sensemaking 0.4.0 → 0.6.0

package/dist/cjs/features/embed.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/embed.ts"],"sourcesContent":["import { existsSync, mkdirSync, readFileSync, renameSync, writeFileSync } from 'node:fs';\nimport { homedir } from 'node:os';\nimport { join } from 'node:path';\nimport type { DatabaseSync } from 'node:sqlite';\nimport type { Config, ResolvedConfig } from '../config.ts';\nimport { embedConfig } from '../config.ts';\nimport { SenseError } from '../errors.ts';\nimport { parseFile } from '../scan.ts';\nimport type { Feature } from './types.ts';\n\n// embeddings(path, chunk, start_line, end_line, scale, vector): heading-based chunks,\n// int8 vectors with a per-vector dequantization scale, vector NULL = not yet embedded.\n// Reconcile stores dirty rows inside its transaction; embedding tops up on the next\n// semantic query (embedPending), so staleness costs recall, never correctness.\n\n// Storage lever fixed by the bake-off (BENCHMARKING.md): int8 at 256 dims is\n// quality-free vs f32-512 when fused. Queries stay f32 at the same dims.\nconst STORE_DIMS = 256;\nconst BATCH = 64;\n\nexport interface EmbedProvider {\n  id: string; // model identity; participates in the cache key, change -> re-embed\n  dims: number;\n  embed(texts: string[]): Promise<Float32Array[]>;\n}\n\ninterface Chunk {\n  startLine: number;\n  endLine: number;\n  text: string;\n}\n\n// Deterministic chunker used at reconcile (line ranges stored) and at embed time (text\n// re-derived from the file -- chunk text is never stored). Heading-delimited with the\n// preamble kept; whole file when no headings. The title/summary prefix mirrors the\n// bm25 column weighting.\nfunction chunksOf(raw: string, search?: { title: string; summary: string }): Chunk[] {\n  const lines = raw.split('\\n');\n  const starts: number[] = [];\n  let inFence = false;\n  for (let i = 0; i < lines.length; i++) {\n    if (/^(```|~~~)/.test(lines[i])) {\n      inFence = !inFence;\n      continue;\n    }\n    if (!inFence && /^#{1,6} +/.test(lines[i])) starts.push(i + 1);\n  }\n  const bounds = starts.length === 0 ? [1] : starts[0] > 1 ? [1, ...starts] : starts;\n  const prefix = [search?.title, search?.summary].filter(Boolean).join('\\n');\n  const chunks: Chunk[] = [];\n  bounds.forEach((start, i) => {\n    const end = i + 1 < bounds.length ? bounds[i + 1] - 1 : lines.length;\n    const text = lines\n      .slice(start - 1, end)\n      .join('\\n')\n      .trim();\n    if (text.length > 0) chunks.push({ startLine: start, endLine: end, text: prefix ? `${prefix}\\n${text}` : text });\n  });\n  return chunks;\n}\n\nexport const embed: Feature = {\n  name: 'embed',\n  schema(db) {\n    db.exec(`CREATE TABLE IF NOT EXISTS embeddings (\"path\" TEXT, chunk INTEGER, start_line INTEGER, end_line INTEGER, scale REAL, vector BLOB, PRIMARY KEY (\"path\", chunk))`);\n  },\n  extract(raw, _body, search) {\n    return chunksOf(raw, search);\n  },\n  remove(db, path) {\n    db.prepare('DELETE FROM embeddings WHERE \"path\" = ?').run(path);\n  },\n  store(db, path, extracted) {\n    const insert = db.prepare('INSERT INTO embeddings (\"path\", chunk, start_line, end_line, scale, vector) VALUES (?, ?, ?, ?, NULL, NULL)');\n    (extracted as Chunk[]).forEach((c, idx) => insert.run(path, idx, c.startLine, c.endLine));\n  },\n};\n\n// --- static type: Model2Vec safetensors + pure-JS tokenizer, model cached in ~/.cache ---\n// Encode convention from model2vec/model.py: no special tokens, drop unk ids, mean-pool,\n// L2-normalize.\n\nfunction fetchToFile(url: string, dest: string): Promise<void> {\n  return fetch(url).then(async (res) => {\n    if (!res.ok) throw new SenseError('EMBED_MODEL', `model download failed: ${url} -> HTTP ${res.status}`);\n    writeFileSync(`${dest}.part`, Buffer.from(await res.arrayBuffer()));\n    renameSync(`${dest}.part`, dest);\n  });\n}\n\nasync function staticProvider(model: string): Promise<EmbedProvider> {\n  let dir = model;\n  if (!existsSync(join(dir, 'model.safetensors'))) {\n    dir = join(homedir(), '.cache', 'sensemaking', 'models', model.replace(/\\//g, '--'));\n    mkdirSync(dir, { recursive: true });\n    for (const file of ['model.safetensors', 'tokenizer.json']) {\n      if (!existsSync(join(dir, file))) {\n        console.error(`fetching ${model}/${file} into ${dir} (once; delete to refetch)`);\n        await fetchToFile(`https://huggingface.co/${model}/resolve/main/${file}`, join(dir, file));\n      }\n    }\n  }\n\n  const raw = readFileSync(join(dir, 'model.safetensors'));\n  const headerLen = Number(raw.readBigUInt64LE(0));\n  const header = JSON.parse(raw.subarray(8, 8 + headerLen).toString('utf8')) as Record<string, { dtype: string; shape: number[]; data_offsets: number[] }>;\n  const entry = Object.entries(header).find(([k]) => k !== '__metadata__');\n  if (!entry || entry[1].dtype !== 'F32') throw new SenseError('EMBED_MODEL', `${model}: expected an F32 safetensors matrix`);\n  const spec = entry[1];\n  const dims = spec.shape[1];\n  const dataStart = raw.byteOffset + 8 + headerLen + spec.data_offsets[0];\n  const matrix = dataStart % 4 === 0 ? new Float32Array(raw.buffer, dataStart, spec.shape[0] * dims) : new Float32Array(raw.buffer.slice(dataStart, dataStart + spec.shape[0] * dims * 4));\n\n  const tokenizerJson = JSON.parse(readFileSync(join(dir, 'tokenizer.json'), 'utf8'));\n  // Lazy import: the tokenizer loads only on the semantic path, never at CLI startup.\n  const { Tokenizer } = await import('@huggingface/tokenizers');\n  const tok = new Tokenizer(tokenizerJson, {});\n  const unkId = tokenizerJson.model?.vocab?.[tokenizerJson.model?.unk_token] ?? -1;\n\n  function one(text: string): Float32Array {\n    const ids = (tok.encode(text, { add_special_tokens: false }).ids as number[]).filter((id) => id !== unkId);\n    const v = new Float32Array(dims);\n    if (ids.length === 0) return v;\n    for (const id of ids) {\n      const off = id * dims;\n      for (let d = 0; d < dims; d++) v[d] += matrix[off + d];\n    }\n    let norm = 0;\n    for (let d = 0; d < dims; d++) {\n      v[d] /= ids.length;\n      norm += v[d] * v[d];\n    }\n    norm = Math.sqrt(norm) + 1e-32;\n    for (let d = 0; d < dims; d++) v[d] /= norm;\n    return v;\n  }\n\n  return { id: `static:${model}`, dims, embed: async (texts) => texts.map(one) };\n}\n\n// --- api type: one POST against any OpenAI-compatible /embeddings endpoint ---\n\nasync function apiProvider(model: string, url: string | undefined, keyEnv: string | undefined): Promise<EmbedProvider> {\n  if (!url) throw new SenseError('EMBED_MODEL', 'features.embed.type \"api\" requires a url');\n  const base = url.replace(/\\/+$/, '');\n  const headers: Record<string, string> = { 'content-type': 'application/json' };\n  const key = keyEnv ? process.env[keyEnv] : undefined;\n  if (key) headers.authorization = `Bearer ${key}`;\n\n  async function post(texts: string[]): Promise<Float32Array[]> {\n    const res = await fetch(`${base}/embeddings`, { method: 'POST', headers, body: JSON.stringify({ model, input: texts }) });\n    if (!res.ok) throw new SenseError('EMBED_MODEL', `${base}/embeddings -> HTTP ${res.status}`);\n    const body = (await res.json()) as { data: Array<{ embedding: number[] }> };\n    return body.data.map((d) => Float32Array.from(d.embedding));\n  }\n\n  const dims = (await post(['dimension probe']))[0].length;\n  return { id: `api:${base}:${model}`, dims, embed: post };\n}\n\nconst providers = new Map<string, Promise<EmbedProvider>>();\n\nfunction getProvider(cfg: Config): Promise<EmbedProvider> {\n  const e = embedConfig(cfg);\n  if (!e) throw new SenseError('EMBED_DISABLED', 'semantic expansion needs features.embed in sense.config.json (e.g. \"features\": { \"embed\": true })');\n  const sig = `${e.type}:${e.model}:${e.url ?? ''}`;\n  let p = providers.get(sig);\n  if (!p) {\n    p = e.type === 'api' ? apiProvider(e.model, e.url, e.key) : staticProvider(e.model);\n    providers.set(sig, p);\n  }\n  return p;\n}\n\n// Slice + re-normalize (Matryoshka); optionally round through int8 storage.\nfunction toStore(full: Float32Array, dims: number, int8: boolean): { v: Float32Array; scale: number } {\n  const v = new Float32Array(dims);\n  let norm = 0;\n  for (let d = 0; d < dims; d++) norm += full[d] * full[d];\n  norm = Math.sqrt(norm) + 1e-32;\n  for (let d = 0; d < dims; d++) v[d] = full[d] / norm;\n  if (!int8) return { v, scale: 1 };\n  let max = 0;\n  for (let d = 0; d < dims; d++) max = Math.max(max, Math.abs(v[d]));\n  return { v, scale: max / 127 || 1 };\n}\n\n// Embed rows whose vector is NULL, re-deriving chunk text from the files through the\n// same parse + chunker that stored the rows.\nexport async function embedPending(db: DatabaseSync, cfg: Config, baseDir: string): Promise<void> {\n  const provider = await getProvider(cfg); // throws EMBED_DISABLED before touching the table\n  const dirty = db.prepare('SELECT \"path\", chunk FROM embeddings WHERE vector IS NULL ORDER BY \"path\", chunk').all() as Array<{ path: string; chunk: number }>;\n  if (dirty.length === 0) return;\n  const storeDims = Math.min(STORE_DIMS, provider.dims);\n\n  const byPath = new Map<string, number[]>();\n  for (const row of dirty) {\n    const list = byPath.get(row.path) ?? [];\n    list.push(row.chunk);\n    byPath.set(row.path, list);\n  }\n\n  const jobs: Array<{ path: string; chunk: number; text: string }> = [];\n  for (const [path, chunkIdxs] of byPath) {\n    let chunks: Chunk[];\n    try {\n      chunks = parseFile({ relPath: path, absPath: join(baseDir, path), mtimeMs: 0, size: 0 }, [embed]).doc.extracted.embed as Chunk[];\n    } catch {\n      continue; // vanished since reconcile; the next reconcile removes its rows\n    }\n    for (const idx of chunkIdxs) if (chunks[idx]) jobs.push({ path, chunk: idx, text: chunks[idx].text });\n  }\n\n  const update = db.prepare('UPDATE embeddings SET scale = ?, vector = ? WHERE \"path\" = ? AND chunk = ?');\n  for (let i = 0; i < jobs.length; i += BATCH) {\n    const batch = jobs.slice(i, i + BATCH);\n    const vectors = await provider.embed(batch.map((j) => j.text));\n    db.exec('BEGIN');\n    try {\n      batch.forEach((job, j) => {\n        const { v, scale } = toStore(vectors[j], storeDims, true);\n        const q = new Int8Array(storeDims);\n        for (let d = 0; d < storeDims; d++) q[d] = Math.round(v[d] / scale);\n        update.run(scale, Buffer.from(q.buffer), job.path, job.chunk);\n      });\n      db.exec('COMMIT');\n    } catch (err) {\n      db.exec('ROLLBACK');\n      throw err;\n    }\n  }\n}\n\n// Top candidates by cosine for a semantic find: best chunk per file, its line range\n// riding along. FTS5 operators in the terms are lexical syntax, not meaning -- stripped\n// before embedding.\nexport async function semanticCandidates(db: DatabaseSync, cfg: Config, terms: string, fetch: number): Promise<Array<{ path: string; lines: string }>> {\n  const baseDir = (cfg as Partial<ResolvedConfig>).baseDir;\n  if (!baseDir) throw new SenseError('EMBED_MODEL', 'semantic expansion needs a config with baseDir (use loadConfig/open)');\n  await embedPending(db, cfg, baseDir);\n\n  const provider = await getProvider(cfg);\n  const storeDims = Math.min(STORE_DIMS, provider.dims);\n  const text = (terms.match(/[\\p{L}\\p{N}]+/gu) ?? []).filter((t) => !['AND', 'OR', 'NOT', 'NEAR'].includes(t)).join(' ');\n  const { v: qv } = toStore((await provider.embed([text]))[0], storeDims, false);\n\n  const rows = db.prepare('SELECT \"path\", start_line, end_line, scale, vector FROM embeddings WHERE vector IS NOT NULL').all() as Array<{\n    path: string;\n    start_line: number;\n    end_line: number;\n    scale: number;\n    vector: Uint8Array;\n  }>;\n\n  const best = new Map<string, { score: number; lines: string }>();\n  for (const row of rows) {\n    const q = new Int8Array(row.vector.buffer, row.vector.byteOffset, Math.min(storeDims, row.vector.byteLength));\n    let dot = 0;\n    for (let d = 0; d < q.length; d++) dot += q[d] * qv[d];\n    const score = dot * row.scale;\n    const existing = best.get(row.path);\n    if (!existing || score > existing.score) best.set(row.path, { score, lines: `L${row.start_line}-${row.end_line}` });\n  }\n  return [...best.entries()]\n    .sort((a, b) => b[1].score - a[1].score)\n    .slice(0, fetch)\n    .map(([path, b]) => ({ path, lines: b.lines }));\n}\n"],"names":["embed","embedPending","semanticCandidates","STORE_DIMS","BATCH","chunksOf","raw","search","lines","split","starts","inFence","i","length","test","push","bounds","prefix","title","summary","filter","Boolean","join","chunks","forEach","start","end","text","slice","trim","startLine","endLine","name","schema","db","exec","extract","_body","remove","path","prepare","run","store","extracted","insert","c","idx","fetchToFile","url","dest","fetch","then","res","ok","SenseError","status","Buffer","from","arrayBuffer","writeFileSync","renameSync","staticProvider","model","tokenizerJson","dir","file","headerLen","header","entry","spec","dims","dataStart","matrix","Tokenizer","tok","unkId","one","ids","encode","add_special_tokens","id","v","Float32Array","off","d","norm","Math","sqrt","existsSync","homedir","replace","mkdirSync","recursive","console","error","readFileSync","Number","readBigUInt64LE","JSON","parse","subarray","toString","Object","entries","find","k","dtype","shape","byteOffset","data_offsets","buffer","vocab","unk_token","texts","map","apiProvider","keyEnv","base","headers","key","post","body","method","stringify","input","json","data","embedding","process","env","undefined","authorization","providers","Map","getProvider","cfg","e","embedConfig","sig","type","p","get","set","toStore","full","int8","scale","max","abs","baseDir","provider","dirty","storeDims","byPath","row","list","jobs","chunkIdxs","update","batch","vectors","j","job","q","Int8Array","round","chunk","err","all","min","parseFile","relPath","absPath","mtimeMs","size","doc","terms","qv","rows","best","dot","score","existing","match","t","includes","vector","byteLength","start_line","end_line","sort","a","b"],"mappings":";;;;;;;;;;;QA6DaA;eAAAA;;QAgISC;eAAAA;;QA+CAC;eAAAA;;;sBA5OyD;sBACvD;wBACH;wBAGO;wBACD;sBACD;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAG1B,sFAAsF;AACtF,uFAAuF;AACvF,oFAAoF;AACpF,+EAA+E;AAE/E,6EAA6E;AAC7E,yEAAyE;AACzE,IAAMC,aAAa;AACnB,IAAMC,QAAQ;AAcd,uFAAuF;AACvF,sFAAsF;AACtF,mFAAmF;AACnF,yBAAyB;AACzB,SAASC,SAASC,GAAW,EAAEC,MAA2C;IACxE,IAAMC,QAAQF,IAAIG,KAAK,CAAC;IACxB,IAAMC,SAAmB,EAAE;IAC3B,IAAIC,UAAU;IACd,IAAK,IAAIC,IAAI,GAAGA,IAAIJ,MAAMK,MAAM,EAAED,IAAK;QACrC,IAAI,aAAaE,IAAI,CAACN,KAAK,CAACI,EAAE,GAAG;YAC/BD,UAAU,CAACA;YACX;QACF;QACA,IAAI,CAACA,WAAW,YAAYG,IAAI,CAACN,KAAK,CAACI,EAAE,GAAGF,OAAOK,IAAI,CAACH,IAAI;IAC9D;IACA,IAAMI,SAASN,OAAOG,MAAM,KAAK,IAAI;QAAC;KAAE,GAAGH,MAAM,CAAC,EAAE,GAAG,IAAI;QAAC;KAAa,CAAd,OAAI,qBAAGA,WAAUA;IAC5E,IAAMO,SAAS;QAACV,mBAAAA,6BAAAA,OAAQW,KAAK;QAAEX,mBAAAA,6BAAAA,OAAQY,OAAO;KAAC,CAACC,MAAM,CAACC,SAASC,IAAI,CAAC;IACrE,IAAMC,SAAkB,EAAE;IAC1BP,OAAOQ,OAAO,CAAC,SAACC,OAAOb;QACrB,IAAMc,MAAMd,IAAI,IAAII,OAAOH,MAAM,GAAGG,MAAM,CAACJ,IAAI,EAAE,GAAG,IAAIJ,MAAMK,MAAM;QACpE,IAAMc,OAAOnB,MACVoB,KAAK,CAACH,QAAQ,GAAGC,KACjBJ,IAAI,CAAC,MACLO,IAAI;QACP,IAAIF,KAAKd,MAAM,GAAG,GAAGU,OAAOR,IAAI,CAAC;YAAEe,WAAWL;YAAOM,SAASL;YAAKC,MAAMV,SAAS,AAAC,GAAaU,OAAXV,QAAO,MAAS,OAALU,QAASA;QAAK;IAChH;IACA,OAAOJ;AACT;AAEO,IAAMvB,QAAiB;IAC5BgC,MAAM;IACNC,QAAAA,SAAAA,OAAOC,EAAE;QACPA,GAAGC,IAAI,CAAC;IACV;IACAC,SAAAA,SAAAA,QAAQ9B,GAAG,EAAE+B,KAAK,EAAE9B,MAAM;QACxB,OAAOF,SAASC,KAAKC;IACvB;IACA+B,QAAAA,SAAAA,OAAOJ,EAAE,EAAEK,IAAI;QACbL,GAAGM,OAAO,CAAC,2CAA2CC,GAAG,CAACF;IAC5D;IACAG,OAAAA,SAAAA,MAAMR,EAAE,EAAEK,IAAI,EAAEI,SAAS;QACvB,IAAMC,SAASV,GAAGM,OAAO,CAAC;QACzBG,UAAsBnB,OAAO,CAAC,SAACqB,GAAGC;mBAAQF,OAAOH,GAAG,CAACF,MAAMO,KAAKD,EAAEf,SAAS,EAAEe,EAAEd,OAAO;;IACzF;AACF;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,gBAAgB;AAEhB,SAASgB,YAAYC,GAAW,EAAEC,IAAY;IAC5C,OAAOC,MAAMF,KAAKG,IAAI,CAAC,SAAOC;;;;;;wBAC5B,IAAI,CAACA,IAAIC,EAAE,EAAE,MAAM,IAAIC,oBAAU,CAAC,eAAe,AAAC,0BAAwCF,OAAfJ,KAAI,aAAsB,OAAXI,IAAIG,MAAM;;4BACrF,GAAO,OAALN,MAAK;;4BAAQO,OAAOC,IAAI;wBAAC;;4BAAML,IAAIM,WAAW;;;wBAA/DC,qBAAa;4BAAiBH,QAAAA;gCAAY;;;wBAC1CI,IAAAA,kBAAU,EAAC,AAAC,GAAO,OAALX,MAAK,UAAQA;;;;;;QAC7B;;AACF;AAEA,SAAeY,eAAeC,KAAa;;kBA2BEC,sBAA7BA,4BAAAA,uBA1BVC,SAIiB,OAARC,MAQP3D,KACA4D,WACAC,QACAC,OAEAC,MACAC,MACAC,WACAC,QAEAT,eAEEU,WACFC,KACAC;QAEN,SAASC,IAAIjD,IAAY;YACvB,IAAMkD,MAAM,AAACH,IAAII,MAAM,CAACnD,MAAM;gBAAEoD,oBAAoB;YAAM,GAAGF,GAAG,CAAczD,MAAM,CAAC,SAAC4D;uBAAOA,OAAOL;;YACpG,IAAMM,IAAI,IAAIC,aAAaZ;YAC3B,IAAIO,IAAIhE,MAAM,KAAK,GAAG,OAAOoE;gBACxB,kCAAA,2BAAA;;gBAAL,QAAK,YAAYJ,wBAAZ,SAAA,6BAAA,QAAA,yBAAA,iCAAiB;oBAAjB,IAAMG,KAAN;oBACH,IAAMG,MAAMH,KAAKV;oBACjB,IAAK,IAAIc,IAAI,GAAGA,IAAId,MAAMc,IAAKH,CAAC,CAACG,EAAE,IAAIZ,MAAM,CAACW,MAAMC,EAAE;gBACxD;;gBAHK;gBAAA;;;yBAAA,6BAAA;wBAAA;;;wBAAA;8BAAA;;;;YAIL,IAAIC,OAAO;YACX,IAAK,IAAID,KAAI,GAAGA,KAAId,MAAMc,KAAK;gBAC7BH,CAAC,CAACG,GAAE,IAAIP,IAAIhE,MAAM;gBAClBwE,QAAQJ,CAAC,CAACG,GAAE,GAAGH,CAAC,CAACG,GAAE;YACrB;YACAC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;YACzB,IAAK,IAAID,KAAI,GAAGA,KAAId,MAAMc,KAAKH,CAAC,CAACG,GAAE,IAAIC;YACvC,OAAOJ;QACT;;;;oBA5CIjB,MAAMF;yBACN,CAAC0B,IAAAA,kBAAU,EAAClE,IAAAA,cAAI,EAAC0C,KAAK,uBAAtB;;;;oBACFA,MAAM1C,IAAAA,cAAI,EAACmE,IAAAA,eAAO,KAAI,UAAU,eAAe,UAAU3B,MAAM4B,OAAO,CAAC,OAAO;oBAC9EC,IAAAA,iBAAS,EAAC3B,KAAK;wBAAE4B,WAAW;oBAAK;4BACd;wBAAC;wBAAqB;;;;+BAAtB;;;;oBAAR3B,OAAQ;yBACb,CAACuB,IAAAA,kBAAU,EAAClE,IAAAA,cAAI,EAAC0C,KAAKC,QAAtB;;;;oBACF4B,QAAQC,KAAK,CAAC,AAAC,YAAoB7B,OAATH,OAAM,KAAgBE,OAAbC,MAAK,UAAY,OAAJD,KAAI;oBACpD;;wBAAMjB,YAAY,AAAC,0BAA+CkB,OAAtBH,OAAM,kBAAqB,OAALG,OAAQ3C,IAAAA,cAAI,EAAC0C,KAAKC;;;oBAApF;;;;;;;;;oBAKA3D,MAAMyF,IAAAA,oBAAY,EAACzE,IAAAA,cAAI,EAAC0C,KAAK;oBAC7BE,YAAY8B,OAAO1F,IAAI2F,eAAe,CAAC;oBACvC9B,SAAS+B,KAAKC,KAAK,CAAC7F,IAAI8F,QAAQ,CAAC,GAAG,IAAIlC,WAAWmC,QAAQ,CAAC;oBAC5DjC,QAAQkC,OAAOC,OAAO,CAACpC,QAAQqC,IAAI,CAAC;iEAAEC;+BAAOA,MAAM;;oBACzD,IAAI,CAACrC,SAASA,KAAK,CAAC,EAAE,CAACsC,KAAK,KAAK,OAAO,MAAM,IAAIpD,oBAAU,CAAC,eAAe,AAAC,GAAQ,OAANQ,OAAM;oBAC/EO,OAAOD,KAAK,CAAC,EAAE;oBACfE,OAAOD,KAAKsC,KAAK,CAAC,EAAE;oBACpBpC,YAAYjE,IAAIsG,UAAU,GAAG,IAAI1C,YAAYG,KAAKwC,YAAY,CAAC,EAAE;oBACjErC,SAASD,YAAY,MAAM,IAAI,IAAIW,aAAa5E,IAAIwG,MAAM,EAAEvC,WAAWF,KAAKsC,KAAK,CAAC,EAAE,GAAGrC,QAAQ,IAAIY,aAAa5E,IAAIwG,MAAM,CAAClF,KAAK,CAAC2C,WAAWA,YAAYF,KAAKsC,KAAK,CAAC,EAAE,GAAGrC,OAAO;oBAE/KP,gBAAgBmC,KAAKC,KAAK,CAACJ,IAAAA,oBAAY,EAACzE,IAAAA,cAAI,EAAC0C,KAAK,mBAAmB;oBAErD;;wBAAM;2EAAA,QAAO;;;;oBAA3BS,YAAc,cAAdA;oBACFC,MAAM,IAAID,UAAUV,eAAe,CAAC;oBACpCY,iBAAQZ,wBAAAA,cAAcD,KAAK,cAAnBC,6CAAAA,6BAAAA,sBAAqBgD,KAAK,cAA1BhD,iDAAAA,0BAA4B,EAACA,uBAAAA,cAAcD,KAAK,cAAnBC,2CAAAA,qBAAqBiD,SAAS,CAAC,uCAAI,CAAC;oBAoB/E;;wBAAO;4BAAEhC,IAAI,AAAC,UAAe,OAANlB;4BAASQ,MAAAA;4BAAMtE,OAAO,SAAPA,MAAciH;;;;;4CAAUA,MAAMC,GAAG,CAACtC;;;;;wBAAK;;;;IAC/E;;AAEA,gFAAgF;AAEhF,SAAeuC,YAAYrD,KAAa,EAAEd,GAAuB,EAAEoE,MAA0B;;YAErFC,MACAC,SACAC,KAUAjD;QAPN,SAAekD,KAAKP,KAAe;;oBAC3B7D,KAEAqE;;;;4BAFM;;gCAAMvE,MAAM,AAAC,GAAO,OAALmE,MAAK,gBAAc;oCAAEK,QAAQ;oCAAQJ,SAAAA;oCAASG,MAAMvB,KAAKyB,SAAS,CAAC;wCAAE7D,OAAAA;wCAAO8D,OAAOX;oCAAM;gCAAG;;;4BAAjH7D,MAAM;4BACZ,IAAI,CAACA,IAAIC,EAAE,EAAE,MAAM,IAAIC,oBAAU,CAAC,eAAe,AAAC,GAA6BF,OAA3BiE,MAAK,wBAAiC,OAAXjE,IAAIG,MAAM;4BAC3E;;gCAAMH,IAAIyE,IAAI;;;4BAAtBJ,OAAQ;4BACd;;gCAAOA,KAAKK,IAAI,CAACZ,GAAG,CAAC,SAAC9B;2CAAMF,aAAazB,IAAI,CAAC2B,EAAE2C,SAAS;;;;;YAC3D;;;;;oBAXA,IAAI,CAAC/E,KAAK,MAAM,IAAIM,oBAAU,CAAC,eAAe;oBACxC+D,OAAOrE,IAAI0C,OAAO,CAAC,QAAQ;oBAC3B4B,UAAkC;wBAAE,gBAAgB;oBAAmB;oBACvEC,MAAMH,SAASY,QAAQC,GAAG,CAACb,OAAO,GAAGc;oBAC3C,IAAIX,KAAKD,QAAQa,aAAa,GAAG,AAAC,UAAa,OAAJZ;oBAS7B;;wBAAMC;4BAAM;;;;oBAApBlD,OAAO,AAAC,aAAgC,CAAC,EAAE,CAACzD,MAAM;oBACxD;;wBAAO;4BAAEmE,IAAI,AAAC,OAAclB,OAARuD,MAAK,KAAS,OAANvD;4BAASQ,MAAAA;4BAAMtE,OAAOwH;wBAAK;;;;IACzD;;AAEA,IAAMY,YAAY,IAAIC;AAEtB,SAASC,YAAYC,GAAW;QAGMC;IAFpC,IAAMA,IAAIC,IAAAA,qBAAW,EAACF;IACtB,IAAI,CAACC,GAAG,MAAM,IAAIlF,oBAAU,CAAC,kBAAkB;IAC/C,IAAMoF,MAAM,AArKd,AAqKe,GAAYF,OAAVA,EAAEG,IAAI,EAAC,YAAGH,EAAE1E,KAAK,EAAC,KAAe,QAAZ0E,SAAAA,EAAExF,GAAG,cAALwF,oBAAAA,SAAS;IAC7C,IAAII,IAAIR,UAAUS,GAAG,CAACH;IACtB,IAAI,CAACE,GAAG;QACNA,IAAIJ,EAAEG,IAAI,KAAK,QAAQxB,YAAYqB,EAAE1E,KAAK,EAAE0E,EAAExF,GAAG,EAAEwF,EAAEjB,GAAG,IAAI1D,eAAe2E,EAAE1E,KAAK;QAClFsE,UAAUU,GAAG,CAACJ,KAAKE;IACrB;IACA,OAAOA;AACT;AAEA,4EAA4E;AAC5E,SAASG,QAAQC,IAAkB,EAAE1E,IAAY,EAAE2E,IAAa;IAC9D,IAAMhE,IAAI,IAAIC,aAAaZ;IAC3B,IAAIe,OAAO;IACX,IAAK,IAAID,IAAI,GAAGA,IAAId,MAAMc,IAAKC,QAAQ2D,IAAI,CAAC5D,EAAE,GAAG4D,IAAI,CAAC5D,EAAE;IACxDC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;IACzB,IAAK,IAAID,KAAI,GAAGA,KAAId,MAAMc,KAAKH,CAAC,CAACG,GAAE,GAAG4D,IAAI,CAAC5D,GAAE,GAAGC;IAChD,IAAI,CAAC4D,MAAM,OAAO;QAAEhE,GAAAA;QAAGiE,OAAO;IAAE;IAChC,IAAIC,MAAM;IACV,IAAK,IAAI/D,KAAI,GAAGA,KAAId,MAAMc,KAAK+D,MAAM7D,KAAK6D,GAAG,CAACA,KAAK7D,KAAK8D,GAAG,CAACnE,CAAC,CAACG,GAAE;IAChE,OAAO;QAAEH,GAAAA;QAAGiE,OAAOC,MAAM,OAAO;IAAE;AACpC;AAIO,SAAelJ,aAAaiC,EAAgB,EAAEqG,GAAW,EAAEc,OAAe;;mBACzEC,UACAC,OAEAC,WAEAC,QACD,2BAAA,mBAAA,gBAAA,WAAA,OAAMC,KACID,aAAPE,MAKFC,MACD,4BAAA,oBAAA,iBAAA,YAAA,qBAAOrH,MAAMsH,WACZtI,QAMC,4BAAA,oBAAA,iBAAA,YAAA,QAAMuB,KAGPgH,QACGlJ;;;;;4BACDmJ,OACAC;;;;oCADAD,QAAQH,KAAKhI,KAAK,CAAChB,GAAGA,IAAIR;oCAChB;;wCAAMkJ,SAAStJ,KAAK,CAAC+J,MAAM7C,GAAG,CAAC,SAAC+C;mDAAMA,EAAEtI,IAAI;;;;oCAAtDqI,UAAU;oCAChB9H,GAAGC,IAAI,CAAC;oCACR,IAAI;wCACF4H,MAAMvI,OAAO,CAAC,SAAC0I,KAAKD;4CAClB,IAAqBlB,WAAAA,QAAQiB,OAAO,CAACC,EAAE,EAAET,WAAW,OAA5CvE,IAAa8D,SAAb9D,GAAGiE,QAAUH,SAAVG;4CACX,IAAMiB,IAAI,IAAIC,UAAUZ;4CACxB,IAAK,IAAIpE,IAAI,GAAGA,IAAIoE,WAAWpE,IAAK+E,CAAC,CAAC/E,EAAE,GAAGE,KAAK+E,KAAK,CAACpF,CAAC,CAACG,EAAE,GAAG8D;4CAC7DY,OAAOrH,GAAG,CAACyG,OAAO1F,OAAOC,IAAI,CAAC0G,EAAErD,MAAM,GAAGoD,IAAI3H,IAAI,EAAE2H,IAAII,KAAK;wCAC9D;wCACApI,GAAGC,IAAI,CAAC;oCACV,EAAE,OAAOoI,KAAK;wCACZrI,GAAGC,IAAI,CAAC;wCACR,MAAMoI;oCACR;;;;;;oBACF;oBAxCiB;;wBAAMjC,YAAYC;;;oBAA7Be,WAAW;oBACXC,QAAQrH,GAAGM,OAAO,CAAC,oFAAoFgI,GAAG;oBAChH,IAAIjB,MAAM1I,MAAM,KAAK,GAAG;;;oBAClB2I,YAAYlE,KAAKmF,GAAG,CAACtK,YAAYmJ,SAAShF,IAAI;oBAE9CmF,SAAS,IAAIpB;oBACd,kCAAA,2BAAA;;wBAAL,IAAK,YAAakB,4BAAb,6BAAA,QAAA,yBAAA,iCAAoB;4BAAdG,MAAN;;4BACGC,QAAOF,cAAAA,OAAOZ,GAAG,CAACa,IAAInH,IAAI,eAAnBkH,yBAAAA;4BACbE,KAAK5I,IAAI,CAAC2I,IAAIY,KAAK;4BACnBb,OAAOX,GAAG,CAACY,IAAInH,IAAI,EAAEoH;wBACvB;;wBAJK;wBAAA;;;iCAAA,6BAAA;gCAAA;;;gCAAA;sCAAA;;;;oBAMCC;oBACD,mCAAA,4BAAA;;wBAAL,IAAK,aAA2BH,6BAA3B,8BAAA,SAAA,0BAAA,kCAAmC;2DAAnC,kBAAOlH,uBAAMsH;4BACZtI,SAAAA,KAAAA;4BACJ,IAAI;gCACFA,SAASmJ,IAAAA,iBAAS,EAAC;oCAAEC,SAASpI;oCAAMqI,SAAStJ,IAAAA,cAAI,EAAC+H,SAAS9G;oCAAOsI,SAAS;oCAAGC,MAAM;gCAAE;oCAAI9K;mCAAQ+K,GAAG,CAACpI,SAAS,CAAC3C,KAAK;4BACvH,EAAE,eAAM;gCACN,UAAU,gEAAgE;4BAC5E;4BACK,mCAAA,4BAAA;;gCAAL,IAAK,aAAa6J,gCAAb,8BAAA,SAAA,0BAAA;oCAAM/G,MAAN;oCAAwB,IAAIvB,MAAM,CAACuB,IAAI,EAAE8G,KAAK7I,IAAI,CAAC;wCAAEwB,MAAAA;wCAAM+H,OAAOxH;wCAAKnB,MAAMJ,MAAM,CAACuB,IAAI,CAACnB,IAAI;oCAAC;;;gCAA9F;gCAAA;;;yCAAA,8BAAA;wCAAA;;;wCAAA;8CAAA;;;;wBACP;;wBARK;wBAAA;;;iCAAA,8BAAA;gCAAA;;;gCAAA;sCAAA;;;;oBAUCmI,SAAS5H,GAAGM,OAAO,CAAC;oBACjB5B,IAAI;;;yBAAGA,CAAAA,IAAIgJ,KAAK/I,MAAM,AAAD;;;;;;;;;;;;oBAAGD,KAAKR;;;;;;;;;;;IAiBxC;;AAKO,SAAeF,mBAAmBgC,EAAgB,EAAEqG,GAAW,EAAEyC,KAAa,EAAE9H,MAAa;;YAOpF8H,cANR3B,SAIAC,UACAE,WACA7H,MACYoH,UAAPkC,IAELC,MAQAC,MACD,2BAAA,mBAAA,gBAAA,WAAA,OAAMzB,KACHS,GACFiB,KACKhG,GACHiG,OACAC;;;;oBAvBFjC,UAAU,AAACd,IAAgCc,OAAO;oBACxD,IAAI,CAACA,SAAS,MAAM,IAAI/F,oBAAU,CAAC,eAAe;oBAClD;;wBAAMrD,aAAaiC,IAAIqG,KAAKc;;;oBAA5B;oBAEiB;;wBAAMf,YAAYC;;;oBAA7Be,WAAW;oBACXE,YAAYlE,KAAKmF,GAAG,CAACtK,YAAYmJ,SAAShF,IAAI;oBAC9C3C,OAAO,EAACqJ,eAAAA,MAAMO,KAAK,CAAC,2naAAZP,0BAAAA,mBAAsC5J,MAAM,CAAC,SAACoK;+BAAM,CAAC;4BAAC;4BAAO;4BAAM;4BAAO;yBAAO,CAACC,QAAQ,CAACD;uBAAIlK,IAAI,CAAC;oBACvF;;wBAAMgI,SAAStJ,KAAK;4BAAE2B;;;;oBAA/BoH,WAAAA;wBAAS,aAA6B,CAAC,EAAE;wBAAES;wBAAW;wBAA7DyB,KAAOlC,SAAV9D;oBAEFiG,OAAOhJ,GAAGM,OAAO,CAAC,+FAA+FgI,GAAG;oBAQpHW,OAAO,IAAI9C;oBACZ,kCAAA,2BAAA;;wBAAL,IAAK,YAAa6C,2BAAb,6BAAA,QAAA,yBAAA,iCAAmB;4BAAbxB,MAAN;4BACGS,IAAI,IAAIC,UAAUV,IAAIgC,MAAM,CAAC5E,MAAM,EAAE4C,IAAIgC,MAAM,CAAC9E,UAAU,EAAEtB,KAAKmF,GAAG,CAACjB,WAAWE,IAAIgC,MAAM,CAACC,UAAU;4BACvGP,MAAM;4BACV,IAAShG,IAAI,GAAGA,IAAI+E,EAAEtJ,MAAM,EAAEuE,IAAKgG,OAAOjB,CAAC,CAAC/E,EAAE,GAAG6F,EAAE,CAAC7F,EAAE;4BAChDiG,QAAQD,MAAM1B,IAAIR,KAAK;4BACvBoC,WAAWH,KAAKtC,GAAG,CAACa,IAAInH,IAAI;4BAClC,IAAI,CAAC+I,YAAYD,QAAQC,SAASD,KAAK,EAAEF,KAAKrC,GAAG,CAACY,IAAInH,IAAI,EAAE;gCAAE8I,OAAAA;gCAAO7K,OAAO,AAAC,IAAqBkJ,OAAlBA,IAAIkC,UAAU,EAAC,KAAgB,OAAblC,IAAImC,QAAQ;4BAAG;wBACnH;;wBAPK;wBAAA;;;iCAAA,6BAAA;gCAAA;;;gCAAA;sCAAA;;;;oBAQL;;wBAAQ,qBAAGV,KAAK5E,OAAO,IACpBuF,IAAI,CAAC,SAACC,GAAGC;mCAAMA,CAAC,CAAC,EAAE,CAACX,KAAK,GAAGU,CAAC,CAAC,EAAE,CAACV,KAAK;2BACtCzJ,KAAK,CAAC,GAAGsB,QACTgE,GAAG,CAAC;qEAAE3E,kBAAMyJ;mCAAQ;gCAAEzJ,MAAAA;gCAAM/B,OAAOwL,EAAExL,KAAK;4BAAC;;;;;IAChD"}

package/dist/cjs/features/index.js

@@ -20,13 +20,15 @@ _export(exports, {
     }
 });
 var _configts = require("../config.js");
+var _embedts = require("./embed.js");
 var _linksts = require("./links.js");
 var _rankts = require("./rank.js");
 var _sectionsts = require("./sections.js");
 var FEATURES = [
     _linksts.links,
     _sectionsts.sections,
-    _rankts.rank
+    _rankts.rank,
+    _embedts.embed
 ];
 function activeFeatures(cfg) {
     return FEATURES.filter(function(feature) {

package/dist/cjs/features/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/index.ts"],"sourcesContent":["import type { Config } from '../config.ts';\nimport { featureEnabled } from '../config.ts';\nimport { links } from './links.ts';\nimport { rank } from './rank.ts';\nimport { sections } from './sections.ts';\nimport type { Feature } from './types.ts';\n\n// Registry order matters: rank reads the links table in afterReconcile.\nexport const FEATURES: Feature[] = [links, sections, rank];\n\nexport function activeFeatures(cfg: Config): Feature[] {\n  return FEATURES.filter((feature) => featureEnabled(cfg, feature.name));\n}\n\nexport { linkEdges } from './links.ts';\nexport type { Section } from './sections.ts';\nexport type { Feature } from './types.ts';\n"],"names":["FEATURES","activeFeatures","linkEdges","links","sections","rank","cfg","filter","feature","featureEnabled","name"],"mappings":";;;;;;;;;;;QAQaA;eAAAA;;QAEGC;eAAAA;;QAIPC;eAAAA,kBAAS;;;wBAba;uBACT;sBACD;0BACI;AAIlB,IAAMF,WAAsB;IAACG,cAAK;IAAEC,oBAAQ;IAAEC,YAAI;CAAC;AAEnD,SAASJ,eAAeK,GAAW;IACxC,OAAON,SAASO,MAAM,CAAC,SAACC;eAAYC,IAAAA,wBAAc,EAACH,KAAKE,QAAQE,IAAI;;AACtE"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/index.ts"],"sourcesContent":["import type { Config } from '../config.ts';\nimport { featureEnabled } from '../config.ts';\nimport { embed } from './embed.ts';\nimport { links } from './links.ts';\nimport { rank } from './rank.ts';\nimport { sections } from './sections.ts';\nimport type { Feature } from './types.ts';\n\n// Registry order matters: rank reads the links table in afterReconcile.\nexport const FEATURES: Feature[] = [links, sections, rank, embed];\n\nexport function activeFeatures(cfg: Config): Feature[] {\n  return FEATURES.filter((feature) => featureEnabled(cfg, feature.name));\n}\n\nexport { linkEdges } from './links.ts';\nexport type { Section } from './sections.ts';\nexport type { Feature } from './types.ts';\n"],"names":["FEATURES","activeFeatures","linkEdges","links","sections","rank","embed","cfg","filter","feature","featureEnabled","name"],"mappings":";;;;;;;;;;;QASaA;eAAAA;;QAEGC;eAAAA;;QAIPC;eAAAA,kBAAS;;;wBAda;uBACT;uBACA;sBACD;0BACI;AAIlB,IAAMF,WAAsB;IAACG,cAAK;IAAEC,oBAAQ;IAAEC,YAAI;IAAEC,cAAK;CAAC;AAE1D,SAASL,eAAeM,GAAW;IACxC,OAAOP,SAASQ,MAAM,CAAC,SAACC;eAAYC,IAAAA,wBAAc,EAACH,KAAKE,QAAQE,IAAI;;AACtE"}

package/dist/cjs/features/types.d.cts

@@ -4,7 +4,10 @@ import type { FileStat } from '../scan.js';
 export interface Feature {
     name: FeatureName;
     schema(db: DatabaseSync): void;
-    extract?(raw: string, body: string): unknown;
+    extract?(raw: string, body: string, search?: {
+        title: string;
+        summary: string;
+    }): unknown;
     remove?(db: DatabaseSync, path: string): void;
     store?(db: DatabaseSync, path: string, extracted: unknown): void;
     afterReconcile?(db: DatabaseSync, files: FileStat[]): void;

package/dist/cjs/features/types.d.ts

@@ -4,7 +4,10 @@ import type { FileStat } from '../scan.js';
 export interface Feature {
     name: FeatureName;
     schema(db: DatabaseSync): void;
-    extract?(raw: string, body: string): unknown;
+    extract?(raw: string, body: string, search?: {
+        title: string;
+        summary: string;
+    }): unknown;
     remove?(db: DatabaseSync, path: string): void;
     store?(db: DatabaseSync, path: string, extracted: unknown): void;
     afterReconcile?(db: DatabaseSync, files: FileStat[]): void;

package/dist/cjs/index.d.cts

@@ -1,18 +1,11 @@
-export type { Config, FeatureName, ResolvedConfig } from './config.js';
-export { CONFIG_FILENAME, enabledFeatures, FEATURE_NAMES, featureEnabled, findConfigPath, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
+export type { Config, EmbedConfig, FeatureName, ResolvedConfig } from './config.js';
+export { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
 export type { OpenResult } from './db.js';
-export { DB_FILENAME, docCount, getMeta, open, rebuild, reconcile, setMeta } from './db.js';
+export { open, rebuild } from './db.js';
 export type { SenseErrorCode } from './errors.js';
 export { SenseError } from './errors.js';
-export { activeFeatures, FEATURES, linkEdges } from './features/index.js';
-export type { Section } from './features/sections.js';
-export type { Feature } from './features/types.js';
-export type { Edge } from './graph.js';
-export { pagerank, personalizedRank } from './graph.js';
 export type { Row } from './output.js';
 export { printRows } from './output.js';
-export type { FileStat, ParsedDoc } from './scan.js';
-export { listFiles, parseFile } from './scan.js';
 export type { FindOptions, Peek, TreeMap } from './verbs.js';
 export { find, mapTree, peek } from './verbs.js';
 export type { WatchEvent, WatchOptions } from './watch.js';

package/dist/cjs/index.d.ts

@@ -1,18 +1,11 @@
-export type { Config, FeatureName, ResolvedConfig } from './config.js';
-export { CONFIG_FILENAME, enabledFeatures, FEATURE_NAMES, featureEnabled, findConfigPath, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
+export type { Config, EmbedConfig, FeatureName, ResolvedConfig } from './config.js';
+export { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
 export type { OpenResult } from './db.js';
-export { DB_FILENAME, docCount, getMeta, open, rebuild, reconcile, setMeta } from './db.js';
+export { open, rebuild } from './db.js';
 export type { SenseErrorCode } from './errors.js';
 export { SenseError } from './errors.js';
-export { activeFeatures, FEATURES, linkEdges } from './features/index.js';
-export type { Section } from './features/sections.js';
-export type { Feature } from './features/types.js';
-export type { Edge } from './graph.js';
-export { pagerank, personalizedRank } from './graph.js';
 export type { Row } from './output.js';
 export { printRows } from './output.js';
-export type { FileStat, ParsedDoc } from './scan.js';
-export { listFiles, parseFile } from './scan.js';
 export type { FindOptions, Peek, TreeMap } from './verbs.js';
 export { find, mapTree, peek } from './verbs.js';
 export type { WatchEvent, WatchOptions } from './watch.js';

package/dist/cjs/index.js

@@ -1,4 +1,5 @@
-// Public library API.
+// Public library API. Deliberately small: every export is a stability promise;
+// internals (feature registry, graph, scan, meta) stay module-private.
 "use strict";
 Object.defineProperty(exports, "__esModule", {
     value: true
@@ -13,15 +14,6 @@ _export(exports, {
     get CONFIG_FILENAME () {
         return _configts.CONFIG_FILENAME;
     },
-    get DB_FILENAME () {
-        return _dbts.DB_FILENAME;
-    },
-    get FEATURES () {
-        return _indexts.FEATURES;
-    },
-    get FEATURE_NAMES () {
-        return _configts.FEATURE_NAMES;
-    },
     get STATE_DIR () {
         return _configts.STATE_DIR;
     },
@@ -31,36 +23,12 @@ _export(exports, {
     get SenseError () {
         return _errorsts.SenseError;
     },
-    get activeFeatures () {
-        return _indexts.activeFeatures;
-    },
-    get docCount () {
-        return _dbts.docCount;
-    },
-    get enabledFeatures () {
-        return _configts.enabledFeatures;
-    },
-    get featureEnabled () {
-        return _configts.featureEnabled;
-    },
     get find () {
         return _verbsts.find;
     },
-    get findConfigPath () {
-        return _configts.findConfigPath;
-    },
-    get getMeta () {
-        return _dbts.getMeta;
-    },
     get initConfig () {
         return _configts.initConfig;
     },
-    get linkEdges () {
-        return _indexts.linkEdges;
-    },
-    get listFiles () {
-        return _scants.listFiles;
-    },
     get loadConfig () {
         return _configts.loadConfig;
     },
@@ -73,41 +41,23 @@ _export(exports, {
     get open () {
         return _dbts.open;
     },
-    get pagerank () {
-        return _graphts.pagerank;
-    },
-    get parseFile () {
-        return _scants.parseFile;
-    },
     get peek () {
         return _verbsts.peek;
     },
-    get personalizedRank () {
-        return _graphts.personalizedRank;
-    },
     get printRows () {
         return _outputts.printRows;
     },
     get rebuild () {
         return _dbts.rebuild;
     },
-    get reconcile () {
-        return _dbts.reconcile;
-    },
     get runWatch () {
         return _watchts.runWatch;
-    },
-    get setMeta () {
-        return _dbts.setMeta;
     }
 });
 var _configts = require("./config.js");
 var _dbts = require("./db.js");
 var _errorsts = require("./errors.js");
-var _indexts = require("./features/index.js");
-var _graphts = require("./graph.js");
 var _outputts = require("./output.js");
-var _scants = require("./scan.js");
 var _verbsts = require("./verbs.js");
 var _watchts = require("./watch.js");
 /* CJS INTEROP */ if (exports.__esModule && exports.default) { try { Object.defineProperty(exports.default, '__esModule', { value: true }); for (var key in exports) { exports.default[key] = exports[key]; } } catch (_) {}; module.exports = exports.default; }

package/dist/cjs/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/index.ts"],"sourcesContent":["// Public library API.\n\nexport type { Config, FeatureName, ResolvedConfig } from './config.ts';\nexport { CONFIG_FILENAME, enabledFeatures, FEATURE_NAMES, featureEnabled, findConfigPath, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.ts';\n\nexport type { OpenResult } from './db.ts';\nexport { DB_FILENAME, docCount, getMeta, open, rebuild, reconcile, setMeta } from './db.ts';\nexport type { SenseErrorCode } from './errors.ts';\nexport { SenseError } from './errors.ts';\nexport { activeFeatures, FEATURES, linkEdges } from './features/index.ts';\nexport type { Section } from './features/sections.ts';\nexport type { Feature } from './features/types.ts';\n\nexport type { Edge } from './graph.ts';\nexport { pagerank, personalizedRank } from './graph.ts';\n\nexport type { Row } from './output.ts';\nexport { printRows } from './output.ts';\n\nexport type { FileStat, ParsedDoc } from './scan.ts';\nexport { listFiles, parseFile } from './scan.ts';\n\nexport type { FindOptions, Peek, TreeMap } from './verbs.ts';\nexport { find, mapTree, peek } from './verbs.ts';\n\nexport type { WatchEvent, WatchOptions } from './watch.ts';\nexport { runWatch } from './watch.ts';\n"],"names":["CONFIG_FILENAME","DB_FILENAME","FEATURES","FEATURE_NAMES","STATE_DIR","SUPPORTED_CONFIG_VERSION","SenseError","activeFeatures","docCount","enabledFeatures","featureEnabled","find","findConfigPath","getMeta","initConfig","linkEdges","listFiles","loadConfig","mapTree","migrateConfig","open","pagerank","parseFile","peek","personalizedRank","printRows","rebuild","reconcile","runWatch","setMeta"],"mappings":"AAAA,sBAAsB;;;;;;;;;;;;QAGbA;eAAAA,yBAAe;;QAGfC;eAAAA,iBAAW;;QAGKC;eAAAA,iBAAQ;;QANUC;eAAAA,uBAAa;;QAAyEC;eAAAA,mBAAS;;QAAEC;eAAAA,kCAAwB;;QAK3JC;eAAAA,oBAAU;;QACVC;eAAAA,uBAAc;;QAHDC;eAAAA,cAAQ;;QAHJC;eAAAA,yBAAe;;QAAiBC;eAAAA,wBAAc;;QAoB/DC;eAAAA,aAAI;;QApB6DC;eAAAA,wBAAc;;QAGxDC;eAAAA,aAAO;;QAHmDC;eAAAA,oBAAU;;QAMjEC;eAAAA,kBAAS;;QAWnCC;eAAAA,iBAAS;;QAjBoFC;eAAAA,oBAAU;;QAoBjGC;eAAAA,gBAAO;;QApB4FC;eAAAA,uBAAa;;QAGtFC;eAAAA,UAAI;;QAQpCC;eAAAA,iBAAQ;;QAMGC;eAAAA,iBAAS;;QAGLC;eAAAA,aAAI;;QATTC;eAAAA,yBAAgB;;QAG1BC;eAAAA,mBAAS;;QAX6BC;eAAAA,aAAO;;QAAEC;eAAAA,eAAS;;QAoBxDC;eAAAA,iBAAQ;;QApBkDC;eAAAA,aAAO;;;wBAHkG;oBAG1F;wBAEvD;uBACyB;uBAKT;wBAGjB;sBAGW;uBAGD;uBAGX"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/index.ts"],"sourcesContent":["// Public library API. Deliberately small: every export is a stability promise;\n// internals (feature registry, graph, scan, meta) stay module-private.\n\nexport type { Config, EmbedConfig, FeatureName, ResolvedConfig } from './config.ts';\nexport { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.ts';\n\nexport type { OpenResult } from './db.ts';\nexport { open, rebuild } from './db.ts';\n\nexport type { SenseErrorCode } from './errors.ts';\nexport { SenseError } from './errors.ts';\n\nexport type { Row } from './output.ts';\nexport { printRows } from './output.ts';\n\nexport type { FindOptions, Peek, TreeMap } from './verbs.ts';\nexport { find, mapTree, peek } from './verbs.ts';\n\nexport type { WatchEvent, WatchOptions } from './watch.ts';\nexport { runWatch } from './watch.ts';\n"],"names":["CONFIG_FILENAME","STATE_DIR","SUPPORTED_CONFIG_VERSION","SenseError","find","initConfig","loadConfig","mapTree","migrateConfig","open","peek","printRows","rebuild","runWatch"],"mappings":"AAAA,+EAA+E;AAC/E,uEAAuE;;;;;;;;;;;;QAG9DA;eAAAA,yBAAe;;QAAyCC;eAAAA,mBAAS;;QAAEC;eAAAA,kCAAwB;;QAM3FC;eAAAA,oBAAU;;QAMVC;eAAAA,aAAI;;QAZaC;eAAAA,oBAAU;;QAAEC;eAAAA,oBAAU;;QAYjCC;eAAAA,gBAAO;;QAZ4BC;eAAAA,uBAAa;;QAGtDC;eAAAA,UAAI;;QASWC;eAAAA,aAAI;;QAHnBC;eAAAA,mBAAS;;QANHC;eAAAA,aAAO;;QAYbC;eAAAA,iBAAQ;;;wBAf2F;oBAG9E;wBAGH;wBAGD;uBAGU;uBAGX"}

package/dist/cjs/output.d.cts

@@ -7,6 +7,10 @@ export declare function renderMap(result: {
     };
     fields: Row[];
     fieldsTotal: number;
+    features: {
+        on: string[];
+        off: string[];
+    };
     hubs: Row[];
     recent: Row[];
 }): string;
@@ -21,4 +25,5 @@ export declare function renderPeek(result: {
     outboundTotal: number;
     backlinksTotal: number;
     unresolvedTotal: number;
+    off: string[];
 }): string;

package/dist/cjs/output.d.ts

@@ -7,6 +7,10 @@ export declare function renderMap(result: {
     };
     fields: Row[];
     fieldsTotal: number;
+    features: {
+        on: string[];
+        off: string[];
+    };
     hubs: Row[];
     recent: Row[];
 }): string;
@@ -21,4 +25,5 @@ export declare function renderPeek(result: {
     outboundTotal: number;
     backlinksTotal: number;
     unresolvedTotal: number;
+    off: string[];
 }): string;

package/dist/cjs/output.js

@@ -120,13 +120,22 @@ function render(rows, format) {
         }).join('  ')
     ].concat(_to_consumable_array(cells.map(formatRow))).join('\n');
 }
+// Text renderers for the layer verbs; cli.ts prints what these return.
+function featuresLine(features) {
+    var off = features.off.length > 0 ? " \xb7 off: ".concat(features.off.map(function(name) {
+        return "".concat(name, " (features.").concat(name, ")");
+    }).join(', ')) : '';
+    return "features: ".concat(features.on.join(', ')).concat(off);
+}
 function renderMap(result) {
     var parts = [
-        "docs: ".concat(result.docs.count, " (").concat(Math.round(result.docs.bytes / 1024), " KB)\n"),
+        "docs: ".concat(result.docs.count, " (").concat(Math.round(result.docs.bytes / 1024), " KB)"),
+        "".concat(featuresLine(result.features), "\n"),
         render(result.fields, 'table')
     ];
     if (result.fieldsTotal > result.fields.length) parts.push("(+".concat(result.fieldsTotal - result.fields.length, " more fields)"));
     if (result.hubs.length > 0) parts.push('\nhubs (by link rank):', render(result.hubs, 'table'));
+    else if (result.features.off.includes('rank')) parts.push('\nhubs: off (features.rank)');
     parts.push('\nrecent:', render(result.recent, 'table'));
     return parts.join('\n');
 }
@@ -176,6 +185,10 @@ function renderPeek(result) {
                 }
             }
         }
+    } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');
+    if (result.off.includes('links')) {
+        lines.push('', 'links: off (features.links)');
+        return lines.join('\n');
     }
     var linkLine = function linkLine(label, shown, total) {
         if (total > 0) lines.push("".concat(label, " (").concat(total, "): ").concat(shown.join(', ')).concat(total > shown.length ? ", +".concat(total - shown.length, " more") : ''));

package/dist/cjs/output.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/output.ts"],"sourcesContent":["// rows -> table (default) | json\n\nexport type Row = Record<string, unknown>;\n\n// Warned about on stderr (keeps --format json stdout machine-readable), never truncated.\nconst OVERSIZE_BYTES = 50_000;\n\n// Flatten embedded newlines so they can't break table column alignment; JSON output is left alone.\nfunction cell(value: unknown): string {\n  return String(value ?? '').replace(/\\s*\\n\\s*/g, ' ');\n}\n\nexport function printRows(rows: Row[], format: 'table' | 'json'): void {\n  const rendered = render(rows, format);\n  console.log(rendered);\n  if (rendered.length > OVERSIZE_BYTES) {\n    console.warn(`warning: result is ${Math.round(rendered.length / 1000)} KB across ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);\n  }\n}\n\nfunction render(rows: Row[], format: 'table' | 'json'): string {\n  if (format === 'json') return JSON.stringify(rows, null, 2);\n  if (rows.length === 0) return '(0 rows)';\n\n  const columns = Object.keys(rows[0]);\n  const cells = rows.map((row) => columns.map((col) => cell(row[col])));\n  const widths = columns.map((col, i) => Math.max(col.length, ...cells.map((row) => row[i].length)));\n\n  const formatRow = (values: string[]) => values.map((value, i) => value.padEnd(widths[i])).join('  ');\n\n  return [formatRow(columns), widths.map((w) => '-'.repeat(w)).join('  '), ...cells.map(formatRow)].join('\\n');\n}\n\n// Text renderers for the layer verbs; cli.ts prints what these return.\n\nexport function renderMap(result: { docs: { count: number; bytes: number }; fields: Row[]; fieldsTotal: number; hubs: Row[]; recent: Row[] }): string {\n  const parts = [`docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)\\n`, render(result.fields, 'table')];\n  if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);\n  if (result.hubs.length > 0) parts.push('\\nhubs (by link rank):', render(result.hubs, 'table'));\n  parts.push('\\nrecent:', render(result.recent, 'table'));\n  return parts.join('\\n');\n}\n\nexport function renderPeek(result: { path: string; tokens: number; frontmatter: Row; sections: Row[]; outbound: string[]; backlinks: string[]; unresolved: string[]; outboundTotal: number; backlinksTotal: number; unresolvedTotal: number }): string {\n  const lines = [`${result.path}  (~${result.tokens} tokens)`];\n  for (const [key, value] of Object.entries(result.frontmatter)) lines.push(`  ${key}: ${value}`);\n  if (result.sections.length > 0) {\n    lines.push('', 'sections:');\n    for (const s of result.sections) lines.push(`  ${'#'.repeat(s.level as number)} ${s.heading}  [L${s.start_line}-${s.end_line}, ~${s.tokens}t]`);\n  }\n  const linkLine = (label: string, shown: string[], total: number) => {\n    if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);\n  };\n  if (result.outboundTotal + result.unresolvedTotal + result.backlinksTotal > 0) lines.push('');\n  linkLine('links out', result.outbound, result.outboundTotal);\n  linkLine('unresolved', result.unresolved, result.unresolvedTotal);\n  linkLine('backlinks', result.backlinks, result.backlinksTotal);\n  return lines.join('\\n');\n}\n"],"names":["printRows","renderMap","renderPeek","OVERSIZE_BYTES","cell","value","String","replace","rows","format","rendered","render","console","log","length","warn","Math","round","JSON","stringify","columns","Object","keys","cells","map","row","col","widths","i","max","formatRow","values","padEnd","join","w","repeat","result","parts","docs","count","bytes","fields","fieldsTotal","push","hubs","recent","lines","path","tokens","entries","frontmatter","key","sections","s","level","heading","start_line","end_line","linkLine","label","shown","total","outboundTotal","unresolvedTotal","backlinksTotal","outbound","unresolved","backlinks"],"mappings":"AAAA,iCAAiC;;;;;;;;;;;;QAYjBA;eAAAA;;QAuBAC;eAAAA;;QAQAC;eAAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAvChB,yFAAyF;AACzF,IAAMC,iBAAiB;AAEvB,mGAAmG;AACnG,SAASC,KAAKC,KAAc;IAC1B,OAAOC,OAAOD,kBAAAA,mBAAAA,QAAS,IAAIE,OAAO,CAAC,aAAa;AAClD;AAEO,SAASP,UAAUQ,IAAW,EAAEC,MAAwB;IAC7D,IAAMC,WAAWC,OAAOH,MAAMC;IAC9BG,QAAQC,GAAG,CAACH;IACZ,IAAIA,SAASI,MAAM,GAAGX,gBAAgB;QACpCS,QAAQG,IAAI,CAAC,AAAC,sBAAqEP,OAAhDQ,KAAKC,KAAK,CAACP,SAASI,MAAM,GAAG,OAAM,eAAyB,OAAZN,KAAKM,MAAM,EAAC;IACjG;AACF;AAEA,SAASH,OAAOH,IAAW,EAAEC,MAAwB;QAMZO;IALvC,IAAIP,WAAW,QAAQ,OAAOS,KAAKC,SAAS,CAACX,MAAM,MAAM;IACzD,IAAIA,KAAKM,MAAM,KAAK,GAAG,OAAO;IAE9B,IAAMM,UAAUC,OAAOC,IAAI,CAACd,IAAI,CAAC,EAAE;IACnC,IAAMe,QAAQf,KAAKgB,GAAG,CAAC,SAACC;eAAQL,QAAQI,GAAG,CAAC,SAACE;mBAAQtB,KAAKqB,GAAG,CAACC,IAAI;;;IAClE,IAAMC,SAASP,QAAQI,GAAG,CAAC,SAACE,KAAKE;eAAMZ,CAAAA,QAAAA,MAAKa,GAAG,OAARb,OAAAA;YAASU,IAAIZ,MAAM;SAAuC,CAA1DE,OAAqB,qBAAGO,MAAMC,GAAG,CAAC,SAACC;mBAAQA,GAAG,CAACG,EAAE,CAACd,MAAM;;;IAE/F,IAAMgB,YAAY,mBAACC;eAAqBA,OAAOP,GAAG,CAAC,SAACnB,OAAOuB;mBAAMvB,MAAM2B,MAAM,CAACL,MAAM,CAACC,EAAE;WAAGK,IAAI,CAAC;;IAE/F,OAAO;QAACH,UAAUV;QAAUO,OAAOH,GAAG,CAAC,SAACU;mBAAM,IAAIC,MAAM,CAACD;WAAID,IAAI,CAAC;KAA+B,CAA1F,OAAkE,qBAAGV,MAAMC,GAAG,CAACM,aAAYG,IAAI,CAAC;AACzG;AAIO,SAAShC,UAAUmC,MAAkH;IAC1I,IAAMC,QAAQ;QAAE,SAA8BrB,OAAtBoB,OAAOE,IAAI,CAACC,KAAK,EAAC,MAAyC,OAArCvB,KAAKC,KAAK,CAACmB,OAAOE,IAAI,CAACE,KAAK,GAAG,OAAM;QAAS7B,OAAOyB,OAAOK,MAAM,EAAE;KAAS;IAC3H,IAAIL,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC3B,MAAM,EAAEuB,MAAMM,IAAI,CAAC,AAAC,KAA8C,OAA1CP,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC3B,MAAM,EAAC;IACzG,IAAIsB,OAAOQ,IAAI,CAAC9B,MAAM,GAAG,GAAGuB,MAAMM,IAAI,CAAC,0BAA0BhC,OAAOyB,OAAOQ,IAAI,EAAE;IACrFP,MAAMM,IAAI,CAAC,aAAahC,OAAOyB,OAAOS,MAAM,EAAE;IAC9C,OAAOR,MAAMJ,IAAI,CAAC;AACpB;AAEO,SAAS/B,WAAWkC,MAAkN;IAC3O,IAAMU,QAAQ;QAAE,GAAoBV,OAAlBA,OAAOW,IAAI,EAAC,QAAoB,OAAdX,OAAOY,MAAM,EAAC;KAAU;QACvD,kCAAA,2BAAA;;QAAL,QAAK,YAAsB3B,OAAO4B,OAAO,CAACb,OAAOc,WAAW,sBAAvD,SAAA,6BAAA,QAAA,yBAAA;YAAA,mCAAA,iBAAOC,sBAAK9C;YAA8CyC,MAAMH,IAAI,CAAC,AAAC,KAAYtC,OAAR8C,KAAI,MAAU,OAAN9C;;;QAAlF;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IACL,IAAI+B,OAAOgB,QAAQ,CAACtC,MAAM,GAAG,GAAG;QAC9BgC,MAAMH,IAAI,CAAC,IAAI;YACV,mCAAA,4BAAA;;YAAL,QAAK,aAAWP,OAAOgB,QAAQ,qBAA1B,UAAA,8BAAA,SAAA,0BAAA;gBAAA,IAAMC,IAAN;gBAA4BP,MAAMH,IAAI,CAAC,AAAC,KAAqCU,OAAjC,IAAIlB,MAAM,CAACkB,EAAEC,KAAK,GAAY,KAAmBD,OAAhBA,EAAEE,OAAO,EAAC,QAAsBF,OAAhBA,EAAEG,UAAU,EAAC,KAAmBH,OAAhBA,EAAEI,QAAQ,EAAC,OAAc,OAATJ,EAAEL,MAAM,EAAC;;;YAAtI;YAAA;;;qBAAA,8BAAA;oBAAA;;;oBAAA;0BAAA;;;;IACP;IACA,IAAMU,WAAW,kBAACC,OAAeC,OAAiBC;QAChD,IAAIA,QAAQ,GAAGf,MAAMH,IAAI,CAAC,AAAC,GAAYkB,OAAVF,OAAM,MAAeC,OAAXC,OAAM,OAAwBA,OAAnBD,MAAM3B,IAAI,CAAC,OAAsE,OAA9D4B,QAAQD,MAAM9C,MAAM,GAAG,AAAC,MAA0B,OAArB+C,QAAQD,MAAM9C,MAAM,EAAC,WAAS;IAClI;IACA,IAAIsB,OAAO0B,aAAa,GAAG1B,OAAO2B,eAAe,GAAG3B,OAAO4B,cAAc,GAAG,GAAGlB,MAAMH,IAAI,CAAC;IAC1Fe,SAAS,aAAatB,OAAO6B,QAAQ,EAAE7B,OAAO0B,aAAa;IAC3DJ,SAAS,cAActB,OAAO8B,UAAU,EAAE9B,OAAO2B,eAAe;IAChEL,SAAS,aAAatB,OAAO+B,SAAS,EAAE/B,OAAO4B,cAAc;IAC7D,OAAOlB,MAAMb,IAAI,CAAC;AACpB"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/output.ts"],"sourcesContent":["// rows -> table (default) | json\n\nexport type Row = Record<string, unknown>;\n\n// Warned about on stderr (keeps --format json stdout machine-readable), never truncated.\nconst OVERSIZE_BYTES = 50_000;\n\n// Flatten embedded newlines so they can't break table column alignment; JSON output is left alone.\nfunction cell(value: unknown): string {\n  return String(value ?? '').replace(/\\s*\\n\\s*/g, ' ');\n}\n\nexport function printRows(rows: Row[], format: 'table' | 'json'): void {\n  const rendered = render(rows, format);\n  console.log(rendered);\n  if (rendered.length > OVERSIZE_BYTES) {\n    console.warn(`warning: result is ${Math.round(rendered.length / 1000)} KB across ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);\n  }\n}\n\nfunction render(rows: Row[], format: 'table' | 'json'): string {\n  if (format === 'json') return JSON.stringify(rows, null, 2);\n  if (rows.length === 0) return '(0 rows)';\n\n  const columns = Object.keys(rows[0]);\n  const cells = rows.map((row) => columns.map((col) => cell(row[col])));\n  const widths = columns.map((col, i) => Math.max(col.length, ...cells.map((row) => row[i].length)));\n\n  const formatRow = (values: string[]) => values.map((value, i) => value.padEnd(widths[i])).join('  ');\n\n  return [formatRow(columns), widths.map((w) => '-'.repeat(w)).join('  '), ...cells.map(formatRow)].join('\\n');\n}\n\n// Text renderers for the layer verbs; cli.ts prints what these return.\n\nfunction featuresLine(features: { on: string[]; off: string[] }): string {\n  const off = features.off.length > 0 ? ` · off: ${features.off.map((name) => `${name} (features.${name})`).join(', ')}` : '';\n  return `features: ${features.on.join(', ')}${off}`;\n}\n\nexport function renderMap(result: { docs: { count: number; bytes: number }; fields: Row[]; fieldsTotal: number; features: { on: string[]; off: string[] }; hubs: Row[]; recent: Row[] }): string {\n  const parts = [`docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)`, `${featuresLine(result.features)}\\n`, render(result.fields, 'table')];\n  if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);\n  if (result.hubs.length > 0) parts.push('\\nhubs (by link rank):', render(result.hubs, 'table'));\n  else if (result.features.off.includes('rank')) parts.push('\\nhubs: off (features.rank)');\n  parts.push('\\nrecent:', render(result.recent, 'table'));\n  return parts.join('\\n');\n}\n\nexport function renderPeek(result: { path: string; tokens: number; frontmatter: Row; sections: Row[]; outbound: string[]; backlinks: string[]; unresolved: string[]; outboundTotal: number; backlinksTotal: number; unresolvedTotal: number; off: string[] }): string {\n  const lines = [`${result.path}  (~${result.tokens} tokens)`];\n  for (const [key, value] of Object.entries(result.frontmatter)) lines.push(`  ${key}: ${value}`);\n  if (result.sections.length > 0) {\n    lines.push('', 'sections:');\n    for (const s of result.sections) lines.push(`  ${'#'.repeat(s.level as number)} ${s.heading}  [L${s.start_line}-${s.end_line}, ~${s.tokens}t]`);\n  } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');\n  if (result.off.includes('links')) {\n    lines.push('', 'links: off (features.links)');\n    return lines.join('\\n');\n  }\n  const linkLine = (label: string, shown: string[], total: number) => {\n    if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);\n  };\n  if (result.outboundTotal + result.unresolvedTotal + result.backlinksTotal > 0) lines.push('');\n  linkLine('links out', result.outbound, result.outboundTotal);\n  linkLine('unresolved', result.unresolved, result.unresolvedTotal);\n  linkLine('backlinks', result.backlinks, result.backlinksTotal);\n  return lines.join('\\n');\n}\n"],"names":["printRows","renderMap","renderPeek","OVERSIZE_BYTES","cell","value","String","replace","rows","format","rendered","render","console","log","length","warn","Math","round","JSON","stringify","columns","Object","keys","cells","map","row","col","widths","i","max","formatRow","values","padEnd","join","w","repeat","featuresLine","features","off","name","on","result","parts","docs","count","bytes","fields","fieldsTotal","push","hubs","includes","recent","lines","path","tokens","entries","frontmatter","key","sections","s","level","heading","start_line","end_line","linkLine","label","shown","total","outboundTotal","unresolvedTotal","backlinksTotal","outbound","unresolved","backlinks"],"mappings":"AAAA,iCAAiC;;;;;;;;;;;;QAYjBA;eAAAA;;QA4BAC;eAAAA;;QASAC;eAAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AA7ChB,yFAAyF;AACzF,IAAMC,iBAAiB;AAEvB,mGAAmG;AACnG,SAASC,KAAKC,KAAc;IAC1B,OAAOC,OAAOD,kBAAAA,mBAAAA,QAAS,IAAIE,OAAO,CAAC,aAAa;AAClD;AAEO,SAASP,UAAUQ,IAAW,EAAEC,MAAwB;IAC7D,IAAMC,WAAWC,OAAOH,MAAMC;IAC9BG,QAAQC,GAAG,CAACH;IACZ,IAAIA,SAASI,MAAM,GAAGX,gBAAgB;QACpCS,QAAQG,IAAI,CAAC,AAAC,sBAAqEP,OAAhDQ,KAAKC,KAAK,CAACP,SAASI,MAAM,GAAG,OAAM,eAAyB,OAAZN,KAAKM,MAAM,EAAC;IACjG;AACF;AAEA,SAASH,OAAOH,IAAW,EAAEC,MAAwB;QAMZO;IALvC,IAAIP,WAAW,QAAQ,OAAOS,KAAKC,SAAS,CAACX,MAAM,MAAM;IACzD,IAAIA,KAAKM,MAAM,KAAK,GAAG,OAAO;IAE9B,IAAMM,UAAUC,OAAOC,IAAI,CAACd,IAAI,CAAC,EAAE;IACnC,IAAMe,QAAQf,KAAKgB,GAAG,CAAC,SAACC;eAAQL,QAAQI,GAAG,CAAC,SAACE;mBAAQtB,KAAKqB,GAAG,CAACC,IAAI;;;IAClE,IAAMC,SAASP,QAAQI,GAAG,CAAC,SAACE,KAAKE;eAAMZ,CAAAA,QAAAA,MAAKa,GAAG,OAARb,OAAAA;YAASU,IAAIZ,MAAM;SAAuC,CAA1DE,OAAqB,qBAAGO,MAAMC,GAAG,CAAC,SAACC;mBAAQA,GAAG,CAACG,EAAE,CAACd,MAAM;;;IAE/F,IAAMgB,YAAY,mBAACC;eAAqBA,OAAOP,GAAG,CAAC,SAACnB,OAAOuB;mBAAMvB,MAAM2B,MAAM,CAACL,MAAM,CAACC,EAAE;WAAGK,IAAI,CAAC;;IAE/F,OAAO;QAACH,UAAUV;QAAUO,OAAOH,GAAG,CAAC,SAACU;mBAAM,IAAIC,MAAM,CAACD;WAAID,IAAI,CAAC;KAA+B,CAA1F,OAAkE,qBAAGV,MAAMC,GAAG,CAACM,aAAYG,IAAI,CAAC;AACzG;AAEA,uEAAuE;AAEvE,SAASG,aAAaC,QAAyC;IAC7D,IAAMC,MAAMD,SAASC,GAAG,CAACxB,MAAM,GAAG,IAAI,AAAC,cAA8E,OAApEuB,SAASC,GAAG,CAACd,GAAG,CAAC,SAACe;eAAS,AAAC,GAAoBA,OAAlBA,MAAK,eAAkB,OAALA,MAAK;OAAIN,IAAI,CAAC,SAAU;IACzH,OAAO,AAAC,aAAqCK,OAAzBD,SAASG,EAAE,CAACP,IAAI,CAAC,OAAY,OAAJK;AAC/C;AAEO,SAASrC,UAAUwC,MAA6J;IACrL,IAAMC,QAAQ;QAAE,SAA8B1B,OAAtByB,OAAOE,IAAI,CAACC,KAAK,EAAC,MAAyC,OAArC5B,KAAKC,KAAK,CAACwB,OAAOE,IAAI,CAACE,KAAK,GAAG,OAAM;QAAQ,GAAgC,OAA9BT,aAAaK,OAAOJ,QAAQ,GAAE;QAAK1B,OAAO8B,OAAOK,MAAM,EAAE;KAAS;IAC/J,IAAIL,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAChC,MAAM,EAAE4B,MAAMM,IAAI,CAAC,AAAC,KAA8C,OAA1CP,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAChC,MAAM,EAAC;IACzG,IAAI2B,OAAOQ,IAAI,CAACnC,MAAM,GAAG,GAAG4B,MAAMM,IAAI,CAAC,0BAA0BrC,OAAO8B,OAAOQ,IAAI,EAAE;SAChF,IAAIR,OAAOJ,QAAQ,CAACC,GAAG,CAACY,QAAQ,CAAC,SAASR,MAAMM,IAAI,CAAC;IAC1DN,MAAMM,IAAI,CAAC,aAAarC,OAAO8B,OAAOU,MAAM,EAAE;IAC9C,OAAOT,MAAMT,IAAI,CAAC;AACpB;AAEO,SAAS/B,WAAWuC,MAAiO;IAC1P,IAAMW,QAAQ;QAAE,GAAoBX,OAAlBA,OAAOY,IAAI,EAAC,QAAoB,OAAdZ,OAAOa,MAAM,EAAC;KAAU;QACvD,kCAAA,2BAAA;;QAAL,QAAK,YAAsBjC,OAAOkC,OAAO,CAACd,OAAOe,WAAW,sBAAvD,SAAA,6BAAA,QAAA,yBAAA;YAAA,mCAAA,iBAAOC,sBAAKpD;YAA8C+C,MAAMJ,IAAI,CAAC,AAAC,KAAY3C,OAARoD,KAAI,MAAU,OAANpD;;;QAAlF;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IACL,IAAIoC,OAAOiB,QAAQ,CAAC5C,MAAM,GAAG,GAAG;QAC9BsC,MAAMJ,IAAI,CAAC,IAAI;YACV,mCAAA,4BAAA;;YAAL,QAAK,aAAWP,OAAOiB,QAAQ,qBAA1B,UAAA,8BAAA,SAAA,0BAAA;gBAAA,IAAMC,IAAN;gBAA4BP,MAAMJ,IAAI,CAAC,AAAC,KAAqCW,OAAjC,IAAIxB,MAAM,CAACwB,EAAEC,KAAK,GAAY,KAAmBD,OAAhBA,EAAEE,OAAO,EAAC,QAAsBF,OAAhBA,EAAEG,UAAU,EAAC,KAAmBH,OAAhBA,EAAEI,QAAQ,EAAC,OAAc,OAATJ,EAAEL,MAAM,EAAC;;;YAAtI;YAAA;;;qBAAA,8BAAA;oBAAA;;;oBAAA;0BAAA;;;;IACP,OAAO,IAAIb,OAAOH,GAAG,CAACY,QAAQ,CAAC,aAAaE,MAAMJ,IAAI,CAAC,IAAI;IAC3D,IAAIP,OAAOH,GAAG,CAACY,QAAQ,CAAC,UAAU;QAChCE,MAAMJ,IAAI,CAAC,IAAI;QACf,OAAOI,MAAMnB,IAAI,CAAC;IACpB;IACA,IAAM+B,WAAW,kBAACC,OAAeC,OAAiBC;QAChD,IAAIA,QAAQ,GAAGf,MAAMJ,IAAI,CAAC,AAAC,GAAYmB,OAAVF,OAAM,MAAeC,OAAXC,OAAM,OAAwBA,OAAnBD,MAAMjC,IAAI,CAAC,OAAsE,OAA9DkC,QAAQD,MAAMpD,MAAM,GAAG,AAAC,MAA0B,OAArBqD,QAAQD,MAAMpD,MAAM,EAAC,WAAS;IAClI;IACA,IAAI2B,OAAO2B,aAAa,GAAG3B,OAAO4B,eAAe,GAAG5B,OAAO6B,cAAc,GAAG,GAAGlB,MAAMJ,IAAI,CAAC;IAC1FgB,SAAS,aAAavB,OAAO8B,QAAQ,EAAE9B,OAAO2B,aAAa;IAC3DJ,SAAS,cAAcvB,OAAO+B,UAAU,EAAE/B,OAAO4B,eAAe;IAChEL,SAAS,aAAavB,OAAOgC,SAAS,EAAEhC,OAAO6B,cAAc;IAC7D,OAAOlB,MAAMnB,IAAI,CAAC;AACpB"}

package/dist/cjs/scan.js

@@ -144,26 +144,27 @@ function parseFile(file) {
             }
         }
     }
+    // title/summary are plain YAML strings -- whitespace-collapse only;
+    // the prose gets the full markdown strip.
+    var search = {
+        title: normalizeText(data.title),
+        summary: normalizeText(data.summary),
+        text: stripText(content)
+    };
     return {
         doc: {
             relPath: file.relPath,
             mtimeMs: file.mtimeMs,
             size: file.size,
             data: mapped,
-            search: {
-                // title/summary are plain YAML strings -- whitespace-collapse only;
-                // the prose gets the full markdown strip.
-                title: normalizeText(data.title),
-                summary: normalizeText(data.summary),
-                text: stripText(content)
-            },
+            search: search,
             extracted: Object.fromEntries(extractors.filter(function(f) {
                 return f.extract;
             }).map(function(f) {
                 var _f_extract;
                 return [
                     f.name,
-                    (_f_extract = f.extract) === null || _f_extract === void 0 ? void 0 : _f_extract.call(f, raw, content)
+                    (_f_extract = f.extract) === null || _f_extract === void 0 ? void 0 : _f_extract.call(f, raw, content, search)
                 ];
             }))
         },

package/dist/cjs/scan.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/scan.ts"],"sourcesContent":["import { readFileSync, statSync } from 'node:fs';\nimport { join } from 'node:path';\nimport fastGlob from 'fast-glob';\nimport removeMarkdown from 'remove-markdown';\nimport { parseDocument } from 'yaml';\nimport type { Config } from './config.ts';\nimport type { Feature } from './features/types.ts';\n\n// Filesystem -> rows. Pure data in, data + warnings out; db.ts does the SQL.\n\n// Reserved: colliding with these would clash with the `frontmatter` table's own columns or the other tables (`content`, `links`, `sections`).\nconst RESERVED_COLUMNS = new Set(['path', '_mtime', '_size', '_rank', 'content', 'links', 'sections']);\n\nfunction normalizeText(value: unknown): string {\n  if (value === null || value === undefined) return '';\n  return String(value).replace(/\\s+/g, ' ').trim();\n}\n\n// Markdown is stripped at index time so snippets read as clean prose (also improves matching, e.g. `**bold**` indexes as `bold`).\n// remove-markdown doesn't cover Obsidian wikilinks or table scaffolding, so those are handled first/after.\nfunction stripText(value: string): string {\n  const withoutWikilinks = value.replace(/\\[\\[([^\\]|]+)\\|([^\\]]+)\\]\\]/g, '$2').replace(/\\[\\[([^\\]]+)\\]\\]/g, '$1');\n  const withoutMarkdown = removeMarkdown(withoutWikilinks);\n  const withoutTables = withoutMarkdown.replace(/^\\s*\\|?[-\\s|:]+\\|\\s*$/gm, '').replace(/\\|/g, ' ');\n  return normalizeText(withoutTables);\n}\n\nexport interface FileStat {\n  relPath: string;\n  absPath: string;\n  mtimeMs: number;\n  size: number;\n}\n\nexport function listFiles(cfg: Config, baseDir: string): FileStat[] {\n  const relPaths = fastGlob.sync(cfg.scan.include, { cwd: baseDir }).sort();\n  return relPaths.map((relPath) => {\n    const absPath = join(baseDir, relPath);\n    const st = statSync(absPath);\n    return { relPath, absPath, mtimeMs: st.mtimeMs, size: st.size };\n  });\n}\n\nexport interface ParsedDoc {\n  relPath: string;\n  mtimeMs: number;\n  size: number;\n  data: Record<string, string | number | null>;\n  // title/summary are duplicated from frontmatter so bm25() can weight them above the body text.\n  search: { title: string; summary: string; text: string };\n  // Per-feature extraction results, keyed by feature name; features store them at reconcile.\n  extracted: Record<string, unknown>;\n}\n\nfunction mapValue(value: unknown): string | number | null {\n  if (value === null || value === undefined) return null;\n  if (typeof value === 'boolean') return value ? 1 : 0;\n  if (typeof value === 'string' || typeof value === 'number') return value;\n  return JSON.stringify(value);\n}\n\n// The delimiter split is all this package used gray-matter for.\nfunction splitFrontmatter(raw: string): { fm: string | null; body: string } {\n  const open = raw.match(/^---\\r?\\n/);\n  if (!open) return { fm: null, body: raw };\n  const rest = raw.slice(open[0].length);\n  const close = rest.match(/^---\\r?(\\n|$)/m);\n  if (!close || close.index === undefined) return { fm: null, body: raw };\n  return { fm: rest.slice(0, close.index), body: rest.slice(close.index + close[0].length) };\n}\n\n// Lenient by design: parseDocument collects syntax errors as data and still yields values,\n// so Obsidian-style frontmatter (e.g. an alias starting with @) survives with a warning.\nfunction parseFrontmatter(relPath: string, fm: string, warnings: string[]): Record<string, unknown> {\n  const doc = parseDocument(fm);\n  if (doc.errors.length > 0) {\n    warnings.push(`warning: ${relPath} frontmatter has ${doc.errors.length} syntax error(s) (${doc.errors[0].message.split('\\n')[0]}); parsed leniently`);\n  }\n  let data: unknown;\n  try {\n    data = doc.toJS();\n  } catch (err) {\n    warnings.push(`warning: ${relPath} has unparseable frontmatter (${(err as Error).message.split('\\n')[0]}); indexing without it`);\n    return {};\n  }\n  if (data === null || data === undefined) return {};\n  if (typeof data !== 'object' || Array.isArray(data)) {\n    warnings.push(`warning: ${relPath} frontmatter is not a key-value mapping; ignoring it`);\n    return {};\n  }\n  return data as Record<string, unknown>;\n}\n\nexport function parseFile(file: FileStat, extractors: Feature[] = []): { doc: ParsedDoc; warnings: string[] } {\n  const raw = readFileSync(file.absPath, 'utf8');\n  const warnings: string[] = [];\n\n  const { fm, body: content } = splitFrontmatter(raw);\n  const data = fm === null ? {} : parseFrontmatter(file.relPath, fm, warnings);\n  const mapped: Record<string, string | number | null> = {};\n\n  for (const key of Object.keys(data)) {\n    if (RESERVED_COLUMNS.has(key)) {\n      warnings.push(`warning: ${file.relPath} has a frontmatter key named \"${key}\", which is reserved; ignoring it`);\n      continue;\n    }\n    mapped[key] = mapValue(data[key]);\n  }\n\n  return {\n    doc: {\n      relPath: file.relPath,\n      mtimeMs: file.mtimeMs,\n      size: file.size,\n      data: mapped,\n      search: {\n        // title/summary are plain YAML strings -- whitespace-collapse only;\n        // the prose gets the full markdown strip.\n        title: normalizeText(data.title),\n        summary: normalizeText(data.summary),\n        text: stripText(content),\n      },\n      extracted: Object.fromEntries(extractors.filter((f) => f.extract).map((f) => [f.name, f.extract?.(raw, content)])),\n    },\n    warnings,\n  };\n}\n"],"names":["listFiles","parseFile","RESERVED_COLUMNS","Set","normalizeText","value","undefined","String","replace","trim","stripText","withoutWikilinks","withoutMarkdown","removeMarkdown","withoutTables","cfg","baseDir","relPaths","fastGlob","sync","scan","include","cwd","sort","map","relPath","absPath","join","st","statSync","mtimeMs","size","mapValue","JSON","stringify","splitFrontmatter","raw","open","match","fm","body","rest","slice","length","close","index","parseFrontmatter","warnings","doc","parseDocument","errors","push","message","split","data","toJS","err","Array","isArray","file","extractors","readFileSync","content","mapped","Object","keys","key","has","search","title","summary","text","extracted","fromEntries","filter","f","extract","name"],"mappings":";;;;;;;;;;;QAkCgBA;eAAAA;;QA2DAC;eAAAA;;;sBA7FuB;wBAClB;+DACA;qEACM;oBACG;;;;;;;;;;AAI9B,6EAA6E;AAE7E,8IAA8I;AAC9I,IAAMC,mBAAmB,IAAIC,IAAI;IAAC;IAAQ;IAAU;IAAS;IAAS;IAAW;IAAS;CAAW;AAErG,SAASC,cAAcC,KAAc;IACnC,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,OAAOC,OAAOF,OAAOG,OAAO,CAAC,QAAQ,KAAKC,IAAI;AAChD;AAEA,kIAAkI;AAClI,2GAA2G;AAC3G,SAASC,UAAUL,KAAa;IAC9B,IAAMM,mBAAmBN,MAAMG,OAAO,CAAC,gCAAgC,MAAMA,OAAO,CAAC,qBAAqB;IAC1G,IAAMI,kBAAkBC,IAAAA,uBAAc,EAACF;IACvC,IAAMG,gBAAgBF,gBAAgBJ,OAAO,CAAC,2BAA2B,IAAIA,OAAO,CAAC,OAAO;IAC5F,OAAOJ,cAAcU;AACvB;AASO,SAASd,UAAUe,GAAW,EAAEC,OAAe;IACpD,IAAMC,WAAWC,iBAAQ,CAACC,IAAI,CAACJ,IAAIK,IAAI,CAACC,OAAO,EAAE;QAAEC,KAAKN;IAAQ,GAAGO,IAAI;IACvE,OAAON,SAASO,GAAG,CAAC,SAACC;QACnB,IAAMC,UAAUC,IAAAA,cAAI,EAACX,SAASS;QAC9B,IAAMG,KAAKC,IAAAA,gBAAQ,EAACH;QACpB,OAAO;YAAED,SAAAA;YAASC,SAAAA;YAASI,SAASF,GAAGE,OAAO;YAAEC,MAAMH,GAAGG,IAAI;QAAC;IAChE;AACF;AAaA,SAASC,SAAS3B,KAAc;IAC9B,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,IAAI,OAAOD,UAAU,WAAW,OAAOA,QAAQ,IAAI;IACnD,IAAI,OAAOA,UAAU,YAAY,OAAOA,UAAU,UAAU,OAAOA;IACnE,OAAO4B,KAAKC,SAAS,CAAC7B;AACxB;AAEA,gEAAgE;AAChE,SAAS8B,iBAAiBC,GAAW;IACnC,IAAMC,OAAOD,IAAIE,KAAK,CAAC;IACvB,IAAI,CAACD,MAAM,OAAO;QAAEE,IAAI;QAAMC,MAAMJ;IAAI;IACxC,IAAMK,OAAOL,IAAIM,KAAK,CAACL,IAAI,CAAC,EAAE,CAACM,MAAM;IACrC,IAAMC,QAAQH,KAAKH,KAAK,CAAC;IACzB,IAAI,CAACM,SAASA,MAAMC,KAAK,KAAKvC,WAAW,OAAO;QAAEiC,IAAI;QAAMC,MAAMJ;IAAI;IACtE,OAAO;QAAEG,IAAIE,KAAKC,KAAK,CAAC,GAAGE,MAAMC,KAAK;QAAGL,MAAMC,KAAKC,KAAK,CAACE,MAAMC,KAAK,GAAGD,KAAK,CAAC,EAAE,CAACD,MAAM;IAAE;AAC3F;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASG,iBAAiBrB,OAAe,EAAEc,EAAU,EAAEQ,QAAkB;IACvE,IAAMC,MAAMC,IAAAA,mBAAa,EAACV;IAC1B,IAAIS,IAAIE,MAAM,CAACP,MAAM,GAAG,GAAG;QACzBI,SAASI,IAAI,CAAC,AAAC,YAAsCH,OAA3BvB,SAAQ,qBAAyDuB,OAAtCA,IAAIE,MAAM,CAACP,MAAM,EAAC,sBAAyD,OAArCK,IAAIE,MAAM,CAAC,EAAE,CAACE,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,EAAC;IAClI;IACA,IAAIC;IACJ,IAAI;QACFA,OAAON,IAAIO,IAAI;IACjB,EAAE,OAAOC,KAAK;QACZT,SAASI,IAAI,CAAC,AAAC,YAAmD,OAAxC1B,SAAQ,kCAAsE,OAAtC,AAAC+B,IAAcJ,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,EAAC;QACxG,OAAO,CAAC;IACV;IACA,IAAIC,SAAS,QAAQA,SAAShD,WAAW,OAAO,CAAC;IACjD,IAAI,CAAA,OAAOgD,qCAAP,SAAOA,KAAG,MAAM,YAAYG,MAAMC,OAAO,CAACJ,OAAO;QACnDP,SAASI,IAAI,CAAC,AAAC,YAAmB,OAAR1B,SAAQ;QAClC,OAAO,CAAC;IACV;IACA,OAAO6B;AACT;AAEO,SAASrD,UAAU0D,IAAc;QAAEC,aAAAA,iEAAwB,EAAE;IAClE,IAAMxB,MAAMyB,IAAAA,oBAAY,EAACF,KAAKjC,OAAO,EAAE;IACvC,IAAMqB,WAAqB,EAAE;IAE7B,IAA8BZ,oBAAAA,iBAAiBC,MAAvCG,KAAsBJ,kBAAtBI,IAAIC,AAAMsB,UAAY3B,kBAAlBK;IACZ,IAAMc,OAAOf,OAAO,OAAO,CAAC,IAAIO,iBAAiBa,KAAKlC,OAAO,EAAEc,IAAIQ;IACnE,IAAMgB,SAAiD,CAAC;QAEnD,kCAAA,2BAAA;;QAAL,QAAK,YAAaC,OAAOC,IAAI,CAACX,0BAAzB,SAAA,6BAAA,QAAA,yBAAA,iCAAgC;YAAhC,IAAMY,MAAN;YACH,IAAIhE,iBAAiBiE,GAAG,CAACD,MAAM;gBAC7BnB,SAASI,IAAI,CAAC,AAAC,YAAwDe,OAA7CP,KAAKlC,OAAO,EAAC,kCAAoC,OAAJyC,KAAI;gBAC3E;YACF;YACAH,MAAM,CAACG,IAAI,GAAGlC,SAASsB,IAAI,CAACY,IAAI;QAClC;;QANK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAQL,OAAO;QACLlB,KAAK;YACHvB,SAASkC,KAAKlC,OAAO;YACrBK,SAAS6B,KAAK7B,OAAO;YACrBC,MAAM4B,KAAK5B,IAAI;YACfuB,MAAMS;YACNK,QAAQ;gBACN,oEAAoE;gBACpE,0CAA0C;gBAC1CC,OAAOjE,cAAckD,KAAKe,KAAK;gBAC/BC,SAASlE,cAAckD,KAAKgB,OAAO;gBACnCC,MAAM7D,UAAUoD;YAClB;YACAU,WAAWR,OAAOS,WAAW,CAACb,WAAWc,MAAM,CAAC,SAACC;uBAAMA,EAAEC,OAAO;eAAEpD,GAAG,CAAC,SAACmD;oBAAeA;uBAAT;oBAACA,EAAEE,IAAI;qBAAEF,aAAAA,EAAEC,OAAO,cAATD,iCAAAA,gBAAAA,GAAYvC,KAAK0B;iBAAS;;QAClH;QACAf,UAAAA;IACF;AACF"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/scan.ts"],"sourcesContent":["import { readFileSync, statSync } from 'node:fs';\nimport { join } from 'node:path';\nimport fastGlob from 'fast-glob';\nimport removeMarkdown from 'remove-markdown';\nimport { parseDocument } from 'yaml';\nimport type { Config } from './config.ts';\nimport type { Feature } from './features/types.ts';\n\n// Filesystem -> rows. Pure data in, data + warnings out; db.ts does the SQL.\n\n// Reserved: colliding with these would clash with the `frontmatter` table's own columns or the other tables (`content`, `links`, `sections`).\nconst RESERVED_COLUMNS = new Set(['path', '_mtime', '_size', '_rank', 'content', 'links', 'sections']);\n\nfunction normalizeText(value: unknown): string {\n  if (value === null || value === undefined) return '';\n  return String(value).replace(/\\s+/g, ' ').trim();\n}\n\n// Markdown is stripped at index time so snippets read as clean prose (also improves matching, e.g. `**bold**` indexes as `bold`).\n// remove-markdown doesn't cover Obsidian wikilinks or table scaffolding, so those are handled first/after.\nfunction stripText(value: string): string {\n  const withoutWikilinks = value.replace(/\\[\\[([^\\]|]+)\\|([^\\]]+)\\]\\]/g, '$2').replace(/\\[\\[([^\\]]+)\\]\\]/g, '$1');\n  const withoutMarkdown = removeMarkdown(withoutWikilinks);\n  const withoutTables = withoutMarkdown.replace(/^\\s*\\|?[-\\s|:]+\\|\\s*$/gm, '').replace(/\\|/g, ' ');\n  return normalizeText(withoutTables);\n}\n\nexport interface FileStat {\n  relPath: string;\n  absPath: string;\n  mtimeMs: number;\n  size: number;\n}\n\nexport function listFiles(cfg: Config, baseDir: string): FileStat[] {\n  const relPaths = fastGlob.sync(cfg.scan.include, { cwd: baseDir }).sort();\n  return relPaths.map((relPath) => {\n    const absPath = join(baseDir, relPath);\n    const st = statSync(absPath);\n    return { relPath, absPath, mtimeMs: st.mtimeMs, size: st.size };\n  });\n}\n\nexport interface ParsedDoc {\n  relPath: string;\n  mtimeMs: number;\n  size: number;\n  data: Record<string, string | number | null>;\n  // title/summary are duplicated from frontmatter so bm25() can weight them above the body text.\n  search: { title: string; summary: string; text: string };\n  // Per-feature extraction results, keyed by feature name; features store them at reconcile.\n  extracted: Record<string, unknown>;\n}\n\nfunction mapValue(value: unknown): string | number | null {\n  if (value === null || value === undefined) return null;\n  if (typeof value === 'boolean') return value ? 1 : 0;\n  if (typeof value === 'string' || typeof value === 'number') return value;\n  return JSON.stringify(value);\n}\n\n// The delimiter split is all this package used gray-matter for.\nfunction splitFrontmatter(raw: string): { fm: string | null; body: string } {\n  const open = raw.match(/^---\\r?\\n/);\n  if (!open) return { fm: null, body: raw };\n  const rest = raw.slice(open[0].length);\n  const close = rest.match(/^---\\r?(\\n|$)/m);\n  if (!close || close.index === undefined) return { fm: null, body: raw };\n  return { fm: rest.slice(0, close.index), body: rest.slice(close.index + close[0].length) };\n}\n\n// Lenient by design: parseDocument collects syntax errors as data and still yields values,\n// so Obsidian-style frontmatter (e.g. an alias starting with @) survives with a warning.\nfunction parseFrontmatter(relPath: string, fm: string, warnings: string[]): Record<string, unknown> {\n  const doc = parseDocument(fm);\n  if (doc.errors.length > 0) {\n    warnings.push(`warning: ${relPath} frontmatter has ${doc.errors.length} syntax error(s) (${doc.errors[0].message.split('\\n')[0]}); parsed leniently`);\n  }\n  let data: unknown;\n  try {\n    data = doc.toJS();\n  } catch (err) {\n    warnings.push(`warning: ${relPath} has unparseable frontmatter (${(err as Error).message.split('\\n')[0]}); indexing without it`);\n    return {};\n  }\n  if (data === null || data === undefined) return {};\n  if (typeof data !== 'object' || Array.isArray(data)) {\n    warnings.push(`warning: ${relPath} frontmatter is not a key-value mapping; ignoring it`);\n    return {};\n  }\n  return data as Record<string, unknown>;\n}\n\nexport function parseFile(file: FileStat, extractors: Feature[] = []): { doc: ParsedDoc; warnings: string[] } {\n  const raw = readFileSync(file.absPath, 'utf8');\n  const warnings: string[] = [];\n\n  const { fm, body: content } = splitFrontmatter(raw);\n  const data = fm === null ? {} : parseFrontmatter(file.relPath, fm, warnings);\n  const mapped: Record<string, string | number | null> = {};\n\n  for (const key of Object.keys(data)) {\n    if (RESERVED_COLUMNS.has(key)) {\n      warnings.push(`warning: ${file.relPath} has a frontmatter key named \"${key}\", which is reserved; ignoring it`);\n      continue;\n    }\n    mapped[key] = mapValue(data[key]);\n  }\n\n  // title/summary are plain YAML strings -- whitespace-collapse only;\n  // the prose gets the full markdown strip.\n  const search = { title: normalizeText(data.title), summary: normalizeText(data.summary), text: stripText(content) };\n\n  return {\n    doc: {\n      relPath: file.relPath,\n      mtimeMs: file.mtimeMs,\n      size: file.size,\n      data: mapped,\n      search,\n      extracted: Object.fromEntries(extractors.filter((f) => f.extract).map((f) => [f.name, f.extract?.(raw, content, search)])),\n    },\n    warnings,\n  };\n}\n"],"names":["listFiles","parseFile","RESERVED_COLUMNS","Set","normalizeText","value","undefined","String","replace","trim","stripText","withoutWikilinks","withoutMarkdown","removeMarkdown","withoutTables","cfg","baseDir","relPaths","fastGlob","sync","scan","include","cwd","sort","map","relPath","absPath","join","st","statSync","mtimeMs","size","mapValue","JSON","stringify","splitFrontmatter","raw","open","match","fm","body","rest","slice","length","close","index","parseFrontmatter","warnings","doc","parseDocument","errors","push","message","split","data","toJS","err","Array","isArray","file","extractors","readFileSync","content","mapped","Object","keys","key","has","search","title","summary","text","extracted","fromEntries","filter","f","extract","name"],"mappings":";;;;;;;;;;;QAkCgBA;eAAAA;;QA2DAC;eAAAA;;;sBA7FuB;wBAClB;+DACA;qEACM;oBACG;;;;;;;;;;AAI9B,6EAA6E;AAE7E,8IAA8I;AAC9I,IAAMC,mBAAmB,IAAIC,IAAI;IAAC;IAAQ;IAAU;IAAS;IAAS;IAAW;IAAS;CAAW;AAErG,SAASC,cAAcC,KAAc;IACnC,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,OAAOC,OAAOF,OAAOG,OAAO,CAAC,QAAQ,KAAKC,IAAI;AAChD;AAEA,kIAAkI;AAClI,2GAA2G;AAC3G,SAASC,UAAUL,KAAa;IAC9B,IAAMM,mBAAmBN,MAAMG,OAAO,CAAC,gCAAgC,MAAMA,OAAO,CAAC,qBAAqB;IAC1G,IAAMI,kBAAkBC,IAAAA,uBAAc,EAACF;IACvC,IAAMG,gBAAgBF,gBAAgBJ,OAAO,CAAC,2BAA2B,IAAIA,OAAO,CAAC,OAAO;IAC5F,OAAOJ,cAAcU;AACvB;AASO,SAASd,UAAUe,GAAW,EAAEC,OAAe;IACpD,IAAMC,WAAWC,iBAAQ,CAACC,IAAI,CAACJ,IAAIK,IAAI,CAACC,OAAO,EAAE;QAAEC,KAAKN;IAAQ,GAAGO,IAAI;IACvE,OAAON,SAASO,GAAG,CAAC,SAACC;QACnB,IAAMC,UAAUC,IAAAA,cAAI,EAACX,SAASS;QAC9B,IAAMG,KAAKC,IAAAA,gBAAQ,EAACH;QACpB,OAAO;YAAED,SAAAA;YAASC,SAAAA;YAASI,SAASF,GAAGE,OAAO;YAAEC,MAAMH,GAAGG,IAAI;QAAC;IAChE;AACF;AAaA,SAASC,SAAS3B,KAAc;IAC9B,IAAIA,UAAU,QAAQA,UAAUC,WAAW,OAAO;IAClD,IAAI,OAAOD,UAAU,WAAW,OAAOA,QAAQ,IAAI;IACnD,IAAI,OAAOA,UAAU,YAAY,OAAOA,UAAU,UAAU,OAAOA;IACnE,OAAO4B,KAAKC,SAAS,CAAC7B;AACxB;AAEA,gEAAgE;AAChE,SAAS8B,iBAAiBC,GAAW;IACnC,IAAMC,OAAOD,IAAIE,KAAK,CAAC;IACvB,IAAI,CAACD,MAAM,OAAO;QAAEE,IAAI;QAAMC,MAAMJ;IAAI;IACxC,IAAMK,OAAOL,IAAIM,KAAK,CAACL,IAAI,CAAC,EAAE,CAACM,MAAM;IACrC,IAAMC,QAAQH,KAAKH,KAAK,CAAC;IACzB,IAAI,CAACM,SAASA,MAAMC,KAAK,KAAKvC,WAAW,OAAO;QAAEiC,IAAI;QAAMC,MAAMJ;IAAI;IACtE,OAAO;QAAEG,IAAIE,KAAKC,KAAK,CAAC,GAAGE,MAAMC,KAAK;QAAGL,MAAMC,KAAKC,KAAK,CAACE,MAAMC,KAAK,GAAGD,KAAK,CAAC,EAAE,CAACD,MAAM;IAAE;AAC3F;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASG,iBAAiBrB,OAAe,EAAEc,EAAU,EAAEQ,QAAkB;IACvE,IAAMC,MAAMC,IAAAA,mBAAa,EAACV;IAC1B,IAAIS,IAAIE,MAAM,CAACP,MAAM,GAAG,GAAG;QACzBI,SAASI,IAAI,CAAC,AAAC,YAAsCH,OAA3BvB,SAAQ,qBAAyDuB,OAAtCA,IAAIE,MAAM,CAACP,MAAM,EAAC,sBAAyD,OAArCK,IAAIE,MAAM,CAAC,EAAE,CAACE,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,EAAC;IAClI;IACA,IAAIC;IACJ,IAAI;QACFA,OAAON,IAAIO,IAAI;IACjB,EAAE,OAAOC,KAAK;QACZT,SAASI,IAAI,CAAC,AAAC,YAAmD,OAAxC1B,SAAQ,kCAAsE,OAAtC,AAAC+B,IAAcJ,OAAO,CAACC,KAAK,CAAC,KAAK,CAAC,EAAE,EAAC;QACxG,OAAO,CAAC;IACV;IACA,IAAIC,SAAS,QAAQA,SAAShD,WAAW,OAAO,CAAC;IACjD,IAAI,CAAA,OAAOgD,qCAAP,SAAOA,KAAG,MAAM,YAAYG,MAAMC,OAAO,CAACJ,OAAO;QACnDP,SAASI,IAAI,CAAC,AAAC,YAAmB,OAAR1B,SAAQ;QAClC,OAAO,CAAC;IACV;IACA,OAAO6B;AACT;AAEO,SAASrD,UAAU0D,IAAc;QAAEC,aAAAA,iEAAwB,EAAE;IAClE,IAAMxB,MAAMyB,IAAAA,oBAAY,EAACF,KAAKjC,OAAO,EAAE;IACvC,IAAMqB,WAAqB,EAAE;IAE7B,IAA8BZ,oBAAAA,iBAAiBC,MAAvCG,KAAsBJ,kBAAtBI,IAAIC,AAAMsB,UAAY3B,kBAAlBK;IACZ,IAAMc,OAAOf,OAAO,OAAO,CAAC,IAAIO,iBAAiBa,KAAKlC,OAAO,EAAEc,IAAIQ;IACnE,IAAMgB,SAAiD,CAAC;QAEnD,kCAAA,2BAAA;;QAAL,QAAK,YAAaC,OAAOC,IAAI,CAACX,0BAAzB,SAAA,6BAAA,QAAA,yBAAA,iCAAgC;YAAhC,IAAMY,MAAN;YACH,IAAIhE,iBAAiBiE,GAAG,CAACD,MAAM;gBAC7BnB,SAASI,IAAI,CAAC,AAAC,YAAwDe,OAA7CP,KAAKlC,OAAO,EAAC,kCAAoC,OAAJyC,KAAI;gBAC3E;YACF;YACAH,MAAM,CAACG,IAAI,GAAGlC,SAASsB,IAAI,CAACY,IAAI;QAClC;;QANK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAQL,oEAAoE;IACpE,0CAA0C;IAC1C,IAAME,SAAS;QAAEC,OAAOjE,cAAckD,KAAKe,KAAK;QAAGC,SAASlE,cAAckD,KAAKgB,OAAO;QAAGC,MAAM7D,UAAUoD;IAAS;IAElH,OAAO;QACLd,KAAK;YACHvB,SAASkC,KAAKlC,OAAO;YACrBK,SAAS6B,KAAK7B,OAAO;YACrBC,MAAM4B,KAAK5B,IAAI;YACfuB,MAAMS;YACNK,QAAAA;YACAI,WAAWR,OAAOS,WAAW,CAACb,WAAWc,MAAM,CAAC,SAACC;uBAAMA,EAAEC,OAAO;eAAEpD,GAAG,CAAC,SAACmD;oBAAeA;uBAAT;oBAACA,EAAEE,IAAI;qBAAEF,aAAAA,EAAEC,OAAO,cAATD,iCAAAA,gBAAAA,GAAYvC,KAAK0B,SAASM;iBAAQ;;QAC1H;QACArB,UAAAA;IACF;AACF"}

package/dist/cjs/verbs.d.cts

@@ -1,11 +1,12 @@
 import type { DatabaseSync } from 'node:sqlite';
-import type { Config } from './config.js';
+import type { Config, FeatureName } from './config.js';
 import type { Row } from './output.js';
 export interface FindOptions {
     k?: number;
     where?: string;
+    semantic?: boolean;
 }
-export declare function find(db: DatabaseSync, cfg: Config, terms: string, opts?: FindOptions): Row[];
+export declare function find(db: DatabaseSync, cfg: Config, terms: string, opts?: FindOptions): Promise<Row[]>;
 export interface TreeMap {
     docs: {
         count: number;
@@ -13,6 +14,10 @@ export interface TreeMap {
     };
     fields: Row[];
     fieldsTotal: number;
+    features: {
+        on: FeatureName[];
+        off: FeatureName[];
+    };
     hubs: Row[];
     recent: Row[];
 }
@@ -28,5 +33,6 @@ export interface Peek {
     outboundTotal: number;
     backlinksTotal: number;
     unresolvedTotal: number;
+    off: FeatureName[];
 }
 export declare function peek(db: DatabaseSync, cfg: Config, pathArg: string): Peek;

package/dist/cjs/verbs.d.ts

@@ -1,11 +1,12 @@
 import type { DatabaseSync } from 'node:sqlite';
-import type { Config } from './config.js';
+import type { Config, FeatureName } from './config.js';
 import type { Row } from './output.js';
 export interface FindOptions {
     k?: number;
     where?: string;
+    semantic?: boolean;
 }
-export declare function find(db: DatabaseSync, cfg: Config, terms: string, opts?: FindOptions): Row[];
+export declare function find(db: DatabaseSync, cfg: Config, terms: string, opts?: FindOptions): Promise<Row[]>;
 export interface TreeMap {
     docs: {
         count: number;
@@ -13,6 +14,10 @@ export interface TreeMap {
     };
     fields: Row[];
     fieldsTotal: number;
+    features: {
+        on: FeatureName[];
+        off: FeatureName[];
+    };
     hubs: Row[];
     recent: Row[];
 }
@@ -28,5 +33,6 @@ export interface Peek {
     outboundTotal: number;
     backlinksTotal: number;
     unresolvedTotal: number;
+    off: FeatureName[];
 }
 export declare function peek(db: DatabaseSync, cfg: Config, pathArg: string): Peek;