diffwiki-core 0.5.0 → 0.6.0-rc.202607240049.586dcb6

package/dist/bm25.cjs

@@ -0,0 +1,85 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/bm25.ts
+var bm25_exports = {};
+__export(bm25_exports, {
+  rankDocs: () => rankDocs,
+  tokenize: () => tokenize
+});
+module.exports = __toCommonJS(bm25_exports);
+var K1 = 1.5;
+var B = 0.75;
+var FIELD_WEIGHTS = { title: 3, tags: 2, collection: 1.5, body: 1 };
+function tokenize(text) {
+  return text.toLowerCase().split(/[^a-z0-9]+/).filter((t) => t.length > 0);
+}
+function termFreqs(tokens) {
+  const freqs = /* @__PURE__ */ new Map();
+  for (const t of tokens) freqs.set(t, (freqs.get(t) ?? 0) + 1);
+  return freqs;
+}
+function bm25FieldScore(queryTerms, fieldTokens, avgFieldLen, idf) {
+  const tf = termFreqs(fieldTokens);
+  const dl = fieldTokens.length;
+  const avgDl = avgFieldLen || 1;
+  let score = 0;
+  for (const term of queryTerms) {
+    const freq = tf.get(term) ?? 0;
+    if (freq === 0) continue;
+    const termIdf = idf.get(term) ?? 0;
+    const numerator = freq * (K1 + 1);
+    const denominator = freq + K1 * (1 - B + B * (dl / avgDl));
+    score += termIdf * (numerator / denominator);
+  }
+  return score;
+}
+function rankDocs(term, docs) {
+  const queryTerms = tokenize(term);
+  if (queryTerms.length === 0) return docs.map((d) => ({ ...d }));
+  const N = docs.length;
+  const docTokenSets = docs.map(
+    (d) => /* @__PURE__ */ new Set([...d.fields.title, ...d.fields.tags, ...d.fields.collection, ...d.fields.body])
+  );
+  const idf = /* @__PURE__ */ new Map();
+  for (const qt of queryTerms) {
+    let docFreq = 0;
+    for (const set of docTokenSets) if (set.has(qt)) docFreq++;
+    idf.set(qt, Math.log((N - docFreq + 0.5) / (docFreq + 0.5) + 1));
+  }
+  const avg = (sel) => docs.reduce((s, d) => s + sel(d), 0) / (N || 1);
+  const avgLens = {
+    title: avg((d) => d.fields.title.length),
+    tags: avg((d) => d.fields.tags.length),
+    collection: avg((d) => d.fields.collection.length),
+    body: avg((d) => d.fields.body.length)
+  };
+  const scored = [];
+  for (const d of docs) {
+    const total = bm25FieldScore(queryTerms, d.fields.title, avgLens.title, idf) * FIELD_WEIGHTS.title + bm25FieldScore(queryTerms, d.fields.tags, avgLens.tags, idf) * FIELD_WEIGHTS.tags + bm25FieldScore(queryTerms, d.fields.collection, avgLens.collection, idf) * FIELD_WEIGHTS.collection + bm25FieldScore(queryTerms, d.fields.body, avgLens.body, idf) * FIELD_WEIGHTS.body;
+    if (total > 0) scored.push({ ...d, score: total });
+  }
+  scored.sort((a, b) => (b.score ?? 0) - (a.score ?? 0));
+  return scored;
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  rankDocs,
+  tokenize
+});

package/dist/bm25.d.cts

@@ -0,0 +1,48 @@
+/**
+ * Pure, runtime-agnostic BM25 ranking.
+ *
+ * This module has ZERO Node/filesystem imports, so the exact same ranking runs on
+ * the server (native search over files) and in the browser (the static site, over
+ * a pre-built index). It is exposed browser-side via the `diffwiki-core/bm25`
+ * subpath export so the client bundle never pulls in Node-only core code.
+ *
+ * @module
+ */
+/** Tokenize a string: lowercase, split on non-alphanumeric, drop empties. */
+declare function tokenize(text: string): string[];
+/** Pre-tokenized, weighted fields of a rankable document. */
+interface RankableFields {
+    title: string[];
+    tags: string[];
+    collection: string[];
+    body: string[];
+}
+/**
+ * A document in the serialisable search index: routing metadata (for linking to
+ * the article) plus the pre-tokenized fields BM25 ranks. The static exporter
+ * writes `SearchDoc[]` to `search-index.json`; the browser ranks it with
+ * {@link rankDocs}.
+ */
+interface SearchDoc {
+    collection: string;
+    slug: string;
+    title: string;
+    tags: string[];
+    fields: RankableFields;
+}
+/**
+ * Rank pre-tokenized docs against a query with BM25 over the four weighted fields
+ * (title ×3, tags ×2, collection ×1.5, body ×1). Pure — no I/O.
+ *
+ * An empty/blank query returns every doc unranked (browse-all), preserving the
+ * native search's behaviour. Otherwise only matching docs (score > 0) are
+ * returned, highest score first. Each returned doc keeps its original shape plus
+ * a `score`.
+ */
+declare function rankDocs<T extends {
+    fields: RankableFields;
+}>(term: string, docs: T[]): Array<T & {
+    score?: number;
+}>;
+
+export { type RankableFields, type SearchDoc, rankDocs, tokenize };

package/dist/bm25.d.ts

@@ -0,0 +1,48 @@
+/**
+ * Pure, runtime-agnostic BM25 ranking.
+ *
+ * This module has ZERO Node/filesystem imports, so the exact same ranking runs on
+ * the server (native search over files) and in the browser (the static site, over
+ * a pre-built index). It is exposed browser-side via the `diffwiki-core/bm25`
+ * subpath export so the client bundle never pulls in Node-only core code.
+ *
+ * @module
+ */
+/** Tokenize a string: lowercase, split on non-alphanumeric, drop empties. */
+declare function tokenize(text: string): string[];
+/** Pre-tokenized, weighted fields of a rankable document. */
+interface RankableFields {
+    title: string[];
+    tags: string[];
+    collection: string[];
+    body: string[];
+}
+/**
+ * A document in the serialisable search index: routing metadata (for linking to
+ * the article) plus the pre-tokenized fields BM25 ranks. The static exporter
+ * writes `SearchDoc[]` to `search-index.json`; the browser ranks it with
+ * {@link rankDocs}.
+ */
+interface SearchDoc {
+    collection: string;
+    slug: string;
+    title: string;
+    tags: string[];
+    fields: RankableFields;
+}
+/**
+ * Rank pre-tokenized docs against a query with BM25 over the four weighted fields
+ * (title ×3, tags ×2, collection ×1.5, body ×1). Pure — no I/O.
+ *
+ * An empty/blank query returns every doc unranked (browse-all), preserving the
+ * native search's behaviour. Otherwise only matching docs (score > 0) are
+ * returned, highest score first. Each returned doc keeps its original shape plus
+ * a `score`.
+ */
+declare function rankDocs<T extends {
+    fields: RankableFields;
+}>(term: string, docs: T[]): Array<T & {
+    score?: number;
+}>;
+
+export { type RankableFields, type SearchDoc, rankDocs, tokenize };

package/dist/bm25.js

@@ -0,0 +1,8 @@
+import {
+  rankDocs,
+  tokenize
+} from "./chunk-FTUF6IXS.js";
+export {
+  rankDocs,
+  tokenize
+};

package/dist/chunk-FTUF6IXS.js

@@ -0,0 +1,60 @@
+// src/bm25.ts
+var K1 = 1.5;
+var B = 0.75;
+var FIELD_WEIGHTS = { title: 3, tags: 2, collection: 1.5, body: 1 };
+function tokenize(text) {
+  return text.toLowerCase().split(/[^a-z0-9]+/).filter((t) => t.length > 0);
+}
+function termFreqs(tokens) {
+  const freqs = /* @__PURE__ */ new Map();
+  for (const t of tokens) freqs.set(t, (freqs.get(t) ?? 0) + 1);
+  return freqs;
+}
+function bm25FieldScore(queryTerms, fieldTokens, avgFieldLen, idf) {
+  const tf = termFreqs(fieldTokens);
+  const dl = fieldTokens.length;
+  const avgDl = avgFieldLen || 1;
+  let score = 0;
+  for (const term of queryTerms) {
+    const freq = tf.get(term) ?? 0;
+    if (freq === 0) continue;
+    const termIdf = idf.get(term) ?? 0;
+    const numerator = freq * (K1 + 1);
+    const denominator = freq + K1 * (1 - B + B * (dl / avgDl));
+    score += termIdf * (numerator / denominator);
+  }
+  return score;
+}
+function rankDocs(term, docs) {
+  const queryTerms = tokenize(term);
+  if (queryTerms.length === 0) return docs.map((d) => ({ ...d }));
+  const N = docs.length;
+  const docTokenSets = docs.map(
+    (d) => /* @__PURE__ */ new Set([...d.fields.title, ...d.fields.tags, ...d.fields.collection, ...d.fields.body])
+  );
+  const idf = /* @__PURE__ */ new Map();
+  for (const qt of queryTerms) {
+    let docFreq = 0;
+    for (const set of docTokenSets) if (set.has(qt)) docFreq++;
+    idf.set(qt, Math.log((N - docFreq + 0.5) / (docFreq + 0.5) + 1));
+  }
+  const avg = (sel) => docs.reduce((s, d) => s + sel(d), 0) / (N || 1);
+  const avgLens = {
+    title: avg((d) => d.fields.title.length),
+    tags: avg((d) => d.fields.tags.length),
+    collection: avg((d) => d.fields.collection.length),
+    body: avg((d) => d.fields.body.length)
+  };
+  const scored = [];
+  for (const d of docs) {
+    const total = bm25FieldScore(queryTerms, d.fields.title, avgLens.title, idf) * FIELD_WEIGHTS.title + bm25FieldScore(queryTerms, d.fields.tags, avgLens.tags, idf) * FIELD_WEIGHTS.tags + bm25FieldScore(queryTerms, d.fields.collection, avgLens.collection, idf) * FIELD_WEIGHTS.collection + bm25FieldScore(queryTerms, d.fields.body, avgLens.body, idf) * FIELD_WEIGHTS.body;
+    if (total > 0) scored.push({ ...d, score: total });
+  }
+  scored.sort((a, b) => (b.score ?? 0) - (a.score ?? 0));
+  return scored;
+}
+
+export {
+  tokenize,
+  rankDocs
+};

package/dist/{chunk-AZ6SZH6P.js → chunk-MLUJY3HI.js}

@@ -96,11 +96,27 @@ var InvalidTargetError = class extends DiffwikiError {
 };
 var PluginError = class extends DiffwikiError {
 };
+var WorktreeNotAllowedError = class extends DiffwikiError {
+  constructor(root) {
+    super(`refusing to add a linked worktree \u2014 add the source repository instead (${root})`);
+  }
+};
+var CollectionAlreadyAddedError = class extends DiffwikiError {
+  constructor(name) {
+    super(`this repository is already added as collection "${name}"`);
+  }
+};
+var InvalidExportSelectionError = class extends DiffwikiError {
+  constructor(problems) {
+    super(`export aborted \u2014 ${problems.length} selection problem(s):
+${problems.map((p) => `  - ${p}`).join("\n")}`);
+  }
+};
 
 // src/collections.ts
 async function fireHook(event, collection) {
   try {
-    const { emitPluginEvent: emitPluginEvent2 } = await import("./events-H3D7FT4O.js");
+    const { emitPluginEvent: emitPluginEvent2 } = await import("./events-UCJWB4WP.js");
     await emitPluginEvent2(event, collection);
   } catch {
   }
@@ -111,6 +127,9 @@ async function listCollections() {
 async function findCollection(name) {
   return (await readRegistry()).collections.find((c) => c.name === name);
 }
+async function findCollectionById(id) {
+  return (await readRegistry()).collections.find((c) => c.id === id);
+}
 async function createCollection(name) {
   if (await findCollection(name)) throw new CollectionExistsError(name);
   const dir = collectionDir(name);
@@ -452,6 +471,9 @@ export {
   ArticleNotFoundError,
   InvalidTargetError,
   PluginError,
+  WorktreeNotAllowedError,
+  CollectionAlreadyAddedError,
+  InvalidExportSelectionError,
   resolveHome,
   registryPath,
   configPath,
@@ -479,6 +501,7 @@ export {
   emitPluginEvent,
   listCollections,
   findCollection,
+  findCollectionById,
   createCollection,
   removeCollection
 };

package/dist/{events-H3D7FT4O.js → events-UCJWB4WP.js}

@@ -1,6 +1,6 @@
 import {
   emitPluginEvent
-} from "./chunk-AZ6SZH6P.js";
+} from "./chunk-MLUJY3HI.js";
 export {
   emitPluginEvent
 };