@simplysm/sd-claude 13.0.77 → 13.0.80

package/claude/skills/sd-document/extract_docx.py

@@ -2,24 +2,13 @@
 """Extract text and images from DOCX files in paragraph flow order."""
 
 import sys
-import io
-import os
-import subprocess
-from pathlib import Path
+from _common import (
+    setup_encoding, make_output_paths, print_header, save_image,
+    ext_from_content_type, print_image_summary, run_cli,
+    normalize_cell, parse_heading_level,
+)
 
-sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
-sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding="utf-8", errors="replace")
-
-
-def ensure_packages():
-    packages = {"python-docx": "docx"}
-    for pip_name, import_name in packages.items():
-        try:
-            __import__(import_name)
-        except ImportError:
-            print(f"Installing package: {pip_name}...", file=sys.stderr)
-            subprocess.check_call([sys.executable, "-m", "pip", "install", pip_name],
-                                  stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+setup_encoding()
 
 
 def extract(file_path):
@@ -27,14 +16,12 @@ def extract(file_path):
     from docx.oxml.ns import qn
 
     doc = Document(file_path)
-    stem = Path(file_path).stem
-    out_dir = Path(file_path).parent / f"{stem}_files"
+    fp, out_dir = make_output_paths(file_path)
     img_idx = 0
 
-    print(f"# {Path(file_path).name}\n")
+    print_header(fp)
 
-    for i, para in enumerate(doc.paragraphs):
-        has_image = False
+    for para in doc.paragraphs:
         for run in para.runs:
             drawings = run._element.findall(f".//{qn('wp:inline')}") + run._element.findall(f".//{qn('wp:anchor')}")
             for drawing in drawings:
@@ -45,48 +32,32 @@ def extract(file_path):
                         rel = doc.part.rels.get(embed)
                         if rel:
                             img_idx += 1
-                            blob = rel.target_part.blob
-                            content_type = rel.target_part.content_type
-                            ext = content_type.split("/")[-1].replace("jpeg", "jpg")
-                            out_dir.mkdir(parents=True, exist_ok=True)
-                            img_path = out_dir / f"img_{img_idx:03d}.{ext}"
-                            img_path.write_bytes(blob)
+                            ext = ext_from_content_type(rel.target_part.content_type)
+                            img_path = save_image(out_dir, img_idx, rel.target_part.blob, ext)
                             print(f"[IMG] {img_path}")
-                            has_image = True
 
         text = para.text.strip()
         if text:
             style = para.style.name if para.style else ""
             prefix = ""
             if "Heading" in style:
-                level = style.replace("Heading ", "").replace("Heading", "1")
-                try:
-                    prefix = "#" * int(level) + " "
-                except ValueError:
+                level = parse_heading_level(style)
+                if level is not None:
+                    prefix = "#" * level + " "
+                else:
                     prefix = "## "
             print(f"{prefix}{text}")
 
-        if has_image or text:
-            pass  # already printed
-        elif not has_image and not text:
-            continue
-
     # Table extraction
     for t_idx, table in enumerate(doc.tables):
         print(f"\n### Table {t_idx + 1}\n")
         for row in table.rows:
-            cells = [cell.text.strip().replace("\n", " ") for cell in row.cells]
+            cells = [normalize_cell(cell.text) for cell in row.cells]
             print("| " + " | ".join(cells) + " |")
 
-    if img_idx > 0:
-        print(f"\n---\n{img_idx} image(s) saved: {out_dir}")
-    else:
-        print("\n---\nNo images")
+    print()
+    print_image_summary(img_idx, out_dir)
 
 
 if __name__ == "__main__":
-    if len(sys.argv) < 2:
-        print("Usage: python extract_docx.py <file.docx>", file=sys.stderr)
-        sys.exit(1)
-    ensure_packages()
-    extract(sys.argv[1])
+    run_cli(extract, "extract_docx.py", {"python-docx": "docx"})

package/claude/skills/sd-document/extract_pdf.py

@@ -2,35 +2,22 @@
 """Extract text, tables, and images from PDF files page by page."""
 
 import sys
-import io
-import subprocess
-from pathlib import Path
+from _common import (
+    setup_encoding, make_output_paths, print_header, save_image,
+    print_image_summary, run_cli, normalize_cell,
+)
 
-sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
-sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding="utf-8", errors="replace")
-
-
-def ensure_packages():
-    packages = {"pdfplumber": "pdfplumber", "pypdf": "pypdf"}
-    for pip_name, import_name in packages.items():
-        try:
-            __import__(import_name)
-        except ImportError:
-            print(f"Installing package: {pip_name}...", file=sys.stderr)
-            subprocess.check_call([sys.executable, "-m", "pip", "install", pip_name],
-                                  stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+setup_encoding()
 
 
 def extract(file_path):
     import pdfplumber
     from pypdf import PdfReader
 
-    stem = Path(file_path).stem
-    out_dir = Path(file_path).parent / f"{stem}_files"
+    fp, out_dir = make_output_paths(file_path)
     img_idx = 0
-    total_text_len = 0
 
-    print(f"# {Path(file_path).name}\n")
+    print_header(fp)
 
     # Text + table extraction (pdfplumber)
     with pdfplumber.open(file_path) as pdf:
@@ -38,17 +25,18 @@ def extract(file_path):
             print(f"## Page {page_num}\n")
 
             text = page.extract_text()
-            if text and text.strip():
-                total_text_len += len(text.strip())
-                print(text.strip())
-                print()
+            if text:
+                stripped = text.strip()
+                if stripped:
+                    print(stripped)
+                    print()
 
             tables = page.extract_tables()
             for t_idx, table in enumerate(tables):
                 if table:
                     print(f"### Table {t_idx + 1}\n")
                     for row in table:
-                        cells = [(c or "").strip().replace("\n", " ") for c in row]
+                        cells = [normalize_cell(c) for c in row]
                         print("| " + " | ".join(cells) + " |")
                     print()
 
@@ -65,38 +53,22 @@ def extract(file_path):
                 filters = obj.get("/Filter", "")
                 if isinstance(filters, list):
                     filters = filters[0] if filters else ""
+                filter_str = str(filters)
                 ext = "png"
-                if "/DCTDecode" in str(filters):
+                if "/DCTDecode" in filter_str:
                     ext = "jpg"
-                elif "/JPXDecode" in str(filters):
+                elif "/JPXDecode" in filter_str:
                     ext = "jp2"
-                out_dir.mkdir(parents=True, exist_ok=True)
-                img_path = out_dir / f"img_{img_idx:03d}.{ext}"
                 try:
-                    img_path.write_bytes(obj.get_data())
-                except Exception:
-                    img_path = out_dir / f"img_{img_idx:03d}.bin"
-                    img_path.write_bytes(obj._data if hasattr(obj, "_data") else b"")
+                    img_path = save_image(out_dir, img_idx, obj.get_data(), ext)
+                except Exception as exc:
+                    print(f"Warning: failed to decode image {img_idx}: {exc}", file=sys.stderr)
+                    img_path = save_image(out_dir, img_idx, obj._data if hasattr(obj, "_data") else b"", "bin")
                 print(f"[IMG] (page={page_num}) {img_path}")
 
-    # OCR notice
-    if total_text_len == 0:
-        print("\n⚠ No text was extracted (may be a scanned PDF).")
-        print("OCR is required:")
-        print("  1. Install Tesseract OCR: https://github.com/tesseract-ocr/tesseract")
-        print("  2. pip install pytesseract pdf2image")
-        print("  3. Extract with pytesseract.image_to_string()")
-
     print()
-    if img_idx > 0:
-        print(f"---\n{img_idx} image(s) saved: {out_dir}")
-    else:
-        print("---\nNo images")
+    print_image_summary(img_idx, out_dir)
 
 
 if __name__ == "__main__":
-    if len(sys.argv) < 2:
-        print("Usage: python extract_pdf.py <file.pdf>", file=sys.stderr)
-        sys.exit(1)
-    ensure_packages()
-    extract(sys.argv[1])
+    run_cli(extract, "extract_pdf.py", {"pdfplumber": "pdfplumber", "pypdf": "pypdf"})

package/claude/skills/sd-document/extract_pptx.py

@@ -2,23 +2,12 @@
 """Extract text and images from PPTX files with per-slide coordinates."""
 
 import sys
-import io
-import subprocess
-from pathlib import Path
+from _common import (
+    setup_encoding, make_output_paths, print_header, save_image,
+    ext_from_content_type, print_image_summary, run_cli,
+)
 
-sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
-sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding="utf-8", errors="replace")
-
-
-def ensure_packages():
-    packages = {"python-pptx": "pptx"}
-    for pip_name, import_name in packages.items():
-        try:
-            __import__(import_name)
-        except ImportError:
-            print(f"Installing package: {pip_name}...", file=sys.stderr)
-            subprocess.check_call([sys.executable, "-m", "pip", "install", pip_name],
-                                  stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+setup_encoding()
 
 
 def emu_to_inches(emu):
@@ -28,50 +17,38 @@ def emu_to_inches(emu):
     return f"{emu / 914400:.1f}"
 
 
+def _pos(shape):
+    return f"(left={emu_to_inches(shape.left)}\", top={emu_to_inches(shape.top)}\")"
+
+
 def extract(file_path):
     from pptx import Presentation
     from pptx.enum.shapes import MSO_SHAPE_TYPE
 
     prs = Presentation(file_path)
-    stem = Path(file_path).stem
-    out_dir = Path(file_path).parent / f"{stem}_files"
+    fp, out_dir = make_output_paths(file_path)
     img_idx = 0
 
-    print(f"# {Path(file_path).name}\n")
+    print_header(fp)
 
     for slide_num, slide in enumerate(prs.slides, 1):
         print(f"## Slide {slide_num}\n")
 
         for shape in slide.shapes:
-            left = emu_to_inches(shape.left)
-            top = emu_to_inches(shape.top)
-            pos = f"(left={left}\", top={top}\")"
-
             if shape.shape_type == MSO_SHAPE_TYPE.PICTURE:
                 img_idx += 1
-                blob = shape.image.blob
-                content_type = shape.image.content_type
-                ext = content_type.split("/")[-1].replace("jpeg", "jpg")
-                out_dir.mkdir(parents=True, exist_ok=True)
-                img_path = out_dir / f"img_{img_idx:03d}.{ext}"
-                img_path.write_bytes(blob)
-                print(f"[IMG] {pos} {img_path}")
+                ext = ext_from_content_type(shape.image.content_type)
+                img_path = save_image(out_dir, img_idx, shape.image.blob, ext)
+                print(f"[IMG] {_pos(shape)} {img_path}")
 
             elif hasattr(shape, "text") and shape.text.strip():
                 text = shape.text.strip().replace("\n", "\n       ")
-                print(f"[TXT] {pos} {text}")
+                print(f"[TXT] {_pos(shape)} {text}")
 
         print()
 
-    if img_idx > 0:
-        print(f"---\n{img_idx} image(s) saved: {out_dir}")
-    else:
-        print("---\nNo images")
+    print_image_summary(img_idx, out_dir)
 
 
 if __name__ == "__main__":
-    if len(sys.argv) < 2:
-        print("Usage: python extract_pptx.py <file.pptx>", file=sys.stderr)
-        sys.exit(1)
-    ensure_packages()
-    extract(sys.argv[1])
+    run_cli(extract, "extract_pptx.py", {"python-pptx": "pptx"})

package/claude/skills/sd-document/extract_xlsx.py

@@ -2,36 +2,22 @@
 """Extract data and images from XLSX files with cell positions."""
 
 import sys
-import io
-import os
-import subprocess
-from pathlib import Path
+from _common import (
+    setup_encoding, make_output_paths, print_header, save_image,
+    print_image_summary, run_cli,
+)
 
-sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
-sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding="utf-8", errors="replace")
-
-
-def ensure_packages():
-    packages = {"openpyxl": "openpyxl"}
-    for pip_name, import_name in packages.items():
-        try:
-            __import__(import_name)
-        except ImportError:
-            print(f"Installing package: {pip_name}...", file=sys.stderr)
-            subprocess.check_call([sys.executable, "-m", "pip", "install", pip_name],
-                                  stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+setup_encoding()
 
 
 def extract(file_path):
     from openpyxl import load_workbook
-    from openpyxl.drawing.image import Image as XlImage
 
     wb = load_workbook(file_path, data_only=True)
-    stem = Path(file_path).stem
-    out_dir = Path(file_path).parent / f"{stem}_files"
+    fp, out_dir = make_output_paths(file_path)
     img_idx = 0
 
-    print(f"# {Path(file_path).name}\n")
+    print_header(fp)
 
     for sheet_name in wb.sheetnames:
         ws = wb[sheet_name]
@@ -51,33 +37,26 @@ def extract(file_path):
                     cells.append("")
                 else:
                     cells.append(str(val).strip())
-            print(f"[{row[0].coordinate.split('1')[0]}{row[0].row}] " + " | ".join(cells))
+            print(f"[{row[0].column_letter}{row[0].row}] " + " | ".join(cells))
 
         # Image extraction
-        if ws._images:
-            for img in ws._images:
+        images = getattr(ws, '_images', [])
+        if images:
+            for img in images:
                 img_idx += 1
                 anchor = ""
-                if hasattr(img.anchor, '_from'):
-                    anchor = f" (near {img.anchor._from.col},{img.anchor._from.row})"
-                ext = "png"
-                out_dir.mkdir(parents=True, exist_ok=True)
-                img_path = out_dir / f"img_{img_idx:03d}.{ext}"
-                with open(img_path, "wb") as f:
-                    f.write(img._data())
+                anchor_from = getattr(getattr(img, 'anchor', None), '_from', None)
+                if anchor_from is not None:
+                    anchor = f" (near {anchor_from.col},{anchor_from.row})"
+                data_fn = getattr(img, '_data', None)
+                blob = data_fn() if callable(data_fn) else b""
+                img_path = save_image(out_dir, img_idx, blob, "png")
                 print(f"[IMG]{anchor} {img_path}")
 
         print()
 
-    if img_idx > 0:
-        print(f"---\n{img_idx} image(s) saved: {out_dir}")
-    else:
-        print("---\nNo images")
+    print_image_summary(img_idx, out_dir)
 
 
 if __name__ == "__main__":
-    if len(sys.argv) < 2:
-        print("Usage: python extract_xlsx.py <file.xlsx>", file=sys.stderr)
-        sys.exit(1)
-    ensure_packages()
-    extract(sys.argv[1])
+    run_cli(extract, "extract_xlsx.py", {"openpyxl": "openpyxl"})

package/claude/skills/sd-email-analyze/SKILL.md

@@ -1,52 +1,44 @@
 ---
 name: sd-email-analyze
-description: "Use when the user's request involves .eml or .msg files. Triggers: email file analysis, email content extraction, attachment extraction, email summary."
-model: haiku
+description: .eml 또는 .msg 파일과 관련하여 "이메일 파일 분석", "이메일 내용 추출", "첨부파일 추출", "이메일 요약"을 요청할 때 사용.
 ---
 
-# Email Analyzer
+# SD Email Analyze — 이메일 파일 분석 및 내용 추출
 
-## Overview
+`.eml` 및 `.msg`(Outlook) 이메일 파일을 파싱하여 메일 헤더, 본문 텍스트, 인라인 이미지, 첨부파일을 추출하고 분석한다.
 
-Python script that parses `.eml` and `.msg` (Outlook) email files. Extracts mail headers, body text, inline images, and attachments to disk. Content analysis of extracted files is delegated to Claude's Read tool and `sd-document` skill.
+ARGUMENTS: 이메일 파일 경로 (필수). `.eml` 또는 `.msg` 파일 경로를 지정한다.
 
-## When to Use
+---
 
-- User provides a `.eml` or `.msg` file to analyze or summarize
-- Email content needs proper decoding
+## Step 1: 이메일 파일 파싱
 
-## Usage
+ARGUMENTS에서 이메일 파일 경로를 추출하여 아래 명령을 실행하라:
 
 ```bash
-python .claude/skills/sd-email-analyze/email-analyzer.py <email_file_path>
+python .claude/skills/sd-email-analyze/email-analyzer.py <이메일_파일_경로>
 ```
 
-First run auto-installs: `extract-msg`.
-
-### After Running
+- 최초 실행 시 `extract-msg`를 자동 설치한다.
+- 실행 결과로 마크다운 보고서가 표준 출력되며, `<이메일_파일명>_files/` 디렉토리에 추출된 파일이 저장된다.
 
-1. Read the markdown output (mail info, body text, file paths)
-2. **Inline images**: Use **Read** tool on each saved path to view
-3. **Attachments**: Use **Read** tool (images) or **sd-document** skill scripts (DOCX, XLSX, PPTX, PDF)
+### 출력 구조
 
-## Output
+1. **메일 정보 테이블**: 제목, 보낸 사람, 받는 사람, 참조, 날짜, 개수
+2. **본문 텍스트**: 일반 텍스트 (일반 텍스트가 없으면 HTML에서 태그 제거)
+3. **인라인 이미지**: 저장된 파일 경로 테이블
+4. **첨부파일**: 저장된 파일 경로 테이블
 
-- `<email_stem>_files/` directory with all extracted files
-- Markdown report to stdout:
-  1. **Mail info table**: Subject, From, To, Cc, Date, counts
-  2. **Body text**: Plain text (HTML stripped if no plain text)
-  3. **Inline images**: Table with saved file paths
-  4. **Attachments**: Table with saved file paths
+## Step 2: 추출된 파일 분석
 
-## Inline Image Handling
+Step 1의 출력에서 추출된 파일 경로를 확인하고 아래를 수행하라:
 
-Two sources extracted:
+1. **인라인 이미지**: 저장된 각 경로에 **Read** 도구를 사용하여 확인
+2. **첨부파일**: **Read** 도구(이미지) 또는 **sd-document** 스킬 스크립트(DOCX, XLSX, PPTX, PDF) 사용
 
-1. **CID images**: MIME parts with Content-ID (`cid:` references in HTML)
-2. **Data URI images**: Base64-encoded images in HTML (`data:image/...;base64,...`)
+### 인라인 이미지 처리
 
-## Common Mistakes
+두 가지 소스에서 추출된다:
 
-- **Wrong Python**: Ensure `python` points to Python 3.8+
-- **Firewall blocking pip**: First run needs internet for `extract-msg` install
-- **Forgetting inline images**: Always check the "Inline images" section and read each path
+1. **CID 이미지**: Content-ID가 있는 MIME 파트 (HTML의 `cid:` 참조)
+2. **Data URI 이미지**: HTML 내 Base64 인코딩 이미지 (`data:image/...;base64,...`)