refactor: HWP 변환을 exe(HWP→HWPX) 단일 경로로 교체, 이미지 경로 URL 인코딩
- hwp.py: COM/pyhwp 제거, HwpToPdfConverter.exe → hwpx 컨버터 재사용으로 단순화 - hwpx.py, hml.py: 이미지 경로의 공백/대괄호 URL 인코딩(%20, %5B, %5D) 추가 (Obsidian 등 Markdown 뷰어에서 [기본이론] 포함 파일명 이미지 표시 오류 수정) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
minsung
@@ -1,181 +1,36 @@
|
||||
#!/usr/bin/env python3
|
||||
"""HWP → Markdown (COM 자동화 우선, pyhwp fallback)"""
|
||||
"""HWP → Markdown (HwpToPdfConverter.exe → HWPX → MD)"""
|
||||
from __future__ import annotations
|
||||
|
||||
import re
|
||||
import shutil
|
||||
import tempfile
|
||||
import subprocess
|
||||
from pathlib import Path
|
||||
|
||||
|
||||
def _com_hwp_to_hml(hwp_path: Path, hml_path: Path, timeout: int = 15) -> bool:
|
||||
import threading
|
||||
result = [False]
|
||||
|
||||
def _run():
|
||||
try:
|
||||
import pythoncom, win32com.client
|
||||
except ImportError:
|
||||
return
|
||||
hwp = None
|
||||
try:
|
||||
pythoncom.CoInitialize()
|
||||
hwp = win32com.client.Dispatch('HWPFrame.HwpObject')
|
||||
try:
|
||||
hwp.RegisterModule('FilePathCheckDLL', 'SecurityModule')
|
||||
except Exception:
|
||||
pass
|
||||
ok = hwp.Open(str(hwp_path).replace('/', '\\'), 'HWP', 'forceopen:true')
|
||||
if not ok:
|
||||
return
|
||||
hwp.SaveAs(str(hml_path).replace('/', '\\'), 'HML', '')
|
||||
result[0] = hml_path.exists()
|
||||
except Exception as e:
|
||||
print(f' COM 오류: {e}')
|
||||
finally:
|
||||
if hwp:
|
||||
try: hwp.Quit()
|
||||
except Exception: pass
|
||||
try: pythoncom.CoUninitialize()
|
||||
except Exception: pass
|
||||
|
||||
t = threading.Thread(target=_run, daemon=True)
|
||||
t.start()
|
||||
t.join(timeout)
|
||||
if t.is_alive():
|
||||
print(f' COM 타임아웃 ({timeout}초) -> pyhwp로 전환')
|
||||
return result[0]
|
||||
_EXE_PATH = Path(__file__).parent.parent / 'HwpToHwpxConverter_260420' / 'HwpToPdfConverter.exe'
|
||||
|
||||
|
||||
def _table_to_md(table_elem) -> str:
|
||||
from bs4 import Tag
|
||||
rows = table_elem.find_all('tr', recursive=False) or table_elem.find_all('tr')
|
||||
if not rows:
|
||||
return ''
|
||||
has_merge = False
|
||||
parsed = []
|
||||
for tr in rows:
|
||||
cells = []
|
||||
for td in tr.find_all(['td', 'th']):
|
||||
cs = int(td.get('colspan', 1))
|
||||
rs = int(td.get('rowspan', 1))
|
||||
if cs > 1 or rs > 1:
|
||||
has_merge = True
|
||||
cells.append((cs, rs, td.get_text(separator='<br>', strip=True)))
|
||||
if cells:
|
||||
parsed.append(cells)
|
||||
if not parsed:
|
||||
return ''
|
||||
if has_merge:
|
||||
lines = ['<table>']
|
||||
for ri, cells in enumerate(parsed):
|
||||
lines.append('<tr>')
|
||||
tag = 'th' if ri == 0 else 'td'
|
||||
for cs, rs, text in cells:
|
||||
attrs = (f' colspan="{cs}"' if cs > 1 else '') + (f' rowspan="{rs}"' if rs > 1 else '')
|
||||
lines.append(f'<{tag}{attrs}>{text}</{tag}>')
|
||||
lines.append('</tr>')
|
||||
lines.append('</table>')
|
||||
return '\n'.join(lines)
|
||||
else:
|
||||
rows_text = [[text for _, _, text in cells] for cells in parsed]
|
||||
mc = max(len(r) for r in rows_text)
|
||||
for r in rows_text:
|
||||
r += [''] * (mc - len(r))
|
||||
def esc(s): return s.replace('|', '\\|')
|
||||
lines = ['| ' + ' | '.join(esc(c) for c in rows_text[0]) + ' |',
|
||||
'| ' + ' | '.join(['---'] * mc) + ' |']
|
||||
for row in rows_text[1:]:
|
||||
lines.append('| ' + ' | '.join(esc(c) for c in row) + ' |')
|
||||
return '\n'.join(lines)
|
||||
|
||||
|
||||
def _detect_structure(text: str):
|
||||
if not text: return 'paragraph', 0, text
|
||||
if re.match(r'^\d+\.\d+\.\d+\s', text): return 'heading', 4, text
|
||||
if re.match(r'^\d+\.\d+\s', text): return 'heading', 3, text
|
||||
if re.match(r'^\d+\.\s.+', text): return 'heading', 2, text
|
||||
if re.match(r'^[\d가-힣]+\)\s+.+', text):return 'heading', 3, text
|
||||
if re.match(r'^[□■]\s*.+', text): return 'heading', 2, text
|
||||
if re.match(r'^[○●◎]\s*.+', text): return 'heading', 3, text
|
||||
if re.match(r'^[▶▷]\s*.+', text): return 'heading', 4, text
|
||||
if re.match(r'^[▪▫\-]\s*.+', text): return 'bullet', 0, text
|
||||
if re.match(r'^[※]', text): return 'paragraph', 0, f'> {text}'
|
||||
return 'paragraph', 0, text
|
||||
|
||||
|
||||
def _pyhwp_hwp_to_md(hwp_path: Path, output_path: Path, base_name: str) -> bool:
|
||||
def _exe_hwp_to_hwpx(hwp_path: Path, timeout: int = 30) -> Path | None:
|
||||
"""exe로 HWP → HWPX 변환. 성공 시 생성된 .hwpx 경로 반환."""
|
||||
if not _EXE_PATH.exists():
|
||||
print(f' [경고] exe 없음: {_EXE_PATH}')
|
||||
return None
|
||||
hwpx_path = hwp_path.with_suffix('.hwpx')
|
||||
existed_before = hwpx_path.exists()
|
||||
try:
|
||||
from hwp5.hwp5html import HTMLTransform
|
||||
from hwp5.xmlmodel import Hwp5File
|
||||
from bs4 import BeautifulSoup
|
||||
except ImportError as e:
|
||||
print(f' pyhwp/bs4 미설치: {e}')
|
||||
return False
|
||||
|
||||
tmp_dir = Path(tempfile.mkdtemp())
|
||||
try:
|
||||
f = Hwp5File(str(hwp_path))
|
||||
HTMLTransform().transform_hwp5_to_dir(f, str(tmp_dir))
|
||||
xhtml_path = tmp_dir / 'index.xhtml'
|
||||
if not xhtml_path.exists():
|
||||
return False
|
||||
|
||||
images_dir = output_path.parent / f'{base_name}_images'
|
||||
images_dir.mkdir(exist_ok=True)
|
||||
img_map = {}
|
||||
bindata_dir = tmp_dir / 'bindata'
|
||||
if bindata_dir.exists():
|
||||
for img in bindata_dir.iterdir():
|
||||
shutil.copy(img, images_dir / img.name)
|
||||
img_map[img.name] = img.name
|
||||
|
||||
soup = BeautifulSoup(xhtml_path.read_text(encoding='utf-8'), 'lxml-xml')
|
||||
for area in soup.find_all(class_=re.compile(r'^(HeaderArea|FooterArea|Header parashape|Footer parashape)$')):
|
||||
area.decompose()
|
||||
|
||||
md_lines = []
|
||||
img_counter = [0]
|
||||
for elem in soup.find_all(['p', 'table']):
|
||||
if elem.find_parent('table'):
|
||||
continue
|
||||
if elem.name == 'table':
|
||||
if not elem.find_parent('p'):
|
||||
md = _table_to_md(elem)
|
||||
if md:
|
||||
md_lines += [md, '']
|
||||
elif elem.name == 'p':
|
||||
for img in elem.find_all('img'):
|
||||
fn = Path(img.get('src', '')).name
|
||||
if fn in img_map:
|
||||
img_counter[0] += 1
|
||||
md_lines += [f'![그림 {img_counter[0]}]({base_name}_images/{fn})', '']
|
||||
inner = elem.find('table')
|
||||
if inner:
|
||||
md = _table_to_md(inner)
|
||||
if md:
|
||||
md_lines += [md, '']
|
||||
continue
|
||||
text = re.sub(r'\s+', ' ', elem.get_text(separator=' ', strip=True)).strip()
|
||||
if not text:
|
||||
continue
|
||||
kind, level, fmt = _detect_structure(text)
|
||||
if kind == 'heading':
|
||||
if md_lines and md_lines[-1] != '':
|
||||
md_lines.append('')
|
||||
md_lines += [f'{"#" * level} {fmt}', '']
|
||||
elif kind == 'bullet':
|
||||
md_lines.append(f'- {re.sub(r"^[▪▫-]\\s*", "", fmt)}')
|
||||
else:
|
||||
md_lines += [fmt, '']
|
||||
|
||||
output_path.write_text(re.sub(r'\n{3,}', '\n\n', '\n'.join(md_lines)), encoding='utf-8')
|
||||
return True
|
||||
subprocess.run(
|
||||
[str(_EXE_PATH), str(hwp_path)],
|
||||
timeout=timeout,
|
||||
capture_output=True,
|
||||
)
|
||||
if hwpx_path.exists() and (not existed_before or hwpx_path.stat().st_mtime > hwp_path.stat().st_mtime):
|
||||
return hwpx_path
|
||||
print(f' [경고] exe 실행 후 .hwpx 파일 없음')
|
||||
return None
|
||||
except subprocess.TimeoutExpired:
|
||||
print(f' [경고] exe 타임아웃 ({timeout}초)')
|
||||
return None
|
||||
except Exception as e:
|
||||
print(f' pyhwp 오류: {e}')
|
||||
return False
|
||||
finally:
|
||||
shutil.rmtree(tmp_dir, ignore_errors=True)
|
||||
print(f' [경고] exe 오류: {e}')
|
||||
return None
|
||||
|
||||
|
||||
def convert_hwp(hwp_path: Path, output_dir: Path) -> dict:
|
||||
@@ -190,22 +45,19 @@ def convert_hwp(hwp_path: Path, output_dir: Path) -> dict:
|
||||
"output": str(md_path), "format": "hwp",
|
||||
}
|
||||
try:
|
||||
hml_path = md_path.with_suffix('.hml')
|
||||
if _com_hwp_to_hml(hwp_path, hml_path):
|
||||
try:
|
||||
from converters.hml import convert_hml
|
||||
r = convert_hml(hml_path, output_dir)
|
||||
hml_path.unlink(missing_ok=True)
|
||||
if r['status'] == 'ok':
|
||||
return result
|
||||
except Exception:
|
||||
hml_path.unlink(missing_ok=True)
|
||||
|
||||
if _pyhwp_hwp_to_md(hwp_path, md_path, hwp_path.stem):
|
||||
return result
|
||||
hwpx_path = _exe_hwp_to_hwpx(hwp_path)
|
||||
if hwpx_path:
|
||||
from converters.hwpx import convert_hwpx
|
||||
r = convert_hwpx(hwpx_path, output_dir)
|
||||
if r['status'] == 'ok':
|
||||
result['images'] = r.get('images', [])
|
||||
return result
|
||||
result['error'] = r.get('error', 'hwpx 변환 실패')
|
||||
else:
|
||||
result['error'] = 'HWP → HWPX 변환 실패 — HwpToPdfConverter.exe 확인 필요'
|
||||
|
||||
result['status'] = 'error'
|
||||
result['error'] = 'COM + pyhwp 모두 실패'
|
||||
print(f' [경고] HWP 변환 실패: {hwp_path.name}')
|
||||
except Exception as e:
|
||||
result['status'] = 'error'
|
||||
result['error'] = str(e)
|
||||
|
||||
Reference in New Issue
Block a user