# -*- coding: utf-8 -*- """系统比对 v2:适配「名称\\n数值 状态\\n范围」布局""" import fitz, re, sys sys.path.insert(0, r'C:/Users/liaoxg.DESKTOP-DGLN79N/.workbuddy/skills/bun-report-parser/scripts') import extract_full_report_v5 as v5 doc = fitz.open(r'E:/cfc/docs/参考资料/560822024-张学红.pdf') all_lines = [] for page in doc: for l in page.get_text().split('\n'): l = l.strip() if l: all_lines.append(l) status_words = ['正常', '过多', '轻度产气', '过低', '不足', '缺乏', '偏⾼', '偏低', '注意'] val_re = re.compile(r'^(\d+)\s*(' + '|'.join(status_words) + r')$') range_re = re.compile(r'^\d+\s*-\s*\d+$') found = [] i = 0 while i < len(all_lines): m = val_re.match(all_lines[i]) if m: val, st = m.group(1), m.group(2) rng = '' if i + 1 < len(all_lines) and range_re.match(all_lines[i + 1]): rng = all_lines[i + 1] # 向上收集名称(可能 1-3 行) name_parts = [] j = i - 1 while j >= 0 and len(name_parts) < 3: prev = all_lines[j] if re.match(r'^\d+/169$', prev): j -= 1 continue if val_re.match(prev) or range_re.match(prev) or prev in status_words: break if prev in ('名称', '评估值', '正常范围', '指标范围', '过量 / 缺乏相关症状', '过量 /缺乏相关症状', '致病菌', '丰度', '评估', '相关症状'): break name_parts.insert(0, prev) if prev.endswith(')') or prev.endswith(')') or len(prev) <= 14: break j -= 1 name = ''.join(name_parts) if name: found.append((name, val, st, rng)) i += 1 whitelists = {'barrier': v5.KNOWN_BARRIER, 'scfa': v5.KNOWN_SCFA, 'neuro': v5.KNOWN_NEUROTRANSMITTER} all_known = set() for s in whitelists.values(): all_known |= set(s) all_known |= set(v5.KNOWN_AMINO) if hasattr(v5, 'KNOWN_AMINO') else set() print('===== 全部动态指标行 %d 条 =====' % len(found)) extra = [] for name, val, st, rng in found: nname = v5._npar(name) hit = any(nname.startswith(v5._npar(k)) or v5._npar(k).startswith(nname) or v5._npar(k) in nname for k in all_known) if not hit: extra.append((name, val, st, rng)) print(f" [{'OK ' if hit else 'NEW'}] {name} = {val} {st} 范围{rng}") print() print('===== 白名单外(漏采)%d 条 =====' % len(extra)) for name, val, st, rng in extra: print(f' >>> {name}: {val} {st} 范围{rng}')