#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""内容模型 → HTML（供 Chrome 打印 PDF），与 DOCX 同一套字号/字体口径"""
import json
import os
import re
import sys
OUT_HTML = '/tmp/stability.html'
CHAPTERS = [
('第一部分 主报告：产品稳定性研究资料', '/tmp/main.json', 'main'),
('第二部分 运输稳定性验证资料（2024 年）', '/tmp/trans.json', 'trans'),
('第三部分 加速稳定性测试报告', '/tmp/acc.json', 'acc'),
('第四部分 冻融稳定性测试报告', '/tmp/frz.json', 'frz'),
]
HDR_KW = ('分组', '样品', '检测项目', '批号', '结果', '项目', '孔数', '名称', '浓度', '时间', '编号', 'Ct')
def augment(m):
last = {}
for pg in m['pages']:
for it in pg['items']:
if it['type'] != 'table' or not it['rows']:
continue
rows = it['rows']
n = max(len(r) for r in rows)
rows = [r + [''] * (n - len(r)) for r in rows]
x0, y0, x1, y1 = it['bbox']
key = (n, round(x0 / 20), round(x1 / 20))
first = ' '.join(rows[0])
head = sum(1 for k in HDR_KW if k in first) >= 2
if head:
last[key] = (pg['page'], [list(r) for r in rows[:3]])
elif key in last and pg['page'] - last[key][0] <= 3:
rows = [list(h) for h in last[key][1]] + rows
rows = [[c.replace('\n', ' ').strip() for c in r] for r in rows]
ded = []
for r in rows:
if ded and r == ded[-1]:
continue
ded.append(r)
it['rows'] = ded # DEDUPE_MARK
FIX = [('ProbeqPCR', 'Probe qPCR'), ('FastAmpliPremix', 'FastAmpli Premix'),
('Prime-ProbeMix', 'Prime-Probe Mix'), ('TemplateDNA', 'Template DNA'),
('ddH O 2', 'ddH2O'), ('UNG Ⅳ(Probe', 'UNG Ⅳ (Probe')]
def esc(s):
s = s or ''
for a, b in FIX:
s = s.replace(a, b)
return s.replace('&', '&').replace('<', '<').replace('>', '>')
def build(toc_pages=None):
parts = ['''
''']
parts.append('''
珠海宝锐生物科技有限公司
2×FastAmpli Premix-UNG Ⅳ（Probe qPCR）
产品稳定性研究资料
货号：M2181-4
文件编号：CA-10-04-R13-1.0 版本：1.0 整理日期：2026 年 10 月
''')
# 目录
toc = []
models = {}
for title, path, pref in CHAPTERS:
if not os.path.exists(path):
continue
m = json.load(open(path, encoding='utf-8'))
augment(m)
models[pref] = m
toc.append((title, 1))
start = 1 if pref == 'main' else 0
for pg in m['pages'][start:]:
for it in pg['items']:
if it['type'] == 'h' and len(it['text']) <= 40 and it['text'] not in [t[0] for t in toc]:
t2 = it['text'].strip()
if re.match(r'^\d+(\.\d+)?[\s、\.]', t2) and t2.count('.') <= 1:
toc.append((t2, 2))
parts.append('
目 录
')
for i, (t, lv) in enumerate(toc):
pn = toc_pages.get(str(i)) if toc_pages else ''
parts.append(f'
{esc(t)}{pn}
')
parts.append('
')
idx = 0
for title, path, pref in CHAPTERS:
if pref not in models:
continue
parts.append(f'
{esc(title)}
'); idx += 1
m = models[pref]
start = 1 if pref == 'main' else 0
for pg in m['pages'][start:]:
for it in pg['items']:
if it['type'] == 'h':
parts.append(f'
{esc(it["text"])}
'); idx += 1
elif it['type'] == 'p':
cls = ' class="cap"' if (len(it['text']) <= 30 and it['text'][:1] in '表图') else ''
parts.append(f'{esc(it["text"])}')
elif it['type'] == 'table':
rows = it['rows']
if not rows:
continue
parts.append('
')
parts.append('
' + ''.join(f'
{esc(c)}
' for c in rows[0]) + '
')
for r in rows[1:]:
parts.append('
' + ''.join(f'
{esc(c)}
' for c in r) + '
')
parts.append('
')
elif it['type'] == 'img':
parts.append(f'')
parts.append('')
open(OUT_HTML, 'w', encoding='utf-8').write('\n'.join(parts))
json.dump([[t, lv] for t, lv in toc], open('/tmp/toc.json', 'w', encoding='utf-8'), ensure_ascii=False)
print('html →', OUT_HTML, '｜toc', len(toc))
if __name__ == '__main__':
build(json.load(open('/tmp/toc_pages.json')) if os.path.exists('/tmp/toc_pages.json') else None)