初始归档：健康管理资料库（大医网/药膳/体质监测/中医理论等12个资料集，7712个文件）

中医入门资料库/05_脚本工具/build_theory_basis.py

@@ -0,0 +1,180 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+构建《中医入门》应用理论依据库 → 02_加工数据/理论依据库.json
+用途: 为体质调理/药膳推荐/处方分析等应用提供可调用的理论依据(原文+行号溯源)
+每条依据含: 原文引文 + 行号(指向 02_加工数据/中医入门.md)
+"""
+import json, os, re, datetime
+
+BOOK = "/home/songyi/Documents/ai_agent_scraper_study/data/中医入门资料库/02_加工数据/中医入门.md"
+OUT  = "/home/songyi/Documents/ai_agent_scraper_study/data/中医入门资料库/02_加工数据/理论依据库.json"
+TODAY = datetime.date.today().isoformat()
+
+lines = open(BOOK, encoding="utf-8").read().splitlines()
+text  = "\n".join(lines)
+def ln(idx):  # 字符偏移 → 行号(1-based)
+    return text[:idx].count("\n") + 1
+
+def find_seg(start, end):
+    i, j = text.find(start), text.find(end)
+    return (i, j) if 0 <= i < j else (i, len(text))
+
+lib = {"metadata": {
+    "生成日期": TODAY,
+    "来源": "中医入门资料库/02_加工数据/中医入门.md（秦伯未《中医入门》体例，用户定级: 重要理论基础）",
+    "用途": "应用调用层——体质调理/药膳推荐/处方分析/配伍安全核查时，调用本库返回《中医入门》理论依据(原文+行号)",
+    "查询接口": "../05_脚本工具/theory_api.py"}}
+
+# ============ 八法 ============
+i0, j0 = find_seg("三、八法", "四、常用治法")
+bafa = {}
+for m in re.finditer(r'^\d+[\.、]\s*([^\s：:，。]{1,2}法)[：:]\s*(.+)$', text[i0:j0], re.M):
+    name = m.group(1)
+    seg_after = text[i0+j0 if False else i0+m.end(): j0]
+    first = seg_after.split("。")[0]
+    bafa[name] = {"定义原文": (m.group(2).split("。")[0] + "。").strip() if m.group(2) else "",
+                  "行号": ln(i0 + m.start())}
+lib["八法"] = bafa
+
+# ============ 常用治法 72 ============
+i0, j0 = find_seg("四、常用治法", "## 第三章")
+zhifa, missing = {}, []
+for m in re.finditer(r'^(\d+)[\.、，,]?\s*([^\s：:，。]{2,12}法)[：:]\s*(.*)$', text[i0:j0], re.M):
+    num, name, rest = int(m.group(1)), m.group(2), m.group(3)
+    use = re.search(r'用于(.*?)(?:药如|。|$)', rest)
+    yao = re.search(r'药如([^.。；;]*)', rest)
+    zhifa[str(num)] = {"名称": name,
+        "适应症": use.group(1).strip("，,。 ") if use else None,
+        "药如": ([x.strip() for x in re.split(r'[、，,]', yao.group(1)) if x.strip()] if yao else []),
+        "行号": ln(i0 + m.start())}
+nums = sorted(int(k) for k in zhifa)
+gaps = [n for n in range(nums[0], nums[-1]+1) if n not in nums]
+lib["治法72"] = {"条目": zhifa, "总数": len(zhifa), "编号范围": [nums[0], nums[-1]],
+                 "缺号": gaps, "缺号说明": "原书编号缺号或OCR漏行，如实记录" if gaps else None}
+
+# ============ 七方 ============
+# 切片止于七方段结尾（其后是张景岳八阵与汪昂22类剂型分类，"22. 救急方"属剂型分类，非七方）
+i0, j0 = find_seg("二、七方", "三、剂型")
+cut = text.find("七方是方剂组成的法则之一", i0)
+if 0 < cut < j0: j0 = cut
+qifang, cur, cur_pos = {}, None, None
+for m in re.finditer(r'^\d+[\.、]\s*([^\s：:，。]{1,3})方[：:]\s*(.+)$|^(二、七方)', text[i0:j0], re.M):
+    if m.group(3):
+        cur = None; continue
+    cur = m.group(1) + "方"
+    para = m.group(2)  # (.+)$ 已捕获整段（每段一行），无需再找空行
+    # 清洗举例药名: 剥离连接词前缀（"如下法中的大承气汤"→"大承气汤"）
+    examples = []
+    for tok in re.findall(r'([\u4e00-\u9fff]{2,9}(?:汤|丸|散|饮|膏))', para):
+        tok = re.split(r'[如用和及等是的中的将]', tok)[-1]
+        if len(tok) >= 3 and tok not in examples: examples.append(tok)
+    qifang[cur] = {"定义原文": para[:150].strip(),
+                   "书内举例": examples,
+                   "行号": ln(i0 + m.start())}
+lib["七方"] = qifang
+
+# ============ 效能分类 15 ============
+i0, j0 = find_seg("二、效能", "#### 1. 扶正类")
+xiaoneng = {}
+for m in re.finditer(r'^(\d+)[\.、]\s*([^\s：:，。]{2,8}药)[：:]\s*(.+)$', text[i0:j0], re.M):
+    seg = m.group(3)
+    reps = []
+    for km in re.finditer(r'([\u4e00-\u9fff]{2,8})如([、\u4e00-\u9fff（）]{2,60})', seg):
+        for h in re.split(r'[、]', km.group(2)):
+            h = re.sub(r'（[^）]*）', '', h).strip("等")
+            if h: reps.append(h)
+    xiaoneng[m.group(2)] = {"定义原文": seg.split("。")[0] + "。", "代表药": reps[:24],
+                            "行号": ln(i0 + m.start())}
+lib["效能分类"] = xiaoneng
+
+# ============ 扶正/祛邪功效药组 ============
+groups = []
+def parse_groups(start, end, kind):
+    i0, j0 = find_seg(start, end)
+    cur_sys = ""
+    for m in re.finditer(r'^\((\d+)\)\s*属于(.+?)：|^[^\n#\d(].*?——.*$', text[i0:j0], re.M):
+        line = m.group(0).strip()
+        if m.group(1):
+            cur_sys = m.group(2); continue
+        eff, _, rhs = line.partition("——")
+        label = f"{kind}·{cur_sys}·{eff.strip()}" if cur_sys else f"{kind}·{eff.strip()}"
+        herbs = []
+        for h in re.split(r'[、，,]', rhs.strip().strip("。")):
+            h = re.sub(r'（[^）]*）', '', h).strip()
+            h = re.sub(r'^(亦可用鲜|亦可用)', '', h).strip()
+            if h: herbs.append(h)
+        groups.append({"组": label, "药": herbs, "行号": ln(i0 + m.start())})
+parse_groups("#### 1. 扶正类", "#### 2. 祛邪类", "扶正类")
+parse_groups("#### 2. 祛邪类", "三、归经", "祛邪类")
+lib["功效药组"] = {"条目": groups, "总数": len(groups)}
+
+# ============ 示例方 40 ============
+i0, j0 = find_seg("一、基本方剂", "## 第四章")
+shili = {}
+for m in re.finditer(r'^(\d+)[\.、]\s*([^：:\n]+)[：:]\s*(.+)$', text[i0:j0], re.M):
+    name = m.group(2).strip()
+    rest = m.group(3)
+    comp = [c.strip() for c in rest.split("。")[0].split("、") if c.strip()]
+    pos = re.search(r'为([^，。]{2,12})主方', rest)
+    zhi = re.search(r'用于([^。]*)', rest)
+    shili[name] = {"序号": int(m.group(1)), "组成_书": comp,
+                   "主方定位": pos.group(1) if pos else None,
+                   "主治": zhi.group(1).strip() if zhi else None,
+                   "行号": ln(i0 + m.start())}
+lib["示例方40"] = shili
+
+# ============ 二十八脉主症 ============
+i0, j0 = find_seg("二十八脉的主症", "## 第三章")
+mai = {}
+for m in re.finditer(r'^([^\s，。]{1,2})脉主(.+)$', text[i0:j0], re.M):
+    mai[m.group(1) + "脉"] = {"主症原文": m.group(2).strip().rstrip("。"), "行号": ln(i0 + m.start())}
+lib["脉象主症"] = {"条目": mai, "总数": len(mai),
+                   "说明": "标题为二十八脉，实列27条，长脉仅见总括韵文（原书编排，如实记录）"}
+
+# ============ 用药安全 ============
+ph_i = text.find("一、配合"); ph_j = text.find("二、用量")
+ph = text[ph_i:ph_j]
+i18 = ph.find("十八反歌：")
+seg18_prose = ph[ph.find("歌中所提十八种药"):ph.find("歌中所提十九种药")]
+seg19_prose = ph[ph.find("歌中所提十九种药"):ph.find("此外，妊娠禁忌")]
+# 畏对提取: 从"如硫黄畏朴硝"起取，避开首句"即表示相畏比较显著"的噪音
+iex = seg19_prose.find("如硫黄畏朴硝")
+pairs = [(a.replace("如", "", 1) if a.startswith("如") else a, b)
+         for a, b in re.findall(r'([\u4e00-\u9fff]{1,4})畏([\u4e00-\u9fff]{1,5})', seg19_prose[iex:])
+         if len(a) >= 2 and len(b) >= 2]
+for extra in [("川乌", "犀角"), ("草乌", "犀角")]:  # "川乌、草乌畏犀角"正则只捕草乌，补川乌
+    if extra not in pairs: pairs.append(extra)
+i19 = ph.find("十九畏歌：")
+seg_preg = ph[ph.find("此外，妊娠禁忌"):ph.find("经验告诉我们")]
+
+safety = {"十八反": {
+    "相反组": [
+        {"组": ["乌头"], "组_通行扩展": ["川乌", "草乌", "附子"],
+         "反": ["半夏", "瓜蒌", "贝母", "白蔹", "白及"], "反_通行扩展": [],
+         "说明": "书列'乌头'；附子为乌头子根，通行本明列附子反半夏/瓜蒌等，安全核查按通行扩展执行"},
+        {"组": ["甘草"], "组_通行扩展": [], "反": ["海藻", "大戟", "甘遂", "芫花"], "反_通行扩展": [],
+         "说明": None},
+        {"组": ["藜芦"], "组_通行扩展": [],
+         "反": ["人参", "沙参", "细辛", "芍药"], "反_通行扩展": ["丹参", "玄参", "苦参"],
+         "说明": "书列'诸参'散文明列人参、沙参；通行本扩展丹参/玄参/苦参"}],
+    "歌诀原文": re.search(r'十八反歌：[^\n]+', ph).group(0),
+    "散文原文": seg18_prose.strip().split("\n")[0],
+    "行号": ln(ph_i + i18),
+    "通行扩展说明": "歌诀'诸参辛芍反藜芦'，通行本藜芦组扩展含丹参/玄参/苦参；乌头组扩展含附子。扩展项在核查结果中标注[通行扩展]"},
+    "十九畏": {
+        "畏对": pairs,
+        "散文原文": seg19_prose.strip().split("\n")[0],
+        "行号": ln(ph_i + i19)},
+    "妊娠禁忌": {"药": [], "原文摘录": seg_preg.strip()[:400], "行号": ln(ph_i + ph.find("此外，妊娠禁忌"))}}
+for m in re.finditer(r'(植物|动物|矿物)药如([^；。]+)', seg_preg):
+    for x in re.split(r'[、]', m.group(2)):
+        x = x.strip()
+        if 2 <= len(x) <= 6: safety["妊娠禁忌"]["药"].append(x)
+lib["用药安全"] = safety
+
+# ============ 输出 ============
+json.dump(lib, open(OUT, "w", encoding="utf-8"), ensure_ascii=False, indent=1)
+print("理论依据库:", OUT, f"({os.path.getsize(OUT)} bytes)")
+print(f"八法 {len(bafa)} | 治法 {len(zhifa)}(缺号{gaps}) | 七方 {len(qifang)} | 效能 {len(xiaoneng)} | "
+      f"药组 {len(groups)} | 示例方 {len(shili)} | 脉 {len(mai)} | 十八反组3 | 十九畏 {len(safety['十九畏']['畏对'])}对 | 妊娠禁忌 {len(safety['妊娠禁忌']['药'])}味")

中医入门资料库/05_脚本工具/reverse_tag_dayi_with_theory.py

@@ -0,0 +1,233 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+中医入门资料库 03_关联融合（反向标注模式）
+方向: 大医网实体(方剂1000/药材1000) → 中医入门理论锚点（用户定级: 重要理论基础）
+理论锚点:
+  药材侧: 气味效能15分类 + 扶正7系38条/祛邪18条功效药组 + 十八反/十九畏/妊娠禁忌
+  方剂侧: 七方(7类) + 剂型(丸散膏丹等) + 40首示例方 + 八法/72治法提及方例
+匹配: 每个大医网药材查询自己在书侧理论锚点中的出现位置（正名/别名/炮制归一）
+      每个大医网方剂查询: 书侧40示例方同名 + 组成药味触发的功效组归属 + 名称含"汤散丸膏"剂型线索
+输出: 03_关联融合/大医网-中医入门_理论锚点反向标注.json
+"""
+import json, os, re, datetime, collections
+
+BASE   = "/home/songyi/Documents/ai_agent_scraper_study/data"
+DY     = os.path.join(BASE, "大医网/01_来源数据")
+BOOK   = os.path.join(BASE, "中医入门资料库/02_加工数据/中医入门.md")
+OUTDIR = os.path.join(BASE, "中医入门资料库/03_关联融合")
+TODAY  = datetime.date.today().isoformat()
+
+# 炮制前缀（候选侧可安全剥离）；产地/品质限定词（土/川/广/云/怀/杭/紫/白等）不剥——
+# "土人参/土白术"等是独立民间药名，剥离会造成假阳性（白木耳式陷阱的镜像）
+PREP_PREFIXES = sorted(["生晒","炙","煨","煅","焦","炒","制","飞","清","法","霜","酒",
+                        "醋","盐","蜜","麸","鲜","干","姜","蒸","煮","泡","燀"], key=len, reverse=True)
+BOOK_ALIAS = {  # 书侧简写 → 药典正名
+    "山萸":"山茱萸","萸肉":"山茱萸","麻仁":"火麻仁","云苓":"茯苓","赤苓":"茯苓",
+    "云茯神":"茯神","银花":"金银花","杭菊花":"菊花","花粉":"天花粉","寸冬":"麦冬",
+    "生地":"生地黄","熟地":"熟地黄","鲜生地":"生地黄","肉果":"肉豆蔻","煨肉果":"肉豆蔻",
+    "淡苁蓉":"肉苁蓉","吴萸":"吴茱萸","川连":"黄连","潼沙苑":"沙苑子","仙灵脾":"淫羊藿",
+    "旱莲草":"墨旱莲","丹皮":"牡丹皮","茅根":"白茅根","鲜茅根":"白茅根","象贝母":"浙贝母",
+    "象贝":"浙贝母","川贝":"川贝母","紫河车":"紫河车","白蒺藜":"刺蒺藜","菟丝饼":"菟丝子",
+    "桑椹子":"桑椹","使君肉":"使君子","山查炭":"山楂","焦麦芽":"麦芽","炒莱菔子":"莱菔子",
+    "驴皮胶":"阿胶","龟版":"龟板","菟蓉":"肉苁蓉","金铃子":"川楝子","延胡":"延胡索",
+    "参三七":"三七","全当归":"当归","当归身":"当归","冬虫草":"冬虫夏草","蝉衣":"蝉蜕",
+    "瓜蒌仁":"瓜蒌子","瓜蒌":"瓜蒌","益智仁":"益智仁","左牡蛎":"牡蛎","生石决":"石决明",
+    "羚羊尖":"羚羊角","炙全蝎":"全蝎","炒谷芽":"谷芽","春砂仁":"砂仁","白蔻仁":"豆蔻",
+    "御米壳":"罂粟壳","煨葛根":"葛根","炙升麻":"升麻","软柴胡":"柴胡","煨草果":"草果",
+    "炒苡仁":"薏苡仁","饭赤豆":"赤小豆","白藓皮":"白鲜皮","炙兜铃":"马兜铃","淡豆豉":"淡豆豉",
+    "黑山栀":"栀子","炒竹茹":"竹茹","炙桑皮":"桑白皮","甜杏仁":"甜杏仁","苦杏仁":"苦杏仁",
+    "炒牛蒡":"牛蒡子","苦桔梗":"桔梗","石草":"石韦","淡菓蓉":"肉苁蓉","复盆子":"覆盆子",
+    "菖蒲":"石菖蒲","萹蓄草":"萹蓄","海金砂":"海金沙","杜仲":"杜仲","虎骨":"虎骨",
+}
+def norm(t):  # 仅剥炮制前缀 + 别名归一（不剥产地/品质限定词）
+    c = t.strip()
+    for p in PREP_PREFIXES:
+        if c.startswith(p) and len(c) > len(p)+1:
+            c = c[len(p):]; break
+    return BOOK_ALIAS.get(c, c)
+
+text = open(BOOK, encoding="utf-8").read()
+
+# ============ 构建书侧理论锚点 ============
+# A. 药材理论锚点: 效能15分类 + 扶正/祛邪功效药组 + 用药安全
+anchors_herb = collections.defaultdict(lambda: {"效能分类": [], "功效药组": [], "用药安全": []})
+def add(anchor, herb_raw, kind, label):
+    n = norm(herb_raw)
+    if len(n) >= 2: anchors_herb[n][kind].append(label)
+    if n != herb_raw and len(herb_raw) >= 2:  # 炮制品/原名也留痕
+        anchors_herb[herb_raw][kind].append(label + "(原词)")
+
+# A1. 效能15分类
+xn = text[text.find("二、效能"):text.find("#### 1. 扶正类")]
+for cat_m in re.finditer(r'^(\d+)\.\s*([^\s：:，。]{2,8}药)：(.+)$', xn, re.M):
+    cat = cat_m.group(2)
+    for m in re.finditer(r'([\u4e00-\u9fff、]{2,24})如([、\u4e00-\u9fff（）a-zA-Z0-9]{2,60})', cat_m.group(3)):
+        for h in re.split(r'[、]', m.group(2)):
+            h = re.sub(r'（[^）]*）', '', h).strip("等")
+            if h: add(None, h, "效能分类", cat)
+
+# A2. 扶正/祛邪功效药组
+def parse_groups(start, end, kind):
+    seg = text[text.find(start):text.find(end)]
+    cur_sys = ""
+    for line in seg.splitlines():
+        s = line.strip()
+        m_sys = re.match(r'^\((\d+)\)\s*属于(.+?)：', s)
+        if m_sys: cur_sys = m_sys.group(2); continue
+        if not re.match(r'^[^#\d(].*——', s): continue
+        eff, _, rhs = s.partition("——")
+        label = f"{kind}·{eff.strip()}"
+        if cur_sys: label = f"{kind}·{cur_sys}·{eff.strip()}"
+        for h in re.split(r'[、，,]', rhs.strip().strip("。")):
+            h = re.sub(r'（[^）]*）', '', h).strip()
+            h = re.sub(r'^(亦可用鲜|亦可用)', '', h).strip()
+            if h: add(None, h, "功效药组", label)
+parse_groups("#### 1. 扶正类", "#### 2. 祛邪类", "扶正类")
+parse_groups("#### 2. 祛邪类", "三、归经", "祛邪类")
+
+# A3. 用药安全: 十八反/十九畏/妊娠禁忌
+ph = text[text.find("一、配合"):text.find("二、用量")]
+i = ph.find("歌中所提十八种药")
+seg18 = ph[i:ph.find("歌中所提十九种药")]  # 不能用"十九畏歌"作边界——开篇引言句已含该词
+for h in ["半夏","瓜蒌","贝母","白蔹","白及","海藻","大戟","甘遂","芫花","甘草","人参","沙参","细辛","芍药","藜芦","乌头"]:
+    if h in seg18: anchors_herb[h]["用药安全"].append("十八反")
+i = ph.find("歌中所提十九种药")
+seg19 = ph[i:ph.find("此外，妊娠禁忌")]
+for pair in re.findall(r'([\u4e00-\u9fff]{1,4})畏([\u4e00-\u9fff]{1,5})', seg19):
+    for h in pair:
+        if len(h) >= 2:
+            anchors_herb[h]["用药安全"].append("十九畏")
+if "川乌" in seg19:  # "川乌、草乌畏犀角"正则只捕到草乌，补川乌
+    anchors_herb["川乌"]["用药安全"].append("十九畏")
+segpreg = ph[ph.find("此外，妊娠禁忌"):ph.find("经验告诉我们")]
+for h in re.findall(r'如([^；。]{2,200}?)[；。]', segpreg):
+    for x in re.split(r'[、]', h):
+        x = x.strip()
+        if 2 <= len(x) <= 5: anchors_herb[x]["用药安全"].append("妊娠禁忌")
+
+# B. 方剂理论锚点: 七方/剂型/40示例方/治法方例
+anchors_fang = {"七方举例": {}, "示例方": {}, "剂型": "丸散膏丹酒汤等"}
+qf = text[text.find("二、七方"):text.find("三、剂型")]
+cur = None
+for para in re.split(r'\n\n+', qf):
+    m = re.match(r'^\d+[\.、]\s*([^\s：:，。]{1,3})方', para.strip())
+    if m: cur = m.group(1) + "方"
+    elif cur:
+        for f in re.findall(r'([\u4e00-\u9fff]{2,8}(?:汤|丸|散|饮|膏))', para):
+            anchors_fang["七方举例"].setdefault(f, []).append(cur)
+sec = text[text.find("一、基本方剂"):text.find("二、处方举例")]
+for n, name, rest in re.findall(r'^(\d+)\.\s*([^：:\n]+)[：:]\s*(.+)$', sec, re.M):
+    anchors_fang["示例方"][name.strip()] = {
+        "序号": int(n), "组成_书": [c.strip() for c in rest.split("。")[0].split("、") if c.strip()],
+        "定位": (re.search(r'为([^，。]{2,12})主方', rest).group(1) if re.search(r'为([^，。]{2,12})主方', rest) else None),
+        "主治": (rest.split("。")[1].replace("用于","").strip() if len(re.findall(r'。', rest))>1 and "用于" in rest else None)}
+# 72治法中的方药举例 → 治法标签（终点用"## 第三章"，因"## 第一节"在第一章也出现）
+zf = text[text.find("四、常用治法"):text.find("## 第三章")]
+zhifa_herb = collections.defaultdict(list)
+for n, name, rest in re.findall(r'^(\d+)\.\s*([^\s：:，。]{2,12}法)：?(.*)$', zf, re.M):
+    m = re.search(r'药如([、\u4e00-\u9fff（）]{2,50})', rest)
+    if m:
+        for h in re.split(r'[、]', m.group(1)):
+            h = re.sub(r'（[^）]*）','',h).strip()
+            if h: zhifa_herb[norm(h)].append(f"治法{n}.{name}")
+print(f"理论锚点: 归一药材 {len(anchors_herb)} 个 | 示例方 {len(anchors_fang['示例方'])} | 七方举例 {len(anchors_fang['七方举例'])} | 治法关联药 {len(zhifa_herb)}")
+
+# ============ 大医网药材 → 理论锚点 ============
+yao_out = []
+for fn in sorted(os.listdir(os.path.join(DY, "中药材"))):
+    d = json.load(open(os.path.join(DY, "中药材", fn), encoding="utf-8"))
+    nm = (d.get("名称") or "").strip()
+    cands = {nm}
+    for a in re.split(r'[、，,;；]', d.get("别名") or ""):
+        a = a.strip()
+        if len(a) >= 2: cands.add(a)
+    hits = []
+    for c in list(cands):
+        cn = norm(c)
+        for key in {c, cn}:
+            if key in anchors_herb:
+                v = anchors_herb[key]
+                kinds = [k for k in ("效能分类","功效药组","用药安全") if v[k]]
+                hits.append({"命中名": key, "书侧原词": list({c} | ({cn} if cn!=c else set())) and (c if key==c else cn),
+                             "类型": kinds,
+                             "效能分类": v["效能分类"], "功效药组": v["功效药组"][:8],
+                             "用药安全": v["用药安全"],
+                             "治法关联": zhifa_herb.get(key, [])[:6]})
+    seen, uniq = set(), []
+    for h in hits:
+        sig = h["命中名"]
+        if sig not in seen: seen.add(sig); uniq.append(h)
+    yao_out.append({
+        "大医网药材": nm, "id": fn.split("_")[0],
+        "别名": [a for a in cands if a != nm][:12],
+        "理论锚点": uniq,
+        "锚点类型": sorted({k for h in uniq for k in h["类型"]}),
+        "关联强度": len(uniq)})
+covered = sum(1 for y in yao_out if y["理论锚点"])
+print(f"药材反向标注: {covered}/1000 有理论锚点 ({covered//10}%)")
+
+# ============ 大医网方剂 → 理论锚点 ============
+fang_out = []
+for fn in sorted(os.listdir(os.path.join(DY, "方剂"))):
+    d = json.load(open(os.path.join(DY, "方剂", fn), encoding="utf-8"))
+    nm = (d.get("名称") or "").strip()
+    nm_clean = re.sub(r'[《》].*$', '', nm).strip()  # 去出处后缀
+    rec = {"大医网方剂": nm, "id": fn.split("_")[0], "理论锚点": []}
+    # B1 40示例方同名
+    if nm_clean in anchors_fang["示例方"]:
+        s = anchors_fang["示例方"][nm_clean]
+        rec["理论锚点"].append({"类型": ["示例方"], "示例方": nm_clean, "书侧序号": s["序号"],
+                                 "书侧组成": s["组成_书"], "主方定位": s["定位"], "主治": s["主治"]})
+    # B2 七方举例
+    if nm_clean in anchors_fang["七方举例"]:
+        rec["理论锚点"].append({"类型": ["七方举例"], "归属": anchors_fang["七方举例"][nm_clean]})
+    # B3 剂型线索（名称后缀 → 剂型理论）
+    m = re.search(r'(丸|散|膏|丹|酒|汤|饮)', nm)
+    if m:
+        rec["理论锚点"].append({"类型": ["剂型理论"], "剂型": m.group(1),
+                                 "理论": "丸散膏丹酒汤等剂型各具性质效用（中医入门·方剂之部·剂型）"})
+    # B4 组成药味 → 功效药组/治法推断（方解-组成仅5%填充，填充者才可算）
+    comp = d.get("方解-组成")
+    if comp:
+        herhs = [norm(x) for x in re.split(r'[、，,]', str(comp)) if x.strip()]
+        groups = collections.Counter()
+        zhifa = []
+        for h in herhs:
+            if h in anchors_herb:
+                for g in anchors_herb[h]["功效药组"]: groups[g] += 1
+            zhifa += zhifa_herb.get(h, [])
+        if groups:
+            rec["理论锚点"].append({"类型": ["功效药组推断"], "top功效组": [g for g,_ in groups.most_common(5)],
+                                     "组内命中": sum(1 for h in herhs if h in anchors_herb), "组成总数": len(herhs)})
+        if zhifa:
+            rec["理论锚点"].append({"类型": ["治法关联"], "治法": sorted(set(zhifa))[:8]})
+    rec["锚点类型"] = sorted({t for a in rec["理论锚点"] for t in a["类型"]})
+    rec["关联强度"] = len(rec["理论锚点"])
+    fang_out.append(rec)
+n_any = sum(1 for f in fang_out if f["理论锚点"])
+n_shili = sum(1 for f in fang_out if "示例方" in f["锚点类型"])
+n_qifang = sum(1 for f in fang_out if "七方举例" in f["锚点类型"])
+n_subst = sum(1 for f in fang_out if set(f["锚点类型"]) - {"剂型理论"})
+n_only_jx = sum(1 for f in fang_out if f["锚点类型"] == ["剂型理论"])
+print(f"方剂反向标注: {n_any}/1000 有锚点 | 实质锚点(示例方/七方/功效组/治法) {n_subst} | 仅剂型线索 {n_only_jx} | 示例方同名 {n_shili} | 七方举例 {n_qifang}")
+
+# ============ 输出 ============
+result = {
+ "metadata": {
+   "timestamp": TODAY,
+   "方向": "大医网实体 → 中医入门理论锚点（反向标注；中医入门为理论依据层，用户定级: 重要理论基础）",
+   "source_theory": "中医入门资料库/02_加工数据/中医入门.md（秦伯未《中医入门》体例）",
+   "source_entity": "大医网/01_来源数据/{方剂1000, 中药材1000}（随机抽样库）",
+   "锚点体系": {"药材": ["效能15分类", "扶正7系+祛邪功效药组", "十八反/十九畏/妊娠禁忌", "72治法药例"],
+               "方剂": ["40首示例方(含主方定位)", "七方举例", "剂型理论", "组成→功效组/治法推断"]},
+   "说明": "大医网为随机抽样1000，书侧示例方未在其中属正常（理论层不依赖抽样覆盖）；药材锚点覆盖率反映抽样命中理论药组的比例",
+   "stats": {"药材有锚点": covered, "药材总数": len(yao_out),
+             "方剂有锚点": n_any, "方剂实质锚点": n_subst, "方剂仅剂型": n_only_jx,
+             "方剂示例方同名": n_shili, "方剂七方举例": n_qifang}},
+ "药材反向标注": yao_out,
+ "方剂反向标注": fang_out,
+}
+out = os.path.join(OUTDIR, "大医网-中医入门_理论锚点反向标注.json")
+json.dump(result, open(out, "w", encoding="utf-8"), ensure_ascii=False, indent=1)
+print("\n输出:", out, f"({os.path.getsize(out)} bytes)")

中医入门资料库/05_脚本工具/theory_api.py

@@ -0,0 +1,215 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+《中医入门》理论依据查询接口（应用调用层）
+
+用法:
+  from theory_api import TCMTheory
+  t = TCMTheory()
+  t.herb("黄芪")            # 药材理论依据: 效能分类/功效药组/治法/用药安全
+  t.formula("大承气汤")      # 方剂理论依据: 主方定位/主治/七方归属/组成药组
+  t.safety_check(["甘草","大戟"])  # 配伍安全核查: 十八反/十九畏/妊娠禁忌
+  t.method("补中益气")       # 治法查询(名称模糊)
+  t.constitution("气虚")     # 体质调理依据(按气虚/阳虚/阴虚等关键词聚合)
+
+依据均含 行号(02_加工数据/中医入门.md) 可人工溯源。
+"""
+import json, os, re
+
+LIB  = os.path.join(os.path.dirname(__file__), "../02_加工数据/理论依据库.json")
+BOOK = os.path.join(os.path.dirname(__file__), "../02_加工数据/中医入门.md")
+
+# 书侧简写 → 正名（与 reverse_tag 脚本一致的归一）
+BOOK_ALIAS = {"山萸":"山茱萸","萸肉":"山茱萸","麻仁":"火麻仁","云苓":"茯苓","赤苓":"茯苓",
+    "云茯神":"茯神","银花":"金银花","杭菊花":"菊花","花粉":"天花粉","寸冬":"麦冬",
+    "生地":"生地黄","熟地":"熟地黄","鲜生地":"生地黄","肉果":"肉豆蔻","淡苁蓉":"肉苁蓉",
+    "吴萸":"吴茱萸","川连":"黄连","潼沙苑":"沙苑子","仙灵脾":"淫羊藿","旱莲草":"墨旱莲",
+    "丹皮":"牡丹皮","茅根":"白茅根","象贝母":"浙贝母","川贝":"川贝母","白蒺藜":"刺蒺藜",
+    "菟丝饼":"菟丝子","桑椹子":"桑椹","使君肉":"使君子","山查炭":"山楂","焦麦芽":"麦芽",
+    "炒莱菔子":"莱菔子","驴皮胶":"阿胶","龟版":"龟板","菟蓉":"肉苁蓉","金铃子":"川楝子",
+    "延胡":"延胡索","参三七":"三七","全当归":"当归","当归身":"当归","冬虫草":"冬虫夏草",
+    "蝉衣":"蝉蜕","瓜蒌仁":"瓜蒌子","左牡蛎":"牡蛎","生石决":"石决明","羚羊尖":"羚羊角",
+    "炙全蝎":"全蝎","炒谷芽":"谷芽","春砂仁":"砂仁","白蔻仁":"豆蔻","御米壳":"罂粟壳",
+    "煨葛根":"葛根","炙升麻":"升麻","软柴胡":"柴胡","煨草果":"草果","炒苡仁":"薏苡仁",
+    "饭赤豆":"赤小豆","白藓皮":"白鲜皮","炙兜铃":"马兜铃","黑山栀":"栀子","炒竹茹":"竹茹",
+    "炙桑皮":"桑白皮","炒牛蒡":"牛蒡子","苦桔梗":"桔梗","石草":"石韦","淡菓蓉":"肉苁蓉",
+    "复盆子":"覆盆子","菖蒲":"石菖蒲","萹蓄草":"萹蓄","海金砂":"海金沙",}
+PREP_PREFIX = ["生晒","炙","煨","煅","焦","炒","制","飞","清","法","霜","酒","醋","盐",
+               "蜜","麸","鲜","干","姜","蒸","煮"]  # 只剥炮制前缀；产地品质限定词不剥(土人参陷阱)
+
+def _norm(name: str):
+    c = name.strip()
+    for p in PREP_PREFIX:
+        if c.startswith(p) and len(c) > len(p) + 1:
+            c = c[len(p):]; break
+    return BOOK_ALIAS.get(c, c)
+
+class TCMTheory:
+    def __init__(self):
+        self.lib = json.load(open(LIB, encoding="utf-8"))
+        self.book_lines = open(BOOK, encoding="utf-8").read().splitlines()
+        # 索引: 药名(正名+别名+炮制品) → 锚点
+        self.herb_index = {}
+        for cat, info in self.lib["效能分类"].items():
+            for h in info["代表药"]:
+                self.herb_index.setdefault(_norm(h), []).append({"通道": "效能分类", "值": cat, "行号": info["行号"]})
+        for g in self.lib["功效药组"]["条目"]:
+            for h in g["药"]:
+                self.herb_index.setdefault(_norm(h), []).append({"通道": "功效药组", "值": g["组"], "行号": g["行号"]})
+        for num, z in self.lib["治法72"]["条目"].items():
+            for h in z["药如"]:
+                self.herb_index.setdefault(_norm(h), []).append({"通道": "治法", "值": f"治法{num}.{z['名称']}", "行号": z["行号"]})
+        self.safety_index = {}
+        for grp in self.lib["用药安全"]["十八反"]["相反组"]:
+            for a in grp["组"]:
+                self.safety_index.setdefault(a, []).append(("十八反", f'与{"、".join(grp["反"])}相反', "书列"))
+            for a in grp.get("组_通行扩展", []):
+                self.safety_index.setdefault(a, []).append(("十八反", f'与{"、".join(grp["反"])}相反', "通行扩展"))
+            for h in grp["反"]:
+                who = "、".join(grp["组"] + grp.get("组_通行扩展", []))
+                self.safety_index.setdefault(h, []).append(("十八反", f"反{who}", "书列"))
+            for h in grp.get("反_通行扩展", []):
+                who = "、".join(grp["组"] + grp.get("组_通行扩展", []))
+                self.safety_index.setdefault(h, []).append(("十八反", f"反{who}", "通行扩展"))
+        for a, b in self.lib["用药安全"]["十九畏"]["畏对"]:
+            self.safety_index.setdefault(a, []).append(("十九畏", f"畏{b}", "书列"))
+            self.safety_index.setdefault(b, []).append(("十九畏", f"为{a}所畏", "书列"))
+        for h in self.lib["用药安全"]["妊娠禁忌"]["药"]:
+            self.safety_index.setdefault(h, []).append(("妊娠禁忌", "妊娠禁忌药", "书列"))
+
+    def _quote(self, line_no, width=60):
+        if 1 <= line_no <= len(self.book_lines):
+            return self.book_lines[line_no - 1][:width]
+        return None
+
+    # ---------- 药材理论依据 ----------
+    def herb(self, name: str):
+        n = _norm(name)
+        hits = self.herb_index.get(n, []) or self.herb_index.get(name.strip(), [])
+        if not hits:
+            return {"药材": name, "结果": "《中医入门》理论药组未收录(该书为入门纲要，未收不等于无理论依据)"}
+        out = {"药材": name, "正名": n, "依据": []}
+        seen = set()
+        for h in hits:
+            key = (h["通道"], h["值"])
+            if key in seen: continue
+            seen.add(key)
+            out["依据"].append({**h, "原文": self._quote(h["行号"])})
+        out["用药安全"] = self._safety_of(n)
+        return out
+
+    # ---------- 方剂理论依据 ----------
+    def formula(self, name: str):
+        s = self.lib["示例方40"].get(name.strip())
+        if not s:
+            # 组成含书侧示例方药的模糊建议
+            return {"方剂": name, "结果": "非书侧40示例方；可按组成逐味调用 herb() 获取理论依据",
+                    "剂型理论": self._jixing_hint(name)}
+        out = {"方剂": name, "主方定位": s["主方定位"], "主治": s["主治"],
+               "组成_书": s["组成_书"], "行号": s["行号"], "原文": self._quote(s["行号"], 120),
+               "七方归属": [q for q, v in self.lib["七方"].items() if name in v["书内举例"]] or None,
+               "组成理论": []}
+        safety_all = []
+        for h in s["组成_书"]:
+            hb = self.herb(h)
+            ch = [f'{d["通道"]}:{d["值"]}' for d in hb.get("依据", [])][:3]
+            out["组成理论"].append({"药": h, "依据": ch})
+            if hb.get("用药安全"): safety_all.append({h: hb["用药安全"]})
+        if safety_all: out["组成用药安全"] = safety_all
+        return out
+
+    # ---------- 配伍安全核查 ----------
+    def safety_check(self, herbs):
+        herbs = [h.strip() for h in herbs if h.strip()]
+        conflicts, warnings = [], []
+        normalized = {h: _norm(h) for h in herbs}
+        names = list(normalized.items())
+        # 十八反: 组内对立（含通行扩展，结果标注来源）
+        for grp in self.lib["用药安全"]["十八反"]["相反组"]:
+            side_a = {(h, n) for h, n in names
+                      if n in grp["组"] or n in grp.get("组_通行扩展", [])}
+            side_b = {(h, n) for h, n in names
+                      if n in grp["反"] or n in grp.get("反_通行扩展", [])}
+            if side_a and side_b:
+                src = "书列"
+                a_ext = any(n in grp.get("组_通行扩展", []) for _, n in side_a)
+                b_ext = any(n in grp.get("反_通行扩展", []) for _, n in side_b)
+                if a_ext or b_ext: src = "通行扩展"
+                conflicts.append({"类型": "十八反", "来源": src,
+                                  "组": sorted(h for h, _ in side_a), "反": sorted(h for h, _ in side_b),
+                                  "行号": self.lib["用药安全"]["十八反"]["行号"]})
+        # 十九畏: 组内畏对
+        pairset = {frozenset(p) for p in self.lib["用药安全"]["十九畏"]["畏对"]}
+        for i in range(len(names)):
+            for j in range(i + 1, len(names)):
+                if frozenset([names[i][1], names[j][1]]) in pairset:
+                    warnings.append({"类型": "十九畏", "对": [names[i][0], names[j][0]],
+                                     "行号": self.lib["用药安全"]["十九畏"]["行号"]})
+        # 妊娠禁忌
+        preg = [h for h, n in names
+                if n in self.lib["用药安全"]["妊娠禁忌"]["药"] or h in self.lib["用药安全"]["妊娠禁忌"]["药"]]
+        return {"核查药味": herbs, "十八反冲突": conflicts or None,
+                "十九畏警式": warnings or None,
+                "妊娠禁忌药": preg or None,
+                "结论": "存在配伍禁忌，须调整" if conflicts else
+                        ("有相畏提示，慎用并权衡" if warnings else
+                         ("含妊娠禁忌药，孕妇须避" if preg else "书侧十八反/十九畏/妊娠禁忌未命中"))}
+
+    # ---------- 治法 ----------
+    def method(self, keyword: str):
+        out = []
+        for num, z in self.lib["治法72"]["条目"].items():
+            hay = z["名称"] + (z["适应症"] or "")
+            if keyword in z["名称"] or keyword in hay or z["名称"][:2] in keyword:
+                out.append({"治法": f"{num}.{z['名称']}", "适应症": z["适应症"],
+                            "药如": z["药如"], "行号": z["行号"],
+                            "原文": self._quote(z["行号"])})
+        for bn, bv in self.lib["八法"].items():
+            if keyword in bn or keyword in bv["定义原文"] or bn[0] in keyword:
+                out.append({"八法": bn, "定义": bv["定义原文"], "行号": bv["行号"]})
+        return {"关键词": keyword, "命中": out or None}
+
+    # ---------- 体质调理依据 ----------
+    CONSTITUTION_SYNONYMS = {
+        "气虚": ["补气", "中气"], "血虚": ["补血", "养血"], "阴虚": ["补阴", "滋阴", "养阴"],
+        "阳虚": ["补阳", "助阳", "温阳", "扶阳"], "痰湿": ["化湿", "化痰", "淡渗"],
+        "湿热": ["清湿热", "化湿", "利湿"], "血瘀": ["活血", "破瘀"], "气滞": ["行气", "破气", "疏肝"],
+        "寒": ["温中", "祛寒", "散寒", "扶阳"], "热": ["清热", "泻火", "凉血"], "食积": ["消食", "消导"]}
+    def constitution(self, keyword: str):
+        """按 气虚/血虚/阴虚/阳虚/痰湿/湿热/血瘀/气滞/寒/热/食积 等关键词聚合理论依据"""
+        syns = self.CONSTITUTION_SYNONYMS.get(keyword, [keyword])
+        res = {"关键词": keyword, "同义词展开": syns, "补益药组": [], "治法": [], "八法": []}
+        for g in self.lib["功效药组"]["条目"]:
+            if any(s in g["组"] for s in syns):
+                res["补益药组"].append({"组": g["组"], "药": g["药"], "行号": g["行号"]})
+        for num, z in self.lib["治法72"]["条目"].items():
+            hay = z["名称"] + (z["适应症"] or "")
+            if any(s in hay for s in syns):
+                res["治法"].append({"治法": f"{num}.{z['名称']}", "药如": z["药如"], "行号": z["行号"]})
+        for bn, bv in self.lib["八法"].items():
+            if any(s in bv["定义原文"] for s in syns):
+                res["八法"].append({"八法": bn, "行号": bv["行号"]})
+        res["提示"] = "示例方与药组的完整组成请调用 formula()/herb() 逐层获取" if (res["补益药组"] or res["治法"]) else None
+        return res
+
+    # ---------- 内部 ----------
+    def _safety_of(self, n):
+        recs = self.safety_index.get(n, [])
+        out = {}
+        for typ, desc, src in recs:
+            out.setdefault(typ, []).append(f"{desc}[{src}]" if src != "书列" else desc)
+        return out or None
+
+    def _jixing_hint(self, name):
+        m = re.search(r'(丸|散|膏|丹|酒|汤|饮)', name)
+        if not m: return None
+        jx = m.group(1)
+        return {"剂型": jx, "说明": "丸散膏丹等剂型各具性质效用（方剂之部·剂型）"}
+
+if __name__ == "__main__":
+    t = TCMTheory()
+    print("== herb 黄芪 ==");        print(json.dumps(t.herb("黄芪"), ensure_ascii=False, indent=1)[:600])
+    print("\n== formula 大承气汤 =="); print(json.dumps(t.formula("大承气汤"), ensure_ascii=False, indent=1)[:700])
+    print("\n== safety 甘草+大戟 =="); print(json.dumps(t.safety_check(["甘草", "大戟", "茯苓"]), ensure_ascii=False, indent=1))
+    print("\n== method 补中 ==");     print(json.dumps(t.method("补中"), ensure_ascii=False, indent=1)[:400])
+    print("\n== constitution 气虚 =="); print(json.dumps(t.constitution("气虚"), ensure_ascii=False, indent=1)[:500])