@tencent-rtc/mcp 1.4.6 → 1.6.0-beta.1

package/knowledge/bm25-cache.js

@@ -0,0 +1,74 @@
+import { attachPoolContent } from './content-loader.js';
+import { buildBM25Index, chunkSearchText } from './bm25.js';
+/**
+ * 按 product 维度 lazy 构建 BM25 索引并进程内缓存。
+ *
+ * 设计要点：
+ * - cache key = product（未指定时用 '*'，对全 manifest 建索引）
+ * - 索引基于该 product 全部 chunks（含 integration corpus），子集 ranking 只看 filteredPool.map(id)
+ * - 首次按 product 触发，后续命中缓存零开销；约 8 个 product 自然分摊冷启动成本
+ * - 不落盘，纯进程常驻；体积粗估 8~12MB（postings + idf）
+ */
+const indexCache = new Map();
+const GLOBAL_KEY = '*';
+function cacheKey(product) {
+    return product ?? GLOBAL_KEY;
+}
+function selectProductChunks(allChunks, product) {
+    if (!product)
+        return allChunks;
+    return allChunks.filter((chunk) => chunk.product === product || chunk.product === '*');
+}
+/**
+ * 获取 product 级 BM25 索引：缓存命中则零开销，否则按 product 切 chunks，
+ * attachPoolContent 拉正文后一次性建索引并入缓存。
+ *
+ * 性能特征：
+ * - 命中：O(1)
+ * - miss：O(P · avgDocLen) 仅一次，P 为该 product 的 chunk 数
+ */
+export function getProductBM25Index(allChunks, product) {
+    const key = cacheKey(product);
+    const cached = indexCache.get(key);
+    if (cached)
+        return cached;
+    const productChunks = selectProductChunks(allChunks, product);
+    const withContent = attachPoolContent(productChunks);
+    const index = buildBM25Index(withContent.map((chunk) => ({
+        id: chunk.id,
+        text: chunkSearchText(chunk),
+    })));
+    indexCache.set(key, index);
+    return index;
+}
+/**
+ * 多 product BM25 索引：products 数组中有任一未缓存时，
+ * 合并所有指定 product 的 chunks 构建索引并缓存（key = 排序 join）。
+ */
+export function getProductsBM25Index(allChunks, products) {
+    if (!products || products.length === 0)
+        return getProductBM25Index(allChunks, undefined);
+    if (products.length === 1)
+        return getProductBM25Index(allChunks, products[0]);
+    const key = [...products].sort().join('|');
+    const cached = indexCache.get(key);
+    if (cached)
+        return cached;
+    const productChunks = allChunks.filter((chunk) => products.includes(chunk.product) || chunk.product === '*');
+    const withContent = attachPoolContent(productChunks);
+    const index = buildBM25Index(withContent.map((chunk) => ({
+        id: chunk.id,
+        text: chunkSearchText(chunk),
+    })));
+    indexCache.set(key, index);
+    return index;
+}
+/** 测试用：清空缓存（生产代码不应调用） */
+export function clearBM25Cache() {
+    indexCache.clear();
+}
+/** 调试用：返回当前缓存的 product 列表 */
+export function listCachedProducts() {
+    return [...indexCache.keys()];
+}
+//# sourceMappingURL=bm25-cache.js.map

package/knowledge/bm25-cache.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"bm25-cache.js","sourceRoot":"","sources":["../../src/knowledge/bm25-cache.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,iBAAiB,EAAE,MAAM,qBAAqB,CAAC;AACxD,OAAO,EAAE,cAAc,EAAE,eAAe,EAAkB,MAAM,WAAW,CAAC;AAE5E;;;;;;;;GAQG;AACH,MAAM,UAAU,GAAG,IAAI,GAAG,EAAqB,CAAC;AAEhD,MAAM,UAAU,GAAG,GAAG,CAAC;AAEvB,SAAS,QAAQ,CAAC,OAAgB;IAChC,OAAO,OAAO,IAAI,UAAU,CAAC;AAC/B,CAAC;AAED,SAAS,mBAAmB,CAAC,SAA0B,EAAE,OAAgB;IACvE,IAAI,CAAC,OAAO;QAAE,OAAO,SAAS,CAAC;IAC/B,OAAO,SAAS,CAAC,MAAM,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,OAAO,KAAK,OAAO,IAAI,KAAK,CAAC,OAAO,KAAK,GAAG,CAAC,CAAC;AACzF,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,UAAU,mBAAmB,CAAC,SAA0B,EAAE,OAAgB;IAC9E,MAAM,GAAG,GAAG,QAAQ,CAAC,OAAO,CAAC,CAAC;IAC9B,MAAM,MAAM,GAAG,UAAU,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;IACnC,IAAI,MAAM;QAAE,OAAO,MAAM,CAAC;IAE1B,MAAM,aAAa,GAAG,mBAAmB,CAAC,SAAS,EAAE,OAAO,CAAC,CAAC;IAC9D,MAAM,WAAW,GAAG,iBAAiB,CAAC,aAAa,CAAC,CAAC;IACrD,MAAM,KAAK,GAAG,cAAc,CAC1B,WAAW,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;QAC1B,EAAE,EAAE,KAAK,CAAC,EAAE;QACZ,IAAI,EAAE,eAAe,CAAC,KAAK,CAAC;KAC7B,CAAC,CAAC,CACJ,CAAC;IACF,UAAU,CAAC,GAAG,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;IAC3B,OAAO,KAAK,CAAC;AACf,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,oBAAoB,CAAC,SAA0B,EAAE,QAAmB;IAClF,IAAI,CAAC,QAAQ,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,mBAAmB,CAAC,SAAS,EAAE,SAAS,CAAC,CAAC;IACzF,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,mBAAmB,CAAC,SAAS,EAAE,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC;IAE9E,MAAM,GAAG,GAAG,CAAC,GAAG,QAAQ,CAAC,CAAC,IAAI,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC3C,MAAM,MAAM,GAAG,UAAU,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;IACnC,IAAI,MAAM;QAAE,OAAO,MAAM,CAAC;IAE1B,MAAM,aAAa,GAAG,SAAS,CAAC,MAAM,CACpC,CAAC,KAAK,EAAE,EAAE,CAAC,QAAQ,CAAC,QAAQ,CAAC,KAAK,CAAC,OAAO,CAAC,IAAI,KAAK,CAAC,OAAO,KAAK,GAAG,CACrE,CAAC;IACF,MAAM,WAAW,GAAG,iBAAiB,CAAC,aAAa,CAAC,CAAC;IACrD,MAAM,KAAK,GAAG,cAAc,CAC1B,WAAW,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;QAC1B,EAAE,EAAE,KAAK,CAAC,EAAE;QACZ,IAAI,EAAE,eAAe,CAAC,KAAK,CAAC;KAC7B,CAAC,CAAC,CACJ,CAAC;IACF,UAAU,CAAC,GAAG,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;IAC3B,OAAO,KAAK,CAAC;AACf,CAAC;AAED,yBAAyB;AACzB,MAAM,UAAU,cAAc;IAC5B,UAAU,CAAC,KAAK,EAAE,CAAC;AACrB,CAAC;AAED,6BAA6B;AAC7B,MAAM,UAAU,kBAAkB;IAChC,OAAO,CAAC,GAAG,UAAU,CAAC,IAAI,EAAE,CAAC,CAAC;AAChC,CAAC"}

package/knowledge/bm25.js

@@ -0,0 +1,148 @@
+const K1 = 1.5;
+const B = 0.75;
+/**
+ * 中文停用词：高频功能词，参与 BM25 会稀释关键术语权重。
+ * 覆盖：疑问代词、助词、人称、连接词、模糊量词、敬辞。
+ */
+const CN_STOPWORDS = new Set([
+    '怎么', '如何', '什么', '哪些', '哪个', '为何', '为啥', '是否',
+    '可以', '能够', '需要', '想要', '应该', '请问', '麻烦',
+    '一下', '一个', '这个', '那个', '这些', '那些', '里面', '外面',
+    '使用', '用法', '用于', '用来', '进行', '实现', '提供', '支持',
+    '问题', '情况', '场景', '方式', '方法', '功能', '能力',
+    '我们', '你们', '他们', '自己', '大家',
+    '以及', '或者', '还是', '但是', '然后', '所以', '因为', '如果',
+    '已经', '正在', '马上', '立即', '总是', '一直',
+]);
+/** 英文停用词：限制在常见 query 噪声词，避免误删技术语义词（如 'use', 'set' 这种命名前缀保留） */
+const EN_STOPWORDS = new Set([
+    'a', 'an', 'the', 'is', 'are', 'was', 'were', 'be', 'been', 'being',
+    'and', 'or', 'but', 'if', 'then', 'so', 'as', 'of', 'at', 'by', 'for', 'in', 'on', 'to',
+    'do', 'does', 'did', 'have', 'has', 'had',
+    'this', 'that', 'these', 'those',
+    'how', 'what', 'why', 'when', 'where', 'which', 'who',
+    'i', 'we', 'you', 'they', 'it', 'me', 'us', 'them',
+    'pls', 'please', 'thanks',
+]);
+function isStopwordCJK(token) {
+    return CN_STOPWORDS.has(token);
+}
+function isStopwordEN(token) {
+    return EN_STOPWORDS.has(token);
+}
+export function tokenize(text) {
+    const tokens = [];
+    const lower = text.toLowerCase();
+    // 英文/数字 token：跳过停用词；保留长度 ≥ 2 的（避免单字符 'i', 'a' 漏入）
+    for (const match of lower.matchAll(/[a-z0-9_]+/g)) {
+        const tok = match[0];
+        if (tok.length === 1)
+            continue;
+        if (isStopwordEN(tok))
+            continue;
+        tokens.push(tok);
+    }
+    // 中文 token：bigram + unigram 两级
+    // - bigram：跳过包含停用词单字的组合（如「怎么调用」的「怎么」）以及任一位是常见停用字
+    // - unigram：跳过纯停用单字（用一个收敛的单字停用集合）
+    const cjk = text.replace(/[^\u4e00-\u9fff]/g, '');
+    for (let i = 0; i < cjk.length - 1; i += 1) {
+        const bigram = cjk.slice(i, i + 2);
+        if (isStopwordCJK(bigram))
+            continue;
+        tokens.push(bigram);
+    }
+    // 单字仅保留有信息量的（中文单字普遍信息量低，全部塞入会拉低 IDF；
+    // 简化策略：只在 bigram 路径之外补充长度 ≥ 3 的连续单字组合的首尾，避免噪声爆炸）
+    // 实践上保留 unigram 但不参与停用词过滤——让 BM25 的 IDF 自然衰减它们的权重。
+    for (let i = 0; i < cjk.length; i += 1) {
+        tokens.push(cjk[i]);
+    }
+    return tokens;
+}
+export function buildBM25Index(documents) {
+    const postings = {};
+    const docLengths = {};
+    const docFreq = {};
+    let totalLength = 0;
+    for (const doc of documents) {
+        const tokens = tokenize(doc.text);
+        docLengths[doc.id] = tokens.length;
+        totalLength += tokens.length;
+        const termFreq = {};
+        for (const token of tokens) {
+            termFreq[token] = (termFreq[token] ?? 0) + 1;
+        }
+        for (const [term, freq] of Object.entries(termFreq)) {
+            if (!postings[term])
+                postings[term] = {};
+            postings[term][doc.id] = freq;
+            docFreq[term] = (docFreq[term] ?? 0) + 1;
+        }
+    }
+    const n = documents.length;
+    const avgDocLength = n > 0 ? totalLength / n : 0;
+    const idf = {};
+    for (const [term, df] of Object.entries(docFreq)) {
+        idf[term] = Math.log(1 + (n - df + 0.5) / (df + 0.5));
+    }
+    return {
+        version: 1,
+        avgDocLength,
+        docLengths,
+        idf,
+        postings,
+    };
+}
+export function scoreBM25(query, chunkId, index) {
+    const tokens = tokenize(query);
+    const docLength = index.docLengths[chunkId] ?? 0;
+    if (docLength === 0)
+        return 0;
+    let score = 0;
+    for (const term of tokens) {
+        const tf = index.postings[term]?.[chunkId] ?? 0;
+        if (tf === 0)
+            continue;
+        const idf = index.idf[term] ?? 0;
+        const numerator = tf * (K1 + 1);
+        const denominator = tf + K1 * (1 - B + B * (docLength / (index.avgDocLength || 1)));
+        score += idf * (numerator / denominator);
+    }
+    return score;
+}
+export function rankBM25(query, chunkIds, index, boosts = {}) {
+    return chunkIds
+        .map((id) => ({
+        id,
+        score: scoreBM25(query, id, index) + (boosts[id] ?? 0),
+    }))
+        .sort((a, b) => b.score - a.score);
+}
+/**
+ * BM25 检索文本：title 重复 3 次以提权（标题语义浓度远高于正文，按业界惯例 boost）。
+ * symbols / error_codes 同样保留以便精确名匹配；正文走原内容。
+ */
+export function chunkSearchText(parts) {
+    const titleBoosted = [parts.title, parts.title, parts.title].filter(Boolean).join(' ');
+    return [titleBoosted, parts.content ?? '', ...parts.symbols, ...parts.error_codes].filter(Boolean).join(' ');
+}
+export function rankPoolBM25(query, pool, boosts = {}) {
+    if (pool.length === 0)
+        return [];
+    const index = buildBM25Index(pool.map((chunk) => ({
+        id: chunk.id,
+        text: chunkSearchText(chunk),
+    })));
+    return rankBM25(query, pool.map((chunk) => chunk.id), index, boosts);
+}
+/**
+ * 基于预构建的 BM25 索引对一个 chunk-id 子集打分，跳过 tokenize + IDF 重算。
+ * 调用方需保证 chunkIds ⊆ index 已收录的 id 集合，否则未收录的 id 会返回 score=0。
+ */
+export function rankSubsetWithIndex(query, chunkIds, index, boosts = {}) {
+    if (chunkIds.length === 0)
+        return [];
+    return rankBM25(query, chunkIds, index, boosts);
+}
+//# sourceMappingURL=bm25.js.map

package/knowledge/bm25.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"bm25.js","sourceRoot":"","sources":["../../src/knowledge/bm25.ts"],"names":[],"mappings":"AAQA,MAAM,EAAE,GAAG,GAAG,CAAC;AACf,MAAM,CAAC,GAAG,IAAI,CAAC;AAEf;;;GAGG;AACH,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC;IAC3B,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IAC9C,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IACxC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IAC9C,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IAC9C,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IACxC,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IAC5B,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IAC9C,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;CACnC,CAAC,CAAC;AAEH,+DAA+D;AAC/D,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC;IAC3B,GAAG,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO;IACnE,KAAK,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI;IACvF,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,KAAK;IACzC,MAAM,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO;IAChC,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,OAAO,EAAE,KAAK;IACrD,GAAG,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM;IAClD,KAAK,EAAE,QAAQ,EAAE,QAAQ;CAC1B,CAAC,CAAC;AAEH,SAAS,aAAa,CAAC,KAAa;IAClC,OAAO,YAAY,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;AACjC,CAAC;AAED,SAAS,YAAY,CAAC,KAAa;IACjC,OAAO,YAAY,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;AACjC,CAAC;AAED,MAAM,UAAU,QAAQ,CAAC,IAAY;IACnC,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;IAEjC,kDAAkD;IAClD,KAAK,MAAM,KAAK,IAAI,KAAK,CAAC,QAAQ,CAAC,aAAa,CAAC,EAAE,CAAC;QAClD,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACrB,IAAI,GAAG,CAAC,MAAM,KAAK,CAAC;YAAE,SAAS;QAC/B,IAAI,YAAY,CAAC,GAAG,CAAC;YAAE,SAAS;QAChC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACnB,CAAC;IAED,+BAA+B;IAC/B,iDAAiD;IACjD,kCAAkC;IAClC,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,mBAAmB,EAAE,EAAE,CAAC,CAAC;IAClD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;QAC3C,MAAM,MAAM,GAAG,GAAG,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC;QACnC,IAAI,aAAa,CAAC,MAAM,CAAC;YAAE,SAAS;QACpC,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IACtB,CAAC;IACD,qCAAqC;IACrC,iDAAiD;IACjD,mDAAmD;IACnD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,GAAG,CAAC,MAAM,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;QACvC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IACtB,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,SAA8C;IAC3E,MAAM,QAAQ,GAA2C,EAAE,CAAC;IAC5D,MAAM,UAAU,GAA2B,EAAE,CAAC;IAC9C,MAAM,OAAO,GAA2B,EAAE,CAAC;IAC3C,IAAI,WAAW,GAAG,CAAC,CAAC;IAEpB,KAAK,MAAM,GAAG,IAAI,SAAS,EAAE,CAAC;QAC5B,MAAM,MAAM,GAAG,QAAQ,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;QAClC,UAAU,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,CAAC;QACnC,WAAW,IAAI,MAAM,CAAC,MAAM,CAAC;QAC7B,MAAM,QAAQ,GAA2B,EAAE,CAAC;QAC5C,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;YAC3B,QAAQ,CAAC,KAAK,CAAC,GAAG,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC;QAC/C,CAAC;QACD,KAAK,MAAM,CAAC,IAAI,EAAE,IAAI,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,QAAQ,CAAC,EAAE,CAAC;YACpD,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC;gBAAE,QAAQ,CAAC,IAAI,CAAC,GAAG,EAAE,CAAC;YACzC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC;YAC9B,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC;QAC3C,CAAC;IACH,CAAC;IAED,MAAM,CAAC,GAAG,SAAS,CAAC,MAAM,CAAC;IAC3B,MAAM,YAAY,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACjD,MAAM,GAAG,GAA2B,EAAE,CAAC;IACvC,KAAK,MAAM,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,OAAO,CAAC,EAAE,CAAC;QACjD,GAAG,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,GAAG,GAAG,CAAC,GAAG,CAAC,EAAE,GAAG,GAAG,CAAC,CAAC,CAAC;IACxD,CAAC;IAED,OAAO;QACL,OAAO,EAAE,CAAC;QACV,YAAY;QACZ,UAAU;QACV,GAAG;QACH,QAAQ;KACT,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,SAAS,CAAC,KAAa,EAAE,OAAe,EAAE,KAAgB;IACxE,MAAM,MAAM,GAAG,QAAQ,CAAC,KAAK,CAAC,CAAC;IAC/B,MAAM,SAAS,GAAG,KAAK,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;IACjD,IAAI,SAAS,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IAC9B,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,KAAK,MAAM,IAAI,IAAI,MAAM,EAAE,CAAC;QAC1B,MAAM,EAAE,GAAG,KAAK,CAAC,QAAQ,CAAC,IAAI,CAAC,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;QAChD,IAAI,EAAE,KAAK,CAAC;YAAE,SAAS;QACvB,MAAM,GAAG,GAAG,KAAK,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACjC,MAAM,SAAS,GAAG,EAAE,GAAG,CAAC,EAAE,GAAG,CAAC,CAAC,CAAC;QAChC,MAAM,WAAW,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,SAAS,GAAG,CAAC,KAAK,CAAC,YAAY,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;QACpF,KAAK,IAAI,GAAG,GAAG,CAAC,SAAS,GAAG,WAAW,CAAC,CAAC;IAC3C,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,MAAM,UAAU,QAAQ,CACtB,KAAa,EACb,QAAkB,EAClB,KAAgB,EAChB,SAAiC,EAAE;IAEnC,OAAO,QAAQ;SACZ,GAAG,CAAC,CAAC,EAAE,EAAE,EAAE,CAAC,CAAC;QACZ,EAAE;QACF,KAAK,EAAE,SAAS,CAAC,KAAK,EAAE,EAAE,EAAE,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,CAAC,CAAC;KACvD,CAAC,CAAC;SACF,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;AACvC,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,eAAe,CAAC,KAK/B;IACC,MAAM,YAAY,GAAG,CAAC,KAAK,CAAC,KAAK,EAAE,KAAK,CAAC,KAAK,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACvF,OAAO,CAAC,YAAY,EAAE,KAAK,CAAC,OAAO,IAAI,EAAE,EAAE,GAAG,KAAK,CAAC,OAAO,EAAE,GAAG,KAAK,CAAC,WAAW,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;AAC/G,CAAC;AAED,MAAM,UAAU,YAAY,CAC1B,KAAa,EACb,IAAS,EACT,SAAiC,EAAE;IAEnC,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACjC,MAAM,KAAK,GAAG,cAAc,CAC1B,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;QACnB,EAAE,EAAE,KAAK,CAAC,EAAE;QACZ,IAAI,EAAE,eAAe,CAAC,KAAK,CAAC;KAC7B,CAAC,CAAC,CACJ,CAAC;IACF,OAAO,QAAQ,CACb,KAAK,EACL,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,EAAE,CAAC,EAC7B,KAAK,EACL,MAAM,CACP,CAAC;AACJ,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,mBAAmB,CACjC,KAAa,EACb,QAAkB,EAClB,KAAgB,EAChB,SAAiC,EAAE;IAEnC,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACrC,OAAO,QAAQ,CAAC,KAAK,EAAE,QAAQ,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC;AAClD,CAAC"}

package/knowledge/build-index.js

@@ -0,0 +1,70 @@
+import fs from 'fs';
+import path from 'path';
+import { fileURLToPath } from 'url';
+import { chunkMarkdown, toManifestChunk } from './chunker.js';
+import { buildDocSourceUrlMap } from './doc-url-resolver.js';
+import { inferPathMeta } from './path-meta.js';
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const RESOURCE_ROOT = path.resolve(__dirname, '../resource');
+function walkMarkdownFiles(dir, baseRel) {
+    if (!fs.existsSync(dir))
+        return [];
+    const files = [];
+    for (const entry of fs.readdirSync(dir, { withFileTypes: true })) {
+        const rel = `${baseRel}/${entry.name}`.replace(/\\/g, '/');
+        const abs = path.join(dir, entry.name);
+        if (entry.isDirectory()) {
+            if (entry.name === 'native_live')
+                continue;
+            files.push(...walkMarkdownFiles(abs, rel));
+        }
+        else if (entry.name.endsWith('.md') && entry.name !== 'workflow.md') {
+            files.push(rel);
+        }
+    }
+    return files;
+}
+function collectComponentSlugs(chunks) {
+    const slugs = {};
+    for (const chunk of chunks) {
+        const componentsMatch = chunk.source.match(/components\/([^/]+)\.md$/);
+        const chatFeaturesMatch = chunk.source.match(/^knowledge\/chat\/preset\/(vue|react)\/features\/([^/]+)\.md$/);
+        const slug = componentsMatch?.[1] ?? chatFeaturesMatch?.[2];
+        if (!slug)
+            continue;
+        slugs[slug] = [slug, slug.replace(/-/g, ''), slug.replace(/-/g, '_')];
+    }
+    return slugs;
+}
+export function buildKnowledgeIndex() {
+    const sources = [
+        ...walkMarkdownFiles(path.join(RESOURCE_ROOT, 'knowledge'), 'knowledge'),
+        ...walkMarkdownFiles(path.join(RESOURCE_ROOT, 'integration'), 'integration'),
+    ];
+    const chunks = [];
+    for (const source of sources) {
+        const abs = path.join(RESOURCE_ROOT, source);
+        const content = fs.readFileSync(abs, 'utf-8');
+        const meta = inferPathMeta(source);
+        chunks.push(...chunkMarkdown(source, content, meta).map(toManifestChunk));
+    }
+    const indexDir = path.join(RESOURCE_ROOT, 'index');
+    fs.mkdirSync(indexDir, { recursive: true });
+    const manifest = {
+        version: 1,
+        built_at: new Date().toISOString(),
+        chunks,
+    };
+    const componentSlugs = collectComponentSlugs(chunks);
+    fs.writeFileSync(path.join(indexDir, 'manifest.json'), JSON.stringify(manifest));
+    fs.writeFileSync(path.join(indexDir, 'component-slugs.json'), JSON.stringify(componentSlugs));
+    const syncManifestPath = path.resolve(__dirname, '../../sync/sync-manifest.json');
+    const docSourceUrls = buildDocSourceUrlMap(syncManifestPath);
+    fs.writeFileSync(path.join(indexDir, 'doc-source-urls.json'), JSON.stringify(docSourceUrls));
+    const staleSearchIndex = path.join(indexDir, 'search-index.json');
+    if (fs.existsSync(staleSearchIndex)) {
+        fs.unlinkSync(staleSearchIndex);
+    }
+    console.warn(`build-knowledge-index: ${chunks.length} chunks (manifest-lite) from ${sources.length} files`);
+}
+//# sourceMappingURL=build-index.js.map

package/knowledge/build-index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"build-index.js","sourceRoot":"","sources":["../../src/knowledge/build-index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,EAAE,aAAa,EAAE,MAAM,KAAK,CAAC;AACpC,OAAO,EAAE,aAAa,EAAE,eAAe,EAAsB,MAAM,cAAc,CAAC;AAClF,OAAO,EAAE,oBAAoB,EAAE,MAAM,uBAAuB,CAAC;AAC7D,OAAO,EAAE,aAAa,EAAE,MAAM,gBAAgB,CAAC;AAE/C,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;AAC/D,MAAM,aAAa,GAAG,IAAI,CAAC,OAAO,CAAC,SAAS,EAAE,aAAa,CAAC,CAAC;AAE7D,SAAS,iBAAiB,CAAC,GAAW,EAAE,OAAe;IACrD,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,GAAG,CAAC;QAAE,OAAO,EAAE,CAAC;IACnC,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,KAAK,MAAM,KAAK,IAAI,EAAE,CAAC,WAAW,CAAC,GAAG,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC;QACjE,MAAM,GAAG,GAAG,GAAG,OAAO,IAAI,KAAK,CAAC,IAAI,EAAE,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;QAC3D,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,CAAC,CAAC;QACvC,IAAI,KAAK,CAAC,WAAW,EAAE,EAAE,CAAC;YACxB,IAAI,KAAK,CAAC,IAAI,KAAK,aAAa;gBAAE,SAAS;YAC3C,KAAK,CAAC,IAAI,CAAC,GAAG,iBAAiB,CAAC,GAAG,EAAE,GAAG,CAAC,CAAC,CAAC;QAC7C,CAAC;aAAM,IAAI,KAAK,CAAC,IAAI,CAAC,QAAQ,CAAC,KAAK,CAAC,IAAI,KAAK,CAAC,IAAI,KAAK,aAAa,EAAE,CAAC;YACtE,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAClB,CAAC;IACH,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,qBAAqB,CAAC,MAAuB;IACpD,MAAM,KAAK,GAA6B,EAAE,CAAC;IAC3C,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,MAAM,eAAe,GAAG,KAAK,CAAC,MAAM,CAAC,KAAK,CAAC,0BAA0B,CAAC,CAAC;QACvE,MAAM,iBAAiB,GAAG,KAAK,CAAC,MAAM,CAAC,KAAK,CAC1C,+DAA+D,CAChE,CAAC;QACF,MAAM,IAAI,GAAG,eAAe,EAAE,CAAC,CAAC,CAAC,IAAI,iBAAiB,EAAE,CAAC,CAAC,CAAC,CAAC;QAC5D,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,EAAE,IAAI,CAAC,OAAO,CAAC,IAAI,EAAE,EAAE,CAAC,EAAE,IAAI,CAAC,OAAO,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,CAAC;IACxE,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,MAAM,UAAU,mBAAmB;IACjC,MAAM,OAAO,GAAG;QACd,GAAG,iBAAiB,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,WAAW,CAAC,EAAE,WAAW,CAAC;QACxE,GAAG,iBAAiB,CAAC,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,aAAa,CAAC,EAAE,aAAa,CAAC;KAC7E,CAAC;IAEF,MAAM,MAAM,GAAoB,EAAE,CAAC;IACnC,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;QAC7B,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,MAAM,CAAC,CAAC;QAC7C,MAAM,OAAO,GAAG,EAAE,CAAC,YAAY,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC;QAC9C,MAAM,IAAI,GAAG,aAAa,CAAC,MAAM,CAAC,CAAC;QACnC,MAAM,CAAC,IAAI,CAAC,GAAG,aAAa,CAAC,MAAM,EAAE,OAAO,EAAE,IAAI,CAAC,CAAC,GAAG,CAAC,eAAe,CAAC,CAAC,CAAC;IAC5E,CAAC;IAED,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,aAAa,EAAE,OAAO,CAAC,CAAC;IACnD,EAAE,CAAC,SAAS,CAAC,QAAQ,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAE5C,MAAM,QAAQ,GAAG;QACf,OAAO,EAAE,CAAC;QACV,QAAQ,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;QAClC,MAAM;KACP,CAAC;IAEF,MAAM,cAAc,GAAG,qBAAqB,CAAC,MAAM,CAAC,CAAC;IAErD,EAAE,CAAC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,eAAe,CAAC,EAAE,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC,CAAC;IACjF,EAAE,CAAC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,sBAAsB,CAAC,EAAE,IAAI,CAAC,SAAS,CAAC,cAAc,CAAC,CAAC,CAAC;IAE9F,MAAM,gBAAgB,GAAG,IAAI,CAAC,OAAO,CAAC,SAAS,EAAE,+BAA+B,CAAC,CAAC;IAClF,MAAM,aAAa,GAAG,oBAAoB,CAAC,gBAAgB,CAAC,CAAC;IAC7D,EAAE,CAAC,aAAa,CAAC,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,sBAAsB,CAAC,EAAE,IAAI,CAAC,SAAS,CAAC,aAAa,CAAC,CAAC,CAAC;IAE7F,MAAM,gBAAgB,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,mBAAmB,CAAC,CAAC;IAClE,IAAI,EAAE,CAAC,UAAU,CAAC,gBAAgB,CAAC,EAAE,CAAC;QACpC,EAAE,CAAC,UAAU,CAAC,gBAAgB,CAAC,CAAC;IAClC,CAAC;IAED,OAAO,CAAC,IAAI,CAAC,0BAA0B,MAAM,CAAC,MAAM,gCAAgC,OAAO,CAAC,MAAM,QAAQ,CAAC,CAAC;AAC9G,CAAC"}

package/knowledge/chunker.js

@@ -0,0 +1,168 @@
+import { extractErrorCodes, extractSymbols } from './extractors.js';
+function slugify(text) {
+    return text
+        .trim()
+        .toLowerCase()
+        .replace(/[^\w\u4e00-\u9fff]+/g, '-')
+        .replace(/^-+|-+$/g, '') || 'section';
+}
+function sourceSlug(source) {
+    return source.replace(/\.md$/, '').replace(/[/\\]/g, '-');
+}
+function trimSliceOffsets(raw, fileStart) {
+    const leading = raw.length - raw.trimStart().length;
+    const trailing = raw.length - raw.trimEnd().length;
+    const char_start = fileStart + leading;
+    const char_end = fileStart + raw.length - trailing;
+    return { body: raw.slice(leading, raw.length - trailing), char_start, char_end };
+}
+function splitByHeading(content, level) {
+    const regex = new RegExp(`^${'#'.repeat(level)}\\s+(.+)$`, 'gm');
+    const sections = [];
+    const matches = [...content.matchAll(regex)];
+    if (matches.length === 0) {
+        return [{ title: 'full', raw: content, start: 0 }];
+    }
+    for (let i = 0; i < matches.length; i += 1) {
+        const match = matches[i];
+        const start = match.index ?? 0;
+        const end = i + 1 < matches.length ? (matches[i + 1].index ?? content.length) : content.length;
+        sections.push({
+            title: match[1].trim(),
+            raw: content.slice(start, end),
+            start,
+        });
+    }
+    return sections;
+}
+function isFaqLike(source, content) {
+    return /faq|error-code|error_code|TXLiteAVCode/i.test(source) || (content.match(/^###\s+/gm) ?? []).length >= 3;
+}
+function isChatUrlMapping(source) {
+    return /knowledge\/chat\/urls\/.+\.md$/i.test(source.replace(/\\/g, '/'));
+}
+function parseUrlMappingRows(content) {
+    const lines = content.split('\n');
+    let sectionTitle = '';
+    let offset = 0;
+    const rows = [];
+    for (const line of lines) {
+        const lineStart = offset;
+        const lineEnd = offset + line.length;
+        const heading = line.match(/^#{2,3}\s+(.+)$/);
+        if (heading) {
+            sectionTitle = heading[1].trim();
+            offset = lineEnd + 1;
+            continue;
+        }
+        const rowMatch = line.match(/^\|\s*([^|]+?)\s*\|\s*`?(https?:\/\/[^|`\s]+)`?\s*\|/);
+        if (rowMatch) {
+            const label = rowMatch[1].trim();
+            if (label && label !== '子问题' && label !== 'Query' && label !== 'Topic' && !/^-+$/.test(label)) {
+                rows.push({
+                    label,
+                    url: rowMatch[2],
+                    sectionTitle,
+                    lineStart,
+                    lineEnd,
+                });
+            }
+        }
+        offset = lineEnd + 1;
+    }
+    return rows;
+}
+function chunkUrlMapping(source, content, meta) {
+    const rows = parseUrlMappingRows(content);
+    if (rows.length === 0) {
+        return [];
+    }
+    const baseSlug = sourceSlug(source);
+    const chunks = [];
+    for (const row of rows) {
+        const sectionLine = row.sectionTitle ? `### ${row.sectionTitle}\n\n` : '';
+        const body = `${sectionLine}| Query | Documentation |\n|-------|---------------|\n| ${row.label} | \`${row.url}\` |`;
+        const headingSlug = slugify(row.label);
+        chunks.push({
+            id: `${baseSlug}#${headingSlug}`,
+            corpus: meta.corpus,
+            product: meta.product,
+            framework: meta.framework,
+            types: meta.types,
+            scenario: meta.scenario,
+            title: row.label,
+            source,
+            content: body,
+            symbols: extractSymbols(body),
+            error_codes: extractErrorCodes(body),
+            char_start: row.lineStart,
+            char_end: row.lineEnd,
+        });
+    }
+    return chunks;
+}
+export function chunkMarkdown(source, content, meta) {
+    const chunks = [];
+    const baseSlug = sourceSlug(source);
+    if (isChatUrlMapping(source)) {
+        const urlChunks = chunkUrlMapping(source, content, meta);
+        if (urlChunks.length > 0) {
+            return urlChunks;
+        }
+    }
+    if (content.length < 2048) {
+        const { body, char_start, char_end } = trimSliceOffsets(content, 0);
+        if (!body)
+            return chunks;
+        chunks.push({
+            id: `${baseSlug}#full`,
+            corpus: meta.corpus,
+            product: meta.product,
+            framework: meta.framework,
+            types: meta.types,
+            scenario: meta.scenario,
+            title: pathBasename(source),
+            source,
+            content: body,
+            symbols: extractSymbols(body),
+            error_codes: extractErrorCodes(body),
+            char_start,
+            char_end,
+        });
+        return chunks;
+    }
+    const level = isFaqLike(source, content) ? 3 : 2;
+    const sections = splitByHeading(content, level);
+    for (const section of sections) {
+        const { body, char_start, char_end } = trimSliceOffsets(section.raw, section.start);
+        if (!body)
+            continue;
+        const headingSlug = slugify(section.title);
+        chunks.push({
+            id: `${baseSlug}#${headingSlug}`,
+            corpus: meta.corpus,
+            product: meta.product,
+            framework: meta.framework,
+            types: meta.types,
+            scenario: meta.scenario,
+            title: section.title,
+            source,
+            content: body,
+            symbols: extractSymbols(body),
+            error_codes: extractErrorCodes(body),
+            char_start,
+            char_end,
+        });
+    }
+    return chunks.length > 0 ? chunks : chunkMarkdown(source, content.slice(0, 2048), meta);
+}
+export function toManifestChunk(chunk) {
+    const { content: _content, ...manifestChunk } = chunk;
+    return manifestChunk;
+}
+function pathBasename(source) {
+    const parts = source.split('/');
+    const file = parts[parts.length - 1] ?? source;
+    return file.replace(/\.md$/, '');
+}
+//# sourceMappingURL=chunker.js.map

package/knowledge/chunker.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"chunker.js","sourceRoot":"","sources":["../../src/knowledge/chunker.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,iBAAiB,EAAE,cAAc,EAAE,MAAM,iBAAiB,CAAC;AAwBpE,SAAS,OAAO,CAAC,IAAY;IAC3B,OAAO,IAAI;SACR,IAAI,EAAE;SACN,WAAW,EAAE;SACb,OAAO,CAAC,sBAAsB,EAAE,GAAG,CAAC;SACpC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,IAAI,SAAS,CAAC;AAC1C,CAAC;AAED,SAAS,UAAU,CAAC,MAAc;IAChC,OAAO,MAAM,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;AAC5D,CAAC;AAED,SAAS,gBAAgB,CAAC,GAAW,EAAE,SAAiB;IACtD,MAAM,OAAO,GAAG,GAAG,CAAC,MAAM,GAAG,GAAG,CAAC,SAAS,EAAE,CAAC,MAAM,CAAC;IACpD,MAAM,QAAQ,GAAG,GAAG,CAAC,MAAM,GAAG,GAAG,CAAC,OAAO,EAAE,CAAC,MAAM,CAAC;IACnD,MAAM,UAAU,GAAG,SAAS,GAAG,OAAO,CAAC;IACvC,MAAM,QAAQ,GAAG,SAAS,GAAG,GAAG,CAAC,MAAM,GAAG,QAAQ,CAAC;IACnD,OAAO,EAAE,IAAI,EAAE,GAAG,CAAC,KAAK,CAAC,OAAO,EAAE,GAAG,CAAC,MAAM,GAAG,QAAQ,CAAC,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAC;AACnF,CAAC;AAED,SAAS,cAAc,CAAC,OAAe,EAAE,KAAa;IACpD,MAAM,KAAK,GAAG,IAAI,MAAM,CAAC,IAAI,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,WAAW,EAAE,IAAI,CAAC,CAAC;IACjE,MAAM,QAAQ,GAAyD,EAAE,CAAC;IAC1E,MAAM,OAAO,GAAG,CAAC,GAAG,OAAO,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC;IAC7C,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACzB,OAAO,CAAC,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAC;IACrD,CAAC;IACD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;QAC3C,MAAM,KAAK,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC;QACzB,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,IAAI,CAAC,CAAC;QAC/B,MAAM,GAAG,GAAG,CAAC,GAAG,CAAC,GAAG,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,KAAK,IAAI,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,CAAC;QAC/F,QAAQ,CAAC,IAAI,CAAC;YACZ,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE;YACtB,GAAG,EAAE,OAAO,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC;YAC9B,KAAK;SACN,CAAC,CAAC;IACL,CAAC;IACD,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,SAAS,SAAS,CAAC,MAAc,EAAE,OAAe;IAChD,OAAO,yCAAyC,CAAC,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,OAAO,CAAC,KAAK,CAAC,WAAW,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC;AAClH,CAAC;AAED,SAAS,gBAAgB,CAAC,MAAc;IACtC,OAAO,iCAAiC,CAAC,IAAI,CAAC,MAAM,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,CAAC;AAC5E,CAAC;AAUD,SAAS,mBAAmB,CAAC,OAAe;IAC1C,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAClC,IAAI,YAAY,GAAG,EAAE,CAAC;IACtB,IAAI,MAAM,GAAG,CAAC,CAAC;IACf,MAAM,IAAI,GAAoB,EAAE,CAAC;IAEjC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,SAAS,GAAG,MAAM,CAAC;QACzB,MAAM,OAAO,GAAG,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC;QAErC,MAAM,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,iBAAiB,CAAC,CAAC;QAC9C,IAAI,OAAO,EAAE,CAAC;YACZ,YAAY,GAAG,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YACjC,MAAM,GAAG,OAAO,GAAG,CAAC,CAAC;YACrB,SAAS;QACX,CAAC;QAED,MAAM,QAAQ,GAAG,IAAI,CAAC,KAAK,CAAC,sDAAsD,CAAC,CAAC;QACpF,IAAI,QAAQ,EAAE,CAAC;YACb,MAAM,KAAK,GAAG,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YACjC,IAAI,KAAK,IAAI,KAAK,KAAK,KAAK,IAAI,KAAK,KAAK,OAAO,IAAI,KAAK,KAAK,OAAO,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;gBAC9F,IAAI,CAAC,IAAI,CAAC;oBACR,KAAK;oBACL,GAAG,EAAE,QAAQ,CAAC,CAAC,CAAC;oBAChB,YAAY;oBACZ,SAAS;oBACT,OAAO;iBACR,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,MAAM,GAAG,OAAO,GAAG,CAAC,CAAC;IACvB,CAAC;IAED,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,eAAe,CAAC,MAAc,EAAE,OAAe,EAAE,IAAc;IACtE,MAAM,IAAI,GAAG,mBAAmB,CAAC,OAAO,CAAC,CAAC;IAC1C,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACtB,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,QAAQ,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC;IACpC,MAAM,MAAM,GAAiB,EAAE,CAAC;IAEhC,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;QACvB,MAAM,WAAW,GAAG,GAAG,CAAC,YAAY,CAAC,CAAC,CAAC,OAAO,GAAG,CAAC,YAAY,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;QAC1E,MAAM,IAAI,GAAG,GAAG,WAAW,2DAA2D,GAAG,CAAC,KAAK,QAAQ,GAAG,CAAC,GAAG,MAAM,CAAC;QACrH,MAAM,WAAW,GAAG,OAAO,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;QAEvC,MAAM,CAAC,IAAI,CAAC;YACV,EAAE,EAAE,GAAG,QAAQ,IAAI,WAAW,EAAE;YAChC,MAAM,EAAE,IAAI,CAAC,MAAM;YACnB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,KAAK,EAAE,IAAI,CAAC,KAAK;YACjB,QAAQ,EAAE,IAAI,CAAC,QAAQ;YACvB,KAAK,EAAE,GAAG,CAAC,KAAK;YAChB,MAAM;YACN,OAAO,EAAE,IAAI;YACb,OAAO,EAAE,cAAc,CAAC,IAAI,CAAC;YAC7B,WAAW,EAAE,iBAAiB,CAAC,IAAI,CAAC;YACpC,UAAU,EAAE,GAAG,CAAC,SAAS;YACzB,QAAQ,EAAE,GAAG,CAAC,OAAO;SACtB,CAAC,CAAC;IACL,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,MAAM,UAAU,aAAa,CAAC,MAAc,EAAE,OAAe,EAAE,IAAc;IAC3E,MAAM,MAAM,GAAiB,EAAE,CAAC;IAChC,MAAM,QAAQ,GAAG,UAAU,CAAC,MAAM,CAAC,CAAC;IAEpC,IAAI,gBAAgB,CAAC,MAAM,CAAC,EAAE,CAAC;QAC7B,MAAM,SAAS,GAAG,eAAe,CAAC,MAAM,EAAE,OAAO,EAAE,IAAI,CAAC,CAAC;QACzD,IAAI,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACzB,OAAO,SAAS,CAAC;QACnB,CAAC;IACH,CAAC;IAED,IAAI,OAAO,CAAC,MAAM,GAAG,IAAI,EAAE,CAAC;QAC1B,MAAM,EAAE,IAAI,EAAE,UAAU,EAAE,QAAQ,EAAE,GAAG,gBAAgB,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QACpE,IAAI,CAAC,IAAI;YAAE,OAAO,MAAM,CAAC;QACzB,MAAM,CAAC,IAAI,CAAC;YACV,EAAE,EAAE,GAAG,QAAQ,OAAO;YACtB,MAAM,EAAE,IAAI,CAAC,MAAM;YACnB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,KAAK,EAAE,IAAI,CAAC,KAAK;YACjB,QAAQ,EAAE,IAAI,CAAC,QAAQ;YACvB,KAAK,EAAE,YAAY,CAAC,MAAM,CAAC;YAC3B,MAAM;YACN,OAAO,EAAE,IAAI;YACb,OAAO,EAAE,cAAc,CAAC,IAAI,CAAC;YAC7B,WAAW,EAAE,iBAAiB,CAAC,IAAI,CAAC;YACpC,UAAU;YACV,QAAQ;SACT,CAAC,CAAC;QACH,OAAO,MAAM,CAAC;IAChB,CAAC;IAED,MAAM,KAAK,GAAG,SAAS,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACjD,MAAM,QAAQ,GAAG,cAAc,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC;IAChD,KAAK,MAAM,OAAO,IAAI,QAAQ,EAAE,CAAC;QAC/B,MAAM,EAAE,IAAI,EAAE,UAAU,EAAE,QAAQ,EAAE,GAAG,gBAAgB,CAAC,OAAO,CAAC,GAAG,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC;QACpF,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,MAAM,WAAW,GAAG,OAAO,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;QAC3C,MAAM,CAAC,IAAI,CAAC;YACV,EAAE,EAAE,GAAG,QAAQ,IAAI,WAAW,EAAE;YAChC,MAAM,EAAE,IAAI,CAAC,MAAM;YACnB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,KAAK,EAAE,IAAI,CAAC,KAAK;YACjB,QAAQ,EAAE,IAAI,CAAC,QAAQ;YACvB,KAAK,EAAE,OAAO,CAAC,KAAK;YACpB,MAAM;YACN,OAAO,EAAE,IAAI;YACb,OAAO,EAAE,cAAc,CAAC,IAAI,CAAC;YAC7B,WAAW,EAAE,iBAAiB,CAAC,IAAI,CAAC;YACpC,UAAU;YACV,QAAQ;SACT,CAAC,CAAC;IACL,CAAC;IAED,OAAO,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,aAAa,CAAC,MAAM,EAAE,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,EAAE,IAAI,CAAC,CAAC;AAC1F,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,KAAiB;IAC/C,MAAM,EACJ,OAAO,EAAE,QAAQ,EACjB,GAAG,aAAa,EACjB,GAAG,KAAK,CAAC;IACV,OAAO,aAAa,CAAC;AACvB,CAAC;AAED,SAAS,YAAY,CAAC,MAAc;IAClC,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;IAChC,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,IAAI,MAAM,CAAC;IAC/C,OAAO,IAAI,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,CAAC;AACnC,CAAC"}

package/knowledge/content-loader.js

@@ -0,0 +1,40 @@
+import fs from 'fs';
+import path from 'path';
+import { fileURLToPath } from 'url';
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
+const fileCache = new Map();
+function resolveResourceRoot() {
+    // npm: knowledge/ → ../resource/ 到包根；dev: dist/knowledge/ → ../resource/ 到 dist/resource/
+    return path.resolve(__dirname, '../resource');
+}
+function readSourceFile(source) {
+    const cached = fileCache.get(source);
+    if (cached !== undefined)
+        return cached;
+    const abs = path.join(resolveResourceRoot(), source);
+    const text = fs.readFileSync(abs, 'utf-8');
+    fileCache.set(source, text);
+    return text;
+}
+export function getSourceFileContent(source) {
+    return readSourceFile(source);
+}
+/** 按 build 时写入的 char 偏移从 md 源文件切片（方案 B：manifest 不存 content） */
+export function getChunkContent(chunk) {
+    const full = readSourceFile(chunk.source);
+    return full.slice(chunk.char_start, chunk.char_end);
+}
+export function attachChunkContent(chunk) {
+    return { ...chunk, content: getChunkContent(chunk) };
+}
+/** 预读 pool 涉及的唯一 source，再批量挂载 content（供 BM25 与 fragments 使用） */
+export function attachPoolContent(pool) {
+    for (const source of new Set(pool.map((chunk) => chunk.source))) {
+        readSourceFile(source);
+    }
+    return pool.map((chunk) => attachChunkContent(chunk));
+}
+export function clearContentCache() {
+    fileCache.clear();
+}
+//# sourceMappingURL=content-loader.js.map

package/knowledge/content-loader.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"content-loader.js","sourceRoot":"","sources":["../../src/knowledge/content-loader.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,IAAI,CAAC;AACpB,OAAO,IAAI,MAAM,MAAM,CAAC;AACxB,OAAO,EAAE,aAAa,EAAE,MAAM,KAAK,CAAC;AAGpC,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;AAE/D,MAAM,SAAS,GAAG,IAAI,GAAG,EAAkB,CAAC;AAE5C,SAAS,mBAAmB;IAC1B,0FAA0F;IAC1F,OAAO,IAAI,CAAC,OAAO,CAAC,SAAS,EAAE,aAAa,CAAC,CAAC;AAChD,CAAC;AAED,SAAS,cAAc,CAAC,MAAc;IACpC,MAAM,MAAM,GAAG,SAAS,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;IACrC,IAAI,MAAM,KAAK,SAAS;QAAE,OAAO,MAAM,CAAC;IACxC,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,CAAC,mBAAmB,EAAE,EAAE,MAAM,CAAC,CAAC;IACrD,MAAM,IAAI,GAAG,EAAE,CAAC,YAAY,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC;IAC3C,SAAS,CAAC,GAAG,CAAC,MAAM,EAAE,IAAI,CAAC,CAAC;IAC5B,OAAO,IAAI,CAAC;AACd,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAC,MAAc;IACjD,OAAO,cAAc,CAAC,MAAM,CAAC,CAAC;AAChC,CAAC;AAED,+DAA+D;AAC/D,MAAM,UAAU,eAAe,CAAC,KAAgE;IAC9F,MAAM,IAAI,GAAG,cAAc,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC;IAC1C,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,UAAU,EAAE,KAAK,CAAC,QAAQ,CAAC,CAAC;AACtD,CAAC;AAID,MAAM,UAAU,kBAAkB,CAAC,KAAoB;IACrD,OAAO,EAAE,GAAG,KAAK,EAAE,OAAO,EAAE,eAAe,CAAC,KAAK,CAAC,EAAE,CAAC;AACvD,CAAC;AAED,gEAAgE;AAChE,MAAM,UAAU,iBAAiB,CAAC,IAAqB;IACrD,KAAK,MAAM,MAAM,IAAI,IAAI,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,EAAE,CAAC;QAChE,cAAc,CAAC,MAAM,CAAC,CAAC;IACzB,CAAC;IACD,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,kBAAkB,CAAC,KAAK,CAAC,CAAC,CAAC;AACxD,CAAC;AAED,MAAM,UAAU,iBAAiB;IAC/B,SAAS,CAAC,KAAK,EAAE,CAAC;AACpB,CAAC"}