#!/usr/bin/env python3 # -*- coding: utf-8 -*- """把四份 PDF 的抽取结果合并为一份格式统一的 DOCX 格式口径（用户 2026-10-26 拍板）： - 字号仅四档：22（封面）/16（标题）/12（正文）/10.5（表格、图注） - 中文 宋体，英文 Times New Roman - 页眉保留产品名+编号，页脚「第 X 页 共 Y 页」 - 图片原样保留 """ import json import os import sys from docx import Document from docx.enum.section import WD_SECTION from docx.enum.table import WD_TABLE_ALIGNMENT from docx.enum.text import WD_ALIGN_PARAGRAPH, WD_LINE_SPACING from docx.oxml import OxmlElement from docx.oxml.ns import qn from docx.shared import Cm, Pt, RGBColor OUT = '/Users/liuxinyuan/Desktop/Hermes输出-工作类/资料/M2181-4_产品稳定性研究资料_合订本_20261026.docx' HEADER_LEFT = '2×FastAmpli Premix-UNG Ⅳ（Probe qPCR）【货号：M2181-4】产品稳定性研究资料' HEADER_RIGHT = 'CA-10-04-R13-1.0' CHAPTERS = [ ('主报告：产品稳定性研究资料', '/tmp/main.json', 'main'), ('运输稳定性验证资料（2024 年）', '/tmp/trans.json', 'trans'), ('加速稳定性测试报告', '/tmp/acc.json', 'acc'), ('冻融稳定性测试报告', '/tmp/frz.json', 'frz'), ] BODY, H1, CAP, COVER = 12, 16, 10.5, 22 def set_font(run, size, bold=False, color=None): run.font.size = Pt(size) run.font.bold = bold run.font.name = 'Times New Roman' r = run._element.rPr.rFonts r.set(qn('w:eastAsia'), '宋体') r.set(qn('w:ascii'), 'Times New Roman') r.set(qn('w:hAnsi'), 'Times New Roman') if color: run.font.color.rgb = color def shade(cell, hexcolor): tcPr = cell._tc.get_or_add_tcPr() shd = OxmlElement('w:shd') shd.set(qn('w:val'), 'clear') shd.set(qn('w:color'), 'auto') shd.set(qn('w:fill'), hexcolor) tcPr.append(shd) def add_field(paragraph, instr, size=CAP): run = paragraph.add_run() set_font(run, size) fld = OxmlElement('w:fldSimple') fld.set(qn('w:instr'), instr) t = OxmlElement('w:t') t.text = '1' r = OxmlElement('w:r') r.append(t) fld.append(r) paragraph._p.append(fld) def style_doc(doc): st = doc.styles['Normal'] st.font.name = 'Times New Roman' st.font.size = Pt(BODY) st.element.rPr.rFonts.set(qn('w:eastAsia'), '宋体') pf = st.paragraph_format pf.line_spacing = 1.5 pf.space_after = Pt(3) sec = doc.sections[0] sec.page_height, sec.page_width = Cm(29.7), Cm(21.0) sec.top_margin = sec.bottom_margin = Cm(2.4) sec.left_margin = sec.right_margin = Cm(2.2) sec.header_distance = Cm(1.2) sec.footer_distance = Cm(1.2) # 页眉 hp = sec.header.paragraphs[0] hp.text = '' tab_stops = hp.paragraph_format.tab_stops tab_stops.add_tab_stop(Cm(16.6), alignment=2) r1 = hp.add_run(HEADER_LEFT) set_font(r1, CAP, color=RGBColor(0x44, 0x44, 0x44)) r2 = hp.add_run('\t' + HEADER_RIGHT) set_font(r2, CAP, color=RGBColor(0x44, 0x44, 0x44)) # 页脚下边框 pPr = hp._p.get_or_add_pPr() pbdr = OxmlElement('w:pBdr') bottom = OxmlElement('w:bottom') bottom.set(qn('w:val'), 'single'); bottom.set(qn('w:sz'), '4') bottom.set(qn('w:space'), '1'); bottom.set(qn('w:color'), 'AAAAAA') pbdr.append(bottom); pPr.append(pbdr) fp = sec.footer.paragraphs[0] fp.alignment = WD_ALIGN_PARAGRAPH.CENTER r = fp.add_run('第 ') set_font(r, CAP) add_field(fp, 'PAGE') r = fp.add_run(' 页 共 ') set_font(r, CAP) add_field(fp, 'NUMPAGES') r = fp.add_run(' 页') set_font(r, CAP) def add_cover(doc): for _ in range(4): doc.add_paragraph() p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.CENTER set_font(p.add_run('珠海宝锐生物科技有限公司'), COVER, bold=True, color=RGBColor(0x0D, 0x47, 0xA1)) p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.CENTER p.paragraph_format.space_before = Pt(24) set_font(p.add_run('2×FastAmpli Premix-UNG Ⅳ（Probe qPCR）'), COVER, bold=True) p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.CENTER set_font(p.add_run('产品稳定性研究资料'), COVER, bold=True) p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.CENTER p.paragraph_format.space_before = Pt(10) set_font(p.add_run('货号：M2181-4'), H1, color=RGBColor(0x44, 0x44, 0x44)) for _ in range(7): doc.add_paragraph() for line in ['文件编号：CA-10-04-R13-1.0', '版本：1.0', '整理日期：2026 年 10 月']: p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.CENTER set_font(p.add_run(line), BODY, color=RGBColor(0x44, 0x44, 0x44)) doc.add_page_break() def add_toc(doc, entries): p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.LEFT set_font(p.add_run('目 录'), H1, bold=True) p.paragraph_format.space_after = Pt(12) # Word 目录域：打开后自动/按 F9 更新页码 para = doc.add_paragraph() run = para.add_run() set_font(run, BODY) fldChar = OxmlElement('w:fldChar'); fldChar.set(qn('w:fldCharType'), 'begin') instr = OxmlElement('w:instrText'); instr.set(qn('xml:space'), 'preserve') instr.text = 'TOC \\o "1-2" \\h \\z \\u' sep = OxmlElement('w:fldChar'); sep.set(qn('w:fldCharType'), 'separate') t = OxmlElement('w:t'); t.text = '（右键“更新域”生成目录页码）' end = OxmlElement('w:fldChar'); end.set(qn('w:fldCharType'), 'end') run._r.append(fldChar); run._r.append(instr); run._r.append(sep); run._r.append(t); run._r.append(end) doc.add_page_break() def add_heading(doc, text, level=1): p = doc.add_paragraph() p.paragraph_format.space_before = Pt(12 if level == 1 else 8) p.paragraph_format.space_after = Pt(6) p.paragraph_format.keep_with_next = True set_font(p.add_run(text), H1, bold=True, color=RGBColor(0x12, 0x44, 0x6E)) def add_body(doc, text): p = doc.add_paragraph() p.paragraph_format.first_line_indent = Pt(BODY * 2) p.alignment = WD_ALIGN_PARAGRAPH.JUSTIFY set_font(p.add_run(text), BODY) def add_caption(doc, text): p = doc.add_paragraph() p.alignment = WD_ALIGN_PARAGRAPH.CENTER set_font(p.add_run(text), CAP, color=RGBColor(0x55, 0x55, 0x55)) def add_table(doc, rows): ncol = max(len(r) for r in rows) rows = [r + [''] * (ncol - len(r)) for r in rows] t = doc.add_table(rows=len(rows), cols=ncol) t.style = 'Table Grid' t.alignment = WD_TABLE_ALIGNMENT.CENTER t.autofit = True for i, row in enumerate(rows): for j, val in enumerate(row): cell = t.cell(i, j) cell.text = '' p = cell.paragraphs[0] p.paragraph_format.line_spacing = 1.15 p.paragraph_format.space_after = Pt(0) set_font(p.add_run(val), CAP, bold=(i == 0)) if i == 0: shade(cell, 'EDF3FA') # 表头重复 trPr = t.rows[0]._tr.get_or_add_trPr() tblHeader = OxmlElement('w:tblHeader') tblHeader.set(qn('w:val'), 'true') trPr.append(tblHeader) doc.add_paragraph().paragraph_format.space_after = Pt(0) def add_image(doc, path, w_px, h_px): width_cm = min(10.5, w_px / 96 * 2.54) p = doc.add_paragraph(); p.alignment = WD_ALIGN_PARAGRAPH.CENTER p.paragraph_format.space_before = Pt(6) p.paragraph_format.space_after = Pt(3) p.add_run().add_picture(path, width=Cm(width_cm)) 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_tables(m): """跨页表格续接：同列宽、同横向范围、近 3 页内，缺表头的补上最近表头""" last = {} # (ncol, x0, x1) -> (page_no, header_rows) for pg in m['pages']: for it in pg['items']: if it['type'] != 'table': continue rows = it['rows'] if not rows: continue ncol = max(len(r) for r in rows) rows = [r + [''] * (ncol - len(r)) for r in rows] x0, y0, x1, y1 = it['bbox'] key = (ncol, round(x0 / 20), round(x1 / 20)) first = ' '.join(rows[0]) is_head = sum(1 for k in HDR_KW if k in first) >= 2 if is_head: last[key] = (pg['page'], [list(r) for r in rows[:3]]) elif key in last and pg['page'] - last[key][0] <= 3: hdr = last[key][1] if not any(' '.join(h) == first for h in hdr): rows = [list(h) for h in hdr] + 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 def build(entries_override=None): doc = Document() style_doc(doc) add_cover(doc) # 先扫描目录条目 entries = [] for ct, path, pref in CHAPTERS: entries.append((ct, 1)) if not os.path.exists(path): continue m = json.load(open(path, encoding='utf-8')) start = 1 if 'main' in pref 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'] != ct: if it['text'] not in [e[0] for e in entries]: entries.append((it['text'], 2)) add_toc(doc, entries_override or entries) for ci, (ct, path, pref) in enumerate(CHAPTERS, 1): if not os.path.exists(path): continue m = json.load(open(path, encoding='utf-8')) augment_tables(m) add_heading(doc, f'第{["一","二","三","四"][ci-1]}部分 {ct}', 1) start = 1 if 'main' in pref else 0 for pg in m['pages'][start:]: for it in pg['items']: if it['type'] == 'h': add_heading(doc, it['text'], 2 if it['text'][0].isdigit() else 2) elif it['type'] == 'p': if len(it['text']) <= 30 and (it['text'].startswith('表') or it['text'].startswith('图')): add_caption(doc, it['text']) else: add_body(doc, it['text']) elif it['type'] == 'table': add_table(doc, it['rows']) elif it['type'] == 'img': add_image(doc, it['path'], it['w'], it['h']) # 原件一页结束 → 不强制分页，保持流式 # 打开时提示更新域 settings = doc.settings.element uf = OxmlElement('w:updateFields'); uf.set(qn('w:val'), 'true') settings.append(uf) doc.save(OUT) print('✅', OUT, '｜目录条目', len(entries_override or entries)) if __name__ == '__main__': build()