"0\"/>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted</w:delText></w:r>\n</w:del>\n<w:r><w:t> more text</w:t></w:r>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n\n<!-- Comment 0 with reply 1 nested inside -->\n<w:commentRangeStart w:id=\"0\"/>\n  <w:commentRangeStart w:id=\"1\"/>\n  <w:r><w:t>text</w:t></w:r>\n  <w:commentRangeEnd w:id=\"1\"/>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"1\"/></w:r>\n```\n\n### Images\n\n1. Add image file to `word/media/`\n2. Add relationship to `word/_rels/document.xml.rels`:\n```xml\n<Relationship Id=\"rId5\" Type=\".../image\" Target=\"media/image1.png\"/>\n```\n3. Add content type to `[Content_Types].xml`:\n```xml\n<Default Extension=\"png\" ContentType=\"image/png\"/>\n```\n4. Reference in document.xml:\n```xml\n<w:drawing>\n  <wp:inline>\n    <wp:extent cx=\"914400\" cy=\"914400\"/>  <!-- EMUs: 914400 = 1 inch -->\n    <a:graphic>\n      <a:graphicData uri=\".../picture\">\n        <pic:pic>\n          <pic:blipFill><a:blip r:embed=\"rId5\"/></pic:blipFill>\n        </pic:pic>\n      </a:graphicData>\n    </a:graphic>\n  </wp:inline>\n</w:drawing>\n```\n\n---\n\n## Dependencies\n\n- **pandoc**: Text extraction\n- **docx**: `npm install -g docx` (new documents)\n- **LibreOffice**: PDF conversion (auto-configured for sandboxed environments via `scripts/office/soffice.py`)\n- **Poppler**: `pdftoppm` for images\n","ruoyi-admin/src/main/resources/skills/pdf/SKILL.md":"---\nname: pdf\ndescription: Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# PDF Processing Guide\n\n## Overview\n\nThis guide covers essential PDF processing operations using Python libraries and command-line tools. For advanced features, JavaScript libraries, and detailed examples, see reference.md. If you need to fill out a PDF form, read forms.md and follow its instructions.\n\n## Quick Start\n\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Read a PDF\nreader = PdfReader(\"document.pdf\")\nprint(f\"Pages: {len(reader.pages)}\")\n\n# Extract text\ntext = \"\"\nfor page in reader.pages:\n    text += page.extract_text()\n```\n\n## Python Libraries\n\n### pypdf - Basic Operations\n\n#### Merge PDFs\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nwriter = PdfWriter()\nfor pdf_file in [\"doc1.pdf\", \"doc2.pdf\", \"doc3.pdf\"]:\n    reader = PdfReader(pdf_file)\n    for page in reader.pages:\n        writer.add_page(page)\n\nwith open(\"merged.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n#### Split PDF\n```python\nreader = PdfReader(\"input.pdf\")\nfor i, page in enumerate(reader.pages):\n    writer = PdfWriter()\n    writer.add_page(page)\n    with open(f\"page_{i+1}.pdf\", \"wb\") as output:\n        writer.write(output)\n```\n\n#### Extract Metadata\n```python\nreader = PdfReader(\"document.pdf\")\nmeta = reader.metadata\nprint(f\"Title: {meta.title}\")\nprint(f\"Author: {meta.author}\")\nprint(f\"Subject: {meta.subject}\")\nprint(f\"Creator: {meta.creator}\")\n```\n\n#### Rotate Pages\n```python\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\npage = reader.pages[0]\npage.rotate(90)  # Rotate 90 degrees clockwise\nwriter.add_page(page)\n\nwith open(\"rotated.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### pdfplumber - Text and Table Extraction\n\n#### Extract Text with Layout\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for page in pdf.pages:\n        text = page.extract_text()\n        print(text)\n```\n\n#### Extract Tables\n```python\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for i, page in enumerate(pdf.pages):\n        tables = page.extract_tables()\n        for j, table in enumerate(tables):\n            print(f\"Table {j+1} on page {i+1}:\")\n            for row in table:\n                print(row)\n```\n\n#### Advanced Table Extraction\n```python\nimport pandas as pd\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    all_tables = []\n    for page in pdf.pages:\n        tables = page.extract_tables()\n        for table in tables:\n            if table:  # Check if table is not empty\n                df = pd.DataFrame(table[1:], columns=table[0])\n                all_tables.append(df)\n\n# Combine all tables\nif all_tables:\n    combined_df = pd.concat(all_tables, ignore_index=True)\n    combined_df.to_excel(\"extracted_tables.xlsx\", index=False)\n```\n\n### reportlab - Create PDFs\n\n#### Basic PDF Creation\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.pdfgen import canvas\n\nc = canvas.Canvas(\"hello.pdf\", pagesize=letter)\nwidth, height = letter\n\n# Add text\nc.drawString(100, height - 100, \"Hello World!\")\nc.drawString(100, height - 120, \"This is a PDF created with reportlab\")\n\n# Add a line\nc.line(100, height - 140, 400, height - 140)\n\n# Save\nc.save()\n```\n\n#### Create PDF with Multiple Pages\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, PageBreak\nfrom reportlab.lib.styles import getSampleStyleSheet\n\ndoc = SimpleDocTemplate(\"report.pdf\", pagesize=letter)\nstyles = getSampleStyleSheet()\nstory = []\n\n# Add content\ntitle = Paragraph(\"Report Title\", styles['Title'])\nstory.append(title)\nstory.append(Spacer(1, 12))\n\nbody = Paragraph(\"This is the body of the report. \" * 20, styles['Normal'])\nstory.append(body)\nstory.append(PageBreak())\n\n# Page 2\nstory.append(Paragraph(\"Page 2\", styles['Heading1']))\nstory.append(Paragraph(\"Content for page 2\", styles['Normal']))\n\n# Build PDF\ndoc.build(story)\n```\n\n## Command-Line Tools\n\n### pdftotext (poppler-utils)\n```bash\n# Extract text\npdftotext input.pdf output.txt\n\n# Extract text preserving layout\npdftotext -layout input.pdf output.txt\n\n# Extract specific pages\npdftotext -f 1 -l 5 input.pdf output.txt  # Pages 1-5\n```\n\n### qpdf\n```bash\n# Merge PDFs\nqpdf --empty --pages file1.pdf file2.pdf -- merged.pdf\n\n# Split pages\nqpdf input.pdf --pages . 1-5 -- pages1-5.pdf\nqpdf input.pdf --pages . 6-10 -- pages6-10.pdf\n\n# Rotate pages\nqpdf input.pdf output.pdf --rotate=+90:1  # Rotate page 1 by 90 degrees\n\n# Remove password\nqpdf --password=mypassword --decrypt encrypted.pdf decrypted.pdf\n```\n\n### pdftk (if available)\n```bash\n# Merge\npdftk file1.pdf file2.pdf cat output merged.pdf\n\n# Split\npdftk input.pdf burst\n\n# Rotate\npdftk input.pdf rotate 1east output rotated.pdf\n```\n\n## Common Tasks\n\n### Extract Text from Scanned PDFs\n```python\n# Requires: pip install pytesseract pdf2image\nimport pytesseract\nfrom pdf2image import convert_from_path\n\n# Convert PDF to images\nimages = convert_from_path('scanned.pdf')\n\n# OCR each page\ntext = \"\"\nfor i, image in enumerate(images):\n    text += f\"Page {i+1}:\\n\"\n    text += pytesseract.image_to_string(image)\n    text += \"\\n\\n\"\n\nprint(text)\n```\n\n### Add Watermark\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Create watermark (or load existing)\nwatermark = PdfReader(\"watermark.pdf\").pages[0]\n\n# Apply to all pages\nreader = PdfReader(\"document.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    page.merge_page(watermark)\n    writer.add_page(page)\n\nwith open(\"watermarked.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### Extract Images\n```bash\n# Using pdfimages (poppler-utils)\npdfimages -j input.pdf output_prefix\n\n# This extracts all images as output_prefix-000.jpg, output_prefix-001.jpg, etc.\n```\n\n### Password Protection\n```python\nfrom pypdf import PdfReader, PdfWriter\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    writer.add_page(page)\n\n# Add password\nwriter.encrypt(\"userpassword\", \"ownerpassword\")\n\nwith open(\"encrypted.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Quick Reference\n\n| Task | Best Tool | Command/Code |\n|------|-----------|--------------|\n| Merge PDFs | pypdf | `writer.add_page(page)` |\n| Split PDFs | pypdf | One page per file |\n| Extract text | pdfplumber | `page.extract_text()` |\n| Extract tables | pdfplumber | `page.extract_tables()` |\n| Create PDFs | reportlab | Canvas or Platypus |\n| Command line merge | qpdf | `qpdf --empty --pages ...` |\n| OCR scanned PDFs | pytesseract | Convert to image first |\n| Fill PDF forms | pdf-lib or pypdf (see forms.md) | See forms.md |\n\n## Next Steps\n\n- For advanced pypdfium2 usage, see reference.md\n- For JavaScript libraries (pdf-lib), see reference.md\n- If you need to fill out a PDF form, follow the instructions in forms.md\n- For troubleshooting guides, see reference.md\n","ruoyi-admin/src/main/resources/skills/xlsx/SKILL.md":"---\nname: xlsx\ndescription: \"Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# Requirements for Outputs\n\n## All Excel files\n\n### Zero Formula Errors\n- Every Excel model MUST be delivered with ZERO formula errors (#REF!, #DIV/0!, #VALUE!, #N/A, #NAME?)\n\n### Preserve Existing Templates (when updating templates)\n- Study and EXACTLY match existing format, style, and conventions when modifying files\n- Never impose standardized formatting on files with established patterns\n- Existing template conventions ALWAYS override these guidelines\n\n## Financial models\n\n### Color Coding Standards\nUnless otherwise stated by the user or existing template\n\n#### Industry-Standard Color Conventions\n- **Blue text (RGB: 0,0,255)**: Hardcoded inputs, and numbers users will change for scenarios\n- **Black text (RGB: 0,0,0)**: ALL formulas and calculations\n- **Green text (RGB: 0,128,0)**: Links pulling from other worksheets within same workbook\n- **Red text (RGB: 255,0,0)**: External links to other files\n- **Yellow background (RGB: 255,255,0)**: Key assumptions needing attention or cells that need to be updated\n\n### Number Formatting Standards\n\n#### Required Format Rules\n- **Years**: Format as text strings (e.g., \"2024\" not \"2,024\")\n- **Currency**: Use $#,##0 format; ALWAYS specify units in headers (\"Revenue ($mm)\")\n- **Zeros**: Use number formatting to make all zeros \"-\", including percentages (e.g., \"$#,##0;($#,##0);-\")\n- **Percentages**: Default to 0.0% format (one decimal)\n- **Multiples**: Format as 0.0x for valuation multiples (EV/EBITDA, P/E)\n- **Negative numbers**: Use parentheses (123) not minus -123\n\n### Formula Construction Rules\n\n#### Assumptions Placement\n- Place ALL assumptions (growth rates, margins, multiples, etc.) in separate assumption cells\n- Use cell references instead of hardcoded values in formulas\n- Example: Use =B5*(1+$B$6) instead of =B5*1.05\n\n#### Formula Error Prevention\n- Verify all cell references are correct\n- Check for off-by-one errors in ranges\n- Ensure consistent formulas across all projection periods\n- Test with edge cases (zero values, negative numbers)\n- Verify no unintended circular references\n\n#### Documentation Requirements for Hardcodes\n- Comment or in cells beside (if end of table). Format: \"Source: [System/Document], [Date], [Specific Reference], [URL if applicable]\"\n- Examples:\n  - \"Source: Company 10-K, FY2024, Page 45, Revenue Note, [SEC EDGAR URL]\"\n  - \"Source: Company 10-Q, Q2 2025, Exhibit 99.1, [SEC EDGAR URL]\"\n  - \"Source: Bloomberg Terminal, 8/15/2025, AAPL US Equity\"\n  - \"Source: FactSet, 8/20/2025, Consensus Estimates Screen\"\n\n# XLSX creation, editing, and analysis\n\n## Overview\n\nA user may ask you to create, edit, or analyze the contents of an .xlsx file. You have different tools and workflows available for different tasks.\n\n## Important Requirements\n\n**LibreOffice Required for Formula Recalculation**: You can assume LibreOffice is installed for recalculating formula values using the `recalc.py` script. The script automatically configures LibreOffice on first run\n\n## Reading and analyzing data\n\n### Data analysis with pandas\nFor data analysis, visualization, and basic operations, use **pandas** which provides powerful data manipulation capabilities:\n\n```python\nimport pandas as pd\n\n# Read Excel\ndf = pd.read_excel('file.xlsx')  # Default: first sheet\nall_sheets = pd.read_excel('file.xlsx', sheet_name=None)  # All sheets as dict\n\n# Analyze\ndf.head()      # Preview data\ndf.info()      # Column info\ndf.describe()  # Statistics\n\n# Write Excel\ndf.to_excel('output.xlsx', index=False)\n```\n\n## Excel File Workflows\n\n## CRITICAL: Use Formulas, Not Hardcoded Values\n\n**Always use Excel formulas instead of calculating values in Python and hardcoding them.** This ensures the spreadsheet remains dynamic and updateable.\n\n### ❌ WRONG - Hardcoding Calculated Values\n```python\n# Bad: Calculating in Python and hardcoding result\ntotal = df['Sales'].sum()\nsheet['B10'] = total  # Hardcodes 5000\n\n# Bad: Computing growth rate in Python\ngrowth = (df.iloc[-1]['Revenue'] - df.iloc[0]['Revenue']) / df.iloc[0]['Revenue']\nsheet['C5'] = growth  # Hardcodes 0.15\n\n# Bad: Python calculation for average\navg = sum(values) / len(values)\nsheet['D20'] = avg  # Hardcodes 42.5\n```\n\n### ✅ CORRECT - Using Excel Formulas\n```python\n# Good: Let Excel calculate the sum\nsheet['B10'] = '=SUM(B2:B9)'\n\n# Good: Growth rate as Excel formula\nsheet['C5'] = '=(C4-C2)/C2'\n\n# Good: Average using Excel function\nsheet['D20'] = '=AVERAGE(D2:D19)'\n```\n\nThis applies to ALL calculations - totals, percentages, ratios, differences, etc. The spreadsheet should be able to recalculate when source data changes.\n\n## Common Workflow\n1. **Choose tool**: pandas for data, openpyxl for formulas/formatting\n2. **Create/Load**: Create new workbook or load existing file\n3. **Modify**: Add/edit data, formulas, and formatting\n4. **Save**: Write to file\n5. **Recalculate formulas (MANDATORY IF USING FORMULAS)**: Use the recalc.py script\n   ```bash\n   python recalc.py output.xlsx\n   ```\n6. **Verify and fix any errors**: \n   - The script returns JSON with error details\n   - If `status` is `errors_found`, check `error_summary` for specific error types and locations\n   - Fix the identified errors and recalculate again\n   - Common errors to fix:\n     - `#REF!`: Invalid cell references\n     - `#DIV/0!`: Division by zero\n     - `#VALUE!`: Wrong data type in formula\n     - `#NAME?`: Unrecognized formula name\n\n### Creating new Excel files\n\n```python\n# Using openpyxl for formulas and formatting\nfrom openpyxl import Workbook\nfrom openpyxl.styles import Font, PatternFill, Alignment\n\nwb = Workbook()\nsheet = wb.active\n\n# Add data\nsheet['A1'] = 'Hello'\nsheet['B1'] = 'World'\nsheet.append(['Row', 'of', 'data'])\n\n# Add formula\nsheet['B2'] = '=SUM(A1:A10)'\n\n# Formatting\nsheet['A1'].font = Font(bold=True, color='FF0000')\nsheet['A1'].fill = PatternFill('solid', start_color='FFFF00')\nsheet['A1'].alignment = Alignment(horizontal='center')\n\n# Column width\nsheet.column_dimensions['A'].width = 20\n\nwb.save('output.xlsx')\n```\n\n### Editing existing Excel files\n\n```python\n# Using openpyxl to preserve formulas and formatting\nfrom openpyxl import load_workbook\n\n# Load existing file\nwb = load_workbook('existing.xlsx')\nsheet = wb.active  # or wb['SheetName'] for specific sheet\n\n# Working with multiple sheets\nfor sheet_name in wb.sheetnames:\n    sheet = wb[sheet_name]\n    print(f\"Sheet: {sheet_name}\")\n\n# Modify cells\nsheet['A1'] = 'New Value'\nsheet.insert_rows(2)  # Insert row at position 2\nsheet.delete_cols(3)  # Delete column 3\n\n# Add new sheet\nnew_sheet = wb.create_sheet('NewSheet')\nnew_sheet['A1'] = 'Data'\n\nwb.save('modified.xlsx')\n```\n\n## Recalculating formulas\n\nExcel files created or modified by openpyxl contain formulas as strings but not calculated values. Use the provided `recalc.py` script to recalculate formulas:\n\n```bash\n# Recalculate existing Excel file\npython recalc.py <excel_file> [timeout_seconds]\n\n# Create new Excel workbook\npython recalc.py --create <output_file> '<json_data>'\n```\n\nExamples:\n```bash\npython recalc.py output.xlsx 30\n\npython recalc.py --create output.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Value\"], [\"Test\", 100]]}}'\n```\n\nThe script:\n- Automatically sets up LibreOffice macro on first run\n- Recalculates all formulas in all sheets\n- Scans ALL cells for Excel errors (#REF!, #DIV/0!, etc.)\n- Returns JSON with detailed error locations and counts\n- Works on Windows, Linux, and macOS\n- Can create new workbooks from JSON configuration\n\n## Creating workbooks with recalc.py\n\nThe `--create` mode allows you to create new Excel workbooks directly from JSON configuration:\n\n```bash\npython recalc.py --create <output_file> '<json_data>'\n```\n\n**Default Behavior**: If you provide a relative path (e.g., `test.xlsx`), the file will be created in the **project root** directory (`ruoyi-ai-v3/`). To save to a specific directory, use an absolute path.\n\n### JSON Format\n\n```json\n{\n  \"Sheet1\": {\n    \"data\": [\n      [\"Header1\", \"Header2\", \"Header3\"],\n      [\"Value1\", \"Value2\", \"=A2*B2\"],\n      [\"Value3\", \"Value4\", \"=SUM(C2:C10)\"]\n    ]\n  },\n  \"Sheet2\": {\n    \"data\": [\n      [\"Name\", \"Score\"],\n      [\"Alice\", 95],\n      [\"Bob\", 87]\n    ]\n  }\n}\n```\n\n### Examples\n\n**Simple data table (saves to project root)**:\n```bash\npython recalc.py --create data.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Age\"], [\"John\", 30], [\"Jane\", 28]]}}'\n# Creates: ruoyi-ai-v3/data.xlsx\n```\n\n**With formulas (saves to project root)**:\n```bash\npython recalc.py --create model.xlsx '{\"Sheet1\": {\"data\": [[\"A\", \"B\", \"Sum\"], [10, 20, \"=A2+B2\"], [5, 15, \"=A3+B3\"]]}}'\n# Creates: ruoyi-ai-v3/model.xlsx\n```\n\n**Save to custom directory (use absolute path)**:\n```bash\npython recalc.py --create \"D:\\Downloads\\custom.xlsx\" '{\"Sheet1\": {\"data\": [[\"A\", \"B\"], [1, 2]]}}'\n# Creates: D:\\Downloads\\custom.xlsx\n```\n\n### File Path Options\n\n| Input | Output Location |\n|-------|-----------------|\n| `test.xlsx` | `ruoyi-ai-v3/test.xlsx` (project root) |\n| `output/test.xlsx` | `ruoyi-ai-v3/output/test.xlsx` |\n| `D:\\Downloads\\test.xlsx` | `D:\\Downloads\\test.xlsx` (absolute path) |\n\n### Features\n\n- First row is automatically formatted as bold header\n- Supports both values and Excel formulas (strings starting with `=`)\n- Multiple sheets can be created in one command\n- Output file extension determines format (.xlsx, .xlsm, etc.)\n\n## Formula Verification Checklist\n\nQuick checks to ensure formulas work correctly:\n\n### Essential Verification\n- [ ] **Test 2-3 sample references**: Verify they pull correct values before building full model\n- [ ] **Column mapping**: Confirm Excel columns match (e.g., column 64 = BL, not BK)\n- [ ] **Row offset**: Remember Excel rows are 1-indexed (DataFrame row 5 = Excel row 6)\n\n### Common Pitfalls\n- [ ] **NaN handling**: Check for null values with `pd.notna()`\n- [ ] **Far-right columns**: FY data often in columns 50+ \n- [ ] **Multiple matches**: Search all occurrences, not just first\n- [ ] **Division by zero**: Check denominators before using `/` in formulas (#DIV/0!)\n- [ ] **Wrong references**: Verify all cell references point to intended cells (#REF!)\n- [ ] **Cross-sheet references**: Use correct format (Sheet1!A1) for linking sheets\n\n### Formula Testing Strategy\n- [ ] **Start small**: Test formulas on 2-3 cells before applying broadly\n- [ ] **Verify dependencies**: Check all cells referenced in formulas exist\n- [ ] **Test edge cases**: Include zero, negative, and very large values\n\n### Interpreting recalc.py Output\nThe script returns JSON with error details:\n```json\n{\n  \"status\": \"success\",           // or \"errors_found\"\n  \"total_errors\": 0,              // Total error count\n  \"total_formulas\": 42,           // Number of formulas in file\n  \"error_summary\": {              // Only present if errors found\n    \"#REF!\": {\n      \"count\": 2,\n      \"locations\": [\"Sheet1!B5\", \"Sheet1!C10\"]\n    }\n  }\n}\n```\n\n## Best Practices\n\n### Library Selection\n- **pandas**: Best for data analysis, bulk operations, and simple data export\n- **openpyxl**: Best for complex formatting, formulas, and Excel-specific features\n\n### Working with openpyxl\n- Cell indices are 1-based (row=1, column=1 refers to cell A1)\n- Use `data_only=True` to read calculated values: `load_workbook('file.xlsx', data_only=True)`\n- **Warning**: If opened with `data_only=True` and saved, formulas are replaced with values and permanently lost\n- For large files: Use `read_only=True` for reading or `write_only=True` for writing\n- Formulas are preserved but not evaluated - use recalc.py to update values\n\n### Working with pandas\n- Specify data types to avoid inference issues: `pd.read_excel('file.xlsx', dtype={'id': str})`\n- For large files, read specific columns: `pd.read_excel('file.xlsx', usecols=['A', 'C', 'E'])`\n- Handle dates properly: `pd.read_excel('file.xlsx', parse_dates=['date_column'])`\n\n## Code Style Guidelines\n**IMPORTANT**: When generating Python code for Excel operations:\n- Write minimal, concise Python code without unnecessary comments\n- Avoid verbose variable names and redundant operations\n- Avoid unnecessary print statements\n\n**For Excel files themselves**:\n- Add comments to cells with complex formulas or important assumptions\n- Document data sources for hardcoded values\n- Include notes for key calculations and model sections"},"files":{"ruoyi-admin/src/main/resources/skills/pdf/forms.md":"**CRITICAL: You MUST complete these steps in order. Do not skip ahead to writing code.**\n\nIf you need to fill out a PDF form, first check to see if the PDF has fillable form fields. Run this script from this file's directory:\n `python scripts/check_fillable_fields <file.pdf>`, and depending on the result go to either the \"Fillable fields\" or \"Non-fillable fields\" and follow those instructions.\n\n# Fillable fields\nIf the PDF has fillable form fields:\n- Run this script from this file's directory: `python scripts/extract_form_field_info.py <input.pdf> <field_info.json>`. It will create a JSON file with a list of fields in this format:\n```\n[\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"rect\": ([left, bottom, right, top] bounding box in PDF coordinates, y=0 is the bottom of the page),\n    \"type\": (\"text\", \"checkbox\", \"radio_group\", or \"choice\"),\n  },\n  // Checkboxes have \"checked_value\" and \"unchecked_value\" properties:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"checkbox\",\n    \"checked_value\": (Set the field to this value to check the checkbox),\n    \"unchecked_value\": (Set the field to this value to uncheck the checkbox),\n  },\n  // Radio groups have a \"radio_options\" list with the possible choices.\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"radio_group\",\n    \"radio_options\": [\n      {\n        \"value\": (set the field to this value to select this radio option),\n        \"rect\": (bounding box for the radio button for this option)\n      },\n      // Other radio options\n    ]\n  },\n  // Multiple choice fields have a \"choice_options\" list with the possible choices:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"choice\",\n    \"choice_options\": [\n      {\n        \"value\": (set the field to this value to select this option),\n        \"text\": (display text of the option)\n      },\n      // Other choice options\n    ],\n  }\n]\n```\n- Convert the PDF to PNGs (one image for each page) with this script (run from this file's directory):\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThen analyze the images to determine the purpose of each form field (make sure to convert the bounding box PDF coordinates to image coordinates).\n- Create a `field_values.json` file in this format with the values to be entered for each field:\n```\n[\n  {\n    \"field_id\": \"last_name\", // Must match the field_id from `extract_form_field_info.py`\n    \"description\": \"The user's last name\",\n    \"page\": 1, // Must match the \"page\" value in field_info.json\n    \"value\": \"Simpson\"\n  },\n  {\n    \"field_id\": \"Checkbox12\",\n    \"description\": \"Checkbox to be checked if the user is 18 or over\",\n    \"page\": 1,\n    \"value\": \"/On\" // If this is a checkbox, use its \"checked_value\" value to check it. If it's a radio button group, use one of the \"value\" values in \"radio_options\".\n  },\n  // more fields\n]\n```\n- Run the `fill_fillable_fields.py` script from this file's directory to create a filled-in PDF:\n`python scripts/fill_fillable_fields.py <input pdf> <field_values.json> <output pdf>`\nThis script will verify that the field IDs and values you provide are valid; if it prints error messages, correct the appropriate fields and try again.\n\n# Non-fillable fields\nIf the PDF doesn't have fillable form fields, you'll need to visually determine where the data should be added and create text annotations. Follow the below steps *exactly*. You MUST perform all of these steps to ensure that the the form is accurately completed. Details for each step are below.\n- Convert the PDF to PNG images and determine field bounding boxes.\n- Create a JSON file with field information and validation images showing the bounding boxes.\n- Validate the the bounding boxes.\n- Use the bounding boxes to fill in the form.\n\n## Step 1: Visual Analysis (REQUIRED)\n- Convert the PDF to PNG images. Run this script from this file's directory:\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThe script will create a PNG image for each page in the PDF.\n- Carefully examine each PNG image and identify all form fields and areas where the user should enter data. For each form field where the user should enter text, determine bounding boxes for both the form field label, and the area where the user should enter text. The label and entry bounding boxes MUST NOT INTERSECT; the text entry box should only include the area where data should be entered. Usually this area will be immediately to the side, above, or below its label. Entry bounding boxes must be tall and wide enough to contain their text.\n\nThese are some examples of form structures that you might see:\n\n*Label inside box*\n```\n┌────────────────────────┐\n│ Name:                  │\n└────────────────────────┘\n```\nThe input area should be to the right of the \"Name\" label and extend to the edge of the box.\n\n*Label before line*\n```\nEmail: _______________________\n```\nThe input area should be above the line and include its entire width.\n\n*Label under line*\n```\n_________________________\nName\n```\nThe input area should be above the line and include the entire width of the line. This is common for signature and date fields.\n\n*Label above line*\n```\nPlease enter any special requests:\n________________________________________________\n```\nThe input area should extend from the bottom of the label to the line, and should include the entire width of the line.\n\n*Checkboxes*\n```\nAre you a US citizen? Yes □  No □\n```\nFor checkboxes:\n- Look for small square boxes (□) - these are the actual checkboxes to target. They may be to the left or right of their labels.\n- Distinguish between label text (\"Yes\", \"No\") and the clickable checkbox squares.\n- The entry bounding box should cover ONLY the small square, not the text label.\n\n### Step 2: Create fields.json and validation images (REQUIRED)\n- Create a file named `fields.json` with information for the form fields and bounding boxes in this format:\n```\n{\n  \"pages\": [\n    {\n      \"page_number\": 1,\n      \"image_width\": (first page image width in pixels),\n      \"image_height\": (first page image height in pixels),\n    },\n    {\n      \"page_number\": 2,\n      \"image_width\": (second page image width in pixels),\n      \"image_height\": (second page image height in pixels),\n    }\n    // additional pages\n  ],\n  \"form_fields\": [\n    // Example for a text field.\n    {\n      \"page_number\": 1,\n      \"description\": \"The user's last name should be entered here\",\n      // Bounding boxes are [left, top, right, bottom]. The bounding boxes for the label and text entry should not overlap.\n      \"field_label\": \"Last name\",\n      \"label_bounding_box\": [30, 125, 95, 142],\n      \"entry_bounding_box\": [100, 125, 280, 142],\n      \"entry_text\": {\n        \"text\": \"Johnson\", // This text will be added as an annotation at the entry_bounding_box location\n        \"font_size\": 14, // optional, defaults to 14\n        \"font_color\": \"000000\", // optional, RRGGBB format, defaults to 000000 (black)\n      }\n    },\n    // Example for a checkbox. TARGET THE SQUARE for the entry bounding box, NOT THE TEXT\n    {\n      \"page_number\": 2,\n      \"description\": \"Checkbox that should be checked if the user is over 18\",\n      \"entry_bounding_box\": [140, 525, 155, 540],  // Small box over checkbox square\n      \"field_label\": \"Yes\",\n      \"label_bounding_box\": [100, 525, 132, 540],  // Box containing \"Yes\" text\n      // Use \"X\" to check a checkbox.\n      \"entry_text\": {\n        \"text\": \"X\",\n      }\n    }\n    // additional form field entries\n  ]\n}\n```\n\nCreate validation images by running this script from this file's directory for each page:\n`python scripts/create_validation_image.py <page_number> <path_to_fields.json> <input_image_path> <output_image_path>\n\nThe validation images will have red rectangles where text should be entered, and blue rectangles covering label text.\n\n### Step 3: Validate Bounding Boxes (REQUIRED)\n#### Automated intersection check\n- Verify that none of bounding boxes intersect and that the entry bounding boxes are tall enough by checking the fields.json file with the `check_bounding_boxes.py` script (run from this file's directory):\n`python scripts/check_bounding_boxes.py <JSON file>`\n\nIf there are errors, reanalyze the relevant fields, adjust the bounding boxes, and iterate until there are no remaining errors. Remember: label (blue) bounding boxes should contain text labels, entry (red) boxes should not.\n\n#### Manual image inspection\n**CRITICAL: Do not proceed without visually inspecting validation images**\n- Red rectangles must ONLY cover input areas\n- Red rectangles MUST NOT contain any text\n- Blue rectangles should contain label text\n- For checkboxes:\n  - Red rectangle MUST be centered on the checkbox square\n  - Blue rectangle should cover the text label for the checkbox\n\n- If any rectangles look wrong, fix fields.json, regenerate the validation images, and verify again. Repeat this process until the bounding boxes are fully accurate.\n\n\n### Step 4: Add annotations to the PDF\nRun this script from this file's directory to create a filled-out PDF using the information in fields.json:\n`python scripts/fill_pdf_form_with_annotations.py <input_pdf_path> <path_to_fields.json> <output_pdf_path>\n","ruoyi-admin/src/main/resources/skills/pdf/reference.md":"# PDF Processing Advanced Reference\n\nThis document contains advanced PDF processing features, detailed examples, and additional libraries not covered in the main skill instructions.\n\n## pypdfium2 Library (Apache/BSD License)\n\n### Overview\npypdfium2 is a Python binding for PDFium (Chromium's PDF library). It's excellent for fast PDF rendering, image generation, and serves as a PyMuPDF replacement.\n\n### Render PDF to Images\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\n\n# Load PDF\npdf = pdfium.PdfDocument(\"document.pdf\")\n\n# Render page to image\npage = pdf[0]  # First page\nbitmap = page.render(\n    scale=2.0,  # Higher resolution\n    rotation=0  # No rotation\n)\n\n# Convert to PIL Image\nimg = bitmap.to_pil()\nimg.save(\"page_1.png\", \"PNG\")\n\n# Process multiple pages\nfor i, page in enumerate(pdf):\n    bitmap = page.render(scale=1.5)\n    img = bitmap.to_pil()\n    img.save(f\"page_{i+1}.jpg\", \"JPEG\", quality=90)\n```\n\n### Extract Text with pypdfium2\n```python\nimport pypdfium2 as pdfium\n\npdf = pdfium.PdfDocument(\"document.pdf\")\nfor i, page in enumerate(pdf):\n    text = page.get_text()\n    print(f\"Page {i+1} text length: {len(text)} chars\")\n```\n\n## JavaScript Libraries\n\n### pdf-lib (MIT License)\n\npdf-lib is a powerful JavaScript library for creating and modifying PDF documents in any JavaScript environment.\n\n#### Load and Manipulate Existing PDF\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function manipulatePDF() {\n    // Load existing PDF\n    const existingPdfBytes = fs.readFileSync('input.pdf');\n    const pdfDoc = await PDFDocument.load(existingPdfBytes);\n\n    // Get page count\n    const pageCount = pdfDoc.getPageCount();\n    console.log(`Document has ${pageCount} pages`);\n\n    // Add new page\n    const newPage = pdfDoc.addPage([600, 400]);\n    newPage.drawText('Added by pdf-lib', {\n        x: 100,\n        y: 300,\n        size: 16\n    });\n\n    // Save modified PDF\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('modified.pdf', pdfBytes);\n}\n```\n\n#### Create Complex PDFs from Scratch\n```javascript\nimport { PDFDocument, rgb, StandardFonts } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function createPDF() {\n    const pdfDoc = await PDFDocument.create();\n\n    // Add fonts\n    const helveticaFont = await pdfDoc.embedFont(StandardFonts.Helvetica);\n    const helveticaBold = await pdfDoc.embedFont(StandardFonts.HelveticaBold);\n\n    // Add page\n    const page = pdfDoc.addPage([595, 842]); // A4 size\n    const { width, height } = page.getSize();\n\n    // Add text with styling\n    page.drawText('Invoice #12345', {\n        x: 50,\n        y: height - 50,\n        size: 18,\n        font: helveticaBold,\n        color: rgb(0.2, 0.2, 0.8)\n    });\n\n    // Add rectangle (header background)\n    page.drawRectangle({\n        x: 40,\n        y: height - 100,\n        width: width - 80,\n        height: 30,\n        color: rgb(0.9, 0.9, 0.9)\n    });\n\n    // Add table-like content\n    const items = [\n        ['Item', 'Qty', 'Price', 'Total'],\n        ['Widget', '2', '$50', '$100'],\n        ['Gadget', '1', '$75', '$75']\n    ];\n\n    let yPos = height - 150;\n    items.forEach(row => {\n        let xPos = 50;\n        row.forEach(cell => {\n            page.drawText(cell, {\n                x: xPos,\n                y: yPos,\n                size: 12,\n                font: helveticaFont\n            });\n            xPos += 120;\n        });\n        yPos -= 25;\n    });\n\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('created.pdf', pdfBytes);\n}\n```\n\n#### Advanced Merge and Split Operations\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function mergePDFs() {\n    // Create new document\n    const mergedPdf = await PDFDocument.create();\n\n    // Load source PDFs\n    const pdf1Bytes = fs.readFileSync('doc1.pdf');\n    const pdf2Bytes = fs.readFileSync('doc2.pdf');\n\n    const pdf1 = await PDFDocument.load(pdf1Bytes);\n    const pdf2 = await PDFDocument.load(pdf2Bytes);\n\n    // Copy pages from first PDF\n    const pdf1Pages = await mergedPdf.copyPages(pdf1, pdf1.getPageIndices());\n    pdf1Pages.forEach(page => mergedPdf.addPage(page));\n\n    // Copy specific pages from second PDF (pages 0, 2, 4)\n    const pdf2Pages = await mergedPdf.copyPages(pdf2, [0, 2, 4]);\n    pdf2Pages.forEach(page => mergedPdf.addPage(page));\n\n    const mergedPdfBytes = await mergedPdf.save();\n    fs.writeFileSync('merged.pdf', mergedPdfBytes);\n}\n```\n\n### pdfjs-dist (Apache License)\n\nPDF.js is Mozilla's JavaScript library for rendering PDFs in the browser.\n\n#### Basic PDF Loading and Rendering\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\n// Configure worker (important for performance)\npdfjsLib.GlobalWorkerOptions.workerSrc = './pdf.worker.js';\n\nasync function renderPDF() {\n    // Load PDF\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    console.log(`Loaded PDF with ${pdf.numPages} pages`);\n\n    // Get first page\n    const page = await pdf.getPage(1);\n    const viewport = page.getViewport({ scale: 1.5 });\n\n    // Render to canvas\n    const canvas = document.createElement('canvas');\n    const context = canvas.getContext('2d');\n    canvas.height = viewport.height;\n    canvas.width = viewport.width;\n\n    const renderContext = {\n        canvasContext: context,\n        viewport: viewport\n    };\n\n    await page.render(renderContext).promise;\n    document.body.appendChild(canvas);\n}\n```\n\n#### Extract Text with Coordinates\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractText() {\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    let fullText = '';\n\n    // Extract text from all pages\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const textContent = await page.getTextContent();\n\n        const pageText = textContent.items\n            .map(item => item.str)\n            .join(' ');\n\n        fullText += `\\n--- Page ${i} ---\\n${pageText}`;\n\n        // Get text with coordinates for advanced processing\n        const textWithCoords = textContent.items.map(item => ({\n            text: item.str,\n            x: item.transform[4],\n            y: item.transform[5],\n            width: item.width,\n            height: item.height\n        }));\n    }\n\n    console.log(fullText);\n    return fullText;\n}\n```\n\n#### Extract Annotations and Forms\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractAnnotations() {\n    const loadingTask = pdfjsLib.getDocument('annotated.pdf');\n    const pdf = await loadingTask.promise;\n\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const annotations = await page.getAnnotations();\n\n        annotations.forEach(annotation => {\n            console.log(`Annotation type: ${annotation.subtype}`);\n            console.log(`Content: ${annotation.contents}`);\n            console.log(`Coordinates: ${JSON.stringify(annotation.rect)}`);\n        });\n    }\n}\n```\n\n## Advanced Command-Line Operations\n\n### poppler-utils Advanced Features\n\n#### Extract Text with Bounding Box Coordinates\n```bash\n# Extract text with bounding box coordinates (essential for structured data)\npdftotext -bbox-layout document.pdf output.xml\n\n# The XML output contains precise coordinates for each text element\n```\n\n#### Advanced Image Conversion\n```bash\n# Convert to PNG images with specific resolution\npdftoppm -png -r 300 document.pdf output_prefix\n\n# Convert specific page range with high resolution\npdftoppm -png -r 600 -f 1 -l 3 document.pdf high_res_pages\n\n# Convert to JPEG with quality setting\npdftoppm -jpeg -jpegopt quality=85 -r 200 document.pdf jpeg_output\n```\n\n#### Extract Embedded Images\n```bash\n# Extract all embedded images with metadata\npdfimages -j -p document.pdf page_images\n\n# List image info without extracting\npdfimages -list document.pdf\n\n# Extract images in their original format\npdfimages -all document.pdf images/img\n```\n\n### qpdf Advanced Features\n\n#### Complex Page Manipulation\n```bash\n# Split PDF into groups of pages\nqpdf --split-pages=3 input.pdf output_group_%02d.pdf\n\n# Extract specific pages with complex ranges\nqpdf input.pdf --pages input.pdf 1,3-5,8,10-end -- extracted.pdf\n\n# Merge specific pages from multiple PDFs\nqpdf --empty --pages doc1.pdf 1-3 doc2.pdf 5-7 doc3.pdf 2,4 -- combined.pdf\n```\n\n#### PDF Optimization and Repair\n```bash\n# Optimize PDF for web (linearize for streaming)\nqpdf --linearize input.pdf optimized.pdf\n\n# Remove unused objects and compress\nqpdf --optimize-level=all input.pdf compressed.pdf\n\n# Attempt to repair corrupted PDF structure\nqpdf --check input.pdf\nqpdf --fix-qdf damaged.pdf repaired.pdf\n\n# Show detailed PDF structure for debugging\nqpdf --show-all-pages input.pdf > structure.txt\n```\n\n#### Advanced Encryption\n```bash\n# Add password protection with specific permissions\nqpdf --encrypt user_pass owner_pass 256 --print=none --modify=none -- input.pdf encrypted.pdf\n\n# Check encryption status\nqpdf --show-encryption encrypted.pdf\n\n# Remove password protection (requires password)\nqpdf --password=secret123 --decrypt encrypted.pdf decrypted.pdf\n```\n\n## Advanced Python Techniques\n\n### pdfplumber Advanced Features\n\n#### Extract Text with Precise Coordinates\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract all text with coordinates\n    chars = page.chars\n    for char in chars[:10]:  # First 10 characters\n        print(f\"Char: '{char['text']}' at x:{char['x0']:.1f} y:{char['y0']:.1f}\")\n    \n    # Extract text by bounding box (left, top, right, bottom)\n    bbox_text = page.within_bbox((100, 100, 400, 200)).extract_text()\n```\n\n#### Advanced Table Extraction with Custom Settings\n```python\nimport pdfplumber\nimport pandas as pd\n\nwith pdfplumber.open(\"complex_table.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract tables with custom settings for complex layouts\n    table_settings = {\n        \"vertical_strategy\": \"lines\",\n        \"horizontal_strategy\": \"lines\",\n        \"snap_tolerance\": 3,\n        \"intersection_tolerance\": 15\n    }\n    tables = page.extract_tables(table_settings)\n    \n    # Visual debugging for table extraction\n    img = page.to_image(resolution=150)\n    img.save(\"debug_layout.png\")\n```\n\n### reportlab Advanced Features\n\n#### Create Professional Reports with Tables\n```python\nfrom reportlab.platypus import SimpleDocTemplate, Table, TableStyle, Paragraph\nfrom reportlab.lib.styles import getSampleStyleSheet\nfrom reportlab.lib import colors\n\n# Sample data\ndata = [\n    ['Product', 'Q1', 'Q2', 'Q3', 'Q4'],\n    ['Widgets', '120', '135', '142', '158'],\n    ['Gadgets', '85', '92', '98', '105']\n]\n\n# Create PDF with table\ndoc = SimpleDocTemplate(\"report.pdf\")\nelements = []\n\n# Add title\nstyles = getSampleStyleSheet()\ntitle = Paragraph(\"Quarterly Sales Report\", styles['Title'])\nelements.append(title)\n\n# Add table with advanced styling\ntable = Table(data)\ntable.setStyle(TableStyle([\n    ('BACKGROUND', (0, 0), (-1, 0), colors.grey),\n    ('TEXTCOLOR', (0, 0), (-1, 0), colors.whitesmoke),\n    ('ALIGN', (0, 0), (-1, -1), 'CENTER'),\n    ('FONTNAME', (0, 0), (-1, 0), 'Helvetica-Bold'),\n    ('FONTSIZE', (0, 0), (-1, 0), 14),\n    ('BOTTOMPADDING', (0, 0), (-1, 0), 12),\n    ('BACKGROUND', (0, 1), (-1, -1), colors.beige),\n    ('GRID', (0, 0), (-1, -1), 1, colors.black)\n]))\nelements.append(table)\n\ndoc.build(elements)\n```\n\n## Complex Workflows\n\n### Extract Figures/Images from PDF\n\n#### Method 1: Using pdfimages (fastest)\n```bash\n# Extract all images with original quality\npdfimages -all document.pdf images/img\n```\n\n#### Method 2: Using pypdfium2 + Image Processing\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\nimport numpy as np\n\ndef extract_figures(pdf_path, output_dir):\n    pdf = pdfium.PdfDocument(pdf_path)\n    \n    for page_num, page in enumerate(pdf):\n        # Render high-resolution page\n        bitmap = page.render(scale=3.0)\n        img = bitmap.to_pil()\n        \n        # Convert to numpy for processing\n        img_array = np.array(img)\n        \n        # Simple figure detection (non-white regions)\n        mask = np.any(img_array != [255, 255, 255], axis=2)\n        \n        # Find contours and extract bounding boxes\n        # (This is simplified - real implementation would need more sophisticated detection)\n        \n        # Save detected figures\n        # ... implementation depends on specific needs\n```\n\n### Batch PDF Processing with Error Handling\n```python\nimport os\nimport glob\nfrom pypdf import PdfReader, PdfWriter\nimport logging\n\nlogging.basicConfig(level=logging.INFO)\nlogger = logging.getLogger(__name__)\n\ndef batch_process_pdfs(input_dir, operation='merge'):\n    pdf_files = glob.glob(os.path.join(input_dir, \"*.pdf\"))\n    \n    if operation == 'merge':\n        writer = PdfWriter()\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                for page in reader.pages:\n                    writer.add_page(page)\n                logger.info(f\"Processed: {pdf_file}\")\n            except Exception as e:\n                logger.error(f\"Failed to process {pdf_file}: {e}\")\n                continue\n        \n        with open(\"batch_merged.pdf\", \"wb\") as output:\n            writer.write(output)\n    \n    elif operation == 'extract_text':\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                text = \"\"\n                for page in reader.pages:\n                    text += page.extract_text()\n                \n                output_file = pdf_file.replace('.pdf', '.txt')\n                with open(output_file, 'w', encoding='utf-8') as f:\n                    f.write(text)\n                logger.info(f\"Extracted text from: {pdf_file}\")\n                \n            except Exception as e:\n                logger.error(f\"Failed to extract text from {pdf_file}: {e}\")\n                continue\n```\n\n### Advanced PDF Cropping\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\n# Crop page (left, bottom, right, top in points)\npage = reader.pages[0]\npage.mediabox.left = 50\npage.mediabox.bottom = 50\npage.mediabox.right = 550\npage.mediabox.top = 750\n\nwriter.add_page(page)\nwith open(\"cropped.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Performance Optimization Tips\n\n### 1. For Large PDFs\n- Use streaming approaches instead of loading entire PDF in memory\n- Use `qpdf --split-pages` for splitting large files\n- Process pages individually with pypdfium2\n\n### 2. For Text Extraction\n- `pdftotext -bbox-layout` is fastest for plain text extraction\n- Use pdfplumber for structured data and tables\n- Avoid `pypdf.extract_text()` for very large documents\n\n### 3. For Image Extraction\n- `pdfimages` is much faster than rendering pages\n- Use low resolution for previews, high resolution for final output\n\n### 4. For Form Filling\n- pdf-lib maintains form structure better than most alternatives\n- Pre-validate form fields before processing\n\n### 5. Memory Management\n```python\n# Process PDFs in chunks\ndef process_large_pdf(pdf_path, chunk_size=10):\n    reader = PdfReader(pdf_path)\n    total_pages = len(reader.pages)\n    \n    for start_idx in range(0, total_pages, chunk_size):\n        end_idx = min(start_idx + chunk_size, total_pages)\n        writer = PdfWriter()\n        \n        for i in range(start_idx, end_idx):\n            writer.add_page(reader.pages[i])\n        \n        # Process chunk\n        with open(f\"chunk_{start_idx//chunk_size}.pdf\", \"wb\") as output:\n            writer.write(output)\n```\n\n## Troubleshooting Common Issues\n\n### Encrypted PDFs\n```python\n# Handle password-protected PDFs\nfrom pypdf import PdfReader\n\ntry:\n    reader = PdfReader(\"encrypted.pdf\")\n    if reader.is_encrypted:\n        reader.decrypt(\"password\")\nexcept Exception as e:\n    print(f\"Failed to decrypt: {e}\")\n```\n\n### Corrupted PDFs\n```bash\n# Use qpdf to repair\nqpdf --check corrupted.pdf\nqpdf --replace-input corrupted.pdf\n```\n\n### Text Extraction Issues\n```python\n# Fallback to OCR for scanned PDFs\nimport pytesseract\nfrom pdf2image import convert_from_path\n\ndef extract_text_with_ocr(pdf_path):\n    images = convert_from_path(pdf_path)\n    text = \"\"\n    for i, image in enumerate(images):\n        text += pytesseract.image_to_string(image)\n    return text\n```\n\n## License Information\n\n- **pypdf**: BSD License\n- **pdfplumber**: MIT License\n- **pypdfium2**: Apache/BSD License\n- **reportlab**: BSD License\n- **poppler-utils**: GPL-2 License\n- **qpdf**: Apache License\n- **pdf-lib**: MIT License\n- **pdfjs-dist**: Apache License","ruoyi-admin/src/main/resources/skills/docx/SKILL.md":"---\nname: docx\ndescription: \"Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# DOCX creation, editing, and analysis\n\n## Overview\n\nA .docx file is a ZIP archive containing XML files.\n\n## Quick Reference\n\n| Task | Approach |\n|------|----------|\n| Read/analyze content | `pandoc` or unpack for raw XML |\n| Create new document | Use `docx-js` - see Creating New Documents below |\n| Edit existing document | Unpack → edit XML → repack - see Editing Existing Documents below |\n\n### Converting .doc to .docx\n\nLegacy `.doc` files must be converted before editing:\n\n```bash\npython scripts/office/soffice.py --headless --convert-to docx document.doc\n```\n\n### Reading Content\n\n```bash\n# Text extraction with tracked changes\npandoc --track-changes=all document.docx -o output.md\n\n# Raw XML access\npython scripts/office/unpack.py document.docx unpacked/\n```\n\n### Converting to Images\n\n```bash\npython scripts/office/soffice.py --headless --convert-to pdf document.docx\npdftoppm -jpeg -r 150 document.pdf page\n```\n\n### Accepting Tracked Changes\n\nTo produce a clean document with all tracked changes accepted (requires LibreOffice):\n\n```bash\npython scripts/accept_changes.py input.docx output.docx\n```\n\n---\n\n## Creating New Documents\n\nGenerate .docx files with JavaScript, then validate. Install: `npm install -g docx`\n\n### Setup\n```javascript\nconst { Document, Packer, Paragraph, TextRun, Table, TableRow, TableCell, ImageRun,\n        Header, Footer, AlignmentType, PageOrientation, LevelFormat, ExternalHyperlink,\n        InternalHyperlink, Bookmark, FootnoteReferenceRun, PositionalTab,\n        PositionalTabAlignment, PositionalTabRelativeTo, PositionalTabLeader,\n        TabStopType, TabStopPosition, Column, SectionType,\n        TableOfContents, HeadingLevel, BorderStyle, WidthType, ShadingType,\n        VerticalAlign, PageNumber, PageBreak } = require('docx');\n\nconst doc = new Document({ sections: [{ children: [/* content */] }] });\nPacker.toBuffer(doc).then(buffer => fs.writeFileSync(\"doc.docx\", buffer));\n```\n\n### Validation\nAfter creating the file, validate it. If validation fails, unpack, fix the XML, and repack.\n```bash\npython scripts/office/validate.py doc.docx\n```\n\n### Page Size\n\n```javascript\n// CRITICAL: docx-js defaults to A4, not US Letter\n// Always set page size explicitly for consistent results\nsections: [{\n  properties: {\n    page: {\n      size: {\n        width: 12240,   // 8.5 inches in DXA\n        height: 15840   // 11 inches in DXA\n      },\n      margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } // 1 inch margins\n    }\n  },\n  children: [/* content */]\n}]\n```\n\n**Common page sizes (DXA units, 1440 DXA = 1 inch):**\n\n| Paper | Width | Height | Content Width (1\" margins) |\n|-------|-------|--------|---------------------------|\n| US Letter | 12,240 | 15,840 | 9,360 |\n| A4 (default) | 11,906 | 16,838 | 9,026 |\n\n**Landscape orientation:** docx-js swaps width/height internally, so pass portrait dimensions and let it handle the swap:\n```javascript\nsize: {\n  width: 12240,   // Pass SHORT edge as width\n  height: 15840,  // Pass LONG edge as height\n  orientation: PageOrientation.LANDSCAPE  // docx-js swaps them in the XML\n},\n// Content width = 15840 - left margin - right margin (uses the long edge)\n```\n\n### Styles (Override Built-in Headings)\n\nUse Arial as the default font (universally supported). Keep titles black for readability.\n\n```javascript\nconst doc = new Document({\n  styles: {\n    default: { document: { run: { font: \"Arial\", size: 24 } } }, // 12pt default\n    paragraphStyles: [\n      // IMPORTANT: Use exact IDs to override built-in styles\n      { id: \"Heading1\", name: \"Heading 1\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 32, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 240, after: 240 }, outlineLevel: 0 } }, // outlineLevel required for TOC\n      { id: \"Heading2\", name: \"Heading 2\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 28, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 180, after: 180 }, outlineLevel: 1 } },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ heading: HeadingLevel.HEADING_1, children: [new TextRun(\"Title\")] }),\n    ]\n  }]\n});\n```\n\n### Lists (NEVER use unicode bullets)\n\n```javascript\n// ❌ WRONG - never manually insert bullet characters\nnew Paragraph({ children: [new TextRun(\"• Item\")] })  // BAD\nnew Paragraph({ children: [new TextRun(\"\\u2022 Item\")] })  // BAD\n\n// ✅ CORRECT - use numbering config with LevelFormat.BULLET\nconst doc = new Document({\n  numbering: {\n    config: [\n      { reference: \"bullets\",\n        levels: [{ level: 0, format: LevelFormat.BULLET, text: \"•\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n      { reference: \"numbers\",\n        levels: [{ level: 0, format: LevelFormat.DECIMAL, text: \"%1.\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ numbering: { reference: \"bullets\", level: 0 },\n        children: [new TextRun(\"Bullet item\")] }),\n      new Paragraph({ numbering: { reference: \"numbers\", level: 0 },\n        children: [new TextRun(\"Numbered item\")] }),\n    ]\n  }]\n});\n\n// ⚠️ Each reference creates INDEPENDENT numbering\n// Same reference = continues (1,2,3 then 4,5,6)\n// Different reference = restarts (1,2,3 then 1,2,3)\n```\n\n### Tables\n\n**CRITICAL: Tables need dual widths** - set both `columnWidths` on the table AND `width` on each cell. Without both, tables render incorrectly on some platforms.\n\n```javascript\n// CRITICAL: Always set table width for consistent rendering\n// CRITICAL: Use ShadingType.CLEAR (not SOLID) to prevent black backgrounds\nconst border = { style: BorderStyle.SINGLE, size: 1, color: \"CCCCCC\" };\nconst borders = { top: border, bottom: border, left: border, right: border };\n\nnew Table({\n  width: { size: 9360, type: WidthType.DXA }, // Always use DXA (percentages break in Google Docs)\n  columnWidths: [4680, 4680], // Must sum to table width (DXA: 1440 = 1 inch)\n  rows: [\n    new TableRow({\n      children: [\n        new TableCell({\n          borders,\n          width: { size: 4680, type: WidthType.DXA }, // Also set on each cell\n          shading: { fill: \"D5E8F0\", type: ShadingType.CLEAR }, // CLEAR not SOLID\n          margins: { top: 80, bottom: 80, left: 120, right: 120 }, // Cell padding (internal, not added to width)\n          children: [new Paragraph({ children: [new TextRun(\"Cell\")] })]\n        })\n      ]\n    })\n  ]\n})\n```\n\n**Table width calculation:**\n\nAlways use `WidthType.DXA` — `WidthType.PERCENTAGE` breaks in Google Docs.\n\n```javascript\n// Table width = sum of columnWidths = content width\n// US Letter with 1\" margins: 12240 - 2880 = 9360 DXA\nwidth: { size: 9360, type: WidthType.DXA },\ncolumnWidths: [7000, 2360]  // Must sum to table width\n```\n\n**Width rules:**\n- **Always use `WidthType.DXA`** — never `WidthType.PERCENTAGE` (incompatible with Google Docs)\n- Table width must equal the sum of `columnWidths`\n- Cell `width` must match corresponding `columnWidth`\n- Cell `margins` are internal padding - they reduce content area, not add to cell width\n- For full-width tables: use content width (page width minus left and right margins)\n\n### Images\n\n```javascript\n// CRITICAL: type parameter is REQUIRED\nnew Paragraph({\n  children: [new ImageRun({\n    type: \"png\", // Required: png, jpg, jpeg, gif, bmp, svg\n    data: fs.readFileSync(\"image.png\"),\n    transformation: { width: 200, height: 150 },\n    altText: { title: \"Title\", description: \"Desc\", name: \"Name\" } // All three required\n  })]\n})\n```\n\n### Page Breaks\n\n```javascript\n// CRITICAL: PageBreak must be inside a Paragraph\nnew Paragraph({ children: [new PageBreak()] })\n\n// Or use pageBreakBefore\nnew Paragraph({ pageBreakBefore: true, children: [new TextRun(\"New page\")] })\n```\n\n### Hyperlinks\n\n```javascript\n// External link\nnew Paragraph({\n  children: [new ExternalHyperlink({\n    children: [new TextRun({ text: \"Click here\", style: \"Hyperlink\" })],\n    link: \"https://example.com\",\n  })]\n})\n\n// Internal link (bookmark + reference)\n// 1. Create bookmark at destination\nnew Paragraph({ heading: HeadingLevel.HEADING_1, children: [\n  new Bookmark({ id: \"chapter1\", children: [new TextRun(\"Chapter 1\")] }),\n]})\n// 2. Link to it\nnew Paragraph({ children: [new InternalHyperlink({\n  children: [new TextRun({ text: \"See Chapter 1\", style: \"Hyperlink\" })],\n  anchor: \"chapter1\",\n})]})\n```\n\n### Footnotes\n\n```javascript\nconst doc = new Document({\n  footnotes: {\n    1: { children: [new Paragraph(\"Source: Annual Report 2024\")] },\n    2: { children: [new Paragraph(\"See appendix for methodology\")] },\n  },\n  sections: [{\n    children: [new Paragraph({\n      children: [\n        new TextRun(\"Revenue grew 15%\"),\n        new FootnoteReferenceRun(1),\n        new TextRun(\" using adjusted metrics\"),\n        new FootnoteReferenceRun(2),\n      ],\n    })]\n  }]\n});\n```\n\n### Tab Stops\n\n```javascript\n// Right-align text on same line (e.g., date opposite a title)\nnew Paragraph({\n  children: [\n    new TextRun(\"Company Name\"),\n    new TextRun(\"\\tJanuary 2025\"),\n  ],\n  tabStops: [{ type: TabStopType.RIGHT, position: TabStopPosition.MAX }],\n})\n\n// Dot leader (e.g., TOC-style)\nnew Paragraph({\n  children: [\n    new TextRun(\"Introduction\"),\n    new TextRun({ children: [\n      new PositionalTab({\n        alignment: PositionalTabAlignment.RIGHT,\n        relativeTo: PositionalTabRelativeTo.MARGIN,\n        leader: PositionalTabLeader.DOT,\n      }),\n      \"3\",\n    ]}),\n  ],\n})\n```\n\n### Multi-Column Layouts\n\n```javascript\n// Equal-width columns\nsections: [{\n  properties: {\n    column: {\n      count: 2,          // number of columns\n      space: 720,        // gap between columns in DXA (720 = 0.5 inch)\n      equalWidth: true,\n      separate: true,    // vertical line between columns\n    },\n  },\n  children: [/* content flows naturally across columns */]\n}]\n\n// Custom-width columns (equalWidth must be false)\nsections: [{\n  properties: {\n    column: {\n      equalWidth: false,\n      children: [\n        new Column({ width: 5400, space: 720 }),\n        new Column({ width: 3240 }),\n      ],\n    },\n  },\n  children: [/* content */]\n}]\n```\n\nForce a column break with a new section using `type: SectionType.NEXT_COLUMN`.\n\n### Table of Contents\n\n```javascript\n// CRITICAL: Headings must use HeadingLevel ONLY - no custom styles\nnew TableOfContents(\"Table of Contents\", { hyperlink: true, headingStyleRange: \"1-3\" })\n```\n\n### Headers/Footers\n\n```javascript\nsections: [{\n  properties: {\n    page: { margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } } // 1440 = 1 inch\n  },\n  headers: {\n    default: new Header({ children: [new Paragraph({ children: [new TextRun(\"Header\")] })] })\n  },\n  footers: {\n    default: new Footer({ children: [new Paragraph({\n      children: [new TextRun(\"Page \"), new TextRun({ children: [PageNumber.CURRENT] })]\n    })] })\n  },\n  children: [/* content */]\n}]\n```\n\n### Critical Rules for docx-js\n\n- **Set page size explicitly** - docx-js defaults to A4; use US Letter (12240 x 15840 DXA) for US documents\n- **Landscape: pass portrait dimensions** - docx-js swaps width/height internally; pass short edge as `width`, long edge as `height`, and set `orientation: PageOrientation.LANDSCAPE`\n- **Never use `\\n`** - use separate Paragraph elements\n- **Never use unicode bullets** - use `LevelFormat.BULLET` with numbering config\n- **PageBreak must be in Paragraph** - standalone creates invalid XML\n- **ImageRun requires `type`** - always specify png/jpg/etc\n- **Always set table `width` with DXA** - never use `WidthType.PERCENTAGE` (breaks in Google Docs)\n- **Tables need dual widths** - `columnWidths` array AND cell `width`, both must match\n- **Table width = sum of columnWidths** - for DXA, ensure they add up exactly\n- **Always add cell margins** - use `margins: { top: 80, bottom: 80, left: 120, right: 120 }` for readable padding\n- **Use `ShadingType.CLEAR`** - never SOLID for table shading\n- **Never use tables as dividers/rules** - cells have minimum height and render as empty boxes (including in headers/footers); use `border: { bottom: { style: BorderStyle.SINGLE, size: 6, color: \"2E75B6\", space: 1 } }` on a Paragraph instead. For two-column footers, use tab stops (see Tab Stops section), not tables\n- **TOC requires HeadingLevel only** - no custom styles on heading paragraphs\n- **Override built-in styles** - use exact IDs: \"Heading1\", \"Heading2\", etc.\n- **Include `outlineLevel`** - required for TOC (0 for H1, 1 for H2, etc.)\n\n---\n\n## Editing Existing Documents\n\n**Follow all 3 steps in order.**\n\n### Step 1: Unpack\n```bash\npython scripts/office/unpack.py document.docx unpacked/\n```\nExtracts XML, pretty-prints, merges adjacent runs, and converts smart quotes to XML entities (`&#x201C;` etc.) so they survive editing. Use `--merge-runs false` to skip run merging.\n\n### Step 2: Edit XML\n\nEdit files in `unpacked/word/`. See XML Reference below for patterns.\n\n**Use \"Claude\" as the author** for tracked changes and comments, unless the user explicitly requests use of a different name.\n\n**Use the Edit tool directly for string replacement. Do not write Python scripts.** Scripts introduce unnecessary complexity. The Edit tool shows exactly what is being replaced.\n\n**CRITICAL: Use smart quotes for new content.** When adding text with apostrophes or quotes, use XML entities to produce smart quotes:\n```xml\n<!-- Use these entities for professional typography -->\n<w:t>Here&#x2019;s a quote: &#x201C;Hello&#x201D;</w:t>\n```\n| Entity | Character |\n|--------|-----------|\n| `&#x2018;` | ‘ (left single) |\n| `&#x2019;` | ’ (right single / apostrophe) |\n| `&#x201C;` | “ (left double) |\n| `&#x201D;` | ” (right double) |\n\n**Adding comments:** Use `comment.py` to handle boilerplate across multiple XML files (text must be pre-escaped XML):\n```bash\npython scripts/comment.py unpacked/ 0 \"Comment text with &amp; and &#x2019;\"\npython scripts/comment.py unpacked/ 1 \"Reply text\" --parent 0  # reply to comment 0\npython scripts/comment.py unpacked/ 0 \"Text\" --author \"Custom Author\"  # custom author name\n```\nThen add markers to document.xml (see Comments in XML Reference).\n\n### Step 3: Pack\n```bash\npython scripts/office/pack.py unpacked/ output.docx --original document.docx\n```\nValidates with auto-repair, condenses XML, and creates DOCX. Use `--validate false` to skip.\n\n**Auto-repair will fix:**\n- `durableId` >= 0x7FFFFFFF (regenerates valid ID)\n- Missing `xml:space=\"preserve\"` on `<w:t>` with whitespace\n\n**Auto-repair won't fix:**\n- Malformed XML, invalid element nesting, missing relationships, schema violations\n\n### Common Pitfalls\n\n- **Replace entire `<w:r>` elements**: When adding tracked changes, replace the whole `<w:r>...</w:r>` block with `<w:del>...<w:ins>...` as siblings. Don't inject tracked change tags inside a run.\n- **Preserve `<w:rPr>` formatting**: Copy the original run's `<w:rPr>` block into your tracked change runs to maintain bold, font size, etc.\n\n---\n\n## XML Reference\n\n### Schema Compliance\n\n- **Element order in `<w:pPr>`**: `<w:pStyle>`, `<w:numPr>`, `<w:spacing>`, `<w:ind>`, `<w:jc>`, `<w:rPr>` last\n- **Whitespace**: Add `xml:space=\"preserve\"` to `<w:t>` with leading/trailing spaces\n- **RSIDs**: Must be 8-digit hex (e.g., `00AB1234`)\n\n### Tracked Changes\n\n**Insertion:**\n```xml\n<w:ins w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:t>inserted text</w:t></w:r>\n</w:ins>\n```\n\n**Deletion:**\n```xml\n<w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n```\n\n**Inside `<w:del>`**: Use `<w:delText>` instead of `<w:t>`, and `<w:delInstrText>` instead of `<w:instrText>`.\n\n**Minimal edits** - only mark what changes:\n```xml\n<!-- Change \"30 days\" to \"60 days\" -->\n<w:r><w:t>The term is </w:t></w:r>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:delText>30</w:delText></w:r>\n</w:del>\n<w:ins w:id=\"2\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:t>60</w:t></w:r>\n</w:ins>\n<w:r><w:t> days.</w:t></w:r>\n```\n\n**Deleting entire paragraphs/list items** - when removing ALL content from a paragraph, also mark the paragraph mark as deleted so it merges with the next paragraph. Add `<w:del/>` inside `<w:pPr><w:rPr>`:\n```xml\n<w:p>\n  <w:pPr>\n    <w:numPr>...</w:numPr>  <!-- list numbering if present -->\n    <w:rPr>\n      <w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\"/>\n    </w:rPr>\n  </w:pPr>\n  <w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n    <w:r><w:delText>Entire paragraph content being deleted...</w:delText></w:r>\n  </w:del>\n</w:p>\n```\nWithout the `<w:del/>` in `<w:pPr><w:rPr>`, accepting changes leaves an empty paragraph/list item.\n\n**Rejecting another author's insertion** - nest deletion inside their insertion:\n```xml\n<w:ins w:author=\"Jane\" w:id=\"5\">\n  <w:del w:author=\"Claude\" w:id=\"10\">\n    <w:r><w:delText>their inserted text</w:delText></w:r>\n  </w:del>\n</w:ins>\n```\n\n**Restoring another author's deletion** - add insertion after (don't modify their deletion):\n```xml\n<w:del w:author=\"Jane\" w:id=\"5\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n<w:ins w:author=\"Claude\" w:id=\"10\">\n  <w:r><w:t>deleted text</w:t></w:r>\n</w:ins>\n```\n\n### Comments\n\nAfter running `comment.py` (see Step 2), add markers to document.xml. For replies, use `--parent` flag and nest markers inside the parent's.\n\n**CRITICAL: `<w:commentRangeStart>` and `<w:commentRangeEnd>` are siblings of `<w:r>`, never inside `<w:r>`.**\n\n```xml\n<!-- Comment markers are direct children of w:p, never inside w:r -->\n<w:commentRangeStart w:id=\"0\"/>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted</w:delText></w:r>\n</w:del>\n<w:r><w:t> more text</w:t></w:r>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n\n<!-- Comment 0 with reply 1 nested inside -->\n<w:commentRangeStart w:id=\"0\"/>\n  <w:commentRangeStart w:id=\"1\"/>\n  <w:r><w:t>text</w:t></w:r>\n  <w:commentRangeEnd w:id=\"1\"/>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"1\"/></w:r>\n```\n\n### Images\n\n1. Add image file to `word/media/`\n2. Add relationship to `word/_rels/document.xml.rels`:\n```xml\n<Relationship Id=\"rId5\" Type=\".../image\" Target=\"media/image1.png\"/>\n```\n3. Add content type to `[Content_Types].xml`:\n```xml\n<Default Extension=\"png\" ContentType=\"image/png\"/>\n```\n4. Reference in document.xml:\n```xml\n<w:drawing>\n  <wp:inline>\n    <wp:extent cx=\"914400\" cy=\"914400\"/>  <!-- EMUs: 914400 = 1 inch -->\n    <a:graphic>\n      <a:graphicData uri=\".../picture\">\n        <pic:pic>\n          <pic:blipFill><a:blip r:embed=\"rId5\"/></pic:blipFill>\n        </pic:pic>\n      </a:graphicData>\n    </a:graphic>\n  </wp:inline>\n</w:drawing>\n```\n\n---\n\n## Dependencies\n\n- **pandoc**: Text extraction\n- **docx**: `npm install -g docx` (new documents)\n- **LibreOffice**: PDF conversion (auto-configured for sandboxed environments via `scripts/office/soffice.py`)\n- **Poppler**: `pdftoppm` for images\n","ruoyi-admin/src/main/resources/skills/pdf/SKILL.md":"---\nname: pdf\ndescription: Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# PDF Processing Guide\n\n## Overview\n\nThis guide covers essential PDF processing operations using Python libraries and command-line tools. For advanced features, JavaScript libraries, and detailed examples, see reference.md. If you need to fill out a PDF form, read forms.md and follow its instructions.\n\n## Quick Start\n\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Read a PDF\nreader = PdfReader(\"document.pdf\")\nprint(f\"Pages: {len(reader.pages)}\")\n\n# Extract text\ntext = \"\"\nfor page in reader.pages:\n    text += page.extract_text()\n```\n\n## Python Libraries\n\n### pypdf - Basic Operations\n\n#### Merge PDFs\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nwriter = PdfWriter()\nfor pdf_file in [\"doc1.pdf\", \"doc2.pdf\", \"doc3.pdf\"]:\n    reader = PdfReader(pdf_file)\n    for page in reader.pages:\n        writer.add_page(page)\n\nwith open(\"merged.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n#### Split PDF\n```python\nreader = PdfReader(\"input.pdf\")\nfor i, page in enumerate(reader.pages):\n    writer = PdfWriter()\n    writer.add_page(page)\n    with open(f\"page_{i+1}.pdf\", \"wb\") as output:\n        writer.write(output)\n```\n\n#### Extract Metadata\n```python\nreader = PdfReader(\"document.pdf\")\nmeta = reader.metadata\nprint(f\"Title: {meta.title}\")\nprint(f\"Author: {meta.author}\")\nprint(f\"Subject: {meta.subject}\")\nprint(f\"Creator: {meta.creator}\")\n```\n\n#### Rotate Pages\n```python\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\npage = reader.pages[0]\npage.rotate(90)  # Rotate 90 degrees clockwise\nwriter.add_page(page)\n\nwith open(\"rotated.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### pdfplumber - Text and Table Extraction\n\n#### Extract Text with Layout\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for page in pdf.pages:\n        text = page.extract_text()\n        print(text)\n```\n\n#### Extract Tables\n```python\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for i, page in enumerate(pdf.pages):\n        tables = page.extract_tables()\n        for j, table in enumerate(tables):\n            print(f\"Table {j+1} on page {i+1}:\")\n            for row in table:\n                print(row)\n```\n\n#### Advanced Table Extraction\n```python\nimport pandas as pd\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    all_tables = []\n    for page in pdf.pages:\n        tables = page.extract_tables()\n        for table in tables:\n            if table:  # Check if table is not empty\n                df = pd.DataFrame(table[1:], columns=table[0])\n                all_tables.append(df)\n\n# Combine all tables\nif all_tables:\n    combined_df = pd.concat(all_tables, ignore_index=True)\n    combined_df.to_excel(\"extracted_tables.xlsx\", index=False)\n```\n\n### reportlab - Create PDFs\n\n#### Basic PDF Creation\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.pdfgen import canvas\n\nc = canvas.Canvas(\"hello.pdf\", pagesize=letter)\nwidth, height = letter\n\n# Add text\nc.drawString(100, height - 100, \"Hello World!\")\nc.drawString(100, height - 120, \"This is a PDF created with reportlab\")\n\n# Add a line\nc.line(100, height - 140, 400, height - 140)\n\n# Save\nc.save()\n```\n\n#### Create PDF with Multiple Pages\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, PageBreak\nfrom reportlab.lib.styles import getSampleStyleSheet\n\ndoc = SimpleDocTemplate(\"report.pdf\", pagesize=letter)\nstyles = getSampleStyleSheet()\nstory = []\n\n# Add content\ntitle = Paragraph(\"Report Title\", styles['Title'])\nstory.append(title)\nstory.append(Spacer(1, 12))\n\nbody = Paragraph(\"This is the body of the report. \" * 20, styles['Normal'])\nstory.append(body)\nstory.append(PageBreak())\n\n# Page 2\nstory.append(Paragraph(\"Page 2\", styles['Heading1']))\nstory.append(Paragraph(\"Content for page 2\", styles['Normal']))\n\n# Build PDF\ndoc.build(story)\n```\n\n## Command-Line Tools\n\n### pdftotext (poppler-utils)\n```bash\n# Extract text\npdftotext input.pdf output.txt\n\n# Extract text preserving layout\npdftotext -layout input.pdf output.txt\n\n# Extract specific pages\npdftotext -f 1 -l 5 input.pdf output.txt  # Pages 1-5\n```\n\n### qpdf\n```bash\n# Merge PDFs\nqpdf --empty --pages file1.pdf file2.pdf -- merged.pdf\n\n# Split pages\nqpdf input.pdf --pages . 1-5 -- pages1-5.pdf\nqpdf input.pdf --pages . 6-10 -- pages6-10.pdf\n\n# Rotate pages\nqpdf input.pdf output.pdf --rotate=+90:1  # Rotate page 1 by 90 degrees\n\n# Remove password\nqpdf --password=mypassword --decrypt encrypted.pdf decrypted.pdf\n```\n\n### pdftk (if available)\n```bash\n# Merge\npdftk file1.pdf file2.pdf cat output merged.pdf\n\n# Split\npdftk input.pdf burst\n\n# Rotate\npdftk input.pdf rotate 1east output rotated.pdf\n```\n\n## Common Tasks\n\n### Extract Text from Scanned PDFs\n```python\n# Requires: pip install pytesseract pdf2image\nimport pytesseract\nfrom pdf2image import convert_from_path\n\n# Convert PDF to images\nimages = convert_from_path('scanned.pdf')\n\n# OCR each page\ntext = \"\"\nfor i, image in enumerate(images):\n    text += f\"Page {i+1}:\\n\"\n    text += pytesseract.image_to_string(image)\n    text += \"\\n\\n\"\n\nprint(text)\n```\n\n### Add Watermark\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Create watermark (or load existing)\nwatermark = PdfReader(\"watermark.pdf\").pages[0]\n\n# Apply to all pages\nreader = PdfReader(\"document.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    page.merge_page(watermark)\n    writer.add_page(page)\n\nwith open(\"watermarked.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### Extract Images\n```bash\n# Using pdfimages (poppler-utils)\npdfimages -j input.pdf output_prefix\n\n# This extracts all images as output_prefix-000.jpg, output_prefix-001.jpg, etc.\n```\n\n### Password Protection\n```python\nfrom pypdf import PdfReader, PdfWriter\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    writer.add_page(page)\n\n# Add password\nwriter.encrypt(\"userpassword\", \"ownerpassword\")\n\nwith open(\"encrypted.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Quick Reference\n\n| Task | Best Tool | Command/Code |\n|------|-----------|--------------|\n| Merge PDFs | pypdf | `writer.add_page(page)` |\n| Split PDFs | pypdf | One page per file |\n| Extract text | pdfplumber | `page.extract_text()` |\n| Extract tables | pdfplumber | `page.extract_tables()` |\n| Create PDFs | reportlab | Canvas or Platypus |\n| Command line merge | qpdf | `qpdf --empty --pages ...` |\n| OCR scanned PDFs | pytesseract | Convert to image first |\n| Fill PDF forms | pdf-lib or pypdf (see forms.md) | See forms.md |\n\n## Next Steps\n\n- For advanced pypdfium2 usage, see reference.md\n- For JavaScript libraries (pdf-lib), see reference.md\n- If you need to fill out a PDF form, follow the instructions in forms.md\n- For troubleshooting guides, see reference.md\n","ruoyi-admin/src/main/resources/skills/xlsx/SKILL.md":"---\nname: xlsx\ndescription: \"Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# Requirements for Outputs\n\n## All Excel files\n\n### Zero Formula Errors\n- Every Excel model MUST be delivered with ZERO formula errors (#REF!, #DIV/0!, #VALUE!, #N/A, #NAME?)\n\n### Preserve Existing Templates (when updating templates)\n- Study and EXACTLY match existing format, style, and conventions when modifying files\n- Never impose standardized formatting on files with established patterns\n- Existing template conventions ALWAYS override these guidelines\n\n## Financial models\n\n### Color Coding Standards\nUnless otherwise stated by the user or existing template\n\n#### Industry-Standard Color Conventions\n- **Blue text (RGB: 0,0,255)**: Hardcoded inputs, and numbers users will change for scenarios\n- **Black text (RGB: 0,0,0)**: ALL formulas and calculations\n- **Green text (RGB: 0,128,0)**: Links pulling from other worksheets within same workbook\n- **Red text (RGB: 255,0,0)**: External links to other files\n- **Yellow background (RGB: 255,255,0)**: Key assumptions needing attention or cells that need to be updated\n\n### Number Formatting Standards\n\n#### Required Format Rules\n- **Years**: Format as text strings (e.g., \"2024\" not \"2,024\")\n- **Currency**: Use $#,##0 format; ALWAYS specify units in headers (\"Revenue ($mm)\")\n- **Zeros**: Use number formatting to make all zeros \"-\", including percentages (e.g., \"$#,##0;($#,##0);-\")\n- **Percentages**: Default to 0.0% format (one decimal)\n- **Multiples**: Format as 0.0x for valuation multiples (EV/EBITDA, P/E)\n- **Negative numbers**: Use parentheses (123) not minus -123\n\n### Formula Construction Rules\n\n#### Assumptions Placement\n- Place ALL assumptions (growth rates, margins, multiples, etc.) in separate assumption cells\n- Use cell references instead of hardcoded values in formulas\n- Example: Use =B5*(1+$B$6) instead of =B5*1.05\n\n#### Formula Error Prevention\n- Verify all cell references are correct\n- Check for off-by-one errors in ranges\n- Ensure consistent formulas across all projection periods\n- Test with edge cases (zero values, negative numbers)\n- Verify no unintended circular references\n\n#### Documentation Requirements for Hardcodes\n- Comment or in cells beside (if end of table). Format: \"Source: [System/Document], [Date], [Specific Reference], [URL if applicable]\"\n- Examples:\n  - \"Source: Company 10-K, FY2024, Page 45, Revenue Note, [SEC EDGAR URL]\"\n  - \"Source: Company 10-Q, Q2 2025, Exhibit 99.1, [SEC EDGAR URL]\"\n  - \"Source: Bloomberg Terminal, 8/15/2025, AAPL US Equity\"\n  - \"Source: FactSet, 8/20/2025, Consensus Estimates Screen\"\n\n# XLSX creation, editing, and analysis\n\n## Overview\n\nA user may ask you to create, edit, or analyze the contents of an .xlsx file. You have different tools and workflows available for different tasks.\n\n## Important Requirements\n\n**LibreOffice Required for Formula Recalculation**: You can assume LibreOffice is installed for recalculating formula values using the `recalc.py` script. The script automatically configures LibreOffice on first run\n\n## Reading and analyzing data\n\n### Data analysis with pandas\nFor data analysis, visualization, and basic operations, use **pandas** which provides powerful data manipulation capabilities:\n\n```python\nimport pandas as pd\n\n# Read Excel\ndf = pd.read_excel('file.xlsx')  # Default: first sheet\nall_sheets = pd.read_excel('file.xlsx', sheet_name=None)  # All sheets as dict\n\n# Analyze\ndf.head()      # Preview data\ndf.info()      # Column info\ndf.describe()  # Statistics\n\n# Write Excel\ndf.to_excel('output.xlsx', index=False)\n```\n\n## Excel File Workflows\n\n## CRITICAL: Use Formulas, Not Hardcoded Values\n\n**Always use Excel formulas instead of calculating values in Python and hardcoding them.** This ensures the spreadsheet remains dynamic and updateable.\n\n### ❌ WRONG - Hardcoding Calculated Values\n```python\n# Bad: Calculating in Python and hardcoding result\ntotal = df['Sales'].sum()\nsheet['B10'] = total  # Hardcodes 5000\n\n# Bad: Computing growth rate in Python\ngrowth = (df.iloc[-1]['Revenue'] - df.iloc[0]['Revenue']) / df.iloc[0]['Revenue']\nsheet['C5'] = growth  # Hardcodes 0.15\n\n# Bad: Python calculation for average\navg = sum(values) / len(values)\nsheet['D20'] = avg  # Hardcodes 42.5\n```\n\n### ✅ CORRECT - Using Excel Formulas\n```python\n# Good: Let Excel calculate the sum\nsheet['B10'] = '=SUM(B2:B9)'\n\n# Good: Growth rate as Excel formula\nsheet['C5'] = '=(C4-C2)/C2'\n\n# Good: Average using Excel function\nsheet['D20'] = '=AVERAGE(D2:D19)'\n```\n\nThis applies to ALL calculations - totals, percentages, ratios, differences, etc. The spreadsheet should be able to recalculate when source data changes.\n\n## Common Workflow\n1. **Choose tool**: pandas for data, openpyxl for formulas/formatting\n2. **Create/Load**: Create new workbook or load existing file\n3. **Modify**: Add/edit data, formulas, and formatting\n4. **Save**: Write to file\n5. **Recalculate formulas (MANDATORY IF USING FORMULAS)**: Use the recalc.py script\n   ```bash\n   python recalc.py output.xlsx\n   ```\n6. **Verify and fix any errors**: \n   - The script returns JSON with error details\n   - If `status` is `errors_found`, check `error_summary` for specific error types and locations\n   - Fix the identified errors and recalculate again\n   - Common errors to fix:\n     - `#REF!`: Invalid cell references\n     - `#DIV/0!`: Division by zero\n     - `#VALUE!`: Wrong data type in formula\n     - `#NAME?`: Unrecognized formula name\n\n### Creating new Excel files\n\n```python\n# Using openpyxl for formulas and formatting\nfrom openpyxl import Workbook\nfrom openpyxl.styles import Font, PatternFill, Alignment\n\nwb = Workbook()\nsheet = wb.active\n\n# Add data\nsheet['A1'] = 'Hello'\nsheet['B1'] = 'World'\nsheet.append(['Row', 'of', 'data'])\n\n# Add formula\nsheet['B2'] = '=SUM(A1:A10)'\n\n# Formatting\nsheet['A1'].font = Font(bold=True, color='FF0000')\nsheet['A1'].fill = PatternFill('solid', start_color='FFFF00')\nsheet['A1'].alignment = Alignment(horizontal='center')\n\n# Column width\nsheet.column_dimensions['A'].width = 20\n\nwb.save('output.xlsx')\n```\n\n### Editing existing Excel files\n\n```python\n# Using openpyxl to preserve formulas and formatting\nfrom openpyxl import load_workbook\n\n# Load existing file\nwb = load_workbook('existing.xlsx')\nsheet = wb.active  # or wb['SheetName'] for specific sheet\n\n# Working with multiple sheets\nfor sheet_name in wb.sheetnames:\n    sheet = wb[sheet_name]\n    print(f\"Sheet: {sheet_name}\")\n\n# Modify cells\nsheet['A1'] = 'New Value'\nsheet.insert_rows(2)  # Insert row at position 2\nsheet.delete_cols(3)  # Delete column 3\n\n# Add new sheet\nnew_sheet = wb.create_sheet('NewSheet')\nnew_sheet['A1'] = 'Data'\n\nwb.save('modified.xlsx')\n```\n\n## Recalculating formulas\n\nExcel files created or modified by openpyxl contain formulas as strings but not calculated values. Use the provided `recalc.py` script to recalculate formulas:\n\n```bash\n# Recalculate existing Excel file\npython recalc.py <excel_file> [timeout_seconds]\n\n# Create new Excel workbook\npython recalc.py --create <output_file> '<json_data>'\n```\n\nExamples:\n```bash\npython recalc.py output.xlsx 30\n\npython recalc.py --create output.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Value\"], [\"Test\", 100]]}}'\n```\n\nThe script:\n- Automatically sets up LibreOffice macro on first run\n- Recalculates all formulas in all sheets\n- Scans ALL cells for Excel errors (#REF!, #DIV/0!, etc.)\n- Returns JSON with detailed error locations and counts\n- Works on Windows, Linux, and macOS\n- Can create new workbooks from JSON configuration\n\n## Creating workbooks with recalc.py\n\nThe `--create` mode allows you to create new Excel workbooks directly from JSON configuration:\n\n```bash\npython recalc.py --create <output_file> '<json_data>'\n```\n\n**Default Behavior**: If you provide a relative path (e.g., `test.xlsx`), the file will be created in the **project root** directory (`ruoyi-ai-v3/`). To save to a specific directory, use an absolute path.\n\n### JSON Format\n\n```json\n{\n  \"Sheet1\": {\n    \"data\": [\n      [\"Header1\", \"Header2\", \"Header3\"],\n      [\"Value1\", \"Value2\", \"=A2*B2\"],\n      [\"Value3\", \"Value4\", \"=SUM(C2:C10)\"]\n    ]\n  },\n  \"Sheet2\": {\n    \"data\": [\n      [\"Name\", \"Score\"],\n      [\"Alice\", 95],\n      [\"Bob\", 87]\n    ]\n  }\n}\n```\n\n### Examples\n\n**Simple data table (saves to project root)**:\n```bash\npython recalc.py --create data.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Age\"], [\"John\", 30], [\"Jane\", 28]]}}'\n# Creates: ruoyi-ai-v3/data.xlsx\n```\n\n**With formulas (saves to project root)**:\n```bash\npython recalc.py --create model.xlsx '{\"Sheet1\": {\"data\": [[\"A\", \"B\", \"Sum\"], [10, 20, \"=A2+B2\"], [5, 15, \"=A3+B3\"]]}}'\n# Creates: ruoyi-ai-v3/model.xlsx\n```\n\n**Save to custom directory (use absolute path)**:\n```bash\npython recalc.py --create \"D:\\Downloads\\custom.xlsx\" '{\"Sheet1\": {\"data\": [[\"A\", \"B\"], [1, 2]]}}'\n# Creates: D:\\Downloads\\custom.xlsx\n```\n\n### File Path Options\n\n| Input | Output Location |\n|-------|-----------------|\n| `test.xlsx` | `ruoyi-ai-v3/test.xlsx` (project root) |\n| `output/test.xlsx` | `ruoyi-ai-v3/output/test.xlsx` |\n| `D:\\Downloads\\test.xlsx` | `D:\\Downloads\\test.xlsx` (absolute path) |\n\n### Features\n\n- First row is automatically formatted as bold header\n- Supports both values and Excel formulas (strings starting with `=`)\n- Multiple sheets can be created in one command\n- Output file extension determines format (.xlsx, .xlsm, etc.)\n\n## Formula Verification Checklist\n\nQuick checks to ensure formulas work correctly:\n\n### Essential Verification\n- [ ] **Test 2-3 sample references**: Verify they pull correct values before building full model\n- [ ] **Column mapping**: Confirm Excel columns match (e.g., column 64 = BL, not BK)\n- [ ] **Row offset**: Remember Excel rows are 1-indexed (DataFrame row 5 = Excel row 6)\n\n### Common Pitfalls\n- [ ] **NaN handling**: Check for null values with `pd.notna()`\n- [ ] **Far-right columns**: FY data often in columns 50+ \n- [ ] **Multiple matches**: Search all occurrences, not just first\n- [ ] **Division by zero**: Check denominators before using `/` in formulas (#DIV/0!)\n- [ ] **Wrong references**: Verify all cell references point to intended cells (#REF!)\n- [ ] **Cross-sheet references**: Use correct format (Sheet1!A1) for linking sheets\n\n### Formula Testing Strategy\n- [ ] **Start small**: Test formulas on 2-3 cells before applying broadly\n- [ ] **Verify dependencies**: Check all cells referenced in formulas exist\n- [ ] **Test edge cases**: Include zero, negative, and very large values\n\n### Interpreting recalc.py Output\nThe script returns JSON with error details:\n```json\n{\n  \"status\": \"success\",           // or \"errors_found\"\n  \"total_errors\": 0,              // Total error count\n  \"total_formulas\": 42,           // Number of formulas in file\n  \"error_summary\": {              // Only present if errors found\n    \"#REF!\": {\n      \"count\": 2,\n      \"locations\": [\"Sheet1!B5\", \"Sheet1!C10\"]\n    }\n  }\n}\n```\n\n## Best Practices\n\n### Library Selection\n- **pandas**: Best for data analysis, bulk operations, and simple data export\n- **openpyxl**: Best for complex formatting, formulas, and Excel-specific features\n\n### Working with openpyxl\n- Cell indices are 1-based (row=1, column=1 refers to cell A1)\n- Use `data_only=True` to read calculated values: `load_workbook('file.xlsx', data_only=True)`\n- **Warning**: If opened with `data_only=True` and saved, formulas are replaced with values and permanently lost\n- For large files: Use `read_only=True` for reading or `write_only=True` for writing\n- Formulas are preserved but not evaluated - use recalc.py to update values\n\n### Working with pandas\n- Specify data types to avoid inference issues: `pd.read_excel('file.xlsx', dtype={'id': str})`\n- For large files, read specific columns: `pd.read_excel('file.xlsx', usecols=['A', 'C', 'E'])`\n- Handle dates properly: `pd.read_excel('file.xlsx', parse_dates=['date_column'])`\n\n## Code Style Guidelines\n**IMPORTANT**: When generating Python code for Excel operations:\n- Write minimal, concise Python code without unnecessary comments\n- Avoid verbose variable names and redundant operations\n- Avoid unnecessary print statements\n\n**For Excel files themselves**:\n- Add comments to cells with complex formulas or important assumptions\n- Document data sources for hardcoded values\n- Include notes for key calculations and model sections"},"items":[{"name":"SKILL.md","path":"ruoyi-admin/src/main/resources/skills/docx/SKILL.md","rawUrl":"https://raw.githubusercontent.com/ageerle/ruoyi-ai/HEAD/ruoyi-admin/src/main/resources/skills/docx/SKILL.md","title":"docx","category":"anthropic-skill","format":"markdown","content":"---\nname: docx\ndescription: \"Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# DOCX creation, editing, and analysis\n\n## Overview\n\nA .docx file is a ZIP archive containing XML files.\n\n## Quick Reference\n\n| Task | Approach |\n|------|----------|\n| Read/analyze content | `pandoc` or unpack for raw XML |\n| Create new document | Use `docx-js` - see Creating New Documents below |\n| Edit existing document | Unpack → edit XML → repack - see Editing Existing Documents below |\n\n### Converting .doc to .docx\n\nLegacy `.doc` files must be converted before editing:\n\n```bash\npython scripts/office/soffice.py --headless --convert-to docx document.doc\n```\n\n### Reading Content\n\n```bash\n# Text extraction with tracked changes\npandoc --track-changes=all document.docx -o output.md\n\n# Raw XML access\npython scripts/office/unpack.py document.docx unpacked/\n```\n\n### Converting to Images\n\n```bash\npython scripts/office/soffice.py --headless --convert-to pdf document.docx\npdftoppm -jpeg -r 150 document.pdf page\n```\n\n### Accepting Tracked Changes\n\nTo produce a clean document with all tracked changes accepted (requires LibreOffice):\n\n```bash\npython scripts/accept_changes.py input.docx output.docx\n```\n\n---\n\n## Creating New Documents\n\nGenerate .docx files with JavaScript, then validate. Install: `npm install -g docx`\n\n### Setup\n```javascript\nconst { Document, Packer, Paragraph, TextRun, Table, TableRow, TableCell, ImageRun,\n        Header, Footer, AlignmentType, PageOrientation, LevelFormat, ExternalHyperlink,\n        InternalHyperlink, Bookmark, FootnoteReferenceRun, PositionalTab,\n        PositionalTabAlignment, PositionalTabRelativeTo, PositionalTabLeader,\n        TabStopType, TabStopPosition, Column, SectionType,\n        TableOfContents, HeadingLevel, BorderStyle, WidthType, ShadingType,\n        VerticalAlign, PageNumber, PageBreak } = require('docx');\n\nconst doc = new Document({ sections: [{ children: [/* content */] }] });\nPacker.toBuffer(doc).then(buffer => fs.writeFileSync(\"doc.docx\", buffer));\n```\n\n### Validation\nAfter creating the file, validate it. If validation fails, unpack, fix the XML, and repack.\n```bash\npython scripts/office/validate.py doc.docx\n```\n\n### Page Size\n\n```javascript\n// CRITICAL: docx-js defaults to A4, not US Letter\n// Always set page size explicitly for consistent results\nsections: [{\n  properties: {\n    page: {\n      size: {\n        width: 12240,   // 8.5 inches in DXA\n        height: 15840   // 11 inches in DXA\n      },\n      margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } // 1 inch margins\n    }\n  },\n  children: [/* content */]\n}]\n```\n\n**Common page sizes (DXA units, 1440 DXA = 1 inch):**\n\n| Paper | Width | Height | Content Width (1\" margins) |\n|-------|-------|--------|---------------------------|\n| US Letter | 12,240 | 15,840 | 9,360 |\n| A4 (default) | 11,906 | 16,838 | 9,026 |\n\n**Landscape orientation:** docx-js swaps width/height internally, so pass portrait dimensions and let it handle the swap:\n```javascript\nsize: {\n  width: 12240,   // Pass SHORT edge as width\n  height: 15840,  // Pass LONG edge as height\n  orientation: PageOrientation.LANDSCAPE  // docx-js swaps them in the XML\n},\n// Content width = 15840 - left margin - right margin (uses the long edge)\n```\n\n### Styles (Override Built-in Headings)\n\nUse Arial as the default font (universally supported). Keep titles black for readability.\n\n```javascript\nconst doc = new Document({\n  styles: {\n    default: { document: { run: { font: \"Arial\", size: 24 } } }, // 12pt default\n    paragraphStyles: [\n      // IMPORTANT: Use exact IDs to override built-in styles\n      { id: \"Heading1\", name: \"Heading 1\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 32, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 240, after: 240 }, outlineLevel: 0 } }, // outlineLevel required for TOC\n      { id: \"Heading2\", name: \"Heading 2\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 28, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 180, after: 180 }, outlineLevel: 1 } },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ heading: HeadingLevel.HEADING_1, children: [new TextRun(\"Title\")] }),\n    ]\n  }]\n});\n```\n\n### Lists (NEVER use unicode bullets)\n\n```javascript\n// ❌ WRONG - never manually insert bullet characters\nnew Paragraph({ children: [new TextRun(\"• Item\")] })  // BAD\nnew Paragraph({ children: [new TextRun(\"\\u2022 Item\")] })  // BAD\n\n// ✅ CORRECT - use numbering config with LevelFormat.BULLET\nconst doc = new Document({\n  numbering: {\n    config: [\n      { reference: \"bullets\",\n        levels: [{ level: 0, format: LevelFormat.BULLET, text: \"•\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n      { reference: \"numbers\",\n        levels: [{ level: 0, format: LevelFormat.DECIMAL, text: \"%1.\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ numbering: { reference: \"bullets\", level: 0 },\n        children: [new TextRun(\"Bullet item\")] }),\n      new Paragraph({ numbering: { reference: \"numbers\", level: 0 },\n        children: [new TextRun(\"Numbered item\")] }),\n    ]\n  }]\n});\n\n// ⚠️ Each reference creates INDEPENDENT numbering\n// Same reference = continues (1,2,3 then 4,5,6)\n// Different reference = restarts (1,2,3 then 1,2,3)\n```\n\n### Tables\n\n**CRITICAL: Tables need dual widths** - set both `columnWidths` on the table AND `width` on each cell. Without both, tables render incorrectly on some platforms.\n\n```javascript\n// CRITICAL: Always set table width for consistent rendering\n// CRITICAL: Use ShadingType.CLEAR (not SOLID) to prevent black backgrounds\nconst border = { style: BorderStyle.SINGLE, size: 1, color: \"CCCCCC\" };\nconst borders = { top: border, bottom: border, left: border, right: border };\n\nnew Table({\n  width: { size: 9360, type: WidthType.DXA }, // Always use DXA (percentages break in Google Docs)\n  columnWidths: [4680, 4680], // Must sum to table width (DXA: 1440 = 1 inch)\n  rows: [\n    new TableRow({\n      children: [\n        new TableCell({\n          borders,\n          width: { size: 4680, type: WidthType.DXA }, // Also set on each cell\n          shading: { fill: \"D5E8F0\", type: ShadingType.CLEAR }, // CLEAR not SOLID\n          margins: { top: 80, bottom: 80, left: 120, right: 120 }, // Cell padding (internal, not added to width)\n          children: [new Paragraph({ children: [new TextRun(\"Cell\")] })]\n        })\n      ]\n    })\n  ]\n})\n```\n\n**Table width calculation:**\n\nAlways use `WidthType.DXA` — `WidthType.PERCENTAGE` breaks in Google Docs.\n\n```javascript\n// Table width = sum of columnWidths = content width\n// US Letter with 1\" margins: 12240 - 2880 = 9360 DXA\nwidth: { size: 9360, type: WidthType.DXA },\ncolumnWidths: [7000, 2360]  // Must sum to table width\n```\n\n**Width rules:**\n- **Always use `WidthType.DXA`** — never `WidthType.PERCENTAGE` (incompatible with Google Docs)\n- Table width must equal the sum of `columnWidths`\n- Cell `width` must match corresponding `columnWidth`\n- Cell `margins` are internal padding - they reduce content area, not add to cell width\n- For full-width tables: use content width (page width minus left and right margins)\n\n### Images\n\n```javascript\n// CRITICAL: type parameter is REQUIRED\nnew Paragraph({\n  children: [new ImageRun({\n    type: \"png\", // Required: png, jpg, jpeg, gif, bmp, svg\n    data: fs.readFileSync(\"image.png\"),\n    transformation: { width: 200, height: 150 },\n    altText: { title: \"Title\", description: \"Desc\", name: \"Name\" } // All three required\n  })]\n})\n```\n\n### Page Breaks\n\n```javascript\n// CRITICAL: PageBreak must be inside a Paragraph\nnew Paragraph({ children: [new PageBreak()] })\n\n// Or use pageBreakBefore\nnew Paragraph({ pageBreakBefore: true, children: [new TextRun(\"New page\")] })\n```\n\n### Hyperlinks\n\n```javascript\n// External link\nnew Paragraph({\n  children: [new ExternalHyperlink({\n    children: [new TextRun({ text: \"Click here\", style: \"Hyperlink\" })],\n    link: \"https://example.com\",\n  })]\n})\n\n// Internal link (bookmark + reference)\n// 1. Create bookmark at destination\nnew Paragraph({ heading: HeadingLevel.HEADING_1, children: [\n  new Bookmark({ id: \"chapter1\", children: [new TextRun(\"Chapter 1\")] }),\n]})\n// 2. Link to it\nnew Paragraph({ children: [new InternalHyperlink({\n  children: [new TextRun({ text: \"See Chapter 1\", style: \"Hyperlink\" })],\n  anchor: \"chapter1\",\n})]})\n```\n\n### Footnotes\n\n```javascript\nconst doc = new Document({\n  footnotes: {\n    1: { children: [new Paragraph(\"Source: Annual Report 2024\")] },\n    2: { children: [new Paragraph(\"See appendix for methodology\")] },\n  },\n  sections: [{\n    children: [new Paragraph({\n      children: [\n        new TextRun(\"Revenue grew 15%\"),\n        new FootnoteReferenceRun(1),\n        new TextRun(\" using adjusted metrics\"),\n        new FootnoteReferenceRun(2),\n      ],\n    })]\n  }]\n});\n```\n\n### Tab Stops\n\n```javascript\n// Right-align text on same line (e.g., date opposite a title)\nnew Paragraph({\n  children: [\n    new TextRun(\"Company Name\"),\n    new TextRun(\"\\tJanuary 2025\"),\n  ],\n  tabStops: [{ type: TabStopType.RIGHT, position: TabStopPosition.MAX }],\n})\n\n// Dot leader (e.g., TOC-style)\nnew Paragraph({\n  children: [\n    new TextRun(\"Introduction\"),\n    new TextRun({ children: [\n      new PositionalTab({\n        alignment: PositionalTabAlignment.RIGHT,\n        relativeTo: PositionalTabRelativeTo.MARGIN,\n        leader: PositionalTabLeader.DOT,\n      }),\n      \"3\",\n    ]}),\n  ],\n})\n```\n\n### Multi-Column Layouts\n\n```javascript\n// Equal-width columns\nsections: [{\n  properties: {\n    column: {\n      count: 2,          // number of columns\n      space: 720,        // gap between columns in DXA (720 = 0.5 inch)\n      equalWidth: true,\n      separate: true,    // vertical line between columns\n    },\n  },\n  children: [/* content flows naturally across columns */]\n}]\n\n// Custom-width columns (equalWidth must be false)\nsections: [{\n  properties: {\n    column: {\n      equalWidth: false,\n      children: [\n        new Column({ width: 5400, space: 720 }),\n        new Column({ width: 3240 }),\n      ],\n    },\n  },\n  children: [/* content */]\n}]\n```\n\nForce a column break with a new section using `type: SectionType.NEXT_COLUMN`.\n\n### Table of Contents\n\n```javascript\n// CRITICAL: Headings must use HeadingLevel ONLY - no custom styles\nnew TableOfContents(\"Table of Contents\", { hyperlink: true, headingStyleRange: \"1-3\" })\n```\n\n### Headers/Footers\n\n```javascript\nsections: [{\n  properties: {\n    page: { margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } } // 1440 = 1 inch\n  },\n  headers: {\n    default: new Header({ children: [new Paragraph({ children: [new TextRun(\"Header\")] })] })\n  },\n  footers: {\n    default: new Footer({ children: [new Paragraph({\n      children: [new TextRun(\"Page \"), new TextRun({ children: [PageNumber.CURRENT] })]\n    })] })\n  },\n  children: [/* content */]\n}]\n```\n\n### Critical Rules for docx-js\n\n- **Set page size explicitly** - docx-js defaults to A4; use US Letter (12240 x 15840 DXA) for US documents\n- **Landscape: pass portrait dimensions** - docx-js swaps width/height internally; pass short edge as `width`, long edge as `height`, and set `orientation: PageOrientation.LANDSCAPE`\n- **Never use `\\n`** - use separate Paragraph elements\n- **Never use unicode bullets** - use `LevelFormat.BULLET` with numbering config\n- **PageBreak must be in Paragraph** - standalone creates invalid XML\n- **ImageRun requires `type`** - always specify png/jpg/etc\n- **Always set table `width` with DXA** - never use `WidthType.PERCENTAGE` (breaks in Google Docs)\n- **Tables need dual widths** - `columnWidths` array AND cell `width`, both must match\n- **Table width = sum of columnWidths** - for DXA, ensure they add up exactly\n- **Always add cell margins** - use `margins: { top: 80, bottom: 80, left: 120, right: 120 }` for readable padding\n- **Use `ShadingType.CLEAR`** - never SOLID for table shading\n- **Never use tables as dividers/rules** - cells have minimum height and render as empty boxes (including in headers/footers); use `border: { bottom: { style: BorderStyle.SINGLE, size: 6, color: \"2E75B6\", space: 1 } }` on a Paragraph instead. For two-column footers, use tab stops (see Tab Stops section), not tables\n- **TOC requires HeadingLevel only** - no custom styles on heading paragraphs\n- **Override built-in styles** - use exact IDs: \"Heading1\", \"Heading2\", etc.\n- **Include `outlineLevel`** - required for TOC (0 for H1, 1 for H2, etc.)\n\n---\n\n## Editing Existing Documents\n\n**Follow all 3 steps in order.**\n\n### Step 1: Unpack\n```bash\npython scripts/office/unpack.py document.docx unpacked/\n```\nExtracts XML, pretty-prints, merges adjacent runs, and converts smart quotes to XML entities (`&#x201C;` etc.) so they survive editing. Use `--merge-runs false` to skip run merging.\n\n### Step 2: Edit XML\n\nEdit files in `unpacked/word/`. See XML Reference below for patterns.\n\n**Use \"Claude\" as the author** for tracked changes and comments, unless the user explicitly requests use of a different name.\n\n**Use the Edit tool directly for string replacement. Do not write Python scripts.** Scripts introduce unnecessary complexity. The Edit tool shows exactly what is being replaced.\n\n**CRITICAL: Use smart quotes for new content.** When adding text with apostrophes or quotes, use XML entities to produce smart quotes:\n```xml\n<!-- Use these entities for professional typography -->\n<w:t>Here&#x2019;s a quote: &#x201C;Hello&#x201D;</w:t>\n```\n| Entity | Character |\n|--------|-----------|\n| `&#x2018;` | ‘ (left single) |\n| `&#x2019;` | ’ (right single / apostrophe) |\n| `&#x201C;` | “ (left double) |\n| `&#x201D;` | ” (right double) |\n\n**Adding comments:** Use `comment.py` to handle boilerplate across multiple XML files (text must be pre-escaped XML):\n```bash\npython scripts/comment.py unpacked/ 0 \"Comment text with &amp; and &#x2019;\"\npython scripts/comment.py unpacked/ 1 \"Reply text\" --parent 0  # reply to comment 0\npython scripts/comment.py unpacked/ 0 \"Text\" --author \"Custom Author\"  # custom author name\n```\nThen add markers to document.xml (see Comments in XML Reference).\n\n### Step 3: Pack\n```bash\npython scripts/office/pack.py unpacked/ output.docx --original document.docx\n```\nValidates with auto-repair, condenses XML, and creates DOCX. Use `--validate false` to skip.\n\n**Auto-repair will fix:**\n- `durableId` >= 0x7FFFFFFF (regenerates valid ID)\n- Missing `xml:space=\"preserve\"` on `<w:t>` with whitespace\n\n**Auto-repair won't fix:**\n- Malformed XML, invalid element nesting, missing relationships, schema violations\n\n### Common Pitfalls\n\n- **Replace entire `<w:r>` elements**: When adding tracked changes, replace the whole `<w:r>...</w:r>` block with `<w:del>...<w:ins>...` as siblings. Don't inject tracked change tags inside a run.\n- **Preserve `<w:rPr>` formatting**: Copy the original run's `<w:rPr>` block into your tracked change runs to maintain bold, font size, etc.\n\n---\n\n## XML Reference\n\n### Schema Compliance\n\n- **Element order in `<w:pPr>`**: `<w:pStyle>`, `<w:numPr>`, `<w:spacing>`, `<w:ind>`, `<w:jc>`, `<w:rPr>` last\n- **Whitespace**: Add `xml:space=\"preserve\"` to `<w:t>` with leading/trailing spaces\n- **RSIDs**: Must be 8-digit hex (e.g., `00AB1234`)\n\n### Tracked Changes\n\n**Insertion:**\n```xml\n<w:ins w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:t>inserted text</w:t></w:r>\n</w:ins>\n```\n\n**Deletion:**\n```xml\n<w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n```\n\n**Inside `<w:del>`**: Use `<w:delText>` instead of `<w:t>`, and `<w:delInstrText>` instead of `<w:instrText>`.\n\n**Minimal edits** - only mark what changes:\n```xml\n<!-- Change \"30 days\" to \"60 days\" -->\n<w:r><w:t>The term is </w:t></w:r>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:delText>30</w:delText></w:r>\n</w:del>\n<w:ins w:id=\"2\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:t>60</w:t></w:r>\n</w:ins>\n<w:r><w:t> days.</w:t></w:r>\n```\n\n**Deleting entire paragraphs/list items** - when removing ALL content from a paragraph, also mark the paragraph mark as deleted so it merges with the next paragraph. Add `<w:del/>` inside `<w:pPr><w:rPr>`:\n```xml\n<w:p>\n  <w:pPr>\n    <w:numPr>...</w:numPr>  <!-- list numbering if present -->\n    <w:rPr>\n      <w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\"/>\n    </w:rPr>\n  </w:pPr>\n  <w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n    <w:r><w:delText>Entire paragraph content being deleted...</w:delText></w:r>\n  </w:del>\n</w:p>\n```\nWithout the `<w:del/>` in `<w:pPr><w:rPr>`, accepting changes leaves an empty paragraph/list item.\n\n**Rejecting another author's insertion** - nest deletion inside their insertion:\n```xml\n<w:ins w:author=\"Jane\" w:id=\"5\">\n  <w:del w:author=\"Claude\" w:id=\"10\">\n    <w:r><w:delText>their inserted text</w:delText></w:r>\n  </w:del>\n</w:ins>\n```\n\n**Restoring another author's deletion** - add insertion after (don't modify their deletion):\n```xml\n<w:del w:author=\"Jane\" w:id=\"5\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n<w:ins w:author=\"Claude\" w:id=\"10\">\n  <w:r><w:t>deleted text</w:t></w:r>\n</w:ins>\n```\n\n### Comments\n\nAfter running `comment.py` (see Step 2), add markers to document.xml. For replies, use `--parent` flag and nest markers inside the parent's.\n\n**CRITICAL: `<w:commentRangeStart>` and `<w:commentRangeEnd>` are siblings of `<w:r>`, never inside `<w:r>`.**\n\n```xml\n<!-- Comment markers are direct children of w:p, never inside w:r -->\n<w:commentRangeStart w:id=\"0\"/>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted</w:delText></w:r>\n</w:del>\n<w:r><w:t> more text</w:t></w:r>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n\n<!-- Comment 0 with reply 1 nested inside -->\n<w:commentRangeStart w:id=\"0\"/>\n  <w:commentRangeStart w:id=\"1\"/>\n  <w:r><w:t>text</w:t></w:r>\n  <w:commentRangeEnd w:id=\"1\"/>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"1\"/></w:r>\n```\n\n### Images\n\n1. Add image file to `word/media/`\n2. Add relationship to `word/_rels/document.xml.rels`:\n```xml\n<Relationship Id=\"rId5\" Type=\".../image\" Target=\"media/image1.png\"/>\n```\n3. Add content type to `[Content_Types].xml`:\n```xml\n<Default Extension=\"png\" ContentType=\"image/png\"/>\n```\n4. Reference in document.xml:\n```xml\n<w:drawing>\n  <wp:inline>\n    <wp:extent cx=\"914400\" cy=\"914400\"/>  <!-- EMUs: 914400 = 1 inch -->\n    <a:graphic>\n      <a:graphicData uri=\".../picture\">\n        <pic:pic>\n          <pic:blipFill><a:blip r:embed=\"rId5\"/></pic:blipFill>\n        </pic:pic>\n      </a:graphicData>\n    </a:graphic>\n  </wp:inline>\n</w:drawing>\n```\n\n---\n\n## Dependencies\n\n- **pandoc**: Text extraction\n- **docx**: `npm install -g docx` (new documents)\n- **LibreOffice**: PDF conversion (auto-configured for sandboxed environments via `scripts/office/soffice.py`)\n- **Poppler**: `pdftoppm` for images\n","frontmatter":{"name":"docx","description":"Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.","license":"Proprietary. LICENSE.txt has complete terms"},"isInternal":false,"tokens":5707,"sizeBytes":20084},{"name":"forms.md","path":"ruoyi-admin/src/main/resources/skills/pdf/forms.md","rawUrl":"https://raw.githubusercontent.com/ageerle/ruoyi-ai/HEAD/ruoyi-admin/src/main/resources/skills/pdf/forms.md","title":"Pdf Skill","category":"anthropic-skill","format":"markdown","content":"**CRITICAL: You MUST complete these steps in order. Do not skip ahead to writing code.**\n\nIf you need to fill out a PDF form, first check to see if the PDF has fillable form fields. Run this script from this file's directory:\n `python scripts/check_fillable_fields <file.pdf>`, and depending on the result go to either the \"Fillable fields\" or \"Non-fillable fields\" and follow those instructions.\n\n# Fillable fields\nIf the PDF has fillable form fields:\n- Run this script from this file's directory: `python scripts/extract_form_field_info.py <input.pdf> <field_info.json>`. It will create a JSON file with a list of fields in this format:\n```\n[\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"rect\": ([left, bottom, right, top] bounding box in PDF coordinates, y=0 is the bottom of the page),\n    \"type\": (\"text\", \"checkbox\", \"radio_group\", or \"choice\"),\n  },\n  // Checkboxes have \"checked_value\" and \"unchecked_value\" properties:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"checkbox\",\n    \"checked_value\": (Set the field to this value to check the checkbox),\n    \"unchecked_value\": (Set the field to this value to uncheck the checkbox),\n  },\n  // Radio groups have a \"radio_options\" list with the possible choices.\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"radio_group\",\n    \"radio_options\": [\n      {\n        \"value\": (set the field to this value to select this radio option),\n        \"rect\": (bounding box for the radio button for this option)\n      },\n      // Other radio options\n    ]\n  },\n  // Multiple choice fields have a \"choice_options\" list with the possible choices:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"choice\",\n    \"choice_options\": [\n      {\n        \"value\": (set the field to this value to select this option),\n        \"text\": (display text of the option)\n      },\n      // Other choice options\n    ],\n  }\n]\n```\n- Convert the PDF to PNGs (one image for each page) with this script (run from this file's directory):\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThen analyze the images to determine the purpose of each form field (make sure to convert the bounding box PDF coordinates to image coordinates).\n- Create a `field_values.json` file in this format with the values to be entered for each field:\n```\n[\n  {\n    \"field_id\": \"last_name\", // Must match the field_id from `extract_form_field_info.py`\n    \"description\": \"The user's last name\",\n    \"page\": 1, // Must match the \"page\" value in field_info.json\n    \"value\": \"Simpson\"\n  },\n  {\n    \"field_id\": \"Checkbox12\",\n    \"description\": \"Checkbox to be checked if the user is 18 or over\",\n    \"page\": 1,\n    \"value\": \"/On\" // If this is a checkbox, use its \"checked_value\" value to check it. If it's a radio button group, use one of the \"value\" values in \"radio_options\".\n  },\n  // more fields\n]\n```\n- Run the `fill_fillable_fields.py` script from this file's directory to create a filled-in PDF:\n`python scripts/fill_fillable_fields.py <input pdf> <field_values.json> <output pdf>`\nThis script will verify that the field IDs and values you provide are valid; if it prints error messages, correct the appropriate fields and try again.\n\n# Non-fillable fields\nIf the PDF doesn't have fillable form fields, you'll need to visually determine where the data should be added and create text annotations. Follow the below steps *exactly*. You MUST perform all of these steps to ensure that the the form is accurately completed. Details for each step are below.\n- Convert the PDF to PNG images and determine field bounding boxes.\n- Create a JSON file with field information and validation images showing the bounding boxes.\n- Validate the the bounding boxes.\n- Use the bounding boxes to fill in the form.\n\n## Step 1: Visual Analysis (REQUIRED)\n- Convert the PDF to PNG images. Run this script from this file's directory:\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThe script will create a PNG image for each page in the PDF.\n- Carefully examine each PNG image and identify all form fields and areas where the user should enter data. For each form field where the user should enter text, determine bounding boxes for both the form field label, and the area where the user should enter text. The label and entry bounding boxes MUST NOT INTERSECT; the text entry box should only include the area where data should be entered. Usually this area will be immediately to the side, above, or below its label. Entry bounding boxes must be tall and wide enough to contain their text.\n\nThese are some examples of form structures that you might see:\n\n*Label inside box*\n```\n┌────────────────────────┐\n│ Name:                  │\n└────────────────────────┘\n```\nThe input area should be to the right of the \"Name\" label and extend to the edge of the box.\n\n*Label before line*\n```\nEmail: _______________________\n```\nThe input area should be above the line and include its entire width.\n\n*Label under line*\n```\n_________________________\nName\n```\nThe input area should be above the line and include the entire width of the line. This is common for signature and date fields.\n\n*Label above line*\n```\nPlease enter any special requests:\n________________________________________________\n```\nThe input area should extend from the bottom of the label to the line, and should include the entire width of the line.\n\n*Checkboxes*\n```\nAre you a US citizen? Yes □  No □\n```\nFor checkboxes:\n- Look for small square boxes (□) - these are the actual checkboxes to target. They may be to the left or right of their labels.\n- Distinguish between label text (\"Yes\", \"No\") and the clickable checkbox squares.\n- The entry bounding box should cover ONLY the small square, not the text label.\n\n### Step 2: Create fields.json and validation images (REQUIRED)\n- Create a file named `fields.json` with information for the form fields and bounding boxes in this format:\n```\n{\n  \"pages\": [\n    {\n      \"page_number\": 1,\n      \"image_width\": (first page image width in pixels),\n      \"image_height\": (first page image height in pixels),\n    },\n    {\n      \"page_number\": 2,\n      \"image_width\": (second page image width in pixels),\n      \"image_height\": (second page image height in pixels),\n    }\n    // additional pages\n  ],\n  \"form_fields\": [\n    // Example for a text field.\n    {\n      \"page_number\": 1,\n      \"description\": \"The user's last name should be entered here\",\n      // Bounding boxes are [left, top, right, bottom]. The bounding boxes for the label and text entry should not overlap.\n      \"field_label\": \"Last name\",\n      \"label_bounding_box\": [30, 125, 95, 142],\n      \"entry_bounding_box\": [100, 125, 280, 142],\n      \"entry_text\": {\n        \"text\": \"Johnson\", // This text will be added as an annotation at the entry_bounding_box location\n        \"font_size\": 14, // optional, defaults to 14\n        \"font_color\": \"000000\", // optional, RRGGBB format, defaults to 000000 (black)\n      }\n    },\n    // Example for a checkbox. TARGET THE SQUARE for the entry bounding box, NOT THE TEXT\n    {\n      \"page_number\": 2,\n      \"description\": \"Checkbox that should be checked if the user is over 18\",\n      \"entry_bounding_box\": [140, 525, 155, 540],  // Small box over checkbox square\n      \"field_label\": \"Yes\",\n      \"label_bounding_box\": [100, 525, 132, 540],  // Box containing \"Yes\" text\n      // Use \"X\" to check a checkbox.\n      \"entry_text\": {\n        \"text\": \"X\",\n      }\n    }\n    // additional form field entries\n  ]\n}\n```\n\nCreate validation images by running this script from this file's directory for each page:\n`python scripts/create_validation_image.py <page_number> <path_to_fields.json> <input_image_path> <output_image_path>\n\nThe validation images will have red rectangles where text should be entered, and blue rectangles covering label text.\n\n### Step 3: Validate Bounding Boxes (REQUIRED)\n#### Automated intersection check\n- Verify that none of bounding boxes intersect and that the entry bounding boxes are tall enough by checking the fields.json file with the `check_bounding_boxes.py` script (run from this file's directory):\n`python scripts/check_bounding_boxes.py <JSON file>`\n\nIf there are errors, reanalyze the relevant fields, adjust the bounding boxes, and iterate until there are no remaining errors. Remember: label (blue) bounding boxes should contain text labels, entry (red) boxes should not.\n\n#### Manual image inspection\n**CRITICAL: Do not proceed without visually inspecting validation images**\n- Red rectangles must ONLY cover input areas\n- Red rectangles MUST NOT contain any text\n- Blue rectangles should contain label text\n- For checkboxes:\n  - Red rectangle MUST be centered on the checkbox square\n  - Blue rectangle should cover the text label for the checkbox\n\n- If any rectangles look wrong, fix fields.json, regenerate the validation images, and verify again. Repeat this process until the bounding boxes are fully accurate.\n\n\n### Step 4: Add annotations to the PDF\nRun this script from this file's directory to create a filled-out PDF using the information in fields.json:\n`python scripts/fill_pdf_form_with_annotations.py <input_pdf_path> <path_to_fields.json> <output_pdf_path>\n","isInternal":false,"tokens":2201,"sizeBytes":9438},{"name":"reference.md","path":"ruoyi-admin/src/main/resources/skills/pdf/reference.md","rawUrl":"https://raw.githubusercontent.com/ageerle/ruoyi-ai/HEAD/ruoyi-admin/src/main/resources/skills/pdf/reference.md","title":"Pdf Skill","category":"anthropic-skill","format":"markdown","content":"# PDF Processing Advanced Reference\n\nThis document contains advanced PDF processing features, detailed examples, and additional libraries not covered in the main skill instructions.\n\n## pypdfium2 Library (Apache/BSD License)\n\n### Overview\npypdfium2 is a Python binding for PDFium (Chromium's PDF library). It's excellent for fast PDF rendering, image generation, and serves as a PyMuPDF replacement.\n\n### Render PDF to Images\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\n\n# Load PDF\npdf = pdfium.PdfDocument(\"document.pdf\")\n\n# Render page to image\npage = pdf[0]  # First page\nbitmap = page.render(\n    scale=2.0,  # Higher resolution\n    rotation=0  # No rotation\n)\n\n# Convert to PIL Image\nimg = bitmap.to_pil()\nimg.save(\"page_1.png\", \"PNG\")\n\n# Process multiple pages\nfor i, page in enumerate(pdf):\n    bitmap = page.render(scale=1.5)\n    img = bitmap.to_pil()\n    img.save(f\"page_{i+1}.jpg\", \"JPEG\", quality=90)\n```\n\n### Extract Text with pypdfium2\n```python\nimport pypdfium2 as pdfium\n\npdf = pdfium.PdfDocument(\"document.pdf\")\nfor i, page in enumerate(pdf):\n    text = page.get_text()\n    print(f\"Page {i+1} text length: {len(text)} chars\")\n```\n\n## JavaScript Libraries\n\n### pdf-lib (MIT License)\n\npdf-lib is a powerful JavaScript library for creating and modifying PDF documents in any JavaScript environment.\n\n#### Load and Manipulate Existing PDF\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function manipulatePDF() {\n    // Load existing PDF\n    const existingPdfBytes = fs.readFileSync('input.pdf');\n    const pdfDoc = await PDFDocument.load(existingPdfBytes);\n\n    // Get page count\n    const pageCount = pdfDoc.getPageCount();\n    console.log(`Document has ${pageCount} pages`);\n\n    // Add new page\n    const newPage = pdfDoc.addPage([600, 400]);\n    newPage.drawText('Added by pdf-lib', {\n        x: 100,\n        y: 300,\n        size: 16\n    });\n\n    // Save modified PDF\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('modified.pdf', pdfBytes);\n}\n```\n\n#### Create Complex PDFs from Scratch\n```javascript\nimport { PDFDocument, rgb, StandardFonts } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function createPDF() {\n    const pdfDoc = await PDFDocument.create();\n\n    // Add fonts\n    const helveticaFont = await pdfDoc.embedFont(StandardFonts.Helvetica);\n    const helveticaBold = await pdfDoc.embedFont(StandardFonts.HelveticaBold);\n\n    // Add page\n    const page = pdfDoc.addPage([595, 842]); // A4 size\n    const { width, height } = page.getSize();\n\n    // Add text with styling\n    page.drawText('Invoice #12345', {\n        x: 50,\n        y: height - 50,\n        size: 18,\n        font: helveticaBold,\n        color: rgb(0.2, 0.2, 0.8)\n    });\n\n    // Add rectangle (header background)\n    page.drawRectangle({\n        x: 40,\n        y: height - 100,\n        width: width - 80,\n        height: 30,\n        color: rgb(0.9, 0.9, 0.9)\n    });\n\n    // Add table-like content\n    const items = [\n        ['Item', 'Qty', 'Price', 'Total'],\n        ['Widget', '2', '$50', '$100'],\n        ['Gadget', '1', '$75', '$75']\n    ];\n\n    let yPos = height - 150;\n    items.forEach(row => {\n        let xPos = 50;\n        row.forEach(cell => {\n            page.drawText(cell, {\n                x: xPos,\n                y: yPos,\n                size: 12,\n                font: helveticaFont\n            });\n            xPos += 120;\n        });\n        yPos -= 25;\n    });\n\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('created.pdf', pdfBytes);\n}\n```\n\n#### Advanced Merge and Split Operations\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function mergePDFs() {\n    // Create new document\n    const mergedPdf = await PDFDocument.create();\n\n    // Load source PDFs\n    const pdf1Bytes = fs.readFileSync('doc1.pdf');\n    const pdf2Bytes = fs.readFileSync('doc2.pdf');\n\n    const pdf1 = await PDFDocument.load(pdf1Bytes);\n    const pdf2 = await PDFDocument.load(pdf2Bytes);\n\n    // Copy pages from first PDF\n    const pdf1Pages = await mergedPdf.copyPages(pdf1, pdf1.getPageIndices());\n    pdf1Pages.forEach(page => mergedPdf.addPage(page));\n\n    // Copy specific pages from second PDF (pages 0, 2, 4)\n    const pdf2Pages = await mergedPdf.copyPages(pdf2, [0, 2, 4]);\n    pdf2Pages.forEach(page => mergedPdf.addPage(page));\n\n    const mergedPdfBytes = await mergedPdf.save();\n    fs.writeFileSync('merged.pdf', mergedPdfBytes);\n}\n```\n\n### pdfjs-dist (Apache License)\n\nPDF.js is Mozilla's JavaScript library for rendering PDFs in the browser.\n\n#### Basic PDF Loading and Rendering\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\n// Configure worker (important for performance)\npdfjsLib.GlobalWorkerOptions.workerSrc = './pdf.worker.js';\n\nasync function renderPDF() {\n    // Load PDF\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    console.log(`Loaded PDF with ${pdf.numPages} pages`);\n\n    // Get first page\n    const page = await pdf.getPage(1);\n    const viewport = page.getViewport({ scale: 1.5 });\n\n    // Render to canvas\n    const canvas = document.createElement('canvas');\n    const context = canvas.getContext('2d');\n    canvas.height = viewport.height;\n    canvas.width = viewport.width;\n\n    const renderContext = {\n        canvasContext: context,\n        viewport: viewport\n    };\n\n    await page.render(renderContext).promise;\n    document.body.appendChild(canvas);\n}\n```\n\n#### Extract Text with Coordinates\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractText() {\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    let fullText = '';\n\n    // Extract text from all pages\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const textContent = await page.getTextContent();\n\n        const pageText = textContent.items\n            .map(item => item.str)\n            .join(' ');\n\n        fullText += `\\n--- Page ${i} ---\\n${pageText}`;\n\n        // Get text with coordinates for advanced processing\n        const textWithCoords = textContent.items.map(item => ({\n            text: item.str,\n            x: item.transform[4],\n            y: item.transform[5],\n            width: item.width,\n            height: item.height\n        }));\n    }\n\n    console.log(fullText);\n    return fullText;\n}\n```\n\n#### Extract Annotations and Forms\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractAnnotations() {\n    const loadingTask = pdfjsLib.getDocument('annotated.pdf');\n    const pdf = await loadingTask.promise;\n\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const annotations = await page.getAnnotations();\n\n        annotations.forEach(annotation => {\n            console.log(`Annotation type: ${annotation.subtype}`);\n            console.log(`Content: ${annotation.contents}`);\n            console.log(`Coordinates: ${JSON.stringify(annotation.rect)}`);\n        });\n    }\n}\n```\n\n## Advanced Command-Line Operations\n\n### poppler-utils Advanced Features\n\n#### Extract Text with Bounding Box Coordinates\n```bash\n# Extract text with bounding box coordinates (essential for structured data)\npdftotext -bbox-layout document.pdf output.xml\n\n# The XML output contains precise coordinates for each text element\n```\n\n#### Advanced Image Conversion\n```bash\n# Convert to PNG images with specific resolution\npdftoppm -png -r 300 document.pdf output_prefix\n\n# Convert specific page range with high resolution\npdftoppm -png -r 600 -f 1 -l 3 document.pdf high_res_pages\n\n# Convert to JPEG with quality setting\npdftoppm -jpeg -jpegopt quality=85 -r 200 document.pdf jpeg_output\n```\n\n#### Extract Embedded Images\n```bash\n# Extract all embedded images with metadata\npdfimages -j -p document.pdf page_images\n\n# List image info without extracting\npdfimages -list document.pdf\n\n# Extract images in their original format\npdfimages -all document.pdf images/img\n```\n\n### qpdf Advanced Features\n\n#### Complex Page Manipulation\n```bash\n# Split PDF into groups of pages\nqpdf --split-pages=3 input.pdf output_group_%02d.pdf\n\n# Extract specific pages with complex ranges\nqpdf input.pdf --pages input.pdf 1,3-5,8,10-end -- extracted.pdf\n\n# Merge specific pages from multiple PDFs\nqpdf --empty --pages doc1.pdf 1-3 doc2.pdf 5-7 doc3.pdf 2,4 -- combined.pdf\n```\n\n#### PDF Optimization and Repair\n```bash\n# Optimize PDF for web (linearize for streaming)\nqpdf --linearize input.pdf optimized.pdf\n\n# Remove unused objects and compress\nqpdf --optimize-level=all input.pdf compressed.pdf\n\n# Attempt to repair corrupted PDF structure\nqpdf --check input.pdf\nqpdf --fix-qdf damaged.pdf repaired.pdf\n\n# Show detailed PDF structure for debugging\nqpdf --show-all-pages input.pdf > structure.txt\n```\n\n#### Advanced Encryption\n```bash\n# Add password protection with specific permissions\nqpdf --encrypt user_pass owner_pass 256 --print=none --modify=none -- input.pdf encrypted.pdf\n\n# Check encryption status\nqpdf --show-encryption encrypted.pdf\n\n# Remove password protection (requires password)\nqpdf --password=secret123 --decrypt encrypted.pdf decrypted.pdf\n```\n\n## Advanced Python Techniques\n\n### pdfplumber Advanced Features\n\n#### Extract Text with Precise Coordinates\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract all text with coordinates\n    chars = page.chars\n    for char in chars[:10]:  # First 10 characters\n        print(f\"Char: '{char['text']}' at x:{char['x0']:.1f} y:{char['y0']:.1f}\")\n    \n    # Extract text by bounding box (left, top, right, bottom)\n    bbox_text = page.within_bbox((100, 100, 400, 200)).extract_text()\n```\n\n#### Advanced Table Extraction with Custom Settings\n```python\nimport pdfplumber\nimport pandas as pd\n\nwith pdfplumber.open(\"complex_table.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract tables with custom settings for complex layouts\n    table_settings = {\n        \"vertical_strategy\": \"lines\",\n        \"horizontal_strategy\": \"lines\",\n        \"snap_tolerance\": 3,\n        \"intersection_tolerance\": 15\n    }\n    tables = page.extract_tables(table_settings)\n    \n    # Visual debugging for table extraction\n    img = page.to_image(resolution=150)\n    img.save(\"debug_layout.png\")\n```\n\n### reportlab Advanced Features\n\n#### Create Professional Reports with Tables\n```python\nfrom reportlab.platypus import SimpleDocTemplate, Table, TableStyle, Paragraph\nfrom reportlab.lib.styles import getSampleStyleSheet\nfrom reportlab.lib import colors\n\n# Sample data\ndata = [\n    ['Product', 'Q1', 'Q2', 'Q3', 'Q4'],\n    ['Widgets', '120', '135', '142', '158'],\n    ['Gadgets', '85', '92', '98', '105']\n]\n\n# Create PDF with table\ndoc = SimpleDocTemplate(\"report.pdf\")\nelements = []\n\n# Add title\nstyles = getSampleStyleSheet()\ntitle = Paragraph(\"Quarterly Sales Report\", styles['Title'])\nelements.append(title)\n\n# Add table with advanced styling\ntable = Table(data)\ntable.setStyle(TableStyle([\n    ('BACKGROUND', (0, 0), (-1, 0), colors.grey),\n    ('TEXTCOLOR', (0, 0), (-1, 0), colors.whitesmoke),\n    ('ALIGN', (0, 0), (-1, -1), 'CENTER'),\n    ('FONTNAME', (0, 0), (-1, 0), 'Helvetica-Bold'),\n    ('FONTSIZE', (0, 0), (-1, 0), 14),\n    ('BOTTOMPADDING', (0, 0), (-1, 0), 12),\n    ('BACKGROUND', (0, 1), (-1, -1), colors.beige),\n    ('GRID', (0, 0), (-1, -1), 1, colors.black)\n]))\nelements.append(table)\n\ndoc.build(elements)\n```\n\n## Complex Workflows\n\n### Extract Figures/Images from PDF\n\n#### Method 1: Using pdfimages (fastest)\n```bash\n# Extract all images with original quality\npdfimages -all document.pdf images/img\n```\n\n#### Method 2: Using pypdfium2 + Image Processing\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\nimport numpy as np\n\ndef extract_figures(pdf_path, output_dir):\n    pdf = pdfium.PdfDocument(pdf_path)\n    \n    for page_num, page in enumerate(pdf):\n        # Render high-resolution page\n        bitmap = page.render(scale=3.0)\n        img = bitmap.to_pil()\n        \n        # Convert to numpy for processing\n        img_array = np.array(img)\n        \n        # Simple figure detection (non-white regions)\n        mask = np.any(img_array != [255, 255, 255], axis=2)\n        \n        # Find contours and extract bounding boxes\n        # (This is simplified - real implementation would need more sophisticated detection)\n        \n        # Save detected figures\n        # ... implementation depends on specific needs\n```\n\n### Batch PDF Processing with Error Handling\n```python\nimport os\nimport glob\nfrom pypdf import PdfReader, PdfWriter\nimport logging\n\nlogging.basicConfig(level=logging.INFO)\nlogger = logging.getLogger(__name__)\n\ndef batch_process_pdfs(input_dir, operation='merge'):\n    pdf_files = glob.glob(os.path.join(input_dir, \"*.pdf\"))\n    \n    if operation == 'merge':\n        writer = PdfWriter()\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                for page in reader.pages:\n                    writer.add_page(page)\n                logger.info(f\"Processed: {pdf_file}\")\n            except Exception as e:\n                logger.error(f\"Failed to process {pdf_file}: {e}\")\n                continue\n        \n        with open(\"batch_merged.pdf\", \"wb\") as output:\n            writer.write(output)\n    \n    elif operation == 'extract_text':\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                text = \"\"\n                for page in reader.pages:\n                    text += page.extract_text()\n                \n                output_file = pdf_file.replace('.pdf', '.txt')\n                with open(output_file, 'w', encoding='utf-8') as f:\n                    f.write(text)\n                logger.info(f\"Extracted text from: {pdf_file}\")\n                \n            except Exception as e:\n                logger.error(f\"Failed to extract text from {pdf_file}: {e}\")\n                continue\n```\n\n### Advanced PDF Cropping\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\n# Crop page (left, bottom, right, top in points)\npage = reader.pages[0]\npage.mediabox.left = 50\npage.mediabox.bottom = 50\npage.mediabox.right = 550\npage.mediabox.top = 750\n\nwriter.add_page(page)\nwith open(\"cropped.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Performance Optimization Tips\n\n### 1. For Large PDFs\n- Use streaming approaches instead of loading entire PDF in memory\n- Use `qpdf --split-pages` for splitting large files\n- Process pages individually with pypdfium2\n\n### 2. For Text Extraction\n- `pdftotext -bbox-layout` is fastest for plain text extraction\n- Use pdfplumber for structured data and tables\n- Avoid `pypdf.extract_text()` for very large documents\n\n### 3. For Image Extraction\n- `pdfimages` is much faster than rendering pages\n- Use low resolution for previews, high resolution for final output\n\n### 4. For Form Filling\n- pdf-lib maintains form structure better than most alternatives\n- Pre-validate form fields before processing\n\n### 5. Memory Management\n```python\n# Process PDFs in chunks\ndef process_large_pdf(pdf_path, chunk_size=10):\n    reader = PdfReader(pdf_path)\n    total_pages = len(reader.pages)\n    \n    for start_idx in range(0, total_pages, chunk_size):\n        end_idx = min(start_idx + chunk_size, total_pages)\n        writer = PdfWriter()\n        \n        for i in range(start_idx, end_idx):\n            writer.add_page(reader.pages[i])\n        \n        # Process chunk\n        with open(f\"chunk_{start_idx//chunk_size}.pdf\", \"wb\") as output:\n            writer.write(output)\n```\n\n## Troubleshooting Common Issues\n\n### Encrypted PDFs\n```python\n# Handle password-protected PDFs\nfrom pypdf import PdfReader\n\ntry:\n    reader = PdfReader(\"encrypted.pdf\")\n    if reader.is_encrypted:\n        reader.decrypt(\"password\")\nexcept Exception as e:\n    print(f\"Failed to decrypt: {e}\")\n```\n\n### Corrupted PDFs\n```bash\n# Use qpdf to repair\nqpdf --check corrupted.pdf\nqpdf --replace-input corrupted.pdf\n```\n\n### Text Extraction Issues\n```python\n# Fallback to OCR for scanned PDFs\nimport pytesseract\nfrom pdf2image import convert_from_path\n\ndef extract_text_with_ocr(pdf_path):\n    images = convert_from_path(pdf_path)\n    text = \"\"\n    for i, image in enumerate(images):\n        text += pytesseract.image_to_string(image)\n    return text\n```\n\n## License Information\n\n- **pypdf**: BSD License\n- **pdfplumber**: MIT License\n- **pypdfium2**: Apache/BSD License\n- **reportlab**: BSD License\n- **poppler-utils**: GPL-2 License\n- **qpdf**: Apache License\n- **pdf-lib**: MIT License\n- **pdfjs-dist**: Apache License","isInternal":false,"tokens":4070,"sizeBytes":16692},{"name":"SKILL.md","path":"ruoyi-admin/src/main/resources/skills/pdf/SKILL.md","rawUrl":"https://raw.githubusercontent.com/ageerle/ruoyi-ai/HEAD/ruoyi-admin/src/main/resources/skills/pdf/SKILL.md","title":"pdf","category":"anthropic-skill","format":"markdown","content":"---\nname: pdf\ndescription: Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# PDF Processing Guide\n\n## Overview\n\nThis guide covers essential PDF processing operations using Python libraries and command-line tools. For advanced features, JavaScript libraries, and detailed examples, see reference.md. If you need to fill out a PDF form, read forms.md and follow its instructions.\n\n## Quick Start\n\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Read a PDF\nreader = PdfReader(\"document.pdf\")\nprint(f\"Pages: {len(reader.pages)}\")\n\n# Extract text\ntext = \"\"\nfor page in reader.pages:\n    text += page.extract_text()\n```\n\n## Python Libraries\n\n### pypdf - Basic Operations\n\n#### Merge PDFs\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nwriter = PdfWriter()\nfor pdf_file in [\"doc1.pdf\", \"doc2.pdf\", \"doc3.pdf\"]:\n    reader = PdfReader(pdf_file)\n    for page in reader.pages:\n        writer.add_page(page)\n\nwith open(\"merged.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n#### Split PDF\n```python\nreader = PdfReader(\"input.pdf\")\nfor i, page in enumerate(reader.pages):\n    writer = PdfWriter()\n    writer.add_page(page)\n    with open(f\"page_{i+1}.pdf\", \"wb\") as output:\n        writer.write(output)\n```\n\n#### Extract Metadata\n```python\nreader = PdfReader(\"document.pdf\")\nmeta = reader.metadata\nprint(f\"Title: {meta.title}\")\nprint(f\"Author: {meta.author}\")\nprint(f\"Subject: {meta.subject}\")\nprint(f\"Creator: {meta.creator}\")\n```\n\n#### Rotate Pages\n```python\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\npage = reader.pages[0]\npage.rotate(90)  # Rotate 90 degrees clockwise\nwriter.add_page(page)\n\nwith open(\"rotated.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### pdfplumber - Text and Table Extraction\n\n#### Extract Text with Layout\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for page in pdf.pages:\n        text = page.extract_text()\n        print(text)\n```\n\n#### Extract Tables\n```python\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for i, page in enumerate(pdf.pages):\n        tables = page.extract_tables()\n        for j, table in enumerate(tables):\n            print(f\"Table {j+1} on page {i+1}:\")\n            for row in table:\n                print(row)\n```\n\n#### Advanced Table Extraction\n```python\nimport pandas as pd\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    all_tables = []\n    for page in pdf.pages:\n        tables = page.extract_tables()\n        for table in tables:\n            if table:  # Check if table is not empty\n                df = pd.DataFrame(table[1:], columns=table[0])\n                all_tables.append(df)\n\n# Combine all tables\nif all_tables:\n    combined_df = pd.concat(all_tables, ignore_index=True)\n    combined_df.to_excel(\"extracted_tables.xlsx\", index=False)\n```\n\n### reportlab - Create PDFs\n\n#### Basic PDF Creation\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.pdfgen import canvas\n\nc = canvas.Canvas(\"hello.pdf\", pagesize=letter)\nwidth, height = letter\n\n# Add text\nc.drawString(100, height - 100, \"Hello World!\")\nc.drawString(100, height - 120, \"This is a PDF created with reportlab\")\n\n# Add a line\nc.line(100, height - 140, 400, height - 140)\n\n# Save\nc.save()\n```\n\n#### Create PDF with Multiple Pages\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, PageBreak\nfrom reportlab.lib.styles import getSampleStyleSheet\n\ndoc = SimpleDocTemplate(\"report.pdf\", pagesize=letter)\nstyles = getSampleStyleSheet()\nstory = []\n\n# Add content\ntitle = Paragraph(\"Report Title\", styles['Title'])\nstory.append(title)\nstory.append(Spacer(1, 12))\n\nbody = Paragraph(\"This is the body of the report. \" * 20, styles['Normal'])\nstory.append(body)\nstory.append(PageBreak())\n\n# Page 2\nstory.append(Paragraph(\"Page 2\", styles['Heading1']))\nstory.append(Paragraph(\"Content for page 2\", styles['Normal']))\n\n# Build PDF\ndoc.build(story)\n```\n\n## Command-Line Tools\n\n### pdftotext (poppler-utils)\n```bash\n# Extract text\npdftotext input.pdf output.txt\n\n# Extract text preserving layout\npdftotext -layout input.pdf output.txt\n\n# Extract specific pages\npdftotext -f 1 -l 5 input.pdf output.txt  # Pages 1-5\n```\n\n### qpdf\n```bash\n# Merge PDFs\nqpdf --empty --pages file1.pdf file2.pdf -- merged.pdf\n\n# Split pages\nqpdf input.pdf --pages . 1-5 -- pages1-5.pdf\nqpdf input.pdf --pages . 6-10 -- pages6-10.pdf\n\n# Rotate pages\nqpdf input.pdf output.pdf --rotate=+90:1  # Rotate page 1 by 90 degrees\n\n# Remove password\nqpdf --password=mypassword --decrypt encrypted.pdf decrypted.pdf\n```\n\n### pdftk (if available)\n```bash\n# Merge\npdftk file1.pdf file2.pdf cat output merged.pdf\n\n# Split\npdftk input.pdf burst\n\n# Rotate\npdftk input.pdf rotate 1east output rotated.pdf\n```\n\n## Common Tasks\n\n### Extract Text from Scanned PDFs\n```python\n# Requires: pip install pytesseract pdf2image\nimport pytesseract\nfrom pdf2image import convert_from_path\n\n# Convert PDF to images\nimages = convert_from_path('scanned.pdf')\n\n# OCR each page\ntext = \"\"\nfor i, image in enumerate(images):\n    text += f\"Page {i+1}:\\n\"\n    text += pytesseract.image_to_string(image)\n    text += \"\\n\\n\"\n\nprint(text)\n```\n\n### Add Watermark\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Create watermark (or load existing)\nwatermark = PdfReader(\"watermark.pdf\").pages[0]\n\n# Apply to all pages\nreader = PdfReader(\"document.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    page.merge_page(watermark)\n    writer.add_page(page)\n\nwith open(\"watermarked.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### Extract Images\n```bash\n# Using pdfimages (poppler-utils)\npdfimages -j input.pdf output_prefix\n\n# This extracts all images as output_prefix-000.jpg, output_prefix-001.jpg, etc.\n```\n\n### Password Protection\n```python\nfrom pypdf import PdfReader, PdfWriter\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    writer.add_page(page)\n\n# Add password\nwriter.encrypt(\"userpassword\", \"ownerpassword\")\n\nwith open(\"encrypted.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Quick Reference\n\n| Task | Best Tool | Command/Code |\n|------|-----------|--------------|\n| Merge PDFs | pypdf | `writer.add_page(page)` |\n| Split PDFs | pypdf | One page per file |\n| Extract text | pdfplumber | `page.extract_text()` |\n| Extract tables | pdfplumber | `page.extract_tables()` |\n| Create PDFs | reportlab | Canvas or Platypus |\n| Command line merge | qpdf | `qpdf --empty --pages ...` |\n| OCR scanned PDFs | pytesseract | Convert to image first |\n| Fill PDF forms | pdf-lib or pypdf (see forms.md) | See forms.md |\n\n## Next Steps\n\n- For advanced pypdfium2 usage, see reference.md\n- For JavaScript libraries (pdf-lib), see reference.md\n- If you need to fill out a PDF form, follow the instructions in forms.md\n- For troubleshooting guides, see reference.md\n","frontmatter":{"name":"pdf","description":"Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.","license":"Proprietary. LICENSE.txt has complete terms"},"isInternal":false,"tokens":1804,"sizeBytes":7068},{"name":"SKILL.md","path":"ruoyi-admin/src/main/resources/skills/xlsx/SKILL.md","rawUrl":"https://raw.githubusercontent.com/ageerle/ruoyi-ai/HEAD/ruoyi-admin/src/main/resources/skills/xlsx/SKILL.md","title":"xlsx","category":"anthropic-skill","format":"markdown","content":"---\nname: xlsx\ndescription: \"Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# Requirements for Outputs\n\n## All Excel files\n\n### Zero Formula Errors\n- Every Excel model MUST be delivered with ZERO formula errors (#REF!, #DIV/0!, #VALUE!, #N/A, #NAME?)\n\n### Preserve Existing Templates (when updating templates)\n- Study and EXACTLY match existing format, style, and conventions when modifying files\n- Never impose standardized formatting on files with established patterns\n- Existing template conventions ALWAYS override these guidelines\n\n## Financial models\n\n### Color Coding Standards\nUnless otherwise stated by the user or existing template\n\n#### Industry-Standard Color Conventions\n- **Blue text (RGB: 0,0,255)**: Hardcoded inputs, and numbers users will change for scenarios\n- **Black text (RGB: 0,0,0)**: ALL formulas and calculations\n- **Green text (RGB: 0,128,0)**: Links pulling from other worksheets within same workbook\n- **Red text (RGB: 255,0,0)**: External links to other files\n- **Yellow background (RGB: 255,255,0)**: Key assumptions needing attention or cells that need to be updated\n\n### Number Formatting Standards\n\n#### Required Format Rules\n- **Years**: Format as text strings (e.g., \"2024\" not \"2,024\")\n- **Currency**: Use $#,##0 format; ALWAYS specify units in headers (\"Revenue ($mm)\")\n- **Zeros**: Use number formatting to make all zeros \"-\", including percentages (e.g., \"$#,##0;($#,##0);-\")\n- **Percentages**: Default to 0.0% format (one decimal)\n- **Multiples**: Format as 0.0x for valuation multiples (EV/EBITDA, P/E)\n- **Negative numbers**: Use parentheses (123) not minus -123\n\n### Formula Construction Rules\n\n#### Assumptions Placement\n- Place ALL assumptions (growth rates, margins, multiples, etc.) in separate assumption cells\n- Use cell references instead of hardcoded values in formulas\n- Example: Use =B5*(1+$B$6) instead of =B5*1.05\n\n#### Formula Error Prevention\n- Verify all cell references are correct\n- Check for off-by-one errors in ranges\n- Ensure consistent formulas across all projection periods\n- Test with edge cases (zero values, negative numbers)\n- Verify no unintended circular references\n\n#### Documentation Requirements for Hardcodes\n- Comment or in cells beside (if end of table). Format: \"Source: [System/Document], [Date], [Specific Reference], [URL if applicable]\"\n- Examples:\n  - \"Source: Company 10-K, FY2024, Page 45, Revenue Note, [SEC EDGAR URL]\"\n  - \"Source: Company 10-Q, Q2 2025, Exhibit 99.1, [SEC EDGAR URL]\"\n  - \"Source: Bloomberg Terminal, 8/15/2025, AAPL US Equity\"\n  - \"Source: FactSet, 8/20/2025, Consensus Estimates Screen\"\n\n# XLSX creation, editing, and analysis\n\n## Overview\n\nA user may ask you to create, edit, or analyze the contents of an .xlsx file. You have different tools and workflows available for different tasks.\n\n## Important Requirements\n\n**LibreOffice Required for Formula Recalculation**: You can assume LibreOffice is installed for recalculating formula values using the `recalc.py` script. The script automatically configures LibreOffice on first run\n\n## Reading and analyzing data\n\n### Data analysis with pandas\nFor data analysis, visualization, and basic operations, use **pandas** which provides powerful data manipulation capabilities:\n\n```python\nimport pandas as pd\n\n# Read Excel\ndf = pd.read_excel('file.xlsx')  # Default: first sheet\nall_sheets = pd.read_excel('file.xlsx', sheet_name=None)  # All sheets as dict\n\n# Analyze\ndf.head()      # Preview data\ndf.info()      # Column info\ndf.describe()  # Statistics\n\n# Write Excel\ndf.to_excel('output.xlsx', index=False)\n```\n\n## Excel File Workflows\n\n## CRITICAL: Use Formulas, Not Hardcoded Values\n\n**Always use Excel formulas instead of calculating values in Python and hardcoding them.** This ensures the spreadsheet remains dynamic and updateable.\n\n### ❌ WRONG - Hardcoding Calculated Values\n```python\n# Bad: Calculating in Python and hardcoding result\ntotal = df['Sales'].sum()\nsheet['B10'] = total  # Hardcodes 5000\n\n# Bad: Computing growth rate in Python\ngrowth = (df.iloc[-1]['Revenue'] - df.iloc[0]['Revenue']) / df.iloc[0]['Revenue']\nsheet['C5'] = growth  # Hardcodes 0.15\n\n# Bad: Python calculation for average\navg = sum(values) / len(values)\nsheet['D20'] = avg  # Hardcodes 42.5\n```\n\n### ✅ CORRECT - Using Excel Formulas\n```python\n# Good: Let Excel calculate the sum\nsheet['B10'] = '=SUM(B2:B9)'\n\n# Good: Growth rate as Excel formula\nsheet['C5'] = '=(C4-C2)/C2'\n\n# Good: Average using Excel function\nsheet['D20'] = '=AVERAGE(D2:D19)'\n```\n\nThis applies to ALL calculations - totals, percentages, ratios, differences, etc. The spreadsheet should be able to recalculate when source data changes.\n\n## Common Workflow\n1. **Choose tool**: pandas for data, openpyxl for formulas/formatting\n2. **Create/Load**: Create new workbook or load existing file\n3. **Modify**: Add/edit data, formulas, and formatting\n4. **Save**: Write to file\n5. **Recalculate formulas (MANDATORY IF USING FORMULAS)**: Use the recalc.py script\n   ```bash\n   python recalc.py output.xlsx\n   ```\n6. **Verify and fix any errors**: \n   - The script returns JSON with error details\n   - If `status` is `errors_found`, check `error_summary` for specific error types and locations\n   - Fix the identified errors and recalculate again\n   - Common errors to fix:\n     - `#REF!`: Invalid cell references\n     - `#DIV/0!`: Division by zero\n     - `#VALUE!`: Wrong data type in formula\n     - `#NAME?`: Unrecognized formula name\n\n### Creating new Excel files\n\n```python\n# Using openpyxl for formulas and formatting\nfrom openpyxl import Workbook\nfrom openpyxl.styles import Font, PatternFill, Alignment\n\nwb = Workbook()\nsheet = wb.active\n\n# Add data\nsheet['A1'] = 'Hello'\nsheet['B1'] = 'World'\nsheet.append(['Row', 'of', 'data'])\n\n# Add formula\nsheet['B2'] = '=SUM(A1:A10)'\n\n# Formatting\nsheet['A1'].font = Font(bold=True, color='FF0000')\nsheet['A1'].fill = PatternFill('solid', start_color='FFFF00')\nsheet['A1'].alignment = Alignment(horizontal='center')\n\n# Column width\nsheet.column_dimensions['A'].width = 20\n\nwb.save('output.xlsx')\n```\n\n### Editing existing Excel files\n\n```python\n# Using openpyxl to preserve formulas and formatting\nfrom openpyxl import load_workbook\n\n# Load existing file\nwb = load_workbook('existing.xlsx')\nsheet = wb.active  # or wb['SheetName'] for specific sheet\n\n# Working with multiple sheets\nfor sheet_name in wb.sheetnames:\n    sheet = wb[sheet_name]\n    print(f\"Sheet: {sheet_name}\")\n\n# Modify cells\nsheet['A1'] = 'New Value'\nsheet.insert_rows(2)  # Insert row at position 2\nsheet.delete_cols(3)  # Delete column 3\n\n# Add new sheet\nnew_sheet = wb.create_sheet('NewSheet')\nnew_sheet['A1'] = 'Data'\n\nwb.save('modified.xlsx')\n```\n\n## Recalculating formulas\n\nExcel files created or modified by openpyxl contain formulas as strings but not calculated values. Use the provided `recalc.py` script to recalculate formulas:\n\n```bash\n# Recalculate existing Excel file\npython recalc.py <excel_file> [timeout_seconds]\n\n# Create new Excel workbook\npython recalc.py --create <output_file> '<json_data>'\n```\n\nExamples:\n```bash\npython recalc.py output.xlsx 30\n\npython recalc.py --create output.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Value\"], [\"Test\", 100]]}}'\n```\n\nThe script:\n- Automatically sets up LibreOffice macro on first run\n- Recalculates all formulas in all sheets\n- Scans ALL cells for Excel errors (#REF!, #DIV/0!, etc.)\n- Returns JSON with detailed error locations and counts\n- Works on Windows, Linux, and macOS\n- Can create new workbooks from JSON configuration\n\n## Creating workbooks with recalc.py\n\nThe `--create` mode allows you to create new Excel workbooks directly from JSON configuration:\n\n```bash\npython recalc.py --create <output_file> '<json_data>'\n```\n\n**Default Behavior**: If you provide a relative path (e.g., `test.xlsx`), the file will be created in the **project root** directory (`ruoyi-ai-v3/`). To save to a specific directory, use an absolute path.\n\n### JSON Format\n\n```json\n{\n  \"Sheet1\": {\n    \"data\": [\n      [\"Header1\", \"Header2\", \"Header3\"],\n      [\"Value1\", \"Value2\", \"=A2*B2\"],\n      [\"Value3\", \"Value4\", \"=SUM(C2:C10)\"]\n    ]\n  },\n  \"Sheet2\": {\n    \"data\": [\n      [\"Name\", \"Score\"],\n      [\"Alice\", 95],\n      [\"Bob\", 87]\n    ]\n  }\n}\n```\n\n### Examples\n\n**Simple data table (saves to project root)**:\n```bash\npython recalc.py --create data.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Age\"], [\"John\", 30], [\"Jane\", 28]]}}'\n# Creates: ruoyi-ai-v3/data.xlsx\n```\n\n**With formulas (saves to project root)**:\n```bash\npython recalc.py --create model.xlsx '{\"Sheet1\": {\"data\": [[\"A\", \"B\", \"Sum\"], [10, 20, \"=A2+B2\"], [5, 15, \"=A3+B3\"]]}}'\n# Creates: ruoyi-ai-v3/model.xlsx\n```\n\n**Save to custom directory (use absolute path)**:\n```bash\npython recalc.py --create \"D:\\Downloads\\custom.xlsx\" '{\"Sheet1\": {\"data\": [[\"A\", \"B\"], [1, 2]]}}'\n# Creates: D:\\Downloads\\custom.xlsx\n```\n\n### File Path Options\n\n| Input | Output Location |\n|-------|-----------------|\n| `test.xlsx` | `ruoyi-ai-v3/test.xlsx` (project root) |\n| `output/test.xlsx` | `ruoyi-ai-v3/output/test.xlsx` |\n| `D:\\Downloads\\test.xlsx` | `D:\\Downloads\\test.xlsx` (absolute path) |\n\n### Features\n\n- First row is automatically formatted as bold header\n- Supports both values and Excel formulas (strings starting with `=`)\n- Multiple sheets can be created in one command\n- Output file extension determines format (.xlsx, .xlsm, etc.)\n\n## Formula Verification Checklist\n\nQuick checks to ensure formulas work correctly:\n\n### Essential Verification\n- [ ] **Test 2-3 sample references**: Verify they pull correct values before building full model\n- [ ] **Column mapping**: Confirm Excel columns match (e.g., column 64 = BL, not BK)\n- [ ] **Row offset**: Remember Excel rows are 1-indexed (DataFrame row 5 = Excel row 6)\n\n### Common Pitfalls\n- [ ] **NaN handling**: Check for null values with `pd.notna()`\n- [ ] **Far-right columns**: FY data often in columns 50+ \n- [ ] **Multiple matches**: Search all occurrences, not just first\n- [ ] **Division by zero**: Check denominators before using `/` in formulas (#DIV/0!)\n- [ ] **Wrong references**: Verify all cell references point to intended cells (#REF!)\n- [ ] **Cross-sheet references**: Use correct format (Sheet1!A1) for linking sheets\n\n### Formula Testing Strategy\n- [ ] **Start small**: Test formulas on 2-3 cells before applying broadly\n- [ ] **Verify dependencies**: Check all cells referenced in formulas exist\n- [ ] **Test edge cases**: Include zero, negative, and very large values\n\n### Interpreting recalc.py Output\nThe script returns JSON with error details:\n```json\n{\n  \"status\": \"success\",           // or \"errors_found\"\n  \"total_errors\": 0,              // Total error count\n  \"total_formulas\": 42,           // Number of formulas in file\n  \"error_summary\": {              // Only present if errors found\n    \"#REF!\": {\n      \"count\": 2,\n      \"locations\": [\"Sheet1!B5\", \"Sheet1!C10\"]\n    }\n  }\n}\n```\n\n## Best Practices\n\n### Library Selection\n- **pandas**: Best for data analysis, bulk operations, and simple data export\n- **openpyxl**: Best for complex formatting, formulas, and Excel-specific features\n\n### Working with openpyxl\n- Cell indices are 1-based (row=1, column=1 refers to cell A1)\n- Use `data_only=True` to read calculated values: `load_workbook('file.xlsx', data_only=True)`\n- **Warning**: If opened with `data_only=True` and saved, formulas are replaced with values and permanently lost\n- For large files: Use `read_only=True` for reading or `write_only=True` for writing\n- Formulas are preserved but not evaluated - use recalc.py to update values\n\n### Working with pandas\n- Specify data types to avoid inference issues: `pd.read_excel('file.xlsx', dtype={'id': str})`\n- For large files, read specific columns: `pd.read_excel('file.xlsx', usecols=['A', 'C', 'E'])`\n- Handle dates properly: `pd.read_excel('file.xlsx', parse_dates=['date_column'])`\n\n## Code Style Guidelines\n**IMPORTANT**: When generating Python code for Excel operations:\n- Write minimal, concise Python code without unnecessary comments\n- Avoid verbose variable names and redundant operations\n- Avoid unnecessary print statements\n\n**For Excel files themselves**:\n- Add comments to cells with complex formulas or important assumptions\n- Document data sources for hardcoded values\n- Include notes for key calculations and model sections","frontmatter":{"name":"xlsx","description":"Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas","license":"Proprietary. LICENSE.txt has complete terms"},"isInternal":false,"tokens":3267,"sizeBytes":12744}],"systemPromptSnippet":"<yakaai_skills repo=\"ageerle/ruoyi-ai\">\n<!-- File: ruoyi-admin/src/main/resources/skills/docx/SKILL.md (Tokens: ~5707 | Category: anthropic-skill) -->\n---\nname: docx\ndescription: \"Use this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# DOCX creation, editing, and analysis\n\n## Overview\n\nA .docx file is a ZIP archive containing XML files.\n\n## Quick Reference\n\n| Task | Approach |\n|------|----------|\n| Read/analyze content | `pandoc` or unpack for raw XML |\n| Create new document | Use `docx-js` - see Creating New Documents below |\n| Edit existing document | Unpack → edit XML → repack - see Editing Existing Documents below |\n\n### Converting .doc to .docx\n\nLegacy `.doc` files must be converted before editing:\n\n```bash\npython scripts/office/soffice.py --headless --convert-to docx document.doc\n```\n\n### Reading Content\n\n```bash\n# Text extraction with tracked changes\npandoc --track-changes=all document.docx -o output.md\n\n# Raw XML access\npython scripts/office/unpack.py document.docx unpacked/\n```\n\n### Converting to Images\n\n```bash\npython scripts/office/soffice.py --headless --convert-to pdf document.docx\npdftoppm -jpeg -r 150 document.pdf page\n```\n\n### Accepting Tracked Changes\n\nTo produce a clean document with all tracked changes accepted (requires LibreOffice):\n\n```bash\npython scripts/accept_changes.py input.docx output.docx\n```\n\n---\n\n## Creating New Documents\n\nGenerate .docx files with JavaScript, then validate. Install: `npm install -g docx`\n\n### Setup\n```javascript\nconst { Document, Packer, Paragraph, TextRun, Table, TableRow, TableCell, ImageRun,\n        Header, Footer, AlignmentType, PageOrientation, LevelFormat, ExternalHyperlink,\n        InternalHyperlink, Bookmark, FootnoteReferenceRun, PositionalTab,\n        PositionalTabAlignment, PositionalTabRelativeTo, PositionalTabLeader,\n        TabStopType, TabStopPosition, Column, SectionType,\n        TableOfContents, HeadingLevel, BorderStyle, WidthType, ShadingType,\n        VerticalAlign, PageNumber, PageBreak } = require('docx');\n\nconst doc = new Document({ sections: [{ children: [/* content */] }] });\nPacker.toBuffer(doc).then(buffer => fs.writeFileSync(\"doc.docx\", buffer));\n```\n\n### Validation\nAfter creating the file, validate it. If validation fails, unpack, fix the XML, and repack.\n```bash\npython scripts/office/validate.py doc.docx\n```\n\n### Page Size\n\n```javascript\n// CRITICAL: docx-js defaults to A4, not US Letter\n// Always set page size explicitly for consistent results\nsections: [{\n  properties: {\n    page: {\n      size: {\n        width: 12240,   // 8.5 inches in DXA\n        height: 15840   // 11 inches in DXA\n      },\n      margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } // 1 inch margins\n    }\n  },\n  children: [/* content */]\n}]\n```\n\n**Common page sizes (DXA units, 1440 DXA = 1 inch):**\n\n| Paper | Width | Height | Content Width (1\" margins) |\n|-------|-------|--------|---------------------------|\n| US Letter | 12,240 | 15,840 | 9,360 |\n| A4 (default) | 11,906 | 16,838 | 9,026 |\n\n**Landscape orientation:** docx-js swaps width/height internally, so pass portrait dimensions and let it handle the swap:\n```javascript\nsize: {\n  width: 12240,   // Pass SHORT edge as width\n  height: 15840,  // Pass LONG edge as height\n  orientation: PageOrientation.LANDSCAPE  // docx-js swaps them in the XML\n},\n// Content width = 15840 - left margin - right margin (uses the long edge)\n```\n\n### Styles (Override Built-in Headings)\n\nUse Arial as the default font (universally supported). Keep titles black for readability.\n\n```javascript\nconst doc = new Document({\n  styles: {\n    default: { document: { run: { font: \"Arial\", size: 24 } } }, // 12pt default\n    paragraphStyles: [\n      // IMPORTANT: Use exact IDs to override built-in styles\n      { id: \"Heading1\", name: \"Heading 1\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 32, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 240, after: 240 }, outlineLevel: 0 } }, // outlineLevel required for TOC\n      { id: \"Heading2\", name: \"Heading 2\", basedOn: \"Normal\", next: \"Normal\", quickFormat: true,\n        run: { size: 28, bold: true, font: \"Arial\" },\n        paragraph: { spacing: { before: 180, after: 180 }, outlineLevel: 1 } },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ heading: HeadingLevel.HEADING_1, children: [new TextRun(\"Title\")] }),\n    ]\n  }]\n});\n```\n\n### Lists (NEVER use unicode bullets)\n\n```javascript\n// ❌ WRONG - never manually insert bullet characters\nnew Paragraph({ children: [new TextRun(\"• Item\")] })  // BAD\nnew Paragraph({ children: [new TextRun(\"\\u2022 Item\")] })  // BAD\n\n// ✅ CORRECT - use numbering config with LevelFormat.BULLET\nconst doc = new Document({\n  numbering: {\n    config: [\n      { reference: \"bullets\",\n        levels: [{ level: 0, format: LevelFormat.BULLET, text: \"•\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n      { reference: \"numbers\",\n        levels: [{ level: 0, format: LevelFormat.DECIMAL, text: \"%1.\", alignment: AlignmentType.LEFT,\n          style: { paragraph: { indent: { left: 720, hanging: 360 } } } }] },\n    ]\n  },\n  sections: [{\n    children: [\n      new Paragraph({ numbering: { reference: \"bullets\", level: 0 },\n        children: [new TextRun(\"Bullet item\")] }),\n      new Paragraph({ numbering: { reference: \"numbers\", level: 0 },\n        children: [new TextRun(\"Numbered item\")] }),\n    ]\n  }]\n});\n\n// ⚠️ Each reference creates INDEPENDENT numbering\n// Same reference = continues (1,2,3 then 4,5,6)\n// Different reference = restarts (1,2,3 then 1,2,3)\n```\n\n### Tables\n\n**CRITICAL: Tables need dual widths** - set both `columnWidths` on the table AND `width` on each cell. Without both, tables render incorrectly on some platforms.\n\n```javascript\n// CRITICAL: Always set table width for consistent rendering\n// CRITICAL: Use ShadingType.CLEAR (not SOLID) to prevent black backgrounds\nconst border = { style: BorderStyle.SINGLE, size: 1, color: \"CCCCCC\" };\nconst borders = { top: border, bottom: border, left: border, right: border };\n\nnew Table({\n  width: { size: 9360, type: WidthType.DXA }, // Always use DXA (percentages break in Google Docs)\n  columnWidths: [4680, 4680], // Must sum to table width (DXA: 1440 = 1 inch)\n  rows: [\n    new TableRow({\n      children: [\n        new TableCell({\n          borders,\n          width: { size: 4680, type: WidthType.DXA }, // Also set on each cell\n          shading: { fill: \"D5E8F0\", type: ShadingType.CLEAR }, // CLEAR not SOLID\n          margins: { top: 80, bottom: 80, left: 120, right: 120 }, // Cell padding (internal, not added to width)\n          children: [new Paragraph({ children: [new TextRun(\"Cell\")] })]\n        })\n      ]\n    })\n  ]\n})\n```\n\n**Table width calculation:**\n\nAlways use `WidthType.DXA` — `WidthType.PERCENTAGE` breaks in Google Docs.\n\n```javascript\n// Table width = sum of columnWidths = content width\n// US Letter with 1\" margins: 12240 - 2880 = 9360 DXA\nwidth: { size: 9360, type: WidthType.DXA },\ncolumnWidths: [7000, 2360]  // Must sum to table width\n```\n\n**Width rules:**\n- **Always use `WidthType.DXA`** — never `WidthType.PERCENTAGE` (incompatible with Google Docs)\n- Table width must equal the sum of `columnWidths`\n- Cell `width` must match corresponding `columnWidth`\n- Cell `margins` are internal padding - they reduce content area, not add to cell width\n- For full-width tables: use content width (page width minus left and right margins)\n\n### Images\n\n```javascript\n// CRITICAL: type parameter is REQUIRED\nnew Paragraph({\n  children: [new ImageRun({\n    type: \"png\", // Required: png, jpg, jpeg, gif, bmp, svg\n    data: fs.readFileSync(\"image.png\"),\n    transformation: { width: 200, height: 150 },\n    altText: { title: \"Title\", description: \"Desc\", name: \"Name\" } // All three required\n  })]\n})\n```\n\n### Page Breaks\n\n```javascript\n// CRITICAL: PageBreak must be inside a Paragraph\nnew Paragraph({ children: [new PageBreak()] })\n\n// Or use pageBreakBefore\nnew Paragraph({ pageBreakBefore: true, children: [new TextRun(\"New page\")] })\n```\n\n### Hyperlinks\n\n```javascript\n// External link\nnew Paragraph({\n  children: [new ExternalHyperlink({\n    children: [new TextRun({ text: \"Click here\", style: \"Hyperlink\" })],\n    link: \"https://example.com\",\n  })]\n})\n\n// Internal link (bookmark + reference)\n// 1. Create bookmark at destination\nnew Paragraph({ heading: HeadingLevel.HEADING_1, children: [\n  new Bookmark({ id: \"chapter1\", children: [new TextRun(\"Chapter 1\")] }),\n]})\n// 2. Link to it\nnew Paragraph({ children: [new InternalHyperlink({\n  children: [new TextRun({ text: \"See Chapter 1\", style: \"Hyperlink\" })],\n  anchor: \"chapter1\",\n})]})\n```\n\n### Footnotes\n\n```javascript\nconst doc = new Document({\n  footnotes: {\n    1: { children: [new Paragraph(\"Source: Annual Report 2024\")] },\n    2: { children: [new Paragraph(\"See appendix for methodology\")] },\n  },\n  sections: [{\n    children: [new Paragraph({\n      children: [\n        new TextRun(\"Revenue grew 15%\"),\n        new FootnoteReferenceRun(1),\n        new TextRun(\" using adjusted metrics\"),\n        new FootnoteReferenceRun(2),\n      ],\n    })]\n  }]\n});\n```\n\n### Tab Stops\n\n```javascript\n// Right-align text on same line (e.g., date opposite a title)\nnew Paragraph({\n  children: [\n    new TextRun(\"Company Name\"),\n    new TextRun(\"\\tJanuary 2025\"),\n  ],\n  tabStops: [{ type: TabStopType.RIGHT, position: TabStopPosition.MAX }],\n})\n\n// Dot leader (e.g., TOC-style)\nnew Paragraph({\n  children: [\n    new TextRun(\"Introduction\"),\n    new TextRun({ children: [\n      new PositionalTab({\n        alignment: PositionalTabAlignment.RIGHT,\n        relativeTo: PositionalTabRelativeTo.MARGIN,\n        leader: PositionalTabLeader.DOT,\n      }),\n      \"3\",\n    ]}),\n  ],\n})\n```\n\n### Multi-Column Layouts\n\n```javascript\n// Equal-width columns\nsections: [{\n  properties: {\n    column: {\n      count: 2,          // number of columns\n      space: 720,        // gap between columns in DXA (720 = 0.5 inch)\n      equalWidth: true,\n      separate: true,    // vertical line between columns\n    },\n  },\n  children: [/* content flows naturally across columns */]\n}]\n\n// Custom-width columns (equalWidth must be false)\nsections: [{\n  properties: {\n    column: {\n      equalWidth: false,\n      children: [\n        new Column({ width: 5400, space: 720 }),\n        new Column({ width: 3240 }),\n      ],\n    },\n  },\n  children: [/* content */]\n}]\n```\n\nForce a column break with a new section using `type: SectionType.NEXT_COLUMN`.\n\n### Table of Contents\n\n```javascript\n// CRITICAL: Headings must use HeadingLevel ONLY - no custom styles\nnew TableOfContents(\"Table of Contents\", { hyperlink: true, headingStyleRange: \"1-3\" })\n```\n\n### Headers/Footers\n\n```javascript\nsections: [{\n  properties: {\n    page: { margin: { top: 1440, right: 1440, bottom: 1440, left: 1440 } } // 1440 = 1 inch\n  },\n  headers: {\n    default: new Header({ children: [new Paragraph({ children: [new TextRun(\"Header\")] })] })\n  },\n  footers: {\n    default: new Footer({ children: [new Paragraph({\n      children: [new TextRun(\"Page \"), new TextRun({ children: [PageNumber.CURRENT] })]\n    })] })\n  },\n  children: [/* content */]\n}]\n```\n\n### Critical Rules for docx-js\n\n- **Set page size explicitly** - docx-js defaults to A4; use US Letter (12240 x 15840 DXA) for US documents\n- **Landscape: pass portrait dimensions** - docx-js swaps width/height internally; pass short edge as `width`, long edge as `height`, and set `orientation: PageOrientation.LANDSCAPE`\n- **Never use `\\n`** - use separate Paragraph elements\n- **Never use unicode bullets** - use `LevelFormat.BULLET` with numbering config\n- **PageBreak must be in Paragraph** - standalone creates invalid XML\n- **ImageRun requires `type`** - always specify png/jpg/etc\n- **Always set table `width` with DXA** - never use `WidthType.PERCENTAGE` (breaks in Google Docs)\n- **Tables need dual widths** - `columnWidths` array AND cell `width`, both must match\n- **Table width = sum of columnWidths** - for DXA, ensure they add up exactly\n- **Always add cell margins** - use `margins: { top: 80, bottom: 80, left: 120, right: 120 }` for readable padding\n- **Use `ShadingType.CLEAR`** - never SOLID for table shading\n- **Never use tables as dividers/rules** - cells have minimum height and render as empty boxes (including in headers/footers); use `border: { bottom: { style: BorderStyle.SINGLE, size: 6, color: \"2E75B6\", space: 1 } }` on a Paragraph instead. For two-column footers, use tab stops (see Tab Stops section), not tables\n- **TOC requires HeadingLevel only** - no custom styles on heading paragraphs\n- **Override built-in styles** - use exact IDs: \"Heading1\", \"Heading2\", etc.\n- **Include `outlineLevel`** - required for TOC (0 for H1, 1 for H2, etc.)\n\n---\n\n## Editing Existing Documents\n\n**Follow all 3 steps in order.**\n\n### Step 1: Unpack\n```bash\npython scripts/office/unpack.py document.docx unpacked/\n```\nExtracts XML, pretty-prints, merges adjacent runs, and converts smart quotes to XML entities (`&#x201C;` etc.) so they survive editing. Use `--merge-runs false` to skip run merging.\n\n### Step 2: Edit XML\n\nEdit files in `unpacked/word/`. See XML Reference below for patterns.\n\n**Use \"Claude\" as the author** for tracked changes and comments, unless the user explicitly requests use of a different name.\n\n**Use the Edit tool directly for string replacement. Do not write Python scripts.** Scripts introduce unnecessary complexity. The Edit tool shows exactly what is being replaced.\n\n**CRITICAL: Use smart quotes for new content.** When adding text with apostrophes or quotes, use XML entities to produce smart quotes:\n```xml\n<!-- Use these entities for professional typography -->\n<w:t>Here&#x2019;s a quote: &#x201C;Hello&#x201D;</w:t>\n```\n| Entity | Character |\n|--------|-----------|\n| `&#x2018;` | ‘ (left single) |\n| `&#x2019;` | ’ (right single / apostrophe) |\n| `&#x201C;` | “ (left double) |\n| `&#x201D;` | ” (right double) |\n\n**Adding comments:** Use `comment.py` to handle boilerplate across multiple XML files (text must be pre-escaped XML):\n```bash\npython scripts/comment.py unpacked/ 0 \"Comment text with &amp; and &#x2019;\"\npython scripts/comment.py unpacked/ 1 \"Reply text\" --parent 0  # reply to comment 0\npython scripts/comment.py unpacked/ 0 \"Text\" --author \"Custom Author\"  # custom author name\n```\nThen add markers to document.xml (see Comments in XML Reference).\n\n### Step 3: Pack\n```bash\npython scripts/office/pack.py unpacked/ output.docx --original document.docx\n```\nValidates with auto-repair, condenses XML, and creates DOCX. Use `--validate false` to skip.\n\n**Auto-repair will fix:**\n- `durableId` >= 0x7FFFFFFF (regenerates valid ID)\n- Missing `xml:space=\"preserve\"` on `<w:t>` with whitespace\n\n**Auto-repair won't fix:**\n- Malformed XML, invalid element nesting, missing relationships, schema violations\n\n### Common Pitfalls\n\n- **Replace entire `<w:r>` elements**: When adding tracked changes, replace the whole `<w:r>...</w:r>` block with `<w:del>...<w:ins>...` as siblings. Don't inject tracked change tags inside a run.\n- **Preserve `<w:rPr>` formatting**: Copy the original run's `<w:rPr>` block into your tracked change runs to maintain bold, font size, etc.\n\n---\n\n## XML Reference\n\n### Schema Compliance\n\n- **Element order in `<w:pPr>`**: `<w:pStyle>`, `<w:numPr>`, `<w:spacing>`, `<w:ind>`, `<w:jc>`, `<w:rPr>` last\n- **Whitespace**: Add `xml:space=\"preserve\"` to `<w:t>` with leading/trailing spaces\n- **RSIDs**: Must be 8-digit hex (e.g., `00AB1234`)\n\n### Tracked Changes\n\n**Insertion:**\n```xml\n<w:ins w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:t>inserted text</w:t></w:r>\n</w:ins>\n```\n\n**Deletion:**\n```xml\n<w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n```\n\n**Inside `<w:del>`**: Use `<w:delText>` instead of `<w:t>`, and `<w:delInstrText>` instead of `<w:instrText>`.\n\n**Minimal edits** - only mark what changes:\n```xml\n<!-- Change \"30 days\" to \"60 days\" -->\n<w:r><w:t>The term is </w:t></w:r>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:delText>30</w:delText></w:r>\n</w:del>\n<w:ins w:id=\"2\" w:author=\"Claude\" w:date=\"...\">\n  <w:r><w:t>60</w:t></w:r>\n</w:ins>\n<w:r><w:t> days.</w:t></w:r>\n```\n\n**Deleting entire paragraphs/list items** - when removing ALL content from a paragraph, also mark the paragraph mark as deleted so it merges with the next paragraph. Add `<w:del/>` inside `<w:pPr><w:rPr>`:\n```xml\n<w:p>\n  <w:pPr>\n    <w:numPr>...</w:numPr>  <!-- list numbering if present -->\n    <w:rPr>\n      <w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\"/>\n    </w:rPr>\n  </w:pPr>\n  <w:del w:id=\"2\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n    <w:r><w:delText>Entire paragraph content being deleted...</w:delText></w:r>\n  </w:del>\n</w:p>\n```\nWithout the `<w:del/>` in `<w:pPr><w:rPr>`, accepting changes leaves an empty paragraph/list item.\n\n**Rejecting another author's insertion** - nest deletion inside their insertion:\n```xml\n<w:ins w:author=\"Jane\" w:id=\"5\">\n  <w:del w:author=\"Claude\" w:id=\"10\">\n    <w:r><w:delText>their inserted text</w:delText></w:r>\n  </w:del>\n</w:ins>\n```\n\n**Restoring another author's deletion** - add insertion after (don't modify their deletion):\n```xml\n<w:del w:author=\"Jane\" w:id=\"5\">\n  <w:r><w:delText>deleted text</w:delText></w:r>\n</w:del>\n<w:ins w:author=\"Claude\" w:id=\"10\">\n  <w:r><w:t>deleted text</w:t></w:r>\n</w:ins>\n```\n\n### Comments\n\nAfter running `comment.py` (see Step 2), add markers to document.xml. For replies, use `--parent` flag and nest markers inside the parent's.\n\n**CRITICAL: `<w:commentRangeStart>` and `<w:commentRangeEnd>` are siblings of `<w:r>`, never inside `<w:r>`.**\n\n```xml\n<!-- Comment markers are direct children of w:p, never inside w:r -->\n<w:commentRangeStart w:id=\"0\"/>\n<w:del w:id=\"1\" w:author=\"Claude\" w:date=\"2025-01-01T00:00:00Z\">\n  <w:r><w:delText>deleted</w:delText></w:r>\n</w:del>\n<w:r><w:t> more text</w:t></w:r>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n\n<!-- Comment 0 with reply 1 nested inside -->\n<w:commentRangeStart w:id=\"0\"/>\n  <w:commentRangeStart w:id=\"1\"/>\n  <w:r><w:t>text</w:t></w:r>\n  <w:commentRangeEnd w:id=\"1\"/>\n<w:commentRangeEnd w:id=\"0\"/>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"0\"/></w:r>\n<w:r><w:rPr><w:rStyle w:val=\"CommentReference\"/></w:rPr><w:commentReference w:id=\"1\"/></w:r>\n```\n\n### Images\n\n1. Add image file to `word/media/`\n2. Add relationship to `word/_rels/document.xml.rels`:\n```xml\n<Relationship Id=\"rId5\" Type=\".../image\" Target=\"media/image1.png\"/>\n```\n3. Add content type to `[Content_Types].xml`:\n```xml\n<Default Extension=\"png\" ContentType=\"image/png\"/>\n```\n4. Reference in document.xml:\n```xml\n<w:drawing>\n  <wp:inline>\n    <wp:extent cx=\"914400\" cy=\"914400\"/>  <!-- EMUs: 914400 = 1 inch -->\n    <a:graphic>\n      <a:graphicData uri=\".../picture\">\n        <pic:pic>\n          <pic:blipFill><a:blip r:embed=\"rId5\"/></pic:blipFill>\n        </pic:pic>\n      </a:graphicData>\n    </a:graphic>\n  </wp:inline>\n</w:drawing>\n```\n\n---\n\n## Dependencies\n\n- **pandoc**: Text extraction\n- **docx**: `npm install -g docx` (new documents)\n- **LibreOffice**: PDF conversion (auto-configured for sandboxed environments via `scripts/office/soffice.py`)\n- **Poppler**: `pdftoppm` for images\n\n<!-- File: ruoyi-admin/src/main/resources/skills/pdf/forms.md (Tokens: ~2201 | Category: anthropic-skill) -->\n**CRITICAL: You MUST complete these steps in order. Do not skip ahead to writing code.**\n\nIf you need to fill out a PDF form, first check to see if the PDF has fillable form fields. Run this script from this file's directory:\n `python scripts/check_fillable_fields <file.pdf>`, and depending on the result go to either the \"Fillable fields\" or \"Non-fillable fields\" and follow those instructions.\n\n# Fillable fields\nIf the PDF has fillable form fields:\n- Run this script from this file's directory: `python scripts/extract_form_field_info.py <input.pdf> <field_info.json>`. It will create a JSON file with a list of fields in this format:\n```\n[\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"rect\": ([left, bottom, right, top] bounding box in PDF coordinates, y=0 is the bottom of the page),\n    \"type\": (\"text\", \"checkbox\", \"radio_group\", or \"choice\"),\n  },\n  // Checkboxes have \"checked_value\" and \"unchecked_value\" properties:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"checkbox\",\n    \"checked_value\": (Set the field to this value to check the checkbox),\n    \"unchecked_value\": (Set the field to this value to uncheck the checkbox),\n  },\n  // Radio groups have a \"radio_options\" list with the possible choices.\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"radio_group\",\n    \"radio_options\": [\n      {\n        \"value\": (set the field to this value to select this radio option),\n        \"rect\": (bounding box for the radio button for this option)\n      },\n      // Other radio options\n    ]\n  },\n  // Multiple choice fields have a \"choice_options\" list with the possible choices:\n  {\n    \"field_id\": (unique ID for the field),\n    \"page\": (page number, 1-based),\n    \"type\": \"choice\",\n    \"choice_options\": [\n      {\n        \"value\": (set the field to this value to select this option),\n        \"text\": (display text of the option)\n      },\n      // Other choice options\n    ],\n  }\n]\n```\n- Convert the PDF to PNGs (one image for each page) with this script (run from this file's directory):\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThen analyze the images to determine the purpose of each form field (make sure to convert the bounding box PDF coordinates to image coordinates).\n- Create a `field_values.json` file in this format with the values to be entered for each field:\n```\n[\n  {\n    \"field_id\": \"last_name\", // Must match the field_id from `extract_form_field_info.py`\n    \"description\": \"The user's last name\",\n    \"page\": 1, // Must match the \"page\" value in field_info.json\n    \"value\": \"Simpson\"\n  },\n  {\n    \"field_id\": \"Checkbox12\",\n    \"description\": \"Checkbox to be checked if the user is 18 or over\",\n    \"page\": 1,\n    \"value\": \"/On\" // If this is a checkbox, use its \"checked_value\" value to check it. If it's a radio button group, use one of the \"value\" values in \"radio_options\".\n  },\n  // more fields\n]\n```\n- Run the `fill_fillable_fields.py` script from this file's directory to create a filled-in PDF:\n`python scripts/fill_fillable_fields.py <input pdf> <field_values.json> <output pdf>`\nThis script will verify that the field IDs and values you provide are valid; if it prints error messages, correct the appropriate fields and try again.\n\n# Non-fillable fields\nIf the PDF doesn't have fillable form fields, you'll need to visually determine where the data should be added and create text annotations. Follow the below steps *exactly*. You MUST perform all of these steps to ensure that the the form is accurately completed. Details for each step are below.\n- Convert the PDF to PNG images and determine field bounding boxes.\n- Create a JSON file with field information and validation images showing the bounding boxes.\n- Validate the the bounding boxes.\n- Use the bounding boxes to fill in the form.\n\n## Step 1: Visual Analysis (REQUIRED)\n- Convert the PDF to PNG images. Run this script from this file's directory:\n`python scripts/convert_pdf_to_images.py <file.pdf> <output_directory>`\nThe script will create a PNG image for each page in the PDF.\n- Carefully examine each PNG image and identify all form fields and areas where the user should enter data. For each form field where the user should enter text, determine bounding boxes for both the form field label, and the area where the user should enter text. The label and entry bounding boxes MUST NOT INTERSECT; the text entry box should only include the area where data should be entered. Usually this area will be immediately to the side, above, or below its label. Entry bounding boxes must be tall and wide enough to contain their text.\n\nThese are some examples of form structures that you might see:\n\n*Label inside box*\n```\n┌────────────────────────┐\n│ Name:                  │\n└────────────────────────┘\n```\nThe input area should be to the right of the \"Name\" label and extend to the edge of the box.\n\n*Label before line*\n```\nEmail: _______________________\n```\nThe input area should be above the line and include its entire width.\n\n*Label under line*\n```\n_________________________\nName\n```\nThe input area should be above the line and include the entire width of the line. This is common for signature and date fields.\n\n*Label above line*\n```\nPlease enter any special requests:\n________________________________________________\n```\nThe input area should extend from the bottom of the label to the line, and should include the entire width of the line.\n\n*Checkboxes*\n```\nAre you a US citizen? Yes □  No □\n```\nFor checkboxes:\n- Look for small square boxes (□) - these are the actual checkboxes to target. They may be to the left or right of their labels.\n- Distinguish between label text (\"Yes\", \"No\") and the clickable checkbox squares.\n- The entry bounding box should cover ONLY the small square, not the text label.\n\n### Step 2: Create fields.json and validation images (REQUIRED)\n- Create a file named `fields.json` with information for the form fields and bounding boxes in this format:\n```\n{\n  \"pages\": [\n    {\n      \"page_number\": 1,\n      \"image_width\": (first page image width in pixels),\n      \"image_height\": (first page image height in pixels),\n    },\n    {\n      \"page_number\": 2,\n      \"image_width\": (second page image width in pixels),\n      \"image_height\": (second page image height in pixels),\n    }\n    // additional pages\n  ],\n  \"form_fields\": [\n    // Example for a text field.\n    {\n      \"page_number\": 1,\n      \"description\": \"The user's last name should be entered here\",\n      // Bounding boxes are [left, top, right, bottom]. The bounding boxes for the label and text entry should not overlap.\n      \"field_label\": \"Last name\",\n      \"label_bounding_box\": [30, 125, 95, 142],\n      \"entry_bounding_box\": [100, 125, 280, 142],\n      \"entry_text\": {\n        \"text\": \"Johnson\", // This text will be added as an annotation at the entry_bounding_box location\n        \"font_size\": 14, // optional, defaults to 14\n        \"font_color\": \"000000\", // optional, RRGGBB format, defaults to 000000 (black)\n      }\n    },\n    // Example for a checkbox. TARGET THE SQUARE for the entry bounding box, NOT THE TEXT\n    {\n      \"page_number\": 2,\n      \"description\": \"Checkbox that should be checked if the user is over 18\",\n      \"entry_bounding_box\": [140, 525, 155, 540],  // Small box over checkbox square\n      \"field_label\": \"Yes\",\n      \"label_bounding_box\": [100, 525, 132, 540],  // Box containing \"Yes\" text\n      // Use \"X\" to check a checkbox.\n      \"entry_text\": {\n        \"text\": \"X\",\n      }\n    }\n    // additional form field entries\n  ]\n}\n```\n\nCreate validation images by running this script from this file's directory for each page:\n`python scripts/create_validation_image.py <page_number> <path_to_fields.json> <input_image_path> <output_image_path>\n\nThe validation images will have red rectangles where text should be entered, and blue rectangles covering label text.\n\n### Step 3: Validate Bounding Boxes (REQUIRED)\n#### Automated intersection check\n- Verify that none of bounding boxes intersect and that the entry bounding boxes are tall enough by checking the fields.json file with the `check_bounding_boxes.py` script (run from this file's directory):\n`python scripts/check_bounding_boxes.py <JSON file>`\n\nIf there are errors, reanalyze the relevant fields, adjust the bounding boxes, and iterate until there are no remaining errors. Remember: label (blue) bounding boxes should contain text labels, entry (red) boxes should not.\n\n#### Manual image inspection\n**CRITICAL: Do not proceed without visually inspecting validation images**\n- Red rectangles must ONLY cover input areas\n- Red rectangles MUST NOT contain any text\n- Blue rectangles should contain label text\n- For checkboxes:\n  - Red rectangle MUST be centered on the checkbox square\n  - Blue rectangle should cover the text label for the checkbox\n\n- If any rectangles look wrong, fix fields.json, regenerate the validation images, and verify again. Repeat this process until the bounding boxes are fully accurate.\n\n\n### Step 4: Add annotations to the PDF\nRun this script from this file's directory to create a filled-out PDF using the information in fields.json:\n`python scripts/fill_pdf_form_with_annotations.py <input_pdf_path> <path_to_fields.json> <output_pdf_path>\n\n<!-- File: ruoyi-admin/src/main/resources/skills/pdf/reference.md (Tokens: ~4070 | Category: anthropic-skill) -->\n# PDF Processing Advanced Reference\n\nThis document contains advanced PDF processing features, detailed examples, and additional libraries not covered in the main skill instructions.\n\n## pypdfium2 Library (Apache/BSD License)\n\n### Overview\npypdfium2 is a Python binding for PDFium (Chromium's PDF library). It's excellent for fast PDF rendering, image generation, and serves as a PyMuPDF replacement.\n\n### Render PDF to Images\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\n\n# Load PDF\npdf = pdfium.PdfDocument(\"document.pdf\")\n\n# Render page to image\npage = pdf[0]  # First page\nbitmap = page.render(\n    scale=2.0,  # Higher resolution\n    rotation=0  # No rotation\n)\n\n# Convert to PIL Image\nimg = bitmap.to_pil()\nimg.save(\"page_1.png\", \"PNG\")\n\n# Process multiple pages\nfor i, page in enumerate(pdf):\n    bitmap = page.render(scale=1.5)\n    img = bitmap.to_pil()\n    img.save(f\"page_{i+1}.jpg\", \"JPEG\", quality=90)\n```\n\n### Extract Text with pypdfium2\n```python\nimport pypdfium2 as pdfium\n\npdf = pdfium.PdfDocument(\"document.pdf\")\nfor i, page in enumerate(pdf):\n    text = page.get_text()\n    print(f\"Page {i+1} text length: {len(text)} chars\")\n```\n\n## JavaScript Libraries\n\n### pdf-lib (MIT License)\n\npdf-lib is a powerful JavaScript library for creating and modifying PDF documents in any JavaScript environment.\n\n#### Load and Manipulate Existing PDF\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function manipulatePDF() {\n    // Load existing PDF\n    const existingPdfBytes = fs.readFileSync('input.pdf');\n    const pdfDoc = await PDFDocument.load(existingPdfBytes);\n\n    // Get page count\n    const pageCount = pdfDoc.getPageCount();\n    console.log(`Document has ${pageCount} pages`);\n\n    // Add new page\n    const newPage = pdfDoc.addPage([600, 400]);\n    newPage.drawText('Added by pdf-lib', {\n        x: 100,\n        y: 300,\n        size: 16\n    });\n\n    // Save modified PDF\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('modified.pdf', pdfBytes);\n}\n```\n\n#### Create Complex PDFs from Scratch\n```javascript\nimport { PDFDocument, rgb, StandardFonts } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function createPDF() {\n    const pdfDoc = await PDFDocument.create();\n\n    // Add fonts\n    const helveticaFont = await pdfDoc.embedFont(StandardFonts.Helvetica);\n    const helveticaBold = await pdfDoc.embedFont(StandardFonts.HelveticaBold);\n\n    // Add page\n    const page = pdfDoc.addPage([595, 842]); // A4 size\n    const { width, height } = page.getSize();\n\n    // Add text with styling\n    page.drawText('Invoice #12345', {\n        x: 50,\n        y: height - 50,\n        size: 18,\n        font: helveticaBold,\n        color: rgb(0.2, 0.2, 0.8)\n    });\n\n    // Add rectangle (header background)\n    page.drawRectangle({\n        x: 40,\n        y: height - 100,\n        width: width - 80,\n        height: 30,\n        color: rgb(0.9, 0.9, 0.9)\n    });\n\n    // Add table-like content\n    const items = [\n        ['Item', 'Qty', 'Price', 'Total'],\n        ['Widget', '2', '$50', '$100'],\n        ['Gadget', '1', '$75', '$75']\n    ];\n\n    let yPos = height - 150;\n    items.forEach(row => {\n        let xPos = 50;\n        row.forEach(cell => {\n            page.drawText(cell, {\n                x: xPos,\n                y: yPos,\n                size: 12,\n                font: helveticaFont\n            });\n            xPos += 120;\n        });\n        yPos -= 25;\n    });\n\n    const pdfBytes = await pdfDoc.save();\n    fs.writeFileSync('created.pdf', pdfBytes);\n}\n```\n\n#### Advanced Merge and Split Operations\n```javascript\nimport { PDFDocument } from 'pdf-lib';\nimport fs from 'fs';\n\nasync function mergePDFs() {\n    // Create new document\n    const mergedPdf = await PDFDocument.create();\n\n    // Load source PDFs\n    const pdf1Bytes = fs.readFileSync('doc1.pdf');\n    const pdf2Bytes = fs.readFileSync('doc2.pdf');\n\n    const pdf1 = await PDFDocument.load(pdf1Bytes);\n    const pdf2 = await PDFDocument.load(pdf2Bytes);\n\n    // Copy pages from first PDF\n    const pdf1Pages = await mergedPdf.copyPages(pdf1, pdf1.getPageIndices());\n    pdf1Pages.forEach(page => mergedPdf.addPage(page));\n\n    // Copy specific pages from second PDF (pages 0, 2, 4)\n    const pdf2Pages = await mergedPdf.copyPages(pdf2, [0, 2, 4]);\n    pdf2Pages.forEach(page => mergedPdf.addPage(page));\n\n    const mergedPdfBytes = await mergedPdf.save();\n    fs.writeFileSync('merged.pdf', mergedPdfBytes);\n}\n```\n\n### pdfjs-dist (Apache License)\n\nPDF.js is Mozilla's JavaScript library for rendering PDFs in the browser.\n\n#### Basic PDF Loading and Rendering\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\n// Configure worker (important for performance)\npdfjsLib.GlobalWorkerOptions.workerSrc = './pdf.worker.js';\n\nasync function renderPDF() {\n    // Load PDF\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    console.log(`Loaded PDF with ${pdf.numPages} pages`);\n\n    // Get first page\n    const page = await pdf.getPage(1);\n    const viewport = page.getViewport({ scale: 1.5 });\n\n    // Render to canvas\n    const canvas = document.createElement('canvas');\n    const context = canvas.getContext('2d');\n    canvas.height = viewport.height;\n    canvas.width = viewport.width;\n\n    const renderContext = {\n        canvasContext: context,\n        viewport: viewport\n    };\n\n    await page.render(renderContext).promise;\n    document.body.appendChild(canvas);\n}\n```\n\n#### Extract Text with Coordinates\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractText() {\n    const loadingTask = pdfjsLib.getDocument('document.pdf');\n    const pdf = await loadingTask.promise;\n\n    let fullText = '';\n\n    // Extract text from all pages\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const textContent = await page.getTextContent();\n\n        const pageText = textContent.items\n            .map(item => item.str)\n            .join(' ');\n\n        fullText += `\\n--- Page ${i} ---\\n${pageText}`;\n\n        // Get text with coordinates for advanced processing\n        const textWithCoords = textContent.items.map(item => ({\n            text: item.str,\n            x: item.transform[4],\n            y: item.transform[5],\n            width: item.width,\n            height: item.height\n        }));\n    }\n\n    console.log(fullText);\n    return fullText;\n}\n```\n\n#### Extract Annotations and Forms\n```javascript\nimport * as pdfjsLib from 'pdfjs-dist';\n\nasync function extractAnnotations() {\n    const loadingTask = pdfjsLib.getDocument('annotated.pdf');\n    const pdf = await loadingTask.promise;\n\n    for (let i = 1; i <= pdf.numPages; i++) {\n        const page = await pdf.getPage(i);\n        const annotations = await page.getAnnotations();\n\n        annotations.forEach(annotation => {\n            console.log(`Annotation type: ${annotation.subtype}`);\n            console.log(`Content: ${annotation.contents}`);\n            console.log(`Coordinates: ${JSON.stringify(annotation.rect)}`);\n        });\n    }\n}\n```\n\n## Advanced Command-Line Operations\n\n### poppler-utils Advanced Features\n\n#### Extract Text with Bounding Box Coordinates\n```bash\n# Extract text with bounding box coordinates (essential for structured data)\npdftotext -bbox-layout document.pdf output.xml\n\n# The XML output contains precise coordinates for each text element\n```\n\n#### Advanced Image Conversion\n```bash\n# Convert to PNG images with specific resolution\npdftoppm -png -r 300 document.pdf output_prefix\n\n# Convert specific page range with high resolution\npdftoppm -png -r 600 -f 1 -l 3 document.pdf high_res_pages\n\n# Convert to JPEG with quality setting\npdftoppm -jpeg -jpegopt quality=85 -r 200 document.pdf jpeg_output\n```\n\n#### Extract Embedded Images\n```bash\n# Extract all embedded images with metadata\npdfimages -j -p document.pdf page_images\n\n# List image info without extracting\npdfimages -list document.pdf\n\n# Extract images in their original format\npdfimages -all document.pdf images/img\n```\n\n### qpdf Advanced Features\n\n#### Complex Page Manipulation\n```bash\n# Split PDF into groups of pages\nqpdf --split-pages=3 input.pdf output_group_%02d.pdf\n\n# Extract specific pages with complex ranges\nqpdf input.pdf --pages input.pdf 1,3-5,8,10-end -- extracted.pdf\n\n# Merge specific pages from multiple PDFs\nqpdf --empty --pages doc1.pdf 1-3 doc2.pdf 5-7 doc3.pdf 2,4 -- combined.pdf\n```\n\n#### PDF Optimization and Repair\n```bash\n# Optimize PDF for web (linearize for streaming)\nqpdf --linearize input.pdf optimized.pdf\n\n# Remove unused objects and compress\nqpdf --optimize-level=all input.pdf compressed.pdf\n\n# Attempt to repair corrupted PDF structure\nqpdf --check input.pdf\nqpdf --fix-qdf damaged.pdf repaired.pdf\n\n# Show detailed PDF structure for debugging\nqpdf --show-all-pages input.pdf > structure.txt\n```\n\n#### Advanced Encryption\n```bash\n# Add password protection with specific permissions\nqpdf --encrypt user_pass owner_pass 256 --print=none --modify=none -- input.pdf encrypted.pdf\n\n# Check encryption status\nqpdf --show-encryption encrypted.pdf\n\n# Remove password protection (requires password)\nqpdf --password=secret123 --decrypt encrypted.pdf decrypted.pdf\n```\n\n## Advanced Python Techniques\n\n### pdfplumber Advanced Features\n\n#### Extract Text with Precise Coordinates\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract all text with coordinates\n    chars = page.chars\n    for char in chars[:10]:  # First 10 characters\n        print(f\"Char: '{char['text']}' at x:{char['x0']:.1f} y:{char['y0']:.1f}\")\n    \n    # Extract text by bounding box (left, top, right, bottom)\n    bbox_text = page.within_bbox((100, 100, 400, 200)).extract_text()\n```\n\n#### Advanced Table Extraction with Custom Settings\n```python\nimport pdfplumber\nimport pandas as pd\n\nwith pdfplumber.open(\"complex_table.pdf\") as pdf:\n    page = pdf.pages[0]\n    \n    # Extract tables with custom settings for complex layouts\n    table_settings = {\n        \"vertical_strategy\": \"lines\",\n        \"horizontal_strategy\": \"lines\",\n        \"snap_tolerance\": 3,\n        \"intersection_tolerance\": 15\n    }\n    tables = page.extract_tables(table_settings)\n    \n    # Visual debugging for table extraction\n    img = page.to_image(resolution=150)\n    img.save(\"debug_layout.png\")\n```\n\n### reportlab Advanced Features\n\n#### Create Professional Reports with Tables\n```python\nfrom reportlab.platypus import SimpleDocTemplate, Table, TableStyle, Paragraph\nfrom reportlab.lib.styles import getSampleStyleSheet\nfrom reportlab.lib import colors\n\n# Sample data\ndata = [\n    ['Product', 'Q1', 'Q2', 'Q3', 'Q4'],\n    ['Widgets', '120', '135', '142', '158'],\n    ['Gadgets', '85', '92', '98', '105']\n]\n\n# Create PDF with table\ndoc = SimpleDocTemplate(\"report.pdf\")\nelements = []\n\n# Add title\nstyles = getSampleStyleSheet()\ntitle = Paragraph(\"Quarterly Sales Report\", styles['Title'])\nelements.append(title)\n\n# Add table with advanced styling\ntable = Table(data)\ntable.setStyle(TableStyle([\n    ('BACKGROUND', (0, 0), (-1, 0), colors.grey),\n    ('TEXTCOLOR', (0, 0), (-1, 0), colors.whitesmoke),\n    ('ALIGN', (0, 0), (-1, -1), 'CENTER'),\n    ('FONTNAME', (0, 0), (-1, 0), 'Helvetica-Bold'),\n    ('FONTSIZE', (0, 0), (-1, 0), 14),\n    ('BOTTOMPADDING', (0, 0), (-1, 0), 12),\n    ('BACKGROUND', (0, 1), (-1, -1), colors.beige),\n    ('GRID', (0, 0), (-1, -1), 1, colors.black)\n]))\nelements.append(table)\n\ndoc.build(elements)\n```\n\n## Complex Workflows\n\n### Extract Figures/Images from PDF\n\n#### Method 1: Using pdfimages (fastest)\n```bash\n# Extract all images with original quality\npdfimages -all document.pdf images/img\n```\n\n#### Method 2: Using pypdfium2 + Image Processing\n```python\nimport pypdfium2 as pdfium\nfrom PIL import Image\nimport numpy as np\n\ndef extract_figures(pdf_path, output_dir):\n    pdf = pdfium.PdfDocument(pdf_path)\n    \n    for page_num, page in enumerate(pdf):\n        # Render high-resolution page\n        bitmap = page.render(scale=3.0)\n        img = bitmap.to_pil()\n        \n        # Convert to numpy for processing\n        img_array = np.array(img)\n        \n        # Simple figure detection (non-white regions)\n        mask = np.any(img_array != [255, 255, 255], axis=2)\n        \n        # Find contours and extract bounding boxes\n        # (This is simplified - real implementation would need more sophisticated detection)\n        \n        # Save detected figures\n        # ... implementation depends on specific needs\n```\n\n### Batch PDF Processing with Error Handling\n```python\nimport os\nimport glob\nfrom pypdf import PdfReader, PdfWriter\nimport logging\n\nlogging.basicConfig(level=logging.INFO)\nlogger = logging.getLogger(__name__)\n\ndef batch_process_pdfs(input_dir, operation='merge'):\n    pdf_files = glob.glob(os.path.join(input_dir, \"*.pdf\"))\n    \n    if operation == 'merge':\n        writer = PdfWriter()\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                for page in reader.pages:\n                    writer.add_page(page)\n                logger.info(f\"Processed: {pdf_file}\")\n            except Exception as e:\n                logger.error(f\"Failed to process {pdf_file}: {e}\")\n                continue\n        \n        with open(\"batch_merged.pdf\", \"wb\") as output:\n            writer.write(output)\n    \n    elif operation == 'extract_text':\n        for pdf_file in pdf_files:\n            try:\n                reader = PdfReader(pdf_file)\n                text = \"\"\n                for page in reader.pages:\n                    text += page.extract_text()\n                \n                output_file = pdf_file.replace('.pdf', '.txt')\n                with open(output_file, 'w', encoding='utf-8') as f:\n                    f.write(text)\n                logger.info(f\"Extracted text from: {pdf_file}\")\n                \n            except Exception as e:\n                logger.error(f\"Failed to extract text from {pdf_file}: {e}\")\n                continue\n```\n\n### Advanced PDF Cropping\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\n# Crop page (left, bottom, right, top in points)\npage = reader.pages[0]\npage.mediabox.left = 50\npage.mediabox.bottom = 50\npage.mediabox.right = 550\npage.mediabox.top = 750\n\nwriter.add_page(page)\nwith open(\"cropped.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Performance Optimization Tips\n\n### 1. For Large PDFs\n- Use streaming approaches instead of loading entire PDF in memory\n- Use `qpdf --split-pages` for splitting large files\n- Process pages individually with pypdfium2\n\n### 2. For Text Extraction\n- `pdftotext -bbox-layout` is fastest for plain text extraction\n- Use pdfplumber for structured data and tables\n- Avoid `pypdf.extract_text()` for very large documents\n\n### 3. For Image Extraction\n- `pdfimages` is much faster than rendering pages\n- Use low resolution for previews, high resolution for final output\n\n### 4. For Form Filling\n- pdf-lib maintains form structure better than most alternatives\n- Pre-validate form fields before processing\n\n### 5. Memory Management\n```python\n# Process PDFs in chunks\ndef process_large_pdf(pdf_path, chunk_size=10):\n    reader = PdfReader(pdf_path)\n    total_pages = len(reader.pages)\n    \n    for start_idx in range(0, total_pages, chunk_size):\n        end_idx = min(start_idx + chunk_size, total_pages)\n        writer = PdfWriter()\n        \n        for i in range(start_idx, end_idx):\n            writer.add_page(reader.pages[i])\n        \n        # Process chunk\n        with open(f\"chunk_{start_idx//chunk_size}.pdf\", \"wb\") as output:\n            writer.write(output)\n```\n\n## Troubleshooting Common Issues\n\n### Encrypted PDFs\n```python\n# Handle password-protected PDFs\nfrom pypdf import PdfReader\n\ntry:\n    reader = PdfReader(\"encrypted.pdf\")\n    if reader.is_encrypted:\n        reader.decrypt(\"password\")\nexcept Exception as e:\n    print(f\"Failed to decrypt: {e}\")\n```\n\n### Corrupted PDFs\n```bash\n# Use qpdf to repair\nqpdf --check corrupted.pdf\nqpdf --replace-input corrupted.pdf\n```\n\n### Text Extraction Issues\n```python\n# Fallback to OCR for scanned PDFs\nimport pytesseract\nfrom pdf2image import convert_from_path\n\ndef extract_text_with_ocr(pdf_path):\n    images = convert_from_path(pdf_path)\n    text = \"\"\n    for i, image in enumerate(images):\n        text += pytesseract.image_to_string(image)\n    return text\n```\n\n## License Information\n\n- **pypdf**: BSD License\n- **pdfplumber**: MIT License\n- **pypdfium2**: Apache/BSD License\n- **reportlab**: BSD License\n- **poppler-utils**: GPL-2 License\n- **qpdf**: Apache License\n- **pdf-lib**: MIT License\n- **pdfjs-dist**: Apache License\n\n<!-- File: ruoyi-admin/src/main/resources/skills/pdf/SKILL.md (Tokens: ~1804 | Category: anthropic-skill) -->\n---\nname: pdf\ndescription: Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# PDF Processing Guide\n\n## Overview\n\nThis guide covers essential PDF processing operations using Python libraries and command-line tools. For advanced features, JavaScript libraries, and detailed examples, see reference.md. If you need to fill out a PDF form, read forms.md and follow its instructions.\n\n## Quick Start\n\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Read a PDF\nreader = PdfReader(\"document.pdf\")\nprint(f\"Pages: {len(reader.pages)}\")\n\n# Extract text\ntext = \"\"\nfor page in reader.pages:\n    text += page.extract_text()\n```\n\n## Python Libraries\n\n### pypdf - Basic Operations\n\n#### Merge PDFs\n```python\nfrom pypdf import PdfWriter, PdfReader\n\nwriter = PdfWriter()\nfor pdf_file in [\"doc1.pdf\", \"doc2.pdf\", \"doc3.pdf\"]:\n    reader = PdfReader(pdf_file)\n    for page in reader.pages:\n        writer.add_page(page)\n\nwith open(\"merged.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n#### Split PDF\n```python\nreader = PdfReader(\"input.pdf\")\nfor i, page in enumerate(reader.pages):\n    writer = PdfWriter()\n    writer.add_page(page)\n    with open(f\"page_{i+1}.pdf\", \"wb\") as output:\n        writer.write(output)\n```\n\n#### Extract Metadata\n```python\nreader = PdfReader(\"document.pdf\")\nmeta = reader.metadata\nprint(f\"Title: {meta.title}\")\nprint(f\"Author: {meta.author}\")\nprint(f\"Subject: {meta.subject}\")\nprint(f\"Creator: {meta.creator}\")\n```\n\n#### Rotate Pages\n```python\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\npage = reader.pages[0]\npage.rotate(90)  # Rotate 90 degrees clockwise\nwriter.add_page(page)\n\nwith open(\"rotated.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### pdfplumber - Text and Table Extraction\n\n#### Extract Text with Layout\n```python\nimport pdfplumber\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for page in pdf.pages:\n        text = page.extract_text()\n        print(text)\n```\n\n#### Extract Tables\n```python\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    for i, page in enumerate(pdf.pages):\n        tables = page.extract_tables()\n        for j, table in enumerate(tables):\n            print(f\"Table {j+1} on page {i+1}:\")\n            for row in table:\n                print(row)\n```\n\n#### Advanced Table Extraction\n```python\nimport pandas as pd\n\nwith pdfplumber.open(\"document.pdf\") as pdf:\n    all_tables = []\n    for page in pdf.pages:\n        tables = page.extract_tables()\n        for table in tables:\n            if table:  # Check if table is not empty\n                df = pd.DataFrame(table[1:], columns=table[0])\n                all_tables.append(df)\n\n# Combine all tables\nif all_tables:\n    combined_df = pd.concat(all_tables, ignore_index=True)\n    combined_df.to_excel(\"extracted_tables.xlsx\", index=False)\n```\n\n### reportlab - Create PDFs\n\n#### Basic PDF Creation\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.pdfgen import canvas\n\nc = canvas.Canvas(\"hello.pdf\", pagesize=letter)\nwidth, height = letter\n\n# Add text\nc.drawString(100, height - 100, \"Hello World!\")\nc.drawString(100, height - 120, \"This is a PDF created with reportlab\")\n\n# Add a line\nc.line(100, height - 140, 400, height - 140)\n\n# Save\nc.save()\n```\n\n#### Create PDF with Multiple Pages\n```python\nfrom reportlab.lib.pagesizes import letter\nfrom reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, PageBreak\nfrom reportlab.lib.styles import getSampleStyleSheet\n\ndoc = SimpleDocTemplate(\"report.pdf\", pagesize=letter)\nstyles = getSampleStyleSheet()\nstory = []\n\n# Add content\ntitle = Paragraph(\"Report Title\", styles['Title'])\nstory.append(title)\nstory.append(Spacer(1, 12))\n\nbody = Paragraph(\"This is the body of the report. \" * 20, styles['Normal'])\nstory.append(body)\nstory.append(PageBreak())\n\n# Page 2\nstory.append(Paragraph(\"Page 2\", styles['Heading1']))\nstory.append(Paragraph(\"Content for page 2\", styles['Normal']))\n\n# Build PDF\ndoc.build(story)\n```\n\n## Command-Line Tools\n\n### pdftotext (poppler-utils)\n```bash\n# Extract text\npdftotext input.pdf output.txt\n\n# Extract text preserving layout\npdftotext -layout input.pdf output.txt\n\n# Extract specific pages\npdftotext -f 1 -l 5 input.pdf output.txt  # Pages 1-5\n```\n\n### qpdf\n```bash\n# Merge PDFs\nqpdf --empty --pages file1.pdf file2.pdf -- merged.pdf\n\n# Split pages\nqpdf input.pdf --pages . 1-5 -- pages1-5.pdf\nqpdf input.pdf --pages . 6-10 -- pages6-10.pdf\n\n# Rotate pages\nqpdf input.pdf output.pdf --rotate=+90:1  # Rotate page 1 by 90 degrees\n\n# Remove password\nqpdf --password=mypassword --decrypt encrypted.pdf decrypted.pdf\n```\n\n### pdftk (if available)\n```bash\n# Merge\npdftk file1.pdf file2.pdf cat output merged.pdf\n\n# Split\npdftk input.pdf burst\n\n# Rotate\npdftk input.pdf rotate 1east output rotated.pdf\n```\n\n## Common Tasks\n\n### Extract Text from Scanned PDFs\n```python\n# Requires: pip install pytesseract pdf2image\nimport pytesseract\nfrom pdf2image import convert_from_path\n\n# Convert PDF to images\nimages = convert_from_path('scanned.pdf')\n\n# OCR each page\ntext = \"\"\nfor i, image in enumerate(images):\n    text += f\"Page {i+1}:\\n\"\n    text += pytesseract.image_to_string(image)\n    text += \"\\n\\n\"\n\nprint(text)\n```\n\n### Add Watermark\n```python\nfrom pypdf import PdfReader, PdfWriter\n\n# Create watermark (or load existing)\nwatermark = PdfReader(\"watermark.pdf\").pages[0]\n\n# Apply to all pages\nreader = PdfReader(\"document.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    page.merge_page(watermark)\n    writer.add_page(page)\n\nwith open(\"watermarked.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n### Extract Images\n```bash\n# Using pdfimages (poppler-utils)\npdfimages -j input.pdf output_prefix\n\n# This extracts all images as output_prefix-000.jpg, output_prefix-001.jpg, etc.\n```\n\n### Password Protection\n```python\nfrom pypdf import PdfReader, PdfWriter\n\nreader = PdfReader(\"input.pdf\")\nwriter = PdfWriter()\n\nfor page in reader.pages:\n    writer.add_page(page)\n\n# Add password\nwriter.encrypt(\"userpassword\", \"ownerpassword\")\n\nwith open(\"encrypted.pdf\", \"wb\") as output:\n    writer.write(output)\n```\n\n## Quick Reference\n\n| Task | Best Tool | Command/Code |\n|------|-----------|--------------|\n| Merge PDFs | pypdf | `writer.add_page(page)` |\n| Split PDFs | pypdf | One page per file |\n| Extract text | pdfplumber | `page.extract_text()` |\n| Extract tables | pdfplumber | `page.extract_tables()` |\n| Create PDFs | reportlab | Canvas or Platypus |\n| Command line merge | qpdf | `qpdf --empty --pages ...` |\n| OCR scanned PDFs | pytesseract | Convert to image first |\n| Fill PDF forms | pdf-lib or pypdf (see forms.md) | See forms.md |\n\n## Next Steps\n\n- For advanced pypdfium2 usage, see reference.md\n- For JavaScript libraries (pdf-lib), see reference.md\n- If you need to fill out a PDF form, follow the instructions in forms.md\n- For troubleshooting guides, see reference.md\n\n<!-- File: ruoyi-admin/src/main/resources/skills/xlsx/SKILL.md (Tokens: ~3267 | Category: anthropic-skill) -->\n---\nname: xlsx\ndescription: \"Comprehensive spreadsheet creation, editing, and analysis with support for formulas, formatting, data analysis, and visualization. When Claude needs to work with spreadsheets (.xlsx, .xlsm, .csv, .tsv, etc) for: (1) Creating new spreadsheets with formulas and formatting, (2) Reading or analyzing data, (3) Modify existing spreadsheets while preserving formulas, (4) Data analysis and visualization in spreadsheets, or (5) Recalculating formulas\"\nlicense: Proprietary. LICENSE.txt has complete terms\n---\n\n# Requirements for Outputs\n\n## All Excel files\n\n### Zero Formula Errors\n- Every Excel model MUST be delivered with ZERO formula errors (#REF!, #DIV/0!, #VALUE!, #N/A, #NAME?)\n\n### Preserve Existing Templates (when updating templates)\n- Study and EXACTLY match existing format, style, and conventions when modifying files\n- Never impose standardized formatting on files with established patterns\n- Existing template conventions ALWAYS override these guidelines\n\n## Financial models\n\n### Color Coding Standards\nUnless otherwise stated by the user or existing template\n\n#### Industry-Standard Color Conventions\n- **Blue text (RGB: 0,0,255)**: Hardcoded inputs, and numbers users will change for scenarios\n- **Black text (RGB: 0,0,0)**: ALL formulas and calculations\n- **Green text (RGB: 0,128,0)**: Links pulling from other worksheets within same workbook\n- **Red text (RGB: 255,0,0)**: External links to other files\n- **Yellow background (RGB: 255,255,0)**: Key assumptions needing attention or cells that need to be updated\n\n### Number Formatting Standards\n\n#### Required Format Rules\n- **Years**: Format as text strings (e.g., \"2024\" not \"2,024\")\n- **Currency**: Use $#,##0 format; ALWAYS specify units in headers (\"Revenue ($mm)\")\n- **Zeros**: Use number formatting to make all zeros \"-\", including percentages (e.g., \"$#,##0;($#,##0);-\")\n- **Percentages**: Default to 0.0% format (one decimal)\n- **Multiples**: Format as 0.0x for valuation multiples (EV/EBITDA, P/E)\n- **Negative numbers**: Use parentheses (123) not minus -123\n\n### Formula Construction Rules\n\n#### Assumptions Placement\n- Place ALL assumptions (growth rates, margins, multiples, etc.) in separate assumption cells\n- Use cell references instead of hardcoded values in formulas\n- Example: Use =B5*(1+$B$6) instead of =B5*1.05\n\n#### Formula Error Prevention\n- Verify all cell references are correct\n- Check for off-by-one errors in ranges\n- Ensure consistent formulas across all projection periods\n- Test with edge cases (zero values, negative numbers)\n- Verify no unintended circular references\n\n#### Documentation Requirements for Hardcodes\n- Comment or in cells beside (if end of table). Format: \"Source: [System/Document], [Date], [Specific Reference], [URL if applicable]\"\n- Examples:\n  - \"Source: Company 10-K, FY2024, Page 45, Revenue Note, [SEC EDGAR URL]\"\n  - \"Source: Company 10-Q, Q2 2025, Exhibit 99.1, [SEC EDGAR URL]\"\n  - \"Source: Bloomberg Terminal, 8/15/2025, AAPL US Equity\"\n  - \"Source: FactSet, 8/20/2025, Consensus Estimates Screen\"\n\n# XLSX creation, editing, and analysis\n\n## Overview\n\nA user may ask you to create, edit, or analyze the contents of an .xlsx file. You have different tools and workflows available for different tasks.\n\n## Important Requirements\n\n**LibreOffice Required for Formula Recalculation**: You can assume LibreOffice is installed for recalculating formula values using the `recalc.py` script. The script automatically configures LibreOffice on first run\n\n## Reading and analyzing data\n\n### Data analysis with pandas\nFor data analysis, visualization, and basic operations, use **pandas** which provides powerful data manipulation capabilities:\n\n```python\nimport pandas as pd\n\n# Read Excel\ndf = pd.read_excel('file.xlsx')  # Default: first sheet\nall_sheets = pd.read_excel('file.xlsx', sheet_name=None)  # All sheets as dict\n\n# Analyze\ndf.head()      # Preview data\ndf.info()      # Column info\ndf.describe()  # Statistics\n\n# Write Excel\ndf.to_excel('output.xlsx', index=False)\n```\n\n## Excel File Workflows\n\n## CRITICAL: Use Formulas, Not Hardcoded Values\n\n**Always use Excel formulas instead of calculating values in Python and hardcoding them.** This ensures the spreadsheet remains dynamic and updateable.\n\n### ❌ WRONG - Hardcoding Calculated Values\n```python\n# Bad: Calculating in Python and hardcoding result\ntotal = df['Sales'].sum()\nsheet['B10'] = total  # Hardcodes 5000\n\n# Bad: Computing growth rate in Python\ngrowth = (df.iloc[-1]['Revenue'] - df.iloc[0]['Revenue']) / df.iloc[0]['Revenue']\nsheet['C5'] = growth  # Hardcodes 0.15\n\n# Bad: Python calculation for average\navg = sum(values) / len(values)\nsheet['D20'] = avg  # Hardcodes 42.5\n```\n\n### ✅ CORRECT - Using Excel Formulas\n```python\n# Good: Let Excel calculate the sum\nsheet['B10'] = '=SUM(B2:B9)'\n\n# Good: Growth rate as Excel formula\nsheet['C5'] = '=(C4-C2)/C2'\n\n# Good: Average using Excel function\nsheet['D20'] = '=AVERAGE(D2:D19)'\n```\n\nThis applies to ALL calculations - totals, percentages, ratios, differences, etc. The spreadsheet should be able to recalculate when source data changes.\n\n## Common Workflow\n1. **Choose tool**: pandas for data, openpyxl for formulas/formatting\n2. **Create/Load**: Create new workbook or load existing file\n3. **Modify**: Add/edit data, formulas, and formatting\n4. **Save**: Write to file\n5. **Recalculate formulas (MANDATORY IF USING FORMULAS)**: Use the recalc.py script\n   ```bash\n   python recalc.py output.xlsx\n   ```\n6. **Verify and fix any errors**: \n   - The script returns JSON with error details\n   - If `status` is `errors_found`, check `error_summary` for specific error types and locations\n   - Fix the identified errors and recalculate again\n   - Common errors to fix:\n     - `#REF!`: Invalid cell references\n     - `#DIV/0!`: Division by zero\n     - `#VALUE!`: Wrong data type in formula\n     - `#NAME?`: Unrecognized formula name\n\n### Creating new Excel files\n\n```python\n# Using openpyxl for formulas and formatting\nfrom openpyxl import Workbook\nfrom openpyxl.styles import Font, PatternFill, Alignment\n\nwb = Workbook()\nsheet = wb.active\n\n# Add data\nsheet['A1'] = 'Hello'\nsheet['B1'] = 'World'\nsheet.append(['Row', 'of', 'data'])\n\n# Add formula\nsheet['B2'] = '=SUM(A1:A10)'\n\n# Formatting\nsheet['A1'].font = Font(bold=True, color='FF0000')\nsheet['A1'].fill = PatternFill('solid', start_color='FFFF00')\nsheet['A1'].alignment = Alignment(horizontal='center')\n\n# Column width\nsheet.column_dimensions['A'].width = 20\n\nwb.save('output.xlsx')\n```\n\n### Editing existing Excel files\n\n```python\n# Using openpyxl to preserve formulas and formatting\nfrom openpyxl import load_workbook\n\n# Load existing file\nwb = load_workbook('existing.xlsx')\nsheet = wb.active  # or wb['SheetName'] for specific sheet\n\n# Working with multiple sheets\nfor sheet_name in wb.sheetnames:\n    sheet = wb[sheet_name]\n    print(f\"Sheet: {sheet_name}\")\n\n# Modify cells\nsheet['A1'] = 'New Value'\nsheet.insert_rows(2)  # Insert row at position 2\nsheet.delete_cols(3)  # Delete column 3\n\n# Add new sheet\nnew_sheet = wb.create_sheet('NewSheet')\nnew_sheet['A1'] = 'Data'\n\nwb.save('modified.xlsx')\n```\n\n## Recalculating formulas\n\nExcel files created or modified by openpyxl contain formulas as strings but not calculated values. Use the provided `recalc.py` script to recalculate formulas:\n\n```bash\n# Recalculate existing Excel file\npython recalc.py <excel_file> [timeout_seconds]\n\n# Create new Excel workbook\npython recalc.py --create <output_file> '<json_data>'\n```\n\nExamples:\n```bash\npython recalc.py output.xlsx 30\n\npython recalc.py --create output.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Value\"], [\"Test\", 100]]}}'\n```\n\nThe script:\n- Automatically sets up LibreOffice macro on first run\n- Recalculates all formulas in all sheets\n- Scans ALL cells for Excel errors (#REF!, #DIV/0!, etc.)\n- Returns JSON with detailed error locations and counts\n- Works on Windows, Linux, and macOS\n- Can create new workbooks from JSON configuration\n\n## Creating workbooks with recalc.py\n\nThe `--create` mode allows you to create new Excel workbooks directly from JSON configuration:\n\n```bash\npython recalc.py --create <output_file> '<json_data>'\n```\n\n**Default Behavior**: If you provide a relative path (e.g., `test.xlsx`), the file will be created in the **project root** directory (`ruoyi-ai-v3/`). To save to a specific directory, use an absolute path.\n\n### JSON Format\n\n```json\n{\n  \"Sheet1\": {\n    \"data\": [\n      [\"Header1\", \"Header2\", \"Header3\"],\n      [\"Value1\", \"Value2\", \"=A2*B2\"],\n      [\"Value3\", \"Value4\", \"=SUM(C2:C10)\"]\n    ]\n  },\n  \"Sheet2\": {\n    \"data\": [\n      [\"Name\", \"Score\"],\n      [\"Alice\", 95],\n      [\"Bob\", 87]\n    ]\n  }\n}\n```\n\n### Examples\n\n**Simple data table (saves to project root)**:\n```bash\npython recalc.py --create data.xlsx '{\"Sheet1\": {\"data\": [[\"Name\", \"Age\"], [\"John\", 30], [\"Jane\", 28]]}}'\n# Creates: ruoyi-ai-v3/data.xlsx\n```\n\n**With formulas (saves to project root)**:\n```bash\npython recalc.py --create model.xlsx '{\"Sheet1\": {\"data\": [[\"A\", \"B\", \"Sum\"], [10, 20, \"=A2+B2\"], [5, 15, \"=A3+B3\"]]}}'\n# Creates: ruoyi-ai-v3/model.xlsx\n```\n\n**Save to custom directory (use absolute path)**:\n```bash\npython recalc.py --create \"D:\\Downloads\\custom.xlsx\" '{\"Sheet1\": {\"data\": [[\"A\", \"B\"], [1, 2]]}}'\n# Creates: D:\\Downloads\\custom.xlsx\n```\n\n### File Path Options\n\n| Input | Output Location |\n|-------|-----------------|\n| `test.xlsx` | `ruoyi-ai-v3/test.xlsx` (project root) |\n| `output/test.xlsx` | `ruoyi-ai-v3/output/test.xlsx` |\n| `D:\\Downloads\\test.xlsx` | `D:\\Downloads\\test.xlsx` (absolute path) |\n\n### Features\n\n- First row is automatically formatted as bold header\n- Supports both values and Excel formulas (strings starting with `=`)\n- Multiple sheets can be created in one command\n- Output file extension determines format (.xlsx, .xlsm, etc.)\n\n## Formula Verification Checklist\n\nQuick checks to ensure formulas work correctly:\n\n### Essential Verification\n- [ ] **Test 2-3 sample references**: Verify they pull correct values before building full model\n- [ ] **Column mapping**: Confirm Excel columns match (e.g., column 64 = BL, not BK)\n- [ ] **Row offset**: Remember Excel rows are 1-indexed (DataFrame row 5 = Excel row 6)\n\n### Common Pitfalls\n- [ ] **NaN handling**: Check for null values with `pd.notna()`\n- [ ] **Far-right columns**: FY data often in columns 50+ \n- [ ] **Multiple matches**: Search all occurrences, not just first\n- [ ] **Division by zero**: Check denominators before using `/` in formulas (#DIV/0!)\n- [ ] **Wrong references**: Verify all cell references point to intended cells (#REF!)\n- [ ] **Cross-sheet references**: Use correct format (Sheet1!A1) for linking sheets\n\n### Formula Testing Strategy\n- [ ] **Start small**: Test formulas on 2-3 cells before applying broadly\n- [ ] **Verify dependencies**: Check all cells referenced in formulas exist\n- [ ] **Test edge cases**: Include zero, negative, and very large values\n\n### Interpreting recalc.py Output\nThe script returns JSON with error details:\n```json\n{\n  \"status\": \"success\",           // or \"errors_found\"\n  \"total_errors\": 0,              // Total error count\n  \"total_formulas\": 42,           // Number of formulas in file\n  \"error_summary\": {              // Only present if errors found\n    \"#REF!\": {\n      \"count\": 2,\n      \"locations\": [\"Sheet1!B5\", \"Sheet1!C10\"]\n    }\n  }\n}\n```\n\n## Best Practices\n\n### Library Selection\n- **pandas**: Best for data analysis, bulk operations, and simple data export\n- **openpyxl**: Best for complex formatting, formulas, and Excel-specific features\n\n### Working with openpyxl\n- Cell indices are 1-based (row=1, column=1 refers to cell A1)\n- Use `data_only=True` to read calculated values: `load_workbook('file.xlsx', data_only=True)`\n- **Warning**: If opened with `data_only=True` and saved, formulas are replaced with values and permanently lost\n- For large files: Use `read_only=True` for reading or `write_only=True` for writing\n- Formulas are preserved but not evaluated - use recalc.py to update values\n\n### Working with pandas\n- Specify data types to avoid inference issues: `pd.read_excel('file.xlsx', dtype={'id': str})`\n- For large files, read specific columns: `pd.read_excel('file.xlsx', usecols=['A', 'C', 'E'])`\n- Handle dates properly: `pd.read_excel('file.xlsx', parse_dates=['date_column'])`\n\n## Code Style Guidelines\n**IMPORTANT**: When generating Python code for Excel operations:\n- Write minimal, concise Python code without unnecessary comments\n- Avoid verbose variable names and redundant operations\n- Avoid unnecessary print statements\n\n**For Excel files themselves**:\n- Add comments to cells with complex formulas or important assumptions\n- Document data sources for hardcoded values\n- Include notes for key calculations and model sections\n</yakaai_skills>"}