sensemaking 0.9.5 → 0.11.0

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

@@ -1 +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 { progress } from '../progress.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  // A doc covered only by semantic-false presets never had extract() run for it (db.ts's\n  // per-file filter skips it), so extracted is undefined here -- store nothing, i.e. no rows.\n  store(db, path, extracted) {\n    if (!extracted) return;\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  enabledForFile(_cfg, file) {\n    return file.embed;\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    // The tokenizer yields undefined (not the unk id) for tokens outside the vocab; an\n    // undefined id would index the matrix at NaN and poison the whole mean-pool -- and the\n    // int8 conversion then stores the NaN vector as all zeros, silently. Keep integers only.\n    const ids = (tok.encode(text, { add_special_tokens: false }).ids as number[]).filter((id) => Number.isInteger(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}:${model}`, dims, embed: post };\n}\n\nconst providers = new Map<string, Promise<EmbedProvider>>();\n\nfunction getProvider(cfg: Config): Promise<EmbedProvider> {\n  const e = embedConfig(cfg);\n  if (!e) throw new SenseError('EMBED_DISABLED', 'semantic search needs at least one preset with vectors on (semantic !== false); every declared preset in sense.config.json has \"semantic\": false');\n  const sig = `${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      // presets/embed are irrelevant here -- re-deriving chunk text for a doc that already\n      // has embeddings rows means it was covered by a semantic-on preset at reconcile time.\n      chunks = parseFile({ relPath: path, absPath: join(baseDir, path), mtimeMs: 0, size: 0, presets: [], embed: true }, [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  // The lazy build is the one long silence a first --semantic query hits (measured 23s at\n  // 26k notes); progress makes it distinguishable from a hang.\n  const report = progress('embedding chunks', jobs.length);\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    report.tick(Math.min(i + BATCH, jobs.length));\n  }\n  report.finish();\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.\n// Returns the cosine similarity alongside each candidate: stored and query vectors are both\n// L2-normalised before quantisation, so the dot product is cosine in [-1, 1]. `find` surfaces\n// it because the fused rank score cannot express match quality -- nearest-neighbour search\n// always returns a nearest neighbour, so without a magnitude an agent cannot tell a real hit\n// from the best of a bad lot.\nexport async function semanticCandidates(db: DatabaseSync, cfg: Config, terms: string, fetch: number): Promise<Array<{ path: string; lines: string; similarity: number }>> {\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, similarity: Math.round(b.score * 1000) / 1000 }));\n}\n"],"names":["existsSync","mkdirSync","readFileSync","renameSync","writeFileSync","homedir","join","embedConfig","SenseError","progress","parseFile","STORE_DIMS","BATCH","chunksOf","raw","search","lines","split","starts","inFence","i","length","test","push","bounds","prefix","title","summary","filter","Boolean","chunks","forEach","start","end","text","slice","trim","startLine","endLine","embed","name","schema","db","exec","extract","_body","remove","path","prepare","run","store","extracted","insert","c","idx","enabledForFile","_cfg","file","fetchToFile","url","dest","fetch","then","res","ok","status","Buffer","from","arrayBuffer","staticProvider","model","tokenizerJson","dir","replace","recursive","console","error","headerLen","Number","readBigUInt64LE","header","JSON","parse","subarray","toString","entry","Object","entries","find","k","dtype","spec","dims","shape","dataStart","byteOffset","data_offsets","matrix","Float32Array","buffer","Tokenizer","tok","unkId","vocab","unk_token","one","ids","encode","add_special_tokens","id","isInteger","v","off","d","norm","Math","sqrt","texts","map","apiProvider","keyEnv","base","headers","key","process","env","undefined","authorization","post","method","body","stringify","input","json","data","embedding","providers","Map","getProvider","cfg","e","sig","type","p","get","set","toStore","full","int8","scale","max","abs","embedPending","baseDir","provider","dirty","all","storeDims","min","byPath","row","list","chunk","jobs","chunkIdxs","relPath","absPath","mtimeMs","size","presets","doc","update","report","batch","vectors","j","job","q","Int8Array","round","err","tick","finish","semanticCandidates","terms","match","t","includes","qv","rows","best","vector","byteLength","dot","score","existing","start_line","end_line","sort","a","b","similarity"],"mappings":"AAAA,SAASA,UAAU,EAAEC,SAAS,EAAEC,YAAY,EAAEC,UAAU,EAAEC,aAAa,QAAQ,UAAU;AACzF,SAASC,OAAO,QAAQ,UAAU;AAClC,SAASC,IAAI,QAAQ,YAAY;AAGjC,SAASC,WAAW,QAAQ,eAAe;AAC3C,SAASC,UAAU,QAAQ,eAAe;AAC1C,SAASC,QAAQ,QAAQ,iBAAiB;AAC1C,SAASC,SAAS,QAAQ,aAAa;AAGvC,sFAAsF;AACtF,uFAAuF;AACvF,oFAAoF;AACpF,+EAA+E;AAE/E,6EAA6E;AAC7E,yEAAyE;AACzE,MAAMC,aAAa;AACnB,MAAMC,QAAQ;AAcd,uFAAuF;AACvF,sFAAsF;AACtF,mFAAmF;AACnF,yBAAyB;AACzB,SAASC,SAASC,GAAW,EAAEC,MAA2C;IACxE,MAAMC,QAAQF,IAAIG,KAAK,CAAC;IACxB,MAAMC,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,MAAMI,SAASN,OAAOG,MAAM,KAAK,IAAI;QAAC;KAAE,GAAGH,MAAM,CAAC,EAAE,GAAG,IAAI;QAAC;WAAMA;KAAO,GAAGA;IAC5E,MAAMO,SAAS;QAACV,mBAAAA,6BAAAA,OAAQW,KAAK;QAAEX,mBAAAA,6BAAAA,OAAQY,OAAO;KAAC,CAACC,MAAM,CAACC,SAASvB,IAAI,CAAC;IACrE,MAAMwB,SAAkB,EAAE;IAC1BN,OAAOO,OAAO,CAAC,CAACC,OAAOZ;QACrB,MAAMa,MAAMb,IAAI,IAAII,OAAOH,MAAM,GAAGG,MAAM,CAACJ,IAAI,EAAE,GAAG,IAAIJ,MAAMK,MAAM;QACpE,MAAMa,OAAOlB,MACVmB,KAAK,CAACH,QAAQ,GAAGC,KACjB3B,IAAI,CAAC,MACL8B,IAAI;QACP,IAAIF,KAAKb,MAAM,GAAG,GAAGS,OAAOP,IAAI,CAAC;YAAEc,WAAWL;YAAOM,SAASL;YAAKC,MAAMT,SAAS,GAAGA,OAAO,EAAE,EAAES,MAAM,GAAGA;QAAK;IAChH;IACA,OAAOJ;AACT;AAEA,OAAO,MAAMS,QAAiB;IAC5BC,MAAM;IACNC,QAAOC,EAAE;QACPA,GAAGC,IAAI,CAAC,CAAC,8JAA8J,CAAC;IAC1K;IACAC,SAAQ9B,GAAG,EAAE+B,KAAK,EAAE9B,MAAM;QACxB,OAAOF,SAASC,KAAKC;IACvB;IACA+B,QAAOJ,EAAE,EAAEK,IAAI;QACbL,GAAGM,OAAO,CAAC,2CAA2CC,GAAG,CAACF;IAC5D;IACA,uFAAuF;IACvF,4FAA4F;IAC5FG,OAAMR,EAAE,EAAEK,IAAI,EAAEI,SAAS;QACvB,IAAI,CAACA,WAAW;QAChB,MAAMC,SAASV,GAAGM,OAAO,CAAC;QACzBG,UAAsBpB,OAAO,CAAC,CAACsB,GAAGC,MAAQF,OAAOH,GAAG,CAACF,MAAMO,KAAKD,EAAEhB,SAAS,EAAEgB,EAAEf,OAAO;IACzF;IACAiB,gBAAeC,IAAI,EAAEC,IAAI;QACvB,OAAOA,KAAKlB,KAAK;IACnB;AACF,EAAE;AAEF,2FAA2F;AAC3F,yFAAyF;AACzF,gBAAgB;AAEhB,SAASmB,YAAYC,GAAW,EAAEC,IAAY;IAC5C,OAAOC,MAAMF,KAAKG,IAAI,CAAC,OAAOC;QAC5B,IAAI,CAACA,IAAIC,EAAE,EAAE,MAAM,IAAIxD,WAAW,eAAe,CAAC,uBAAuB,EAAEmD,IAAI,SAAS,EAAEI,IAAIE,MAAM,EAAE;QACtG7D,cAAc,GAAGwD,KAAK,KAAK,CAAC,EAAEM,OAAOC,IAAI,CAAC,MAAMJ,IAAIK,WAAW;QAC/DjE,WAAW,GAAGyD,KAAK,KAAK,CAAC,EAAEA;IAC7B;AACF;AAEA,eAAeS,eAAeC,KAAa;;QA2BEC,sBAA7BA,4BAAAA;IA1Bd,IAAIC,MAAMF;IACV,IAAI,CAACtE,WAAWM,KAAKkE,KAAK,uBAAuB;QAC/CA,MAAMlE,KAAKD,WAAW,UAAU,eAAe,UAAUiE,MAAMG,OAAO,CAAC,OAAO;QAC9ExE,UAAUuE,KAAK;YAAEE,WAAW;QAAK;QACjC,KAAK,MAAMjB,QAAQ;YAAC;YAAqB;SAAiB,CAAE;YAC1D,IAAI,CAACzD,WAAWM,KAAKkE,KAAKf,QAAQ;gBAChCkB,QAAQC,KAAK,CAAC,CAAC,SAAS,EAAEN,MAAM,CAAC,EAAEb,KAAK,MAAM,EAAEe,IAAI,0BAA0B,CAAC;gBAC/E,MAAMd,YAAY,CAAC,uBAAuB,EAAEY,MAAM,cAAc,EAAEb,MAAM,EAAEnD,KAAKkE,KAAKf;YACtF;QACF;IACF;IAEA,MAAM3C,MAAMZ,aAAaI,KAAKkE,KAAK;IACnC,MAAMK,YAAYC,OAAOhE,IAAIiE,eAAe,CAAC;IAC7C,MAAMC,SAASC,KAAKC,KAAK,CAACpE,IAAIqE,QAAQ,CAAC,GAAG,IAAIN,WAAWO,QAAQ,CAAC;IAClE,MAAMC,QAAQC,OAAOC,OAAO,CAACP,QAAQQ,IAAI,CAAC,CAAC,CAACC,EAAE,GAAKA,MAAM;IACzD,IAAI,CAACJ,SAASA,KAAK,CAAC,EAAE,CAACK,KAAK,KAAK,OAAO,MAAM,IAAIlF,WAAW,eAAe,GAAG8D,MAAM,oCAAoC,CAAC;IAC1H,MAAMqB,OAAON,KAAK,CAAC,EAAE;IACrB,MAAMO,OAAOD,KAAKE,KAAK,CAAC,EAAE;IAC1B,MAAMC,YAAYhF,IAAIiF,UAAU,GAAG,IAAIlB,YAAYc,KAAKK,YAAY,CAAC,EAAE;IACvE,MAAMC,SAASH,YAAY,MAAM,IAAI,IAAII,aAAapF,IAAIqF,MAAM,EAAEL,WAAWH,KAAKE,KAAK,CAAC,EAAE,GAAGD,QAAQ,IAAIM,aAAapF,IAAIqF,MAAM,CAAChE,KAAK,CAAC2D,WAAWA,YAAYH,KAAKE,KAAK,CAAC,EAAE,GAAGD,OAAO;IAErL,MAAMrB,gBAAgBU,KAAKC,KAAK,CAAChF,aAAaI,KAAKkE,KAAK,mBAAmB;IAC3E,oFAAoF;IACpF,MAAM,EAAE4B,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC;IACnC,MAAMC,MAAM,IAAID,UAAU7B,eAAe,CAAC;IAC1C,MAAM+B,iBAAQ/B,wBAAAA,cAAcD,KAAK,cAAnBC,6CAAAA,6BAAAA,sBAAqBgC,KAAK,cAA1BhC,iDAAAA,0BAA4B,EAACA,uBAAAA,cAAcD,KAAK,cAAnBC,2CAAAA,qBAAqBiC,SAAS,CAAC,uCAAI,CAAC;IAE/E,SAASC,IAAIvE,IAAY;QACvB,mFAAmF;QACnF,uFAAuF;QACvF,yFAAyF;QACzF,MAAMwE,MAAM,AAACL,IAAIM,MAAM,CAACzE,MAAM;YAAE0E,oBAAoB;QAAM,GAAGF,GAAG,CAAc9E,MAAM,CAAC,CAACiF,KAAO/B,OAAOgC,SAAS,CAACD,OAAOA,OAAOP;QAC5H,MAAMS,IAAI,IAAIb,aAAaN;QAC3B,IAAIc,IAAIrF,MAAM,KAAK,GAAG,OAAO0F;QAC7B,KAAK,MAAMF,MAAMH,IAAK;YACpB,MAAMM,MAAMH,KAAKjB;YACjB,IAAK,IAAIqB,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,IAAIhB,MAAM,CAACe,MAAMC,EAAE;QACxD;QACA,IAAIC,OAAO;QACX,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAK;YAC7BF,CAAC,CAACE,EAAE,IAAIP,IAAIrF,MAAM;YAClB6F,QAAQH,CAAC,CAACE,EAAE,GAAGF,CAAC,CAACE,EAAE;QACrB;QACAC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;QACzB,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,IAAIC;QACvC,OAAOH;IACT;IAEA,OAAO;QAAEF,IAAI,CAAC,OAAO,EAAEvC,OAAO;QAAEsB;QAAMrD,OAAO,OAAO8E,QAAUA,MAAMC,GAAG,CAACb;IAAK;AAC/E;AAEA,gFAAgF;AAEhF,eAAec,YAAYjD,KAAa,EAAEX,GAAuB,EAAE6D,MAA0B;IAC3F,IAAI,CAAC7D,KAAK,MAAM,IAAInD,WAAW,eAAe;IAC9C,MAAMiH,OAAO9D,IAAIc,OAAO,CAAC,QAAQ;IACjC,MAAMiD,UAAkC;QAAE,gBAAgB;IAAmB;IAC7E,MAAMC,MAAMH,SAASI,QAAQC,GAAG,CAACL,OAAO,GAAGM;IAC3C,IAAIH,KAAKD,QAAQK,aAAa,GAAG,CAAC,OAAO,EAAEJ,KAAK;IAEhD,eAAeK,KAAKX,KAAe;QACjC,MAAMtD,MAAM,MAAMF,MAAM,GAAG4D,KAAK,WAAW,CAAC,EAAE;YAAEQ,QAAQ;YAAQP;YAASQ,MAAMjD,KAAKkD,SAAS,CAAC;gBAAE7D;gBAAO8D,OAAOf;YAAM;QAAG;QACvH,IAAI,CAACtD,IAAIC,EAAE,EAAE,MAAM,IAAIxD,WAAW,eAAe,GAAGiH,KAAK,oBAAoB,EAAE1D,IAAIE,MAAM,EAAE;QAC3F,MAAMiE,OAAQ,MAAMnE,IAAIsE,IAAI;QAC5B,OAAOH,KAAKI,IAAI,CAAChB,GAAG,CAAC,CAACL,IAAMf,aAAa/B,IAAI,CAAC8C,EAAEsB,SAAS;IAC3D;IAEA,MAAM3C,OAAO,AAAC,CAAA,MAAMoC,KAAK;QAAC;KAAkB,CAAA,CAAE,CAAC,EAAE,CAAC3G,MAAM;IACxD,OAAO;QAAEwF,IAAI,CAAC,IAAI,EAAEY,KAAK,CAAC,EAAEnD,OAAO;QAAEsB;QAAMrD,OAAOyF;IAAK;AACzD;AAEA,MAAMQ,YAAY,IAAIC;AAEtB,SAASC,YAAYC,GAAW;QAGMC;IAFpC,MAAMA,IAAIrI,YAAYoI;IACtB,IAAI,CAACC,GAAG,MAAM,IAAIpI,WAAW,kBAAkB;IAC/C,MAAMqI,MAAM,GAAGD,EAAEE,IAAI,CAAC,CAAC,EAAEF,EAAEtE,KAAK,CAAC,CAAC,GAAEsE,SAAAA,EAAEjF,GAAG,cAALiF,oBAAAA,SAAS,IAAI;IACjD,IAAIG,IAAIP,UAAUQ,GAAG,CAACH;IACtB,IAAI,CAACE,GAAG;QACNA,IAAIH,EAAEE,IAAI,KAAK,QAAQvB,YAAYqB,EAAEtE,KAAK,EAAEsE,EAAEjF,GAAG,EAAEiF,EAAEjB,GAAG,IAAItD,eAAeuE,EAAEtE,KAAK;QAClFkE,UAAUS,GAAG,CAACJ,KAAKE;IACrB;IACA,OAAOA;AACT;AAEA,4EAA4E;AAC5E,SAASG,QAAQC,IAAkB,EAAEvD,IAAY,EAAEwD,IAAa;IAC9D,MAAMrC,IAAI,IAAIb,aAAaN;IAC3B,IAAIsB,OAAO;IACX,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKC,QAAQiC,IAAI,CAAClC,EAAE,GAAGkC,IAAI,CAAClC,EAAE;IACxDC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;IACzB,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,GAAGkC,IAAI,CAAClC,EAAE,GAAGC;IAChD,IAAI,CAACkC,MAAM,OAAO;QAAErC;QAAGsC,OAAO;IAAE;IAChC,IAAIC,MAAM;IACV,IAAK,IAAIrC,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKqC,MAAMnC,KAAKmC,GAAG,CAACA,KAAKnC,KAAKoC,GAAG,CAACxC,CAAC,CAACE,EAAE;IAChE,OAAO;QAAEF;QAAGsC,OAAOC,MAAM,OAAO;IAAE;AACpC;AAEA,qFAAqF;AACrF,6CAA6C;AAC7C,OAAO,eAAeE,aAAa9G,EAAgB,EAAEiG,GAAW,EAAEc,OAAe;IAC/E,MAAMC,WAAW,MAAMhB,YAAYC,MAAM,kDAAkD;IAC3F,MAAMgB,QAAQjH,GAAGM,OAAO,CAAC,oFAAoF4G,GAAG;IAChH,IAAID,MAAMtI,MAAM,KAAK,GAAG;IACxB,MAAMwI,YAAY1C,KAAK2C,GAAG,CAACnJ,YAAY+I,SAAS9D,IAAI;IAEpD,MAAMmE,SAAS,IAAItB;IACnB,KAAK,MAAMuB,OAAOL,MAAO;YACVI;QAAb,MAAME,QAAOF,cAAAA,OAAOf,GAAG,CAACgB,IAAIjH,IAAI,eAAnBgH,yBAAAA,cAAwB,EAAE;QACvCE,KAAK1I,IAAI,CAACyI,IAAIE,KAAK;QACnBH,OAAOd,GAAG,CAACe,IAAIjH,IAAI,EAAEkH;IACvB;IAEA,MAAME,OAA6D,EAAE;IACrE,KAAK,MAAM,CAACpH,MAAMqH,UAAU,IAAIL,OAAQ;QACtC,IAAIjI;QACJ,IAAI;YACF,qFAAqF;YACrF,sFAAsF;YACtFA,SAASpB,UAAU;gBAAE2J,SAAStH;gBAAMuH,SAAShK,KAAKmJ,SAAS1G;gBAAOwH,SAAS;gBAAGC,MAAM;gBAAGC,SAAS,EAAE;gBAAElI,OAAO;YAAK,GAAG;gBAACA;aAAM,EAAEmI,GAAG,CAACvH,SAAS,CAACZ,KAAK;QACjJ,EAAE,OAAM;YACN,UAAU,gEAAgE;QAC5E;QACA,KAAK,MAAMe,OAAO8G,UAAW,IAAItI,MAAM,CAACwB,IAAI,EAAE6G,KAAK5I,IAAI,CAAC;YAAEwB;YAAMmH,OAAO5G;YAAKpB,MAAMJ,MAAM,CAACwB,IAAI,CAACpB,IAAI;QAAC;IACrG;IAEA,MAAMyI,SAASjI,GAAGM,OAAO,CAAC;IAC1B,wFAAwF;IACxF,6DAA6D;IAC7D,MAAM4H,SAASnK,SAAS,oBAAoB0J,KAAK9I,MAAM;IACvD,IAAK,IAAID,IAAI,GAAGA,IAAI+I,KAAK9I,MAAM,EAAED,KAAKR,MAAO;QAC3C,MAAMiK,QAAQV,KAAKhI,KAAK,CAACf,GAAGA,IAAIR;QAChC,MAAMkK,UAAU,MAAMpB,SAASnH,KAAK,CAACsI,MAAMvD,GAAG,CAAC,CAACyD,IAAMA,EAAE7I,IAAI;QAC5DQ,GAAGC,IAAI,CAAC;QACR,IAAI;YACFkI,MAAM9I,OAAO,CAAC,CAACiJ,KAAKD;gBAClB,MAAM,EAAEhE,CAAC,EAAEsC,KAAK,EAAE,GAAGH,QAAQ4B,OAAO,CAACC,EAAE,EAAElB,WAAW;gBACpD,MAAMoB,IAAI,IAAIC,UAAUrB;gBACxB,IAAK,IAAI5C,IAAI,GAAGA,IAAI4C,WAAW5C,IAAKgE,CAAC,CAAChE,EAAE,GAAGE,KAAKgE,KAAK,CAACpE,CAAC,CAACE,EAAE,GAAGoC;gBAC7DsB,OAAO1H,GAAG,CAACoG,OAAOnF,OAAOC,IAAI,CAAC8G,EAAE9E,MAAM,GAAG6E,IAAIjI,IAAI,EAAEiI,IAAId,KAAK;YAC9D;YACAxH,GAAGC,IAAI,CAAC;QACV,EAAE,OAAOyI,KAAK;YACZ1I,GAAGC,IAAI,CAAC;YACR,MAAMyI;QACR;QACAR,OAAOS,IAAI,CAAClE,KAAK2C,GAAG,CAAC1I,IAAIR,OAAOuJ,KAAK9I,MAAM;IAC7C;IACAuJ,OAAOU,MAAM;AACf;AAEA,oFAAoF;AACpF,wFAAwF;AACxF,oBAAoB;AACpB,4FAA4F;AAC5F,8FAA8F;AAC9F,2FAA2F;AAC3F,6FAA6F;AAC7F,8BAA8B;AAC9B,OAAO,eAAeC,mBAAmB7I,EAAgB,EAAEiG,GAAW,EAAE6C,KAAa,EAAE3H,MAAa;QAOpF2H;IANd,MAAM/B,UAAU,AAACd,IAAgCc,OAAO;IACxD,IAAI,CAACA,SAAS,MAAM,IAAIjJ,WAAW,eAAe;IAClD,MAAMgJ,aAAa9G,IAAIiG,KAAKc;IAE5B,MAAMC,WAAW,MAAMhB,YAAYC;IACnC,MAAMkB,YAAY1C,KAAK2C,GAAG,CAACnJ,YAAY+I,SAAS9D,IAAI;IACpD,MAAM1D,OAAO,EAACsJ,eAAAA,MAAMC,KAAK,CAAC,gCAAZD,0BAAAA,eAAkC,EAAE,EAAE5J,MAAM,CAAC,CAAC8J,IAAM,CAAC;YAAC;YAAO;YAAM;YAAO;SAAO,CAACC,QAAQ,CAACD,IAAIpL,IAAI,CAAC;IAClH,MAAM,EAAEyG,GAAG6E,EAAE,EAAE,GAAG1C,QAAQ,AAAC,CAAA,MAAMQ,SAASnH,KAAK,CAAC;QAACL;KAAK,CAAA,CAAE,CAAC,EAAE,EAAE2H,WAAW;IAExE,MAAMgC,OAAOnJ,GAAGM,OAAO,CAAC,+FAA+F4G,GAAG;IAQ1H,MAAMkC,OAAO,IAAIrD;IACjB,KAAK,MAAMuB,OAAO6B,KAAM;QACtB,MAAMZ,IAAI,IAAIC,UAAUlB,IAAI+B,MAAM,CAAC5F,MAAM,EAAE6D,IAAI+B,MAAM,CAAChG,UAAU,EAAEoB,KAAK2C,GAAG,CAACD,WAAWG,IAAI+B,MAAM,CAACC,UAAU;QAC3G,IAAIC,MAAM;QACV,IAAK,IAAIhF,IAAI,GAAGA,IAAIgE,EAAE5J,MAAM,EAAE4F,IAAKgF,OAAOhB,CAAC,CAAChE,EAAE,GAAG2E,EAAE,CAAC3E,EAAE;QACtD,MAAMiF,QAAQD,MAAMjC,IAAIX,KAAK;QAC7B,MAAM8C,WAAWL,KAAK9C,GAAG,CAACgB,IAAIjH,IAAI;QAClC,IAAI,CAACoJ,YAAYD,QAAQC,SAASD,KAAK,EAAEJ,KAAK7C,GAAG,CAACe,IAAIjH,IAAI,EAAE;YAAEmJ;YAAOlL,OAAO,CAAC,CAAC,EAAEgJ,IAAIoC,UAAU,CAAC,CAAC,EAAEpC,IAAIqC,QAAQ,EAAE;QAAC;IACnH;IACA,OAAO;WAAIP,KAAKvG,OAAO;KAAG,CACvB+G,IAAI,CAAC,CAACC,GAAGC,IAAMA,CAAC,CAAC,EAAE,CAACN,KAAK,GAAGK,CAAC,CAAC,EAAE,CAACL,KAAK,EACtC/J,KAAK,CAAC,GAAG0B,QACTyD,GAAG,CAAC,CAAC,CAACvE,MAAMyJ,EAAE,GAAM,CAAA;YAAEzJ;YAAM/B,OAAOwL,EAAExL,KAAK;YAAEyL,YAAYtF,KAAKgE,KAAK,CAACqB,EAAEN,KAAK,GAAG,QAAQ;QAAK,CAAA;AAC/F"}
+{"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 { progress } from '../progress.ts';\nimport { parseFile } from '../scan.ts';\nimport type { Feature } from './types.ts';\n\n// Heading-based chunks, int8 vectors with a per-vector scale, NULL vector = not yet embedded.\n// Reconcile writes dirty rows; embedding tops up on the next search, so staleness costs recall.\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;\n// Seed chunks that participate in a `related` scan; see similarNotes for the measurement.\nconst TARGET_CHUNK_CAP = 16;\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, so embed time can re-derive text from stored line ranges. Heading-delimited,\n// preamble kept, whole body when no headings; the title/summary prefix mirrors bm25 weighting.\n//\n// Chunks the body, not the raw file, with `offset` shifting line numbers back onto the raw\n// file so a range stays a direct Read range (sections is 1-indexed over raw too). Chunking raw\n// put the frontmatter block in its own leading chunk on every note with a heading, which made\n// `lines` point at YAML and made frontmatter-only notes near-identical to each other. It also\n// disagreed with FTS, which indexes the body alone.\nfunction chunksOf(body: string, search?: { title: string; summary: string }, offset = 0): Chunk[] {\n  const lines = body.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    // A body with nothing in it yields no chunks at all, so a frontmatter-only note has no\n    // vectors rather than a vector of its own YAML.\n    if (text.length > 0) chunks.push({ startLine: start + offset, endLine: end + offset, 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    // Lines the frontmatter occupies, so body line 1 maps back to its raw line number.\n    return chunksOf(body, search, raw.split('\\n').length - body.split('\\n').length);\n  },\n  remove(db, path) {\n    db.prepare('DELETE FROM embeddings WHERE \"path\" = ?').run(path);\n  },\n  // A tree with no embedding model never had extract() run for the doc (db.ts's per-file\n  // filter skips it), so extracted is undefined here -- store nothing, i.e. no rows.\n  store(db, path, extracted) {\n    if (!extracted) return;\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  enabledForFile(_cfg, file) {\n    return file.embed;\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\nconst MODEL_FILES = ['model.safetensors', 'tokenizer.json'];\n// The pair, for messages that name what a local model directory must contain.\nexport const MODEL_FILENAMES = MODEL_FILES.join(' and ');\n\n// A Hugging Face repo id: one slash, HF's charset. Anything else is a path, used as one rather\n// than flattened into a cache key Windows would reject.\nconst HF_MODEL_ID = /^[A-Za-z0-9][\\w.-]*\\/[A-Za-z0-9][\\w.-]*$/;\n\n// Downloadable iff it names a Hugging Face repo; a local path is the caller's to populate.\nexport function isDownloadable(model: string): boolean {\n  return HF_MODEL_ID.test(model);\n}\n\n// Machine-wide, not per-tree: `.sense/` is the index a rebuild throws away, while a 124 MB\n// model is shared by every tree on the machine. Honors XDG_CACHE_HOME where it is set;\n// elsewhere ~/.cache, which is also where Hugging Face's own libraries keep theirs.\nfunction cacheRoot(): string {\n  const xdg = process.env.XDG_CACHE_HOME;\n  return join(xdg && xdg.length > 0 ? xdg : join(homedir(), '.cache'), 'sensemaking', 'models');\n}\n\n// A model is either a local directory the caller pointed at, or the cache `sense download`\n// fills. Nothing here fetches: an absent model degrades search to its other signals rather\n// than pulling 124 MB out of a command that reads like a query.\nexport function modelDir(model: string): string {\n  if (!HF_MODEL_ID.test(model)) return model;\n  return join(cacheRoot(), model.replace(/\\//g, '--'));\n}\n\n// Both files, so an interrupted download reads as absent and the next one resumes it.\nexport function hasModelFiles(model: string): boolean {\n  const dir = modelDir(model);\n  return MODEL_FILES.every((file) => existsSync(join(dir, file)));\n}\n\n// api providers have nothing to download, so they are never \"missing\" here; an unreachable\n// endpoint surfaces as an EMBED_MODEL error at call time instead.\nexport function modelPresent(cfg: Config): boolean {\n  const e = embedConfig(cfg);\n  if (!e) return false;\n  return e.type === 'api' || hasModelFiles(e.model);\n}\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(`${dest}.part`, dest);\n  });\n}\n\n// The only code path that touches the network for weights. `sense download` calls it; no\n// query ever does. Idempotent: a file already on disk is left alone.\nexport async function downloadModel(model: string, onFile?: (file: string, dir: string) => void): Promise<string> {\n  if (!isDownloadable(model)) {\n    throw new SenseError('EMBED_MODEL', `embed.model \"${model}\" is a local path, not a Hugging Face model id, so there is nothing to download; put model.safetensors and tokenizer.json in that directory, or name a model id like \"minishlab/potion-retrieval-32M\"`);\n  }\n  const dir = modelDir(model);\n  mkdirSync(dir, { recursive: true });\n  for (const file of MODEL_FILES) {\n    if (existsSync(join(dir, file))) continue;\n    onFile?.(file, dir);\n    await fetchToFile(`https://huggingface.co/${model}/resolve/main/${file}`, join(dir, file));\n  }\n  return dir;\n}\n\nasync function staticProvider(model: string): Promise<EmbedProvider> {\n  const dir = modelDir(model);\n  for (const file of MODEL_FILES) {\n    if (!existsSync(join(dir, file))) {\n      // A local path the caller controls gets told what is missing where; only a repo id can\n      // be fixed by downloading.\n      const fix = isDownloadable(model) ? 'run `sense download`' : `expected ${MODEL_FILENAMES} in that directory`;\n      throw new SenseError('EMBED_MODEL_MISSING', `embed model ${model} is not available (looked in ${dir}); ${fix}`);\n    }\n  }\n\n  const raw = readFileSync(join(dir, 'model.safetensors'));\n  const headerLen = Number(raw.readBigUInt64LE(0));\n  const header = JSON.parse(raw.subarray(8, 8 + headerLen).toString('utf8')) as Record<string, { dtype: string; shape: number[]; data_offsets: number[] }>;\n  const entry = Object.entries(header).find(([k]) => k !== '__metadata__');\n  if (!entry || entry[1].dtype !== 'F32') throw new SenseError('EMBED_MODEL', `${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    // The tokenizer yields undefined (not the unk id) for tokens outside the vocab; an\n    // undefined id would index the matrix at NaN and poison the whole mean-pool -- and the\n    // int8 conversion then stores the NaN vector as all zeros, silently. Keep integers only.\n    const ids = (tok.encode(text, { add_special_tokens: false }).ids as number[]).filter((id) => Number.isInteger(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}:${model}`, dims, embed: post };\n}\n\nconst providers = new Map<string, Promise<EmbedProvider>>();\n\nfunction getProvider(cfg: Config): Promise<EmbedProvider> {\n  const e = embedConfig(cfg);\n  if (!e) throw new SenseError('EMBED_DISABLED', 'this tree has no embedding model: add an \"embed\" block naming one to sense.config.json, then run `sense download`');\n  const sig = `${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      // presets/embed are irrelevant here -- re-deriving chunk text for a doc that already\n      // has embeddings rows means the tree had an embedding model at reconcile time.\n      chunks = parseFile({ relPath: path, absPath: join(baseDir, path), mtimeMs: 0, size: 0, presets: [], embed: true }, [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  // The lazy build is the one long silence a first search hits (measured 23s at\n  // 26k notes); progress makes it distinguishable from a hang.\n  const report = progress('embedding chunks', jobs.length);\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    report.tick(Math.min(i + BATCH, jobs.length));\n  }\n  report.finish();\n}\n\n// Dequantised int8 dot products land a little either side of a true cosine, so an identical\n// pair prints 1.001 and undermines a column whose whole job is being a bounded number.\nfunction asCosine(score: number): number {\n  return Math.round(Math.min(1, Math.max(-1, score)) * 1000) / 1000;\n}\n\n// Best chunk per file by cosine, its line range riding along; FTS5 operators are stripped as\n// lexical syntax. Similarity comes back because the fused score cannot express match quality,\n// and nearest-neighbour search always returns a neighbour however far away.\nexport async function semanticCandidates(db: DatabaseSync, cfg: Config, terms: string, fetch: number): Promise<Array<{ path: string; lines: string; similarity: number }>> {\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, similarity: asCosine(b.score) }));\n}\n\n// Whether a note has any chunk with a vector. Distinguishes \"nothing is near this note\" from\n// \"this note has no text\", which look the same in an empty result.\nexport function hasEmbedding(db: DatabaseSync, path: string): boolean {\n  const row = db.prepare('SELECT 1 AS ok FROM embeddings WHERE \"path\" = ? AND vector IS NOT NULL LIMIT 1').get(path) as { ok: number } | undefined;\n  return row !== undefined;\n}\n\n// Note-to-note similarity is the max cosine over (target chunk, other chunk) pairs, one linear\n// scan of stored vectors. Reads only what is stored, so it stays sync.\nexport function similarNotes(db: DatabaseSync, _cfg: Config, path: string, opts: { exclude: Set<string>; allowed?: Set<string>; k: number }): Array<{ path: string; similarity: number }> {\n  // Cost is target_chunks x stored_chunks, so a heading-dense seed multiplies a full-corpus\n  // scan (12.7s at 201 chunks/note). Sample evenly, so late sections still get a vote.\n  const targetRows = db.prepare('SELECT scale, vector FROM embeddings WHERE \"path\" = ? AND vector IS NOT NULL ORDER BY chunk').all(path) as Array<{ scale: number; vector: Uint8Array }>;\n  if (targetRows.length === 0) return [];\n  const step = Math.max(1, Math.ceil(targetRows.length / TARGET_CHUNK_CAP));\n  const target = targetRows.filter((_, i) => i % step === 0).map((row) => ({ v: new Int8Array(row.vector.buffer, row.vector.byteOffset, row.vector.byteLength), scale: row.scale }));\n\n  const rows = db.prepare('SELECT \"path\", scale, vector FROM embeddings WHERE vector IS NOT NULL').all() as Array<{ path: string; scale: number; vector: Uint8Array }>;\n  const best = new Map<string, number>();\n  for (const row of rows) {\n    if (row.path === path || opts.exclude.has(row.path) || (opts.allowed && !opts.allowed.has(row.path))) continue;\n    const other = new Int8Array(row.vector.buffer, row.vector.byteOffset, row.vector.byteLength);\n    for (const t of target) {\n      let dot = 0;\n      const len = Math.min(t.v.length, other.length);\n      for (let d = 0; d < len; d++) dot += t.v[d] * other[d];\n      const score = dot * t.scale * row.scale;\n      const existing = best.get(row.path);\n      if (existing === undefined || score > existing) best.set(row.path, score);\n    }\n  }\n\n  return [...best.entries()]\n    .sort((a, b) => b[1] - a[1])\n    .slice(0, opts.k)\n    .map(([p, score]) => ({ path: p, similarity: asCosine(score) }));\n}\n"],"names":["existsSync","mkdirSync","readFileSync","renameSync","writeFileSync","homedir","join","embedConfig","SenseError","progress","parseFile","STORE_DIMS","TARGET_CHUNK_CAP","BATCH","chunksOf","body","search","offset","lines","split","starts","inFence","i","length","test","push","bounds","prefix","title","summary","filter","Boolean","chunks","forEach","start","end","text","slice","trim","startLine","endLine","embed","name","schema","db","exec","extract","raw","remove","path","prepare","run","store","extracted","insert","c","idx","enabledForFile","_cfg","file","MODEL_FILES","MODEL_FILENAMES","HF_MODEL_ID","isDownloadable","model","cacheRoot","xdg","process","env","XDG_CACHE_HOME","modelDir","replace","hasModelFiles","dir","every","modelPresent","cfg","e","type","fetchToFile","url","dest","fetch","then","res","ok","status","Buffer","from","arrayBuffer","downloadModel","onFile","recursive","staticProvider","tokenizerJson","fix","headerLen","Number","readBigUInt64LE","header","JSON","parse","subarray","toString","entry","Object","entries","find","k","dtype","spec","dims","shape","dataStart","byteOffset","data_offsets","matrix","Float32Array","buffer","Tokenizer","tok","unkId","vocab","unk_token","one","ids","encode","add_special_tokens","id","isInteger","v","off","d","norm","Math","sqrt","texts","map","apiProvider","keyEnv","base","headers","key","undefined","authorization","post","method","stringify","input","json","data","embedding","providers","Map","getProvider","sig","p","get","set","toStore","full","int8","scale","max","abs","embedPending","baseDir","provider","dirty","all","storeDims","min","byPath","row","list","chunk","jobs","chunkIdxs","relPath","absPath","mtimeMs","size","presets","doc","update","report","batch","vectors","j","job","q","Int8Array","round","err","tick","finish","asCosine","score","semanticCandidates","terms","match","t","includes","qv","rows","best","vector","byteLength","dot","existing","start_line","end_line","sort","a","b","similarity","hasEmbedding","similarNotes","opts","targetRows","step","ceil","target","_","exclude","has","allowed","other","len"],"mappings":"AAAA,SAASA,UAAU,EAAEC,SAAS,EAAEC,YAAY,EAAEC,UAAU,EAAEC,aAAa,QAAQ,UAAU;AACzF,SAASC,OAAO,QAAQ,UAAU;AAClC,SAASC,IAAI,QAAQ,YAAY;AAGjC,SAASC,WAAW,QAAQ,eAAe;AAC3C,SAASC,UAAU,QAAQ,eAAe;AAC1C,SAASC,QAAQ,QAAQ,iBAAiB;AAC1C,SAASC,SAAS,QAAQ,aAAa;AAGvC,8FAA8F;AAC9F,gGAAgG;AAEhG,6EAA6E;AAC7E,yEAAyE;AACzE,MAAMC,aAAa;AACnB,0FAA0F;AAC1F,MAAMC,mBAAmB;AACzB,MAAMC,QAAQ;AAcd,8FAA8F;AAC9F,+FAA+F;AAC/F,EAAE;AACF,2FAA2F;AAC3F,+FAA+F;AAC/F,8FAA8F;AAC9F,8FAA8F;AAC9F,oDAAoD;AACpD,SAASC,SAASC,IAAY,EAAEC,MAA2C,EAAEC,SAAS,CAAC;IACrF,MAAMC,QAAQH,KAAKI,KAAK,CAAC;IACzB,MAAMC,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,MAAMI,SAASN,OAAOG,MAAM,KAAK,IAAI;QAAC;KAAE,GAAGH,MAAM,CAAC,EAAE,GAAG,IAAI;QAAC;WAAMA;KAAO,GAAGA;IAC5E,MAAMO,SAAS;QAACX,mBAAAA,6BAAAA,OAAQY,KAAK;QAAEZ,mBAAAA,6BAAAA,OAAQa,OAAO;KAAC,CAACC,MAAM,CAACC,SAASzB,IAAI,CAAC;IACrE,MAAM0B,SAAkB,EAAE;IAC1BN,OAAOO,OAAO,CAAC,CAACC,OAAOZ;QACrB,MAAMa,MAAMb,IAAI,IAAII,OAAOH,MAAM,GAAGG,MAAM,CAACJ,IAAI,EAAE,GAAG,IAAIJ,MAAMK,MAAM;QACpE,MAAMa,OAAOlB,MACVmB,KAAK,CAACH,QAAQ,GAAGC,KACjB7B,IAAI,CAAC,MACLgC,IAAI;QACP,uFAAuF;QACvF,gDAAgD;QAChD,IAAIF,KAAKb,MAAM,GAAG,GAAGS,OAAOP,IAAI,CAAC;YAAEc,WAAWL,QAAQjB;YAAQuB,SAASL,MAAMlB;YAAQmB,MAAMT,SAAS,GAAGA,OAAO,EAAE,EAAES,MAAM,GAAGA;QAAK;IAClI;IACA,OAAOJ;AACT;AAEA,OAAO,MAAMS,QAAiB;IAC5BC,MAAM;IACNC,QAAOC,EAAE;QACPA,GAAGC,IAAI,CAAC,CAAC,8JAA8J,CAAC;IAC1K;IACAC,SAAQC,GAAG,EAAEhC,IAAI,EAAEC,MAAM;QACvB,mFAAmF;QACnF,OAAOF,SAASC,MAAMC,QAAQ+B,IAAI5B,KAAK,CAAC,MAAMI,MAAM,GAAGR,KAAKI,KAAK,CAAC,MAAMI,MAAM;IAChF;IACAyB,QAAOJ,EAAE,EAAEK,IAAI;QACbL,GAAGM,OAAO,CAAC,2CAA2CC,GAAG,CAACF;IAC5D;IACA,uFAAuF;IACvF,mFAAmF;IACnFG,OAAMR,EAAE,EAAEK,IAAI,EAAEI,SAAS;QACvB,IAAI,CAACA,WAAW;QAChB,MAAMC,SAASV,GAAGM,OAAO,CAAC;QACzBG,UAAsBpB,OAAO,CAAC,CAACsB,GAAGC,MAAQF,OAAOH,GAAG,CAACF,MAAMO,KAAKD,EAAEhB,SAAS,EAAEgB,EAAEf,OAAO;IACzF;IACAiB,gBAAeC,IAAI,EAAEC,IAAI;QACvB,OAAOA,KAAKlB,KAAK;IACnB;AACF,EAAE;AAEF,2FAA2F;AAC3F,yFAAyF;AACzF,gBAAgB;AAEhB,MAAMmB,cAAc;IAAC;IAAqB;CAAiB;AAC3D,8EAA8E;AAC9E,OAAO,MAAMC,kBAAkBD,YAAYtD,IAAI,CAAC,SAAS;AAEzD,+FAA+F;AAC/F,wDAAwD;AACxD,MAAMwD,cAAc;AAEpB,2FAA2F;AAC3F,OAAO,SAASC,eAAeC,KAAa;IAC1C,OAAOF,YAAYtC,IAAI,CAACwC;AAC1B;AAEA,2FAA2F;AAC3F,uFAAuF;AACvF,oFAAoF;AACpF,SAASC;IACP,MAAMC,MAAMC,QAAQC,GAAG,CAACC,cAAc;IACtC,OAAO/D,KAAK4D,OAAOA,IAAI3C,MAAM,GAAG,IAAI2C,MAAM5D,KAAKD,WAAW,WAAW,eAAe;AACtF;AAEA,2FAA2F;AAC3F,2FAA2F;AAC3F,gEAAgE;AAChE,OAAO,SAASiE,SAASN,KAAa;IACpC,IAAI,CAACF,YAAYtC,IAAI,CAACwC,QAAQ,OAAOA;IACrC,OAAO1D,KAAK2D,aAAaD,MAAMO,OAAO,CAAC,OAAO;AAChD;AAEA,sFAAsF;AACtF,OAAO,SAASC,cAAcR,KAAa;IACzC,MAAMS,MAAMH,SAASN;IACrB,OAAOJ,YAAYc,KAAK,CAAC,CAACf,OAAS3D,WAAWM,KAAKmE,KAAKd;AAC1D;AAEA,2FAA2F;AAC3F,kEAAkE;AAClE,OAAO,SAASgB,aAAaC,GAAW;IACtC,MAAMC,IAAItE,YAAYqE;IACtB,IAAI,CAACC,GAAG,OAAO;IACf,OAAOA,EAAEC,IAAI,KAAK,SAASN,cAAcK,EAAEb,KAAK;AAClD;AAEA,SAASe,YAAYC,GAAW,EAAEC,IAAY;IAC5C,OAAOC,MAAMF,KAAKG,IAAI,CAAC,OAAOC;QAC5B,IAAI,CAACA,IAAIC,EAAE,EAAE,MAAM,IAAI7E,WAAW,eAAe,CAAC,uBAAuB,EAAEwE,IAAI,SAAS,EAAEI,IAAIE,MAAM,EAAE;QACtGlF,cAAc,GAAG6E,KAAK,KAAK,CAAC,EAAEM,OAAOC,IAAI,CAAC,MAAMJ,IAAIK,WAAW;QAC/DtF,WAAW,GAAG8E,KAAK,KAAK,CAAC,EAAEA;IAC7B;AACF;AAEA,yFAAyF;AACzF,qEAAqE;AACrE,OAAO,eAAeS,cAAc1B,KAAa,EAAE2B,MAA4C;IAC7F,IAAI,CAAC5B,eAAeC,QAAQ;QAC1B,MAAM,IAAIxD,WAAW,eAAe,CAAC,aAAa,EAAEwD,MAAM,qMAAqM,CAAC;IAClQ;IACA,MAAMS,MAAMH,SAASN;IACrB/D,UAAUwE,KAAK;QAAEmB,WAAW;IAAK;IACjC,KAAK,MAAMjC,QAAQC,YAAa;QAC9B,IAAI5D,WAAWM,KAAKmE,KAAKd,QAAQ;QACjCgC,mBAAAA,6BAAAA,OAAShC,MAAMc;QACf,MAAMM,YAAY,CAAC,uBAAuB,EAAEf,MAAM,cAAc,EAAEL,MAAM,EAAErD,KAAKmE,KAAKd;IACtF;IACA,OAAOc;AACT;AAEA,eAAeoB,eAAe7B,KAAa;;QAyBE8B,sBAA7BA,4BAAAA;IAxBd,MAAMrB,MAAMH,SAASN;IACrB,KAAK,MAAML,QAAQC,YAAa;QAC9B,IAAI,CAAC5D,WAAWM,KAAKmE,KAAKd,QAAQ;YAChC,uFAAuF;YACvF,2BAA2B;YAC3B,MAAMoC,MAAMhC,eAAeC,SAAS,yBAAyB,CAAC,SAAS,EAAEH,gBAAgB,kBAAkB,CAAC;YAC5G,MAAM,IAAIrD,WAAW,uBAAuB,CAAC,YAAY,EAAEwD,MAAM,6BAA6B,EAAES,IAAI,GAAG,EAAEsB,KAAK;QAChH;IACF;IAEA,MAAMhD,MAAM7C,aAAaI,KAAKmE,KAAK;IACnC,MAAMuB,YAAYC,OAAOlD,IAAImD,eAAe,CAAC;IAC7C,MAAMC,SAASC,KAAKC,KAAK,CAACtD,IAAIuD,QAAQ,CAAC,GAAG,IAAIN,WAAWO,QAAQ,CAAC;IAClE,MAAMC,QAAQC,OAAOC,OAAO,CAACP,QAAQQ,IAAI,CAAC,CAAC,CAACC,EAAE,GAAKA,MAAM;IACzD,IAAI,CAACJ,SAASA,KAAK,CAAC,EAAE,CAACK,KAAK,KAAK,OAAO,MAAM,IAAIrG,WAAW,eAAe,GAAGwD,MAAM,oCAAoC,CAAC;IAC1H,MAAM8C,OAAON,KAAK,CAAC,EAAE;IACrB,MAAMO,OAAOD,KAAKE,KAAK,CAAC,EAAE;IAC1B,MAAMC,YAAYlE,IAAImE,UAAU,GAAG,IAAIlB,YAAYc,KAAKK,YAAY,CAAC,EAAE;IACvE,MAAMC,SAASH,YAAY,MAAM,IAAI,IAAII,aAAatE,IAAIuE,MAAM,EAAEL,WAAWH,KAAKE,KAAK,CAAC,EAAE,GAAGD,QAAQ,IAAIM,aAAatE,IAAIuE,MAAM,CAACjF,KAAK,CAAC4E,WAAWA,YAAYH,KAAKE,KAAK,CAAC,EAAE,GAAGD,OAAO;IAErL,MAAMjB,gBAAgBM,KAAKC,KAAK,CAACnG,aAAaI,KAAKmE,KAAK,mBAAmB;IAC3E,oFAAoF;IACpF,MAAM,EAAE8C,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC;IACnC,MAAMC,MAAM,IAAID,UAAUzB,eAAe,CAAC;IAC1C,MAAM2B,iBAAQ3B,wBAAAA,cAAc9B,KAAK,cAAnB8B,6CAAAA,6BAAAA,sBAAqB4B,KAAK,cAA1B5B,iDAAAA,0BAA4B,EAACA,uBAAAA,cAAc9B,KAAK,cAAnB8B,2CAAAA,qBAAqB6B,SAAS,CAAC,uCAAI,CAAC;IAE/E,SAASC,IAAIxF,IAAY;QACvB,mFAAmF;QACnF,uFAAuF;QACvF,yFAAyF;QACzF,MAAMyF,MAAM,AAACL,IAAIM,MAAM,CAAC1F,MAAM;YAAE2F,oBAAoB;QAAM,GAAGF,GAAG,CAAc/F,MAAM,CAAC,CAACkG,KAAO/B,OAAOgC,SAAS,CAACD,OAAOA,OAAOP;QAC5H,MAAMS,IAAI,IAAIb,aAAaN;QAC3B,IAAIc,IAAItG,MAAM,KAAK,GAAG,OAAO2G;QAC7B,KAAK,MAAMF,MAAMH,IAAK;YACpB,MAAMM,MAAMH,KAAKjB;YACjB,IAAK,IAAIqB,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,IAAIhB,MAAM,CAACe,MAAMC,EAAE;QACxD;QACA,IAAIC,OAAO;QACX,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAK;YAC7BF,CAAC,CAACE,EAAE,IAAIP,IAAItG,MAAM;YAClB8G,QAAQH,CAAC,CAACE,EAAE,GAAGF,CAAC,CAACE,EAAE;QACrB;QACAC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;QACzB,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,IAAIC;QACvC,OAAOH;IACT;IAEA,OAAO;QAAEF,IAAI,CAAC,OAAO,EAAEhE,OAAO;QAAE+C;QAAMtE,OAAO,OAAO+F,QAAUA,MAAMC,GAAG,CAACb;IAAK;AAC/E;AAEA,gFAAgF;AAEhF,eAAec,YAAY1E,KAAa,EAAEgB,GAAuB,EAAE2D,MAA0B;IAC3F,IAAI,CAAC3D,KAAK,MAAM,IAAIxE,WAAW,eAAe;IAC9C,MAAMoI,OAAO5D,IAAIT,OAAO,CAAC,QAAQ;IACjC,MAAMsE,UAAkC;QAAE,gBAAgB;IAAmB;IAC7E,MAAMC,MAAMH,SAASxE,QAAQC,GAAG,CAACuE,OAAO,GAAGI;IAC3C,IAAID,KAAKD,QAAQG,aAAa,GAAG,CAAC,OAAO,EAAEF,KAAK;IAEhD,eAAeG,KAAKT,KAAe;QACjC,MAAMpD,MAAM,MAAMF,MAAM,GAAG0D,KAAK,WAAW,CAAC,EAAE;YAAEM,QAAQ;YAAQL;YAAS9H,MAAMqF,KAAK+C,SAAS,CAAC;gBAAEnF;gBAAOoF,OAAOZ;YAAM;QAAG;QACvH,IAAI,CAACpD,IAAIC,EAAE,EAAE,MAAM,IAAI7E,WAAW,eAAe,GAAGoI,KAAK,oBAAoB,EAAExD,IAAIE,MAAM,EAAE;QAC3F,MAAMvE,OAAQ,MAAMqE,IAAIiE,IAAI;QAC5B,OAAOtI,KAAKuI,IAAI,CAACb,GAAG,CAAC,CAACL,IAAMf,aAAa7B,IAAI,CAAC4C,EAAEmB,SAAS;IAC3D;IAEA,MAAMxC,OAAO,AAAC,CAAA,MAAMkC,KAAK;QAAC;KAAkB,CAAA,CAAE,CAAC,EAAE,CAAC1H,MAAM;IACxD,OAAO;QAAEyG,IAAI,CAAC,IAAI,EAAEY,KAAK,CAAC,EAAE5E,OAAO;QAAE+C;QAAMtE,OAAOwG;IAAK;AACzD;AAEA,MAAMO,YAAY,IAAIC;AAEtB,SAASC,YAAY9E,GAAW;QAGMC;IAFpC,MAAMA,IAAItE,YAAYqE;IACtB,IAAI,CAACC,GAAG,MAAM,IAAIrE,WAAW,kBAAkB;IAC/C,MAAMmJ,MAAM,GAAG9E,EAAEC,IAAI,CAAC,CAAC,EAAED,EAAEb,KAAK,CAAC,CAAC,GAAEa,SAAAA,EAAEG,GAAG,cAALH,oBAAAA,SAAS,IAAI;IACjD,IAAI+E,IAAIJ,UAAUK,GAAG,CAACF;IACtB,IAAI,CAACC,GAAG;QACNA,IAAI/E,EAAEC,IAAI,KAAK,QAAQ4D,YAAY7D,EAAEb,KAAK,EAAEa,EAAEG,GAAG,EAAEH,EAAEiE,GAAG,IAAIjD,eAAehB,EAAEb,KAAK;QAClFwF,UAAUM,GAAG,CAACH,KAAKC;IACrB;IACA,OAAOA;AACT;AAEA,4EAA4E;AAC5E,SAASG,QAAQC,IAAkB,EAAEjD,IAAY,EAAEkD,IAAa;IAC9D,MAAM/B,IAAI,IAAIb,aAAaN;IAC3B,IAAIsB,OAAO;IACX,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKC,QAAQ2B,IAAI,CAAC5B,EAAE,GAAG4B,IAAI,CAAC5B,EAAE;IACxDC,OAAOC,KAAKC,IAAI,CAACF,QAAQ;IACzB,IAAK,IAAID,IAAI,GAAGA,IAAIrB,MAAMqB,IAAKF,CAAC,CAACE,EAAE,GAAG4B,IAAI,CAAC5B,EAAE,GAAGC;IAChD,IAAI,CAAC4B,MAAM,OAAO;QAAE/B;QAAGgC,OAAO;IAAE;IAChC,IAAIC,MAAM;IACV,IAAK,IAAI/B,IAAI,GAAGA,IAAIrB,MAAMqB,IAAK+B,MAAM7B,KAAK6B,GAAG,CAACA,KAAK7B,KAAK8B,GAAG,CAAClC,CAAC,CAACE,EAAE;IAChE,OAAO;QAAEF;QAAGgC,OAAOC,MAAM,OAAO;IAAE;AACpC;AAEA,qFAAqF;AACrF,6CAA6C;AAC7C,OAAO,eAAeE,aAAazH,EAAgB,EAAEgC,GAAW,EAAE0F,OAAe;IAC/E,MAAMC,WAAW,MAAMb,YAAY9E,MAAM,kDAAkD;IAC3F,MAAM4F,QAAQ5H,GAAGM,OAAO,CAAC,oFAAoFuH,GAAG;IAChH,IAAID,MAAMjJ,MAAM,KAAK,GAAG;IACxB,MAAMmJ,YAAYpC,KAAKqC,GAAG,CAAChK,YAAY4J,SAASxD,IAAI;IAEpD,MAAM6D,SAAS,IAAInB;IACnB,KAAK,MAAMoB,OAAOL,MAAO;YACVI;QAAb,MAAME,QAAOF,cAAAA,OAAOf,GAAG,CAACgB,IAAI5H,IAAI,eAAnB2H,yBAAAA,cAAwB,EAAE;QACvCE,KAAKrJ,IAAI,CAACoJ,IAAIE,KAAK;QACnBH,OAAOd,GAAG,CAACe,IAAI5H,IAAI,EAAE6H;IACvB;IAEA,MAAME,OAA6D,EAAE;IACrE,KAAK,MAAM,CAAC/H,MAAMgI,UAAU,IAAIL,OAAQ;QACtC,IAAI5I;QACJ,IAAI;YACF,qFAAqF;YACrF,+EAA+E;YAC/EA,SAAStB,UAAU;gBAAEwK,SAASjI;gBAAMkI,SAAS7K,KAAKgK,SAASrH;gBAAOmI,SAAS;gBAAGC,MAAM;gBAAGC,SAAS,EAAE;gBAAE7I,OAAO;YAAK,GAAG;gBAACA;aAAM,EAAE8I,GAAG,CAAClI,SAAS,CAACZ,KAAK;QACjJ,EAAE,OAAM;YACN,UAAU,gEAAgE;QAC5E;QACA,KAAK,MAAMe,OAAOyH,UAAW,IAAIjJ,MAAM,CAACwB,IAAI,EAAEwH,KAAKvJ,IAAI,CAAC;YAAEwB;YAAM8H,OAAOvH;YAAKpB,MAAMJ,MAAM,CAACwB,IAAI,CAACpB,IAAI;QAAC;IACrG;IAEA,MAAMoJ,SAAS5I,GAAGM,OAAO,CAAC;IAC1B,8EAA8E;IAC9E,6DAA6D;IAC7D,MAAMuI,SAAShL,SAAS,oBAAoBuK,KAAKzJ,MAAM;IACvD,IAAK,IAAID,IAAI,GAAGA,IAAI0J,KAAKzJ,MAAM,EAAED,KAAKT,MAAO;QAC3C,MAAM6K,QAAQV,KAAK3I,KAAK,CAACf,GAAGA,IAAIT;QAChC,MAAM8K,UAAU,MAAMpB,SAAS9H,KAAK,CAACiJ,MAAMjD,GAAG,CAAC,CAACmD,IAAMA,EAAExJ,IAAI;QAC5DQ,GAAGC,IAAI,CAAC;QACR,IAAI;YACF6I,MAAMzJ,OAAO,CAAC,CAAC4J,KAAKD;gBAClB,MAAM,EAAE1D,CAAC,EAAEgC,KAAK,EAAE,GAAGH,QAAQ4B,OAAO,CAACC,EAAE,EAAElB,WAAW;gBACpD,MAAMoB,IAAI,IAAIC,UAAUrB;gBACxB,IAAK,IAAItC,IAAI,GAAGA,IAAIsC,WAAWtC,IAAK0D,CAAC,CAAC1D,EAAE,GAAGE,KAAK0D,KAAK,CAAC9D,CAAC,CAACE,EAAE,GAAG8B;gBAC7DsB,OAAOrI,GAAG,CAAC+G,OAAO3E,OAAOC,IAAI,CAACsG,EAAExE,MAAM,GAAGuE,IAAI5I,IAAI,EAAE4I,IAAId,KAAK;YAC9D;YACAnI,GAAGC,IAAI,CAAC;QACV,EAAE,OAAOoJ,KAAK;YACZrJ,GAAGC,IAAI,CAAC;YACR,MAAMoJ;QACR;QACAR,OAAOS,IAAI,CAAC5D,KAAKqC,GAAG,CAACrJ,IAAIT,OAAOmK,KAAKzJ,MAAM;IAC7C;IACAkK,OAAOU,MAAM;AACf;AAEA,4FAA4F;AAC5F,uFAAuF;AACvF,SAASC,SAASC,KAAa;IAC7B,OAAO/D,KAAK0D,KAAK,CAAC1D,KAAKqC,GAAG,CAAC,GAAGrC,KAAK6B,GAAG,CAAC,CAAC,GAAGkC,UAAU,QAAQ;AAC/D;AAEA,6FAA6F;AAC7F,8FAA8F;AAC9F,4EAA4E;AAC5E,OAAO,eAAeC,mBAAmB1J,EAAgB,EAAEgC,GAAW,EAAE2H,KAAa,EAAErH,MAAa;QAOpFqH;IANd,MAAMjC,UAAU,AAAC1F,IAAgC0F,OAAO;IACxD,IAAI,CAACA,SAAS,MAAM,IAAI9J,WAAW,eAAe;IAClD,MAAM6J,aAAazH,IAAIgC,KAAK0F;IAE5B,MAAMC,WAAW,MAAMb,YAAY9E;IACnC,MAAM8F,YAAYpC,KAAKqC,GAAG,CAAChK,YAAY4J,SAASxD,IAAI;IACpD,MAAM3E,OAAO,EAACmK,eAAAA,MAAMC,KAAK,CAAC,gCAAZD,0BAAAA,eAAkC,EAAE,EAAEzK,MAAM,CAAC,CAAC2K,IAAM,CAAC;YAAC;YAAO;YAAM;YAAO;SAAO,CAACC,QAAQ,CAACD,IAAInM,IAAI,CAAC;IAClH,MAAM,EAAE4H,GAAGyE,EAAE,EAAE,GAAG5C,QAAQ,AAAC,CAAA,MAAMQ,SAAS9H,KAAK,CAAC;QAACL;KAAK,CAAA,CAAE,CAAC,EAAE,EAAEsI,WAAW;IAExE,MAAMkC,OAAOhK,GAAGM,OAAO,CAAC,+FAA+FuH,GAAG;IAQ1H,MAAMoC,OAAO,IAAIpD;IACjB,KAAK,MAAMoB,OAAO+B,KAAM;QACtB,MAAMd,IAAI,IAAIC,UAAUlB,IAAIiC,MAAM,CAACxF,MAAM,EAAEuD,IAAIiC,MAAM,CAAC5F,UAAU,EAAEoB,KAAKqC,GAAG,CAACD,WAAWG,IAAIiC,MAAM,CAACC,UAAU;QAC3G,IAAIC,MAAM;QACV,IAAK,IAAI5E,IAAI,GAAGA,IAAI0D,EAAEvK,MAAM,EAAE6G,IAAK4E,OAAOlB,CAAC,CAAC1D,EAAE,GAAGuE,EAAE,CAACvE,EAAE;QACtD,MAAMiE,QAAQW,MAAMnC,IAAIX,KAAK;QAC7B,MAAM+C,WAAWJ,KAAKhD,GAAG,CAACgB,IAAI5H,IAAI;QAClC,IAAI,CAACgK,YAAYZ,QAAQY,SAASZ,KAAK,EAAEQ,KAAK/C,GAAG,CAACe,IAAI5H,IAAI,EAAE;YAAEoJ;YAAOnL,OAAO,CAAC,CAAC,EAAE2J,IAAIqC,UAAU,CAAC,CAAC,EAAErC,IAAIsC,QAAQ,EAAE;QAAC;IACnH;IACA,OAAO;WAAIN,KAAKnG,OAAO;KAAG,CACvB0G,IAAI,CAAC,CAACC,GAAGC,IAAMA,CAAC,CAAC,EAAE,CAACjB,KAAK,GAAGgB,CAAC,CAAC,EAAE,CAAChB,KAAK,EACtChK,KAAK,CAAC,GAAG6C,QACTuD,GAAG,CAAC,CAAC,CAACxF,MAAMqK,EAAE,GAAM,CAAA;YAAErK;YAAM/B,OAAOoM,EAAEpM,KAAK;YAAEqM,YAAYnB,SAASkB,EAAEjB,KAAK;QAAE,CAAA;AAC/E;AAEA,6FAA6F;AAC7F,mEAAmE;AACnE,OAAO,SAASmB,aAAa5K,EAAgB,EAAEK,IAAY;IACzD,MAAM4H,MAAMjI,GAAGM,OAAO,CAAC,kFAAkF2G,GAAG,CAAC5G;IAC7G,OAAO4H,QAAQ9B;AACjB;AAEA,+FAA+F;AAC/F,uEAAuE;AACvE,OAAO,SAAS0E,aAAa7K,EAAgB,EAAEc,IAAY,EAAET,IAAY,EAAEyK,IAAgE;IACzI,0FAA0F;IAC1F,qFAAqF;IACrF,MAAMC,aAAa/K,GAAGM,OAAO,CAAC,+FAA+FuH,GAAG,CAACxH;IACjI,IAAI0K,WAAWpM,MAAM,KAAK,GAAG,OAAO,EAAE;IACtC,MAAMqM,OAAOtF,KAAK6B,GAAG,CAAC,GAAG7B,KAAKuF,IAAI,CAACF,WAAWpM,MAAM,GAAGX;IACvD,MAAMkN,SAASH,WAAW7L,MAAM,CAAC,CAACiM,GAAGzM,IAAMA,IAAIsM,SAAS,GAAGnF,GAAG,CAAC,CAACoC,MAAS,CAAA;YAAE3C,GAAG,IAAI6D,UAAUlB,IAAIiC,MAAM,CAACxF,MAAM,EAAEuD,IAAIiC,MAAM,CAAC5F,UAAU,EAAE2D,IAAIiC,MAAM,CAACC,UAAU;YAAG7C,OAAOW,IAAIX,KAAK;QAAC,CAAA;IAE/K,MAAM0C,OAAOhK,GAAGM,OAAO,CAAC,yEAAyEuH,GAAG;IACpG,MAAMoC,OAAO,IAAIpD;IACjB,KAAK,MAAMoB,OAAO+B,KAAM;QACtB,IAAI/B,IAAI5H,IAAI,KAAKA,QAAQyK,KAAKM,OAAO,CAACC,GAAG,CAACpD,IAAI5H,IAAI,KAAMyK,KAAKQ,OAAO,IAAI,CAACR,KAAKQ,OAAO,CAACD,GAAG,CAACpD,IAAI5H,IAAI,GAAI;QACtG,MAAMkL,QAAQ,IAAIpC,UAAUlB,IAAIiC,MAAM,CAACxF,MAAM,EAAEuD,IAAIiC,MAAM,CAAC5F,UAAU,EAAE2D,IAAIiC,MAAM,CAACC,UAAU;QAC3F,KAAK,MAAMN,KAAKqB,OAAQ;YACtB,IAAId,MAAM;YACV,MAAMoB,MAAM9F,KAAKqC,GAAG,CAAC8B,EAAEvE,CAAC,CAAC3G,MAAM,EAAE4M,MAAM5M,MAAM;YAC7C,IAAK,IAAI6G,IAAI,GAAGA,IAAIgG,KAAKhG,IAAK4E,OAAOP,EAAEvE,CAAC,CAACE,EAAE,GAAG+F,KAAK,CAAC/F,EAAE;YACtD,MAAMiE,QAAQW,MAAMP,EAAEvC,KAAK,GAAGW,IAAIX,KAAK;YACvC,MAAM+C,WAAWJ,KAAKhD,GAAG,CAACgB,IAAI5H,IAAI;YAClC,IAAIgK,aAAalE,aAAasD,QAAQY,UAAUJ,KAAK/C,GAAG,CAACe,IAAI5H,IAAI,EAAEoJ;QACrE;IACF;IAEA,OAAO;WAAIQ,KAAKnG,OAAO;KAAG,CACvB0G,IAAI,CAAC,CAACC,GAAGC,IAAMA,CAAC,CAAC,EAAE,GAAGD,CAAC,CAAC,EAAE,EAC1BhL,KAAK,CAAC,GAAGqL,KAAK9G,CAAC,EACf6B,GAAG,CAAC,CAAC,CAACmB,GAAGyC,MAAM,GAAM,CAAA;YAAEpJ,MAAM2G;YAAG2D,YAAYnB,SAASC;QAAO,CAAA;AACjE"}

package/dist/esm/features/links.js

@@ -73,13 +73,9 @@ function resolveAll(db, files) {
     const rows = db.prepare('SELECT src, target, dst FROM links').all();
     return resolveRows(db, rows, pathSet, byBase);
 }
-// Re-resolve only what this reconcile could have affected: (a) links whose src was just
-// re-stored (their rows are fresh, dst reset to NULL by store()), (b) links whose target's
-// basename matches a file that appeared or disappeared this reconcile (resolution and
-// lexicographic-tie ambiguity hinge on exactly that set), and (c) links whose dst was a
-// path that just vanished. All three are indexed lookups (src is the links PK's leading
-// column; target_base and dst each have their own index), so this reads only the rows this
-// reconcile could plausibly have changed, not the whole table.
+// Re-resolve only what this reconcile could have touched: re-stored sources, links whose
+// target basename matches an added/vanished file, and links whose dst just vanished. All
+// three are indexed lookups, so this never reads the whole table.
 function resolveIncremental(db, delta) {
     const pathSet = new Set(delta.files.map((f)=>f.relPath));
     const byBase = buildByBase(delta.files);
@@ -109,18 +105,15 @@ function resolveIncremental(db, delta) {
         ...candidates.values()
     ], pathSet, byBase);
 }
-// Resolved edges, for rank and for find's graph expansion.
+// Resolved edges, for rank and for search's graph expansion.
 export function linkEdges(db) {
     return db.prepare('SELECT src, dst FROM links WHERE dst IS NOT NULL').all().map((r)=>[
             r.src,
             r.dst
         ]);
 }
-// A removed path's outbound link rows are already gone by the time afterReconcile runs
-// (remove() below deletes them for vanished files AND for reparsed existing files), so
-// afterReconcile can't tell from the links table alone whether deleted rows carried edges.
-// remove() records that here, keyed on the reconcile's own delta object -- state dies with
-// the reconcile, including on ROLLBACK, and distinct reconciles never share it.
+// remove() has already deleted the rows by the time afterReconcile runs, so it records here
+// whether they carried edges. Keyed on the delta object, so the state dies with the reconcile.
 const removedWithLinks = new WeakMap();
 function recordRemoved(delta, path) {
     let set = removedWithLinks.get(delta);
@@ -180,10 +173,8 @@ export const links = {
         // reparsed-to-zero-links case (no surviving rows to re-resolve).
         const removeHadLinks = ((_ref = (_removedWithLinks_get = removedWithLinks.get(delta)) === null || _removedWithLinks_get === void 0 ? void 0 : _removedWithLinks_get.size) !== null && _ref !== void 0 ? _ref : 0) > 0;
         const churn = delta.reparsed.length + delta.vanished.length;
-        // Fallback rule: correctness over cleverness once too much of the tree moved in one
-        // reconcile -- a full pass is the one guaranteed to match resolveTarget's ambiguity
-        // rules exactly. A cold build reparses ~every file, so it clears this threshold on its
-        // own without a separate check.
+        // Past this share of the tree, a full pass is the only one guaranteed to match
+        // resolveTarget's ambiguity rules. A cold build clears the threshold on its own.
         const large = delta.files.length === 0 || churn > 0.2 * delta.files.length;
         const dstChanged = large ? resolveAll(db, delta.files) : resolveIncremental(db, delta);
         delta.linksChanged = dstChanged || removeHadLinks;

package/dist/esm/features/links.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/links.ts"],"sourcesContent":["import posix from 'node:path/posix';\nimport type { DatabaseSync } from 'node:sqlite';\nimport type { FileStat } from '../scan.ts';\nimport type { Feature, ReconcileDelta } from './types.ts';\n\n// links(src, target, target_base, dst): target as written, target_base its baseKey\n// (indexed, for the incremental resolve below), dst the resolved path or NULL (a\n// queryable dead link).\n\n// Wikilinks ([[target]], [[target#anchor|alias]], embeds) plus relative markdown links to .md files.\nfunction extract(_raw: string, body: string): string[] {\n  const targets = new Set<string>();\n  for (const m of body.matchAll(/\\[\\[([^\\]|#]+)(?:#[^\\]|]*)?(?:\\|[^\\]]*)?\\]\\]/g)) {\n    const target = m[1].trim();\n    if (target) targets.add(target);\n  }\n  for (const m of body.matchAll(/\\]\\(([^)]+\\.md)(?:#[^)]*)?\\)/g)) {\n    const target = m[1].trim();\n    if (target && !/^[a-z]+:\\/\\//i.test(target)) targets.add(target);\n  }\n  return [...targets];\n}\n\nfunction cleanTarget(target: string): string {\n  return target.replace(/\\\\/g, '/').replace(/^\\.\\//, '');\n}\n\nfunction baseKey(path: string): string {\n  return posix.basename(path).replace(/\\.md$/i, '').toLowerCase();\n}\n\n// Obsidian-style: exact relative path (with/without .md), path relative to the linking\n// note's directory, then basename match (lexicographically first on ties).\nfunction resolveTarget(src: string, target: string, pathSet: Set<string>, byBase: Map<string, string[]>): string | null {\n  const clean = cleanTarget(target);\n  const fromSrc = posix.normalize(posix.join(posix.dirname(src), clean));\n  for (const candidate of [clean, `${clean}.md`, fromSrc, `${fromSrc}.md`]) {\n    if (pathSet.has(candidate)) return candidate;\n  }\n  return byBase.get(baseKey(clean))?.[0] ?? null;\n}\n\nfunction buildByBase(files: FileStat[]): Map<string, string[]> {\n  const byBase = new Map<string, string[]>();\n  for (const path of files.map((f) => f.relPath).sort()) {\n    const key = baseKey(path);\n    const list = byBase.get(key);\n    if (list) list.push(path);\n    else byBase.set(key, [path]);\n  }\n  return byBase;\n}\n\ninterface LinkRow {\n  src: string;\n  target: string;\n  dst: string | null;\n}\n\n// Applies resolveTarget to `rows`, writing only rows whose dst actually changes.\nfunction resolveRows(db: DatabaseSync, rows: LinkRow[], pathSet: Set<string>, byBase: Map<string, string[]>): boolean {\n  const update = db.prepare('UPDATE links SET dst = ? WHERE src = ? AND target = ?');\n  let changed = false;\n  for (const row of rows) {\n    const dst = resolveTarget(row.src, row.target, pathSet, byBase);\n    if (dst !== row.dst) {\n      update.run(dst, row.src, row.target);\n      changed = true;\n    }\n  }\n  return changed;\n}\n\n// A new or deleted file can change any note's resolution, so re-resolve the whole table.\n// Fallback for cold builds and large deltas -- see afterReconcile's threshold.\nfunction resolveAll(db: DatabaseSync, files: FileStat[]): boolean {\n  const pathSet = new Set(files.map((f) => f.relPath));\n  const byBase = buildByBase(files);\n  const rows = db.prepare('SELECT src, target, dst FROM links').all() as unknown as LinkRow[];\n  return resolveRows(db, rows, pathSet, byBase);\n}\n\n// Re-resolve only what this reconcile could have affected: (a) links whose src was just\n// re-stored (their rows are fresh, dst reset to NULL by store()), (b) links whose target's\n// basename matches a file that appeared or disappeared this reconcile (resolution and\n// lexicographic-tie ambiguity hinge on exactly that set), and (c) links whose dst was a\n// path that just vanished. All three are indexed lookups (src is the links PK's leading\n// column; target_base and dst each have their own index), so this reads only the rows this\n// reconcile could plausibly have changed, not the whole table.\nfunction resolveIncremental(db: DatabaseSync, delta: ReconcileDelta): boolean {\n  const pathSet = new Set(delta.files.map((f) => f.relPath));\n  const byBase = buildByBase(delta.files);\n\n  const changedBasenames = new Set<string>();\n  for (const p of delta.added) changedBasenames.add(baseKey(p));\n  for (const p of delta.vanished) changedBasenames.add(baseKey(p));\n\n  // \\0 can't appear in a path or target, so the key never collides; a space can.\n  const candidates = new Map<string, LinkRow>();\n  const collect = (rows: LinkRow[]) => {\n    for (const row of rows) candidates.set(`${row.src}\\0${row.target}`, row);\n  };\n  // Chunked so a large-but-under-threshold delta never exceeds SQLite's bound-variable\n  // limit (SQLITE_MAX_VARIABLE_NUMBER, 32766).\n  const collectIn = (column: string, keys: string[]) => {\n    for (let i = 0; i < keys.length; i += 500) {\n      const chunk = keys.slice(i, i + 500);\n      const placeholders = chunk.map(() => '?').join(', ');\n      collect(db.prepare(`SELECT src, target, dst FROM links WHERE ${column} IN (${placeholders})`).all(...chunk) as unknown as LinkRow[]);\n    }\n  };\n\n  collectIn('src', delta.reparsed);\n  collectIn('target_base', [...changedBasenames]);\n  collectIn('dst', delta.vanished);\n\n  return resolveRows(db, [...candidates.values()], pathSet, byBase);\n}\n\n// Resolved edges, for rank and for find's graph expansion.\nexport function linkEdges(db: DatabaseSync): [string, string][] {\n  return (db.prepare('SELECT src, dst FROM links WHERE dst IS NOT NULL').all() as Array<{ src: string; dst: string }>).map((r) => [r.src, r.dst]);\n}\n\n// A removed path's outbound link rows are already gone by the time afterReconcile runs\n// (remove() below deletes them for vanished files AND for reparsed existing files), so\n// afterReconcile can't tell from the links table alone whether deleted rows carried edges.\n// remove() records that here, keyed on the reconcile's own delta object -- state dies with\n// the reconcile, including on ROLLBACK, and distinct reconciles never share it.\nconst removedWithLinks = new WeakMap<ReconcileDelta, Set<string>>();\n\nfunction recordRemoved(delta: ReconcileDelta, path: string): void {\n  let set = removedWithLinks.get(delta);\n  if (!set) {\n    set = new Set();\n    removedWithLinks.set(delta, set);\n  }\n  set.add(path);\n}\n\n// remove() runs per path; the vanished list is an array, so cache the Set per delta.\nconst vanishedSets = new WeakMap<ReconcileDelta, Set<string>>();\nfunction vanishedSet(delta: ReconcileDelta): Set<string> {\n  let set = vanishedSets.get(delta);\n  if (!set) {\n    set = new Set(delta.vanished);\n    vanishedSets.set(delta, set);\n  }\n  return set;\n}\n\nexport const links: Feature = {\n  name: 'links',\n  schema(db) {\n    db.exec('CREATE TABLE IF NOT EXISTS links (src TEXT, target TEXT, target_base TEXT, dst TEXT, PRIMARY KEY (src, target))');\n    db.exec('CREATE INDEX IF NOT EXISTS links_dst ON links(dst)');\n    db.exec('CREATE INDEX IF NOT EXISTS links_target_base ON links(target_base)');\n  },\n  extract,\n  // Reparsed docs are diffed in store() below instead of wiped here: deleting and\n  // re-inserting resets dst to NULL, which made every touch-only reparse read as an edge\n  // change and recompute PageRank -- measured as most of the remaining update cost.\n  remove(db, path, delta) {\n    if (!vanishedSet(delta).has(path)) return;\n    const result = db.prepare('DELETE FROM links WHERE src = ?').run(path);\n    if (Number(result.changes) > 0) recordRemoved(delta, path);\n  },\n  store(db, path, extracted, delta) {\n    const targets = extracted as string[];\n    // Stale rows (targets the new parse no longer contains) are real edge removals.\n    let stale: ReturnType<ReturnType<DatabaseSync['prepare']>['run']>;\n    if (targets.length === 0) {\n      stale = db.prepare('DELETE FROM links WHERE src = ?').run(path);\n    } else {\n      const placeholders = targets.map(() => '?').join(', ');\n      stale = db.prepare(`DELETE FROM links WHERE src = ? AND target NOT IN (${placeholders})`).run(path, ...targets);\n    }\n    if (Number(stale.changes) > 0) recordRemoved(delta, path);\n    // Upsert preserves dst on surviving rows, so an unchanged link resolves to the same\n    // value and reports no change; only genuinely new rows start at NULL.\n    const insert = db.prepare('INSERT INTO links (src, target, target_base, dst) VALUES (?, ?, ?, NULL) ON CONFLICT(src, target) DO UPDATE SET target_base = excluded.target_base');\n    for (const target of targets) insert.run(path, target, baseKey(cleanTarget(target)));\n  },\n  afterReconcile(db, delta) {\n    // Any deleted rows that carried edges mean the edge set shrank -- whether the file\n    // vanished or was reparsed down to fewer/no links. dstChanged alone misses the\n    // reparsed-to-zero-links case (no surviving rows to re-resolve).\n    const removeHadLinks = (removedWithLinks.get(delta)?.size ?? 0) > 0;\n\n    const churn = delta.reparsed.length + delta.vanished.length;\n    // Fallback rule: correctness over cleverness once too much of the tree moved in one\n    // reconcile -- a full pass is the one guaranteed to match resolveTarget's ambiguity\n    // rules exactly. A cold build reparses ~every file, so it clears this threshold on its\n    // own without a separate check.\n    const large = delta.files.length === 0 || churn > 0.2 * delta.files.length;\n\n    const dstChanged = large ? resolveAll(db, delta.files) : resolveIncremental(db, delta);\n    delta.linksChanged = dstChanged || removeHadLinks;\n  },\n};\n"],"names":["posix","extract","_raw","body","targets","Set","m","matchAll","target","trim","add","test","cleanTarget","replace","baseKey","path","basename","toLowerCase","resolveTarget","src","pathSet","byBase","clean","fromSrc","normalize","join","dirname","candidate","has","get","buildByBase","files","Map","map","f","relPath","sort","key","list","push","set","resolveRows","db","rows","update","prepare","changed","row","dst","run","resolveAll","all","resolveIncremental","delta","changedBasenames","p","added","vanished","candidates","collect","collectIn","column","keys","i","length","chunk","slice","placeholders","reparsed","values","linkEdges","r","removedWithLinks","WeakMap","recordRemoved","vanishedSets","vanishedSet","links","name","schema","exec","remove","result","Number","changes","store","extracted","stale","insert","afterReconcile","removeHadLinks","size","churn","large","dstChanged","linksChanged"],"mappings":"AAAA,OAAOA,WAAW,kBAAkB;AAKpC,mFAAmF;AACnF,iFAAiF;AACjF,wBAAwB;AAExB,qGAAqG;AACrG,SAASC,QAAQC,IAAY,EAAEC,IAAY;IACzC,MAAMC,UAAU,IAAIC;IACpB,KAAK,MAAMC,KAAKH,KAAKI,QAAQ,CAAC,iDAAkD;QAC9E,MAAMC,SAASF,CAAC,CAAC,EAAE,CAACG,IAAI;QACxB,IAAID,QAAQJ,QAAQM,GAAG,CAACF;IAC1B;IACA,KAAK,MAAMF,KAAKH,KAAKI,QAAQ,CAAC,iCAAkC;QAC9D,MAAMC,SAASF,CAAC,CAAC,EAAE,CAACG,IAAI;QACxB,IAAID,UAAU,CAAC,gBAAgBG,IAAI,CAACH,SAASJ,QAAQM,GAAG,CAACF;IAC3D;IACA,OAAO;WAAIJ;KAAQ;AACrB;AAEA,SAASQ,YAAYJ,MAAc;IACjC,OAAOA,OAAOK,OAAO,CAAC,OAAO,KAAKA,OAAO,CAAC,SAAS;AACrD;AAEA,SAASC,QAAQC,IAAY;IAC3B,OAAOf,MAAMgB,QAAQ,CAACD,MAAMF,OAAO,CAAC,UAAU,IAAII,WAAW;AAC/D;AAEA,uFAAuF;AACvF,2EAA2E;AAC3E,SAASC,cAAcC,GAAW,EAAEX,MAAc,EAAEY,OAAoB,EAAEC,MAA6B;;QAM9FA;IALP,MAAMC,QAAQV,YAAYJ;IAC1B,MAAMe,UAAUvB,MAAMwB,SAAS,CAACxB,MAAMyB,IAAI,CAACzB,MAAM0B,OAAO,CAACP,MAAMG;IAC/D,KAAK,MAAMK,aAAa;QAACL;QAAO,GAAGA,MAAM,GAAG,CAAC;QAAEC;QAAS,GAAGA,QAAQ,GAAG,CAAC;KAAC,CAAE;QACxE,IAAIH,QAAQQ,GAAG,CAACD,YAAY,OAAOA;IACrC;IACA,gBAAON,cAAAA,OAAOQ,GAAG,CAACf,QAAQQ,qBAAnBD,kCAAAA,WAA4B,CAAC,EAAE,uCAAI;AAC5C;AAEA,SAASS,YAAYC,KAAiB;IACpC,MAAMV,SAAS,IAAIW;IACnB,KAAK,MAAMjB,QAAQgB,MAAME,GAAG,CAAC,CAACC,IAAMA,EAAEC,OAAO,EAAEC,IAAI,GAAI;QACrD,MAAMC,MAAMvB,QAAQC;QACpB,MAAMuB,OAAOjB,OAAOQ,GAAG,CAACQ;QACxB,IAAIC,MAAMA,KAAKC,IAAI,CAACxB;aACfM,OAAOmB,GAAG,CAACH,KAAK;YAACtB;SAAK;IAC7B;IACA,OAAOM;AACT;AAQA,iFAAiF;AACjF,SAASoB,YAAYC,EAAgB,EAAEC,IAAe,EAAEvB,OAAoB,EAAEC,MAA6B;IACzG,MAAMuB,SAASF,GAAGG,OAAO,CAAC;IAC1B,IAAIC,UAAU;IACd,KAAK,MAAMC,OAAOJ,KAAM;QACtB,MAAMK,MAAM9B,cAAc6B,IAAI5B,GAAG,EAAE4B,IAAIvC,MAAM,EAAEY,SAASC;QACxD,IAAI2B,QAAQD,IAAIC,GAAG,EAAE;YACnBJ,OAAOK,GAAG,CAACD,KAAKD,IAAI5B,GAAG,EAAE4B,IAAIvC,MAAM;YACnCsC,UAAU;QACZ;IACF;IACA,OAAOA;AACT;AAEA,yFAAyF;AACzF,+EAA+E;AAC/E,SAASI,WAAWR,EAAgB,EAAEX,KAAiB;IACrD,MAAMX,UAAU,IAAIf,IAAI0B,MAAME,GAAG,CAAC,CAACC,IAAMA,EAAEC,OAAO;IAClD,MAAMd,SAASS,YAAYC;IAC3B,MAAMY,OAAOD,GAAGG,OAAO,CAAC,sCAAsCM,GAAG;IACjE,OAAOV,YAAYC,IAAIC,MAAMvB,SAASC;AACxC;AAEA,wFAAwF;AACxF,2FAA2F;AAC3F,sFAAsF;AACtF,wFAAwF;AACxF,wFAAwF;AACxF,2FAA2F;AAC3F,+DAA+D;AAC/D,SAAS+B,mBAAmBV,EAAgB,EAAEW,KAAqB;IACjE,MAAMjC,UAAU,IAAIf,IAAIgD,MAAMtB,KAAK,CAACE,GAAG,CAAC,CAACC,IAAMA,EAAEC,OAAO;IACxD,MAAMd,SAASS,YAAYuB,MAAMtB,KAAK;IAEtC,MAAMuB,mBAAmB,IAAIjD;IAC7B,KAAK,MAAMkD,KAAKF,MAAMG,KAAK,CAAEF,iBAAiB5C,GAAG,CAACI,QAAQyC;IAC1D,KAAK,MAAMA,KAAKF,MAAMI,QAAQ,CAAEH,iBAAiB5C,GAAG,CAACI,QAAQyC;IAE7D,+EAA+E;IAC/E,MAAMG,aAAa,IAAI1B;IACvB,MAAM2B,UAAU,CAAChB;QACf,KAAK,MAAMI,OAAOJ,KAAMe,WAAWlB,GAAG,CAAC,GAAGO,IAAI5B,GAAG,CAAC,EAAE,EAAE4B,IAAIvC,MAAM,EAAE,EAAEuC;IACtE;IACA,qFAAqF;IACrF,6CAA6C;IAC7C,MAAMa,YAAY,CAACC,QAAgBC;QACjC,IAAK,IAAIC,IAAI,GAAGA,IAAID,KAAKE,MAAM,EAAED,KAAK,IAAK;YACzC,MAAME,QAAQH,KAAKI,KAAK,CAACH,GAAGA,IAAI;YAChC,MAAMI,eAAeF,MAAMhC,GAAG,CAAC,IAAM,KAAKR,IAAI,CAAC;YAC/CkC,QAAQjB,GAAGG,OAAO,CAAC,CAAC,yCAAyC,EAAEgB,OAAO,KAAK,EAAEM,aAAa,CAAC,CAAC,EAAEhB,GAAG,IAAIc;QACvG;IACF;IAEAL,UAAU,OAAOP,MAAMe,QAAQ;IAC/BR,UAAU,eAAe;WAAIN;KAAiB;IAC9CM,UAAU,OAAOP,MAAMI,QAAQ;IAE/B,OAAOhB,YAAYC,IAAI;WAAIgB,WAAWW,MAAM;KAAG,EAAEjD,SAASC;AAC5D;AAEA,2DAA2D;AAC3D,OAAO,SAASiD,UAAU5B,EAAgB;IACxC,OAAO,AAACA,GAAGG,OAAO,CAAC,oDAAoDM,GAAG,GAA2ClB,GAAG,CAAC,CAACsC,IAAM;YAACA,EAAEpD,GAAG;YAAEoD,EAAEvB,GAAG;SAAC;AAChJ;AAEA,uFAAuF;AACvF,uFAAuF;AACvF,2FAA2F;AAC3F,2FAA2F;AAC3F,gFAAgF;AAChF,MAAMwB,mBAAmB,IAAIC;AAE7B,SAASC,cAAcrB,KAAqB,EAAEtC,IAAY;IACxD,IAAIyB,MAAMgC,iBAAiB3C,GAAG,CAACwB;IAC/B,IAAI,CAACb,KAAK;QACRA,MAAM,IAAInC;QACVmE,iBAAiBhC,GAAG,CAACa,OAAOb;IAC9B;IACAA,IAAI9B,GAAG,CAACK;AACV;AAEA,qFAAqF;AACrF,MAAM4D,eAAe,IAAIF;AACzB,SAASG,YAAYvB,KAAqB;IACxC,IAAIb,MAAMmC,aAAa9C,GAAG,CAACwB;IAC3B,IAAI,CAACb,KAAK;QACRA,MAAM,IAAInC,IAAIgD,MAAMI,QAAQ;QAC5BkB,aAAanC,GAAG,CAACa,OAAOb;IAC1B;IACA,OAAOA;AACT;AAEA,OAAO,MAAMqC,QAAiB;IAC5BC,MAAM;IACNC,QAAOrC,EAAE;QACPA,GAAGsC,IAAI,CAAC;QACRtC,GAAGsC,IAAI,CAAC;QACRtC,GAAGsC,IAAI,CAAC;IACV;IACA/E;IACA,gFAAgF;IAChF,uFAAuF;IACvF,kFAAkF;IAClFgF,QAAOvC,EAAE,EAAE3B,IAAI,EAAEsC,KAAK;QACpB,IAAI,CAACuB,YAAYvB,OAAOzB,GAAG,CAACb,OAAO;QACnC,MAAMmE,SAASxC,GAAGG,OAAO,CAAC,mCAAmCI,GAAG,CAAClC;QACjE,IAAIoE,OAAOD,OAAOE,OAAO,IAAI,GAAGV,cAAcrB,OAAOtC;IACvD;IACAsE,OAAM3C,EAAE,EAAE3B,IAAI,EAAEuE,SAAS,EAAEjC,KAAK;QAC9B,MAAMjD,UAAUkF;QAChB,gFAAgF;QAChF,IAAIC;QACJ,IAAInF,QAAQ4D,MAAM,KAAK,GAAG;YACxBuB,QAAQ7C,GAAGG,OAAO,CAAC,mCAAmCI,GAAG,CAAClC;QAC5D,OAAO;YACL,MAAMoD,eAAe/D,QAAQ6B,GAAG,CAAC,IAAM,KAAKR,IAAI,CAAC;YACjD8D,QAAQ7C,GAAGG,OAAO,CAAC,CAAC,mDAAmD,EAAEsB,aAAa,CAAC,CAAC,EAAElB,GAAG,CAAClC,SAASX;QACzG;QACA,IAAI+E,OAAOI,MAAMH,OAAO,IAAI,GAAGV,cAAcrB,OAAOtC;QACpD,oFAAoF;QACpF,sEAAsE;QACtE,MAAMyE,SAAS9C,GAAGG,OAAO,CAAC;QAC1B,KAAK,MAAMrC,UAAUJ,QAASoF,OAAOvC,GAAG,CAAClC,MAAMP,QAAQM,QAAQF,YAAYJ;IAC7E;IACAiF,gBAAe/C,EAAE,EAAEW,KAAK;;YAIEmB;QAHxB,mFAAmF;QACnF,+EAA+E;QAC/E,iEAAiE;QACjE,MAAMkB,iBAAiB,UAAClB,wBAAAA,iBAAiB3C,GAAG,CAACwB,oBAArBmB,4CAAAA,sBAA6BmB,IAAI,uCAAI,KAAK;QAElE,MAAMC,QAAQvC,MAAMe,QAAQ,CAACJ,MAAM,GAAGX,MAAMI,QAAQ,CAACO,MAAM;QAC3D,oFAAoF;QACpF,oFAAoF;QACpF,uFAAuF;QACvF,gCAAgC;QAChC,MAAM6B,QAAQxC,MAAMtB,KAAK,CAACiC,MAAM,KAAK,KAAK4B,QAAQ,MAAMvC,MAAMtB,KAAK,CAACiC,MAAM;QAE1E,MAAM8B,aAAaD,QAAQ3C,WAAWR,IAAIW,MAAMtB,KAAK,IAAIqB,mBAAmBV,IAAIW;QAChFA,MAAM0C,YAAY,GAAGD,cAAcJ;IACrC;AACF,EAAE"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/links.ts"],"sourcesContent":["import posix from 'node:path/posix';\nimport type { DatabaseSync } from 'node:sqlite';\nimport type { FileStat } from '../scan.ts';\nimport type { Feature, ReconcileDelta } from './types.ts';\n\n// links(src, target, target_base, dst): target as written, target_base its baseKey\n// (indexed, for the incremental resolve below), dst the resolved path or NULL (a\n// queryable dead link).\n\n// Wikilinks ([[target]], [[target#anchor|alias]], embeds) plus relative markdown links to .md files.\nfunction extract(_raw: string, body: string): string[] {\n  const targets = new Set<string>();\n  for (const m of body.matchAll(/\\[\\[([^\\]|#]+)(?:#[^\\]|]*)?(?:\\|[^\\]]*)?\\]\\]/g)) {\n    const target = m[1].trim();\n    if (target) targets.add(target);\n  }\n  for (const m of body.matchAll(/\\]\\(([^)]+\\.md)(?:#[^)]*)?\\)/g)) {\n    const target = m[1].trim();\n    if (target && !/^[a-z]+:\\/\\//i.test(target)) targets.add(target);\n  }\n  return [...targets];\n}\n\nfunction cleanTarget(target: string): string {\n  return target.replace(/\\\\/g, '/').replace(/^\\.\\//, '');\n}\n\nfunction baseKey(path: string): string {\n  return posix.basename(path).replace(/\\.md$/i, '').toLowerCase();\n}\n\n// Obsidian-style: exact relative path (with/without .md), path relative to the linking\n// note's directory, then basename match (lexicographically first on ties).\nfunction resolveTarget(src: string, target: string, pathSet: Set<string>, byBase: Map<string, string[]>): string | null {\n  const clean = cleanTarget(target);\n  const fromSrc = posix.normalize(posix.join(posix.dirname(src), clean));\n  for (const candidate of [clean, `${clean}.md`, fromSrc, `${fromSrc}.md`]) {\n    if (pathSet.has(candidate)) return candidate;\n  }\n  return byBase.get(baseKey(clean))?.[0] ?? null;\n}\n\nfunction buildByBase(files: FileStat[]): Map<string, string[]> {\n  const byBase = new Map<string, string[]>();\n  for (const path of files.map((f) => f.relPath).sort()) {\n    const key = baseKey(path);\n    const list = byBase.get(key);\n    if (list) list.push(path);\n    else byBase.set(key, [path]);\n  }\n  return byBase;\n}\n\ninterface LinkRow {\n  src: string;\n  target: string;\n  dst: string | null;\n}\n\n// Applies resolveTarget to `rows`, writing only rows whose dst actually changes.\nfunction resolveRows(db: DatabaseSync, rows: LinkRow[], pathSet: Set<string>, byBase: Map<string, string[]>): boolean {\n  const update = db.prepare('UPDATE links SET dst = ? WHERE src = ? AND target = ?');\n  let changed = false;\n  for (const row of rows) {\n    const dst = resolveTarget(row.src, row.target, pathSet, byBase);\n    if (dst !== row.dst) {\n      update.run(dst, row.src, row.target);\n      changed = true;\n    }\n  }\n  return changed;\n}\n\n// A new or deleted file can change any note's resolution, so re-resolve the whole table.\n// Fallback for cold builds and large deltas -- see afterReconcile's threshold.\nfunction resolveAll(db: DatabaseSync, files: FileStat[]): boolean {\n  const pathSet = new Set(files.map((f) => f.relPath));\n  const byBase = buildByBase(files);\n  const rows = db.prepare('SELECT src, target, dst FROM links').all() as unknown as LinkRow[];\n  return resolveRows(db, rows, pathSet, byBase);\n}\n\n// Re-resolve only what this reconcile could have touched: re-stored sources, links whose\n// target basename matches an added/vanished file, and links whose dst just vanished. All\n// three are indexed lookups, so this never reads the whole table.\nfunction resolveIncremental(db: DatabaseSync, delta: ReconcileDelta): boolean {\n  const pathSet = new Set(delta.files.map((f) => f.relPath));\n  const byBase = buildByBase(delta.files);\n\n  const changedBasenames = new Set<string>();\n  for (const p of delta.added) changedBasenames.add(baseKey(p));\n  for (const p of delta.vanished) changedBasenames.add(baseKey(p));\n\n  // \\0 can't appear in a path or target, so the key never collides; a space can.\n  const candidates = new Map<string, LinkRow>();\n  const collect = (rows: LinkRow[]) => {\n    for (const row of rows) candidates.set(`${row.src}\\0${row.target}`, row);\n  };\n  // Chunked so a large-but-under-threshold delta never exceeds SQLite's bound-variable\n  // limit (SQLITE_MAX_VARIABLE_NUMBER, 32766).\n  const collectIn = (column: string, keys: string[]) => {\n    for (let i = 0; i < keys.length; i += 500) {\n      const chunk = keys.slice(i, i + 500);\n      const placeholders = chunk.map(() => '?').join(', ');\n      collect(db.prepare(`SELECT src, target, dst FROM links WHERE ${column} IN (${placeholders})`).all(...chunk) as unknown as LinkRow[]);\n    }\n  };\n\n  collectIn('src', delta.reparsed);\n  collectIn('target_base', [...changedBasenames]);\n  collectIn('dst', delta.vanished);\n\n  return resolveRows(db, [...candidates.values()], pathSet, byBase);\n}\n\n// Resolved edges, for rank and for search's graph expansion.\nexport function linkEdges(db: DatabaseSync): [string, string][] {\n  return (db.prepare('SELECT src, dst FROM links WHERE dst IS NOT NULL').all() as Array<{ src: string; dst: string }>).map((r) => [r.src, r.dst]);\n}\n\n// remove() has already deleted the rows by the time afterReconcile runs, so it records here\n// whether they carried edges. Keyed on the delta object, so the state dies with the reconcile.\nconst removedWithLinks = new WeakMap<ReconcileDelta, Set<string>>();\n\nfunction recordRemoved(delta: ReconcileDelta, path: string): void {\n  let set = removedWithLinks.get(delta);\n  if (!set) {\n    set = new Set();\n    removedWithLinks.set(delta, set);\n  }\n  set.add(path);\n}\n\n// remove() runs per path; the vanished list is an array, so cache the Set per delta.\nconst vanishedSets = new WeakMap<ReconcileDelta, Set<string>>();\nfunction vanishedSet(delta: ReconcileDelta): Set<string> {\n  let set = vanishedSets.get(delta);\n  if (!set) {\n    set = new Set(delta.vanished);\n    vanishedSets.set(delta, set);\n  }\n  return set;\n}\n\nexport const links: Feature = {\n  name: 'links',\n  schema(db) {\n    db.exec('CREATE TABLE IF NOT EXISTS links (src TEXT, target TEXT, target_base TEXT, dst TEXT, PRIMARY KEY (src, target))');\n    db.exec('CREATE INDEX IF NOT EXISTS links_dst ON links(dst)');\n    db.exec('CREATE INDEX IF NOT EXISTS links_target_base ON links(target_base)');\n  },\n  extract,\n  // Reparsed docs are diffed in store() below instead of wiped here: deleting and\n  // re-inserting resets dst to NULL, which made every touch-only reparse read as an edge\n  // change and recompute PageRank -- measured as most of the remaining update cost.\n  remove(db, path, delta) {\n    if (!vanishedSet(delta).has(path)) return;\n    const result = db.prepare('DELETE FROM links WHERE src = ?').run(path);\n    if (Number(result.changes) > 0) recordRemoved(delta, path);\n  },\n  store(db, path, extracted, delta) {\n    const targets = extracted as string[];\n    // Stale rows (targets the new parse no longer contains) are real edge removals.\n    let stale: ReturnType<ReturnType<DatabaseSync['prepare']>['run']>;\n    if (targets.length === 0) {\n      stale = db.prepare('DELETE FROM links WHERE src = ?').run(path);\n    } else {\n      const placeholders = targets.map(() => '?').join(', ');\n      stale = db.prepare(`DELETE FROM links WHERE src = ? AND target NOT IN (${placeholders})`).run(path, ...targets);\n    }\n    if (Number(stale.changes) > 0) recordRemoved(delta, path);\n    // Upsert preserves dst on surviving rows, so an unchanged link resolves to the same\n    // value and reports no change; only genuinely new rows start at NULL.\n    const insert = db.prepare('INSERT INTO links (src, target, target_base, dst) VALUES (?, ?, ?, NULL) ON CONFLICT(src, target) DO UPDATE SET target_base = excluded.target_base');\n    for (const target of targets) insert.run(path, target, baseKey(cleanTarget(target)));\n  },\n  afterReconcile(db, delta) {\n    // Any deleted rows that carried edges mean the edge set shrank -- whether the file\n    // vanished or was reparsed down to fewer/no links. dstChanged alone misses the\n    // reparsed-to-zero-links case (no surviving rows to re-resolve).\n    const removeHadLinks = (removedWithLinks.get(delta)?.size ?? 0) > 0;\n\n    const churn = delta.reparsed.length + delta.vanished.length;\n    // Past this share of the tree, a full pass is the only one guaranteed to match\n    // resolveTarget's ambiguity rules. A cold build clears the threshold on its own.\n    const large = delta.files.length === 0 || churn > 0.2 * delta.files.length;\n\n    const dstChanged = large ? resolveAll(db, delta.files) : resolveIncremental(db, delta);\n    delta.linksChanged = dstChanged || removeHadLinks;\n  },\n};\n"],"names":["posix","extract","_raw","body","targets","Set","m","matchAll","target","trim","add","test","cleanTarget","replace","baseKey","path","basename","toLowerCase","resolveTarget","src","pathSet","byBase","clean","fromSrc","normalize","join","dirname","candidate","has","get","buildByBase","files","Map","map","f","relPath","sort","key","list","push","set","resolveRows","db","rows","update","prepare","changed","row","dst","run","resolveAll","all","resolveIncremental","delta","changedBasenames","p","added","vanished","candidates","collect","collectIn","column","keys","i","length","chunk","slice","placeholders","reparsed","values","linkEdges","r","removedWithLinks","WeakMap","recordRemoved","vanishedSets","vanishedSet","links","name","schema","exec","remove","result","Number","changes","store","extracted","stale","insert","afterReconcile","removeHadLinks","size","churn","large","dstChanged","linksChanged"],"mappings":"AAAA,OAAOA,WAAW,kBAAkB;AAKpC,mFAAmF;AACnF,iFAAiF;AACjF,wBAAwB;AAExB,qGAAqG;AACrG,SAASC,QAAQC,IAAY,EAAEC,IAAY;IACzC,MAAMC,UAAU,IAAIC;IACpB,KAAK,MAAMC,KAAKH,KAAKI,QAAQ,CAAC,iDAAkD;QAC9E,MAAMC,SAASF,CAAC,CAAC,EAAE,CAACG,IAAI;QACxB,IAAID,QAAQJ,QAAQM,GAAG,CAACF;IAC1B;IACA,KAAK,MAAMF,KAAKH,KAAKI,QAAQ,CAAC,iCAAkC;QAC9D,MAAMC,SAASF,CAAC,CAAC,EAAE,CAACG,IAAI;QACxB,IAAID,UAAU,CAAC,gBAAgBG,IAAI,CAACH,SAASJ,QAAQM,GAAG,CAACF;IAC3D;IACA,OAAO;WAAIJ;KAAQ;AACrB;AAEA,SAASQ,YAAYJ,MAAc;IACjC,OAAOA,OAAOK,OAAO,CAAC,OAAO,KAAKA,OAAO,CAAC,SAAS;AACrD;AAEA,SAASC,QAAQC,IAAY;IAC3B,OAAOf,MAAMgB,QAAQ,CAACD,MAAMF,OAAO,CAAC,UAAU,IAAII,WAAW;AAC/D;AAEA,uFAAuF;AACvF,2EAA2E;AAC3E,SAASC,cAAcC,GAAW,EAAEX,MAAc,EAAEY,OAAoB,EAAEC,MAA6B;;QAM9FA;IALP,MAAMC,QAAQV,YAAYJ;IAC1B,MAAMe,UAAUvB,MAAMwB,SAAS,CAACxB,MAAMyB,IAAI,CAACzB,MAAM0B,OAAO,CAACP,MAAMG;IAC/D,KAAK,MAAMK,aAAa;QAACL;QAAO,GAAGA,MAAM,GAAG,CAAC;QAAEC;QAAS,GAAGA,QAAQ,GAAG,CAAC;KAAC,CAAE;QACxE,IAAIH,QAAQQ,GAAG,CAACD,YAAY,OAAOA;IACrC;IACA,gBAAON,cAAAA,OAAOQ,GAAG,CAACf,QAAQQ,qBAAnBD,kCAAAA,WAA4B,CAAC,EAAE,uCAAI;AAC5C;AAEA,SAASS,YAAYC,KAAiB;IACpC,MAAMV,SAAS,IAAIW;IACnB,KAAK,MAAMjB,QAAQgB,MAAME,GAAG,CAAC,CAACC,IAAMA,EAAEC,OAAO,EAAEC,IAAI,GAAI;QACrD,MAAMC,MAAMvB,QAAQC;QACpB,MAAMuB,OAAOjB,OAAOQ,GAAG,CAACQ;QACxB,IAAIC,MAAMA,KAAKC,IAAI,CAACxB;aACfM,OAAOmB,GAAG,CAACH,KAAK;YAACtB;SAAK;IAC7B;IACA,OAAOM;AACT;AAQA,iFAAiF;AACjF,SAASoB,YAAYC,EAAgB,EAAEC,IAAe,EAAEvB,OAAoB,EAAEC,MAA6B;IACzG,MAAMuB,SAASF,GAAGG,OAAO,CAAC;IAC1B,IAAIC,UAAU;IACd,KAAK,MAAMC,OAAOJ,KAAM;QACtB,MAAMK,MAAM9B,cAAc6B,IAAI5B,GAAG,EAAE4B,IAAIvC,MAAM,EAAEY,SAASC;QACxD,IAAI2B,QAAQD,IAAIC,GAAG,EAAE;YACnBJ,OAAOK,GAAG,CAACD,KAAKD,IAAI5B,GAAG,EAAE4B,IAAIvC,MAAM;YACnCsC,UAAU;QACZ;IACF;IACA,OAAOA;AACT;AAEA,yFAAyF;AACzF,+EAA+E;AAC/E,SAASI,WAAWR,EAAgB,EAAEX,KAAiB;IACrD,MAAMX,UAAU,IAAIf,IAAI0B,MAAME,GAAG,CAAC,CAACC,IAAMA,EAAEC,OAAO;IAClD,MAAMd,SAASS,YAAYC;IAC3B,MAAMY,OAAOD,GAAGG,OAAO,CAAC,sCAAsCM,GAAG;IACjE,OAAOV,YAAYC,IAAIC,MAAMvB,SAASC;AACxC;AAEA,yFAAyF;AACzF,yFAAyF;AACzF,kEAAkE;AAClE,SAAS+B,mBAAmBV,EAAgB,EAAEW,KAAqB;IACjE,MAAMjC,UAAU,IAAIf,IAAIgD,MAAMtB,KAAK,CAACE,GAAG,CAAC,CAACC,IAAMA,EAAEC,OAAO;IACxD,MAAMd,SAASS,YAAYuB,MAAMtB,KAAK;IAEtC,MAAMuB,mBAAmB,IAAIjD;IAC7B,KAAK,MAAMkD,KAAKF,MAAMG,KAAK,CAAEF,iBAAiB5C,GAAG,CAACI,QAAQyC;IAC1D,KAAK,MAAMA,KAAKF,MAAMI,QAAQ,CAAEH,iBAAiB5C,GAAG,CAACI,QAAQyC;IAE7D,+EAA+E;IAC/E,MAAMG,aAAa,IAAI1B;IACvB,MAAM2B,UAAU,CAAChB;QACf,KAAK,MAAMI,OAAOJ,KAAMe,WAAWlB,GAAG,CAAC,GAAGO,IAAI5B,GAAG,CAAC,EAAE,EAAE4B,IAAIvC,MAAM,EAAE,EAAEuC;IACtE;IACA,qFAAqF;IACrF,6CAA6C;IAC7C,MAAMa,YAAY,CAACC,QAAgBC;QACjC,IAAK,IAAIC,IAAI,GAAGA,IAAID,KAAKE,MAAM,EAAED,KAAK,IAAK;YACzC,MAAME,QAAQH,KAAKI,KAAK,CAACH,GAAGA,IAAI;YAChC,MAAMI,eAAeF,MAAMhC,GAAG,CAAC,IAAM,KAAKR,IAAI,CAAC;YAC/CkC,QAAQjB,GAAGG,OAAO,CAAC,CAAC,yCAAyC,EAAEgB,OAAO,KAAK,EAAEM,aAAa,CAAC,CAAC,EAAEhB,GAAG,IAAIc;QACvG;IACF;IAEAL,UAAU,OAAOP,MAAMe,QAAQ;IAC/BR,UAAU,eAAe;WAAIN;KAAiB;IAC9CM,UAAU,OAAOP,MAAMI,QAAQ;IAE/B,OAAOhB,YAAYC,IAAI;WAAIgB,WAAWW,MAAM;KAAG,EAAEjD,SAASC;AAC5D;AAEA,6DAA6D;AAC7D,OAAO,SAASiD,UAAU5B,EAAgB;IACxC,OAAO,AAACA,GAAGG,OAAO,CAAC,oDAAoDM,GAAG,GAA2ClB,GAAG,CAAC,CAACsC,IAAM;YAACA,EAAEpD,GAAG;YAAEoD,EAAEvB,GAAG;SAAC;AAChJ;AAEA,4FAA4F;AAC5F,+FAA+F;AAC/F,MAAMwB,mBAAmB,IAAIC;AAE7B,SAASC,cAAcrB,KAAqB,EAAEtC,IAAY;IACxD,IAAIyB,MAAMgC,iBAAiB3C,GAAG,CAACwB;IAC/B,IAAI,CAACb,KAAK;QACRA,MAAM,IAAInC;QACVmE,iBAAiBhC,GAAG,CAACa,OAAOb;IAC9B;IACAA,IAAI9B,GAAG,CAACK;AACV;AAEA,qFAAqF;AACrF,MAAM4D,eAAe,IAAIF;AACzB,SAASG,YAAYvB,KAAqB;IACxC,IAAIb,MAAMmC,aAAa9C,GAAG,CAACwB;IAC3B,IAAI,CAACb,KAAK;QACRA,MAAM,IAAInC,IAAIgD,MAAMI,QAAQ;QAC5BkB,aAAanC,GAAG,CAACa,OAAOb;IAC1B;IACA,OAAOA;AACT;AAEA,OAAO,MAAMqC,QAAiB;IAC5BC,MAAM;IACNC,QAAOrC,EAAE;QACPA,GAAGsC,IAAI,CAAC;QACRtC,GAAGsC,IAAI,CAAC;QACRtC,GAAGsC,IAAI,CAAC;IACV;IACA/E;IACA,gFAAgF;IAChF,uFAAuF;IACvF,kFAAkF;IAClFgF,QAAOvC,EAAE,EAAE3B,IAAI,EAAEsC,KAAK;QACpB,IAAI,CAACuB,YAAYvB,OAAOzB,GAAG,CAACb,OAAO;QACnC,MAAMmE,SAASxC,GAAGG,OAAO,CAAC,mCAAmCI,GAAG,CAAClC;QACjE,IAAIoE,OAAOD,OAAOE,OAAO,IAAI,GAAGV,cAAcrB,OAAOtC;IACvD;IACAsE,OAAM3C,EAAE,EAAE3B,IAAI,EAAEuE,SAAS,EAAEjC,KAAK;QAC9B,MAAMjD,UAAUkF;QAChB,gFAAgF;QAChF,IAAIC;QACJ,IAAInF,QAAQ4D,MAAM,KAAK,GAAG;YACxBuB,QAAQ7C,GAAGG,OAAO,CAAC,mCAAmCI,GAAG,CAAClC;QAC5D,OAAO;YACL,MAAMoD,eAAe/D,QAAQ6B,GAAG,CAAC,IAAM,KAAKR,IAAI,CAAC;YACjD8D,QAAQ7C,GAAGG,OAAO,CAAC,CAAC,mDAAmD,EAAEsB,aAAa,CAAC,CAAC,EAAElB,GAAG,CAAClC,SAASX;QACzG;QACA,IAAI+E,OAAOI,MAAMH,OAAO,IAAI,GAAGV,cAAcrB,OAAOtC;QACpD,oFAAoF;QACpF,sEAAsE;QACtE,MAAMyE,SAAS9C,GAAGG,OAAO,CAAC;QAC1B,KAAK,MAAMrC,UAAUJ,QAASoF,OAAOvC,GAAG,CAAClC,MAAMP,QAAQM,QAAQF,YAAYJ;IAC7E;IACAiF,gBAAe/C,EAAE,EAAEW,KAAK;;YAIEmB;QAHxB,mFAAmF;QACnF,+EAA+E;QAC/E,iEAAiE;QACjE,MAAMkB,iBAAiB,UAAClB,wBAAAA,iBAAiB3C,GAAG,CAACwB,oBAArBmB,4CAAAA,sBAA6BmB,IAAI,uCAAI,KAAK;QAElE,MAAMC,QAAQvC,MAAMe,QAAQ,CAACJ,MAAM,GAAGX,MAAMI,QAAQ,CAACO,MAAM;QAC3D,+EAA+E;QAC/E,iFAAiF;QACjF,MAAM6B,QAAQxC,MAAMtB,KAAK,CAACiC,MAAM,KAAK,KAAK4B,QAAQ,MAAMvC,MAAMtB,KAAK,CAACiC,MAAM;QAE1E,MAAM8B,aAAaD,QAAQ3C,WAAWR,IAAIW,MAAMtB,KAAK,IAAIqB,mBAAmBV,IAAIW;QAChFA,MAAM0C,YAAY,GAAGD,cAAcJ;IACrC;AACF,EAAE"}

package/dist/esm/features/rank.js

@@ -9,12 +9,8 @@ export const rank = {
         if (!columns.some((c)=>c.name === '_rank')) db.exec('ALTER TABLE frontmatter ADD COLUMN "_rank" REAL');
     },
     afterReconcile (db, delta) {
-        // links.afterReconcile runs first (registry order) and sets delta.linksChanged;
-        // `undefined` (links off, or no opinion) recomputes, same as always -- only an
-        // explicit `false` allows a skip. Rank also depends on the node set (every note is a
-        // PageRank node, edges or not), so any add or vanish recomputes even with edges
-        // untouched -- otherwise a new linkless note keeps _rank NULL and every other rank
-        // stays normalized over the old count.
+        // Only an explicit linksChanged:false skips. Rank also depends on the node set, so any
+        // add or vanish recomputes -- otherwise a new linkless note keeps _rank NULL.
         if (delta.linksChanged === false && delta.added.length === 0 && delta.vanished.length === 0) return;
         const nodes = db.prepare('SELECT "path" FROM frontmatter').all().map((r)=>r.path);
         const ranks = pagerank(nodes, linkEdges(db));

package/dist/esm/features/rank.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/rank.ts"],"sourcesContent":["import { pagerank } from '../graph.ts';\nimport { linkEdges } from './links.ts';\nimport type { Feature } from './types.ts';\n\n// frontmatter._rank: PageRank over resolved links -- a static \"how load-bearing is this\n// note\" prior. Depends on links (enforced in config.featureEnabled).\n\nexport const rank: Feature = {\n  name: 'rank',\n  schema(db) {\n    const columns = db.prepare('PRAGMA table_info(frontmatter)').all() as Array<{ name: string }>;\n    if (!columns.some((c) => c.name === '_rank')) db.exec('ALTER TABLE frontmatter ADD COLUMN \"_rank\" REAL');\n  },\n  afterReconcile(db, delta) {\n    // links.afterReconcile runs first (registry order) and sets delta.linksChanged;\n    // `undefined` (links off, or no opinion) recomputes, same as always -- only an\n    // explicit `false` allows a skip. Rank also depends on the node set (every note is a\n    // PageRank node, edges or not), so any add or vanish recomputes even with edges\n    // untouched -- otherwise a new linkless note keeps _rank NULL and every other rank\n    // stays normalized over the old count.\n    if (delta.linksChanged === false && delta.added.length === 0 && delta.vanished.length === 0) return;\n    const nodes = (db.prepare('SELECT \"path\" FROM frontmatter').all() as Array<{ path: string }>).map((r) => r.path);\n    const ranks = pagerank(nodes, linkEdges(db));\n    const update = db.prepare('UPDATE frontmatter SET \"_rank\" = ? WHERE \"path\" = ?');\n    for (const [path, value] of ranks) update.run(value, path);\n  },\n};\n"],"names":["pagerank","linkEdges","rank","name","schema","db","columns","prepare","all","some","c","exec","afterReconcile","delta","linksChanged","added","length","vanished","nodes","map","r","path","ranks","update","value","run"],"mappings":"AAAA,SAASA,QAAQ,QAAQ,cAAc;AACvC,SAASC,SAAS,QAAQ,aAAa;AAGvC,wFAAwF;AACxF,qEAAqE;AAErE,OAAO,MAAMC,OAAgB;IAC3BC,MAAM;IACNC,QAAOC,EAAE;QACP,MAAMC,UAAUD,GAAGE,OAAO,CAAC,kCAAkCC,GAAG;QAChE,IAAI,CAACF,QAAQG,IAAI,CAAC,CAACC,IAAMA,EAAEP,IAAI,KAAK,UAAUE,GAAGM,IAAI,CAAC;IACxD;IACAC,gBAAeP,EAAE,EAAEQ,KAAK;QACtB,gFAAgF;QAChF,+EAA+E;QAC/E,qFAAqF;QACrF,gFAAgF;QAChF,mFAAmF;QACnF,uCAAuC;QACvC,IAAIA,MAAMC,YAAY,KAAK,SAASD,MAAME,KAAK,CAACC,MAAM,KAAK,KAAKH,MAAMI,QAAQ,CAACD,MAAM,KAAK,GAAG;QAC7F,MAAME,QAAQ,AAACb,GAAGE,OAAO,CAAC,kCAAkCC,GAAG,GAA+BW,GAAG,CAAC,CAACC,IAAMA,EAAEC,IAAI;QAC/G,MAAMC,QAAQtB,SAASkB,OAAOjB,UAAUI;QACxC,MAAMkB,SAASlB,GAAGE,OAAO,CAAC;QAC1B,KAAK,MAAM,CAACc,MAAMG,MAAM,IAAIF,MAAOC,OAAOE,GAAG,CAACD,OAAOH;IACvD;AACF,EAAE"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/rank.ts"],"sourcesContent":["import { pagerank } from '../graph.ts';\nimport { linkEdges } from './links.ts';\nimport type { Feature } from './types.ts';\n\n// frontmatter._rank: PageRank over resolved links -- a static \"how load-bearing is this\n// note\" prior. Depends on links (enforced in config.featureEnabled).\n\nexport const rank: Feature = {\n  name: 'rank',\n  schema(db) {\n    const columns = db.prepare('PRAGMA table_info(frontmatter)').all() as Array<{ name: string }>;\n    if (!columns.some((c) => c.name === '_rank')) db.exec('ALTER TABLE frontmatter ADD COLUMN \"_rank\" REAL');\n  },\n  afterReconcile(db, delta) {\n    // Only an explicit linksChanged:false skips. Rank also depends on the node set, so any\n    // add or vanish recomputes -- otherwise a new linkless note keeps _rank NULL.\n    if (delta.linksChanged === false && delta.added.length === 0 && delta.vanished.length === 0) return;\n    const nodes = (db.prepare('SELECT \"path\" FROM frontmatter').all() as Array<{ path: string }>).map((r) => r.path);\n    const ranks = pagerank(nodes, linkEdges(db));\n    const update = db.prepare('UPDATE frontmatter SET \"_rank\" = ? WHERE \"path\" = ?');\n    for (const [path, value] of ranks) update.run(value, path);\n  },\n};\n"],"names":["pagerank","linkEdges","rank","name","schema","db","columns","prepare","all","some","c","exec","afterReconcile","delta","linksChanged","added","length","vanished","nodes","map","r","path","ranks","update","value","run"],"mappings":"AAAA,SAASA,QAAQ,QAAQ,cAAc;AACvC,SAASC,SAAS,QAAQ,aAAa;AAGvC,wFAAwF;AACxF,qEAAqE;AAErE,OAAO,MAAMC,OAAgB;IAC3BC,MAAM;IACNC,QAAOC,EAAE;QACP,MAAMC,UAAUD,GAAGE,OAAO,CAAC,kCAAkCC,GAAG;QAChE,IAAI,CAACF,QAAQG,IAAI,CAAC,CAACC,IAAMA,EAAEP,IAAI,KAAK,UAAUE,GAAGM,IAAI,CAAC;IACxD;IACAC,gBAAeP,EAAE,EAAEQ,KAAK;QACtB,uFAAuF;QACvF,8EAA8E;QAC9E,IAAIA,MAAMC,YAAY,KAAK,SAASD,MAAME,KAAK,CAACC,MAAM,KAAK,KAAKH,MAAMI,QAAQ,CAACD,MAAM,KAAK,GAAG;QAC7F,MAAME,QAAQ,AAACb,GAAGE,OAAO,CAAC,kCAAkCC,GAAG,GAA+BW,GAAG,CAAC,CAACC,IAAMA,EAAEC,IAAI;QAC/G,MAAMC,QAAQtB,SAASkB,OAAOjB,UAAUI;QACxC,MAAMkB,SAASlB,GAAGE,OAAO,CAAC;QAC1B,KAAK,MAAM,CAACc,MAAMG,MAAM,IAAIF,MAAOC,OAAOE,GAAG,CAACD,OAAOH;IACvD;AACF,EAAE"}

package/dist/esm/features/types.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/types.ts"],"sourcesContent":["import type { DatabaseSync } from 'node:sqlite';\nimport type { Config, FeatureName } from '../config.ts';\nimport type { FileStat } from '../scan.ts';\n\n// What changed this reconcile, so afterReconcile hooks can act on the delta instead of\n// re-reading the whole tree. `reparsed`/`added`/`vanished` are relPaths; `added` is the\n// subset of `reparsed` that are brand-new files (not just modified). One delta object per\n// reconcile: hooks communicate through it (links sets `linksChanged` for rank), and any\n// state keyed on it dies with the reconcile -- including on ROLLBACK.\nexport interface ReconcileDelta {\n  files: FileStat[];\n  reparsed: string[];\n  added: string[];\n  vanished: string[];\n  // Set by the links feature: the resolved edge set may differ from before this reconcile.\n  linksChanged?: boolean;\n}\n\n// A feature owns its schema, per-file extraction/storage, and any whole-tree pass.\n// Adding a feature = one module here + one entry in the registry (index.ts).\nexport interface Feature {\n  name: FeatureName;\n  schema(db: DatabaseSync): void;\n  // Pure, per file: raw is the full file, body the prose after frontmatter,\n  // search the normalized title/summary strings.\n  extract?(raw: string, body: string, search?: { title: string; summary: string }): unknown;\n  remove?(db: DatabaseSync, path: string, delta: ReconcileDelta): void;\n  store?(db: DatabaseSync, path: string, extracted: unknown, delta: ReconcileDelta): void;\n  // After all rows are current, inside the reconcile transaction (resolution, rank).\n  // Hooks run in registry order and communicate through fields on `delta`.\n  afterReconcile?(db: DatabaseSync, delta: ReconcileDelta): void;\n  // Per-file opt-out (currently only embed, via FileStat.embed): absent means the feature\n  // applies to every file it's otherwise on for. Checked per file before extract/store so a\n  // doc covered only by semantic-false presets gets no rows for it at all.\n  enabledForFile?(cfg: Config, file: FileStat): boolean;\n}\n"],"names":[],"mappings":"AAkBA,mFAAmF;AACnF,6EAA6E;AAC7E,WAeC"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/features/types.ts"],"sourcesContent":["import type { DatabaseSync } from 'node:sqlite';\nimport type { Config, FeatureName } from '../config.ts';\nimport type { FileStat } from '../scan.ts';\n\n// What changed this reconcile, so hooks act on the delta rather than re-reading the tree.\n// `added` is the brand-new subset of `reparsed`. Hooks communicate through this one object.\nexport interface ReconcileDelta {\n  files: FileStat[];\n  reparsed: string[];\n  added: string[];\n  vanished: string[];\n  // Set by the links feature: the resolved edge set may differ from before this reconcile.\n  linksChanged?: boolean;\n}\n\n// A feature owns its schema, per-file extraction/storage, and any whole-tree pass.\n// Adding a feature = one module here + one entry in the registry (index.ts).\nexport interface Feature {\n  name: FeatureName;\n  schema(db: DatabaseSync): void;\n  // Pure, per file: raw is the full file, body the prose after frontmatter,\n  // search the normalized title/summary strings.\n  extract?(raw: string, body: string, search?: { title: string; summary: string }): unknown;\n  remove?(db: DatabaseSync, path: string, delta: ReconcileDelta): void;\n  store?(db: DatabaseSync, path: string, extracted: unknown, delta: ReconcileDelta): void;\n  // After all rows are current, inside the reconcile transaction (resolution, rank).\n  // Hooks run in registry order and communicate through fields on `delta`.\n  afterReconcile?(db: DatabaseSync, delta: ReconcileDelta): void;\n  // Per-file opt-out (currently only embed, via FileStat.embed): absent means the feature\n  // applies to every file it's otherwise on for. Checked per file before extract/store so a\n  // doc in a tree with no embedding model gets no rows for it at all.\n  enabledForFile?(cfg: Config, file: FileStat): boolean;\n}\n"],"names":[],"mappings":"AAeA,mFAAmF;AACnF,6EAA6E;AAC7E,WAeC"}

package/dist/esm/index.d.ts

@@ -1,9 +1,9 @@
 export type { Peek, PresetCoverage, SearchOptions, TreeMap } from './commands.js';
 export { mapTree, peek, presetCoverage, search } from './commands.js';
-export type { Config, EmbedConfig, FeatureName, Preset, ResolvedConfig, SavedQuery, SavedSearch } from './config.js';
+export type { Config, EmbedConfig, FeatureName, Preset, ResolvedConfig, SavedQuery, SavedSearch, SearchOverrides } from './config.js';
 export { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
 export type { OpenResult } from './db.js';
-export { open, rebuild } from './db.js';
+export { clearCache, open } from './db.js';
 export type { SenseErrorCode } from './errors.js';
 export { SenseError } from './errors.js';
 export type { Row } from './output.js';

package/dist/esm/index.js

@@ -2,7 +2,7 @@
 // internals (feature registry, graph, scan, meta) stay module-private.
 export { mapTree, peek, presetCoverage, search } from './commands.js';
 export { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.js';
-export { open, rebuild } from './db.js';
+export { clearCache, open } from './db.js';
 export { SenseError } from './errors.js';
 export { printRows } from './output.js';
 export { runWatch } from './watch.js';

package/dist/esm/index.js.map

@@ -1 +1 @@
-{"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 { Peek, PresetCoverage, SearchOptions, TreeMap } from './commands.ts';\nexport { mapTree, peek, presetCoverage, search } from './commands.ts';\nexport type { Config, EmbedConfig, FeatureName, Preset, ResolvedConfig, SavedQuery, SavedSearch } from './config.ts';\nexport { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.ts';\nexport type { OpenResult } from './db.ts';\nexport { open, rebuild } from './db.ts';\nexport type { SenseErrorCode } from './errors.ts';\nexport { SenseError } from './errors.ts';\nexport type { Row } from './output.ts';\nexport { printRows } from './output.ts';\n\nexport type { WatchEvent, WatchOptions } from './watch.ts';\nexport { runWatch } from './watch.ts';\n"],"names":["mapTree","peek","presetCoverage","search","CONFIG_FILENAME","initConfig","loadConfig","migrateConfig","STATE_DIR","SUPPORTED_CONFIG_VERSION","open","rebuild","SenseError","printRows","runWatch"],"mappings":"AAAA,+EAA+E;AAC/E,uEAAuE;AAGvE,SAASA,OAAO,EAAEC,IAAI,EAAEC,cAAc,EAAEC,MAAM,QAAQ,gBAAgB;AAEtE,SAASC,eAAe,EAAEC,UAAU,EAAEC,UAAU,EAAEC,aAAa,EAAEC,SAAS,EAAEC,wBAAwB,QAAQ,cAAc;AAE1H,SAASC,IAAI,EAAEC,OAAO,QAAQ,UAAU;AAExC,SAASC,UAAU,QAAQ,cAAc;AAEzC,SAASC,SAAS,QAAQ,cAAc;AAGxC,SAASC,QAAQ,QAAQ,aAAa"}
+{"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 { Peek, PresetCoverage, SearchOptions, TreeMap } from './commands.ts';\nexport { mapTree, peek, presetCoverage, search } from './commands.ts';\nexport type { Config, EmbedConfig, FeatureName, Preset, ResolvedConfig, SavedQuery, SavedSearch, SearchOverrides } from './config.ts';\nexport { CONFIG_FILENAME, initConfig, loadConfig, migrateConfig, STATE_DIR, SUPPORTED_CONFIG_VERSION } from './config.ts';\nexport type { OpenResult } from './db.ts';\nexport { clearCache, open } from './db.ts';\nexport type { SenseErrorCode } from './errors.ts';\nexport { SenseError } from './errors.ts';\nexport type { Row } from './output.ts';\nexport { printRows } from './output.ts';\n\nexport type { WatchEvent, WatchOptions } from './watch.ts';\nexport { runWatch } from './watch.ts';\n"],"names":["mapTree","peek","presetCoverage","search","CONFIG_FILENAME","initConfig","loadConfig","migrateConfig","STATE_DIR","SUPPORTED_CONFIG_VERSION","clearCache","open","SenseError","printRows","runWatch"],"mappings":"AAAA,+EAA+E;AAC/E,uEAAuE;AAGvE,SAASA,OAAO,EAAEC,IAAI,EAAEC,cAAc,EAAEC,MAAM,QAAQ,gBAAgB;AAEtE,SAASC,eAAe,EAAEC,UAAU,EAAEC,UAAU,EAAEC,aAAa,EAAEC,SAAS,EAAEC,wBAAwB,QAAQ,cAAc;AAE1H,SAASC,UAAU,EAAEC,IAAI,QAAQ,UAAU;AAE3C,SAASC,UAAU,QAAQ,cAAc;AAEzC,SAASC,SAAS,QAAQ,cAAc;AAGxC,SAASC,QAAQ,QAAQ,aAAa"}

package/dist/esm/output.d.ts

@@ -9,6 +9,7 @@ export declare function presetsLines(presets: Array<{
     name: string;
     files: number;
     embedded: number;
+    semantic?: boolean;
 }>): string[];
 export declare function renderMap(result: {
     docs: {
@@ -25,6 +26,7 @@ export declare function renderMap(result: {
         name: string;
         files: number;
         embedded: number;
+        semantic?: boolean;
     }>;
     hubs: Row[];
     recent: Row[];
@@ -33,6 +35,7 @@ export declare function renderPeek(result: {
     path: string;
     tokens: number;
     frontmatter: Row;
+    parseError?: string | null;
     sections: Row[];
     outbound: string[];
     backlinks: string[];

package/dist/esm/output.js

@@ -12,11 +12,8 @@ export function printRows(rows, format) {
         console.warn(`warning: result is ${Math.round(rendered.length / 1000)} KB across ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);
     }
 }
-// Table output is for reading; a row wider than the terminal wraps into an unreadable
-// block (a `find` header measured 330 chars once `summary` is present, which drove one
-// user to pipe --format json through a script for every interactive query). Widest
-// columns give up space first, down to a floor, and only as far as the terminal needs.
-// --format json is never truncated: it is the machine-readable surface.
+// Tables are for reading, and a row wider than the terminal wraps unreadably: widest columns
+// give up space first, down to a floor. --format json is never truncated.
 const MIN_COLUMN = 12;
 const TRUNCATED = '…';
 function fitWidths(widths, limit) {
@@ -48,19 +45,21 @@ function renderRows(rows, format, width = process.stdout.columns) {
     ].join('\n');
 }
 // Text renderers for the top-level commands; cli.ts prints what these return.
-// `embed` is not a features-block key (it's derived from presets), so its off hint reads
-// differently from links/sections/rank's plain `features.<name>`.
+// Names the config key that turns a feature back on. Only the three features-block toggles
+// reach here; `embed` has its own status line.
 export function featureOffHint(name) {
-    return name === 'embed' ? 'no preset has semantic on' : `features.${name}`;
+    return `features.${name}`;
 }
 export function featuresLine(features) {
     const off = features.off.length > 0 ? ` · off: ${features.off.map((name)=>`${name} (${featureOffHint(name)})`).join(', ')}` : '';
     return `features: ${features.on.join(', ')}${off}`;
 }
 export function presetsLines(presets) {
+    // "0 embedded" is ambiguous between a scope that declined vectors and one that has not built
+    // them yet, so a semantic-off preset says which.
     return [
         'presets:',
-        ...presets.map((p)=>`  ${p.name}: ${p.files} file(s), ${p.embedded} embedded`)
+        ...presets.map((p)=>`  ${p.name}: ${p.files} file(s), ${p.embedded} embedded${p.semantic === false ? ' (semantic: false)' : ''}`)
     ];
 }
 export function renderMap(result) {
@@ -82,6 +81,9 @@ export function renderPeek(result) {
     const lines = [
         `${result.path}  (~${result.tokens} tokens)`
     ];
+    // Otherwise a refused parse is indistinguishable from a note that has no frontmatter, which
+    // is the confusion the whole quarantine design exists to remove.
+    if (result.parseError) lines.push(`  frontmatter: did not parse, so none of it is indexed (${result.parseError})`);
     for (const [key, value] of Object.entries(result.frontmatter))lines.push(`  ${key}: ${value}`);
     if (result.sections.length > 0) {
         lines.push('', 'sections:');
@@ -90,14 +92,14 @@ export function renderPeek(result) {
     } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');
     if (result.off.includes('links')) {
         lines.push('', 'links: off (features.links)');
-        return lines.join('\n');
+    } else {
+        const linkLine = (label, shown, total)=>{
+            if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);
+        };
+        if (result.outboundTotal + result.unresolvedTotal + result.backlinksTotal > 0) lines.push('');
+        linkLine('links out', result.outbound, result.outboundTotal);
+        linkLine('unresolved', result.unresolved, result.unresolvedTotal);
+        linkLine('backlinks', result.backlinks, result.backlinksTotal);
     }
-    const linkLine = (label, shown, total)=>{
-        if (total > 0) lines.push(`${label} (${total}): ${shown.join(', ')}${total > shown.length ? `, +${total - shown.length} more` : ''}`);
-    };
-    if (result.outboundTotal + result.unresolvedTotal + result.backlinksTotal > 0) lines.push('');
-    linkLine('links out', result.outbound, result.outboundTotal);
-    linkLine('unresolved', result.unresolved, result.unresolvedTotal);
-    linkLine('backlinks', result.backlinks, result.backlinksTotal);
     return lines.join('\n');
 }

package/dist/esm/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 = renderRows(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 ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);\n  }\n}\n\n// Table output is for reading; a row wider than the terminal wraps into an unreadable\n// block (a `find` header measured 330 chars once `summary` is present, which drove one\n// user to pipe --format json through a script for every interactive query). Widest\n// columns give up space first, down to a floor, and only as far as the terminal needs.\n// --format json is never truncated: it is the machine-readable surface.\nconst MIN_COLUMN = 12;\nconst TRUNCATED = '…';\n\nfunction fitWidths(widths: number[], limit: number): number[] {\n  const gaps = (widths.length - 1) * 2;\n  const fitted = [...widths];\n  for (;;) {\n    const total = fitted.reduce((a, b) => a + b, 0) + gaps;\n    if (total <= limit) return fitted;\n    const widest = fitted.indexOf(Math.max(...fitted));\n    if (fitted[widest] <= MIN_COLUMN) return fitted;\n    fitted[widest] = Math.max(MIN_COLUMN, fitted[widest] - Math.max(1, total - limit));\n  }\n}\n\nfunction renderRows(rows: Row[], format: 'table' | 'json', width = process.stdout.columns): string {\n  if (format === 'json') return JSON.stringify(rows, null, 2);\n  if (rows.length === 0) return '(0 rows)';\n\n  const columns = Object.keys(rows[0]);\n  const cells = rows.map((row) => columns.map((col) => cell(row[col])));\n  const natural = columns.map((col, i) => Math.max(col.length, ...cells.map((row) => row[i].length)));\n  const widths = width && width > MIN_COLUMN ? fitWidths(natural, width) : natural;\n\n  const clip = (value: string, w: number) => (value.length <= w ? value.padEnd(w) : `${value.slice(0, w - 1)}${TRUNCATED}`);\n  const formatRow = (values: string[]) =>\n    values\n      .map((value, i) => clip(value, widths[i]))\n      .join('  ')\n      .trimEnd();\n\n  return [formatRow(columns), widths.map((w) => '-'.repeat(w)).join('  '), ...cells.map(formatRow)].join('\\n');\n}\n\n// Text renderers for the top-level commands; cli.ts prints what these return.\n\n// `embed` is not a features-block key (it's derived from presets), so its off hint reads\n// differently from links/sections/rank's plain `features.<name>`.\nexport function featureOffHint(name: string): string {\n  return name === 'embed' ? 'no preset has semantic on' : `features.${name}`;\n}\n\nexport function featuresLine(features: { on: string[]; off: string[] }): string {\n  const off = features.off.length > 0 ? ` · off: ${features.off.map((name) => `${name} (${featureOffHint(name)})`).join(', ')}` : '';\n  return `features: ${features.on.join(', ')}${off}`;\n}\n\nexport function presetsLines(presets: Array<{ name: string; files: number; embedded: number }>): string[] {\n  return ['presets:', ...presets.map((p) => `  ${p.name}: ${p.files} file(s), ${p.embedded} embedded`)];\n}\n\nexport function renderMap(result: { docs: { count: number; bytes: number }; fields: Row[]; fieldsTotal: number; features: { on: string[]; off: string[] }; presets: Array<{ name: string; files: number; embedded: number }>; hubs: Row[]; recent: Row[] }): string {\n  const parts = [`docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)`, featuresLine(result.features), '', ...presetsLines(result.presets), '', renderRows(result.fields, 'table')];\n  if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);\n  if (result.hubs.length > 0) parts.push('\\nhubs (by link rank):', renderRows(result.hubs, 'table'));\n  else if (result.features.off.includes('rank')) parts.push(`\\nhubs: off (${featureOffHint('rank')})`);\n  parts.push('\\nrecent:', renderRows(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[]; sectionsTotal: number; 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    if (result.sectionsTotal > result.sections.length) lines.push(`  (+${result.sectionsTotal - result.sections.length} more sections -- sections table has all of them)`);\n  } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');\n  if (result.off.includes('links')) {\n    lines.push('', 'links: off (features.links)');\n    return lines.join('\\n');\n  }\n  const linkLine = (label: string, shown: string[], total: number) => {\n    if (total > 0) lines.push(`${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":["OVERSIZE_BYTES","cell","value","String","replace","printRows","rows","format","rendered","renderRows","console","log","length","warn","Math","round","MIN_COLUMN","TRUNCATED","fitWidths","widths","limit","gaps","fitted","total","reduce","a","b","widest","indexOf","max","width","process","stdout","columns","JSON","stringify","Object","keys","cells","map","row","col","natural","i","clip","w","padEnd","slice","formatRow","values","join","trimEnd","repeat","featureOffHint","name","featuresLine","features","off","on","presetsLines","presets","p","files","embedded","renderMap","result","parts","docs","count","bytes","fields","fieldsTotal","push","hubs","includes","recent","renderPeek","lines","path","tokens","key","entries","frontmatter","sections","s","level","heading","start_line","end_line","sectionsTotal","linkLine","label","shown","outboundTotal","unresolvedTotal","backlinksTotal","outbound","unresolved","backlinks"],"mappings":"AAAA,iCAAiC;AAIjC,yFAAyF;AACzF,MAAMA,iBAAiB;AAEvB,mGAAmG;AACnG,SAASC,KAAKC,KAAc;IAC1B,OAAOC,OAAOD,kBAAAA,mBAAAA,QAAS,IAAIE,OAAO,CAAC,aAAa;AAClD;AAEA,OAAO,SAASC,UAAUC,IAAW,EAAEC,MAAwB;IAC7D,MAAMC,WAAWC,WAAWH,MAAMC;IAClCG,QAAQC,GAAG,CAACH;IACZ,IAAIA,SAASI,MAAM,GAAGZ,gBAAgB;QACpCU,QAAQG,IAAI,CAAC,CAAC,mBAAmB,EAAEC,KAAKC,KAAK,CAACP,SAASI,MAAM,GAAG,MAAM,WAAW,EAAEN,KAAKM,MAAM,CAAC,8EAA8E,CAAC;IAChL;AACF;AAEA,sFAAsF;AACtF,uFAAuF;AACvF,mFAAmF;AACnF,uFAAuF;AACvF,wEAAwE;AACxE,MAAMI,aAAa;AACnB,MAAMC,YAAY;AAElB,SAASC,UAAUC,MAAgB,EAAEC,KAAa;IAChD,MAAMC,OAAO,AAACF,CAAAA,OAAOP,MAAM,GAAG,CAAA,IAAK;IACnC,MAAMU,SAAS;WAAIH;KAAO;IAC1B,OAAS;QACP,MAAMI,QAAQD,OAAOE,MAAM,CAAC,CAACC,GAAGC,IAAMD,IAAIC,GAAG,KAAKL;QAClD,IAAIE,SAASH,OAAO,OAAOE;QAC3B,MAAMK,SAASL,OAAOM,OAAO,CAACd,KAAKe,GAAG,IAAIP;QAC1C,IAAIA,MAAM,CAACK,OAAO,IAAIX,YAAY,OAAOM;QACzCA,MAAM,CAACK,OAAO,GAAGb,KAAKe,GAAG,CAACb,YAAYM,MAAM,CAACK,OAAO,GAAGb,KAAKe,GAAG,CAAC,GAAGN,QAAQH;IAC7E;AACF;AAEA,SAASX,WAAWH,IAAW,EAAEC,MAAwB,EAAEuB,QAAQC,QAAQC,MAAM,CAACC,OAAO;IACvF,IAAI1B,WAAW,QAAQ,OAAO2B,KAAKC,SAAS,CAAC7B,MAAM,MAAM;IACzD,IAAIA,KAAKM,MAAM,KAAK,GAAG,OAAO;IAE9B,MAAMqB,UAAUG,OAAOC,IAAI,CAAC/B,IAAI,CAAC,EAAE;IACnC,MAAMgC,QAAQhC,KAAKiC,GAAG,CAAC,CAACC,MAAQP,QAAQM,GAAG,CAAC,CAACE,MAAQxC,KAAKuC,GAAG,CAACC,IAAI;IAClE,MAAMC,UAAUT,QAAQM,GAAG,CAAC,CAACE,KAAKE,IAAM7B,KAAKe,GAAG,CAACY,IAAI7B,MAAM,KAAK0B,MAAMC,GAAG,CAAC,CAACC,MAAQA,GAAG,CAACG,EAAE,CAAC/B,MAAM;IAChG,MAAMO,SAASW,SAASA,QAAQd,aAAaE,UAAUwB,SAASZ,SAASY;IAEzE,MAAME,OAAO,CAAC1C,OAAe2C,IAAe3C,MAAMU,MAAM,IAAIiC,IAAI3C,MAAM4C,MAAM,CAACD,KAAK,GAAG3C,MAAM6C,KAAK,CAAC,GAAGF,IAAI,KAAK5B,WAAW;IACxH,MAAM+B,YAAY,CAACC,SACjBA,OACGV,GAAG,CAAC,CAACrC,OAAOyC,IAAMC,KAAK1C,OAAOiB,MAAM,CAACwB,EAAE,GACvCO,IAAI,CAAC,MACLC,OAAO;IAEZ,OAAO;QAACH,UAAUf;QAAUd,OAAOoB,GAAG,CAAC,CAACM,IAAM,IAAIO,MAAM,CAACP,IAAIK,IAAI,CAAC;WAAUZ,MAAMC,GAAG,CAACS;KAAW,CAACE,IAAI,CAAC;AACzG;AAEA,8EAA8E;AAE9E,yFAAyF;AACzF,kEAAkE;AAClE,OAAO,SAASG,eAAeC,IAAY;IACzC,OAAOA,SAAS,UAAU,8BAA8B,CAAC,SAAS,EAAEA,MAAM;AAC5E;AAEA,OAAO,SAASC,aAAaC,QAAyC;IACpE,MAAMC,MAAMD,SAASC,GAAG,CAAC7C,MAAM,GAAG,IAAI,CAAC,QAAQ,EAAE4C,SAASC,GAAG,CAAClB,GAAG,CAAC,CAACe,OAAS,GAAGA,KAAK,EAAE,EAAED,eAAeC,MAAM,CAAC,CAAC,EAAEJ,IAAI,CAAC,OAAO,GAAG;IAChI,OAAO,CAAC,UAAU,EAAEM,SAASE,EAAE,CAACR,IAAI,CAAC,QAAQO,KAAK;AACpD;AAEA,OAAO,SAASE,aAAaC,OAAiE;IAC5F,OAAO;QAAC;WAAeA,QAAQrB,GAAG,CAAC,CAACsB,IAAM,CAAC,EAAE,EAAEA,EAAEP,IAAI,CAAC,EAAE,EAAEO,EAAEC,KAAK,CAAC,UAAU,EAAED,EAAEE,QAAQ,CAAC,SAAS,CAAC;KAAE;AACvG;AAEA,OAAO,SAASC,UAAUC,MAAgO;IACxP,MAAMC,QAAQ;QAAC,CAAC,MAAM,EAAED,OAAOE,IAAI,CAACC,KAAK,CAAC,EAAE,EAAEtD,KAAKC,KAAK,CAACkD,OAAOE,IAAI,CAACE,KAAK,GAAG,MAAM,IAAI,CAAC;QAAEd,aAAaU,OAAOT,QAAQ;QAAG;WAAOG,aAAaM,OAAOL,OAAO;QAAG;QAAInD,WAAWwD,OAAOK,MAAM,EAAE;KAAS;IACrM,IAAIL,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC1D,MAAM,EAAEsD,MAAMM,IAAI,CAAC,CAAC,EAAE,EAAEP,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC1D,MAAM,CAAC,aAAa,CAAC;IACvH,IAAIqD,OAAOQ,IAAI,CAAC7D,MAAM,GAAG,GAAGsD,MAAMM,IAAI,CAAC,0BAA0B/D,WAAWwD,OAAOQ,IAAI,EAAE;SACpF,IAAIR,OAAOT,QAAQ,CAACC,GAAG,CAACiB,QAAQ,CAAC,SAASR,MAAMM,IAAI,CAAC,CAAC,aAAa,EAAEnB,eAAe,QAAQ,CAAC,CAAC;IACnGa,MAAMM,IAAI,CAAC,aAAa/D,WAAWwD,OAAOU,MAAM,EAAE;IAClD,OAAOT,MAAMhB,IAAI,CAAC;AACpB;AAEA,OAAO,SAAS0B,WAAWX,MAAwP;IACjR,MAAMY,QAAQ;QAAC,GAAGZ,OAAOa,IAAI,CAAC,IAAI,EAAEb,OAAOc,MAAM,CAAC,QAAQ,CAAC;KAAC;IAC5D,KAAK,MAAM,CAACC,KAAK9E,MAAM,IAAIkC,OAAO6C,OAAO,CAAChB,OAAOiB,WAAW,EAAGL,MAAML,IAAI,CAAC,CAAC,EAAE,EAAEQ,IAAI,EAAE,EAAE9E,OAAO;IAC9F,IAAI+D,OAAOkB,QAAQ,CAACvE,MAAM,GAAG,GAAG;QAC9BiE,MAAML,IAAI,CAAC,IAAI;QACf,KAAK,MAAMY,KAAKnB,OAAOkB,QAAQ,CAAEN,MAAML,IAAI,CAAC,CAAC,EAAE,EAAE,IAAIpB,MAAM,CAACgC,EAAEC,KAAK,EAAY,CAAC,EAAED,EAAEE,OAAO,CAAC,IAAI,EAAEF,EAAEG,UAAU,CAAC,CAAC,EAAEH,EAAEI,QAAQ,CAAC,GAAG,EAAEJ,EAAEL,MAAM,CAAC,EAAE,CAAC;QAC9I,IAAId,OAAOwB,aAAa,GAAGxB,OAAOkB,QAAQ,CAACvE,MAAM,EAAEiE,MAAML,IAAI,CAAC,CAAC,IAAI,EAAEP,OAAOwB,aAAa,GAAGxB,OAAOkB,QAAQ,CAACvE,MAAM,CAAC,iDAAiD,CAAC;IACvK,OAAO,IAAIqD,OAAOR,GAAG,CAACiB,QAAQ,CAAC,aAAaG,MAAML,IAAI,CAAC,IAAI;IAC3D,IAAIP,OAAOR,GAAG,CAACiB,QAAQ,CAAC,UAAU;QAChCG,MAAML,IAAI,CAAC,IAAI;QACf,OAAOK,MAAM3B,IAAI,CAAC;IACpB;IACA,MAAMwC,WAAW,CAACC,OAAeC,OAAiBrE;QAChD,IAAIA,QAAQ,GAAGsD,MAAML,IAAI,CAAC,GAAGmB,MAAM,EAAE,EAAEpE,MAAM,GAAG,EAAEqE,MAAM1C,IAAI,CAAC,QAAQ3B,QAAQqE,MAAMhF,MAAM,GAAG,CAAC,GAAG,EAAEW,QAAQqE,MAAMhF,MAAM,CAAC,KAAK,CAAC,GAAG,IAAI;IACtI;IACA,IAAIqD,OAAO4B,aAAa,GAAG5B,OAAO6B,eAAe,GAAG7B,OAAO8B,cAAc,GAAG,GAAGlB,MAAML,IAAI,CAAC;IAC1FkB,SAAS,aAAazB,OAAO+B,QAAQ,EAAE/B,OAAO4B,aAAa;IAC3DH,SAAS,cAAczB,OAAOgC,UAAU,EAAEhC,OAAO6B,eAAe;IAChEJ,SAAS,aAAazB,OAAOiC,SAAS,EAAEjC,OAAO8B,cAAc;IAC7D,OAAOlB,MAAM3B,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 = renderRows(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 ${rows.length} row(s); consider a LIMIT, fewer columns, or snippet() instead of whole values`);\n  }\n}\n\n// Tables are for reading, and a row wider than the terminal wraps unreadably: widest columns\n// give up space first, down to a floor. --format json is never truncated.\nconst MIN_COLUMN = 12;\nconst TRUNCATED = '…';\n\nfunction fitWidths(widths: number[], limit: number): number[] {\n  const gaps = (widths.length - 1) * 2;\n  const fitted = [...widths];\n  for (;;) {\n    const total = fitted.reduce((a, b) => a + b, 0) + gaps;\n    if (total <= limit) return fitted;\n    const widest = fitted.indexOf(Math.max(...fitted));\n    if (fitted[widest] <= MIN_COLUMN) return fitted;\n    fitted[widest] = Math.max(MIN_COLUMN, fitted[widest] - Math.max(1, total - limit));\n  }\n}\n\nfunction renderRows(rows: Row[], format: 'table' | 'json', width = process.stdout.columns): string {\n  if (format === 'json') return JSON.stringify(rows, null, 2);\n  if (rows.length === 0) return '(0 rows)';\n\n  const columns = Object.keys(rows[0]);\n  const cells = rows.map((row) => columns.map((col) => cell(row[col])));\n  const natural = columns.map((col, i) => Math.max(col.length, ...cells.map((row) => row[i].length)));\n  const widths = width && width > MIN_COLUMN ? fitWidths(natural, width) : natural;\n\n  const clip = (value: string, w: number) => (value.length <= w ? value.padEnd(w) : `${value.slice(0, w - 1)}${TRUNCATED}`);\n  const formatRow = (values: string[]) =>\n    values\n      .map((value, i) => clip(value, widths[i]))\n      .join('  ')\n      .trimEnd();\n\n  return [formatRow(columns), widths.map((w) => '-'.repeat(w)).join('  '), ...cells.map(formatRow)].join('\\n');\n}\n\n// Text renderers for the top-level commands; cli.ts prints what these return.\n\n// Names the config key that turns a feature back on. Only the three features-block toggles\n// reach here; `embed` has its own status line.\nexport function featureOffHint(name: string): string {\n  return `features.${name}`;\n}\n\nexport function featuresLine(features: { on: string[]; off: string[] }): string {\n  const off = features.off.length > 0 ? ` · off: ${features.off.map((name) => `${name} (${featureOffHint(name)})`).join(', ')}` : '';\n  return `features: ${features.on.join(', ')}${off}`;\n}\n\nexport function presetsLines(presets: Array<{ name: string; files: number; embedded: number; semantic?: boolean }>): string[] {\n  // \"0 embedded\" is ambiguous between a scope that declined vectors and one that has not built\n  // them yet, so a semantic-off preset says which.\n  return ['presets:', ...presets.map((p) => `  ${p.name}: ${p.files} file(s), ${p.embedded} embedded${p.semantic === false ? ' (semantic: false)' : ''}`)];\n}\n\nexport function renderMap(result: { docs: { count: number; bytes: number }; fields: Row[]; fieldsTotal: number; features: { on: string[]; off: string[] }; presets: Array<{ name: string; files: number; embedded: number; semantic?: boolean }>; hubs: Row[]; recent: Row[] }): string {\n  const parts = [`docs: ${result.docs.count} (${Math.round(result.docs.bytes / 1024)} KB)`, featuresLine(result.features), '', ...presetsLines(result.presets), '', renderRows(result.fields, 'table')];\n  if (result.fieldsTotal > result.fields.length) parts.push(`(+${result.fieldsTotal - result.fields.length} more fields)`);\n  if (result.hubs.length > 0) parts.push('\\nhubs (by link rank):', renderRows(result.hubs, 'table'));\n  else if (result.features.off.includes('rank')) parts.push(`\\nhubs: off (${featureOffHint('rank')})`);\n  parts.push('\\nrecent:', renderRows(result.recent, 'table'));\n  return parts.join('\\n');\n}\n\nexport function renderPeek(result: { path: string; tokens: number; frontmatter: Row; parseError?: string | null; sections: Row[]; outbound: string[]; backlinks: string[]; unresolved: string[]; sectionsTotal: number; outboundTotal: number; backlinksTotal: number; unresolvedTotal: number; off: string[] }): string {\n  const lines = [`${result.path}  (~${result.tokens} tokens)`];\n  // Otherwise a refused parse is indistinguishable from a note that has no frontmatter, which\n  // is the confusion the whole quarantine design exists to remove.\n  if (result.parseError) lines.push(`  frontmatter: did not parse, so none of it is indexed (${result.parseError})`);\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    if (result.sectionsTotal > result.sections.length) lines.push(`  (+${result.sectionsTotal - result.sections.length} more sections -- sections table has all of them)`);\n  } else if (result.off.includes('sections')) lines.push('', 'sections: off (features.sections)');\n  if (result.off.includes('links')) {\n    lines.push('', 'links: off (features.links)');\n  } else {\n    const linkLine = (label: string, shown: string[], total: number) => {\n      if (total > 0) lines.push(`${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  }\n  return lines.join('\\n');\n}\n"],"names":["OVERSIZE_BYTES","cell","value","String","replace","printRows","rows","format","rendered","renderRows","console","log","length","warn","Math","round","MIN_COLUMN","TRUNCATED","fitWidths","widths","limit","gaps","fitted","total","reduce","a","b","widest","indexOf","max","width","process","stdout","columns","JSON","stringify","Object","keys","cells","map","row","col","natural","i","clip","w","padEnd","slice","formatRow","values","join","trimEnd","repeat","featureOffHint","name","featuresLine","features","off","on","presetsLines","presets","p","files","embedded","semantic","renderMap","result","parts","docs","count","bytes","fields","fieldsTotal","push","hubs","includes","recent","renderPeek","lines","path","tokens","parseError","key","entries","frontmatter","sections","s","level","heading","start_line","end_line","sectionsTotal","linkLine","label","shown","outboundTotal","unresolvedTotal","backlinksTotal","outbound","unresolved","backlinks"],"mappings":"AAAA,iCAAiC;AAIjC,yFAAyF;AACzF,MAAMA,iBAAiB;AAEvB,mGAAmG;AACnG,SAASC,KAAKC,KAAc;IAC1B,OAAOC,OAAOD,kBAAAA,mBAAAA,QAAS,IAAIE,OAAO,CAAC,aAAa;AAClD;AAEA,OAAO,SAASC,UAAUC,IAAW,EAAEC,MAAwB;IAC7D,MAAMC,WAAWC,WAAWH,MAAMC;IAClCG,QAAQC,GAAG,CAACH;IACZ,IAAIA,SAASI,MAAM,GAAGZ,gBAAgB;QACpCU,QAAQG,IAAI,CAAC,CAAC,mBAAmB,EAAEC,KAAKC,KAAK,CAACP,SAASI,MAAM,GAAG,MAAM,WAAW,EAAEN,KAAKM,MAAM,CAAC,8EAA8E,CAAC;IAChL;AACF;AAEA,6FAA6F;AAC7F,0EAA0E;AAC1E,MAAMI,aAAa;AACnB,MAAMC,YAAY;AAElB,SAASC,UAAUC,MAAgB,EAAEC,KAAa;IAChD,MAAMC,OAAO,AAACF,CAAAA,OAAOP,MAAM,GAAG,CAAA,IAAK;IACnC,MAAMU,SAAS;WAAIH;KAAO;IAC1B,OAAS;QACP,MAAMI,QAAQD,OAAOE,MAAM,CAAC,CAACC,GAAGC,IAAMD,IAAIC,GAAG,KAAKL;QAClD,IAAIE,SAASH,OAAO,OAAOE;QAC3B,MAAMK,SAASL,OAAOM,OAAO,CAACd,KAAKe,GAAG,IAAIP;QAC1C,IAAIA,MAAM,CAACK,OAAO,IAAIX,YAAY,OAAOM;QACzCA,MAAM,CAACK,OAAO,GAAGb,KAAKe,GAAG,CAACb,YAAYM,MAAM,CAACK,OAAO,GAAGb,KAAKe,GAAG,CAAC,GAAGN,QAAQH;IAC7E;AACF;AAEA,SAASX,WAAWH,IAAW,EAAEC,MAAwB,EAAEuB,QAAQC,QAAQC,MAAM,CAACC,OAAO;IACvF,IAAI1B,WAAW,QAAQ,OAAO2B,KAAKC,SAAS,CAAC7B,MAAM,MAAM;IACzD,IAAIA,KAAKM,MAAM,KAAK,GAAG,OAAO;IAE9B,MAAMqB,UAAUG,OAAOC,IAAI,CAAC/B,IAAI,CAAC,EAAE;IACnC,MAAMgC,QAAQhC,KAAKiC,GAAG,CAAC,CAACC,MAAQP,QAAQM,GAAG,CAAC,CAACE,MAAQxC,KAAKuC,GAAG,CAACC,IAAI;IAClE,MAAMC,UAAUT,QAAQM,GAAG,CAAC,CAACE,KAAKE,IAAM7B,KAAKe,GAAG,CAACY,IAAI7B,MAAM,KAAK0B,MAAMC,GAAG,CAAC,CAACC,MAAQA,GAAG,CAACG,EAAE,CAAC/B,MAAM;IAChG,MAAMO,SAASW,SAASA,QAAQd,aAAaE,UAAUwB,SAASZ,SAASY;IAEzE,MAAME,OAAO,CAAC1C,OAAe2C,IAAe3C,MAAMU,MAAM,IAAIiC,IAAI3C,MAAM4C,MAAM,CAACD,KAAK,GAAG3C,MAAM6C,KAAK,CAAC,GAAGF,IAAI,KAAK5B,WAAW;IACxH,MAAM+B,YAAY,CAACC,SACjBA,OACGV,GAAG,CAAC,CAACrC,OAAOyC,IAAMC,KAAK1C,OAAOiB,MAAM,CAACwB,EAAE,GACvCO,IAAI,CAAC,MACLC,OAAO;IAEZ,OAAO;QAACH,UAAUf;QAAUd,OAAOoB,GAAG,CAAC,CAACM,IAAM,IAAIO,MAAM,CAACP,IAAIK,IAAI,CAAC;WAAUZ,MAAMC,GAAG,CAACS;KAAW,CAACE,IAAI,CAAC;AACzG;AAEA,8EAA8E;AAE9E,2FAA2F;AAC3F,+CAA+C;AAC/C,OAAO,SAASG,eAAeC,IAAY;IACzC,OAAO,CAAC,SAAS,EAAEA,MAAM;AAC3B;AAEA,OAAO,SAASC,aAAaC,QAAyC;IACpE,MAAMC,MAAMD,SAASC,GAAG,CAAC7C,MAAM,GAAG,IAAI,CAAC,QAAQ,EAAE4C,SAASC,GAAG,CAAClB,GAAG,CAAC,CAACe,OAAS,GAAGA,KAAK,EAAE,EAAED,eAAeC,MAAM,CAAC,CAAC,EAAEJ,IAAI,CAAC,OAAO,GAAG;IAChI,OAAO,CAAC,UAAU,EAAEM,SAASE,EAAE,CAACR,IAAI,CAAC,QAAQO,KAAK;AACpD;AAEA,OAAO,SAASE,aAAaC,OAAqF;IAChH,6FAA6F;IAC7F,iDAAiD;IACjD,OAAO;QAAC;WAAeA,QAAQrB,GAAG,CAAC,CAACsB,IAAM,CAAC,EAAE,EAAEA,EAAEP,IAAI,CAAC,EAAE,EAAEO,EAAEC,KAAK,CAAC,UAAU,EAAED,EAAEE,QAAQ,CAAC,SAAS,EAAEF,EAAEG,QAAQ,KAAK,QAAQ,uBAAuB,IAAI;KAAE;AAC1J;AAEA,OAAO,SAASC,UAAUC,MAAoP;IAC5Q,MAAMC,QAAQ;QAAC,CAAC,MAAM,EAAED,OAAOE,IAAI,CAACC,KAAK,CAAC,EAAE,EAAEvD,KAAKC,KAAK,CAACmD,OAAOE,IAAI,CAACE,KAAK,GAAG,MAAM,IAAI,CAAC;QAAEf,aAAaW,OAAOV,QAAQ;QAAG;WAAOG,aAAaO,OAAON,OAAO;QAAG;QAAInD,WAAWyD,OAAOK,MAAM,EAAE;KAAS;IACrM,IAAIL,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC3D,MAAM,EAAEuD,MAAMM,IAAI,CAAC,CAAC,EAAE,EAAEP,OAAOM,WAAW,GAAGN,OAAOK,MAAM,CAAC3D,MAAM,CAAC,aAAa,CAAC;IACvH,IAAIsD,OAAOQ,IAAI,CAAC9D,MAAM,GAAG,GAAGuD,MAAMM,IAAI,CAAC,0BAA0BhE,WAAWyD,OAAOQ,IAAI,EAAE;SACpF,IAAIR,OAAOV,QAAQ,CAACC,GAAG,CAACkB,QAAQ,CAAC,SAASR,MAAMM,IAAI,CAAC,CAAC,aAAa,EAAEpB,eAAe,QAAQ,CAAC,CAAC;IACnGc,MAAMM,IAAI,CAAC,aAAahE,WAAWyD,OAAOU,MAAM,EAAE;IAClD,OAAOT,MAAMjB,IAAI,CAAC;AACpB;AAEA,OAAO,SAAS2B,WAAWX,MAAoR;IAC7S,MAAMY,QAAQ;QAAC,GAAGZ,OAAOa,IAAI,CAAC,IAAI,EAAEb,OAAOc,MAAM,CAAC,QAAQ,CAAC;KAAC;IAC5D,4FAA4F;IAC5F,iEAAiE;IACjE,IAAId,OAAOe,UAAU,EAAEH,MAAML,IAAI,CAAC,CAAC,wDAAwD,EAAEP,OAAOe,UAAU,CAAC,CAAC,CAAC;IACjH,KAAK,MAAM,CAACC,KAAKhF,MAAM,IAAIkC,OAAO+C,OAAO,CAACjB,OAAOkB,WAAW,EAAGN,MAAML,IAAI,CAAC,CAAC,EAAE,EAAES,IAAI,EAAE,EAAEhF,OAAO;IAC9F,IAAIgE,OAAOmB,QAAQ,CAACzE,MAAM,GAAG,GAAG;QAC9BkE,MAAML,IAAI,CAAC,IAAI;QACf,KAAK,MAAMa,KAAKpB,OAAOmB,QAAQ,CAAEP,MAAML,IAAI,CAAC,CAAC,EAAE,EAAE,IAAIrB,MAAM,CAACkC,EAAEC,KAAK,EAAY,CAAC,EAAED,EAAEE,OAAO,CAAC,IAAI,EAAEF,EAAEG,UAAU,CAAC,CAAC,EAAEH,EAAEI,QAAQ,CAAC,GAAG,EAAEJ,EAAEN,MAAM,CAAC,EAAE,CAAC;QAC9I,IAAId,OAAOyB,aAAa,GAAGzB,OAAOmB,QAAQ,CAACzE,MAAM,EAAEkE,MAAML,IAAI,CAAC,CAAC,IAAI,EAAEP,OAAOyB,aAAa,GAAGzB,OAAOmB,QAAQ,CAACzE,MAAM,CAAC,iDAAiD,CAAC;IACvK,OAAO,IAAIsD,OAAOT,GAAG,CAACkB,QAAQ,CAAC,aAAaG,MAAML,IAAI,CAAC,IAAI;IAC3D,IAAIP,OAAOT,GAAG,CAACkB,QAAQ,CAAC,UAAU;QAChCG,MAAML,IAAI,CAAC,IAAI;IACjB,OAAO;QACL,MAAMmB,WAAW,CAACC,OAAeC,OAAiBvE;YAChD,IAAIA,QAAQ,GAAGuD,MAAML,IAAI,CAAC,GAAGoB,MAAM,EAAE,EAAEtE,MAAM,GAAG,EAAEuE,MAAM5C,IAAI,CAAC,QAAQ3B,QAAQuE,MAAMlF,MAAM,GAAG,CAAC,GAAG,EAAEW,QAAQuE,MAAMlF,MAAM,CAAC,KAAK,CAAC,GAAG,IAAI;QACtI;QACA,IAAIsD,OAAO6B,aAAa,GAAG7B,OAAO8B,eAAe,GAAG9B,OAAO+B,cAAc,GAAG,GAAGnB,MAAML,IAAI,CAAC;QAC1FmB,SAAS,aAAa1B,OAAOgC,QAAQ,EAAEhC,OAAO6B,aAAa;QAC3DH,SAAS,cAAc1B,OAAOiC,UAAU,EAAEjC,OAAO8B,eAAe;QAChEJ,SAAS,aAAa1B,OAAOkC,SAAS,EAAElC,OAAO+B,cAAc;IAC/D;IACA,OAAOnB,MAAM5B,IAAI,CAAC;AACpB"}

package/dist/esm/progress.js

@@ -1,7 +1,5 @@
-// Long-work progress on stderr only: stdout stays the parseable answer, so progress can
-// never corrupt --format json (the same channel warnings already use). TTY-aware: a human
-// terminal gets one self-overwriting line; a non-TTY consumer (agent, pipe, CI) gets at
-// most a few append-only lines -- an agent transcript gains lines, not redraw frames.
+// Progress on stderr only, so it can never corrupt --format json. A TTY gets one
+// self-overwriting line; a pipe gets a few append-only ones, not redraw frames.
 const NON_TTY_MILESTONES = 4;
 // `total` known upfront; `label` names the work ("embedding notes", "reparsing files").
 // Silent entirely below `threshold` -- short work needs no narration.

package/dist/esm/progress.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/progress.ts"],"sourcesContent":["// Long-work progress on stderr only: stdout stays the parseable answer, so progress can\n// never corrupt --format json (the same channel warnings already use). TTY-aware: a human\n// terminal gets one self-overwriting line; a non-TTY consumer (agent, pipe, CI) gets at\n// most a few append-only lines -- an agent transcript gains lines, not redraw frames.\n\nconst NON_TTY_MILESTONES = 4;\n\nexport interface Progress {\n  tick(done: number): void;\n  finish(): void;\n}\n\n// `total` known upfront; `label` names the work (\"embedding notes\", \"reparsing files\").\n// Silent entirely below `threshold` -- short work needs no narration.\nexport function progress(label: string, total: number, threshold = 200): Progress {\n  if (total < threshold) return { tick() {}, finish() {} };\n\n  const started = Date.now();\n  const isTTY = process.stderr.isTTY === true;\n  let lastShown = 0;\n  let wrote = false;\n\n  const line = (done: number) => {\n    const elapsed = (Date.now() - started) / 1000;\n    // Estimate from observed rate, only once there is a rate to observe.\n    const eta = done > 0 && done < total ? `, ~${Math.max(1, Math.round((elapsed / done) * (total - done)))}s left` : '';\n    return `${label}: ${done}/${total}${eta}`;\n  };\n\n  return {\n    tick(done: number) {\n      if (isTTY) {\n        process.stderr.write(`\\r${line(done)}`);\n        wrote = true;\n        return;\n      }\n      // Sparse milestones: first tick, then every 1/NON_TTY_MILESTONES of the total.\n      if (done === lastShown) return;\n      if (lastShown === 0 || done - lastShown >= total / NON_TTY_MILESTONES) {\n        process.stderr.write(`${line(done)}\\n`);\n        lastShown = done;\n        wrote = true;\n      }\n    },\n    finish() {\n      if (!wrote) return;\n      if (isTTY) process.stderr.write(`\\r${line(total)}\\n`);\n      else if (lastShown < total) process.stderr.write(`${line(total)}\\n`);\n    },\n  };\n}\n"],"names":["NON_TTY_MILESTONES","progress","label","total","threshold","tick","finish","started","Date","now","isTTY","process","stderr","lastShown","wrote","line","done","elapsed","eta","Math","max","round","write"],"mappings":"AAAA,wFAAwF;AACxF,0FAA0F;AAC1F,wFAAwF;AACxF,sFAAsF;AAEtF,MAAMA,qBAAqB;AAO3B,wFAAwF;AACxF,sEAAsE;AACtE,OAAO,SAASC,SAASC,KAAa,EAAEC,KAAa,EAAEC,YAAY,GAAG;IACpE,IAAID,QAAQC,WAAW,OAAO;QAAEC,SAAQ;QAAGC,WAAU;IAAE;IAEvD,MAAMC,UAAUC,KAAKC,GAAG;IACxB,MAAMC,QAAQC,QAAQC,MAAM,CAACF,KAAK,KAAK;IACvC,IAAIG,YAAY;IAChB,IAAIC,QAAQ;IAEZ,MAAMC,OAAO,CAACC;QACZ,MAAMC,UAAU,AAACT,CAAAA,KAAKC,GAAG,KAAKF,OAAM,IAAK;QACzC,qEAAqE;QACrE,MAAMW,MAAMF,OAAO,KAAKA,OAAOb,QAAQ,CAAC,GAAG,EAAEgB,KAAKC,GAAG,CAAC,GAAGD,KAAKE,KAAK,CAAC,AAACJ,UAAUD,OAASb,CAAAA,QAAQa,IAAG,IAAK,MAAM,CAAC,GAAG;QAClH,OAAO,GAAGd,MAAM,EAAE,EAAEc,KAAK,CAAC,EAAEb,QAAQe,KAAK;IAC3C;IAEA,OAAO;QACLb,MAAKW,IAAY;YACf,IAAIN,OAAO;gBACTC,QAAQC,MAAM,CAACU,KAAK,CAAC,CAAC,EAAE,EAAEP,KAAKC,OAAO;gBACtCF,QAAQ;gBACR;YACF;YACA,+EAA+E;YAC/E,IAAIE,SAASH,WAAW;YACxB,IAAIA,cAAc,KAAKG,OAAOH,aAAaV,QAAQH,oBAAoB;gBACrEW,QAAQC,MAAM,CAACU,KAAK,CAAC,GAAGP,KAAKC,MAAM,EAAE,CAAC;gBACtCH,YAAYG;gBACZF,QAAQ;YACV;QACF;QACAR;YACE,IAAI,CAACQ,OAAO;YACZ,IAAIJ,OAAOC,QAAQC,MAAM,CAACU,KAAK,CAAC,CAAC,EAAE,EAAEP,KAAKZ,OAAO,EAAE,CAAC;iBAC/C,IAAIU,YAAYV,OAAOQ,QAAQC,MAAM,CAACU,KAAK,CAAC,GAAGP,KAAKZ,OAAO,EAAE,CAAC;QACrE;IACF;AACF"}
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/progress.ts"],"sourcesContent":["// Progress on stderr only, so it can never corrupt --format json. A TTY gets one\n// self-overwriting line; a pipe gets a few append-only ones, not redraw frames.\n\nconst NON_TTY_MILESTONES = 4;\n\nexport interface Progress {\n  tick(done: number): void;\n  finish(): void;\n}\n\n// `total` known upfront; `label` names the work (\"embedding notes\", \"reparsing files\").\n// Silent entirely below `threshold` -- short work needs no narration.\nexport function progress(label: string, total: number, threshold = 200): Progress {\n  if (total < threshold) return { tick() {}, finish() {} };\n\n  const started = Date.now();\n  const isTTY = process.stderr.isTTY === true;\n  let lastShown = 0;\n  let wrote = false;\n\n  const line = (done: number) => {\n    const elapsed = (Date.now() - started) / 1000;\n    // Estimate from observed rate, only once there is a rate to observe.\n    const eta = done > 0 && done < total ? `, ~${Math.max(1, Math.round((elapsed / done) * (total - done)))}s left` : '';\n    return `${label}: ${done}/${total}${eta}`;\n  };\n\n  return {\n    tick(done: number) {\n      if (isTTY) {\n        process.stderr.write(`\\r${line(done)}`);\n        wrote = true;\n        return;\n      }\n      // Sparse milestones: first tick, then every 1/NON_TTY_MILESTONES of the total.\n      if (done === lastShown) return;\n      if (lastShown === 0 || done - lastShown >= total / NON_TTY_MILESTONES) {\n        process.stderr.write(`${line(done)}\\n`);\n        lastShown = done;\n        wrote = true;\n      }\n    },\n    finish() {\n      if (!wrote) return;\n      if (isTTY) process.stderr.write(`\\r${line(total)}\\n`);\n      else if (lastShown < total) process.stderr.write(`${line(total)}\\n`);\n    },\n  };\n}\n"],"names":["NON_TTY_MILESTONES","progress","label","total","threshold","tick","finish","started","Date","now","isTTY","process","stderr","lastShown","wrote","line","done","elapsed","eta","Math","max","round","write"],"mappings":"AAAA,iFAAiF;AACjF,gFAAgF;AAEhF,MAAMA,qBAAqB;AAO3B,wFAAwF;AACxF,sEAAsE;AACtE,OAAO,SAASC,SAASC,KAAa,EAAEC,KAAa,EAAEC,YAAY,GAAG;IACpE,IAAID,QAAQC,WAAW,OAAO;QAAEC,SAAQ;QAAGC,WAAU;IAAE;IAEvD,MAAMC,UAAUC,KAAKC,GAAG;IACxB,MAAMC,QAAQC,QAAQC,MAAM,CAACF,KAAK,KAAK;IACvC,IAAIG,YAAY;IAChB,IAAIC,QAAQ;IAEZ,MAAMC,OAAO,CAACC;QACZ,MAAMC,UAAU,AAACT,CAAAA,KAAKC,GAAG,KAAKF,OAAM,IAAK;QACzC,qEAAqE;QACrE,MAAMW,MAAMF,OAAO,KAAKA,OAAOb,QAAQ,CAAC,GAAG,EAAEgB,KAAKC,GAAG,CAAC,GAAGD,KAAKE,KAAK,CAAC,AAACJ,UAAUD,OAASb,CAAAA,QAAQa,IAAG,IAAK,MAAM,CAAC,GAAG;QAClH,OAAO,GAAGd,MAAM,EAAE,EAAEc,KAAK,CAAC,EAAEb,QAAQe,KAAK;IAC3C;IAEA,OAAO;QACLb,MAAKW,IAAY;YACf,IAAIN,OAAO;gBACTC,QAAQC,MAAM,CAACU,KAAK,CAAC,CAAC,EAAE,EAAEP,KAAKC,OAAO;gBACtCF,QAAQ;gBACR;YACF;YACA,+EAA+E;YAC/E,IAAIE,SAASH,WAAW;YACxB,IAAIA,cAAc,KAAKG,OAAOH,aAAaV,QAAQH,oBAAoB;gBACrEW,QAAQC,MAAM,CAACU,KAAK,CAAC,GAAGP,KAAKC,MAAM,EAAE,CAAC;gBACtCH,YAAYG;gBACZF,QAAQ;YACV;QACF;QACAR;YACE,IAAI,CAACQ,OAAO;YACZ,IAAIJ,OAAOC,QAAQC,MAAM,CAACU,KAAK,CAAC,CAAC,EAAE,EAAEP,KAAKZ,OAAO,EAAE,CAAC;iBAC/C,IAAIU,YAAYV,OAAOQ,QAAQC,MAAM,CAACU,KAAK,CAAC,GAAGP,KAAKZ,OAAO,EAAE,CAAC;QACrE;IACF;AACF"}

package/dist/esm/scan.d.ts

@@ -9,6 +9,7 @@ export interface FileStat {
     presets: string[];
     embed: boolean;
 }
+export declare function toPosixPath(relPath: string, separator?: string): string;
 export declare function listFiles(cfg: Config, baseDir: string): FileStat[];
 export interface ParsedDoc {
     relPath: string;
@@ -16,6 +17,7 @@ export interface ParsedDoc {
     size: number;
     presets: string[];
     data: Record<string, string | number | bigint | null>;
+    parseError: string | null;
     search: {
         title: string;
         summary: string;