sensemaking 0.27.0 → 0.27.1

package/dist/cjs/text/segment.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/text/segment.ts"],"sourcesContent":["import { stemmer } from 'stemmer';\n\n// Word boundaries FTS5's tokenizers cannot find on their own. Contract: seg(query) must appear in\n// seg(document) for any substring query; word mode is context-dependent and breaks that, so grapheme mode (UAX #29) is used.\n\n// Scripts written without word spaces: a closed set of writing systems.\n// Script_Extensions, not Script: the katakana long-vowel mark ー is Script=Common.\nexport const UNSPACED_SCRIPTS = '\\\\p{scx=Han}\\\\p{scx=Hiragana}\\\\p{scx=Katakana}\\\\p{scx=Thai}\\\\p{scx=Khmer}\\\\p{scx=Lao}\\\\p{scx=Myanmar}';\n// A run is script BASE characters with their combining marks attached; a bare mark after a\n// Latin letter (decomposed é) never starts one.\nconst RUN_BODY = `(?:[${UNSPACED_SCRIPTS}]\\\\p{M}*)+`;\nconst RUN = new RegExp(`(${RUN_BODY})`, 'gu');\nconst HAS_RUN = new RegExp(`[${UNSPACED_SCRIPTS}]`, 'u');\n\n// Whether text holds a script that marks no word boundaries, the predicate every store's\n// sidecar population and query split turns on.\nexport function hasUnspacedRun(text: string): boolean {\n  return HAS_RUN.test(text);\n}\n\nexport interface SearchToken {\n  text: string;\n  start: number;\n  end: number;\n  unspaced: boolean;\n}\n\n// Search tokens share the native separator contract: letters, marks, and numbers form tokens;\n// underscore, apostrophe, and hyphen separate them. Unspaced-script runs stay whole so callers\n// can apply substring semantics, while adjacent Latin text still gets its own token.\nexport function searchTokens(text: string): SearchToken[] {\n  const tokens: SearchToken[] = [];\n  for (const match of text.matchAll(/[\\p{L}\\p{N}][\\p{L}\\p{M}\\p{N}]*/gu)) {\n    const start = match.index;\n    const end = start + match[0].length;\n    let cursor = start;\n    for (const run of text.slice(start, end).matchAll(RUN)) {\n      const runStart = start + run.index;\n      if (runStart > cursor) tokens.push({ text: text.slice(cursor, runStart), start: cursor, end: runStart, unspaced: false });\n      const runEnd = runStart + run[0].length;\n      tokens.push({ text: run[0], start: runStart, end: runEnd, unspaced: true });\n      cursor = runEnd;\n    }\n    if (cursor < end) tokens.push({ text: text.slice(cursor, end), start: cursor, end, unspaced: false });\n  }\n  return tokens;\n}\n\n// Phrase membership is checked after native ranking. A CJK token matches inside one authored\n// run; Latin tokens compare folded Porter stems. Each call scans one field once per candidate.\nexport function matchesSearchPhrase(text: string, phrase: SearchToken[]): boolean {\n  const haystack = searchTokens(text);\n  if (phrase.length === 0 || haystack.length < phrase.length) return false;\n  for (let start = 0; start <= haystack.length - phrase.length; start++) {\n    let matches = true;\n    for (let i = 0; i < phrase.length; i++) {\n      const query = phrase[i];\n      const authored = haystack[start + i];\n      if (query.unspaced) {\n        if (!authored.unspaced || !authored.text.includes(query.text)) matches = false;\n      } else if (authored.unspaced || stemmer(foldForSearch(authored.text)) !== stemmer(foldForSearch(query.text))) {\n        matches = false;\n      }\n      if (!matches) break;\n    }\n    if (matches) return true;\n  }\n  return false;\n}\n\n// Character spans of unspaced-script runs in text, the same boundary the index segments on --\n// the snippet marker uses this to keep substring semantics inside a run and word/stem matching outside it.\nexport function unspacedRuns(text: string): Array<{ start: number; end: number }> {\n  return Array.from(text.matchAll(RUN), (m) => ({ start: m.index, end: m.index + m[0].length }));\n}\n// Grapheme clusters, ECMA-402/UAX #29: base char plus its marks, ZWJ sequences, Hangul jamo.\n// Built on first use and kept: construction is 6.6 ms, and a tree with no unspaced-script run\n// never segments at all, so no command pays it at module load.\nlet graphemeSegmenter: Intl.Segmenter | undefined;\n// unicode61 drops Unicode punctuation as a separator. Some of it (。、「」) has Script_Extensions\n// into an unspaced script, so RUN keeps it -- a grapheme matching this becomes a split point.\nconst PUNCTUATION = /\\p{P}/u;\n// Token barrier between separate runs (or a punctuation split within one) so their graphemes\n// are never phrase-adjacent. U+A7F7: a letter (so FTS5 keeps it as a token) no one types.\nconst BARRIER = 'ꟷ';\n\nfunction graphemes(run: string): string[] {\n  graphemeSegmenter ??= new Intl.Segmenter(undefined, { granularity: 'grapheme' });\n  return Array.from(graphemeSegmenter.segment(run), (s) => s.segment);\n}\n\n// A run's graphemes, cut into punctuation-free groups at every punctuation grapheme (dropped,\n// matching unicode61). The one place punctuation is classified, so index and query agree.\nfunction splitOnPunctuation(run: string): string[][] {\n  const groups: string[][] = [[]];\n  for (const g of graphemes(run)) {\n    if (PUNCTUATION.test(g)) groups.push([]);\n    else groups[groups.length - 1].push(g);\n  }\n  return groups.filter((g) => g.length > 0);\n}\n\n// Index side: '' when the field has no unspaced-script run. Otherwise each run explodes into\n// its graphemes, barrier-delimited from its neighbors and from a punctuation split within itself.\nexport function segmentField(text: string): string {\n  if (!HAS_RUN.test(text)) return '';\n  const out = text.replace(RUN, (run) => {\n    const body = splitOnPunctuation(run)\n      .map((g) => g.join(' '))\n      .join(` ${BARRIER} `);\n    return ` ${BARRIER} ${body} ${BARRIER} `;\n  });\n  return out.replace(/\\s+/g, ' ').trim();\n}\n\n// A `title:`/`summary:`/`text:` qualifier directly before a run that is about to become a\n// quoted grapheme phrase, so the rewrite can retarget it at the matching `_seg` column.\nconst QUALIFIER = /(^|[\\s(])(-?)(title|summary|text)\\s*:\\s*$/;\n\n// Raw title/summary/text drop punctuation as unicode61's token separator, so `数数` and `数。数`\n// falsely match there; only the barriered `_seg` columns are safe from it, hence this fallback target.\nconst SIDECAR_COLUMNS = '{title_seg summary_seg text_seg}:';\nconst AUTHORED_COLUMNS = '{title summary text}:';\nconst QUOTED_RUN_TOKEN = new RegExp(`\"(${RUN_BODY})\"`, 'gu');\n\ninterface PositionalSpan {\n  start: number;\n  end: number;\n  qualified: boolean;\n}\n\nconst FTS_WORD = /[\\p{L}\\p{N}_]/u;\n\nfunction isFtsWord(char: string | undefined): boolean {\n  return char !== undefined && FTS_WORD.test(char);\n}\n\nfunction quoteEnd(text: string, start: number): number {\n  for (let i = start + 1; i < text.length; i++) {\n    if (text[i] !== '\"') continue;\n    if (text[i + 1] === '\"') {\n      i++;\n      continue;\n    }\n    return i + 1;\n  }\n  return text.length;\n}\n\nfunction nearEnd(text: string, open: number): number {\n  let depth = 1;\n  let quoted = false;\n  for (let i = open + 1; i < text.length; i++) {\n    const char = text[i];\n    if (char === '\"') {\n      if (quoted && text[i + 1] === '\"') {\n        i++;\n      } else {\n        quoted = !quoted;\n      }\n    } else if (!quoted && char === '(') {\n      depth++;\n    } else if (!quoted && char === ')' && --depth === 0) {\n      return i + 1;\n    }\n  }\n  return text.length;\n}\n\n// Lexical routing only: SQLite remains responsible for validating the expression. The scanner\n// recognizes NEAR groups and the one FTS string governed by ^, while quoted text shields both.\nfunction positionalSpans(terms: string): PositionalSpan[] {\n  const spans: PositionalSpan[] = [];\n  for (let i = 0; i < terms.length; i++) {\n    if (terms[i] === '\"') {\n      i = quoteEnd(terms, i) - 1;\n      continue;\n    }\n    if (terms.startsWith('NEAR', i) && !isFtsWord(terms[i - 1])) {\n      let open = i + 4;\n      while (/\\s/u.test(terms[open] ?? '')) open++;\n      if (terms[open] === '(') {\n        const end = nearEnd(terms, open);\n        if (HAS_RUN.test(terms.slice(open + 1, end))) spans.push({ start: i, end, qualified: QUALIFIER.test(terms.slice(0, i)) });\n        i = end - 1;\n        continue;\n      }\n    }\n    if (terms[i] === '^' && !isFtsWord(terms[i - 1])) {\n      let operand = i + 1;\n      while (/\\s/u.test(terms[operand] ?? '')) operand++;\n      const end =\n        terms[operand] === '\"'\n          ? quoteEnd(terms, operand)\n          : (() => {\n              let cursor = operand;\n              while (/[\\p{L}\\p{N}_]/u.test(terms[cursor] ?? '')) cursor++;\n              return cursor;\n            })();\n      if (end > operand && HAS_RUN.test(terms.slice(operand, end))) spans.push({ start: i, end, qualified: QUALIFIER.test(terms.slice(0, i)) });\n      i = Math.max(i, end - 1);\n    }\n  }\n  return spans;\n}\n\nfunction routePositionalSpans(terms: string): string {\n  const spans = positionalSpans(terms);\n  if (spans.length === 0) return terms;\n  let out = '';\n  let cursor = 0;\n  for (const span of spans) {\n    out += terms.slice(cursor, span.start);\n    if (!span.qualified) out += AUTHORED_COLUMNS;\n    out += terms.slice(span.start, span.end);\n    cursor = span.end;\n  }\n  return out + terms.slice(cursor);\n}\n\nfunction inPositionalSpan(start: number, spans: PositionalSpan[]): boolean {\n  return spans.some((span) => start >= span.start && start < span.end);\n}\n\n// A run's punctuation-free groups, each its own quoted phrase (bare token if one grapheme),\n// space-joined -- the same split points segmentField barriers, so query and index agree.\nfunction runQuery(run: string, columnPrefix: string, quoteSingle = false): string {\n  return splitOnPunctuation(run)\n    .map((g) => `${columnPrefix}${g.length > 1 || quoteSingle ? `\"${g.join(' ')}\"` : g[0]}`)\n    .join(' ');\n}\n\n// Retarget complete quoted run tokens without interpreting the expression around them. Positional\n// spans have already been routed to authored columns, so sidecar barriers cannot change meaning.\nfunction quotedRunTokens(terms: string): string {\n  const spans = positionalSpans(terms);\n  let out = '';\n  let cursor = 0;\n  for (const match of terms.matchAll(QUOTED_RUN_TOKEN)) {\n    const start = match.index;\n    const run = match[1];\n    if (inPositionalSpan(start, spans) || PUNCTUATION.test(run)) continue;\n    out += terms.slice(cursor, start);\n    const qualifier = out.match(QUALIFIER);\n    if (qualifier) {\n      out = `${out.slice(0, qualifier.index)}${qualifier[1]}${runQuery(run, `${qualifier[2]}${qualifier[3]}_seg:`, true)}`;\n    } else {\n      out += runQuery(run, SIDECAR_COLUMNS, true);\n    }\n    cursor = start + match[0].length;\n  }\n  return cursor === 0 ? terms : out + terms.slice(cursor);\n}\n\n// Query side: each unspaced run becomes phrases of its graphemes, matching segmentField. A\n// qualifier maps to its `_seg` column; unqualified maps to all three (SIDECAR_COLUMNS).\nexport function segmentMatch(terms: string): string {\n  if (!HAS_RUN.test(terms)) return terms;\n  terms = routePositionalSpans(terms);\n  terms = quotedRunTokens(terms);\n  const spans = positionalSpans(terms);\n  let out = '';\n  let quoted = false;\n  let offset = 0;\n  const pieces = terms.split(RUN); // split keeps captured runs at odd indices\n  for (let i = 0; i < pieces.length; i++) {\n    if (i % 2 === 0) {\n      for (const ch of pieces[i]) if (ch === '\"') quoted = !quoted;\n      out += pieces[i];\n      offset += pieces[i].length;\n      continue;\n    }\n    const protectedRun = inPositionalSpan(offset, spans);\n    if (quoted || protectedRun) {\n      out += pieces[i]; // an author's phrase is matched as written\n      offset += pieces[i].length;\n      continue;\n    }\n    const m = out.match(QUALIFIER);\n    if (m) {\n      out = `${out.slice(0, m.index)}${m[1]}${runQuery(pieces[i], `${m[2]}${m[3]}_seg:`)}`;\n    } else {\n      out += runQuery(pieces[i], SIDECAR_COLUMNS);\n    }\n    offset += pieces[i].length;\n  }\n  return out;\n}\n\n// FTS tokenizers compare letters without combining marks. Keep the authored string for output and\n// offsets, but use this key when JS needs to compare a matched word to its source segment.\nexport function foldForSearch(text: string): string {\n  return text.normalize('NFD').replace(/\\p{M}/gu, '').toLowerCase();\n}\n"],"names":["UNSPACED_SCRIPTS","foldForSearch","hasUnspacedRun","matchesSearchPhrase","searchTokens","segmentField","segmentMatch","unspacedRuns","RUN_BODY","RUN","RegExp","HAS_RUN","text","test","tokens","matchAll","match","start","index","end","length","cursor","slice","run","runStart","push","unspaced","runEnd","phrase","haystack","matches","i","query","authored","includes","stemmer","Array","from","m","graphemeSegmenter","PUNCTUATION","BARRIER","graphemes","Intl","Segmenter","undefined","granularity","segment","s","splitOnPunctuation","groups","g","filter","out","replace","body","map","join","trim","QUALIFIER","SIDECAR_COLUMNS","AUTHORED_COLUMNS","QUOTED_RUN_TOKEN","FTS_WORD","isFtsWord","char","quoteEnd","nearEnd","open","depth","quoted","positionalSpans","terms","startsWith","spans","qualified","operand","Math","max","routePositionalSpans","span","inPositionalSpan","some","runQuery","columnPrefix","quoteSingle","quotedRunTokens","qualifier","offset","pieces","split","ch","protectedRun","normalize","toLowerCase"],"mappings":";;;;;;;;;;;QAOaA;eAAAA;;QA4RGC;eAAAA;;QAnRAC;eAAAA;;QAkCAC;eAAAA;;QApBAC;eAAAA;;QA0EAC;eAAAA;;QAwJAC;eAAAA;;QAxLAC;eAAAA;;;uBAxEQ;AAOjB,IAAMP,mBAAmB;AAChC,2FAA2F;AAC3F,gDAAgD;AAChD,IAAMQ,WAAW,AAAC,OAAuB,OAAjBR,kBAAiB;AACzC,IAAMS,MAAM,IAAIC,OAAO,AAAC,IAAY,OAATF,UAAS,MAAI;AACxC,IAAMG,UAAU,IAAID,OAAO,AAAC,IAAoB,OAAjBV,kBAAiB,MAAI;AAI7C,SAASE,eAAeU,IAAY;IACzC,OAAOD,QAAQE,IAAI,CAACD;AACtB;AAYO,SAASR,aAAaQ,IAAY;IACvC,IAAME,SAAwB,EAAE;QAC3B,kCAAA,2BAAA;;QAAL,QAAK,YAAeF,KAAKG,QAAQ,CAAC,oo+BAA7B,SAAA,6BAAA,QAAA,yBAAA,iCAAkE;YAAlE,IAAMC,QAAN;YACH,IAAMC,QAAQD,MAAME,KAAK;YACzB,IAAMC,MAAMF,QAAQD,KAAK,CAAC,EAAE,CAACI,MAAM;YACnC,IAAIC,SAASJ;gBACR,mCAAA,4BAAA;;gBAAL,QAAK,aAAaL,KAAKU,KAAK,CAACL,OAAOE,KAAKJ,QAAQ,CAACN,yBAA7C,UAAA,8BAAA,SAAA,0BAAA,kCAAmD;oBAAnD,IAAMc,MAAN;oBACH,IAAMC,WAAWP,QAAQM,IAAIL,KAAK;oBAClC,IAAIM,WAAWH,QAAQP,OAAOW,IAAI,CAAC;wBAAEb,MAAMA,KAAKU,KAAK,CAACD,QAAQG;wBAAWP,OAAOI;wBAAQF,KAAKK;wBAAUE,UAAU;oBAAM;oBACvH,IAAMC,SAASH,WAAWD,GAAG,CAAC,EAAE,CAACH,MAAM;oBACvCN,OAAOW,IAAI,CAAC;wBAAEb,MAAMW,GAAG,CAAC,EAAE;wBAAEN,OAAOO;wBAAUL,KAAKQ;wBAAQD,UAAU;oBAAK;oBACzEL,SAASM;gBACX;;gBANK;gBAAA;;;yBAAA,8BAAA;wBAAA;;;wBAAA;8BAAA;;;;YAOL,IAAIN,SAASF,KAAKL,OAAOW,IAAI,CAAC;gBAAEb,MAAMA,KAAKU,KAAK,CAACD,QAAQF;gBAAMF,OAAOI;gBAAQF,KAAAA;gBAAKO,UAAU;YAAM;QACrG;;QAZK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAaL,OAAOZ;AACT;AAIO,SAASX,oBAAoBS,IAAY,EAAEgB,MAAqB;IACrE,IAAMC,WAAWzB,aAAaQ;IAC9B,IAAIgB,OAAOR,MAAM,KAAK,KAAKS,SAAST,MAAM,GAAGQ,OAAOR,MAAM,EAAE,OAAO;IACnE,IAAK,IAAIH,QAAQ,GAAGA,SAASY,SAAST,MAAM,GAAGQ,OAAOR,MAAM,EAAEH,QAAS;QACrE,IAAIa,UAAU;QACd,IAAK,IAAIC,IAAI,GAAGA,IAAIH,OAAOR,MAAM,EAAEW,IAAK;YACtC,IAAMC,QAAQJ,MAAM,CAACG,EAAE;YACvB,IAAME,WAAWJ,QAAQ,CAACZ,QAAQc,EAAE;YACpC,IAAIC,MAAMN,QAAQ,EAAE;gBAClB,IAAI,CAACO,SAASP,QAAQ,IAAI,CAACO,SAASrB,IAAI,CAACsB,QAAQ,CAACF,MAAMpB,IAAI,GAAGkB,UAAU;YAC3E,OAAO,IAAIG,SAASP,QAAQ,IAAIS,IAAAA,gBAAO,EAAClC,cAAcgC,SAASrB,IAAI,OAAOuB,IAAAA,gBAAO,EAAClC,cAAc+B,MAAMpB,IAAI,IAAI;gBAC5GkB,UAAU;YACZ;YACA,IAAI,CAACA,SAAS;QAChB;QACA,IAAIA,SAAS,OAAO;IACtB;IACA,OAAO;AACT;AAIO,SAASvB,aAAaK,IAAY;IACvC,OAAOwB,MAAMC,IAAI,CAACzB,KAAKG,QAAQ,CAACN,MAAM,SAAC6B;eAAO;YAAErB,OAAOqB,EAAEpB,KAAK;YAAEC,KAAKmB,EAAEpB,KAAK,GAAGoB,CAAC,CAAC,EAAE,CAAClB,MAAM;QAAC;;AAC7F;AACA,6FAA6F;AAC7F,8FAA8F;AAC9F,+DAA+D;AAC/D,IAAImB;AACJ,8FAA8F;AAC9F,8FAA8F;AAC9F,IAAMC,cAAc;AACpB,6FAA6F;AAC7F,0FAA0F;AAC1F,IAAMC,UAAU;AAEhB,SAASC,UAAUnB,GAAW;IAC5BgB,8BAAAA,+BAAAA,oBAAAA,oBAAsB,IAAII,KAAKC,SAAS,CAACC,WAAW;QAAEC,aAAa;IAAW;IAC9E,OAAOV,MAAMC,IAAI,CAACE,kBAAkBQ,OAAO,CAACxB,MAAM,SAACyB;eAAMA,EAAED,OAAO;;AACpE;AAEA,8FAA8F;AAC9F,0FAA0F;AAC1F,SAASE,mBAAmB1B,GAAW;IACrC,IAAM2B,SAAqB;QAAC,EAAE;KAAC;QAC1B,kCAAA,2BAAA;;QAAL,QAAK,YAAWR,UAAUnB,yBAArB,SAAA,6BAAA,QAAA,yBAAA,iCAA2B;YAA3B,IAAM4B,IAAN;YACH,IAAIX,YAAY3B,IAAI,CAACsC,IAAID,OAAOzB,IAAI,CAAC,EAAE;iBAClCyB,MAAM,CAACA,OAAO9B,MAAM,GAAG,EAAE,CAACK,IAAI,CAAC0B;QACtC;;QAHK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAIL,OAAOD,OAAOE,MAAM,CAAC,SAACD;eAAMA,EAAE/B,MAAM,GAAG;;AACzC;AAIO,SAASf,aAAaO,IAAY;IACvC,IAAI,CAACD,QAAQE,IAAI,CAACD,OAAO,OAAO;IAChC,IAAMyC,MAAMzC,KAAK0C,OAAO,CAAC7C,KAAK,SAACc;QAC7B,IAAMgC,OAAON,mBAAmB1B,KAC7BiC,GAAG,CAAC,SAACL;mBAAMA,EAAEM,IAAI,CAAC;WAClBA,IAAI,CAAC,AAAC,IAAW,OAARhB,SAAQ;QACpB,OAAO,AAAC,IAAcc,OAAXd,SAAQ,KAAWA,OAARc,MAAK,KAAW,OAARd,SAAQ;IACxC;IACA,OAAOY,IAAIC,OAAO,CAAC,QAAQ,KAAKI,IAAI;AACtC;AAEA,0FAA0F;AAC1F,wFAAwF;AACxF,IAAMC,YAAY;AAElB,4FAA4F;AAC5F,uGAAuG;AACvG,IAAMC,kBAAkB;AACxB,IAAMC,mBAAmB;AACzB,IAAMC,mBAAmB,IAAIpD,OAAO,AAAC,KAAa,OAATF,UAAS,OAAK;AAQvD,IAAMuD,WAAW;AAEjB,SAASC,UAAUC,IAAwB;IACzC,OAAOA,SAASpB,aAAakB,SAASlD,IAAI,CAACoD;AAC7C;AAEA,SAASC,SAAStD,IAAY,EAAEK,KAAa;IAC3C,IAAK,IAAIc,IAAId,QAAQ,GAAGc,IAAInB,KAAKQ,MAAM,EAAEW,IAAK;QAC5C,IAAInB,IAAI,CAACmB,EAAE,KAAK,KAAK;QACrB,IAAInB,IAAI,CAACmB,IAAI,EAAE,KAAK,KAAK;YACvBA;YACA;QACF;QACA,OAAOA,IAAI;IACb;IACA,OAAOnB,KAAKQ,MAAM;AACpB;AAEA,SAAS+C,QAAQvD,IAAY,EAAEwD,IAAY;IACzC,IAAIC,QAAQ;IACZ,IAAIC,SAAS;IACb,IAAK,IAAIvC,IAAIqC,OAAO,GAAGrC,IAAInB,KAAKQ,MAAM,EAAEW,IAAK;QAC3C,IAAMkC,OAAOrD,IAAI,CAACmB,EAAE;QACpB,IAAIkC,SAAS,KAAK;YAChB,IAAIK,UAAU1D,IAAI,CAACmB,IAAI,EAAE,KAAK,KAAK;gBACjCA;YACF,OAAO;gBACLuC,SAAS,CAACA;YACZ;QACF,OAAO,IAAI,CAACA,UAAUL,SAAS,KAAK;YAClCI;QACF,OAAO,IAAI,CAACC,UAAUL,SAAS,OAAO,EAAEI,UAAU,GAAG;YACnD,OAAOtC,IAAI;QACb;IACF;IACA,OAAOnB,KAAKQ,MAAM;AACpB;AAEA,8FAA8F;AAC9F,+FAA+F;AAC/F,SAASmD,gBAAgBC,KAAa;;QAGlC,IAAIA,KAAK,CAACzC,GAAE,KAAK,KAAK;YACpBA,KAAImC,SAASM,OAAOzC,MAAK;YACzB,eAAA;QACF;QACA,IAAIyC,MAAMC,UAAU,CAAC,QAAQ1C,OAAM,CAACiC,UAAUQ,KAAK,CAACzC,KAAI,EAAE,GAAG;gBAEzCyC;YADlB,IAAIJ,OAAOrC,KAAI;YACf,MAAO,mBAAMlB,IAAI,EAAC2D,cAAAA,KAAK,CAACJ,KAAK,cAAXI,yBAAAA,cAAe,IAAKJ;YACtC,IAAII,KAAK,CAACJ,KAAK,KAAK,KAAK;gBACvB,IAAMjD,MAAMgD,QAAQK,OAAOJ;gBAC3B,IAAIzD,QAAQE,IAAI,CAAC2D,MAAMlD,KAAK,CAAC8C,OAAO,GAAGjD,OAAOuD,MAAMjD,IAAI,CAAC;oBAAER,OAAOc;oBAAGZ,KAAAA;oBAAKwD,WAAWhB,UAAU9C,IAAI,CAAC2D,MAAMlD,KAAK,CAAC,GAAGS;gBAAI;gBACvHA,KAAIZ,MAAM;gBACV,eAAA;YACF;QACF;QACA,IAAIqD,KAAK,CAACzC,GAAE,KAAK,OAAO,CAACiC,UAAUQ,KAAK,CAACzC,KAAI,EAAE,GAAG;gBAE9ByC;YADlB,IAAII,UAAU7C,KAAI;YAClB,MAAO,mBAAMlB,IAAI,EAAC2D,iBAAAA,KAAK,CAACI,QAAQ,cAAdJ,4BAAAA,iBAAkB,IAAKI;YACzC,IAAMzD,OACJqD,KAAK,CAACI,QAAQ,KAAK,MACfV,SAASM,OAAOI,WAChB,AAAC;oBAE8BJ;gBAD7B,IAAInD,SAASuD;gBACb,MAAO,4maAAiB/D,IAAI,EAAC2D,gBAAAA,KAAK,CAACnD,OAAO,cAAbmD,2BAAAA,gBAAiB,IAAKnD;gBACnD,OAAOA;YACT;YACN,IAAIF,OAAMyD,WAAWjE,QAAQE,IAAI,CAAC2D,MAAMlD,KAAK,CAACsD,SAASzD,QAAOuD,MAAMjD,IAAI,CAAC;gBAAER,OAAOc;gBAAGZ,KAAAA;gBAAKwD,WAAWhB,UAAU9C,IAAI,CAAC2D,MAAMlD,KAAK,CAAC,GAAGS;YAAI;YACvIA,KAAI8C,KAAKC,GAAG,CAAC/C,IAAGZ,OAAM;QACxB;;IACF;IA9BA,IAAMuD,QAA0B,EAAE;IAClC,IAAK,IAAI3C,IAAI,GAAGA,IAAIyC,MAAMpD,MAAM,EAAEW;IA8BlC,OAAO2C;AACT;AAEA,SAASK,qBAAqBP,KAAa;IACzC,IAAME,QAAQH,gBAAgBC;IAC9B,IAAIE,MAAMtD,MAAM,KAAK,GAAG,OAAOoD;IAC/B,IAAInB,MAAM;IACV,IAAIhC,SAAS;QACR,kCAAA,2BAAA;;QAAL,QAAK,YAAcqD,0BAAd,SAAA,6BAAA,QAAA,yBAAA,iCAAqB;YAArB,IAAMM,OAAN;YACH3B,OAAOmB,MAAMlD,KAAK,CAACD,QAAQ2D,KAAK/D,KAAK;YACrC,IAAI,CAAC+D,KAAKL,SAAS,EAAEtB,OAAOQ;YAC5BR,OAAOmB,MAAMlD,KAAK,CAAC0D,KAAK/D,KAAK,EAAE+D,KAAK7D,GAAG;YACvCE,SAAS2D,KAAK7D,GAAG;QACnB;;QALK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAML,OAAOkC,MAAMmB,MAAMlD,KAAK,CAACD;AAC3B;AAEA,SAAS4D,iBAAiBhE,KAAa,EAAEyD,KAAuB;IAC9D,OAAOA,MAAMQ,IAAI,CAAC,SAACF;eAAS/D,SAAS+D,KAAK/D,KAAK,IAAIA,QAAQ+D,KAAK7D,GAAG;;AACrE;AAEA,4FAA4F;AAC5F,yFAAyF;AACzF,SAASgE,SAAS5D,GAAW,EAAE6D,YAAoB;QAAEC,cAAAA,iEAAc;IACjE,OAAOpC,mBAAmB1B,KACvBiC,GAAG,CAAC,SAACL;eAAM,AAAC,GAAiBA,OAAfiC,cAAuE,OAAxDjC,EAAE/B,MAAM,GAAG,KAAKiE,cAAc,AAAC,IAAe,OAAZlC,EAAEM,IAAI,CAAC,MAAK,OAAKN,CAAC,CAAC,EAAE;OACpFM,IAAI,CAAC;AACV;AAEA,kGAAkG;AAClG,iGAAiG;AACjG,SAAS6B,gBAAgBd,KAAa;IACpC,IAAME,QAAQH,gBAAgBC;IAC9B,IAAInB,MAAM;IACV,IAAIhC,SAAS;QACR,kCAAA,2BAAA;;QAAL,QAAK,YAAemD,MAAMzD,QAAQ,CAAC+C,sCAA9B,SAAA,6BAAA,QAAA,yBAAA,iCAAiD;YAAjD,IAAM9C,QAAN;YACH,IAAMC,QAAQD,MAAME,KAAK;YACzB,IAAMK,MAAMP,KAAK,CAAC,EAAE;YACpB,IAAIiE,iBAAiBhE,OAAOyD,UAAUlC,YAAY3B,IAAI,CAACU,MAAM;YAC7D8B,OAAOmB,MAAMlD,KAAK,CAACD,QAAQJ;YAC3B,IAAMsE,YAAYlC,IAAIrC,KAAK,CAAC2C;YAC5B,IAAI4B,WAAW;gBACblC,MAAM,AAAC,GAAkCkC,OAAhClC,IAAI/B,KAAK,CAAC,GAAGiE,UAAUrE,KAAK,GAAmBiE,OAAfI,SAAS,CAAC,EAAE,EAA8D,OAA3DJ,SAAS5D,KAAK,AAAC,GAAiBgE,OAAfA,SAAS,CAAC,EAAE,EAAgB,OAAbA,SAAS,CAAC,EAAE,EAAC,UAAQ;YAC/G,OAAO;gBACLlC,OAAO8B,SAAS5D,KAAKqC,iBAAiB;YACxC;YACAvC,SAASJ,QAAQD,KAAK,CAAC,EAAE,CAACI,MAAM;QAClC;;QAZK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAaL,OAAOC,WAAW,IAAImD,QAAQnB,MAAMmB,MAAMlD,KAAK,CAACD;AAClD;AAIO,SAASf,aAAakE,KAAa;IACxC,IAAI,CAAC7D,QAAQE,IAAI,CAAC2D,QAAQ,OAAOA;IACjCA,QAAQO,qBAAqBP;IAC7BA,QAAQc,gBAAgBd;IACxB,IAAME,QAAQH,gBAAgBC;IAC9B,IAAInB,MAAM;IACV,IAAIiB,SAAS;IACb,IAAIkB,SAAS;IACb,IAAMC,SAASjB,MAAMkB,KAAK,CAACjF,MAAM,2CAA2C;IAC5E,IAAK,IAAIsB,IAAI,GAAGA,IAAI0D,OAAOrE,MAAM,EAAEW,IAAK;QACtC,IAAIA,IAAI,MAAM,GAAG;gBACV,kCAAA,2BAAA;;gBAAL,QAAK,YAAY0D,MAAM,CAAC1D,EAAE,qBAArB,SAAA,6BAAA,QAAA,yBAAA;oBAAA,IAAM4D,KAAN;oBAAuB,IAAIA,OAAO,KAAKrB,SAAS,CAACA;;;gBAAjD;gBAAA;;;yBAAA,6BAAA;wBAAA;;;wBAAA;8BAAA;;;;YACLjB,OAAOoC,MAAM,CAAC1D,EAAE;YAChByD,UAAUC,MAAM,CAAC1D,EAAE,CAACX,MAAM;YAC1B;QACF;QACA,IAAMwE,eAAeX,iBAAiBO,QAAQd;QAC9C,IAAIJ,UAAUsB,cAAc;YAC1BvC,OAAOoC,MAAM,CAAC1D,EAAE,EAAE,2CAA2C;YAC7DyD,UAAUC,MAAM,CAAC1D,EAAE,CAACX,MAAM;YAC1B;QACF;QACA,IAAMkB,IAAIe,IAAIrC,KAAK,CAAC2C;QACpB,IAAIrB,GAAG;YACLe,MAAM,AAAC,GAA0Bf,OAAxBe,IAAI/B,KAAK,CAAC,GAAGgB,EAAEpB,KAAK,GAAWiE,OAAP7C,CAAC,CAAC,EAAE,EAA8C,OAA3C6C,SAASM,MAAM,CAAC1D,EAAE,EAAE,AAAC,GAASO,OAAPA,CAAC,CAAC,EAAE,EAAQ,OAALA,CAAC,CAAC,EAAE,EAAC;QAC7E,OAAO;YACLe,OAAO8B,SAASM,MAAM,CAAC1D,EAAE,EAAE6B;QAC7B;QACA4B,UAAUC,MAAM,CAAC1D,EAAE,CAACX,MAAM;IAC5B;IACA,OAAOiC;AACT;AAIO,SAASpD,cAAcW,IAAY;IACxC,OAAOA,KAAKiF,SAAS,CAAC,OAAOvC,OAAO,CAAC,27JAAW,IAAIwC,WAAW;AACjE"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/text/segment.ts"],"sourcesContent":["import { stemmer } from 'stemmer';\n\n// Word boundaries FTS5's tokenizers cannot find on their own. Contract: seg(query) must appear in\n// seg(document) for any substring query; word mode is context-dependent and breaks that, so grapheme mode (UAX #29) is used.\n\n// Scripts written without word spaces: a closed set of writing systems.\n// Script_Extensions, not Script: the katakana long-vowel mark ー is Script=Common.\nexport const UNSPACED_SCRIPTS = '\\\\p{scx=Han}\\\\p{scx=Hiragana}\\\\p{scx=Katakana}\\\\p{scx=Thai}\\\\p{scx=Khmer}\\\\p{scx=Lao}\\\\p{scx=Myanmar}';\n// A run is script BASE characters with their combining marks attached; a bare mark after a\n// Latin letter (decomposed é) never starts one.\nconst RUN_BODY = `(?:[${UNSPACED_SCRIPTS}]\\\\p{M}*)+`;\nconst RUN = new RegExp(`(${RUN_BODY})`, 'gu');\nconst HAS_RUN = new RegExp(`[${UNSPACED_SCRIPTS}]`, 'u');\n\n// Whether text holds a script that marks no word boundaries, the predicate every store's\n// sidecar population and query split turns on.\nexport function hasUnspacedRun(text: string): boolean {\n  return HAS_RUN.test(text);\n}\n\nexport interface SearchToken {\n  text: string;\n  start: number;\n  end: number;\n  unspaced: boolean;\n}\n\n// Search tokens share the native separator contract: letters, marks, and numbers form tokens;\n// underscore, apostrophe, and hyphen separate them. Unspaced-script runs stay whole so callers\n// can apply substring semantics, while adjacent Latin text still gets its own token.\nexport function searchTokens(text: string): SearchToken[] {\n  const tokens: SearchToken[] = [];\n  for (const match of text.matchAll(/[\\p{L}\\p{N}][\\p{L}\\p{M}\\p{N}]*/gu)) {\n    const start = match.index;\n    const end = start + match[0].length;\n    let cursor = start;\n    for (const run of text.slice(start, end).matchAll(RUN)) {\n      const runStart = start + run.index;\n      if (runStart > cursor) tokens.push({ text: text.slice(cursor, runStart), start: cursor, end: runStart, unspaced: false });\n      const runEnd = runStart + run[0].length;\n      tokens.push({ text: run[0], start: runStart, end: runEnd, unspaced: true });\n      cursor = runEnd;\n    }\n    if (cursor < end) tokens.push({ text: text.slice(cursor, end), start: cursor, end, unspaced: false });\n  }\n  return tokens;\n}\n\n// Phrase membership is checked after native ranking. A CJK token matches inside one authored\n// run; Latin tokens compare folded Porter stems. Each call scans one field once per candidate.\nexport function matchesSearchPhrase(text: string, phrase: SearchToken[]): boolean {\n  const haystack = searchTokens(text);\n  if (phrase.length === 0 || haystack.length < phrase.length) return false;\n  for (let start = 0; start <= haystack.length - phrase.length; start++) {\n    let matches = true;\n    for (let i = 0; i < phrase.length; i++) {\n      const query = phrase[i];\n      const authored = haystack[start + i];\n      if (query.unspaced) {\n        if (!authored.unspaced || !authored.text.includes(query.text)) matches = false;\n      } else if (authored.unspaced || stemmer(foldForSearch(authored.text)) !== stemmer(foldForSearch(query.text))) {\n        matches = false;\n      }\n      if (!matches) break;\n    }\n    if (matches) return true;\n  }\n  return false;\n}\n\n// Character spans of unspaced-script runs in text, the same boundary the index segments on --\n// the snippet marker uses this to keep substring semantics inside a run and word/stem matching outside it.\nexport function unspacedRuns(text: string): Array<{ start: number; end: number }> {\n  return Array.from(text.matchAll(RUN), (m) => ({ start: m.index, end: m.index + m[0].length }));\n}\n// Grapheme clusters, ECMA-402/UAX #29: base char plus its marks, ZWJ sequences, Hangul jamo.\n// Built on first use and kept: construction is 6.6 ms, and a tree with no unspaced-script run\n// never segments at all, so no command pays it at module load.\nlet graphemeSegmenter: Intl.Segmenter | undefined;\n// unicode61 drops Unicode punctuation as a separator. Some of it (。、「」) has Script_Extensions\n// into an unspaced script, so RUN keeps it -- a grapheme matching this becomes a split point.\nconst PUNCTUATION = /\\p{P}/u;\n// Token barrier between separate runs (or a punctuation split within one) so their graphemes\n// are never phrase-adjacent. U+A7F7: a letter (so FTS5 keeps it as a token) no one types.\nconst BARRIER = 'ꟷ';\n\nfunction graphemes(run: string): string[] {\n  graphemeSegmenter ??= new Intl.Segmenter(undefined, { granularity: 'grapheme' });\n  return Array.from(graphemeSegmenter.segment(run), (s) => s.segment);\n}\n\n// A run's graphemes, cut into punctuation-free groups at every punctuation grapheme (dropped,\n// matching unicode61). The one place punctuation is classified, so index and query agree.\nfunction splitOnPunctuation(run: string): string[][] {\n  const groups: string[][] = [[]];\n  for (const g of graphemes(run)) {\n    if (PUNCTUATION.test(g)) groups.push([]);\n    else groups[groups.length - 1].push(g);\n  }\n  return groups.filter((g) => g.length > 0);\n}\n\n// Index side: '' when the field has no unspaced-script run. Otherwise each run explodes into\n// its graphemes, barrier-delimited from its neighbors and from a punctuation split within itself.\nexport function segmentField(text: string): string {\n  if (!HAS_RUN.test(text)) return '';\n  const out = text.replace(RUN, (run) => {\n    const body = splitOnPunctuation(run)\n      .map((g) => g.join(' '))\n      .join(` ${BARRIER} `);\n    return ` ${BARRIER} ${body} ${BARRIER} `;\n  });\n  return out.replace(/\\s+/g, ' ').trim();\n}\n\n// A `title:`/`summary:`/`text:` qualifier directly before a run that is about to become a\n// quoted grapheme phrase, so the rewrite can retarget it at the matching `_seg` column.\nconst QUALIFIER = /(^|[\\s(])(-?)(title|summary|text)\\s*:\\s*$/;\n\n// Raw title/summary/text drop punctuation as unicode61's token separator, so `数数` and `数。数`\n// falsely match there; only the barriered `_seg` columns are safe from it, hence this fallback target.\nconst SIDECAR_COLUMNS = '{title_seg summary_seg text_seg}:';\nconst AUTHORED_COLUMNS = '{title summary text}:';\nconst QUOTED_RUN_TOKEN = new RegExp(`\"(${RUN_BODY})\"`, 'gu');\n\ninterface PositionalSpan {\n  start: number;\n  end: number;\n  qualified: boolean;\n}\n\nconst FTS_WORD = /[\\p{L}\\p{N}_]/u;\n\nfunction isFtsWord(char: string | undefined): boolean {\n  return char !== undefined && FTS_WORD.test(char);\n}\n\nfunction quoteEnd(text: string, start: number): number {\n  for (let i = start + 1; i < text.length; i++) {\n    if (text[i] !== '\"') continue;\n    if (text[i + 1] === '\"') {\n      i++;\n      continue;\n    }\n    return i + 1;\n  }\n  return text.length;\n}\n\nfunction nearEnd(text: string, open: number): number {\n  let depth = 1;\n  let quoted = false;\n  for (let i = open + 1; i < text.length; i++) {\n    const char = text[i];\n    if (char === '\"') {\n      if (quoted && text[i + 1] === '\"') {\n        i++;\n      } else {\n        quoted = !quoted;\n      }\n    } else if (!quoted && char === '(') {\n      depth++;\n    } else if (!quoted && char === ')' && --depth === 0) {\n      return i + 1;\n    }\n  }\n  return text.length;\n}\n\n// Lexical routing only: SQLite remains responsible for validating the expression. The scanner\n// recognizes NEAR groups and the one FTS string governed by ^, while quoted text shields both.\nfunction positionalSpans(terms: string): PositionalSpan[] {\n  const spans: PositionalSpan[] = [];\n  for (let i = 0; i < terms.length; i++) {\n    if (terms[i] === '\"') {\n      i = quoteEnd(terms, i) - 1;\n      continue;\n    }\n    if (terms.startsWith('NEAR', i) && !isFtsWord(terms[i - 1])) {\n      let open = i + 4;\n      while (/\\s/u.test(terms[open] ?? '')) open++;\n      if (terms[open] === '(') {\n        const end = nearEnd(terms, open);\n        if (HAS_RUN.test(terms.slice(open + 1, end))) spans.push({ start: i, end, qualified: QUALIFIER.test(terms.slice(0, i)) });\n        i = end - 1;\n        continue;\n      }\n    }\n    if (terms[i] === '^' && !isFtsWord(terms[i - 1])) {\n      let operand = i + 1;\n      while (/\\s/u.test(terms[operand] ?? '')) operand++;\n      const end =\n        terms[operand] === '\"'\n          ? quoteEnd(terms, operand)\n          : (() => {\n              let cursor = operand;\n              while (/[\\p{L}\\p{N}_]/u.test(terms[cursor] ?? '')) cursor++;\n              return cursor;\n            })();\n      if (end > operand && HAS_RUN.test(terms.slice(operand, end))) spans.push({ start: i, end, qualified: QUALIFIER.test(terms.slice(0, i)) });\n      i = Math.max(i, end - 1);\n    }\n  }\n  return spans;\n}\n\nfunction routePositionalSpans(terms: string): string {\n  const spans = positionalSpans(terms);\n  if (spans.length === 0) return terms;\n  let out = '';\n  let cursor = 0;\n  for (const span of spans) {\n    out += terms.slice(cursor, span.start);\n    if (!span.qualified) out += AUTHORED_COLUMNS;\n    out += terms.slice(span.start, span.end);\n    cursor = span.end;\n  }\n  return out + terms.slice(cursor);\n}\n\nfunction inPositionalSpan(start: number, spans: PositionalSpan[]): boolean {\n  return spans.some((span) => start >= span.start && start < span.end);\n}\n\n// A run's punctuation-free groups, each its own quoted phrase (bare token if one grapheme),\n// space-joined -- the same split points segmentField barriers, so query and index agree.\nfunction runQuery(run: string, columnPrefix: string, quoteSingle = false): string {\n  return splitOnPunctuation(run)\n    .map((g) => `${columnPrefix}${g.length > 1 || quoteSingle ? `\"${g.join(' ')}\"` : g[0]}`)\n    .join(' ');\n}\n\n// Retarget complete quoted run tokens without interpreting the expression around them. Positional\n// spans have already been routed to authored columns, so sidecar barriers cannot change meaning.\nfunction quotedRunTokens(terms: string): string {\n  const spans = positionalSpans(terms);\n  let out = '';\n  let cursor = 0;\n  for (const match of terms.matchAll(QUOTED_RUN_TOKEN)) {\n    const start = match.index;\n    const run = match[1];\n    if (inPositionalSpan(start, spans) || PUNCTUATION.test(run)) continue;\n    out += terms.slice(cursor, start);\n    const qualifier = out.match(QUALIFIER);\n    if (qualifier) {\n      out = `${out.slice(0, qualifier.index)}${qualifier[1]}${runQuery(run, `${qualifier[2]}${qualifier[3]}_seg:`, true)}`;\n    } else {\n      out += runQuery(run, SIDECAR_COLUMNS, true);\n    }\n    cursor = start + match[0].length;\n  }\n  return cursor === 0 ? terms : out + terms.slice(cursor);\n}\n\n// Query side: each unspaced run becomes phrases of its graphemes, matching segmentField. A\n// qualifier maps to its `_seg` column; unqualified maps to all three (SIDECAR_COLUMNS).\nexport function segmentMatch(terms: string): string {\n  if (!HAS_RUN.test(terms)) return terms;\n  terms = routePositionalSpans(terms);\n  terms = quotedRunTokens(terms);\n  const spans = positionalSpans(terms);\n  let out = '';\n  let quoted = false;\n  let offset = 0;\n  const pieces = terms.split(RUN); // split keeps captured runs at odd indices\n  for (let i = 0; i < pieces.length; i++) {\n    if (i % 2 === 0) {\n      for (const ch of pieces[i]) if (ch === '\"') quoted = !quoted;\n      out += pieces[i];\n      offset += pieces[i].length;\n      continue;\n    }\n    const protectedRun = inPositionalSpan(offset, spans);\n    if (quoted || protectedRun) {\n      out += pieces[i]; // an author's phrase is matched as written\n      offset += pieces[i].length;\n      continue;\n    }\n    const m = out.match(QUALIFIER);\n    if (m) {\n      out = `${out.slice(0, m.index)}${m[1]}${runQuery(pieces[i], `${m[2]}${m[3]}_seg:`)}`;\n    } else {\n      out += runQuery(pieces[i], SIDECAR_COLUMNS);\n    }\n    offset += pieces[i].length;\n  }\n  return out;\n}\n\n// FTS tokenizers compare letters without combining marks. Keep the authored string for output and\n// offsets, but use this key when JS needs to compare a matched word to its source segment.\nexport function foldForSearch(text: string): string {\n  return text.normalize('NFD').replace(/\\p{M}/gu, '').toLowerCase();\n}\n"],"names":["UNSPACED_SCRIPTS","foldForSearch","hasUnspacedRun","matchesSearchPhrase","searchTokens","segmentField","segmentMatch","unspacedRuns","RUN_BODY","RUN","RegExp","HAS_RUN","text","test","tokens","matchAll","match","start","index","end","length","cursor","slice","run","runStart","push","unspaced","runEnd","phrase","haystack","matches","i","query","authored","includes","stemmer","Array","from","m","graphemeSegmenter","PUNCTUATION","BARRIER","graphemes","Intl","Segmenter","undefined","granularity","segment","s","splitOnPunctuation","groups","g","filter","out","replace","body","map","join","trim","QUALIFIER","SIDECAR_COLUMNS","AUTHORED_COLUMNS","QUOTED_RUN_TOKEN","FTS_WORD","isFtsWord","char","quoteEnd","nearEnd","open","depth","quoted","positionalSpans","terms","startsWith","spans","qualified","operand","Math","max","routePositionalSpans","span","inPositionalSpan","some","runQuery","columnPrefix","quoteSingle","quotedRunTokens","qualifier","offset","pieces","split","ch","protectedRun","normalize","toLowerCase"],"mappings":";;;;;;;;;;;QAOaA;eAAAA;;QA4RGC;eAAAA;;QAnRAC;eAAAA;;QAkCAC;eAAAA;;QApBAC;eAAAA;;QA0EAC;eAAAA;;QAwJAC;eAAAA;;QAxLAC;eAAAA;;;uBAxEQ;AAOjB,IAAMP,mBAAmB;AAChC,2FAA2F;AAC3F,gDAAgD;AAChD,IAAMQ,WAAW,AAAC,OAAuB,OAAjBR,kBAAiB;AACzC,IAAMS,MAAM,IAAIC,OAAO,AAAC,IAAY,OAATF,UAAS,MAAI;AACxC,IAAMG,UAAU,IAAID,OAAO,AAAC,IAAoB,OAAjBV,kBAAiB,MAAI;AAI7C,SAASE,eAAeU,IAAY;IACzC,OAAOD,QAAQE,IAAI,CAACD;AACtB;AAYO,SAASR,aAAaQ,IAAY;IACvC,IAAME,SAAwB,EAAE;QAC3B,kCAAA,2BAAA;;QAAL,QAAK,YAAeF,KAAKG,QAAQ,CAAC,oo+BAA7B,SAAA,6BAAA,QAAA,yBAAA,iCAAkE;YAAlE,IAAMC,QAAN;YACH,IAAMC,QAAQD,MAAME,KAAK;YACzB,IAAMC,MAAMF,QAAQD,KAAK,CAAC,EAAE,CAACI,MAAM;YACnC,IAAIC,SAASJ;gBACR,mCAAA,4BAAA;;gBAAL,QAAK,aAAaL,KAAKU,KAAK,CAACL,OAAOE,KAAKJ,QAAQ,CAACN,yBAA7C,UAAA,8BAAA,SAAA,0BAAA,kCAAmD;oBAAnD,IAAMc,MAAN;oBACH,IAAMC,WAAWP,QAAQM,IAAIL,KAAK;oBAClC,IAAIM,WAAWH,QAAQP,OAAOW,IAAI,CAAC;wBAAEb,MAAMA,KAAKU,KAAK,CAACD,QAAQG;wBAAWP,OAAOI;wBAAQF,KAAKK;wBAAUE,UAAU;oBAAM;oBACvH,IAAMC,SAASH,WAAWD,GAAG,CAAC,EAAE,CAACH,MAAM;oBACvCN,OAAOW,IAAI,CAAC;wBAAEb,MAAMW,GAAG,CAAC,EAAE;wBAAEN,OAAOO;wBAAUL,KAAKQ;wBAAQD,UAAU;oBAAK;oBACzEL,SAASM;gBACX;;gBANK;gBAAA;;;yBAAA,8BAAA;wBAAA;;;wBAAA;8BAAA;;;;YAOL,IAAIN,SAASF,KAAKL,OAAOW,IAAI,CAAC;gBAAEb,MAAMA,KAAKU,KAAK,CAACD,QAAQF;gBAAMF,OAAOI;gBAAQF,KAAAA;gBAAKO,UAAU;YAAM;QACrG;;QAZK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAaL,OAAOZ;AACT;AAIO,SAASX,oBAAoBS,IAAY,EAAEgB,MAAqB;IACrE,IAAMC,WAAWzB,aAAaQ;IAC9B,IAAIgB,OAAOR,MAAM,KAAK,KAAKS,SAAST,MAAM,GAAGQ,OAAOR,MAAM,EAAE,OAAO;IACnE,IAAK,IAAIH,QAAQ,GAAGA,SAASY,SAAST,MAAM,GAAGQ,OAAOR,MAAM,EAAEH,QAAS;QACrE,IAAIa,UAAU;QACd,IAAK,IAAIC,IAAI,GAAGA,IAAIH,OAAOR,MAAM,EAAEW,IAAK;YACtC,IAAMC,QAAQJ,MAAM,CAACG,EAAE;YACvB,IAAME,WAAWJ,QAAQ,CAACZ,QAAQc,EAAE;YACpC,IAAIC,MAAMN,QAAQ,EAAE;gBAClB,IAAI,CAACO,SAASP,QAAQ,IAAI,CAACO,SAASrB,IAAI,CAACsB,QAAQ,CAACF,MAAMpB,IAAI,GAAGkB,UAAU;YAC3E,OAAO,IAAIG,SAASP,QAAQ,IAAIS,IAAAA,gBAAO,EAAClC,cAAcgC,SAASrB,IAAI,OAAOuB,IAAAA,gBAAO,EAAClC,cAAc+B,MAAMpB,IAAI,IAAI;gBAC5GkB,UAAU;YACZ;YACA,IAAI,CAACA,SAAS;QAChB;QACA,IAAIA,SAAS,OAAO;IACtB;IACA,OAAO;AACT;AAIO,SAASvB,aAAaK,IAAY;IACvC,OAAOwB,MAAMC,IAAI,CAACzB,KAAKG,QAAQ,CAACN,MAAM,SAAC6B;eAAO;YAAErB,OAAOqB,EAAEpB,KAAK;YAAEC,KAAKmB,EAAEpB,KAAK,GAAGoB,CAAC,CAAC,EAAE,CAAClB,MAAM;QAAC;;AAC7F;AACA,6FAA6F;AAC7F,8FAA8F;AAC9F,+DAA+D;AAC/D,IAAImB;AACJ,8FAA8F;AAC9F,8FAA8F;AAC9F,IAAMC,cAAc;AACpB,6FAA6F;AAC7F,0FAA0F;AAC1F,IAAMC,UAAU;AAEhB,SAASC,UAAUnB,GAAW;IAC5BgB,8BAAAA,+BAAAA,oBAAAA,oBAAsB,IAAII,KAAKC,SAAS,CAACC,WAAW;QAAEC,aAAa;IAAW;IAC9E,OAAOV,MAAMC,IAAI,CAACE,kBAAkBQ,OAAO,CAACxB,MAAM,SAACyB;eAAMA,EAAED,OAAO;;AACpE;AAEA,8FAA8F;AAC9F,0FAA0F;AAC1F,SAASE,mBAAmB1B,GAAW;IACrC,IAAM2B,SAAqB;QAAC,EAAE;KAAC;QAC1B,kCAAA,2BAAA;;QAAL,QAAK,YAAWR,UAAUnB,yBAArB,SAAA,6BAAA,QAAA,yBAAA,iCAA2B;YAA3B,IAAM4B,IAAN;YACH,IAAIX,YAAY3B,IAAI,CAACsC,IAAID,OAAOzB,IAAI,CAAC,EAAE;iBAClCyB,MAAM,CAACA,OAAO9B,MAAM,GAAG,EAAE,CAACK,IAAI,CAAC0B;QACtC;;QAHK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAIL,OAAOD,OAAOE,MAAM,CAAC,SAACD;eAAMA,EAAE/B,MAAM,GAAG;;AACzC;AAIO,SAASf,aAAaO,IAAY;IACvC,IAAI,CAACD,QAAQE,IAAI,CAACD,OAAO,OAAO;IAChC,IAAMyC,MAAMzC,KAAK0C,OAAO,CAAC7C,KAAK,SAACc;QAC7B,IAAMgC,OAAON,mBAAmB1B,KAC7BiC,GAAG,CAAC,SAACL;mBAAMA,EAAEM,IAAI,CAAC;WAClBA,IAAI,CAAC,AAAC,IAAW,OAARhB,SAAQ;QACpB,OAAO,AAAC,IAAcc,OAAXd,SAAQ,KAAWA,OAARc,MAAK,KAAW,OAARd,SAAQ;IACxC;IACA,OAAOY,IAAIC,OAAO,CAAC,QAAQ,KAAKI,IAAI;AACtC;AAEA,0FAA0F;AAC1F,wFAAwF;AACxF,IAAMC,YAAY;AAElB,4FAA4F;AAC5F,uGAAuG;AACvG,IAAMC,kBAAkB;AACxB,IAAMC,mBAAmB;AACzB,IAAMC,mBAAmB,IAAIpD,OAAO,AAAC,KAAa,OAATF,UAAS,OAAK;AAQvD,IAAMuD,WAAW;AAEjB,SAASC,UAAUC,IAAwB;IACzC,OAAOA,SAASpB,aAAakB,SAASlD,IAAI,CAACoD;AAC7C;AAEA,SAASC,SAAStD,IAAY,EAAEK,KAAa;IAC3C,IAAK,IAAIc,IAAId,QAAQ,GAAGc,IAAInB,KAAKQ,MAAM,EAAEW,IAAK;QAC5C,IAAInB,IAAI,CAACmB,EAAE,KAAK,KAAK;QACrB,IAAInB,IAAI,CAACmB,IAAI,EAAE,KAAK,KAAK;YACvBA;YACA;QACF;QACA,OAAOA,IAAI;IACb;IACA,OAAOnB,KAAKQ,MAAM;AACpB;AAEA,SAAS+C,QAAQvD,IAAY,EAAEwD,IAAY;IACzC,IAAIC,QAAQ;IACZ,IAAIC,SAAS;IACb,IAAK,IAAIvC,IAAIqC,OAAO,GAAGrC,IAAInB,KAAKQ,MAAM,EAAEW,IAAK;QAC3C,IAAMkC,OAAOrD,IAAI,CAACmB,EAAE;QACpB,IAAIkC,SAAS,KAAK;YAChB,IAAIK,UAAU1D,IAAI,CAACmB,IAAI,EAAE,KAAK,KAAK;gBACjCA;YACF,OAAO;gBACLuC,SAAS,CAACA;YACZ;QACF,OAAO,IAAI,CAACA,UAAUL,SAAS,KAAK;YAClCI;QACF,OAAO,IAAI,CAACC,UAAUL,SAAS,OAAO,EAAEI,UAAU,GAAG;YACnD,OAAOtC,IAAI;QACb;IACF;IACA,OAAOnB,KAAKQ,MAAM;AACpB;AAEA,8FAA8F;AAC9F,+FAA+F;AAC/F,SAASmD,gBAAgBC,KAAa;;QAGlC,IAAIA,KAAK,CAACzC,GAAE,KAAK,KAAK;YACpBA,KAAImC,SAASM,OAAOzC,MAAK;YACzB,eAAA;QACF;QACA,IAAIyC,MAAMC,UAAU,CAAC,QAAQ1C,OAAM,CAACiC,UAAUQ,KAAK,CAACzC,KAAI,EAAE,GAAG;gBAEzCyC;YADlB,IAAIJ,OAAOrC,KAAI;YACf,MAAO,mBAAMlB,IAAI,EAAC2D,cAAAA,KAAK,CAACJ,KAAK,cAAXI,yBAAAA,cAAe,IAAKJ;YACtC,IAAII,KAAK,CAACJ,KAAK,KAAK,KAAK;gBACvB,IAAMjD,MAAMgD,QAAQK,OAAOJ;gBAC3B,IAAIzD,QAAQE,IAAI,CAAC2D,MAAMlD,KAAK,CAAC8C,OAAO,GAAGjD,OAAOuD,MAAMjD,IAAI,CAAC;oBAAER,OAAOc;oBAAGZ,KAAAA;oBAAKwD,WAAWhB,UAAU9C,IAAI,CAAC2D,MAAMlD,KAAK,CAAC,GAAGS;gBAAI;gBACvHA,KAAIZ,MAAM;gBACV,eAAA;YACF;QACF;QACA,IAAIqD,KAAK,CAACzC,GAAE,KAAK,OAAO,CAACiC,UAAUQ,KAAK,CAACzC,KAAI,EAAE,GAAG;gBAE9ByC;YADlB,IAAII,UAAU7C,KAAI;YAClB,MAAO,mBAAMlB,IAAI,EAAC2D,iBAAAA,KAAK,CAACI,QAAQ,cAAdJ,4BAAAA,iBAAkB,IAAKI;YACzC,IAAMzD,OACJqD,KAAK,CAACI,QAAQ,KAAK,MACfV,SAASM,OAAOI,WAChB,AAAC;oBAE8BJ;gBAD7B,IAAInD,SAASuD;gBACb,MAAO,4maAAiB/D,IAAI,EAAC2D,gBAAAA,KAAK,CAACnD,OAAO,cAAbmD,2BAAAA,gBAAiB,IAAKnD;gBACnD,OAAOA;YACT;YACN,IAAIF,OAAMyD,WAAWjE,QAAQE,IAAI,CAAC2D,MAAMlD,KAAK,CAACsD,SAASzD,QAAOuD,MAAMjD,IAAI,CAAC;gBAAER,OAAOc;gBAAGZ,KAAAA;gBAAKwD,WAAWhB,UAAU9C,IAAI,CAAC2D,MAAMlD,KAAK,CAAC,GAAGS;YAAI;YACvIA,KAAI8C,KAAKC,GAAG,CAAC/C,IAAGZ,OAAM;QACxB;;IACF;IA9BA,IAAMuD,QAA0B,EAAE;IAClC,IAAK,IAAI3C,IAAI,GAAGA,IAAIyC,MAAMpD,MAAM,EAAEW;IA8BlC,OAAO2C;AACT;AAEA,SAASK,qBAAqBP,KAAa;IACzC,IAAME,QAAQH,gBAAgBC;IAC9B,IAAIE,MAAMtD,MAAM,KAAK,GAAG,OAAOoD;IAC/B,IAAInB,MAAM;IACV,IAAIhC,SAAS;QACR,kCAAA,2BAAA;;QAAL,QAAK,YAAcqD,0BAAd,SAAA,6BAAA,QAAA,yBAAA,iCAAqB;YAArB,IAAMM,OAAN;YACH3B,OAAOmB,MAAMlD,KAAK,CAACD,QAAQ2D,KAAK/D,KAAK;YACrC,IAAI,CAAC+D,KAAKL,SAAS,EAAEtB,OAAOQ;YAC5BR,OAAOmB,MAAMlD,KAAK,CAAC0D,KAAK/D,KAAK,EAAE+D,KAAK7D,GAAG;YACvCE,SAAS2D,KAAK7D,GAAG;QACnB;;QALK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAML,OAAOkC,MAAMmB,MAAMlD,KAAK,CAACD;AAC3B;AAEA,SAAS4D,iBAAiBhE,KAAa,EAAEyD,KAAuB;IAC9D,OAAOA,MAAMQ,IAAI,CAAC,SAACF;eAAS/D,SAAS+D,KAAK/D,KAAK,IAAIA,QAAQ+D,KAAK7D,GAAG;;AACrE;AAEA,4FAA4F;AAC5F,yFAAyF;AACzF,SAASgE,SAAS5D,GAAW,EAAE6D,YAAoB;QAAEC,cAAAA,iEAAc;IACjE,OAAOpC,mBAAmB1B,KACvBiC,GAAG,CAAC,SAACL;eAAM,AAAC,GAAiBA,OAAfiC,cAAuE,OAAxDjC,EAAE/B,MAAM,GAAG,KAAKiE,cAAc,AAAC,IAAe,OAAZlC,EAAEM,IAAI,CAAC,MAAK,OAAKN,CAAC,CAAC,EAAE;OACpFM,IAAI,CAAC;AACV;AAEA,kGAAkG;AAClG,iGAAiG;AACjG,SAAS6B,gBAAgBd,KAAa;IACpC,IAAME,QAAQH,gBAAgBC;IAC9B,IAAInB,MAAM;IACV,IAAIhC,SAAS;QACR,kCAAA,2BAAA;;QAAL,QAAK,YAAemD,MAAMzD,QAAQ,CAAC+C,sCAA9B,SAAA,6BAAA,QAAA,yBAAA,iCAAiD;YAAjD,IAAM9C,QAAN;YACH,IAAMC,QAAQD,MAAME,KAAK;YACzB,IAAMK,MAAMP,KAAK,CAAC,EAAE;YACpB,IAAIiE,iBAAiBhE,OAAOyD,UAAUlC,YAAY3B,IAAI,CAACU,MAAM;YAC7D8B,OAAOmB,MAAMlD,KAAK,CAACD,QAAQJ;YAC3B,IAAMsE,YAAYlC,IAAIrC,KAAK,CAAC2C;YAC5B,IAAI4B,WAAW;gBACblC,MAAM,AAAC,GAAkCkC,OAAhClC,IAAI/B,KAAK,CAAC,GAAGiE,UAAUrE,KAAK,GAAmBiE,OAAfI,SAAS,CAAC,EAAE,EAA8D,OAA3DJ,SAAS5D,KAAK,AAAC,GAAiBgE,OAAfA,SAAS,CAAC,EAAE,EAAgB,OAAbA,SAAS,CAAC,EAAE,EAAC,UAAQ;YAC/G,OAAO;gBACLlC,OAAO8B,SAAS5D,KAAKqC,iBAAiB;YACxC;YACAvC,SAASJ,QAAQD,KAAK,CAAC,EAAE,CAACI,MAAM;QAClC;;QAZK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAaL,OAAOC,WAAW,IAAImD,QAAQnB,MAAMmB,MAAMlD,KAAK,CAACD;AAClD;AAIO,SAASf,aAAakE,KAAa;IACxC,IAAI,CAAC7D,QAAQE,IAAI,CAAC2D,QAAQ,OAAOA;IACjCA,QAAQO,qBAAqBP;IAC7BA,QAAQc,gBAAgBd;IACxB,IAAME,QAAQH,gBAAgBC;IAC9B,IAAInB,MAAM;IACV,IAAIiB,SAAS;IACb,IAAIkB,SAAS;IACb,IAAMC,SAASjB,MAAMkB,KAAK,CAACjF,MAAM,2CAA2C;IAC5E,IAAK,IAAIsB,IAAI,GAAGA,IAAI0D,OAAOrE,MAAM,EAAEW,IAAK;QACtC,IAAIA,IAAI,MAAM,GAAG;gBACV,kCAAA,2BAAA;;gBAAL,QAAK,YAAY0D,MAAM,CAAC1D,EAAE,qBAArB,SAAA,6BAAA,QAAA,yBAAA;oBAAA,IAAM4D,KAAN;oBAAuB,IAAIA,OAAO,KAAKrB,SAAS,CAACA;;;gBAAjD;gBAAA;;;yBAAA,6BAAA;wBAAA;;;wBAAA;8BAAA;;;;YACLjB,OAAOoC,MAAM,CAAC1D,EAAE;YAChByD,UAAUC,MAAM,CAAC1D,EAAE,CAACX,MAAM;YAC1B;QACF;QACA,IAAMwE,eAAeX,iBAAiBO,QAAQd;QAC9C,IAAIJ,UAAUsB,cAAc;YAC1BvC,OAAOoC,MAAM,CAAC1D,EAAE,EAAE,2CAA2C;YAC7DyD,UAAUC,MAAM,CAAC1D,EAAE,CAACX,MAAM;YAC1B;QACF;QACA,IAAMkB,IAAIe,IAAIrC,KAAK,CAAC2C;QACpB,IAAIrB,GAAG;YACLe,MAAM,AAAC,GAA0Bf,OAAxBe,IAAI/B,KAAK,CAAC,GAAGgB,EAAEpB,KAAK,GAAWiE,OAAP7C,CAAC,CAAC,EAAE,EAA8C,OAA3C6C,SAASM,MAAM,CAAC1D,EAAE,EAAE,AAAC,GAASO,OAAPA,CAAC,CAAC,EAAE,EAAQ,OAALA,CAAC,CAAC,EAAE,EAAC;QAC7E,OAAO;YACLe,OAAO8B,SAASM,MAAM,CAAC1D,EAAE,EAAE6B;QAC7B;QACA4B,UAAUC,MAAM,CAAC1D,EAAE,CAACX,MAAM;IAC5B;IACA,OAAOiC;AACT;AAIO,SAASpD,cAAcW,IAAY;IACxC,OAAOA,KAAKiF,SAAS,CAAC,OAAOvC,OAAO,CAAC,27JAAW,IAAIwC,WAAW;AACjE"}

package/dist/cjs/text/strip.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/text/strip.ts"],"sourcesContent":["import type { Block } from '../chunk/index.ts';\nimport { extractTexts, parse } from '../chunk/index.ts';\nimport { normalizeText } from '../scan/frontmatter.ts';\n\n// Flat prose from blocks a caller already parsed, via the same extractor the chunker uses\n// (src/chunk) -- words and vectors read one definition of \"the prose of this note\".\nexport function textFromBlocks(blocks: Block[]): string {\n  return normalizeText(extractTexts(blocks));\n}\n\n// Thin parse + delegate, for callers with no blocks of their own already.\nexport function stripText(value: string): string {\n  return textFromBlocks(parse(value));\n}\n"],"names":["stripText","textFromBlocks","blocks","normalizeText","extractTexts","value","parse"],"mappings":";;;;;;;;;;;QAWgBA;eAAAA;;QALAC;eAAAA;;;uBALoB;6BACN;AAIvB,SAASA,eAAeC,MAAe;IAC5C,OAAOC,IAAAA,4BAAa,EAACC,IAAAA,qBAAY,EAACF;AACpC;AAGO,SAASF,UAAUK,KAAa;IACrC,OAAOJ,eAAeK,IAAAA,cAAK,EAACD;AAC9B"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/text/strip.ts"],"sourcesContent":["import type { Block } from '../chunk/index.ts';\nimport { extractTexts, parse } from '../chunk/index.ts';\nimport { normalizeText } from '../scan/frontmatter.ts';\n\n// Flat prose from blocks a caller already parsed, via the same extractor the chunker uses\n// (src/chunk) -- words and vectors read one definition of \"the prose of this note\".\nexport function textFromBlocks(blocks: Block[]): string {\n  return normalizeText(extractTexts(blocks));\n}\n\n// Thin parse + delegate, for callers with no blocks of their own already.\nexport function stripText(value: string): string {\n  return textFromBlocks(parse(value));\n}\n"],"names":["stripText","textFromBlocks","blocks","normalizeText","extractTexts","value","parse"],"mappings":";;;;;;;;;;;QAWgBA;eAAAA;;QALAC;eAAAA;;;uBALoB;6BACN;AAIvB,SAASA,eAAeC,MAAe;IAC5C,OAAOC,IAAAA,4BAAa,EAACC,IAAAA,qBAAY,EAACF;AACpC;AAGO,SAASF,UAAUK,KAAa;IACrC,OAAOJ,eAAeK,IAAAA,cAAK,EAACD;AAC9B"}

package/dist/cjs/watch-claim.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/watch-claim.ts"],"sourcesContent":["import { randomUUID } from 'node:crypto';\nimport { existsSync } from 'node:fs';\nimport { join } from 'node:path';\nimport { DatabaseSync } from 'node:sqlite';\nimport { Worker } from 'node:worker_threads';\nimport { SenseError } from './errors.ts';\nimport { resolveWorkerFile } from './lib/worker-file.ts';\nimport { reviveError, type WorkerErrorPayload } from './scan/worker-error.ts';\nimport { createConnection } from './store/sqlite/connection.ts';\nimport { BEGIN_WRITE, withTransaction } from './store/transaction.ts';\n\nexport const WATCH_CLAIM_FILENAME = '.sense-watch.db';\nexport const WATCH_HEARTBEAT_INTERVAL_MS = 5_000;\nexport const WATCH_STALE_HEARTBEAT_MS = 15_000;\nconst CLAIM_BUSY_TIMEOUT_MS = 30_000;\n\nexport interface WatchClaimRecord {\n  token: string;\n  pid: number;\n  heartbeatMs: number;\n}\n\ninterface ClaimRow {\n  token: string;\n  pid: number;\n  heartbeat_ms: number;\n}\n\nexport interface WatchClaimWorkerData {\n  configDir: string;\n  force: boolean;\n  heartbeatIntervalMs: number;\n  pid: number;\n  token: string;\n}\n\nexport type WatchClaimWorkerMessage = { type: 'acquired' } | { type: 'stopped' } | { type: 'failure'; error: WorkerErrorPayload };\n\nexport interface WatchClaimController {\n  close(): Promise<void>;\n}\n\nfunction claimPath(configDir: string): string {\n  return join(configDir, WATCH_CLAIM_FILENAME);\n}\n\nfunction changedOne(changes: number | bigint): boolean {\n  return changes === 1 || changes === BigInt(1);\n}\n\n// This database is separate from the selected search store so its heartbeat remains writable while\n// a DuckDB, Turso, or SQLite reconciliation owns the cache connection or blocks the main thread.\nexport class WatchClaimDatabase {\n  private readonly db: DatabaseSync;\n  private readonly conn: ReturnType<typeof createConnection>;\n\n  constructor(configDir: string) {\n    const db = new DatabaseSync(claimPath(configDir));\n    try {\n      db.exec(`PRAGMA busy_timeout = ${CLAIM_BUSY_TIMEOUT_MS}`);\n      db.exec('CREATE TABLE IF NOT EXISTS watch_claim (singleton INTEGER PRIMARY KEY CHECK (singleton = 1), token TEXT NOT NULL, pid INTEGER NOT NULL, heartbeat_ms INTEGER NOT NULL)');\n    } catch (err) {\n      try {\n        db.close();\n      } catch (closeError) {\n        throw new AggregateError([err, closeError], 'watch claim setup and database cleanup both failed');\n      }\n      throw err;\n    }\n    this.db = db;\n    this.conn = createConnection(db);\n  }\n\n  async acquire(token: string, pid: number, force: boolean): Promise<void> {\n    await withTransaction(\n      this.conn,\n      async () => {\n        const current = this.db.prepare('SELECT token, pid, heartbeat_ms FROM watch_claim WHERE singleton = 1').get() as ClaimRow | undefined;\n        if (current && !force) {\n          const age = Date.now() - current.heartbeat_ms;\n          if (age >= 0 && age < WATCH_STALE_HEARTBEAT_MS) throw new SenseError('WATCH_ACTIVE', `another watcher appears active (heartbeat ${Math.round(age / 1000)}s ago); use --force to override`);\n        }\n        this.db.prepare('INSERT INTO watch_claim (singleton, token, pid, heartbeat_ms) VALUES (1, ?, ?, ?) ON CONFLICT(singleton) DO UPDATE SET token = excluded.token, pid = excluded.pid, heartbeat_ms = excluded.heartbeat_ms').run(token, pid, Date.now());\n      },\n      BEGIN_WRITE\n    );\n  }\n\n  renew(token: string): boolean {\n    return changedOne(this.db.prepare('UPDATE watch_claim SET heartbeat_ms = ? WHERE singleton = 1 AND token = ?').run(Date.now(), token).changes);\n  }\n\n  release(token: string): boolean {\n    return changedOne(this.db.prepare('DELETE FROM watch_claim WHERE singleton = 1 AND token = ?').run(token).changes);\n  }\n\n  read(): WatchClaimRecord | null {\n    const row = this.db.prepare('SELECT token, pid, heartbeat_ms FROM watch_claim WHERE singleton = 1').get() as ClaimRow | undefined;\n    return row ? { token: row.token, pid: row.pid, heartbeatMs: row.heartbeat_ms } : null;\n  }\n\n  close(): void {\n    this.db.close();\n  }\n}\n\n// Status opens the coordinator read-only and therefore cannot create either the file or its schema.\nexport function readWatchClaim(configDir: string): WatchClaimRecord | null {\n  const path = claimPath(configDir);\n  if (!existsSync(path)) return null;\n  let db: DatabaseSync;\n  try {\n    db = new DatabaseSync(path, { readOnly: true });\n  } catch (err) {\n    if (!existsSync(path)) return null;\n    throw err;\n  }\n  try {\n    db.exec(`PRAGMA busy_timeout = ${CLAIM_BUSY_TIMEOUT_MS}`);\n    const row = db.prepare('SELECT token, pid, heartbeat_ms FROM watch_claim WHERE singleton = 1').get() as ClaimRow | undefined;\n    return row ? { token: row.token, pid: row.pid, heartbeatMs: row.heartbeat_ms } : null;\n  } catch (err) {\n    if (/no such table: watch_claim/.test((err as Error).message)) return null;\n    throw err;\n  } finally {\n    db.close();\n  }\n}\n\nexport async function startWatchClaim(configDir: string, force: boolean, heartbeatIntervalMs: number, onFailure: (err: Error) => void): Promise<WatchClaimController> {\n  const workerData: WatchClaimWorkerData = { configDir, force, heartbeatIntervalMs, pid: process.pid, token: randomUUID() };\n  const worker = new Worker(resolveWorkerFile('watch-heartbeat'), { workerData });\n  let acquired = false;\n  let closing = false;\n  let exited = false;\n  let failureReported = false;\n  let closeError: Error | null = null;\n  let closePromise: Promise<void> | undefined;\n  let resolveExit!: () => void;\n  const exit = new Promise<void>((resolve) => {\n    resolveExit = resolve;\n  });\n  let resolveStart!: () => void;\n  let rejectStart!: (err: Error) => void;\n  const start = new Promise<void>((resolve, reject) => {\n    resolveStart = resolve;\n    rejectStart = reject;\n  });\n\n  const fail = (err: Error): void => {\n    if (failureReported) return;\n    failureReported = true;\n    if (!acquired) rejectStart(err);\n    else if (closing) closeError ??= err;\n    else onFailure(err);\n  };\n  worker.on('message', (message: WatchClaimWorkerMessage) => {\n    if (message.type === 'acquired') {\n      acquired = true;\n      resolveStart();\n    } else if (message.type === 'failure') {\n      fail(reviveError(message.error));\n    }\n  });\n  worker.on('error', fail);\n  worker.on('exit', (code) => {\n    exited = true;\n    resolveExit();\n    if (closing && code !== 0) closeError ??= new Error(`watch heartbeat worker stopped with exit code ${code}`);\n    else if (!closing && !failureReported) fail(new Error(`watch heartbeat worker stopped unexpectedly with exit code ${code}`));\n  });\n\n  try {\n    await start;\n  } catch (err) {\n    closing = true;\n    if (!exited) await worker.terminate();\n    await exit;\n    throw err;\n  }\n\n  return {\n    close(): Promise<void> {\n      closePromise ??= (async () => {\n        closing = true;\n        if (!exited) {\n          try {\n            worker.postMessage({ type: 'stop' });\n          } catch (err) {\n            closeError ??= err instanceof Error ? err : new Error(String(err));\n            await worker.terminate();\n          }\n        }\n        await exit;\n        if (closeError) throw closeError;\n      })();\n      return closePromise;\n    },\n  };\n}\n"],"names":["WATCH_CLAIM_FILENAME","WATCH_HEARTBEAT_INTERVAL_MS","WATCH_STALE_HEARTBEAT_MS","WatchClaimDatabase","readWatchClaim","startWatchClaim","CLAIM_BUSY_TIMEOUT_MS","claimPath","configDir","join","changedOne","changes","BigInt","db","DatabaseSync","exec","err","close","closeError","AggregateError","conn","createConnection","acquire","token","pid","force","withTransaction","current","age","prepare","get","Date","now","heartbeat_ms","SenseError","Math","round","run","BEGIN_WRITE","renew","release","read","row","heartbeatMs","path","existsSync","readOnly","test","message","heartbeatIntervalMs","onFailure","workerData","worker","acquired","closing","exited","failureReported","closePromise","resolveExit","exit","resolveStart","rejectStart","start","fail","process","randomUUID","Worker","resolveWorkerFile","Promise","resolve","reject","on","type","reviveError","error","code","Error","terminate","postMessage","String"],"mappings":";;;;;;;;;;;QAWaA;eAAAA;;QACAC;eAAAA;;QACAC;eAAAA;;QAuCAC;eAAAA;;QAuDGC;eAAAA;;QAsBMC;eAAAA;;;0BAjIK;sBACA;wBACN;0BACQ;kCACN;wBACI;4BACO;6BACmB;4BACpB;6BACY;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAEtC,IAAML,uBAAuB;AAC7B,IAAMC,8BAA8B;AACpC,IAAMC,2BAA2B;AACxC,IAAMI,wBAAwB;AA4B9B,SAASC,UAAUC,SAAiB;IAClC,OAAOC,IAAAA,cAAI,EAACD,WAAWR;AACzB;AAEA,SAASU,WAAWC,OAAwB;IAC1C,OAAOA,YAAY,KAAKA,YAAYC,OAAO;AAC7C;AAIO,IAAA,AAAMT,mCAAN;;aAAMA,mBAICK,SAAiB;gCAJlBL;QAKT,IAAMU,KAAK,IAAIC,wBAAY,CAACP,UAAUC;QACtC,IAAI;YACFK,GAAGE,IAAI,CAAC,AAAC,yBAA8C,OAAtBT;YACjCO,GAAGE,IAAI,CAAC;QACV,EAAE,OAAOC,KAAK;YACZ,IAAI;gBACFH,GAAGI,KAAK;YACV,EAAE,OAAOC,YAAY;gBACnB,MAAM,IAAIC,eAAe;oBAACH;oBAAKE;iBAAW,EAAE;YAC9C;YACA,MAAMF;QACR;QACA,IAAI,CAACH,EAAE,GAAGA;QACV,IAAI,CAACO,IAAI,GAAGC,IAAAA,8BAAgB,EAACR;;iBAlBpBV;IAqBX,OAAMmB,OAaL,GAbD,SAAMA,QAAQC,KAAa,EAAEC,GAAW,EAAEC,KAAc;;;;;;;wBACtD;;4BAAMC,IAAAA,8BAAe,EACnB,IAAI,CAACN,IAAI,EACT;;wCACQO,SAEEC;;wCAFFD,UAAU,IAAI,CAACd,EAAE,CAACgB,OAAO,CAAC,wEAAwEC,GAAG;wCAC3G,IAAIH,WAAW,CAACF,OAAO;4CACfG,MAAMG,KAAKC,GAAG,KAAKL,QAAQM,YAAY;4CAC7C,IAAIL,OAAO,KAAKA,MAAM1B,0BAA0B,MAAM,IAAIgC,oBAAU,CAAC,gBAAgB,AAAC,6CAAmE,OAAvBC,KAAKC,KAAK,CAACR,MAAM,OAAM;wCAC3J;wCACA,IAAI,CAACf,EAAE,CAACgB,OAAO,CAAC,2MAA2MQ,GAAG,CAACd,OAAOC,KAAKO,KAAKC,GAAG;;;;;gCACrP;+BACAM,0BAAW;;;wBAVb;;;;;;QAYF;;IAEAC,OAAAA,KAEC,GAFDA,SAAAA,MAAMhB,KAAa;QACjB,OAAOb,WAAW,IAAI,CAACG,EAAE,CAACgB,OAAO,CAAC,6EAA6EQ,GAAG,CAACN,KAAKC,GAAG,IAAIT,OAAOZ,OAAO;IAC/I;IAEA6B,OAAAA,OAEC,GAFDA,SAAAA,QAAQjB,KAAa;QACnB,OAAOb,WAAW,IAAI,CAACG,EAAE,CAACgB,OAAO,CAAC,6DAA6DQ,GAAG,CAACd,OAAOZ,OAAO;IACnH;IAEA8B,OAAAA,IAGC,GAHDA,SAAAA;QACE,IAAMC,MAAM,IAAI,CAAC7B,EAAE,CAACgB,OAAO,CAAC,wEAAwEC,GAAG;QACvG,OAAOY,MAAM;YAAEnB,OAAOmB,IAAInB,KAAK;YAAEC,KAAKkB,IAAIlB,GAAG;YAAEmB,aAAaD,IAAIT,YAAY;QAAC,IAAI;IACnF;IAEAhB,OAAAA,KAEC,GAFDA,SAAAA;QACE,IAAI,CAACJ,EAAE,CAACI,KAAK;IACf;WAnDWd;;AAuDN,SAASC,eAAeI,SAAiB;IAC9C,IAAMoC,OAAOrC,UAAUC;IACvB,IAAI,CAACqC,IAAAA,kBAAU,EAACD,OAAO,OAAO;IAC9B,IAAI/B;IACJ,IAAI;QACFA,KAAK,IAAIC,wBAAY,CAAC8B,MAAM;YAAEE,UAAU;QAAK;IAC/C,EAAE,OAAO9B,KAAK;QACZ,IAAI,CAAC6B,IAAAA,kBAAU,EAACD,OAAO,OAAO;QAC9B,MAAM5B;IACR;IACA,IAAI;QACFH,GAAGE,IAAI,CAAC,AAAC,yBAA8C,OAAtBT;QACjC,IAAMoC,MAAM7B,GAAGgB,OAAO,CAAC,wEAAwEC,GAAG;QAClG,OAAOY,MAAM;YAAEnB,OAAOmB,IAAInB,KAAK;YAAEC,KAAKkB,IAAIlB,GAAG;YAAEmB,aAAaD,IAAIT,YAAY;QAAC,IAAI;IACnF,EAAE,OAAOjB,KAAK;QACZ,IAAI,6BAA6B+B,IAAI,CAAC,AAAC/B,IAAcgC,OAAO,GAAG,OAAO;QACtE,MAAMhC;IACR,SAAU;QACRH,GAAGI,KAAK;IACV;AACF;AAEO,SAAeZ,gBAAgBG,SAAiB,EAAEiB,KAAc,EAAEwB,mBAA2B,EAAEC,SAA+B;;YAC7HC,YACAC,QACFC,UACAC,SACAC,QACAC,iBACAtC,YACAuC,cACAC,aACEC,MAGFC,cACAC,aACEC,OAKAC,MAyBG/C;;;;oBA5CHmC,aAAmC;wBAAE3C,WAAAA;wBAAWiB,OAAAA;wBAAOwB,qBAAAA;wBAAqBzB,KAAKwC,QAAQxC,GAAG;wBAAED,OAAO0C,IAAAA,sBAAU;oBAAG;oBAClHb,SAAS,IAAIc,0BAAM,CAACC,IAAAA,+BAAiB,EAAC,oBAAoB;wBAAEhB,YAAAA;oBAAW;oBACzEE,WAAW;oBACXC,UAAU;oBACVC,SAAS;oBACTC,kBAAkB;oBAClBtC,aAA2B;oBAGzByC,OAAO,IAAIS,QAAc,SAACC;wBAC9BX,cAAcW;oBAChB;oBAGMP,QAAQ,IAAIM,QAAc,SAACC,SAASC;wBACxCV,eAAeS;wBACfR,cAAcS;oBAChB;oBAEMP,OAAO,cAAC/C;wBACZ,IAAIwC,iBAAiB;wBACrBA,kBAAkB;wBAClB,IAAI,CAACH,UAAUQ,YAAY7C;6BACtB,IAAIsC,SAASpC,uBAAAA,wBAAAA,aAAAA,aAAeF;6BAC5BkC,UAAUlC;oBACjB;oBACAoC,OAAOmB,EAAE,CAAC,WAAW,SAACvB;wBACpB,IAAIA,QAAQwB,IAAI,KAAK,YAAY;4BAC/BnB,WAAW;4BACXO;wBACF,OAAO,IAAIZ,QAAQwB,IAAI,KAAK,WAAW;4BACrCT,KAAKU,IAAAA,0BAAW,EAACzB,QAAQ0B,KAAK;wBAChC;oBACF;oBACAtB,OAAOmB,EAAE,CAAC,SAASR;oBACnBX,OAAOmB,EAAE,CAAC,QAAQ,SAACI;wBACjBpB,SAAS;wBACTG;wBACA,IAAIJ,WAAWqB,SAAS,GAAGzD,uBAAAA,wBAAAA,aAAAA,aAAe,IAAI0D,MAAM,AAAC,iDAAqD,OAALD;6BAChG,IAAI,CAACrB,WAAW,CAACE,iBAAiBO,KAAK,IAAIa,MAAM,AAAC,8DAAkE,OAALD;oBACtH;;;;;;;;;oBAGE;;wBAAMb;;;oBAAN;;;;;;oBACO9C;oBACPsC,UAAU;yBACN,CAACC,QAAD;;;;oBAAS;;wBAAMH,OAAOyB,SAAS;;;oBAAtB;;;oBACb;;wBAAMlB;;;oBAAN;oBACA,MAAM3C;;oBAGR;;wBAAO;4BACLC,OAAAA,SAAAA;gCACEwC,yBAAAA,0BAAAA,eAAAA,eAAiB,AAAC;;4CAKLzC;;;;oDAJXsC,UAAU;yDACN,CAACC,QAAD;;;;;;;;;;;;oDAEAH,OAAO0B,WAAW,CAAC;wDAAEN,MAAM;oDAAO;;;;;;oDAC3BxD;oDACPE,uBAAAA,wBAAAA,aAAAA,aAAeF,AAAG,YAAHA,KAAe4D,SAAQ5D,MAAM,IAAI4D,MAAMG,OAAO/D;oDAC7D;;wDAAMoC,OAAOyB,SAAS;;;oDAAtB;;;;;;oDAGJ;;wDAAMlB;;;oDAAN;oDACA,IAAIzC,YAAY,MAAMA;;;;;;oCACxB;;gCACA,OAAOuC;4BACT;wBACF;;;;IACF"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/watch-claim.ts"],"sourcesContent":["import { randomUUID } from 'node:crypto';\nimport { existsSync } from 'node:fs';\nimport { join } from 'node:path';\nimport { DatabaseSync } from 'node:sqlite';\nimport { Worker } from 'node:worker_threads';\nimport { SenseError } from './errors.ts';\nimport { resolveWorkerFile } from './lib/worker-file.ts';\nimport { reviveError, type WorkerErrorPayload } from './scan/worker-error.ts';\nimport { createConnection } from './store/sqlite/connection.ts';\nimport { BEGIN_WRITE, withTransaction } from './store/transaction.ts';\n\nexport const WATCH_CLAIM_FILENAME = '.sense-watch.db';\nexport const WATCH_HEARTBEAT_INTERVAL_MS = 5_000;\nexport const WATCH_STALE_HEARTBEAT_MS = 15_000;\nconst CLAIM_BUSY_TIMEOUT_MS = 30_000;\n\nexport interface WatchClaimRecord {\n  token: string;\n  pid: number;\n  heartbeatMs: number;\n}\n\ninterface ClaimRow {\n  token: string;\n  pid: number;\n  heartbeat_ms: number;\n}\n\nexport interface WatchClaimWorkerData {\n  configDir: string;\n  force: boolean;\n  heartbeatIntervalMs: number;\n  pid: number;\n  token: string;\n}\n\nexport type WatchClaimWorkerMessage = { type: 'acquired' } | { type: 'stopped' } | { type: 'failure'; error: WorkerErrorPayload };\n\nexport interface WatchClaimController {\n  close(): Promise<void>;\n}\n\nfunction claimPath(configDir: string): string {\n  return join(configDir, WATCH_CLAIM_FILENAME);\n}\n\nfunction changedOne(changes: number | bigint): boolean {\n  return changes === 1 || changes === BigInt(1);\n}\n\n// This database is separate from the selected search store so its heartbeat remains writable while\n// a DuckDB, Turso, or SQLite reconciliation owns the cache connection or blocks the main thread.\nexport class WatchClaimDatabase {\n  private readonly db: DatabaseSync;\n  private readonly conn: ReturnType<typeof createConnection>;\n\n  constructor(configDir: string) {\n    const db = new DatabaseSync(claimPath(configDir));\n    try {\n      db.exec(`PRAGMA busy_timeout = ${CLAIM_BUSY_TIMEOUT_MS}`);\n      db.exec('CREATE TABLE IF NOT EXISTS watch_claim (singleton INTEGER PRIMARY KEY CHECK (singleton = 1), token TEXT NOT NULL, pid INTEGER NOT NULL, heartbeat_ms INTEGER NOT NULL)');\n    } catch (err) {\n      try {\n        db.close();\n      } catch (closeError) {\n        throw new AggregateError([err, closeError], 'watch claim setup and database cleanup both failed');\n      }\n      throw err;\n    }\n    this.db = db;\n    this.conn = createConnection(db);\n  }\n\n  async acquire(token: string, pid: number, force: boolean): Promise<void> {\n    await withTransaction(\n      this.conn,\n      async () => {\n        const current = this.db.prepare('SELECT token, pid, heartbeat_ms FROM watch_claim WHERE singleton = 1').get() as ClaimRow | undefined;\n        if (current && !force) {\n          const age = Date.now() - current.heartbeat_ms;\n          if (age >= 0 && age < WATCH_STALE_HEARTBEAT_MS) throw new SenseError('WATCH_ACTIVE', `another watcher appears active (heartbeat ${Math.round(age / 1000)}s ago); use --force to override`);\n        }\n        this.db.prepare('INSERT INTO watch_claim (singleton, token, pid, heartbeat_ms) VALUES (1, ?, ?, ?) ON CONFLICT(singleton) DO UPDATE SET token = excluded.token, pid = excluded.pid, heartbeat_ms = excluded.heartbeat_ms').run(token, pid, Date.now());\n      },\n      BEGIN_WRITE\n    );\n  }\n\n  renew(token: string): boolean {\n    return changedOne(this.db.prepare('UPDATE watch_claim SET heartbeat_ms = ? WHERE singleton = 1 AND token = ?').run(Date.now(), token).changes);\n  }\n\n  release(token: string): boolean {\n    return changedOne(this.db.prepare('DELETE FROM watch_claim WHERE singleton = 1 AND token = ?').run(token).changes);\n  }\n\n  read(): WatchClaimRecord | null {\n    const row = this.db.prepare('SELECT token, pid, heartbeat_ms FROM watch_claim WHERE singleton = 1').get() as ClaimRow | undefined;\n    return row ? { token: row.token, pid: row.pid, heartbeatMs: row.heartbeat_ms } : null;\n  }\n\n  close(): void {\n    this.db.close();\n  }\n}\n\n// Status opens the coordinator read-only and therefore cannot create either the file or its schema.\nexport function readWatchClaim(configDir: string): WatchClaimRecord | null {\n  const path = claimPath(configDir);\n  if (!existsSync(path)) return null;\n  let db: DatabaseSync;\n  try {\n    db = new DatabaseSync(path, { readOnly: true });\n  } catch (err) {\n    if (!existsSync(path)) return null;\n    throw err;\n  }\n  try {\n    db.exec(`PRAGMA busy_timeout = ${CLAIM_BUSY_TIMEOUT_MS}`);\n    const row = db.prepare('SELECT token, pid, heartbeat_ms FROM watch_claim WHERE singleton = 1').get() as ClaimRow | undefined;\n    return row ? { token: row.token, pid: row.pid, heartbeatMs: row.heartbeat_ms } : null;\n  } catch (err) {\n    if (/no such table: watch_claim/.test((err as Error).message)) return null;\n    throw err;\n  } finally {\n    db.close();\n  }\n}\n\nexport async function startWatchClaim(configDir: string, force: boolean, heartbeatIntervalMs: number, onFailure: (err: Error) => void): Promise<WatchClaimController> {\n  const workerData: WatchClaimWorkerData = { configDir, force, heartbeatIntervalMs, pid: process.pid, token: randomUUID() };\n  const worker = new Worker(resolveWorkerFile('watch-heartbeat'), { workerData });\n  let acquired = false;\n  let closing = false;\n  let exited = false;\n  let failureReported = false;\n  let closeError: Error | null = null;\n  let closePromise: Promise<void> | undefined;\n  let resolveExit!: () => void;\n  const exit = new Promise<void>((resolve) => {\n    resolveExit = resolve;\n  });\n  let resolveStart!: () => void;\n  let rejectStart!: (err: Error) => void;\n  const start = new Promise<void>((resolve, reject) => {\n    resolveStart = resolve;\n    rejectStart = reject;\n  });\n\n  const fail = (err: Error): void => {\n    if (failureReported) return;\n    failureReported = true;\n    if (!acquired) rejectStart(err);\n    else if (closing) closeError ??= err;\n    else onFailure(err);\n  };\n  worker.on('message', (message: WatchClaimWorkerMessage) => {\n    if (message.type === 'acquired') {\n      acquired = true;\n      resolveStart();\n    } else if (message.type === 'failure') {\n      fail(reviveError(message.error));\n    }\n  });\n  worker.on('error', fail);\n  worker.on('exit', (code) => {\n    exited = true;\n    resolveExit();\n    if (closing && code !== 0) closeError ??= new Error(`watch heartbeat worker stopped with exit code ${code}`);\n    else if (!closing && !failureReported) fail(new Error(`watch heartbeat worker stopped unexpectedly with exit code ${code}`));\n  });\n\n  try {\n    await start;\n  } catch (err) {\n    closing = true;\n    if (!exited) await worker.terminate();\n    await exit;\n    throw err;\n  }\n\n  return {\n    close(): Promise<void> {\n      closePromise ??= (async () => {\n        closing = true;\n        if (!exited) {\n          try {\n            worker.postMessage({ type: 'stop' });\n          } catch (err) {\n            closeError ??= err instanceof Error ? err : new Error(String(err));\n            await worker.terminate();\n          }\n        }\n        await exit;\n        if (closeError) throw closeError;\n      })();\n      return closePromise;\n    },\n  };\n}\n"],"names":["WATCH_CLAIM_FILENAME","WATCH_HEARTBEAT_INTERVAL_MS","WATCH_STALE_HEARTBEAT_MS","WatchClaimDatabase","readWatchClaim","startWatchClaim","CLAIM_BUSY_TIMEOUT_MS","claimPath","configDir","join","changedOne","changes","BigInt","db","DatabaseSync","exec","err","close","closeError","AggregateError","conn","createConnection","acquire","token","pid","force","withTransaction","current","age","prepare","get","Date","now","heartbeat_ms","SenseError","Math","round","run","BEGIN_WRITE","renew","release","read","row","heartbeatMs","path","existsSync","readOnly","test","message","heartbeatIntervalMs","onFailure","workerData","worker","acquired","closing","exited","failureReported","closePromise","resolveExit","exit","resolveStart","rejectStart","start","fail","process","randomUUID","Worker","resolveWorkerFile","Promise","resolve","reject","on","type","reviveError","error","code","Error","terminate","postMessage","String"],"mappings":";;;;;;;;;;;QAWaA;eAAAA;;QACAC;eAAAA;;QACAC;eAAAA;;QAuCAC;eAAAA;;QAuDGC;eAAAA;;QAsBMC;eAAAA;;;0BAjIK;sBACA;wBACN;0BACQ;kCACN;wBACI;4BACO;6BACmB;4BACpB;6BACY;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAEtC,IAAML,uBAAuB;AAC7B,IAAMC,8BAA8B;AACpC,IAAMC,2BAA2B;AACxC,IAAMI,wBAAwB;AA4B9B,SAASC,UAAUC,SAAiB;IAClC,OAAOC,IAAAA,cAAI,EAACD,WAAWR;AACzB;AAEA,SAASU,WAAWC,OAAwB;IAC1C,OAAOA,YAAY,KAAKA,YAAYC,OAAO;AAC7C;AAIO,IAAA,AAAMT,mCAAN;;aAAMA,mBAICK,SAAiB;gCAJlBL;QAKT,IAAMU,KAAK,IAAIC,wBAAY,CAACP,UAAUC;QACtC,IAAI;YACFK,GAAGE,IAAI,CAAC,AAAC,yBAA8C,OAAtBT;YACjCO,GAAGE,IAAI,CAAC;QACV,EAAE,OAAOC,KAAK;YACZ,IAAI;gBACFH,GAAGI,KAAK;YACV,EAAE,OAAOC,YAAY;gBACnB,MAAM,IAAIC,eAAe;oBAACH;oBAAKE;iBAAW,EAAE;YAC9C;YACA,MAAMF;QACR;QACA,IAAI,CAACH,EAAE,GAAGA;QACV,IAAI,CAACO,IAAI,GAAGC,IAAAA,8BAAgB,EAACR;;iBAlBpBV;IAqBX,OAAMmB,OAaL,GAbD,SAAMA,QAAQC,KAAa,EAAEC,GAAW,EAAEC,KAAc;;;;;;;wBACtD;;4BAAMC,IAAAA,8BAAe,EACnB,IAAI,CAACN,IAAI,EACT;;wCACQO,SAEEC;;wCAFFD,UAAU,IAAI,CAACd,EAAE,CAACgB,OAAO,CAAC,wEAAwEC,GAAG;wCAC3G,IAAIH,WAAW,CAACF,OAAO;4CACfG,MAAMG,KAAKC,GAAG,KAAKL,QAAQM,YAAY;4CAC7C,IAAIL,OAAO,KAAKA,MAAM1B,0BAA0B,MAAM,IAAIgC,oBAAU,CAAC,gBAAgB,AAAC,6CAAmE,OAAvBC,KAAKC,KAAK,CAACR,MAAM,OAAM;wCAC3J;wCACA,IAAI,CAACf,EAAE,CAACgB,OAAO,CAAC,2MAA2MQ,GAAG,CAACd,OAAOC,KAAKO,KAAKC,GAAG;;;;;gCACrP;+BACAM,0BAAW;;;wBAVb;;;;;;QAYF;;IAEAC,OAAAA,KAEC,GAFDA,SAAAA,MAAMhB,KAAa;QACjB,OAAOb,WAAW,IAAI,CAACG,EAAE,CAACgB,OAAO,CAAC,6EAA6EQ,GAAG,CAACN,KAAKC,GAAG,IAAIT,OAAOZ,OAAO;IAC/I;IAEA6B,OAAAA,OAEC,GAFDA,SAAAA,QAAQjB,KAAa;QACnB,OAAOb,WAAW,IAAI,CAACG,EAAE,CAACgB,OAAO,CAAC,6DAA6DQ,GAAG,CAACd,OAAOZ,OAAO;IACnH;IAEA8B,OAAAA,IAGC,GAHDA,SAAAA;QACE,IAAMC,MAAM,IAAI,CAAC7B,EAAE,CAACgB,OAAO,CAAC,wEAAwEC,GAAG;QACvG,OAAOY,MAAM;YAAEnB,OAAOmB,IAAInB,KAAK;YAAEC,KAAKkB,IAAIlB,GAAG;YAAEmB,aAAaD,IAAIT,YAAY;QAAC,IAAI;IACnF;IAEAhB,OAAAA,KAEC,GAFDA,SAAAA;QACE,IAAI,CAACJ,EAAE,CAACI,KAAK;IACf;WAnDWd;;AAuDN,SAASC,eAAeI,SAAiB;IAC9C,IAAMoC,OAAOrC,UAAUC;IACvB,IAAI,CAACqC,IAAAA,kBAAU,EAACD,OAAO,OAAO;IAC9B,IAAI/B;IACJ,IAAI;QACFA,KAAK,IAAIC,wBAAY,CAAC8B,MAAM;YAAEE,UAAU;QAAK;IAC/C,EAAE,OAAO9B,KAAK;QACZ,IAAI,CAAC6B,IAAAA,kBAAU,EAACD,OAAO,OAAO;QAC9B,MAAM5B;IACR;IACA,IAAI;QACFH,GAAGE,IAAI,CAAC,AAAC,yBAA8C,OAAtBT;QACjC,IAAMoC,MAAM7B,GAAGgB,OAAO,CAAC,wEAAwEC,GAAG;QAClG,OAAOY,MAAM;YAAEnB,OAAOmB,IAAInB,KAAK;YAAEC,KAAKkB,IAAIlB,GAAG;YAAEmB,aAAaD,IAAIT,YAAY;QAAC,IAAI;IACnF,EAAE,OAAOjB,KAAK;QACZ,IAAI,6BAA6B+B,IAAI,CAAC,AAAC/B,IAAcgC,OAAO,GAAG,OAAO;QACtE,MAAMhC;IACR,SAAU;QACRH,GAAGI,KAAK;IACV;AACF;AAEO,SAAeZ,gBAAgBG,SAAiB,EAAEiB,KAAc,EAAEwB,mBAA2B,EAAEC,SAA+B;;YAC7HC,YACAC,QACFC,UACAC,SACAC,QACAC,iBACAtC,YACAuC,cACAC,aACEC,MAGFC,cACAC,aACEC,OAKAC,MAyBG/C;;;;oBA5CHmC,aAAmC;wBAAE3C,WAAAA;wBAAWiB,OAAAA;wBAAOwB,qBAAAA;wBAAqBzB,KAAKwC,QAAQxC,GAAG;wBAAED,OAAO0C,IAAAA,sBAAU;oBAAG;oBAClHb,SAAS,IAAIc,0BAAM,CAACC,IAAAA,+BAAiB,EAAC,oBAAoB;wBAAEhB,YAAAA;oBAAW;oBACzEE,WAAW;oBACXC,UAAU;oBACVC,SAAS;oBACTC,kBAAkB;oBAClBtC,aAA2B;oBAGzByC,OAAO,IAAIS,QAAc,SAACC;wBAC9BX,cAAcW;oBAChB;oBAGMP,QAAQ,IAAIM,QAAc,SAACC,SAASC;wBACxCV,eAAeS;wBACfR,cAAcS;oBAChB;oBAEMP,OAAO,cAAC/C;wBACZ,IAAIwC,iBAAiB;wBACrBA,kBAAkB;wBAClB,IAAI,CAACH,UAAUQ,YAAY7C;6BACtB,IAAIsC,SAASpC,uBAAAA,wBAAAA,aAAAA,aAAeF;6BAC5BkC,UAAUlC;oBACjB;oBACAoC,OAAOmB,EAAE,CAAC,WAAW,SAACvB;wBACpB,IAAIA,QAAQwB,IAAI,KAAK,YAAY;4BAC/BnB,WAAW;4BACXO;wBACF,OAAO,IAAIZ,QAAQwB,IAAI,KAAK,WAAW;4BACrCT,KAAKU,IAAAA,0BAAW,EAACzB,QAAQ0B,KAAK;wBAChC;oBACF;oBACAtB,OAAOmB,EAAE,CAAC,SAASR;oBACnBX,OAAOmB,EAAE,CAAC,QAAQ,SAACI;wBACjBpB,SAAS;wBACTG;wBACA,IAAIJ,WAAWqB,SAAS,GAAGzD,uBAAAA,wBAAAA,aAAAA,aAAe,IAAI0D,MAAM,AAAC,iDAAqD,OAALD;6BAChG,IAAI,CAACrB,WAAW,CAACE,iBAAiBO,KAAK,IAAIa,MAAM,AAAC,8DAAkE,OAALD;oBACtH;;;;;;;;;oBAGE;;wBAAMb;;;oBAAN;;;;;;oBACO9C;oBACPsC,UAAU;yBACN,CAACC,QAAD;;;;oBAAS;;wBAAMH,OAAOyB,SAAS;;;oBAAtB;;;oBACb;;wBAAMlB;;;oBAAN;oBACA,MAAM3C;;oBAGR;;wBAAO;4BACLC,OAAAA,SAAAA;gCACEwC,yBAAAA,0BAAAA,eAAAA,eAAiB,AAAC;;4CAKLzC;;;;oDAJXsC,UAAU;yDACN,CAACC,QAAD;;;;;;;;;;;;oDAEAH,OAAO0B,WAAW,CAAC;wDAAEN,MAAM;oDAAO;;;;;;oDAC3BxD;oDACPE,uBAAAA,wBAAAA,aAAAA,aAAeF,AAAG,YAAHA,KAAe4D,SAAQ5D,MAAM,IAAI4D,MAAMG,OAAO/D;oDAC7D;;wDAAMoC,OAAOyB,SAAS;;;oDAAtB;;;;;;oDAGJ;;wDAAMlB;;;oDAAN;oDACA,IAAIzC,YAAY,MAAMA;;;;;;oCACxB;;gCACA,OAAOuC;4BACT;wBACF;;;;IACF"}

package/dist/cjs/watch.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/watch.ts"],"sourcesContent":["import { watch as fsWatch } from 'node:fs';\nimport type { ResolvedConfig } from './config/index.ts';\nimport { SenseError } from './errors.ts';\nimport { guardedTick } from './lib/guarded-tick.ts';\nimport { shouldReconcileWatchNotification } from './lib/watch-notification.ts';\nimport { docCount, openStore } from './store/index.ts';\nimport { startWatchClaim, WATCH_HEARTBEAT_INTERVAL_MS } from './watch-claim.ts';\n\n// Watch prepares the index for snapshot queries; filesystem events request debounced builds\n// and periodic reconciliation recovers missed notifications.\nconst DEBOUNCE_MS = 200;\n\nexport type WatchEvent = { type: 'started'; rootDir: string; dbPath: string } | { type: 'reconciled'; parsed: number; total: number; warnings: string[] } | { type: 'reconcile-error'; message: string };\n\n/** Options for the root {@link runWatch} API. */\nexport interface WatchOptions {\n  force?: boolean;\n  onEvent?: (event: WatchEvent) => void;\n  /** Caller abort cancels provider work, skips final preparation, and resolves after cleanup. */\n  signal?: AbortSignal;\n  debounceMs?: number;\n  heartbeatIntervalMs?: number;\n}\n\nfunction isCallerAbort(signal: AbortSignal | undefined, reason: unknown): boolean {\n  return signal?.aborted === true && Object.is(reason, signal.reason);\n}\n\nfunction errorMessage(reason: unknown): string {\n  return reason instanceof Error ? reason.message : String(reason);\n}\n\n/**\n * Runs in the foreground. Caller abort waits for active native work and cleanup; SIGINT/SIGTERM\n * also run one final freshness pass. Cleanup failures reject either shutdown path.\n */\nexport async function runWatch(cfg: ResolvedConfig, opts: WatchOptions = {}): Promise<void> {\n  if (opts.signal?.aborted) return;\n  const onEvent = opts.onEvent ?? (() => {});\n  const debounceMs = opts.debounceMs ?? DEBOUNCE_MS;\n  const reconcileIntervalMs = opts.heartbeatIntervalMs ?? WATCH_HEARTBEAT_INTERVAL_MS;\n  if (!Number.isFinite(reconcileIntervalMs) || reconcileIntervalMs <= 0) throw new SenseError('CONFIG_INVALID', 'watch heartbeatIntervalMs must be a finite positive number');\n  // Reconciliation may be parked for a caller, but live ownership must renew before fixed expiry.\n  const claimIntervalMs = Math.min(reconcileIntervalMs, WATCH_HEARTBEAT_INTERVAL_MS);\n  const rootDir = cfg.rootDir ?? cfg.baseDir;\n  const configDir = cfg.configDir ?? rootDir;\n  let claimFailure: Error | null = null;\n  let requestShutdown: ((reason: Error) => void) | undefined;\n  const claim = await startWatchClaim(configDir, opts.force ?? false, claimIntervalMs, (err) => {\n    claimFailure ??= err;\n    requestShutdown?.(err);\n  });\n  let claimCloseStarted = false;\n  const closeClaim = (): Promise<void> => {\n    claimCloseStarted = true;\n    return claim.close();\n  };\n  let watcher: ReturnType<typeof fsWatch> | undefined;\n  let startup = true;\n  let startupNotification = false;\n  let scheduleReconcile: (() => void) | undefined;\n  let stopping = false;\n  let callerCanceled = false;\n  try {\n    opts.signal?.throwIfAborted();\n    // Subscribe before the first build. Native watchers have no ready barrier, so startup also\n    // performs an unconditional catch-up pass after the initial build.\n    watcher = fsWatch(rootDir, { recursive: true }, (_event, filename) => {\n      if (!shouldReconcileWatchNotification(filename)) return;\n      if (startup) startupNotification = true;\n      else scheduleReconcile?.();\n    });\n    watcher.on('error', (err) => {\n      claimFailure ??= err;\n      requestShutdown?.(err);\n    });\n\n    const startupEvents: Array<{ parsed: number; total: number; warnings: string[] }> = [];\n    const { store: initialStore, dbPath, warnings: initialWarnings, parsed: initialParsed } = await openStore(cfg, { signal: opts.signal });\n    let initialTotal = 0;\n    try {\n      // total is read before close: an event fires only once nothing is held open again.\n      initialTotal = initialWarnings.length > 0 || initialParsed > 0 ? await docCount(initialStore) : 0;\n    } finally {\n      await initialStore.close();\n    }\n    opts.signal?.throwIfAborted();\n    if (claimFailure) throw claimFailure;\n\n    startupEvents.push({ parsed: initialParsed, total: initialTotal, warnings: initialWarnings });\n    // Always reopen once after subscribing, then drain notifications observed while startup was\n    // in progress. Each pass closes before the next one starts, so no stores overlap.\n    do {\n      startupNotification = false;\n      opts.signal?.throwIfAborted();\n      const { store, parsed, warnings } = await openStore(cfg, { signal: opts.signal });\n      try {\n        if (parsed > 0 || warnings.length > 0) startupEvents.push({ parsed, total: await docCount(store), warnings });\n      } finally {\n        await store.close();\n      }\n      opts.signal?.throwIfAborted();\n      if (claimFailure) throw claimFailure;\n    } while (startupNotification && !opts.signal?.aborted);\n    startup = false;\n\n    // At most one open+reconcile+close cycle runs at a time; a trigger that arrives mid-cycle is\n    // coalesced into a single rerun instead of opening a second, overlapping connection.\n    let running: Promise<void> | null = null;\n    let queuedAlways: boolean | null = null;\n\n    const tick = async (alwaysEmit: boolean): Promise<void> => {\n      try {\n        const { store, parsed, warnings } = await openStore(cfg, { signal: opts.signal });\n        try {\n          if (alwaysEmit || parsed > 0 || warnings.length > 0) {\n            onEvent({ type: 'reconciled', parsed, total: await docCount(store), warnings });\n          }\n        } finally {\n          await store.close();\n        }\n      } catch (err) {\n        if (isCallerAbort(opts.signal, err)) return;\n        onEvent({ type: 'reconcile-error', message: errorMessage(err) });\n      }\n    };\n\n    const startCycle = (alwaysEmit: boolean) => {\n      running = tick(alwaysEmit).then(() => {\n        running = null;\n        if (queuedAlways !== null && !stopping) {\n          const again = queuedAlways;\n          queuedAlways = null;\n          startCycle(again);\n        }\n      });\n    };\n\n    const requestCycle = (alwaysEmit: boolean) => {\n      if (stopping) return;\n      if (running) {\n        queuedAlways = queuedAlways || alwaysEmit;\n        return;\n      }\n      startCycle(alwaysEmit);\n    };\n\n    let debounceTimer: NodeJS.Timeout | null = null;\n    // Ignore our own state files, or cache and claim writes would retrigger reconciliation forever.\n    // An unresolvable filename reconciles because parsing nothing costs less than missing a real edit.\n    scheduleReconcile = () => {\n      if (debounceTimer) clearTimeout(debounceTimer);\n      debounceTimer = setTimeout(() => {\n        debounceTimer = null;\n        requestCycle(true);\n      }, debounceMs);\n    };\n    const reconcileTimer = setInterval(\n      guardedTick(\n        async () => requestCycle(false),\n        () => stopping\n      ),\n      reconcileIntervalMs\n    );\n\n    return await new Promise<void>((resolveShutdown, rejectShutdown) => {\n      let shutdownReason: Error | null = null;\n      const shutdown = (reason?: Error, canceledByCaller = false) => {\n        shutdownReason ??= reason ?? null;\n        callerCanceled ||= canceledByCaller;\n        if (stopping) return;\n        stopping = true;\n        const shutdownTask = (async () => {\n          process.off('SIGINT', stopNormally);\n          process.off('SIGTERM', stopNormally);\n          opts.signal?.removeEventListener('abort', stopForCaller);\n          clearInterval(reconcileTimer);\n          if (debounceTimer) clearTimeout(debounceTimer);\n          watcher?.close();\n          await running;\n          let finalError: unknown;\n          try {\n            // A normal stop performs one final freshness pass. Ownership loss drains existing work\n            // but starts nothing new because the replacement watcher now owns that responsibility.\n            if (!shutdownReason && !callerCanceled) {\n              const { store: finalStore } = await openStore(cfg);\n              await finalStore.close();\n            }\n          } catch (err) {\n            finalError = err;\n          }\n          try {\n            await closeClaim();\n          } catch (err) {\n            if (finalError) throw new AggregateError([finalError, err], 'watch shutdown and claim cleanup both failed');\n            throw err;\n          }\n          if (finalError) throw finalError;\n          if (shutdownReason) throw shutdownReason;\n        })();\n        shutdownTask.then(resolveShutdown, rejectShutdown);\n      };\n      const stopNormally = () => shutdown();\n      const stopForCaller = () => shutdown(undefined, true);\n      requestShutdown = shutdown;\n      process.once('SIGINT', stopNormally);\n      process.once('SIGTERM', stopNormally);\n      if (claimFailure) shutdown(claimFailure);\n      else if (opts.signal?.aborted) stopForCaller();\n      else {\n        opts.signal?.addEventListener('abort', stopForCaller, { once: true });\n        try {\n          onEvent({ type: 'started', rootDir, dbPath });\n          if (!stopping) {\n            for (const event of startupEvents) {\n              if (stopping) break;\n              if (event.warnings.length > 0 || event.parsed > 0) onEvent({ type: 'reconciled', ...event });\n            }\n          }\n        } catch (err) {\n          shutdown(err instanceof Error ? err : new Error(String(err)));\n        }\n      }\n    });\n  } catch (err) {\n    const canceledByCaller = isCallerAbort(opts.signal, err);\n    watcher?.close();\n    if (claimCloseStarted) throw err;\n    try {\n      await closeClaim();\n    } catch (closeError) {\n      throw new AggregateError([err, closeError], 'watch setup and claim cleanup both failed');\n    }\n    if (canceledByCaller) return;\n    throw err;\n  } finally {\n    requestShutdown = undefined;\n  }\n}\n"],"names":["runWatch","DEBOUNCE_MS","isCallerAbort","signal","reason","aborted","Object","is","errorMessage","Error","message","String","cfg","opts","onEvent","debounceMs","reconcileIntervalMs","claimIntervalMs","rootDir","configDir","claimFailure","requestShutdown","claim","claimCloseStarted","closeClaim","watcher","startup","startupNotification","scheduleReconcile","stopping","callerCanceled","startupEvents","initialStore","dbPath","initialWarnings","initialParsed","initialTotal","store","parsed","warnings","running","queuedAlways","tick","startCycle","requestCycle","debounceTimer","reconcileTimer","err","canceledByCaller","closeError","heartbeatIntervalMs","WATCH_HEARTBEAT_INTERVAL_MS","Number","isFinite","SenseError","Math","min","baseDir","startWatchClaim","force","close","throwIfAborted","fsWatch","recursive","_event","filename","shouldReconcileWatchNotification","on","openStore","length","docCount","push","total","alwaysEmit","type","then","again","clearTimeout","setTimeout","setInterval","guardedTick","Promise","resolveShutdown","rejectShutdown","shutdownReason","shutdown","shutdownTask","finalError","finalStore","process","off","stopNormally","removeEventListener","stopForCaller","clearInterval","AggregateError","undefined","once","addEventListener","event"],"mappings":";;;;+BAoCsBA;;;eAAAA;;;sBApCW;wBAEN;6BACC;mCACqB;uBACb;4BACyB;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAE7D,4FAA4F;AAC5F,6DAA6D;AAC7D,IAAMC,cAAc;AAcpB,SAASC,cAAcC,MAA+B,EAAEC,MAAe;IACrE,OAAOD,CAAAA,mBAAAA,6BAAAA,OAAQE,OAAO,MAAK,QAAQC,OAAOC,EAAE,CAACH,QAAQD,OAAOC,MAAM;AACpE;AAEA,SAASI,aAAaJ,MAAe;IACnC,OAAOA,AAAM,YAANA,QAAkBK,SAAQL,OAAOM,OAAO,GAAGC,OAAOP;AAC3D;AAMO,SAAeJ;wCAASY,GAAmB;YAAEC,MAElCA,eACGA,kBACSA,2BAIZD,cACEA,gBAG6BC,aAX3CA,cACEC,SACAC,YACAC,qBAGAC,iBACAC,SACAC,WACFC,cACAC,iBACEC,OAIFC,mBACEC,YAIFC,SACAC,SACAC,qBACAC,mBACAC,UACAC,gBAEFjB,eAsBAA,eAiBiCA,eA1B3BkB,eACoF,MAA3EC,cAAcC,QAAkBC,iBAAyBC,eACpEC,oBAeFvB,eAOAA,eANoC,OAA5BwB,OAAOC,QAAQC,oBAarBC,SACAC,cAEEC,MAgBAC,YAWAC,cASFC,eAUEC,gBAmECC,KACDC,kBAKGC;;;;;oBAlMuCpC,OAAAA,oEAAqB,CAAC;oBACxE,KAAIA,eAAAA,KAAKV,MAAM,cAAXU,mCAAAA,aAAaR,OAAO,EAAE;;;oBACpBS,WAAUD,gBAAAA,KAAKC,OAAO,cAAZD,2BAAAA,gBAAiB,YAAO;oBAClCE,cAAaF,mBAAAA,KAAKE,UAAU,cAAfF,8BAAAA,mBAAmBZ;oBAChCe,uBAAsBH,4BAAAA,KAAKqC,mBAAmB,cAAxBrC,uCAAAA,4BAA4BsC,yCAA2B;oBACnF,IAAI,CAACC,OAAOC,QAAQ,CAACrC,wBAAwBA,uBAAuB,GAAG,MAAM,IAAIsC,oBAAU,CAAC,kBAAkB;oBAC9G,gGAAgG;oBAC1FrC,kBAAkBsC,KAAKC,GAAG,CAACxC,qBAAqBmC,yCAA2B;oBAC3EjC,WAAUN,eAAAA,IAAIM,OAAO,cAAXN,0BAAAA,eAAeA,IAAI6C,OAAO;oBACpCtC,aAAYP,iBAAAA,IAAIO,SAAS,cAAbP,4BAAAA,iBAAiBM;oBAC/BE,eAA6B;oBAEnB;;wBAAMsC,IAAAA,6BAAe,EAACvC,YAAWN,cAAAA,KAAK8C,KAAK,cAAV9C,yBAAAA,cAAc,OAAOI,iBAAiB,SAAC8B;4BACpF3B,yBAAAA,0BAAAA,eAAAA,eAAiB2B;4BACjB1B,4BAAAA,sCAAAA,gBAAkB0B;wBACpB;;;oBAHMzB,QAAQ;oBAIVC,oBAAoB;oBAClBC,aAAa;wBACjBD,oBAAoB;wBACpB,OAAOD,MAAMsC,KAAK;oBACpB;oBAEIlC,UAAU;oBACVC,sBAAsB;oBAEtBE,WAAW;oBACXC,iBAAiB;;;;;;;;;qBAEnBjB,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBAC3B,2FAA2F;oBAC3F,mEAAmE;oBACnEpC,UAAUqC,IAAAA,aAAO,EAAC5C,SAAS;wBAAE6C,WAAW;oBAAK,GAAG,SAACC,QAAQC;wBACvD,IAAI,CAACC,IAAAA,qDAAgC,EAACD,WAAW;wBACjD,IAAIvC,SAASC,sBAAsB;6BAC9BC,8BAAAA,wCAAAA;oBACP;oBACAH,QAAQ0C,EAAE,CAAC,SAAS,SAACpB;wBACnB3B,yBAAAA,0BAAAA,eAAAA,eAAiB2B;wBACjB1B,4BAAAA,sCAAAA,gBAAkB0B;oBACpB;oBAEMhB;oBACoF;;wBAAMqC,IAAAA,kBAAS,EAACxD,KAAK;4BAAET,QAAQU,KAAKV,MAAM;wBAAC;;;oBAA3C,OAAA,eAA3E6B,eAA2E,KAAlFK,OAAqBJ,SAA6D,KAA7DA,QAAkBC,kBAA2C,KAArDK,UAAmCJ,gBAAkB,KAA1BG;oBAC5DF,eAAe;;;;;;;;;yBAGFF,CAAAA,gBAAgBmC,MAAM,GAAG,KAAKlC,gBAAgB,CAAA,GAA9CD;;;;oBAAkD;;wBAAMoC,IAAAA,iBAAQ,EAACtC;;;2BAAf;;;;;;2BAA+B;;;oBADhG,mFAAmF;oBACnFI;;;;;;oBAEA;;wBAAMJ,aAAa4B,KAAK;;;oBAAxB;;;;;qBAEF/C,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBAC3B,IAAIzC,cAAc,MAAMA;oBAExBW,cAAcwC,IAAI,CAAC;wBAAEjC,QAAQH;wBAAeqC,OAAOpC;wBAAcG,UAAUL;oBAAgB;;;oBAIzFP,sBAAsB;qBACtBd,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBACS;;wBAAMO,IAAAA,kBAAS,EAACxD,KAAK;4BAAET,QAAQU,KAAKV,MAAM;wBAAC;;;oBAA3C,QAAA,eAA5BkC,QAA4B,MAA5BA,OAAOC,SAAqB,MAArBA,QAAQC,WAAa,MAAbA;;;;;;;;;yBAEjBD,CAAAA,SAAS,KAAKC,SAAS8B,MAAM,GAAG,CAAA,GAAhC/B;;;;wBAAmCP,cAAcwC,IAAI;;wBAAGjC,QAAAA;;oBAAe;;wBAAMgC,IAAAA,iBAAQ,EAACjC;;;oBAAnDN,QAAAA;+BAA6ByC,QAAO,qBAAuBjC,WAAAA;;;;;;;;;oBAElG;;wBAAMF,MAAMuB,KAAK;;;oBAAjB;;;;;qBAEF/C,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBAC3B,IAAIzC,cAAc,MAAMA;;;wBACjBO,uBAAuB,GAACd,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAaR,OAAO,GAA5CsB;;;;;;oBACTD,UAAU;oBAEV,6FAA6F;oBAC7F,qFAAqF;oBACjFc,UAAgC;oBAChCC,eAA+B;oBAE7BC,OAAO,cAAO+B;;gCAEoB,MAA5BpC,OAAOC,QAAQC,gBAQhBQ;;;;;;;;;;wCAR6B;;4CAAMqB,IAAAA,kBAAS,EAACxD,KAAK;gDAAET,QAAQU,KAAKV,MAAM;4CAAC;;;wCAA3C,OAAA,eAA5BkC,QAA4B,KAA5BA,OAAOC,SAAqB,KAArBA,QAAQC,WAAa,KAAbA;;;;;;;;;6CAEjBkC,CAAAA,cAAcnC,SAAS,KAAKC,SAAS8B,MAAM,GAAG,CAAA,GAA9CI;;;;;4CACQC,MAAM;4CAAcpC,QAAAA;;wCAAe;;4CAAMgC,IAAAA,iBAAQ,EAACjC;;;wCAA5DvB;kDAAsC0D,QAAO,oBAAuBjC,WAAAA;;;;;;;;;wCAGtE;;4CAAMF,MAAMuB,KAAK;;;wCAAjB;;;;;;;;;;wCAEKb;wCACP,IAAI7C,cAAcW,KAAKV,MAAM,EAAE4C,MAAM;;;wCACrCjC,QAAQ;4CAAE4D,MAAM;4CAAmBhE,SAASF,aAAauC;wCAAK;;;;;;;;;;;wBAElE;;oBAEMJ,aAAa,qBAAC8B;wBAClBjC,UAAUE,KAAK+B,YAAYE,IAAI,CAAC;4BAC9BnC,UAAU;4BACV,IAAIC,iBAAiB,QAAQ,CAACZ,UAAU;gCACtC,IAAM+C,QAAQnC;gCACdA,eAAe;gCACfE,WAAWiC;4BACb;wBACF;oBACF;oBAEMhC,eAAe,sBAAC6B;wBACpB,IAAI5C,UAAU;wBACd,IAAIW,SAAS;4BACXC,eAAeA,gBAAgBgC;4BAC/B;wBACF;wBACA9B,WAAW8B;oBACb;oBAEI5B,gBAAuC;oBAC3C,gGAAgG;oBAChG,mGAAmG;oBACnGjB,oBAAoB;wBAClB,IAAIiB,eAAegC,aAAahC;wBAChCA,gBAAgBiC,WAAW;4BACzBjC,gBAAgB;4BAChBD,aAAa;wBACf,GAAG7B;oBACL;oBACM+B,iBAAiBiC,YACrBC,IAAAA,0BAAW,EACT;;;;;oCAAYpC,aAAa;;;;uBACzB;+BAAMf;wBAERb;oBAGK;;wBAAM,IAAIiE,QAAc,SAACC,iBAAiBC;gCA2CtCtE;4BA1CT,IAAIuE,iBAA+B;4BACnC,IAAMC,WAAW,kBAACjF;oCAAgB4C,oFAAmB;gCACnDoC,2BAAAA,4BAAAA,iBAAAA,iBAAmBhF,mBAAAA,oBAAAA,SAAU;gCAC7B0B,mBAAAA,iBAAmBkB;gCACnB,IAAInB,UAAU;gCACdA,WAAW;gCACX,IAAMyD,eAAe,AAAC;;4CAGpBzE,cAKI0E,YAK8B,MAAfC,YAGVzC,KAKAA;;;;oDApBT0C,QAAQC,GAAG,CAAC,UAAUC;oDACtBF,QAAQC,GAAG,CAAC,WAAWC;qDACvB9E,eAAAA,KAAKV,MAAM,cAAXU,mCAAAA,aAAa+E,mBAAmB,CAAC,SAASC;oDAC1CC,cAAchD;oDACd,IAAID,eAAegC,aAAahC;oDAChCpB,oBAAAA,8BAAAA,QAASmC,KAAK;oDACd;;wDAAMpB;;;oDAAN;;;;;;;;;yDAKM,CAAA,CAAC4C,kBAAkB,CAACtD,cAAa,GAAjC;;;;oDAC4B;;wDAAMsC,IAAAA,kBAAS,EAACxD;;;oDAAhB,OAAA,eAAf4E,aAAe,KAAtBnD;oDACR;;wDAAMmD,WAAW5B,KAAK;;;oDAAtB;;;;;;;;oDAEKb;oDACPwC,aAAaxC;;;;;;;;;;;;oDAGb;;wDAAMvB;;;oDAAN;;;;;;oDACOuB;oDACP,IAAIwC,YAAY,MAAM,IAAIQ;wDAAgBR;wDAAYxC;uDAAM;oDAC5D,MAAMA;;oDAER,IAAIwC,YAAY,MAAMA;oDACtB,IAAIH,gBAAgB,MAAMA;;;;;;oCAC5B;;gCACAE,aAAaX,IAAI,CAACO,iBAAiBC;4BACrC;4BACA,IAAMQ,eAAe;uCAAMN;;4BAC3B,IAAMQ,gBAAgB;uCAAMR,SAASW,WAAW;;4BAChD3E,kBAAkBgE;4BAClBI,QAAQQ,IAAI,CAAC,UAAUN;4BACvBF,QAAQQ,IAAI,CAAC,WAAWN;4BACxB,IAAIvE,cAAciE,SAASjE;iCACtB,KAAIP,eAAAA,KAAKV,MAAM,cAAXU,mCAAAA,aAAaR,OAAO,EAAEwF;iCAC1B;oCACHhF;iCAAAA,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAaqF,gBAAgB,CAAC,SAASL,eAAe;oCAAEI,MAAM;gCAAK;gCACnE,IAAI;oCACFnF,QAAQ;wCAAE4D,MAAM;wCAAWxD,SAAAA;wCAASe,QAAAA;oCAAO;oCAC3C,IAAI,CAACJ,UAAU;4CACR,kCAAA,2BAAA;;4CAAL,QAAK,YAAeE,kCAAf,SAAA,6BAAA,QAAA,yBAAA,iCAA8B;gDAA9B,IAAMoE,QAAN;gDACH,IAAItE,UAAU;gDACd,IAAIsE,MAAM5D,QAAQ,CAAC8B,MAAM,GAAG,KAAK8B,MAAM7D,MAAM,GAAG,GAAGxB,QAAQ;oDAAE4D,MAAM;mDAAiByB;4CACtF;;4CAHK;4CAAA;;;qDAAA,6BAAA;oDAAA;;;oDAAA;0DAAA;;;;oCAIP;gCACF,EAAE,OAAOpD,KAAK;oCACZsC,SAAStC,AAAG,YAAHA,KAAetC,SAAQsC,MAAM,IAAItC,MAAME,OAAOoC;gCACzD;4BACF;wBACF;;;oBA1DA;;wBAAO;;;oBA2DAA;oBACDC,mBAAmB9C,cAAcW,KAAKV,MAAM,EAAE4C;oBACpDtB,oBAAAA,8BAAAA,QAASmC,KAAK;oBACd,IAAIrC,mBAAmB,MAAMwB;;;;;;;;;oBAE3B;;wBAAMvB;;;oBAAN;;;;;;oBACOyB;oBACP,MAAM,IAAI8C;wBAAgBhD;wBAAKE;uBAAa;;oBAE9C,IAAID,kBAAkB;;;oBACtB,MAAMD;;oBAEN1B,kBAAkB2E;;;;;;;;;;IAEtB"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/watch.ts"],"sourcesContent":["import { watch as fsWatch } from 'node:fs';\nimport type { ResolvedConfig } from './config/index.ts';\nimport { SenseError } from './errors.ts';\nimport { guardedTick } from './lib/guarded-tick.ts';\nimport { shouldReconcileWatchNotification } from './lib/watch-notification.ts';\nimport { docCount, openStore } from './store/index.ts';\nimport { startWatchClaim, WATCH_HEARTBEAT_INTERVAL_MS } from './watch-claim.ts';\n\n// Watch prepares the index for snapshot queries; filesystem events request debounced builds\n// and periodic reconciliation recovers missed notifications.\nconst DEBOUNCE_MS = 200;\n\nexport type WatchEvent = { type: 'started'; rootDir: string; dbPath: string } | { type: 'reconciled'; parsed: number; total: number; warnings: string[] } | { type: 'reconcile-error'; message: string };\n\n/** Options for the root {@link runWatch} API. */\nexport interface WatchOptions {\n  force?: boolean;\n  onEvent?: (event: WatchEvent) => void;\n  /** Caller abort cancels provider work, skips final preparation, and resolves after cleanup. */\n  signal?: AbortSignal;\n  debounceMs?: number;\n  heartbeatIntervalMs?: number;\n}\n\nfunction isCallerAbort(signal: AbortSignal | undefined, reason: unknown): boolean {\n  return signal?.aborted === true && Object.is(reason, signal.reason);\n}\n\nfunction errorMessage(reason: unknown): string {\n  return reason instanceof Error ? reason.message : String(reason);\n}\n\n/**\n * Runs in the foreground. Caller abort waits for active native work and cleanup; SIGINT/SIGTERM\n * also run one final freshness pass. Cleanup failures reject either shutdown path.\n */\nexport async function runWatch(cfg: ResolvedConfig, opts: WatchOptions = {}): Promise<void> {\n  if (opts.signal?.aborted) return;\n  const onEvent = opts.onEvent ?? (() => {});\n  const debounceMs = opts.debounceMs ?? DEBOUNCE_MS;\n  const reconcileIntervalMs = opts.heartbeatIntervalMs ?? WATCH_HEARTBEAT_INTERVAL_MS;\n  if (!Number.isFinite(reconcileIntervalMs) || reconcileIntervalMs <= 0) throw new SenseError('CONFIG_INVALID', 'watch heartbeatIntervalMs must be a finite positive number');\n  // Reconciliation may be parked for a caller, but live ownership must renew before fixed expiry.\n  const claimIntervalMs = Math.min(reconcileIntervalMs, WATCH_HEARTBEAT_INTERVAL_MS);\n  const rootDir = cfg.rootDir ?? cfg.baseDir;\n  const configDir = cfg.configDir ?? rootDir;\n  let claimFailure: Error | null = null;\n  let requestShutdown: ((reason: Error) => void) | undefined;\n  const claim = await startWatchClaim(configDir, opts.force ?? false, claimIntervalMs, (err) => {\n    claimFailure ??= err;\n    requestShutdown?.(err);\n  });\n  let claimCloseStarted = false;\n  const closeClaim = (): Promise<void> => {\n    claimCloseStarted = true;\n    return claim.close();\n  };\n  let watcher: ReturnType<typeof fsWatch> | undefined;\n  let startup = true;\n  let startupNotification = false;\n  let scheduleReconcile: (() => void) | undefined;\n  let stopping = false;\n  let callerCanceled = false;\n  try {\n    opts.signal?.throwIfAborted();\n    // Subscribe before the first build. Native watchers have no ready barrier, so startup also\n    // performs an unconditional catch-up pass after the initial build.\n    watcher = fsWatch(rootDir, { recursive: true }, (_event, filename) => {\n      if (!shouldReconcileWatchNotification(filename)) return;\n      if (startup) startupNotification = true;\n      else scheduleReconcile?.();\n    });\n    watcher.on('error', (err) => {\n      claimFailure ??= err;\n      requestShutdown?.(err);\n    });\n\n    const startupEvents: Array<{ parsed: number; total: number; warnings: string[] }> = [];\n    const { store: initialStore, dbPath, warnings: initialWarnings, parsed: initialParsed } = await openStore(cfg, { signal: opts.signal });\n    let initialTotal = 0;\n    try {\n      // total is read before close: an event fires only once nothing is held open again.\n      initialTotal = initialWarnings.length > 0 || initialParsed > 0 ? await docCount(initialStore) : 0;\n    } finally {\n      await initialStore.close();\n    }\n    opts.signal?.throwIfAborted();\n    if (claimFailure) throw claimFailure;\n\n    startupEvents.push({ parsed: initialParsed, total: initialTotal, warnings: initialWarnings });\n    // Always reopen once after subscribing, then drain notifications observed while startup was\n    // in progress. Each pass closes before the next one starts, so no stores overlap.\n    do {\n      startupNotification = false;\n      opts.signal?.throwIfAborted();\n      const { store, parsed, warnings } = await openStore(cfg, { signal: opts.signal });\n      try {\n        if (parsed > 0 || warnings.length > 0) startupEvents.push({ parsed, total: await docCount(store), warnings });\n      } finally {\n        await store.close();\n      }\n      opts.signal?.throwIfAborted();\n      if (claimFailure) throw claimFailure;\n    } while (startupNotification && !opts.signal?.aborted);\n    startup = false;\n\n    // At most one open+reconcile+close cycle runs at a time; a trigger that arrives mid-cycle is\n    // coalesced into a single rerun instead of opening a second, overlapping connection.\n    let running: Promise<void> | null = null;\n    let queuedAlways: boolean | null = null;\n\n    const tick = async (alwaysEmit: boolean): Promise<void> => {\n      try {\n        const { store, parsed, warnings } = await openStore(cfg, { signal: opts.signal });\n        try {\n          if (alwaysEmit || parsed > 0 || warnings.length > 0) {\n            onEvent({ type: 'reconciled', parsed, total: await docCount(store), warnings });\n          }\n        } finally {\n          await store.close();\n        }\n      } catch (err) {\n        if (isCallerAbort(opts.signal, err)) return;\n        onEvent({ type: 'reconcile-error', message: errorMessage(err) });\n      }\n    };\n\n    const startCycle = (alwaysEmit: boolean) => {\n      running = tick(alwaysEmit).then(() => {\n        running = null;\n        if (queuedAlways !== null && !stopping) {\n          const again = queuedAlways;\n          queuedAlways = null;\n          startCycle(again);\n        }\n      });\n    };\n\n    const requestCycle = (alwaysEmit: boolean) => {\n      if (stopping) return;\n      if (running) {\n        queuedAlways = queuedAlways || alwaysEmit;\n        return;\n      }\n      startCycle(alwaysEmit);\n    };\n\n    let debounceTimer: NodeJS.Timeout | null = null;\n    // Ignore our own state files, or cache and claim writes would retrigger reconciliation forever.\n    // An unresolvable filename reconciles because parsing nothing costs less than missing a real edit.\n    scheduleReconcile = () => {\n      if (debounceTimer) clearTimeout(debounceTimer);\n      debounceTimer = setTimeout(() => {\n        debounceTimer = null;\n        requestCycle(true);\n      }, debounceMs);\n    };\n    const reconcileTimer = setInterval(\n      guardedTick(\n        async () => requestCycle(false),\n        () => stopping\n      ),\n      reconcileIntervalMs\n    );\n\n    return await new Promise<void>((resolveShutdown, rejectShutdown) => {\n      let shutdownReason: Error | null = null;\n      const shutdown = (reason?: Error, canceledByCaller = false) => {\n        shutdownReason ??= reason ?? null;\n        callerCanceled ||= canceledByCaller;\n        if (stopping) return;\n        stopping = true;\n        const shutdownTask = (async () => {\n          process.off('SIGINT', stopNormally);\n          process.off('SIGTERM', stopNormally);\n          opts.signal?.removeEventListener('abort', stopForCaller);\n          clearInterval(reconcileTimer);\n          if (debounceTimer) clearTimeout(debounceTimer);\n          watcher?.close();\n          await running;\n          let finalError: unknown;\n          try {\n            // A normal stop performs one final freshness pass. Ownership loss drains existing work\n            // but starts nothing new because the replacement watcher now owns that responsibility.\n            if (!shutdownReason && !callerCanceled) {\n              const { store: finalStore } = await openStore(cfg);\n              await finalStore.close();\n            }\n          } catch (err) {\n            finalError = err;\n          }\n          try {\n            await closeClaim();\n          } catch (err) {\n            if (finalError) throw new AggregateError([finalError, err], 'watch shutdown and claim cleanup both failed');\n            throw err;\n          }\n          if (finalError) throw finalError;\n          if (shutdownReason) throw shutdownReason;\n        })();\n        shutdownTask.then(resolveShutdown, rejectShutdown);\n      };\n      const stopNormally = () => shutdown();\n      const stopForCaller = () => shutdown(undefined, true);\n      requestShutdown = shutdown;\n      process.once('SIGINT', stopNormally);\n      process.once('SIGTERM', stopNormally);\n      if (claimFailure) shutdown(claimFailure);\n      else if (opts.signal?.aborted) stopForCaller();\n      else {\n        opts.signal?.addEventListener('abort', stopForCaller, { once: true });\n        try {\n          onEvent({ type: 'started', rootDir, dbPath });\n          if (!stopping) {\n            for (const event of startupEvents) {\n              if (stopping) break;\n              if (event.warnings.length > 0 || event.parsed > 0) onEvent({ type: 'reconciled', ...event });\n            }\n          }\n        } catch (err) {\n          shutdown(err instanceof Error ? err : new Error(String(err)));\n        }\n      }\n    });\n  } catch (err) {\n    const canceledByCaller = isCallerAbort(opts.signal, err);\n    watcher?.close();\n    if (claimCloseStarted) throw err;\n    try {\n      await closeClaim();\n    } catch (closeError) {\n      throw new AggregateError([err, closeError], 'watch setup and claim cleanup both failed');\n    }\n    if (canceledByCaller) return;\n    throw err;\n  } finally {\n    requestShutdown = undefined;\n  }\n}\n"],"names":["runWatch","DEBOUNCE_MS","isCallerAbort","signal","reason","aborted","Object","is","errorMessage","Error","message","String","cfg","opts","onEvent","debounceMs","reconcileIntervalMs","claimIntervalMs","rootDir","configDir","claimFailure","requestShutdown","claim","claimCloseStarted","closeClaim","watcher","startup","startupNotification","scheduleReconcile","stopping","callerCanceled","startupEvents","initialStore","dbPath","initialWarnings","initialParsed","initialTotal","store","parsed","warnings","running","queuedAlways","tick","startCycle","requestCycle","debounceTimer","reconcileTimer","err","canceledByCaller","closeError","heartbeatIntervalMs","WATCH_HEARTBEAT_INTERVAL_MS","Number","isFinite","SenseError","Math","min","baseDir","startWatchClaim","force","close","throwIfAborted","fsWatch","recursive","_event","filename","shouldReconcileWatchNotification","on","openStore","length","docCount","push","total","alwaysEmit","type","then","again","clearTimeout","setTimeout","setInterval","guardedTick","Promise","resolveShutdown","rejectShutdown","shutdownReason","shutdown","shutdownTask","finalError","finalStore","process","off","stopNormally","removeEventListener","stopForCaller","clearInterval","AggregateError","undefined","once","addEventListener","event"],"mappings":";;;;+BAoCsBA;;;eAAAA;;;sBApCW;wBAEN;6BACC;mCACqB;uBACb;4BACyB;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAE7D,4FAA4F;AAC5F,6DAA6D;AAC7D,IAAMC,cAAc;AAcpB,SAASC,cAAcC,MAA+B,EAAEC,MAAe;IACrE,OAAOD,CAAAA,mBAAAA,6BAAAA,OAAQE,OAAO,MAAK,QAAQC,OAAOC,EAAE,CAACH,QAAQD,OAAOC,MAAM;AACpE;AAEA,SAASI,aAAaJ,MAAe;IACnC,OAAOA,AAAM,YAANA,QAAkBK,SAAQL,OAAOM,OAAO,GAAGC,OAAOP;AAC3D;AAMO,SAAeJ;wCAASY,GAAmB;YAAEC,MAElCA,eACGA,kBACSA,2BAIZD,cACEA,gBAG6BC,aAX3CA,cACEC,SACAC,YACAC,qBAGAC,iBACAC,SACAC,WACFC,cACAC,iBACEC,OAIFC,mBACEC,YAIFC,SACAC,SACAC,qBACAC,mBACAC,UACAC,gBAEFjB,eAsBAA,eAiBiCA,eA1B3BkB,eACoF,MAA3EC,cAAcC,QAAkBC,iBAAyBC,eACpEC,oBAeFvB,eAOAA,eANoC,OAA5BwB,OAAOC,QAAQC,oBAarBC,SACAC,cAEEC,MAgBAC,YAWAC,cASFC,eAUEC,gBAmECC,KACDC,kBAKGC;;;;;oBAlMuCpC,OAAAA,oEAAqB,CAAC;oBACxE,KAAIA,eAAAA,KAAKV,MAAM,cAAXU,mCAAAA,aAAaR,OAAO,EAAE;;;oBACpBS,WAAUD,gBAAAA,KAAKC,OAAO,cAAZD,2BAAAA,gBAAiB,YAAO;oBAClCE,cAAaF,mBAAAA,KAAKE,UAAU,cAAfF,8BAAAA,mBAAmBZ;oBAChCe,uBAAsBH,4BAAAA,KAAKqC,mBAAmB,cAAxBrC,uCAAAA,4BAA4BsC,yCAA2B;oBACnF,IAAI,CAACC,OAAOC,QAAQ,CAACrC,wBAAwBA,uBAAuB,GAAG,MAAM,IAAIsC,oBAAU,CAAC,kBAAkB;oBAC9G,gGAAgG;oBAC1FrC,kBAAkBsC,KAAKC,GAAG,CAACxC,qBAAqBmC,yCAA2B;oBAC3EjC,WAAUN,eAAAA,IAAIM,OAAO,cAAXN,0BAAAA,eAAeA,IAAI6C,OAAO;oBACpCtC,aAAYP,iBAAAA,IAAIO,SAAS,cAAbP,4BAAAA,iBAAiBM;oBAC/BE,eAA6B;oBAEnB;;wBAAMsC,IAAAA,6BAAe,EAACvC,YAAWN,cAAAA,KAAK8C,KAAK,cAAV9C,yBAAAA,cAAc,OAAOI,iBAAiB,SAAC8B;4BACpF3B,yBAAAA,0BAAAA,eAAAA,eAAiB2B;4BACjB1B,4BAAAA,sCAAAA,gBAAkB0B;wBACpB;;;oBAHMzB,QAAQ;oBAIVC,oBAAoB;oBAClBC,aAAa;wBACjBD,oBAAoB;wBACpB,OAAOD,MAAMsC,KAAK;oBACpB;oBAEIlC,UAAU;oBACVC,sBAAsB;oBAEtBE,WAAW;oBACXC,iBAAiB;;;;;;;;;qBAEnBjB,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBAC3B,2FAA2F;oBAC3F,mEAAmE;oBACnEpC,UAAUqC,IAAAA,aAAO,EAAC5C,SAAS;wBAAE6C,WAAW;oBAAK,GAAG,SAACC,QAAQC;wBACvD,IAAI,CAACC,IAAAA,qDAAgC,EAACD,WAAW;wBACjD,IAAIvC,SAASC,sBAAsB;6BAC9BC,8BAAAA,wCAAAA;oBACP;oBACAH,QAAQ0C,EAAE,CAAC,SAAS,SAACpB;wBACnB3B,yBAAAA,0BAAAA,eAAAA,eAAiB2B;wBACjB1B,4BAAAA,sCAAAA,gBAAkB0B;oBACpB;oBAEMhB;oBACoF;;wBAAMqC,IAAAA,kBAAS,EAACxD,KAAK;4BAAET,QAAQU,KAAKV,MAAM;wBAAC;;;oBAA3C,OAAA,eAA3E6B,eAA2E,KAAlFK,OAAqBJ,SAA6D,KAA7DA,QAAkBC,kBAA2C,KAArDK,UAAmCJ,gBAAkB,KAA1BG;oBAC5DF,eAAe;;;;;;;;;yBAGFF,CAAAA,gBAAgBmC,MAAM,GAAG,KAAKlC,gBAAgB,CAAA,GAA9CD;;;;oBAAkD;;wBAAMoC,IAAAA,iBAAQ,EAACtC;;;2BAAf;;;;;;2BAA+B;;;oBADhG,mFAAmF;oBACnFI;;;;;;oBAEA;;wBAAMJ,aAAa4B,KAAK;;;oBAAxB;;;;;qBAEF/C,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBAC3B,IAAIzC,cAAc,MAAMA;oBAExBW,cAAcwC,IAAI,CAAC;wBAAEjC,QAAQH;wBAAeqC,OAAOpC;wBAAcG,UAAUL;oBAAgB;;;oBAIzFP,sBAAsB;qBACtBd,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBACS;;wBAAMO,IAAAA,kBAAS,EAACxD,KAAK;4BAAET,QAAQU,KAAKV,MAAM;wBAAC;;;oBAA3C,QAAA,eAA5BkC,QAA4B,MAA5BA,OAAOC,SAAqB,MAArBA,QAAQC,WAAa,MAAbA;;;;;;;;;yBAEjBD,CAAAA,SAAS,KAAKC,SAAS8B,MAAM,GAAG,CAAA,GAAhC/B;;;;wBAAmCP,cAAcwC,IAAI;;wBAAGjC,QAAAA;;oBAAe;;wBAAMgC,IAAAA,iBAAQ,EAACjC;;;oBAAnDN,QAAAA;+BAA6ByC,QAAO,qBAAuBjC,WAAAA;;;;;;;;;oBAElG;;wBAAMF,MAAMuB,KAAK;;;oBAAjB;;;;;qBAEF/C,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAagD,cAAc;oBAC3B,IAAIzC,cAAc,MAAMA;;;wBACjBO,uBAAuB,GAACd,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAaR,OAAO,GAA5CsB;;;;;;oBACTD,UAAU;oBAEV,6FAA6F;oBAC7F,qFAAqF;oBACjFc,UAAgC;oBAChCC,eAA+B;oBAE7BC,OAAO,cAAO+B;;gCAEoB,MAA5BpC,OAAOC,QAAQC,gBAQhBQ;;;;;;;;;;wCAR6B;;4CAAMqB,IAAAA,kBAAS,EAACxD,KAAK;gDAAET,QAAQU,KAAKV,MAAM;4CAAC;;;wCAA3C,OAAA,eAA5BkC,QAA4B,KAA5BA,OAAOC,SAAqB,KAArBA,QAAQC,WAAa,KAAbA;;;;;;;;;6CAEjBkC,CAAAA,cAAcnC,SAAS,KAAKC,SAAS8B,MAAM,GAAG,CAAA,GAA9CI;;;;;4CACQC,MAAM;4CAAcpC,QAAAA;;wCAAe;;4CAAMgC,IAAAA,iBAAQ,EAACjC;;;wCAA5DvB;kDAAsC0D,QAAO,oBAAuBjC,WAAAA;;;;;;;;;wCAGtE;;4CAAMF,MAAMuB,KAAK;;;wCAAjB;;;;;;;;;;wCAEKb;wCACP,IAAI7C,cAAcW,KAAKV,MAAM,EAAE4C,MAAM;;;wCACrCjC,QAAQ;4CAAE4D,MAAM;4CAAmBhE,SAASF,aAAauC;wCAAK;;;;;;;;;;;wBAElE;;oBAEMJ,aAAa,qBAAC8B;wBAClBjC,UAAUE,KAAK+B,YAAYE,IAAI,CAAC;4BAC9BnC,UAAU;4BACV,IAAIC,iBAAiB,QAAQ,CAACZ,UAAU;gCACtC,IAAM+C,QAAQnC;gCACdA,eAAe;gCACfE,WAAWiC;4BACb;wBACF;oBACF;oBAEMhC,eAAe,sBAAC6B;wBACpB,IAAI5C,UAAU;wBACd,IAAIW,SAAS;4BACXC,eAAeA,gBAAgBgC;4BAC/B;wBACF;wBACA9B,WAAW8B;oBACb;oBAEI5B,gBAAuC;oBAC3C,gGAAgG;oBAChG,mGAAmG;oBACnGjB,oBAAoB;wBAClB,IAAIiB,eAAegC,aAAahC;wBAChCA,gBAAgBiC,WAAW;4BACzBjC,gBAAgB;4BAChBD,aAAa;wBACf,GAAG7B;oBACL;oBACM+B,iBAAiBiC,YACrBC,IAAAA,0BAAW,EACT;;;;;oCAAYpC,aAAa;;;;uBACzB;+BAAMf;wBAERb;oBAGK;;wBAAM,IAAIiE,QAAc,SAACC,iBAAiBC;gCA2CtCtE;4BA1CT,IAAIuE,iBAA+B;4BACnC,IAAMC,WAAW,kBAACjF;oCAAgB4C,oFAAmB;gCACnDoC,2BAAAA,4BAAAA,iBAAAA,iBAAmBhF,mBAAAA,oBAAAA,SAAU;gCAC7B0B,mBAAAA,iBAAmBkB;gCACnB,IAAInB,UAAU;gCACdA,WAAW;gCACX,IAAMyD,eAAe,AAAC;;4CAGpBzE,cAKI0E,YAK8B,MAAfC,YAGVzC,KAKAA;;;;oDApBT0C,QAAQC,GAAG,CAAC,UAAUC;oDACtBF,QAAQC,GAAG,CAAC,WAAWC;qDACvB9E,eAAAA,KAAKV,MAAM,cAAXU,mCAAAA,aAAa+E,mBAAmB,CAAC,SAASC;oDAC1CC,cAAchD;oDACd,IAAID,eAAegC,aAAahC;oDAChCpB,oBAAAA,8BAAAA,QAASmC,KAAK;oDACd;;wDAAMpB;;;oDAAN;;;;;;;;;yDAKM,CAAA,CAAC4C,kBAAkB,CAACtD,cAAa,GAAjC;;;;oDAC4B;;wDAAMsC,IAAAA,kBAAS,EAACxD;;;oDAAhB,OAAA,eAAf4E,aAAe,KAAtBnD;oDACR;;wDAAMmD,WAAW5B,KAAK;;;oDAAtB;;;;;;;;oDAEKb;oDACPwC,aAAaxC;;;;;;;;;;;;oDAGb;;wDAAMvB;;;oDAAN;;;;;;oDACOuB;oDACP,IAAIwC,YAAY,MAAM,IAAIQ;wDAAgBR;wDAAYxC;uDAAM;oDAC5D,MAAMA;;oDAER,IAAIwC,YAAY,MAAMA;oDACtB,IAAIH,gBAAgB,MAAMA;;;;;;oCAC5B;;gCACAE,aAAaX,IAAI,CAACO,iBAAiBC;4BACrC;4BACA,IAAMQ,eAAe;uCAAMN;;4BAC3B,IAAMQ,gBAAgB;uCAAMR,SAASW,WAAW;;4BAChD3E,kBAAkBgE;4BAClBI,QAAQQ,IAAI,CAAC,UAAUN;4BACvBF,QAAQQ,IAAI,CAAC,WAAWN;4BACxB,IAAIvE,cAAciE,SAASjE;iCACtB,KAAIP,eAAAA,KAAKV,MAAM,cAAXU,mCAAAA,aAAaR,OAAO,EAAEwF;iCAC1B;oCACHhF;iCAAAA,gBAAAA,KAAKV,MAAM,cAAXU,oCAAAA,cAAaqF,gBAAgB,CAAC,SAASL,eAAe;oCAAEI,MAAM;gCAAK;gCACnE,IAAI;oCACFnF,QAAQ;wCAAE4D,MAAM;wCAAWxD,SAAAA;wCAASe,QAAAA;oCAAO;oCAC3C,IAAI,CAACJ,UAAU;4CACR,kCAAA,2BAAA;;4CAAL,QAAK,YAAeE,kCAAf,SAAA,6BAAA,QAAA,yBAAA,iCAA8B;gDAA9B,IAAMoE,QAAN;gDACH,IAAItE,UAAU;gDACd,IAAIsE,MAAM5D,QAAQ,CAAC8B,MAAM,GAAG,KAAK8B,MAAM7D,MAAM,GAAG,GAAGxB,QAAQ;oDAAE4D,MAAM;mDAAiByB;4CACtF;;4CAHK;4CAAA;;;qDAAA,6BAAA;oDAAA;;;oDAAA;0DAAA;;;;oCAIP;gCACF,EAAE,OAAOpD,KAAK;oCACZsC,SAAStC,AAAG,YAAHA,KAAetC,SAAQsC,MAAM,IAAItC,MAAME,OAAOoC;gCACzD;4BACF;wBACF;;;oBA1DA;;wBAAO;;;oBA2DAA;oBACDC,mBAAmB9C,cAAcW,KAAKV,MAAM,EAAE4C;oBACpDtB,oBAAAA,8BAAAA,QAASmC,KAAK;oBACd,IAAIrC,mBAAmB,MAAMwB;;;;;;;;;oBAE3B;;wBAAMvB;;;oBAAN;;;;;;oBACOyB;oBACP,MAAM,IAAI8C;wBAAgBhD;wBAAKE;uBAAa;;oBAE9C,IAAID,kBAAkB;;;oBACtB,MAAMD;;oBAEN1B,kBAAkB2E;;;;;;;;;;IAEtB"}

package/dist/cjs/workers/parse.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/workers/parse.ts"],"sourcesContent":["import Tinypool from 'tinypool';\nimport type { Config, FeatureName } from '../config/index.ts';\nimport { FEATURES } from '../features/index.ts';\nimport type { ParsedDoc } from '../scan/index.ts';\nimport { parseFile } from '../scan/index.ts';\nimport type { FileStat } from '../scan/list.ts';\nimport { featuresForFile } from '../scan/reparse.ts';\nimport type { WorkerErrorPayload } from '../scan/worker-error.ts';\nimport { serializeError } from '../scan/worker-error.ts';\n\n// Constant for the whole dispatch, so it crosses once per worker instead of once per task. A\n// Feature carries closures and cannot cross the thread boundary; its name can, and the registry here resolves it back.\nexport interface ParseWorkerData {\n  cfg: Config;\n  featureNames: FeatureName[];\n}\n\n// Each task is a small file batch. Results stay per-file so one failure does not stop the\n// remaining files in that batch, and no token tree crosses the thread boundary.\nexport type ParseTask = FileStat[];\n\n// parseMs is the worker's own hrtime for parseFile, excluding dispatch and the return clone --\n// what separates it from the pool's dispatch-to-drain `parse` stage.\nexport type ParseFileResult = { ok: true; doc: ParsedDoc; warnings: string[]; parseMs: number } | { ok: false; error: WorkerErrorPayload };\nexport type ParseTaskResult = ParseFileResult[];\n\n// Read once per worker, not per task.\nconst { cfg, featureNames } = Tinypool.workerData as ParseWorkerData;\n// Filtering the registry (rather than mapping the names) keeps registry order, which is the\n// order `extracted` keys land in on the serial path.\nconst selected = FEATURES.filter((feature) => featureNames.includes(feature.name));\n\nexport default function parseTask(files: ParseTask): ParseTaskResult {\n  return files.map((file): ParseFileResult => {\n    try {\n      const start = process.hrtime.bigint();\n      const { doc, warnings } = parseFile(file, featuresForFile(selected, cfg, file), cfg);\n      const parseMs = Number(process.hrtime.bigint() - start) / 1e6;\n      return { ok: true, doc, warnings, parseMs };\n    } catch (err) {\n      return { ok: false, error: serializeError(err) };\n    }\n  });\n}\n"],"names":["parseTask","Tinypool","workerData","cfg","featureNames","selected","FEATURES","filter","feature","includes","name","files","map","file","start","process","hrtime","bigint","parseFile","featuresForFile","doc","warnings","parseMs","Number","ok","err","error","serializeError"],"mappings":";;;;+BAgCA;;;eAAwBA;;;+DAhCH;uBAEI;wBAEC;yBAEM;6BAED;;;;;;AAkB/B,sCAAsC;AACtC,IAA8BC,uBAAAA,iBAAQ,CAACC,UAAU,EAAzCC,MAAsBF,qBAAtBE,KAAKC,eAAiBH,qBAAjBG;AACb,4FAA4F;AAC5F,qDAAqD;AACrD,IAAMC,WAAWC,iBAAQ,CAACC,MAAM,CAAC,SAACC;WAAYJ,aAAaK,QAAQ,CAACD,QAAQE,IAAI;;AAEjE,SAASV,UAAUW,KAAgB;IAChD,OAAOA,MAAMC,GAAG,CAAC,SAACC;QAChB,IAAI;YACF,IAAMC,QAAQC,QAAQC,MAAM,CAACC,MAAM;YACnC,IAA0BC,aAAAA,IAAAA,mBAAS,EAACL,MAAMM,IAAAA,0BAAe,EAACd,UAAUF,KAAKU,OAAOV,MAAxEiB,MAAkBF,WAAlBE,KAAKC,WAAaH,WAAbG;YACb,IAAMC,UAAUC,OAAOR,QAAQC,MAAM,CAACC,MAAM,KAAKH,SAAS;YAC1D,OAAO;gBAAEU,IAAI;gBAAMJ,KAAAA;gBAAKC,UAAAA;gBAAUC,SAAAA;YAAQ;QAC5C,EAAE,OAAOG,KAAK;YACZ,OAAO;gBAAED,IAAI;gBAAOE,OAAOC,IAAAA,6BAAc,EAACF;YAAK;QACjD;IACF;AACF"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/workers/parse.ts"],"sourcesContent":["import Tinypool from 'tinypool';\nimport type { Config, FeatureName } from '../config/index.ts';\nimport { FEATURES } from '../features/index.ts';\nimport type { ParsedDoc } from '../scan/index.ts';\nimport { parseFile } from '../scan/index.ts';\nimport type { FileStat } from '../scan/list.ts';\nimport { featuresForFile } from '../scan/reparse.ts';\nimport type { WorkerErrorPayload } from '../scan/worker-error.ts';\nimport { serializeError } from '../scan/worker-error.ts';\n\n// Constant for the whole dispatch, so it crosses once per worker instead of once per task. A\n// Feature carries closures and cannot cross the thread boundary; its name can, and the registry here resolves it back.\nexport interface ParseWorkerData {\n  cfg: Config;\n  featureNames: FeatureName[];\n}\n\n// Each task is a small file batch. Results stay per-file so one failure does not stop the\n// remaining files in that batch, and no token tree crosses the thread boundary.\nexport type ParseTask = FileStat[];\n\n// parseMs is the worker's own hrtime for parseFile, excluding dispatch and the return clone --\n// what separates it from the pool's dispatch-to-drain `parse` stage.\nexport type ParseFileResult = { ok: true; doc: ParsedDoc; warnings: string[]; parseMs: number } | { ok: false; error: WorkerErrorPayload };\nexport type ParseTaskResult = ParseFileResult[];\n\n// Read once per worker, not per task.\nconst { cfg, featureNames } = Tinypool.workerData as ParseWorkerData;\n// Filtering the registry (rather than mapping the names) keeps registry order, which is the\n// order `extracted` keys land in on the serial path.\nconst selected = FEATURES.filter((feature) => featureNames.includes(feature.name));\n\nexport default function parseTask(files: ParseTask): ParseTaskResult {\n  return files.map((file): ParseFileResult => {\n    try {\n      const start = process.hrtime.bigint();\n      const { doc, warnings } = parseFile(file, featuresForFile(selected, cfg, file), cfg);\n      const parseMs = Number(process.hrtime.bigint() - start) / 1e6;\n      return { ok: true, doc, warnings, parseMs };\n    } catch (err) {\n      return { ok: false, error: serializeError(err) };\n    }\n  });\n}\n"],"names":["parseTask","Tinypool","workerData","cfg","featureNames","selected","FEATURES","filter","feature","includes","name","files","map","file","start","process","hrtime","bigint","parseFile","featuresForFile","doc","warnings","parseMs","Number","ok","err","error","serializeError"],"mappings":";;;;+BAgCA;;;eAAwBA;;;+DAhCH;uBAEI;wBAEC;yBAEM;6BAED;;;;;;AAkB/B,sCAAsC;AACtC,IAA8BC,uBAAAA,iBAAQ,CAACC,UAAU,EAAzCC,MAAsBF,qBAAtBE,KAAKC,eAAiBH,qBAAjBG;AACb,4FAA4F;AAC5F,qDAAqD;AACrD,IAAMC,WAAWC,iBAAQ,CAACC,MAAM,CAAC,SAACC;WAAYJ,aAAaK,QAAQ,CAACD,QAAQE,IAAI;;AAEjE,SAASV,UAAUW,KAAgB;IAChD,OAAOA,MAAMC,GAAG,CAAC,SAACC;QAChB,IAAI;YACF,IAAMC,QAAQC,QAAQC,MAAM,CAACC,MAAM;YACnC,IAA0BC,aAAAA,IAAAA,mBAAS,EAACL,MAAMM,IAAAA,0BAAe,EAACd,UAAUF,KAAKU,OAAOV,MAAxEiB,MAAkBF,WAAlBE,KAAKC,WAAaH,WAAbG;YACb,IAAMC,UAAUC,OAAOR,QAAQC,MAAM,CAACC,MAAM,KAAKH,SAAS;YAC1D,OAAO;gBAAEU,IAAI;gBAAMJ,KAAAA;gBAAKC,UAAAA;gBAAUC,SAAAA;YAAQ;QAC5C,EAAE,OAAOG,KAAK;YACZ,OAAO;gBAAED,IAAI;gBAAOE,OAAOC,IAAAA,6BAAc,EAACF;YAAK;QACjD;IACF;AACF"}

package/dist/cjs/workers/watch-heartbeat.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/workers/watch-heartbeat.ts"],"sourcesContent":["import { parentPort, workerData } from 'node:worker_threads';\nimport { SenseError } from '../errors.ts';\nimport { serializeError } from '../scan/worker-error.ts';\nimport { WatchClaimDatabase, type WatchClaimWorkerData, type WatchClaimWorkerMessage } from '../watch-claim.ts';\n\nfunction requireParentPort() {\n  const port = parentPort;\n  if (!port) throw new Error('watch heartbeat worker requires a parent port');\n  return port;\n}\n\nconst port = requireParentPort();\nconst data = workerData as WatchClaimWorkerData;\nlet database: WatchClaimDatabase | undefined;\nlet timer: NodeJS.Timeout | undefined;\nlet stopping = false;\nlet renewing = Promise.resolve();\n\nfunction send(message: WatchClaimWorkerMessage): void {\n  port.postMessage(message);\n}\n\nfunction asError(value: unknown): Error {\n  return value instanceof Error ? value : new Error(String(value));\n}\n\nfunction closeDatabase(): Error | undefined {\n  if (timer) clearInterval(timer);\n  timer = undefined;\n  try {\n    database?.close();\n    database = undefined;\n    return undefined;\n  } catch (err) {\n    return asError(err);\n  }\n}\n\nfunction addFailure(primary: Error, secondary: Error | undefined, message: string): Error {\n  return secondary ? new AggregateError([primary, secondary], message) : primary;\n}\n\nfunction fail(err: unknown): void {\n  if (stopping) return;\n  stopping = true;\n  let failure = asError(err);\n  try {\n    database?.release(data.token);\n  } catch (releaseError) {\n    failure = addFailure(failure, asError(releaseError), 'watch heartbeat and claim release both failed');\n  }\n  const closeError = closeDatabase();\n  failure = addFailure(failure, closeError, 'watch heartbeat and database cleanup both failed');\n  try {\n    send({ type: 'failure', error: serializeError(failure) });\n  } finally {\n    port.close();\n  }\n}\n\nasync function stop(): Promise<void> {\n  if (stopping) return;\n  stopping = true;\n  if (timer) clearInterval(timer);\n  let failure: Error | undefined;\n  try {\n    await renewing;\n    if (!database?.release(data.token)) throw new SenseError('WATCH_ACTIVE', 'watch ownership was replaced by another watcher');\n  } catch (err) {\n    failure = asError(err);\n  } finally {\n    const closeError = closeDatabase();\n    if (closeError) failure = failure ? addFailure(failure, closeError, 'watch heartbeat shutdown and database cleanup both failed') : closeError;\n    try {\n      if (failure) send({ type: 'failure', error: serializeError(failure) });\n      else send({ type: 'stopped' });\n    } finally {\n      port.close();\n    }\n  }\n}\n\nasync function main(): Promise<void> {\n  try {\n    database = new WatchClaimDatabase(data.configDir);\n    await database.acquire(data.token, data.pid, data.force);\n    send({ type: 'acquired' });\n    timer = setInterval(() => {\n      if (stopping) return;\n      renewing = renewing.then(() => {\n        if (!database?.renew(data.token)) throw new SenseError('WATCH_ACTIVE', 'watch ownership was replaced by another watcher');\n      });\n      renewing.catch(fail);\n    }, data.heartbeatIntervalMs);\n    port.on('message', (message: { type?: string }) => {\n      if (message.type === 'stop') void stop();\n    });\n  } catch (err) {\n    fail(err);\n  }\n}\n\nvoid main();\n"],"names":["requireParentPort","port","parentPort","Error","data","workerData","database","timer","stopping","renewing","Promise","resolve","send","message","postMessage","asError","value","String","closeDatabase","clearInterval","undefined","close","err","addFailure","primary","secondary","AggregateError","fail","failure","release","token","releaseError","closeError","type","error","serializeError","stop","SenseError","main","WatchClaimDatabase","configDir","acquire","pid","force","setInterval","then","renew","catch","heartbeatIntervalMs","on"],"mappings":";;;;kCAAuC;wBACZ;6BACI;4BAC6D;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAE5F,SAASA;IACP,IAAMC,OAAOC,8BAAU;IACvB,IAAI,CAACD,MAAM,MAAM,IAAIE,MAAM;IAC3B,OAAOF;AACT;AAEA,IAAMA,OAAOD;AACb,IAAMI,OAAOC,8BAAU;AACvB,IAAIC;AACJ,IAAIC;AACJ,IAAIC,WAAW;AACf,IAAIC,WAAWC,QAAQC,OAAO;AAE9B,SAASC,KAAKC,OAAgC;IAC5CZ,KAAKa,WAAW,CAACD;AACnB;AAEA,SAASE,QAAQC,KAAc;IAC7B,OAAOA,AAAK,YAALA,OAAiBb,SAAQa,QAAQ,IAAIb,MAAMc,OAAOD;AAC3D;AAEA,SAASE;IACP,IAAIX,OAAOY,cAAcZ;IACzBA,QAAQa;IACR,IAAI;QACFd,qBAAAA,+BAAAA,SAAUe,KAAK;QACff,WAAWc;QACX,OAAOA;IACT,EAAE,OAAOE,KAAK;QACZ,OAAOP,QAAQO;IACjB;AACF;AAEA,SAASC,WAAWC,OAAc,EAAEC,SAA4B,EAAEZ,OAAe;IAC/E,OAAOY,YAAY,IAAIC,eAAe;QAACF;QAASC;KAAU,EAAEZ,WAAWW;AACzE;AAEA,SAASG,KAAKL,GAAY;IACxB,IAAId,UAAU;IACdA,WAAW;IACX,IAAIoB,UAAUb,QAAQO;IACtB,IAAI;QACFhB,qBAAAA,+BAAAA,SAAUuB,OAAO,CAACzB,KAAK0B,KAAK;IAC9B,EAAE,OAAOC,cAAc;QACrBH,UAAUL,WAAWK,SAASb,QAAQgB,eAAe;IACvD;IACA,IAAMC,aAAad;IACnBU,UAAUL,WAAWK,SAASI,YAAY;IAC1C,IAAI;QACFpB,KAAK;YAAEqB,MAAM;YAAWC,OAAOC,IAAAA,6BAAc,EAACP;QAAS;IACzD,SAAU;QACR3B,KAAKoB,KAAK;IACZ;AACF;AAEA,SAAee;;YAITR,SAIKN,KAGDU;;;;oBAVR,IAAIxB,UAAU;;;oBACdA,WAAW;oBACX,IAAID,OAAOY,cAAcZ;;;;;;;;;oBAGvB;;wBAAME;;;oBAAN;oBACA,IAAI,EAACH,qBAAAA,+BAAAA,SAAUuB,OAAO,CAACzB,KAAK0B,KAAK,IAAG,MAAM,IAAIO,oBAAU,CAAC,gBAAgB;;;;;;oBAClEf;oBACPM,UAAUb,QAAQO;;;;;;oBAEZU,aAAad;oBACnB,IAAIc,YAAYJ,UAAUA,UAAUL,WAAWK,SAASI,YAAY,+DAA+DA;oBACnI,IAAI;wBACF,IAAIJ,SAAShB,KAAK;4BAAEqB,MAAM;4BAAWC,OAAOC,IAAAA,6BAAc,EAACP;wBAAS;6BAC/DhB,KAAK;4BAAEqB,MAAM;wBAAU;oBAC9B,SAAU;wBACRhC,KAAKoB,KAAK;oBACZ;;;;;;;;;;IAEJ;;AAEA,SAAeiB;;YAeJhB;;;;;;;;;;oBAbPhB,WAAW,IAAIiC,gCAAkB,CAACnC,KAAKoC,SAAS;oBAChD;;wBAAMlC,SAASmC,OAAO,CAACrC,KAAK0B,KAAK,EAAE1B,KAAKsC,GAAG,EAAEtC,KAAKuC,KAAK;;;oBAAvD;oBACA/B,KAAK;wBAAEqB,MAAM;oBAAW;oBACxB1B,QAAQqC,YAAY;wBAClB,IAAIpC,UAAU;wBACdC,WAAWA,SAASoC,IAAI,CAAC;4BACvB,IAAI,EAACvC,qBAAAA,+BAAAA,SAAUwC,KAAK,CAAC1C,KAAK0B,KAAK,IAAG,MAAM,IAAIO,oBAAU,CAAC,gBAAgB;wBACzE;wBACA5B,SAASsC,KAAK,CAACpB;oBACjB,GAAGvB,KAAK4C,mBAAmB;oBAC3B/C,KAAKgD,EAAE,CAAC,WAAW,SAACpC;wBAClB,IAAIA,QAAQoB,IAAI,KAAK,QAAQ,KAAKG;oBACpC;;;;;;oBACOd;oBACPK,KAAKL;;;;;;;;;;;IAET;;AAEA,KAAKgB"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/workers/watch-heartbeat.ts"],"sourcesContent":["import { parentPort, workerData } from 'node:worker_threads';\nimport { SenseError } from '../errors.ts';\nimport { serializeError } from '../scan/worker-error.ts';\nimport { WatchClaimDatabase, type WatchClaimWorkerData, type WatchClaimWorkerMessage } from '../watch-claim.ts';\n\nfunction requireParentPort() {\n  const port = parentPort;\n  if (!port) throw new Error('watch heartbeat worker requires a parent port');\n  return port;\n}\n\nconst port = requireParentPort();\nconst data = workerData as WatchClaimWorkerData;\nlet database: WatchClaimDatabase | undefined;\nlet timer: NodeJS.Timeout | undefined;\nlet stopping = false;\nlet renewing = Promise.resolve();\n\nfunction send(message: WatchClaimWorkerMessage): void {\n  port.postMessage(message);\n}\n\nfunction asError(value: unknown): Error {\n  return value instanceof Error ? value : new Error(String(value));\n}\n\nfunction closeDatabase(): Error | undefined {\n  if (timer) clearInterval(timer);\n  timer = undefined;\n  try {\n    database?.close();\n    database = undefined;\n    return undefined;\n  } catch (err) {\n    return asError(err);\n  }\n}\n\nfunction addFailure(primary: Error, secondary: Error | undefined, message: string): Error {\n  return secondary ? new AggregateError([primary, secondary], message) : primary;\n}\n\nfunction fail(err: unknown): void {\n  if (stopping) return;\n  stopping = true;\n  let failure = asError(err);\n  try {\n    database?.release(data.token);\n  } catch (releaseError) {\n    failure = addFailure(failure, asError(releaseError), 'watch heartbeat and claim release both failed');\n  }\n  const closeError = closeDatabase();\n  failure = addFailure(failure, closeError, 'watch heartbeat and database cleanup both failed');\n  try {\n    send({ type: 'failure', error: serializeError(failure) });\n  } finally {\n    port.close();\n  }\n}\n\nasync function stop(): Promise<void> {\n  if (stopping) return;\n  stopping = true;\n  if (timer) clearInterval(timer);\n  let failure: Error | undefined;\n  try {\n    await renewing;\n    if (!database?.release(data.token)) throw new SenseError('WATCH_ACTIVE', 'watch ownership was replaced by another watcher');\n  } catch (err) {\n    failure = asError(err);\n  } finally {\n    const closeError = closeDatabase();\n    if (closeError) failure = failure ? addFailure(failure, closeError, 'watch heartbeat shutdown and database cleanup both failed') : closeError;\n    try {\n      if (failure) send({ type: 'failure', error: serializeError(failure) });\n      else send({ type: 'stopped' });\n    } finally {\n      port.close();\n    }\n  }\n}\n\nasync function main(): Promise<void> {\n  try {\n    database = new WatchClaimDatabase(data.configDir);\n    await database.acquire(data.token, data.pid, data.force);\n    send({ type: 'acquired' });\n    timer = setInterval(() => {\n      if (stopping) return;\n      renewing = renewing.then(() => {\n        if (!database?.renew(data.token)) throw new SenseError('WATCH_ACTIVE', 'watch ownership was replaced by another watcher');\n      });\n      renewing.catch(fail);\n    }, data.heartbeatIntervalMs);\n    port.on('message', (message: { type?: string }) => {\n      if (message.type === 'stop') void stop();\n    });\n  } catch (err) {\n    fail(err);\n  }\n}\n\nvoid main();\n"],"names":["requireParentPort","port","parentPort","Error","data","workerData","database","timer","stopping","renewing","Promise","resolve","send","message","postMessage","asError","value","String","closeDatabase","clearInterval","undefined","close","err","addFailure","primary","secondary","AggregateError","fail","failure","release","token","releaseError","closeError","type","error","serializeError","stop","SenseError","main","WatchClaimDatabase","configDir","acquire","pid","force","setInterval","then","renew","catch","heartbeatIntervalMs","on"],"mappings":";;;;kCAAuC;wBACZ;6BACI;4BAC6D;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAE5F,SAASA;IACP,IAAMC,OAAOC,8BAAU;IACvB,IAAI,CAACD,MAAM,MAAM,IAAIE,MAAM;IAC3B,OAAOF;AACT;AAEA,IAAMA,OAAOD;AACb,IAAMI,OAAOC,8BAAU;AACvB,IAAIC;AACJ,IAAIC;AACJ,IAAIC,WAAW;AACf,IAAIC,WAAWC,QAAQC,OAAO;AAE9B,SAASC,KAAKC,OAAgC;IAC5CZ,KAAKa,WAAW,CAACD;AACnB;AAEA,SAASE,QAAQC,KAAc;IAC7B,OAAOA,AAAK,YAALA,OAAiBb,SAAQa,QAAQ,IAAIb,MAAMc,OAAOD;AAC3D;AAEA,SAASE;IACP,IAAIX,OAAOY,cAAcZ;IACzBA,QAAQa;IACR,IAAI;QACFd,qBAAAA,+BAAAA,SAAUe,KAAK;QACff,WAAWc;QACX,OAAOA;IACT,EAAE,OAAOE,KAAK;QACZ,OAAOP,QAAQO;IACjB;AACF;AAEA,SAASC,WAAWC,OAAc,EAAEC,SAA4B,EAAEZ,OAAe;IAC/E,OAAOY,YAAY,IAAIC,eAAe;QAACF;QAASC;KAAU,EAAEZ,WAAWW;AACzE;AAEA,SAASG,KAAKL,GAAY;IACxB,IAAId,UAAU;IACdA,WAAW;IACX,IAAIoB,UAAUb,QAAQO;IACtB,IAAI;QACFhB,qBAAAA,+BAAAA,SAAUuB,OAAO,CAACzB,KAAK0B,KAAK;IAC9B,EAAE,OAAOC,cAAc;QACrBH,UAAUL,WAAWK,SAASb,QAAQgB,eAAe;IACvD;IACA,IAAMC,aAAad;IACnBU,UAAUL,WAAWK,SAASI,YAAY;IAC1C,IAAI;QACFpB,KAAK;YAAEqB,MAAM;YAAWC,OAAOC,IAAAA,6BAAc,EAACP;QAAS;IACzD,SAAU;QACR3B,KAAKoB,KAAK;IACZ;AACF;AAEA,SAAee;;YAITR,SAIKN,KAGDU;;;;oBAVR,IAAIxB,UAAU;;;oBACdA,WAAW;oBACX,IAAID,OAAOY,cAAcZ;;;;;;;;;oBAGvB;;wBAAME;;;oBAAN;oBACA,IAAI,EAACH,qBAAAA,+BAAAA,SAAUuB,OAAO,CAACzB,KAAK0B,KAAK,IAAG,MAAM,IAAIO,oBAAU,CAAC,gBAAgB;;;;;;oBAClEf;oBACPM,UAAUb,QAAQO;;;;;;oBAEZU,aAAad;oBACnB,IAAIc,YAAYJ,UAAUA,UAAUL,WAAWK,SAASI,YAAY,+DAA+DA;oBACnI,IAAI;wBACF,IAAIJ,SAAShB,KAAK;4BAAEqB,MAAM;4BAAWC,OAAOC,IAAAA,6BAAc,EAACP;wBAAS;6BAC/DhB,KAAK;4BAAEqB,MAAM;wBAAU;oBAC9B,SAAU;wBACRhC,KAAKoB,KAAK;oBACZ;;;;;;;;;;IAEJ;;AAEA,SAAeiB;;YAeJhB;;;;;;;;;;oBAbPhB,WAAW,IAAIiC,gCAAkB,CAACnC,KAAKoC,SAAS;oBAChD;;wBAAMlC,SAASmC,OAAO,CAACrC,KAAK0B,KAAK,EAAE1B,KAAKsC,GAAG,EAAEtC,KAAKuC,KAAK;;;oBAAvD;oBACA/B,KAAK;wBAAEqB,MAAM;oBAAW;oBACxB1B,QAAQqC,YAAY;wBAClB,IAAIpC,UAAU;wBACdC,WAAWA,SAASoC,IAAI,CAAC;4BACvB,IAAI,EAACvC,qBAAAA,+BAAAA,SAAUwC,KAAK,CAAC1C,KAAK0B,KAAK,IAAG,MAAM,IAAIO,oBAAU,CAAC,gBAAgB;wBACzE;wBACA5B,SAASsC,KAAK,CAACpB;oBACjB,GAAGvB,KAAK4C,mBAAmB;oBAC3B/C,KAAKgD,EAAE,CAAC,WAAW,SAACpC;wBAClB,IAAIA,QAAQoB,IAAI,KAAK,QAAQ,KAAKG;oBACpC;;;;;;oBACOd;oBACPK,KAAKL;;;;;;;;;;;IAET;;AAEA,KAAKgB"}

package/dist/esm/chunk/extract.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/chunk/extract.ts"],"sourcesContent":["import type { Token } from 'markdown-it';\nimport { parser } from './parser.ts';\nimport type { Block } from './types.ts';\n\n// markdown-it, like mdast, has no wikilink, embed, or %%comment%% token -- literal text resolved\n// by regex passes over non-code text (code/inline code is placeholder-held and spliced back verbatim).\nconst COMMENT_PAIR = /%%[\\s\\S]*?%%/g;\n// An embed always yields its target, whatever follows a pipe (a resize suffix, never an alias);\n// must run before the wikilink regex below, bang included, or that regex matches first.\nconst EMBED = /!\\[\\[([^\\]]+)\\]\\]/g;\nconst WIKILINK = /\\[\\[([^\\]]+)\\]\\]/g;\n// A trailing \" ^id\" is an Obsidian block anchor, valid only at a line's end; a caret elsewhere\n// (mid-line) is ordinary text and is left alone.\nconst BLOCK_ID = / \\^[A-Za-z0-9-]+$/gm;\n// A callout marker is only ever the first thing in a blockquote's text (Obsidian's grammar),\n// so this anchors to blockquote output alone -- never applied to prose in general.\nconst CALLOUT_MARKER = /^\\[!\\w[\\w-]*\\][+-]?[ \\t]?/;\n// An html token's content is raw HTML, block or inline: a full <!-- --> comment is dropped, any\n// remaining tags are stripped, and the text a browser would still render survives.\nconst HTML_COMMENT = /<!--[\\s\\S]*?-->/g;\nconst HTML_TAG = /<[^>]+>/g;\n\nfunction stripHtml(value: string): string {\n  return value.replace(HTML_COMMENT, '').replace(HTML_TAG, '');\n}\n\n// A `#anchor` keeps its text (hash dropped), unlike links.ts's parseWikilinkInner (the resolver\n// authority), which discards it entirely -- replicated here since that return value doesn't fit.\nfunction anchorText(base: string): string {\n  const hashIdx = base.indexOf('#');\n  if (hashIdx === -1) return base.trim();\n  const target = base.slice(0, hashIdx).trim();\n  const anchor = base.slice(hashIdx + 1).trim();\n  return target ? `${target} ${anchor}` : anchor;\n}\n\nfunction wikilinkText(inner: string): string {\n  const pipeIdx = inner.indexOf('|');\n  return pipeIdx === -1 ? anchorText(inner) : inner.slice(pipeIdx + 1).trim();\n}\n\nfunction embedText(inner: string): string {\n  const pipeIdx = inner.indexOf('|');\n  return anchorText(pipeIdx === -1 ? inner : inner.slice(0, pipeIdx));\n}\n\nfunction resolveFlavor(text: string): string {\n  return text\n    .replace(HTML_COMMENT, '')\n    .replace(COMMENT_PAIR, '')\n    .replace(EMBED, (_, inner: string) => embedText(inner))\n    .replace(WIKILINK, (_, inner: string) => wikilinkText(inner))\n    .replace(BLOCK_ID, '');\n}\n\n// U+E000 (Private Use Area) never appears in real markdown text, so it is a collision-free\n// placeholder delimiter, spliced back to the literal code value after flavor resolution.\nfunction codePlaceholder(value: string, code: string[]): string {\n  const idx = code.push(value) - 1;\n  return `\\uE000${idx}\\uE000`;\n}\n\n// GFM's autolink literal ends where its path machine stops: a trail of punctuation is a genuine\n// end (link stops before it) only when trailEndsAt accepts it, else the marks are part of the link.\nconst PUNCT = new Set(['!', '\"', '&', \"'\", ')', '*', ',', '.', ';', ':', '<', '?', '_', '~', ']']);\nconst TRAIL_PUNCT = new Set(['!', '\"', \"'\", ')', '*', ',', '.', ';', ':', '?', '_', '~']);\nconst isAlpha = (c: string | undefined): boolean => c !== undefined && ((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'));\nfunction trailEndsAt(s: string, j: number): boolean {\n  for (;;) {\n    const c = s[j];\n    if (c === undefined) return true;\n    if (TRAIL_PUNCT.has(c)) {\n      j += 1;\n      continue;\n    }\n    if (c === '&') {\n      j += 1;\n      if (!isAlpha(s[j])) return false;\n      j += 1;\n      while (isAlpha(s[j])) j += 1;\n      if (s[j] !== ';') return false;\n      j += 1;\n      continue;\n    }\n    if (c === ']') {\n      j += 1;\n      const d = s[j];\n      return d === undefined || d === '(' || d === '[' || /\\s/u.test(d ?? ' ');\n    }\n    if (c === '<') return true;\n    if (c !== undefined && /^\\s$/u.test(c)) return true;\n    return false;\n  }\n}\n// Where the GFM autolink ends in S (the linkifier's text plus the following text), old-style:\n// punctuation that fails the trail test, or a ) with closes <= opens, extends the link.\nfunction gfmAutolinkEnd(s: string): number {\n  let open = 0;\n  let close = 0;\n  for (let i = 0; i < s.length; i++) {\n    const c = s[i];\n    if (c === '(') {\n      open += 1;\n      continue;\n    }\n    if (c === ')' && close < open) {\n      close += 1;\n      continue;\n    }\n    if (PUNCT.has(c) && trailEndsAt(s, i)) return i;\n    if (c !== undefined && /^\\s$/u.test(c)) return i;\n  }\n  return s.length;\n}\n// linkify-it links more than GFM's autolink literal (bare domains, ftp://, //host); only the\n// GFM forms (http(s)://, www., email) drop their text.\nfunction isGfmUrl(text: string): boolean {\n  return /^https?:\\/\\//i.test(text) || /^www\\./i.test(text);\n}\n// Only a bare address is GFM's email autolink: a url keeps its @, since a path can hold one.\nfunction isGfmEmail(text: string): boolean {\n  return !isGfmUrl(text) && text.includes('@');\n}\n// The task-lists plugin injects its checkbox as the item's first inline child.\nfunction isTaskCheckbox(token: Token | undefined): boolean {\n  return token?.type === 'html_inline' && /^<input\\b/i.test(token.content);\n}\n// A linkify span can swallow the & of a following entity (span ...c& + text amp; z); the\n// re-emitted lead is then a reference the entity rule never saw, so decode it old-style.\nconst LEAD_ENTITY = /^&[a-z#][a-z0-9]{1,31};/i;\n\n// Split a flat token range into its top-level blocks: nesting 1 opens, nesting 0 is one block.\nfunction topLevelBlocks(tokens: Token[]): Token[][] {\n  const out: Token[][] = [];\n  let i = 0;\n  while (i < tokens.length) {\n    const token = tokens[i];\n    if (token.nesting === 1) {\n      let depth = 1;\n      let j = i + 1;\n      while (j < tokens.length && depth > 0) {\n        depth += tokens[j].nesting === 1 ? 1 : tokens[j].nesting === -1 ? -1 : 0;\n        j += 1;\n      }\n      out.push(tokens.slice(i, j));\n      i = j;\n    } else if (token.nesting === 0) {\n      out.push(tokens.slice(i, i + 1));\n      i += 1;\n    } else {\n      i += 1;\n    }\n  }\n  return out;\n}\n\n// A first line with no source content (a bare task checkbox, whitespace) contributes no line to\n// the value, so its softbreak is dropped; a first line whose content extraction removes (an\n// autolink, an image) still leaves its line ending behind.\nfunction paragraphText(children: Token[], code: string[]): string {\n  let k = 0;\n  const first = children[0];\n  if (isTaskCheckbox(first)) k = 1;\n  while (k < children.length && children[k].type === 'text' && children[k].content.trim() === '') k += 1;\n  if (k > 0 && children[k]?.type === 'softbreak') return inlineText(children.slice(k + 1), code);\n  return inlineText(children, code);\n}\n\nfunction blockText(tokens: Token[], code: string[]): string {\n  const first = tokens[0];\n  if (!first) return '';\n  switch (first.type) {\n    case 'heading_open':\n    case 'paragraph_open': {\n      const inline = tokens.find((t) => t.type === 'inline');\n      return inline ? paragraphText(inline.children ?? [], code) : '';\n    }\n    case 'fence':\n    case 'code_block':\n      // Fence and indented code tokens carry the closing newline their lines were joined with.\n      return codePlaceholder(first.content.replace(/\\n$/, ''), code);\n    case 'table_open':\n      return tableText(tokens, code);\n    case 'ordered_list_open':\n    case 'bullet_list_open':\n      return listItemsText(tokens, code);\n    case 'blockquote_open':\n      return blocksIn(tokens.slice(1, -1), code).replace(CALLOUT_MARKER, '');\n    case 'html_block':\n      // A non-final-line html block carries its last line's terminator, which the mdast value lacked.\n      return stripHtml(first.content.replace(/\\n$/, ''));\n    case 'footnote_reference_open':\n      return blocksIn(tokens.slice(1, -1), code);\n    default:\n      return '';\n  }\n}\n\n// Sibling blocks (list items, blockquote children, footnote bodies) joined one per line;\n// empties dropped so a blank child never widens the gap between its neighbors.\nfunction blocksIn(tokens: Token[], code: string[]): string {\n  return topLevelBlocks(tokens)\n    .map((b) => blockText(b, code))\n    .filter((s) => s.length > 0)\n    .join('\\n');\n}\n\n// The task-lists plugin slices 3 of the marker's 4 chars, leaving its trailing space on the text\n// after the checkbox token. Returns a copy with that space dropped, never a mutation: a block is\n// extracted more than once (parse pre-extracts headings, group.ts extracts again).\nfunction withoutTaskMarker(item: Token[]): Token[] {\n  for (let i = 0; i < item.length; i++) {\n    const token = item[i];\n    // The checkbox, if any, sits in the item's first inline only.\n    if (token.type !== 'inline' || !token.children?.length) continue;\n    const second = token.children[1];\n    if (!isTaskCheckbox(token.children[0]) || second?.type !== 'text' || !second.content.startsWith(' ')) return item;\n    const children = token.children.slice();\n    children[1] = { ...second, content: second.content.slice(1) } as Token;\n    const copy = item.slice();\n    copy[i] = { ...token, children } as Token;\n    return copy;\n  }\n  return item;\n}\n\nfunction listItemsText(tokens: Token[], code: string[]): string {\n  const items: string[] = [];\n  for (const item of topLevelBlocks(tokens.slice(1, -1))) {\n    items.push(blocksIn(withoutTaskMarker(item).slice(1, -1), code));\n  }\n  return items.filter((s) => s.length > 0).join('\\n');\n}\n\n// Table rows joined by newline, cells by space (mdast's tableRow/tableCell joiners).\nfunction tableText(tokens: Token[], code: string[]): string {\n  const rows: string[] = [];\n  let cells: string[] = [];\n  for (const token of tokens) {\n    if (token.type === 'tr_open') cells = [];\n    else if (token.type === 'inline') cells.push(inlineText(token.children ?? [], code));\n    else if (token.type === 'tr_close') rows.push(cells.join(' '));\n  }\n  return rows.join('\\n');\n}\n\n// Inline content (headings, paragraphs, emphasis, links) concatenated with no separator: the\n// source text tokens already carry whatever spacing separates them.\nfunction inlineText(tokens: Token[], code: string[], inAlt = false): string {\n  let out = '';\n  for (let i = 0; i < tokens.length; i++) {\n    const token = tokens[i];\n    if (token.type === 'link_open') {\n      let skip = 0;\n      let j = i + 1;\n      let depth = 1;\n      while (j < tokens.length && depth > 0) {\n        if (tokens[j].type === 'link_open') depth += 1;\n        else if (tokens[j].type === 'link_close') depth -= 1;\n        j += 1;\n      }\n      const text = inlineText(tokens.slice(i + 1, j - 1), code, inAlt);\n      const markup = token.markup;\n      const next = tokens[j];\n      if (inAlt) {\n        // An image's alt is plain text: mdast kept every link's own text there, autolinks included.\n        out += text;\n      } else if (markup === 'linkify' && isGfmUrl(text)) {\n        // GFM drop: the next text token resumes at the old link's boundary, re-emitting any\n        // trimmed trail. <www.x> has no trail to absorb (the brackets bound it), keep the >.\n        // Emails keep their whole remainder: GFM's email machine has no trail, so the span\n        // already ends at the old boundary and the entity rule has decoded the rest.\n        const prev = tokens[i - 1];\n        const bracketed = prev?.type === 'text' && prev.content.endsWith('<') && next?.type === 'text' && next.content.startsWith('>');\n        const tail = next?.type === 'text' ? text + next.content : text;\n        const k = bracketed ? text.length : gfmAutolinkEnd(tail);\n        out += tail.slice(k).replace(LEAD_ENTITY, (m) => parser().utils.unescapeAll(m));\n        // The trail came out of the next text token, so that token is consumed here, not re-read.\n        if (next?.type === 'text') skip = 1;\n      } else if (markup === 'linkify' && isGfmEmail(text)) {\n        // Email drop: nothing after the span to trim, so the span alone is dropped.\n      } else if (markup === 'autolink') {\n        // <...> leaf: dropped whole, its brackets live inside the token so nothing follows to trim.\n      } else {\n        out += text; // user link or non-GFM linkify target: keep the display text\n      }\n      i = j - 1 + skip;\n    } else if (token.type === 'text') {\n      out += token.content;\n    } else if (token.type === 'softbreak') {\n      out += '\\n';\n    } else if (token.type === 'hardbreak') {\n      out += ' ';\n    } else if (token.type === 'code_inline') {\n      out += codePlaceholder(token.content, code);\n    } else if (token.type === 'html_inline') {\n      out += stripHtml(token.content);\n    } else if (token.type === 'image') {\n      out += inlineText(token.children ?? [], code, true);\n    }\n    // Emphasis, strikethrough and footnote markers carry no text of their own.\n  }\n  return out;\n}\n\nconst CODE_PLACEHOLDER = /\\uE000(\\d+)\\uE000/g;\n\n// Plain text of one parsed block: heading/list/table structure is kept as text, markup (emphasis,\n// link targets, task and callout markers) is dropped. Pure, synchronous.\nexport function extractText(block: Block): string {\n  return extractTexts([block]);\n}\n\n// Blocks resolved together: a blank line inside %%...%% splits it across blocks, and a per-block\n// strip then never sees the closing %%. Code stays placeholder-held across all of them.\nexport function extractTexts(blocks: Block[]): string {\n  const code: string[] = [];\n  const joined = blocks\n    .map((b) => blockText(b.node, code))\n    .filter((s) => s.length > 0)\n    .join('\\n');\n  const resolved = resolveFlavor(joined);\n  return code.length === 0 ? resolved : resolved.replace(CODE_PLACEHOLDER, (_, i: string) => code[Number(i)]);\n}\n"],"names":["parser","COMMENT_PAIR","EMBED","WIKILINK","BLOCK_ID","CALLOUT_MARKER","HTML_COMMENT","HTML_TAG","stripHtml","value","replace","anchorText","base","hashIdx","indexOf","trim","target","slice","anchor","wikilinkText","inner","pipeIdx","embedText","resolveFlavor","text","_","codePlaceholder","code","idx","push","PUNCT","Set","TRAIL_PUNCT","isAlpha","c","undefined","trailEndsAt","s","j","has","d","test","gfmAutolinkEnd","open","close","i","length","isGfmUrl","isGfmEmail","includes","isTaskCheckbox","token","type","content","LEAD_ENTITY","topLevelBlocks","tokens","out","nesting","depth","paragraphText","children","k","first","inlineText","blockText","inline","find","t","tableText","listItemsText","blocksIn","map","b","filter","join","withoutTaskMarker","item","second","startsWith","copy","items","rows","cells","inAlt","skip","markup","next","prev","bracketed","endsWith","tail","m","utils","unescapeAll","CODE_PLACEHOLDER","extractText","block","extractTexts","blocks","joined","node","resolved","Number"],"mappings":"AACA,SAASA,MAAM,QAAQ,cAAc;AAGrC,iGAAiG;AACjG,uGAAuG;AACvG,MAAMC,eAAe;AACrB,gGAAgG;AAChG,wFAAwF;AACxF,MAAMC,QAAQ;AACd,MAAMC,WAAW;AACjB,+FAA+F;AAC/F,iDAAiD;AACjD,MAAMC,WAAW;AACjB,6FAA6F;AAC7F,mFAAmF;AACnF,MAAMC,iBAAiB;AACvB,gGAAgG;AAChG,mFAAmF;AACnF,MAAMC,eAAe;AACrB,MAAMC,WAAW;AAEjB,SAASC,UAAUC,KAAa;IAC9B,OAAOA,MAAMC,OAAO,CAACJ,cAAc,IAAII,OAAO,CAACH,UAAU;AAC3D;AAEA,gGAAgG;AAChG,iGAAiG;AACjG,SAASI,WAAWC,IAAY;IAC9B,MAAMC,UAAUD,KAAKE,OAAO,CAAC;IAC7B,IAAID,YAAY,CAAC,GAAG,OAAOD,KAAKG,IAAI;IACpC,MAAMC,SAASJ,KAAKK,KAAK,CAAC,GAAGJ,SAASE,IAAI;IAC1C,MAAMG,SAASN,KAAKK,KAAK,CAACJ,UAAU,GAAGE,IAAI;IAC3C,OAAOC,SAAS,GAAGA,OAAO,CAAC,EAAEE,QAAQ,GAAGA;AAC1C;AAEA,SAASC,aAAaC,KAAa;IACjC,MAAMC,UAAUD,MAAMN,OAAO,CAAC;IAC9B,OAAOO,YAAY,CAAC,IAAIV,WAAWS,SAASA,MAAMH,KAAK,CAACI,UAAU,GAAGN,IAAI;AAC3E;AAEA,SAASO,UAAUF,KAAa;IAC9B,MAAMC,UAAUD,MAAMN,OAAO,CAAC;IAC9B,OAAOH,WAAWU,YAAY,CAAC,IAAID,QAAQA,MAAMH,KAAK,CAAC,GAAGI;AAC5D;AAEA,SAASE,cAAcC,IAAY;IACjC,OAAOA,KACJd,OAAO,CAACJ,cAAc,IACtBI,OAAO,CAACT,cAAc,IACtBS,OAAO,CAACR,OAAO,CAACuB,GAAGL,QAAkBE,UAAUF,QAC/CV,OAAO,CAACP,UAAU,CAACsB,GAAGL,QAAkBD,aAAaC,QACrDV,OAAO,CAACN,UAAU;AACvB;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASsB,gBAAgBjB,KAAa,EAAEkB,IAAc;IACpD,MAAMC,MAAMD,KAAKE,IAAI,CAACpB,SAAS;IAC/B,OAAO,CAAC,MAAM,EAAEmB,IAAI,MAAM,CAAC;AAC7B;AAEA,gGAAgG;AAChG,oGAAoG;AACpG,MAAME,QAAQ,IAAIC,IAAI;IAAC;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;CAAI;AACjG,MAAMC,cAAc,IAAID,IAAI;IAAC;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;CAAI;AACxF,MAAME,UAAU,CAACC,IAAmCA,MAAMC,aAAc,CAAA,AAACD,KAAK,OAAOA,KAAK,OAASA,KAAK,OAAOA,KAAK,GAAG;AACvH,SAASE,YAAYC,CAAS,EAAEC,CAAS;IACvC,OAAS;QACP,MAAMJ,IAAIG,CAAC,CAACC,EAAE;QACd,IAAIJ,MAAMC,WAAW,OAAO;QAC5B,IAAIH,YAAYO,GAAG,CAACL,IAAI;YACtBI,KAAK;YACL;QACF;QACA,IAAIJ,MAAM,KAAK;YACbI,KAAK;YACL,IAAI,CAACL,QAAQI,CAAC,CAACC,EAAE,GAAG,OAAO;YAC3BA,KAAK;YACL,MAAOL,QAAQI,CAAC,CAACC,EAAE,EAAGA,KAAK;YAC3B,IAAID,CAAC,CAACC,EAAE,KAAK,KAAK,OAAO;YACzBA,KAAK;YACL;QACF;QACA,IAAIJ,MAAM,KAAK;YACbI,KAAK;YACL,MAAME,IAAIH,CAAC,CAACC,EAAE;YACd,OAAOE,MAAML,aAAaK,MAAM,OAAOA,MAAM,OAAO,MAAMC,IAAI,CAACD,cAAAA,eAAAA,IAAK;QACtE;QACA,IAAIN,MAAM,KAAK,OAAO;QACtB,IAAIA,MAAMC,aAAa,QAAQM,IAAI,CAACP,IAAI,OAAO;QAC/C,OAAO;IACT;AACF;AACA,8FAA8F;AAC9F,wFAAwF;AACxF,SAASQ,eAAeL,CAAS;IAC/B,IAAIM,OAAO;IACX,IAAIC,QAAQ;IACZ,IAAK,IAAIC,IAAI,GAAGA,IAAIR,EAAES,MAAM,EAAED,IAAK;QACjC,MAAMX,IAAIG,CAAC,CAACQ,EAAE;QACd,IAAIX,MAAM,KAAK;YACbS,QAAQ;YACR;QACF;QACA,IAAIT,MAAM,OAAOU,QAAQD,MAAM;YAC7BC,SAAS;YACT;QACF;QACA,IAAId,MAAMS,GAAG,CAACL,MAAME,YAAYC,GAAGQ,IAAI,OAAOA;QAC9C,IAAIX,MAAMC,aAAa,QAAQM,IAAI,CAACP,IAAI,OAAOW;IACjD;IACA,OAAOR,EAAES,MAAM;AACjB;AACA,6FAA6F;AAC7F,uDAAuD;AACvD,SAASC,SAASvB,IAAY;IAC5B,OAAO,gBAAgBiB,IAAI,CAACjB,SAAS,UAAUiB,IAAI,CAACjB;AACtD;AACA,6FAA6F;AAC7F,SAASwB,WAAWxB,IAAY;IAC9B,OAAO,CAACuB,SAASvB,SAASA,KAAKyB,QAAQ,CAAC;AAC1C;AACA,+EAA+E;AAC/E,SAASC,eAAeC,KAAwB;IAC9C,OAAOA,CAAAA,kBAAAA,4BAAAA,MAAOC,IAAI,MAAK,iBAAiB,aAAaX,IAAI,CAACU,MAAME,OAAO;AACzE;AACA,yFAAyF;AACzF,yFAAyF;AACzF,MAAMC,cAAc;AAEpB,+FAA+F;AAC/F,SAASC,eAAeC,MAAe;IACrC,MAAMC,MAAiB,EAAE;IACzB,IAAIZ,IAAI;IACR,MAAOA,IAAIW,OAAOV,MAAM,CAAE;QACxB,MAAMK,QAAQK,MAAM,CAACX,EAAE;QACvB,IAAIM,MAAMO,OAAO,KAAK,GAAG;YACvB,IAAIC,QAAQ;YACZ,IAAIrB,IAAIO,IAAI;YACZ,MAAOP,IAAIkB,OAAOV,MAAM,IAAIa,QAAQ,EAAG;gBACrCA,SAASH,MAAM,CAAClB,EAAE,CAACoB,OAAO,KAAK,IAAI,IAAIF,MAAM,CAAClB,EAAE,CAACoB,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI;gBACvEpB,KAAK;YACP;YACAmB,IAAI5B,IAAI,CAAC2B,OAAOvC,KAAK,CAAC4B,GAAGP;YACzBO,IAAIP;QACN,OAAO,IAAIa,MAAMO,OAAO,KAAK,GAAG;YAC9BD,IAAI5B,IAAI,CAAC2B,OAAOvC,KAAK,CAAC4B,GAAGA,IAAI;YAC7BA,KAAK;QACP,OAAO;YACLA,KAAK;QACP;IACF;IACA,OAAOY;AACT;AAEA,gGAAgG;AAChG,4FAA4F;AAC5F,2DAA2D;AAC3D,SAASG,cAAcC,QAAiB,EAAElC,IAAc;QAKzCkC;IAJb,IAAIC,IAAI;IACR,MAAMC,QAAQF,QAAQ,CAAC,EAAE;IACzB,IAAIX,eAAea,QAAQD,IAAI;IAC/B,MAAOA,IAAID,SAASf,MAAM,IAAIe,QAAQ,CAACC,EAAE,CAACV,IAAI,KAAK,UAAUS,QAAQ,CAACC,EAAE,CAACT,OAAO,CAACtC,IAAI,OAAO,GAAI+C,KAAK;IACrG,IAAIA,IAAI,KAAKD,EAAAA,cAAAA,QAAQ,CAACC,EAAE,cAAXD,kCAAAA,YAAaT,IAAI,MAAK,aAAa,OAAOY,WAAWH,SAAS5C,KAAK,CAAC6C,IAAI,IAAInC;IACzF,OAAOqC,WAAWH,UAAUlC;AAC9B;AAEA,SAASsC,UAAUT,MAAe,EAAE7B,IAAc;IAChD,MAAMoC,QAAQP,MAAM,CAAC,EAAE;IACvB,IAAI,CAACO,OAAO,OAAO;IACnB,OAAQA,MAAMX,IAAI;QAChB,KAAK;QACL,KAAK;YAAkB;oBAESc;gBAD9B,MAAMA,SAASV,OAAOW,IAAI,CAAC,CAACC,IAAMA,EAAEhB,IAAI,KAAK;gBAC7C,OAAOc,SAASN,eAAcM,mBAAAA,OAAOL,QAAQ,cAAfK,8BAAAA,mBAAmB,EAAE,EAAEvC,QAAQ;YAC/D;QACA,KAAK;QACL,KAAK;YACH,yFAAyF;YACzF,OAAOD,gBAAgBqC,MAAMV,OAAO,CAAC3C,OAAO,CAAC,OAAO,KAAKiB;QAC3D,KAAK;YACH,OAAO0C,UAAUb,QAAQ7B;QAC3B,KAAK;QACL,KAAK;YACH,OAAO2C,cAAcd,QAAQ7B;QAC/B,KAAK;YACH,OAAO4C,SAASf,OAAOvC,KAAK,CAAC,GAAG,CAAC,IAAIU,MAAMjB,OAAO,CAACL,gBAAgB;QACrE,KAAK;YACH,gGAAgG;YAChG,OAAOG,UAAUuD,MAAMV,OAAO,CAAC3C,OAAO,CAAC,OAAO;QAChD,KAAK;YACH,OAAO6D,SAASf,OAAOvC,KAAK,CAAC,GAAG,CAAC,IAAIU;QACvC;YACE,OAAO;IACX;AACF;AAEA,yFAAyF;AACzF,+EAA+E;AAC/E,SAAS4C,SAASf,MAAe,EAAE7B,IAAc;IAC/C,OAAO4B,eAAeC,QACnBgB,GAAG,CAAC,CAACC,IAAMR,UAAUQ,GAAG9C,OACxB+C,MAAM,CAAC,CAACrC,IAAMA,EAAES,MAAM,GAAG,GACzB6B,IAAI,CAAC;AACV;AAEA,iGAAiG;AACjG,iGAAiG;AACjG,mFAAmF;AACnF,SAASC,kBAAkBC,IAAa;IACtC,IAAK,IAAIhC,IAAI,GAAGA,IAAIgC,KAAK/B,MAAM,EAAED,IAAK;YAGJM;QAFhC,MAAMA,QAAQ0B,IAAI,CAAChC,EAAE;QACrB,8DAA8D;QAC9D,IAAIM,MAAMC,IAAI,KAAK,YAAY,GAACD,kBAAAA,MAAMU,QAAQ,cAAdV,sCAAAA,gBAAgBL,MAAM,GAAE;QACxD,MAAMgC,SAAS3B,MAAMU,QAAQ,CAAC,EAAE;QAChC,IAAI,CAACX,eAAeC,MAAMU,QAAQ,CAAC,EAAE,KAAKiB,CAAAA,mBAAAA,6BAAAA,OAAQ1B,IAAI,MAAK,UAAU,CAAC0B,OAAOzB,OAAO,CAAC0B,UAAU,CAAC,MAAM,OAAOF;QAC7G,MAAMhB,WAAWV,MAAMU,QAAQ,CAAC5C,KAAK;QACrC4C,QAAQ,CAAC,EAAE,GAAG;YAAE,GAAGiB,MAAM;YAAEzB,SAASyB,OAAOzB,OAAO,CAACpC,KAAK,CAAC;QAAG;QAC5D,MAAM+D,OAAOH,KAAK5D,KAAK;QACvB+D,IAAI,CAACnC,EAAE,GAAG;YAAE,GAAGM,KAAK;YAAEU;QAAS;QAC/B,OAAOmB;IACT;IACA,OAAOH;AACT;AAEA,SAASP,cAAcd,MAAe,EAAE7B,IAAc;IACpD,MAAMsD,QAAkB,EAAE;IAC1B,KAAK,MAAMJ,QAAQtB,eAAeC,OAAOvC,KAAK,CAAC,GAAG,CAAC,IAAK;QACtDgE,MAAMpD,IAAI,CAAC0C,SAASK,kBAAkBC,MAAM5D,KAAK,CAAC,GAAG,CAAC,IAAIU;IAC5D;IACA,OAAOsD,MAAMP,MAAM,CAAC,CAACrC,IAAMA,EAAES,MAAM,GAAG,GAAG6B,IAAI,CAAC;AAChD;AAEA,qFAAqF;AACrF,SAASN,UAAUb,MAAe,EAAE7B,IAAc;IAChD,MAAMuD,OAAiB,EAAE;IACzB,IAAIC,QAAkB,EAAE;IACxB,KAAK,MAAMhC,SAASK,OAAQ;YAE8BL;QADxD,IAAIA,MAAMC,IAAI,KAAK,WAAW+B,QAAQ,EAAE;aACnC,IAAIhC,MAAMC,IAAI,KAAK,UAAU+B,MAAMtD,IAAI,CAACmC,YAAWb,kBAAAA,MAAMU,QAAQ,cAAdV,6BAAAA,kBAAkB,EAAE,EAAExB;aACzE,IAAIwB,MAAMC,IAAI,KAAK,YAAY8B,KAAKrD,IAAI,CAACsD,MAAMR,IAAI,CAAC;IAC3D;IACA,OAAOO,KAAKP,IAAI,CAAC;AACnB;AAEA,6FAA6F;AAC7F,oEAAoE;AACpE,SAASX,WAAWR,MAAe,EAAE7B,IAAc,EAAEyD,QAAQ,KAAK;IAChE,IAAI3B,MAAM;IACV,IAAK,IAAIZ,IAAI,GAAGA,IAAIW,OAAOV,MAAM,EAAED,IAAK;QACtC,MAAMM,QAAQK,MAAM,CAACX,EAAE;QACvB,IAAIM,MAAMC,IAAI,KAAK,aAAa;YAC9B,IAAIiC,OAAO;YACX,IAAI/C,IAAIO,IAAI;YACZ,IAAIc,QAAQ;YACZ,MAAOrB,IAAIkB,OAAOV,MAAM,IAAIa,QAAQ,EAAG;gBACrC,IAAIH,MAAM,CAAClB,EAAE,CAACc,IAAI,KAAK,aAAaO,SAAS;qBACxC,IAAIH,MAAM,CAAClB,EAAE,CAACc,IAAI,KAAK,cAAcO,SAAS;gBACnDrB,KAAK;YACP;YACA,MAAMd,OAAOwC,WAAWR,OAAOvC,KAAK,CAAC4B,IAAI,GAAGP,IAAI,IAAIX,MAAMyD;YAC1D,MAAME,SAASnC,MAAMmC,MAAM;YAC3B,MAAMC,OAAO/B,MAAM,CAAClB,EAAE;YACtB,IAAI8C,OAAO;gBACT,4FAA4F;gBAC5F3B,OAAOjC;YACT,OAAO,IAAI8D,WAAW,aAAavC,SAASvB,OAAO;gBACjD,oFAAoF;gBACpF,qFAAqF;gBACrF,mFAAmF;gBACnF,6EAA6E;gBAC7E,MAAMgE,OAAOhC,MAAM,CAACX,IAAI,EAAE;gBAC1B,MAAM4C,YAAYD,CAAAA,iBAAAA,2BAAAA,KAAMpC,IAAI,MAAK,UAAUoC,KAAKnC,OAAO,CAACqC,QAAQ,CAAC,QAAQH,CAAAA,iBAAAA,2BAAAA,KAAMnC,IAAI,MAAK,UAAUmC,KAAKlC,OAAO,CAAC0B,UAAU,CAAC;gBAC1H,MAAMY,OAAOJ,CAAAA,iBAAAA,2BAAAA,KAAMnC,IAAI,MAAK,SAAS5B,OAAO+D,KAAKlC,OAAO,GAAG7B;gBAC3D,MAAMsC,IAAI2B,YAAYjE,KAAKsB,MAAM,GAAGJ,eAAeiD;gBACnDlC,OAAOkC,KAAK1E,KAAK,CAAC6C,GAAGpD,OAAO,CAAC4C,aAAa,CAACsC,IAAM5F,SAAS6F,KAAK,CAACC,WAAW,CAACF;gBAC5E,0FAA0F;gBAC1F,IAAIL,CAAAA,iBAAAA,2BAAAA,KAAMnC,IAAI,MAAK,QAAQiC,OAAO;YACpC,OAAO,IAAIC,WAAW,aAAatC,WAAWxB,OAAO;YACnD,4EAA4E;YAC9E,OAAO,IAAI8D,WAAW,YAAY;YAChC,4FAA4F;YAC9F,OAAO;gBACL7B,OAAOjC,MAAM,6DAA6D;YAC5E;YACAqB,IAAIP,IAAI,IAAI+C;QACd,OAAO,IAAIlC,MAAMC,IAAI,KAAK,QAAQ;YAChCK,OAAON,MAAME,OAAO;QACtB,OAAO,IAAIF,MAAMC,IAAI,KAAK,aAAa;YACrCK,OAAO;QACT,OAAO,IAAIN,MAAMC,IAAI,KAAK,aAAa;YACrCK,OAAO;QACT,OAAO,IAAIN,MAAMC,IAAI,KAAK,eAAe;YACvCK,OAAO/B,gBAAgByB,MAAME,OAAO,EAAE1B;QACxC,OAAO,IAAIwB,MAAMC,IAAI,KAAK,eAAe;YACvCK,OAAOjD,UAAU2C,MAAME,OAAO;QAChC,OAAO,IAAIF,MAAMC,IAAI,KAAK,SAAS;gBACfD;YAAlBM,OAAOO,YAAWb,kBAAAA,MAAMU,QAAQ,cAAdV,6BAAAA,kBAAkB,EAAE,EAAExB,MAAM;QAChD;IACA,2EAA2E;IAC7E;IACA,OAAO8B;AACT;AAEA,MAAMsC,mBAAmB;AAEzB,kGAAkG;AAClG,yEAAyE;AACzE,OAAO,SAASC,YAAYC,KAAY;IACtC,OAAOC,aAAa;QAACD;KAAM;AAC7B;AAEA,iGAAiG;AACjG,wFAAwF;AACxF,OAAO,SAASC,aAAaC,MAAe;IAC1C,MAAMxE,OAAiB,EAAE;IACzB,MAAMyE,SAASD,OACZ3B,GAAG,CAAC,CAACC,IAAMR,UAAUQ,EAAE4B,IAAI,EAAE1E,OAC7B+C,MAAM,CAAC,CAACrC,IAAMA,EAAES,MAAM,GAAG,GACzB6B,IAAI,CAAC;IACR,MAAM2B,WAAW/E,cAAc6E;IAC/B,OAAOzE,KAAKmB,MAAM,KAAK,IAAIwD,WAAWA,SAAS5F,OAAO,CAACqF,kBAAkB,CAACtE,GAAGoB,IAAclB,IAAI,CAAC4E,OAAO1D,GAAG;AAC5G"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/chunk/extract.ts"],"sourcesContent":["import type { Token } from 'markdown-it';\nimport { parser } from './parser.ts';\nimport type { Block } from './types.ts';\n\n// markdown-it, like mdast, has no wikilink, embed, or %%comment%% token -- literal text resolved\n// by regex passes over non-code text (code/inline code is placeholder-held and spliced back verbatim).\nconst COMMENT_PAIR = /%%[\\s\\S]*?%%/g;\n// An embed always yields its target, whatever follows a pipe (a resize suffix, never an alias);\n// must run before the wikilink regex below, bang included, or that regex matches first.\nconst EMBED = /!\\[\\[([^\\]]+)\\]\\]/g;\nconst WIKILINK = /\\[\\[([^\\]]+)\\]\\]/g;\n// A trailing \" ^id\" is an Obsidian block anchor, valid only at a line's end; a caret elsewhere\n// (mid-line) is ordinary text and is left alone.\nconst BLOCK_ID = / \\^[A-Za-z0-9-]+$/gm;\n// A callout marker is only ever the first thing in a blockquote's text (Obsidian's grammar),\n// so this anchors to blockquote output alone -- never applied to prose in general.\nconst CALLOUT_MARKER = /^\\[!\\w[\\w-]*\\][+-]?[ \\t]?/;\n// An html token's content is raw HTML, block or inline: a full <!-- --> comment is dropped, any\n// remaining tags are stripped, and the text a browser would still render survives.\nconst HTML_COMMENT = /<!--[\\s\\S]*?-->/g;\nconst HTML_TAG = /<[^>]+>/g;\n\nfunction stripHtml(value: string): string {\n  return value.replace(HTML_COMMENT, '').replace(HTML_TAG, '');\n}\n\n// A `#anchor` keeps its text (hash dropped), unlike links.ts's parseWikilinkInner (the resolver\n// authority), which discards it entirely -- replicated here since that return value doesn't fit.\nfunction anchorText(base: string): string {\n  const hashIdx = base.indexOf('#');\n  if (hashIdx === -1) return base.trim();\n  const target = base.slice(0, hashIdx).trim();\n  const anchor = base.slice(hashIdx + 1).trim();\n  return target ? `${target} ${anchor}` : anchor;\n}\n\nfunction wikilinkText(inner: string): string {\n  const pipeIdx = inner.indexOf('|');\n  return pipeIdx === -1 ? anchorText(inner) : inner.slice(pipeIdx + 1).trim();\n}\n\nfunction embedText(inner: string): string {\n  const pipeIdx = inner.indexOf('|');\n  return anchorText(pipeIdx === -1 ? inner : inner.slice(0, pipeIdx));\n}\n\nfunction resolveFlavor(text: string): string {\n  return text\n    .replace(HTML_COMMENT, '')\n    .replace(COMMENT_PAIR, '')\n    .replace(EMBED, (_, inner: string) => embedText(inner))\n    .replace(WIKILINK, (_, inner: string) => wikilinkText(inner))\n    .replace(BLOCK_ID, '');\n}\n\n// U+E000 (Private Use Area) never appears in real markdown text, so it is a collision-free\n// placeholder delimiter, spliced back to the literal code value after flavor resolution.\nfunction codePlaceholder(value: string, code: string[]): string {\n  const idx = code.push(value) - 1;\n  return `\\uE000${idx}\\uE000`;\n}\n\n// GFM's autolink literal ends where its path machine stops: a trail of punctuation is a genuine\n// end (link stops before it) only when trailEndsAt accepts it, else the marks are part of the link.\nconst PUNCT = new Set(['!', '\"', '&', \"'\", ')', '*', ',', '.', ';', ':', '<', '?', '_', '~', ']']);\nconst TRAIL_PUNCT = new Set(['!', '\"', \"'\", ')', '*', ',', '.', ';', ':', '?', '_', '~']);\nconst isAlpha = (c: string | undefined): boolean => c !== undefined && ((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'));\nfunction trailEndsAt(s: string, j: number): boolean {\n  for (;;) {\n    const c = s[j];\n    if (c === undefined) return true;\n    if (TRAIL_PUNCT.has(c)) {\n      j += 1;\n      continue;\n    }\n    if (c === '&') {\n      j += 1;\n      if (!isAlpha(s[j])) return false;\n      j += 1;\n      while (isAlpha(s[j])) j += 1;\n      if (s[j] !== ';') return false;\n      j += 1;\n      continue;\n    }\n    if (c === ']') {\n      j += 1;\n      const d = s[j];\n      return d === undefined || d === '(' || d === '[' || /\\s/u.test(d ?? ' ');\n    }\n    if (c === '<') return true;\n    if (c !== undefined && /^\\s$/u.test(c)) return true;\n    return false;\n  }\n}\n// Where the GFM autolink ends in S (the linkifier's text plus the following text), old-style:\n// punctuation that fails the trail test, or a ) with closes <= opens, extends the link.\nfunction gfmAutolinkEnd(s: string): number {\n  let open = 0;\n  let close = 0;\n  for (let i = 0; i < s.length; i++) {\n    const c = s[i];\n    if (c === '(') {\n      open += 1;\n      continue;\n    }\n    if (c === ')' && close < open) {\n      close += 1;\n      continue;\n    }\n    if (PUNCT.has(c) && trailEndsAt(s, i)) return i;\n    if (c !== undefined && /^\\s$/u.test(c)) return i;\n  }\n  return s.length;\n}\n// linkify-it links more than GFM's autolink literal (bare domains, ftp://, //host); only the\n// GFM forms (http(s)://, www., email) drop their text.\nfunction isGfmUrl(text: string): boolean {\n  return /^https?:\\/\\//i.test(text) || /^www\\./i.test(text);\n}\n// Only a bare address is GFM's email autolink: a url keeps its @, since a path can hold one.\nfunction isGfmEmail(text: string): boolean {\n  return !isGfmUrl(text) && text.includes('@');\n}\n// The task-lists plugin injects its checkbox as the item's first inline child.\nfunction isTaskCheckbox(token: Token | undefined): boolean {\n  return token?.type === 'html_inline' && /^<input\\b/i.test(token.content);\n}\n// A linkify span can swallow the & of a following entity (span ...c& + text amp; z); the\n// re-emitted lead is then a reference the entity rule never saw, so decode it old-style.\nconst LEAD_ENTITY = /^&[a-z#][a-z0-9]{1,31};/i;\n\n// Split a flat token range into its top-level blocks: nesting 1 opens, nesting 0 is one block.\nfunction topLevelBlocks(tokens: Token[]): Token[][] {\n  const out: Token[][] = [];\n  let i = 0;\n  while (i < tokens.length) {\n    const token = tokens[i];\n    if (token.nesting === 1) {\n      let depth = 1;\n      let j = i + 1;\n      while (j < tokens.length && depth > 0) {\n        depth += tokens[j].nesting === 1 ? 1 : tokens[j].nesting === -1 ? -1 : 0;\n        j += 1;\n      }\n      out.push(tokens.slice(i, j));\n      i = j;\n    } else if (token.nesting === 0) {\n      out.push(tokens.slice(i, i + 1));\n      i += 1;\n    } else {\n      i += 1;\n    }\n  }\n  return out;\n}\n\n// A first line with no source content (a bare task checkbox, whitespace) contributes no line to\n// the value, so its softbreak is dropped; a first line whose content extraction removes (an\n// autolink, an image) still leaves its line ending behind.\nfunction paragraphText(children: Token[], code: string[]): string {\n  let k = 0;\n  const first = children[0];\n  if (isTaskCheckbox(first)) k = 1;\n  while (k < children.length && children[k].type === 'text' && children[k].content.trim() === '') k += 1;\n  if (k > 0 && children[k]?.type === 'softbreak') return inlineText(children.slice(k + 1), code);\n  return inlineText(children, code);\n}\n\nfunction blockText(tokens: Token[], code: string[]): string {\n  const first = tokens[0];\n  if (!first) return '';\n  switch (first.type) {\n    case 'heading_open':\n    case 'paragraph_open': {\n      const inline = tokens.find((t) => t.type === 'inline');\n      return inline ? paragraphText(inline.children ?? [], code) : '';\n    }\n    case 'fence':\n    case 'code_block':\n      // Fence and indented code tokens carry the closing newline their lines were joined with.\n      return codePlaceholder(first.content.replace(/\\n$/, ''), code);\n    case 'table_open':\n      return tableText(tokens, code);\n    case 'ordered_list_open':\n    case 'bullet_list_open':\n      return listItemsText(tokens, code);\n    case 'blockquote_open':\n      return blocksIn(tokens.slice(1, -1), code).replace(CALLOUT_MARKER, '');\n    case 'html_block':\n      // A non-final-line html block carries its last line's terminator, which the mdast value lacked.\n      return stripHtml(first.content.replace(/\\n$/, ''));\n    case 'footnote_reference_open':\n      return blocksIn(tokens.slice(1, -1), code);\n    default:\n      return '';\n  }\n}\n\n// Sibling blocks (list items, blockquote children, footnote bodies) joined one per line;\n// empties dropped so a blank child never widens the gap between its neighbors.\nfunction blocksIn(tokens: Token[], code: string[]): string {\n  return topLevelBlocks(tokens)\n    .map((b) => blockText(b, code))\n    .filter((s) => s.length > 0)\n    .join('\\n');\n}\n\n// The task-lists plugin slices 3 of the marker's 4 chars, leaving its trailing space on the text\n// after the checkbox token. Returns a copy with that space dropped, never a mutation: a block is\n// extracted more than once (parse pre-extracts headings, group.ts extracts again).\nfunction withoutTaskMarker(item: Token[]): Token[] {\n  for (let i = 0; i < item.length; i++) {\n    const token = item[i];\n    // The checkbox, if any, sits in the item's first inline only.\n    if (token.type !== 'inline' || !token.children?.length) continue;\n    const second = token.children[1];\n    if (!isTaskCheckbox(token.children[0]) || second?.type !== 'text' || !second.content.startsWith(' ')) return item;\n    const children = token.children.slice();\n    children[1] = { ...second, content: second.content.slice(1) } as Token;\n    const copy = item.slice();\n    copy[i] = { ...token, children } as Token;\n    return copy;\n  }\n  return item;\n}\n\nfunction listItemsText(tokens: Token[], code: string[]): string {\n  const items: string[] = [];\n  for (const item of topLevelBlocks(tokens.slice(1, -1))) {\n    items.push(blocksIn(withoutTaskMarker(item).slice(1, -1), code));\n  }\n  return items.filter((s) => s.length > 0).join('\\n');\n}\n\n// Table rows joined by newline, cells by space (mdast's tableRow/tableCell joiners).\nfunction tableText(tokens: Token[], code: string[]): string {\n  const rows: string[] = [];\n  let cells: string[] = [];\n  for (const token of tokens) {\n    if (token.type === 'tr_open') cells = [];\n    else if (token.type === 'inline') cells.push(inlineText(token.children ?? [], code));\n    else if (token.type === 'tr_close') rows.push(cells.join(' '));\n  }\n  return rows.join('\\n');\n}\n\n// Inline content (headings, paragraphs, emphasis, links) concatenated with no separator: the\n// source text tokens already carry whatever spacing separates them.\nfunction inlineText(tokens: Token[], code: string[], inAlt = false): string {\n  let out = '';\n  for (let i = 0; i < tokens.length; i++) {\n    const token = tokens[i];\n    if (token.type === 'link_open') {\n      let skip = 0;\n      let j = i + 1;\n      let depth = 1;\n      while (j < tokens.length && depth > 0) {\n        if (tokens[j].type === 'link_open') depth += 1;\n        else if (tokens[j].type === 'link_close') depth -= 1;\n        j += 1;\n      }\n      const text = inlineText(tokens.slice(i + 1, j - 1), code, inAlt);\n      const markup = token.markup;\n      const next = tokens[j];\n      if (inAlt) {\n        // An image's alt is plain text: mdast kept every link's own text there, autolinks included.\n        out += text;\n      } else if (markup === 'linkify' && isGfmUrl(text)) {\n        // GFM drop: the next text token resumes at the old link's boundary, re-emitting any\n        // trimmed trail. <www.x> has no trail to absorb (the brackets bound it), keep the >.\n        // Emails keep their whole remainder: GFM's email machine has no trail, so the span\n        // already ends at the old boundary and the entity rule has decoded the rest.\n        const prev = tokens[i - 1];\n        const bracketed = prev?.type === 'text' && prev.content.endsWith('<') && next?.type === 'text' && next.content.startsWith('>');\n        const tail = next?.type === 'text' ? text + next.content : text;\n        const k = bracketed ? text.length : gfmAutolinkEnd(tail);\n        out += tail.slice(k).replace(LEAD_ENTITY, (m) => parser().utils.unescapeAll(m));\n        // The trail came out of the next text token, so that token is consumed here, not re-read.\n        if (next?.type === 'text') skip = 1;\n      } else if (markup === 'linkify' && isGfmEmail(text)) {\n        // Email drop: nothing after the span to trim, so the span alone is dropped.\n      } else if (markup === 'autolink') {\n        // <...> leaf: dropped whole, its brackets live inside the token so nothing follows to trim.\n      } else {\n        out += text; // user link or non-GFM linkify target: keep the display text\n      }\n      i = j - 1 + skip;\n    } else if (token.type === 'text') {\n      out += token.content;\n    } else if (token.type === 'softbreak') {\n      out += '\\n';\n    } else if (token.type === 'hardbreak') {\n      out += ' ';\n    } else if (token.type === 'code_inline') {\n      out += codePlaceholder(token.content, code);\n    } else if (token.type === 'html_inline') {\n      out += stripHtml(token.content);\n    } else if (token.type === 'image') {\n      out += inlineText(token.children ?? [], code, true);\n    }\n    // Emphasis, strikethrough and footnote markers carry no text of their own.\n  }\n  return out;\n}\n\nconst CODE_PLACEHOLDER = /\\uE000(\\d+)\\uE000/g;\n\n// Plain text of one parsed block: heading/list/table structure is kept as text, markup (emphasis,\n// link targets, task and callout markers) is dropped. Pure, synchronous.\nexport function extractText(block: Block): string {\n  return extractTexts([block]);\n}\n\n// Blocks resolved together: a blank line inside %%...%% splits it across blocks, and a per-block\n// strip then never sees the closing %%. Code stays placeholder-held across all of them.\nexport function extractTexts(blocks: Block[]): string {\n  const code: string[] = [];\n  const joined = blocks\n    .map((b) => blockText(b.node, code))\n    .filter((s) => s.length > 0)\n    .join('\\n');\n  const resolved = resolveFlavor(joined);\n  return code.length === 0 ? resolved : resolved.replace(CODE_PLACEHOLDER, (_, i: string) => code[Number(i)]);\n}\n"],"names":["parser","COMMENT_PAIR","EMBED","WIKILINK","BLOCK_ID","CALLOUT_MARKER","HTML_COMMENT","HTML_TAG","stripHtml","value","replace","anchorText","base","hashIdx","indexOf","trim","target","slice","anchor","wikilinkText","inner","pipeIdx","embedText","resolveFlavor","text","_","codePlaceholder","code","idx","push","PUNCT","Set","TRAIL_PUNCT","isAlpha","c","undefined","trailEndsAt","s","j","has","d","test","gfmAutolinkEnd","open","close","i","length","isGfmUrl","isGfmEmail","includes","isTaskCheckbox","token","type","content","LEAD_ENTITY","topLevelBlocks","tokens","out","nesting","depth","paragraphText","children","k","first","inlineText","blockText","inline","find","t","tableText","listItemsText","blocksIn","map","b","filter","join","withoutTaskMarker","item","second","startsWith","copy","items","rows","cells","inAlt","skip","markup","next","prev","bracketed","endsWith","tail","m","utils","unescapeAll","CODE_PLACEHOLDER","extractText","block","extractTexts","blocks","joined","node","resolved","Number"],"mappings":"AACA,SAASA,MAAM,QAAQ,cAAc;AAGrC,iGAAiG;AACjG,uGAAuG;AACvG,MAAMC,eAAe;AACrB,gGAAgG;AAChG,wFAAwF;AACxF,MAAMC,QAAQ;AACd,MAAMC,WAAW;AACjB,+FAA+F;AAC/F,iDAAiD;AACjD,MAAMC,WAAW;AACjB,6FAA6F;AAC7F,mFAAmF;AACnF,MAAMC,iBAAiB;AACvB,gGAAgG;AAChG,mFAAmF;AACnF,MAAMC,eAAe;AACrB,MAAMC,WAAW;AAEjB,SAASC,UAAUC,KAAa;IAC9B,OAAOA,MAAMC,OAAO,CAACJ,cAAc,IAAII,OAAO,CAACH,UAAU;AAC3D;AAEA,gGAAgG;AAChG,iGAAiG;AACjG,SAASI,WAAWC,IAAY;IAC9B,MAAMC,UAAUD,KAAKE,OAAO,CAAC;IAC7B,IAAID,YAAY,CAAC,GAAG,OAAOD,KAAKG,IAAI;IACpC,MAAMC,SAASJ,KAAKK,KAAK,CAAC,GAAGJ,SAASE,IAAI;IAC1C,MAAMG,SAASN,KAAKK,KAAK,CAACJ,UAAU,GAAGE,IAAI;IAC3C,OAAOC,SAAS,GAAGA,OAAO,CAAC,EAAEE,QAAQ,GAAGA;AAC1C;AAEA,SAASC,aAAaC,KAAa;IACjC,MAAMC,UAAUD,MAAMN,OAAO,CAAC;IAC9B,OAAOO,YAAY,CAAC,IAAIV,WAAWS,SAASA,MAAMH,KAAK,CAACI,UAAU,GAAGN,IAAI;AAC3E;AAEA,SAASO,UAAUF,KAAa;IAC9B,MAAMC,UAAUD,MAAMN,OAAO,CAAC;IAC9B,OAAOH,WAAWU,YAAY,CAAC,IAAID,QAAQA,MAAMH,KAAK,CAAC,GAAGI;AAC5D;AAEA,SAASE,cAAcC,IAAY;IACjC,OAAOA,KACJd,OAAO,CAACJ,cAAc,IACtBI,OAAO,CAACT,cAAc,IACtBS,OAAO,CAACR,OAAO,CAACuB,GAAGL,QAAkBE,UAAUF,QAC/CV,OAAO,CAACP,UAAU,CAACsB,GAAGL,QAAkBD,aAAaC,QACrDV,OAAO,CAACN,UAAU;AACvB;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASsB,gBAAgBjB,KAAa,EAAEkB,IAAc;IACpD,MAAMC,MAAMD,KAAKE,IAAI,CAACpB,SAAS;IAC/B,OAAO,CAAC,MAAM,EAAEmB,IAAI,MAAM,CAAC;AAC7B;AAEA,gGAAgG;AAChG,oGAAoG;AACpG,MAAME,QAAQ,IAAIC,IAAI;IAAC;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;CAAI;AACjG,MAAMC,cAAc,IAAID,IAAI;IAAC;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;IAAK;CAAI;AACxF,MAAME,UAAU,CAACC,IAAmCA,MAAMC,aAAc,CAAA,AAACD,KAAK,OAAOA,KAAK,OAASA,KAAK,OAAOA,KAAK,GAAG;AACvH,SAASE,YAAYC,CAAS,EAAEC,CAAS;IACvC,OAAS;QACP,MAAMJ,IAAIG,CAAC,CAACC,EAAE;QACd,IAAIJ,MAAMC,WAAW,OAAO;QAC5B,IAAIH,YAAYO,GAAG,CAACL,IAAI;YACtBI,KAAK;YACL;QACF;QACA,IAAIJ,MAAM,KAAK;YACbI,KAAK;YACL,IAAI,CAACL,QAAQI,CAAC,CAACC,EAAE,GAAG,OAAO;YAC3BA,KAAK;YACL,MAAOL,QAAQI,CAAC,CAACC,EAAE,EAAGA,KAAK;YAC3B,IAAID,CAAC,CAACC,EAAE,KAAK,KAAK,OAAO;YACzBA,KAAK;YACL;QACF;QACA,IAAIJ,MAAM,KAAK;YACbI,KAAK;YACL,MAAME,IAAIH,CAAC,CAACC,EAAE;YACd,OAAOE,MAAML,aAAaK,MAAM,OAAOA,MAAM,OAAO,MAAMC,IAAI,CAACD,cAAAA,eAAAA,IAAK;QACtE;QACA,IAAIN,MAAM,KAAK,OAAO;QACtB,IAAIA,MAAMC,aAAa,QAAQM,IAAI,CAACP,IAAI,OAAO;QAC/C,OAAO;IACT;AACF;AACA,8FAA8F;AAC9F,wFAAwF;AACxF,SAASQ,eAAeL,CAAS;IAC/B,IAAIM,OAAO;IACX,IAAIC,QAAQ;IACZ,IAAK,IAAIC,IAAI,GAAGA,IAAIR,EAAES,MAAM,EAAED,IAAK;QACjC,MAAMX,IAAIG,CAAC,CAACQ,EAAE;QACd,IAAIX,MAAM,KAAK;YACbS,QAAQ;YACR;QACF;QACA,IAAIT,MAAM,OAAOU,QAAQD,MAAM;YAC7BC,SAAS;YACT;QACF;QACA,IAAId,MAAMS,GAAG,CAACL,MAAME,YAAYC,GAAGQ,IAAI,OAAOA;QAC9C,IAAIX,MAAMC,aAAa,QAAQM,IAAI,CAACP,IAAI,OAAOW;IACjD;IACA,OAAOR,EAAES,MAAM;AACjB;AACA,6FAA6F;AAC7F,uDAAuD;AACvD,SAASC,SAASvB,IAAY;IAC5B,OAAO,gBAAgBiB,IAAI,CAACjB,SAAS,UAAUiB,IAAI,CAACjB;AACtD;AACA,6FAA6F;AAC7F,SAASwB,WAAWxB,IAAY;IAC9B,OAAO,CAACuB,SAASvB,SAASA,KAAKyB,QAAQ,CAAC;AAC1C;AACA,+EAA+E;AAC/E,SAASC,eAAeC,KAAwB;IAC9C,OAAOA,CAAAA,kBAAAA,4BAAAA,MAAOC,IAAI,MAAK,iBAAiB,aAAaX,IAAI,CAACU,MAAME,OAAO;AACzE;AACA,yFAAyF;AACzF,yFAAyF;AACzF,MAAMC,cAAc;AAEpB,+FAA+F;AAC/F,SAASC,eAAeC,MAAe;IACrC,MAAMC,MAAiB,EAAE;IACzB,IAAIZ,IAAI;IACR,MAAOA,IAAIW,OAAOV,MAAM,CAAE;QACxB,MAAMK,QAAQK,MAAM,CAACX,EAAE;QACvB,IAAIM,MAAMO,OAAO,KAAK,GAAG;YACvB,IAAIC,QAAQ;YACZ,IAAIrB,IAAIO,IAAI;YACZ,MAAOP,IAAIkB,OAAOV,MAAM,IAAIa,QAAQ,EAAG;gBACrCA,SAASH,MAAM,CAAClB,EAAE,CAACoB,OAAO,KAAK,IAAI,IAAIF,MAAM,CAAClB,EAAE,CAACoB,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI;gBACvEpB,KAAK;YACP;YACAmB,IAAI5B,IAAI,CAAC2B,OAAOvC,KAAK,CAAC4B,GAAGP;YACzBO,IAAIP;QACN,OAAO,IAAIa,MAAMO,OAAO,KAAK,GAAG;YAC9BD,IAAI5B,IAAI,CAAC2B,OAAOvC,KAAK,CAAC4B,GAAGA,IAAI;YAC7BA,KAAK;QACP,OAAO;YACLA,KAAK;QACP;IACF;IACA,OAAOY;AACT;AAEA,gGAAgG;AAChG,4FAA4F;AAC5F,2DAA2D;AAC3D,SAASG,cAAcC,QAAiB,EAAElC,IAAc;QAKzCkC;IAJb,IAAIC,IAAI;IACR,MAAMC,QAAQF,QAAQ,CAAC,EAAE;IACzB,IAAIX,eAAea,QAAQD,IAAI;IAC/B,MAAOA,IAAID,SAASf,MAAM,IAAIe,QAAQ,CAACC,EAAE,CAACV,IAAI,KAAK,UAAUS,QAAQ,CAACC,EAAE,CAACT,OAAO,CAACtC,IAAI,OAAO,GAAI+C,KAAK;IACrG,IAAIA,IAAI,KAAKD,EAAAA,cAAAA,QAAQ,CAACC,EAAE,cAAXD,kCAAAA,YAAaT,IAAI,MAAK,aAAa,OAAOY,WAAWH,SAAS5C,KAAK,CAAC6C,IAAI,IAAInC;IACzF,OAAOqC,WAAWH,UAAUlC;AAC9B;AAEA,SAASsC,UAAUT,MAAe,EAAE7B,IAAc;IAChD,MAAMoC,QAAQP,MAAM,CAAC,EAAE;IACvB,IAAI,CAACO,OAAO,OAAO;IACnB,OAAQA,MAAMX,IAAI;QAChB,KAAK;QACL,KAAK;YAAkB;oBAESc;gBAD9B,MAAMA,SAASV,OAAOW,IAAI,CAAC,CAACC,IAAMA,EAAEhB,IAAI,KAAK;gBAC7C,OAAOc,SAASN,eAAcM,mBAAAA,OAAOL,QAAQ,cAAfK,8BAAAA,mBAAmB,EAAE,EAAEvC,QAAQ;YAC/D;QACA,KAAK;QACL,KAAK;YACH,yFAAyF;YACzF,OAAOD,gBAAgBqC,MAAMV,OAAO,CAAC3C,OAAO,CAAC,OAAO,KAAKiB;QAC3D,KAAK;YACH,OAAO0C,UAAUb,QAAQ7B;QAC3B,KAAK;QACL,KAAK;YACH,OAAO2C,cAAcd,QAAQ7B;QAC/B,KAAK;YACH,OAAO4C,SAASf,OAAOvC,KAAK,CAAC,GAAG,CAAC,IAAIU,MAAMjB,OAAO,CAACL,gBAAgB;QACrE,KAAK;YACH,gGAAgG;YAChG,OAAOG,UAAUuD,MAAMV,OAAO,CAAC3C,OAAO,CAAC,OAAO;QAChD,KAAK;YACH,OAAO6D,SAASf,OAAOvC,KAAK,CAAC,GAAG,CAAC,IAAIU;QACvC;YACE,OAAO;IACX;AACF;AAEA,yFAAyF;AACzF,+EAA+E;AAC/E,SAAS4C,SAASf,MAAe,EAAE7B,IAAc;IAC/C,OAAO4B,eAAeC,QACnBgB,GAAG,CAAC,CAACC,IAAMR,UAAUQ,GAAG9C,OACxB+C,MAAM,CAAC,CAACrC,IAAMA,EAAES,MAAM,GAAG,GACzB6B,IAAI,CAAC;AACV;AAEA,iGAAiG;AACjG,iGAAiG;AACjG,mFAAmF;AACnF,SAASC,kBAAkBC,IAAa;IACtC,IAAK,IAAIhC,IAAI,GAAGA,IAAIgC,KAAK/B,MAAM,EAAED,IAAK;YAGJM;QAFhC,MAAMA,QAAQ0B,IAAI,CAAChC,EAAE;QACrB,8DAA8D;QAC9D,IAAIM,MAAMC,IAAI,KAAK,YAAY,GAACD,kBAAAA,MAAMU,QAAQ,cAAdV,sCAAAA,gBAAgBL,MAAM,GAAE;QACxD,MAAMgC,SAAS3B,MAAMU,QAAQ,CAAC,EAAE;QAChC,IAAI,CAACX,eAAeC,MAAMU,QAAQ,CAAC,EAAE,KAAKiB,CAAAA,mBAAAA,6BAAAA,OAAQ1B,IAAI,MAAK,UAAU,CAAC0B,OAAOzB,OAAO,CAAC0B,UAAU,CAAC,MAAM,OAAOF;QAC7G,MAAMhB,WAAWV,MAAMU,QAAQ,CAAC5C,KAAK;QACrC4C,QAAQ,CAAC,EAAE,GAAG;YAAE,GAAGiB,MAAM;YAAEzB,SAASyB,OAAOzB,OAAO,CAACpC,KAAK,CAAC;QAAG;QAC5D,MAAM+D,OAAOH,KAAK5D,KAAK;QACvB+D,IAAI,CAACnC,EAAE,GAAG;YAAE,GAAGM,KAAK;YAAEU;QAAS;QAC/B,OAAOmB;IACT;IACA,OAAOH;AACT;AAEA,SAASP,cAAcd,MAAe,EAAE7B,IAAc;IACpD,MAAMsD,QAAkB,EAAE;IAC1B,KAAK,MAAMJ,QAAQtB,eAAeC,OAAOvC,KAAK,CAAC,GAAG,CAAC,IAAK;QACtDgE,MAAMpD,IAAI,CAAC0C,SAASK,kBAAkBC,MAAM5D,KAAK,CAAC,GAAG,CAAC,IAAIU;IAC5D;IACA,OAAOsD,MAAMP,MAAM,CAAC,CAACrC,IAAMA,EAAES,MAAM,GAAG,GAAG6B,IAAI,CAAC;AAChD;AAEA,qFAAqF;AACrF,SAASN,UAAUb,MAAe,EAAE7B,IAAc;IAChD,MAAMuD,OAAiB,EAAE;IACzB,IAAIC,QAAkB,EAAE;IACxB,KAAK,MAAMhC,SAASK,OAAQ;YAE8BL;QADxD,IAAIA,MAAMC,IAAI,KAAK,WAAW+B,QAAQ,EAAE;aACnC,IAAIhC,MAAMC,IAAI,KAAK,UAAU+B,MAAMtD,IAAI,CAACmC,YAAWb,kBAAAA,MAAMU,QAAQ,cAAdV,6BAAAA,kBAAkB,EAAE,EAAExB;aACzE,IAAIwB,MAAMC,IAAI,KAAK,YAAY8B,KAAKrD,IAAI,CAACsD,MAAMR,IAAI,CAAC;IAC3D;IACA,OAAOO,KAAKP,IAAI,CAAC;AACnB;AAEA,6FAA6F;AAC7F,oEAAoE;AACpE,SAASX,WAAWR,MAAe,EAAE7B,IAAc,EAAEyD,QAAQ,KAAK;IAChE,IAAI3B,MAAM;IACV,IAAK,IAAIZ,IAAI,GAAGA,IAAIW,OAAOV,MAAM,EAAED,IAAK;QACtC,MAAMM,QAAQK,MAAM,CAACX,EAAE;QACvB,IAAIM,MAAMC,IAAI,KAAK,aAAa;YAC9B,IAAIiC,OAAO;YACX,IAAI/C,IAAIO,IAAI;YACZ,IAAIc,QAAQ;YACZ,MAAOrB,IAAIkB,OAAOV,MAAM,IAAIa,QAAQ,EAAG;gBACrC,IAAIH,MAAM,CAAClB,EAAE,CAACc,IAAI,KAAK,aAAaO,SAAS;qBACxC,IAAIH,MAAM,CAAClB,EAAE,CAACc,IAAI,KAAK,cAAcO,SAAS;gBACnDrB,KAAK;YACP;YACA,MAAMd,OAAOwC,WAAWR,OAAOvC,KAAK,CAAC4B,IAAI,GAAGP,IAAI,IAAIX,MAAMyD;YAC1D,MAAME,SAASnC,MAAMmC,MAAM;YAC3B,MAAMC,OAAO/B,MAAM,CAAClB,EAAE;YACtB,IAAI8C,OAAO;gBACT,4FAA4F;gBAC5F3B,OAAOjC;YACT,OAAO,IAAI8D,WAAW,aAAavC,SAASvB,OAAO;gBACjD,oFAAoF;gBACpF,qFAAqF;gBACrF,mFAAmF;gBACnF,6EAA6E;gBAC7E,MAAMgE,OAAOhC,MAAM,CAACX,IAAI,EAAE;gBAC1B,MAAM4C,YAAYD,CAAAA,iBAAAA,2BAAAA,KAAMpC,IAAI,MAAK,UAAUoC,KAAKnC,OAAO,CAACqC,QAAQ,CAAC,QAAQH,CAAAA,iBAAAA,2BAAAA,KAAMnC,IAAI,MAAK,UAAUmC,KAAKlC,OAAO,CAAC0B,UAAU,CAAC;gBAC1H,MAAMY,OAAOJ,CAAAA,iBAAAA,2BAAAA,KAAMnC,IAAI,MAAK,SAAS5B,OAAO+D,KAAKlC,OAAO,GAAG7B;gBAC3D,MAAMsC,IAAI2B,YAAYjE,KAAKsB,MAAM,GAAGJ,eAAeiD;gBACnDlC,OAAOkC,KAAK1E,KAAK,CAAC6C,GAAGpD,OAAO,CAAC4C,aAAa,CAACsC,IAAM5F,SAAS6F,KAAK,CAACC,WAAW,CAACF;gBAC5E,0FAA0F;gBAC1F,IAAIL,CAAAA,iBAAAA,2BAAAA,KAAMnC,IAAI,MAAK,QAAQiC,OAAO;YACpC,OAAO,IAAIC,WAAW,aAAatC,WAAWxB,OAAO;YACnD,4EAA4E;YAC9E,OAAO,IAAI8D,WAAW,YAAY;YAChC,4FAA4F;YAC9F,OAAO;gBACL7B,OAAOjC,MAAM,6DAA6D;YAC5E;YACAqB,IAAIP,IAAI,IAAI+C;QACd,OAAO,IAAIlC,MAAMC,IAAI,KAAK,QAAQ;YAChCK,OAAON,MAAME,OAAO;QACtB,OAAO,IAAIF,MAAMC,IAAI,KAAK,aAAa;YACrCK,OAAO;QACT,OAAO,IAAIN,MAAMC,IAAI,KAAK,aAAa;YACrCK,OAAO;QACT,OAAO,IAAIN,MAAMC,IAAI,KAAK,eAAe;YACvCK,OAAO/B,gBAAgByB,MAAME,OAAO,EAAE1B;QACxC,OAAO,IAAIwB,MAAMC,IAAI,KAAK,eAAe;YACvCK,OAAOjD,UAAU2C,MAAME,OAAO;QAChC,OAAO,IAAIF,MAAMC,IAAI,KAAK,SAAS;gBACfD;YAAlBM,OAAOO,YAAWb,kBAAAA,MAAMU,QAAQ,cAAdV,6BAAAA,kBAAkB,EAAE,EAAExB,MAAM;QAChD;IACA,2EAA2E;IAC7E;IACA,OAAO8B;AACT;AAEA,MAAMsC,mBAAmB;AAEzB,kGAAkG;AAClG,yEAAyE;AACzE,OAAO,SAASC,YAAYC,KAAY;IACtC,OAAOC,aAAa;QAACD;KAAM;AAC7B;AAEA,iGAAiG;AACjG,wFAAwF;AACxF,OAAO,SAASC,aAAaC,MAAe;IAC1C,MAAMxE,OAAiB,EAAE;IACzB,MAAMyE,SAASD,OACZ3B,GAAG,CAAC,CAACC,IAAMR,UAAUQ,EAAE4B,IAAI,EAAE1E,OAC7B+C,MAAM,CAAC,CAACrC,IAAMA,EAAES,MAAM,GAAG,GACzB6B,IAAI,CAAC;IACR,MAAM2B,WAAW/E,cAAc6E;IAC/B,OAAOzE,KAAKmB,MAAM,KAAK,IAAIwD,WAAWA,SAAS5F,OAAO,CAACqF,kBAAkB,CAACtE,GAAGoB,IAAclB,IAAI,CAAC4E,OAAO1D,GAAG;AAC5G"}

package/dist/esm/chunk/group.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/chunk/group.ts"],"sourcesContent":["import { extractText } from './extract.ts';\nimport { parse } from './parse.ts';\nimport { DEFAULT_TARGET_TOKENS, estimateTokens } from './tokens.ts';\nimport type { Block, BlockType, Chunk, ChunkOptions } from './types.ts';\n\nconst PGC_GROUP_SIZE = 2;\nconst OVERSIZE_TRIGGER_MULTIPLE = 2;\n\ninterface ResolvedOptions {\n  targetTokens: number;\n  text: 'extracted' | 'raw';\n}\n\nfunction resolveOptions(opts?: ChunkOptions): ResolvedOptions {\n  return {\n    targetTokens: opts?.targetTokens ?? DEFAULT_TARGET_TOKENS,\n    text: opts?.text ?? 'raw',\n  };\n}\n\n// A heading of any depth ends the current scope and starts a new one (D1); the heading block\n// itself is carried into the new scope, where it joins that scope's first group (F7/F10).\nfunction splitScopes(blocks: Block[]): Block[][] {\n  const scopes: Block[][] = [];\n  let current: Block[] = [];\n  for (const block of blocks) {\n    if (block.type === 'heading' && current.length > 0) {\n      scopes.push(current);\n      current = [];\n    }\n    current.push(block);\n  }\n  if (current.length > 0) scopes.push(current);\n  return scopes;\n}\n\ninterface Part {\n  startLine: number;\n  endLine: number;\n  text: string;\n  // True for a sub-line split piece: its text is already the final slice, not the whole line --\n  // group()'s extent-based raw re-slice must not touch it (siblings share startLine === endLine).\n  final?: boolean;\n}\n\nfunction finalize(parts: Part[]): (Chunk & { final?: boolean }) | undefined {\n  if (parts.length === 0) return undefined;\n  const first = parts[0];\n  const last = parts[parts.length - 1];\n  return { startLine: first.startLine, endLine: last.endLine, text: parts.map((p) => p.text).join('\\n'), final: parts.some((p) => p.final) };\n}\n\nconst NEWLINE_TOKENS = estimateTokens('\\n');\n// Built on first use and kept: each construction is ~3.5 ms, and only an oversize block is ever\n// split, so no command pays for a segmenter it never reaches.\nconst SEGMENTERS = new Map<string, Intl.Segmenter>();\n\nfunction segmentsOf(text: string, granularity: 'sentence' | 'word'): string[] {\n  let segmenter = SEGMENTERS.get(granularity);\n  if (!segmenter) {\n    segmenter = new Intl.Segmenter(undefined, { granularity });\n    SEGMENTERS.set(granularity, segmenter);\n  }\n  return Array.from(segmenter.segment(text), (s) => s.segment);\n}\n\n// Greedily packs segments (already contiguous, tiling the source text with no gaps) into groups\n// of at most `working` estimated tokens; a lone segment over `working` still stands alone.\nfunction pack(segments: string[], working: number): string[] {\n  const groups: string[] = [];\n  let current = '';\n  let tokens = 0;\n  for (const segment of segments) {\n    const segmentTokens = estimateTokens(segment);\n    if (current.length > 0 && tokens + segmentTokens > working) {\n      groups.push(current);\n      current = '';\n      tokens = 0;\n    }\n    current += segment;\n    tokens += segmentTokens;\n  }\n  if (current.length > 0) groups.push(current);\n  return groups;\n}\n\n// Line-split alone can't shrink a lone dense line (the CJK case): falls back to sentence then\n// word boundaries (Intl.Segmenter, the same grapheme-safe engine as segment.ts), mode-agnostic on `text`.\nfunction splitLineText(text: string, working: number): string[] {\n  const sentences = segmentsOf(text, 'sentence');\n  const out: string[] = [];\n  let current = '';\n  let tokens = 0;\n  const flush = () => {\n    if (current.length > 0) {\n      out.push(current);\n      current = '';\n      tokens = 0;\n    }\n  };\n  for (const sentence of sentences) {\n    const sentenceTokens = estimateTokens(sentence);\n    if (sentenceTokens > working) {\n      flush();\n      out.push(...pack(segmentsOf(sentence, 'word'), working));\n      continue;\n    }\n    if (current.length > 0 && tokens + sentenceTokens > working) flush();\n    current += sentence;\n    tokens += sentenceTokens;\n  }\n  flush();\n  return out;\n}\n\nconst ATOMIC_TYPES: ReadonlySet<BlockType> = new Set(['code', 'table', 'list']);\n\n// Atomic (code/table/list) pieces are always a raw line slice: re-parsing a table's later pieces\n// without their header/delimiter rows would demote them to paragraph text.\nfunction piece(pieceLines: string[], startLine: number, endLine: number, blockType: BlockType, textMode: 'extracted' | 'raw'): Part {\n  const text =\n    ATOMIC_TYPES.has(blockType) || textMode === 'raw'\n      ? pieceLines.join('\\n')\n      : parse(pieceLines.join('\\n'))\n          .map((b) => extractText(b))\n          .join('\\n');\n  return { startLine, endLine, text };\n}\n\n// A one-line piece over working can't shrink via another line-boundary pass (rule 5's gap), so it\n// splits at sentence/word boundaries instead; `final` stops group() re-deriving its text by extent (F5).\nfunction finalizePiece(pieceLines: string[], startLine: number, endLine: number, working: number, blockType: BlockType, textMode: 'extracted' | 'raw'): Part[] {\n  const p = piece(pieceLines, startLine, endLine, blockType, textMode);\n  if (pieceLines.length === 1 && estimateTokens(p.text) > working) {\n    return splitLineText(p.text, working).map((text) => ({ startLine, endLine, text, final: true }));\n  }\n  return [p];\n}\n\n// A block over 2x working size splits at line boundaries into pieces each <= working size, never\n// mid-line. `seed`: pending tokens (e.g. a heading) the first piece must join, checked against the limit.\nfunction splitOversizeBlock(lines: string[], startLine: number, endLine: number, working: number, blockType: BlockType, textMode: 'extracted' | 'raw', seed = 0): Part[] {\n  const pieces: Part[] = [];\n  let pieceLines: string[] = [];\n  let pieceStart = startLine;\n  let tokens = seed;\n  for (let line = startLine; line <= endLine; line++) {\n    const lineText = lines[line - 1];\n    const sep = pieceLines.length > 0 || tokens > 0 ? NEWLINE_TOKENS : 0;\n    const lineTokens = estimateTokens(lineText);\n    if (pieceLines.length > 0 && tokens + sep + lineTokens > working) {\n      pieces.push(...finalizePiece(pieceLines, pieceStart, line - 1, working, blockType, textMode));\n      pieceLines = [];\n      tokens = 0;\n      pieceStart = line;\n      pieceLines.push(lineText);\n      tokens += lineTokens;\n      continue;\n    }\n    pieceLines.push(lineText);\n    tokens += sep + lineTokens;\n  }\n  if (pieceLines.length > 0) pieces.push(...finalizePiece(pieceLines, pieceStart, endLine, working, blockType, textMode));\n  return pieces;\n}\n\n// One heading scope's groups (D1): a heading opens the first group, and an oversize block\n// (rule 5, including an oversize heading) splits into pieces that each close their own group.\nfunction groupScope(scopeBlocks: Block[], lines: string[], resolved: ResolvedOptions): (Chunk & { final?: boolean })[] {\n  const working = resolved.targetTokens;\n  const trigger = working * OVERSIZE_TRIGGER_MULTIPLE;\n  const finished: (Chunk & { final?: boolean })[] = [];\n  let parts: Part[] = [];\n  let paragraphCount = 0;\n  let tokens = 0;\n\n  function close(): void {\n    const group = finalize(parts);\n    if (group) finished.push(group);\n    parts = [];\n    paragraphCount = 0;\n    tokens = 0;\n  }\n\n  // tokens tracks the active text mode's own estimate (a newline between parts costs\n  // NEWLINE_TOKENS too), so packing decisions size the text the chunk will actually ship as.\n  function addPart(text: string, sizeText: string, startLine: number, endLine: number): void {\n    tokens += (parts.length > 0 ? NEWLINE_TOKENS : 0) + estimateTokens(sizeText);\n    parts.push({ startLine, endLine, text });\n  }\n\n  for (const block of scopeBlocks) {\n    const raw = lines.slice(block.startLine - 1, block.endLine).join('\\n');\n    const blockTokens = estimateTokens(raw);\n\n    if (blockTokens > trigger) {\n      const seed = parts.length > 0 ? tokens : 0;\n      for (const p of splitOversizeBlock(lines, block.startLine, block.endLine, working, block.type, resolved.text, seed)) {\n        parts.push(p);\n        close();\n      }\n      continue;\n    }\n\n    const extracted = extractText(block);\n    const sizeText = resolved.text === 'raw' ? raw : extracted;\n\n    if (block.type === 'heading') {\n      addPart(extracted, sizeText, block.startLine, block.endLine);\n      continue;\n    }\n\n    // The 2x-working invariant holds even under pgc's paper-faithful 2-paragraph pairing --\n    // close first if the pair about to form would cross it.\n    const pairOversize = parts.length > 0 && tokens + NEWLINE_TOKENS + blockTokens > trigger;\n    if (pairOversize) close();\n\n    addPart(extracted, sizeText, block.startLine, block.endLine);\n    paragraphCount++;\n\n    if (paragraphCount >= PGC_GROUP_SIZE) close();\n  }\n  close();\n\n  return finished;\n}\n\n// Groups already-parsed blocks per opts (D1/D3), against the same body the blocks were parsed\n// from (line lookups for oversize splitting).\nexport function group(blocks: Block[], body: string, opts?: ChunkOptions): Chunk[] {\n  const resolved = resolveOptions(opts);\n  const lines = body.split('\\n');\n  const chunks: (Chunk & { final?: boolean })[] = [];\n  for (const scope of splitScopes(blocks)) chunks.push(...groupScope(scope, lines, resolved));\n  // 'raw': the chunk's own source lines verbatim, replacing the flavor-resolved join above (D9).\n  // A `final` chunk already carries its own slice's raw text; re-slicing by extent would return the whole shared line.\n  const texted =\n    resolved.text === 'raw'\n      ? chunks.map((c) =>\n          c.final\n            ? c\n            : {\n                ...c,\n                text: lines\n                  .slice(c.startLine - 1, c.endLine)\n                  .join('\\n')\n                  .trim(),\n              }\n        )\n      : chunks;\n  // A group can be all-blank (flavor-stripped to nothing, or a raw slice of pure syntax); it never produces a chunk.\n  return texted.filter((c) => c.text.trim().length > 0).map((c) => ({ startLine: c.startLine, endLine: c.endLine, text: c.text }));\n}\n"],"names":["extractText","parse","DEFAULT_TARGET_TOKENS","estimateTokens","PGC_GROUP_SIZE","OVERSIZE_TRIGGER_MULTIPLE","resolveOptions","opts","targetTokens","text","splitScopes","blocks","scopes","current","block","type","length","push","finalize","parts","undefined","first","last","startLine","endLine","map","p","join","final","some","NEWLINE_TOKENS","SEGMENTERS","Map","segmentsOf","granularity","segmenter","get","Intl","Segmenter","set","Array","from","segment","s","pack","segments","working","groups","tokens","segmentTokens","splitLineText","sentences","out","flush","sentence","sentenceTokens","ATOMIC_TYPES","Set","piece","pieceLines","blockType","textMode","has","b","finalizePiece","splitOversizeBlock","lines","seed","pieces","pieceStart","line","lineText","sep","lineTokens","groupScope","scopeBlocks","resolved","trigger","finished","paragraphCount","close","group","addPart","sizeText","raw","slice","blockTokens","extracted","pairOversize","body","split","chunks","scope","texted","c","trim","filter"],"mappings":"AAAA,SAASA,WAAW,QAAQ,eAAe;AAC3C,SAASC,KAAK,QAAQ,aAAa;AACnC,SAASC,qBAAqB,EAAEC,cAAc,QAAQ,cAAc;AAGpE,MAAMC,iBAAiB;AACvB,MAAMC,4BAA4B;AAOlC,SAASC,eAAeC,IAAmB;;IACzC,OAAO;QACLC,YAAY,UAAED,iBAAAA,2BAAAA,KAAMC,YAAY,uCAAIN;QACpCO,IAAI,WAAEF,iBAAAA,2BAAAA,KAAME,IAAI,yCAAI;IACtB;AACF;AAEA,6FAA6F;AAC7F,0FAA0F;AAC1F,SAASC,YAAYC,MAAe;IAClC,MAAMC,SAAoB,EAAE;IAC5B,IAAIC,UAAmB,EAAE;IACzB,KAAK,MAAMC,SAASH,OAAQ;QAC1B,IAAIG,MAAMC,IAAI,KAAK,aAAaF,QAAQG,MAAM,GAAG,GAAG;YAClDJ,OAAOK,IAAI,CAACJ;YACZA,UAAU,EAAE;QACd;QACAA,QAAQI,IAAI,CAACH;IACf;IACA,IAAID,QAAQG,MAAM,GAAG,GAAGJ,OAAOK,IAAI,CAACJ;IACpC,OAAOD;AACT;AAWA,SAASM,SAASC,KAAa;IAC7B,IAAIA,MAAMH,MAAM,KAAK,GAAG,OAAOI;IAC/B,MAAMC,QAAQF,KAAK,CAAC,EAAE;IACtB,MAAMG,OAAOH,KAAK,CAACA,MAAMH,MAAM,GAAG,EAAE;IACpC,OAAO;QAAEO,WAAWF,MAAME,SAAS;QAAEC,SAASF,KAAKE,OAAO;QAAEf,MAAMU,MAAMM,GAAG,CAAC,CAACC,IAAMA,EAAEjB,IAAI,EAAEkB,IAAI,CAAC;QAAOC,OAAOT,MAAMU,IAAI,CAAC,CAACH,IAAMA,EAAEE,KAAK;IAAE;AAC3I;AAEA,MAAME,iBAAiB3B,eAAe;AACtC,gGAAgG;AAChG,8DAA8D;AAC9D,MAAM4B,aAAa,IAAIC;AAEvB,SAASC,WAAWxB,IAAY,EAAEyB,WAAgC;IAChE,IAAIC,YAAYJ,WAAWK,GAAG,CAACF;IAC/B,IAAI,CAACC,WAAW;QACdA,YAAY,IAAIE,KAAKC,SAAS,CAAClB,WAAW;YAAEc;QAAY;QACxDH,WAAWQ,GAAG,CAACL,aAAaC;IAC9B;IACA,OAAOK,MAAMC,IAAI,CAACN,UAAUO,OAAO,CAACjC,OAAO,CAACkC,IAAMA,EAAED,OAAO;AAC7D;AAEA,gGAAgG;AAChG,2FAA2F;AAC3F,SAASE,KAAKC,QAAkB,EAAEC,OAAe;IAC/C,MAAMC,SAAmB,EAAE;IAC3B,IAAIlC,UAAU;IACd,IAAImC,SAAS;IACb,KAAK,MAAMN,WAAWG,SAAU;QAC9B,MAAMI,gBAAgB9C,eAAeuC;QACrC,IAAI7B,QAAQG,MAAM,GAAG,KAAKgC,SAASC,gBAAgBH,SAAS;YAC1DC,OAAO9B,IAAI,CAACJ;YACZA,UAAU;YACVmC,SAAS;QACX;QACAnC,WAAW6B;QACXM,UAAUC;IACZ;IACA,IAAIpC,QAAQG,MAAM,GAAG,GAAG+B,OAAO9B,IAAI,CAACJ;IACpC,OAAOkC;AACT;AAEA,8FAA8F;AAC9F,0GAA0G;AAC1G,SAASG,cAAczC,IAAY,EAAEqC,OAAe;IAClD,MAAMK,YAAYlB,WAAWxB,MAAM;IACnC,MAAM2C,MAAgB,EAAE;IACxB,IAAIvC,UAAU;IACd,IAAImC,SAAS;IACb,MAAMK,QAAQ;QACZ,IAAIxC,QAAQG,MAAM,GAAG,GAAG;YACtBoC,IAAInC,IAAI,CAACJ;YACTA,UAAU;YACVmC,SAAS;QACX;IACF;IACA,KAAK,MAAMM,YAAYH,UAAW;QAChC,MAAMI,iBAAiBpD,eAAemD;QACtC,IAAIC,iBAAiBT,SAAS;YAC5BO;YACAD,IAAInC,IAAI,IAAI2B,KAAKX,WAAWqB,UAAU,SAASR;YAC/C;QACF;QACA,IAAIjC,QAAQG,MAAM,GAAG,KAAKgC,SAASO,iBAAiBT,SAASO;QAC7DxC,WAAWyC;QACXN,UAAUO;IACZ;IACAF;IACA,OAAOD;AACT;AAEA,MAAMI,eAAuC,IAAIC,IAAI;IAAC;IAAQ;IAAS;CAAO;AAE9E,iGAAiG;AACjG,2EAA2E;AAC3E,SAASC,MAAMC,UAAoB,EAAEpC,SAAiB,EAAEC,OAAe,EAAEoC,SAAoB,EAAEC,QAA6B;IAC1H,MAAMpD,OACJ+C,aAAaM,GAAG,CAACF,cAAcC,aAAa,QACxCF,WAAWhC,IAAI,CAAC,QAChB1B,MAAM0D,WAAWhC,IAAI,CAAC,OACnBF,GAAG,CAAC,CAACsC,IAAM/D,YAAY+D,IACvBpC,IAAI,CAAC;IACd,OAAO;QAAEJ;QAAWC;QAASf;IAAK;AACpC;AAEA,kGAAkG;AAClG,yGAAyG;AACzG,SAASuD,cAAcL,UAAoB,EAAEpC,SAAiB,EAAEC,OAAe,EAAEsB,OAAe,EAAEc,SAAoB,EAAEC,QAA6B;IACnJ,MAAMnC,IAAIgC,MAAMC,YAAYpC,WAAWC,SAASoC,WAAWC;IAC3D,IAAIF,WAAW3C,MAAM,KAAK,KAAKb,eAAeuB,EAAEjB,IAAI,IAAIqC,SAAS;QAC/D,OAAOI,cAAcxB,EAAEjB,IAAI,EAAEqC,SAASrB,GAAG,CAAC,CAAChB,OAAU,CAAA;gBAAEc;gBAAWC;gBAASf;gBAAMmB,OAAO;YAAK,CAAA;IAC/F;IACA,OAAO;QAACF;KAAE;AACZ;AAEA,iGAAiG;AACjG,0GAA0G;AAC1G,SAASuC,mBAAmBC,KAAe,EAAE3C,SAAiB,EAAEC,OAAe,EAAEsB,OAAe,EAAEc,SAAoB,EAAEC,QAA6B,EAAEM,OAAO,CAAC;IAC7J,MAAMC,SAAiB,EAAE;IACzB,IAAIT,aAAuB,EAAE;IAC7B,IAAIU,aAAa9C;IACjB,IAAIyB,SAASmB;IACb,IAAK,IAAIG,OAAO/C,WAAW+C,QAAQ9C,SAAS8C,OAAQ;QAClD,MAAMC,WAAWL,KAAK,CAACI,OAAO,EAAE;QAChC,MAAME,MAAMb,WAAW3C,MAAM,GAAG,KAAKgC,SAAS,IAAIlB,iBAAiB;QACnE,MAAM2C,aAAatE,eAAeoE;QAClC,IAAIZ,WAAW3C,MAAM,GAAG,KAAKgC,SAASwB,MAAMC,aAAa3B,SAAS;YAChEsB,OAAOnD,IAAI,IAAI+C,cAAcL,YAAYU,YAAYC,OAAO,GAAGxB,SAASc,WAAWC;YACnFF,aAAa,EAAE;YACfX,SAAS;YACTqB,aAAaC;YACbX,WAAW1C,IAAI,CAACsD;YAChBvB,UAAUyB;YACV;QACF;QACAd,WAAW1C,IAAI,CAACsD;QAChBvB,UAAUwB,MAAMC;IAClB;IACA,IAAId,WAAW3C,MAAM,GAAG,GAAGoD,OAAOnD,IAAI,IAAI+C,cAAcL,YAAYU,YAAY7C,SAASsB,SAASc,WAAWC;IAC7G,OAAOO;AACT;AAEA,0FAA0F;AAC1F,8FAA8F;AAC9F,SAASM,WAAWC,WAAoB,EAAET,KAAe,EAAEU,QAAyB;IAClF,MAAM9B,UAAU8B,SAASpE,YAAY;IACrC,MAAMqE,UAAU/B,UAAUzC;IAC1B,MAAMyE,WAA4C,EAAE;IACpD,IAAI3D,QAAgB,EAAE;IACtB,IAAI4D,iBAAiB;IACrB,IAAI/B,SAAS;IAEb,SAASgC;QACP,MAAMC,QAAQ/D,SAASC;QACvB,IAAI8D,OAAOH,SAAS7D,IAAI,CAACgE;QACzB9D,QAAQ,EAAE;QACV4D,iBAAiB;QACjB/B,SAAS;IACX;IAEA,mFAAmF;IACnF,2FAA2F;IAC3F,SAASkC,QAAQzE,IAAY,EAAE0E,QAAgB,EAAE5D,SAAiB,EAAEC,OAAe;QACjFwB,UAAU,AAAC7B,CAAAA,MAAMH,MAAM,GAAG,IAAIc,iBAAiB,CAAA,IAAK3B,eAAegF;QACnEhE,MAAMF,IAAI,CAAC;YAAEM;YAAWC;YAASf;QAAK;IACxC;IAEA,KAAK,MAAMK,SAAS6D,YAAa;QAC/B,MAAMS,MAAMlB,MAAMmB,KAAK,CAACvE,MAAMS,SAAS,GAAG,GAAGT,MAAMU,OAAO,EAAEG,IAAI,CAAC;QACjE,MAAM2D,cAAcnF,eAAeiF;QAEnC,IAAIE,cAAcT,SAAS;YACzB,MAAMV,OAAOhD,MAAMH,MAAM,GAAG,IAAIgC,SAAS;YACzC,KAAK,MAAMtB,KAAKuC,mBAAmBC,OAAOpD,MAAMS,SAAS,EAAET,MAAMU,OAAO,EAAEsB,SAAShC,MAAMC,IAAI,EAAE6D,SAASnE,IAAI,EAAE0D,MAAO;gBACnHhD,MAAMF,IAAI,CAACS;gBACXsD;YACF;YACA;QACF;QAEA,MAAMO,YAAYvF,YAAYc;QAC9B,MAAMqE,WAAWP,SAASnE,IAAI,KAAK,QAAQ2E,MAAMG;QAEjD,IAAIzE,MAAMC,IAAI,KAAK,WAAW;YAC5BmE,QAAQK,WAAWJ,UAAUrE,MAAMS,SAAS,EAAET,MAAMU,OAAO;YAC3D;QACF;QAEA,wFAAwF;QACxF,wDAAwD;QACxD,MAAMgE,eAAerE,MAAMH,MAAM,GAAG,KAAKgC,SAASlB,iBAAiBwD,cAAcT;QACjF,IAAIW,cAAcR;QAElBE,QAAQK,WAAWJ,UAAUrE,MAAMS,SAAS,EAAET,MAAMU,OAAO;QAC3DuD;QAEA,IAAIA,kBAAkB3E,gBAAgB4E;IACxC;IACAA;IAEA,OAAOF;AACT;AAEA,8FAA8F;AAC9F,8CAA8C;AAC9C,OAAO,SAASG,MAAMtE,MAAe,EAAE8E,IAAY,EAAElF,IAAmB;IACtE,MAAMqE,WAAWtE,eAAeC;IAChC,MAAM2D,QAAQuB,KAAKC,KAAK,CAAC;IACzB,MAAMC,SAA0C,EAAE;IAClD,KAAK,MAAMC,SAASlF,YAAYC,QAASgF,OAAO1E,IAAI,IAAIyD,WAAWkB,OAAO1B,OAAOU;IACjF,+FAA+F;IAC/F,qHAAqH;IACrH,MAAMiB,SACJjB,SAASnE,IAAI,KAAK,QACdkF,OAAOlE,GAAG,CAAC,CAACqE,IACVA,EAAElE,KAAK,GACHkE,IACA;YACE,GAAGA,CAAC;YACJrF,MAAMyD,MACHmB,KAAK,CAACS,EAAEvE,SAAS,GAAG,GAAGuE,EAAEtE,OAAO,EAChCG,IAAI,CAAC,MACLoE,IAAI;QACT,KAENJ;IACN,mHAAmH;IACnH,OAAOE,OAAOG,MAAM,CAAC,CAACF,IAAMA,EAAErF,IAAI,CAACsF,IAAI,GAAG/E,MAAM,GAAG,GAAGS,GAAG,CAAC,CAACqE,IAAO,CAAA;YAAEvE,WAAWuE,EAAEvE,SAAS;YAAEC,SAASsE,EAAEtE,OAAO;YAAEf,MAAMqF,EAAErF,IAAI;QAAC,CAAA;AAC/H"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/chunk/group.ts"],"sourcesContent":["import { extractText } from './extract.ts';\nimport { parse } from './parse.ts';\nimport { DEFAULT_TARGET_TOKENS, estimateTokens } from './tokens.ts';\nimport type { Block, BlockType, Chunk, ChunkOptions } from './types.ts';\n\nconst PGC_GROUP_SIZE = 2;\nconst OVERSIZE_TRIGGER_MULTIPLE = 2;\n\ninterface ResolvedOptions {\n  targetTokens: number;\n  text: 'extracted' | 'raw';\n}\n\nfunction resolveOptions(opts?: ChunkOptions): ResolvedOptions {\n  return {\n    targetTokens: opts?.targetTokens ?? DEFAULT_TARGET_TOKENS,\n    text: opts?.text ?? 'raw',\n  };\n}\n\n// A heading of any depth ends the current scope and starts a new one (D1); the heading block\n// itself is carried into the new scope, where it joins that scope's first group (F7/F10).\nfunction splitScopes(blocks: Block[]): Block[][] {\n  const scopes: Block[][] = [];\n  let current: Block[] = [];\n  for (const block of blocks) {\n    if (block.type === 'heading' && current.length > 0) {\n      scopes.push(current);\n      current = [];\n    }\n    current.push(block);\n  }\n  if (current.length > 0) scopes.push(current);\n  return scopes;\n}\n\ninterface Part {\n  startLine: number;\n  endLine: number;\n  text: string;\n  // True for a sub-line split piece: its text is already the final slice, not the whole line --\n  // group()'s extent-based raw re-slice must not touch it (siblings share startLine === endLine).\n  final?: boolean;\n}\n\nfunction finalize(parts: Part[]): (Chunk & { final?: boolean }) | undefined {\n  if (parts.length === 0) return undefined;\n  const first = parts[0];\n  const last = parts[parts.length - 1];\n  return { startLine: first.startLine, endLine: last.endLine, text: parts.map((p) => p.text).join('\\n'), final: parts.some((p) => p.final) };\n}\n\nconst NEWLINE_TOKENS = estimateTokens('\\n');\n// Built on first use and kept: each construction is ~3.5 ms, and only an oversize block is ever\n// split, so no command pays for a segmenter it never reaches.\nconst SEGMENTERS = new Map<string, Intl.Segmenter>();\n\nfunction segmentsOf(text: string, granularity: 'sentence' | 'word'): string[] {\n  let segmenter = SEGMENTERS.get(granularity);\n  if (!segmenter) {\n    segmenter = new Intl.Segmenter(undefined, { granularity });\n    SEGMENTERS.set(granularity, segmenter);\n  }\n  return Array.from(segmenter.segment(text), (s) => s.segment);\n}\n\n// Greedily packs segments (already contiguous, tiling the source text with no gaps) into groups\n// of at most `working` estimated tokens; a lone segment over `working` still stands alone.\nfunction pack(segments: string[], working: number): string[] {\n  const groups: string[] = [];\n  let current = '';\n  let tokens = 0;\n  for (const segment of segments) {\n    const segmentTokens = estimateTokens(segment);\n    if (current.length > 0 && tokens + segmentTokens > working) {\n      groups.push(current);\n      current = '';\n      tokens = 0;\n    }\n    current += segment;\n    tokens += segmentTokens;\n  }\n  if (current.length > 0) groups.push(current);\n  return groups;\n}\n\n// Line-split alone can't shrink a lone dense line (the CJK case): falls back to sentence then\n// word boundaries (Intl.Segmenter, the same grapheme-safe engine as segment.ts), mode-agnostic on `text`.\nfunction splitLineText(text: string, working: number): string[] {\n  const sentences = segmentsOf(text, 'sentence');\n  const out: string[] = [];\n  let current = '';\n  let tokens = 0;\n  const flush = () => {\n    if (current.length > 0) {\n      out.push(current);\n      current = '';\n      tokens = 0;\n    }\n  };\n  for (const sentence of sentences) {\n    const sentenceTokens = estimateTokens(sentence);\n    if (sentenceTokens > working) {\n      flush();\n      out.push(...pack(segmentsOf(sentence, 'word'), working));\n      continue;\n    }\n    if (current.length > 0 && tokens + sentenceTokens > working) flush();\n    current += sentence;\n    tokens += sentenceTokens;\n  }\n  flush();\n  return out;\n}\n\nconst ATOMIC_TYPES: ReadonlySet<BlockType> = new Set(['code', 'table', 'list']);\n\n// Atomic (code/table/list) pieces are always a raw line slice: re-parsing a table's later pieces\n// without their header/delimiter rows would demote them to paragraph text.\nfunction piece(pieceLines: string[], startLine: number, endLine: number, blockType: BlockType, textMode: 'extracted' | 'raw'): Part {\n  const text =\n    ATOMIC_TYPES.has(blockType) || textMode === 'raw'\n      ? pieceLines.join('\\n')\n      : parse(pieceLines.join('\\n'))\n          .map((b) => extractText(b))\n          .join('\\n');\n  return { startLine, endLine, text };\n}\n\n// A one-line piece over working can't shrink via another line-boundary pass (rule 5's gap), so it\n// splits at sentence/word boundaries instead; `final` stops group() re-deriving its text by extent (F5).\nfunction finalizePiece(pieceLines: string[], startLine: number, endLine: number, working: number, blockType: BlockType, textMode: 'extracted' | 'raw'): Part[] {\n  const p = piece(pieceLines, startLine, endLine, blockType, textMode);\n  if (pieceLines.length === 1 && estimateTokens(p.text) > working) {\n    return splitLineText(p.text, working).map((text) => ({ startLine, endLine, text, final: true }));\n  }\n  return [p];\n}\n\n// A block over 2x working size splits at line boundaries into pieces each <= working size, never\n// mid-line. `seed`: pending tokens (e.g. a heading) the first piece must join, checked against the limit.\nfunction splitOversizeBlock(lines: string[], startLine: number, endLine: number, working: number, blockType: BlockType, textMode: 'extracted' | 'raw', seed = 0): Part[] {\n  const pieces: Part[] = [];\n  let pieceLines: string[] = [];\n  let pieceStart = startLine;\n  let tokens = seed;\n  for (let line = startLine; line <= endLine; line++) {\n    const lineText = lines[line - 1];\n    const sep = pieceLines.length > 0 || tokens > 0 ? NEWLINE_TOKENS : 0;\n    const lineTokens = estimateTokens(lineText);\n    if (pieceLines.length > 0 && tokens + sep + lineTokens > working) {\n      pieces.push(...finalizePiece(pieceLines, pieceStart, line - 1, working, blockType, textMode));\n      pieceLines = [];\n      tokens = 0;\n      pieceStart = line;\n      pieceLines.push(lineText);\n      tokens += lineTokens;\n      continue;\n    }\n    pieceLines.push(lineText);\n    tokens += sep + lineTokens;\n  }\n  if (pieceLines.length > 0) pieces.push(...finalizePiece(pieceLines, pieceStart, endLine, working, blockType, textMode));\n  return pieces;\n}\n\n// One heading scope's groups (D1): a heading opens the first group, and an oversize block\n// (rule 5, including an oversize heading) splits into pieces that each close their own group.\nfunction groupScope(scopeBlocks: Block[], lines: string[], resolved: ResolvedOptions): (Chunk & { final?: boolean })[] {\n  const working = resolved.targetTokens;\n  const trigger = working * OVERSIZE_TRIGGER_MULTIPLE;\n  const finished: (Chunk & { final?: boolean })[] = [];\n  let parts: Part[] = [];\n  let paragraphCount = 0;\n  let tokens = 0;\n\n  function close(): void {\n    const group = finalize(parts);\n    if (group) finished.push(group);\n    parts = [];\n    paragraphCount = 0;\n    tokens = 0;\n  }\n\n  // tokens tracks the active text mode's own estimate (a newline between parts costs\n  // NEWLINE_TOKENS too), so packing decisions size the text the chunk will actually ship as.\n  function addPart(text: string, sizeText: string, startLine: number, endLine: number): void {\n    tokens += (parts.length > 0 ? NEWLINE_TOKENS : 0) + estimateTokens(sizeText);\n    parts.push({ startLine, endLine, text });\n  }\n\n  for (const block of scopeBlocks) {\n    const raw = lines.slice(block.startLine - 1, block.endLine).join('\\n');\n    const blockTokens = estimateTokens(raw);\n\n    if (blockTokens > trigger) {\n      const seed = parts.length > 0 ? tokens : 0;\n      for (const p of splitOversizeBlock(lines, block.startLine, block.endLine, working, block.type, resolved.text, seed)) {\n        parts.push(p);\n        close();\n      }\n      continue;\n    }\n\n    const extracted = extractText(block);\n    const sizeText = resolved.text === 'raw' ? raw : extracted;\n\n    if (block.type === 'heading') {\n      addPart(extracted, sizeText, block.startLine, block.endLine);\n      continue;\n    }\n\n    // The 2x-working invariant holds even under pgc's paper-faithful 2-paragraph pairing --\n    // close first if the pair about to form would cross it.\n    const pairOversize = parts.length > 0 && tokens + NEWLINE_TOKENS + blockTokens > trigger;\n    if (pairOversize) close();\n\n    addPart(extracted, sizeText, block.startLine, block.endLine);\n    paragraphCount++;\n\n    if (paragraphCount >= PGC_GROUP_SIZE) close();\n  }\n  close();\n\n  return finished;\n}\n\n// Groups already-parsed blocks per opts (D1/D3), against the same body the blocks were parsed\n// from (line lookups for oversize splitting).\nexport function group(blocks: Block[], body: string, opts?: ChunkOptions): Chunk[] {\n  const resolved = resolveOptions(opts);\n  const lines = body.split('\\n');\n  const chunks: (Chunk & { final?: boolean })[] = [];\n  for (const scope of splitScopes(blocks)) chunks.push(...groupScope(scope, lines, resolved));\n  // 'raw': the chunk's own source lines verbatim, replacing the flavor-resolved join above (D9).\n  // A `final` chunk already carries its own slice's raw text; re-slicing by extent would return the whole shared line.\n  const texted =\n    resolved.text === 'raw'\n      ? chunks.map((c) =>\n          c.final\n            ? c\n            : {\n                ...c,\n                text: lines\n                  .slice(c.startLine - 1, c.endLine)\n                  .join('\\n')\n                  .trim(),\n              }\n        )\n      : chunks;\n  // A group can be all-blank (flavor-stripped to nothing, or a raw slice of pure syntax); it never produces a chunk.\n  return texted.filter((c) => c.text.trim().length > 0).map((c) => ({ startLine: c.startLine, endLine: c.endLine, text: c.text }));\n}\n"],"names":["extractText","parse","DEFAULT_TARGET_TOKENS","estimateTokens","PGC_GROUP_SIZE","OVERSIZE_TRIGGER_MULTIPLE","resolveOptions","opts","targetTokens","text","splitScopes","blocks","scopes","current","block","type","length","push","finalize","parts","undefined","first","last","startLine","endLine","map","p","join","final","some","NEWLINE_TOKENS","SEGMENTERS","Map","segmentsOf","granularity","segmenter","get","Intl","Segmenter","set","Array","from","segment","s","pack","segments","working","groups","tokens","segmentTokens","splitLineText","sentences","out","flush","sentence","sentenceTokens","ATOMIC_TYPES","Set","piece","pieceLines","blockType","textMode","has","b","finalizePiece","splitOversizeBlock","lines","seed","pieces","pieceStart","line","lineText","sep","lineTokens","groupScope","scopeBlocks","resolved","trigger","finished","paragraphCount","close","group","addPart","sizeText","raw","slice","blockTokens","extracted","pairOversize","body","split","chunks","scope","texted","c","trim","filter"],"mappings":"AAAA,SAASA,WAAW,QAAQ,eAAe;AAC3C,SAASC,KAAK,QAAQ,aAAa;AACnC,SAASC,qBAAqB,EAAEC,cAAc,QAAQ,cAAc;AAGpE,MAAMC,iBAAiB;AACvB,MAAMC,4BAA4B;AAOlC,SAASC,eAAeC,IAAmB;;IACzC,OAAO;QACLC,YAAY,UAAED,iBAAAA,2BAAAA,KAAMC,YAAY,uCAAIN;QACpCO,IAAI,WAAEF,iBAAAA,2BAAAA,KAAME,IAAI,yCAAI;IACtB;AACF;AAEA,6FAA6F;AAC7F,0FAA0F;AAC1F,SAASC,YAAYC,MAAe;IAClC,MAAMC,SAAoB,EAAE;IAC5B,IAAIC,UAAmB,EAAE;IACzB,KAAK,MAAMC,SAASH,OAAQ;QAC1B,IAAIG,MAAMC,IAAI,KAAK,aAAaF,QAAQG,MAAM,GAAG,GAAG;YAClDJ,OAAOK,IAAI,CAACJ;YACZA,UAAU,EAAE;QACd;QACAA,QAAQI,IAAI,CAACH;IACf;IACA,IAAID,QAAQG,MAAM,GAAG,GAAGJ,OAAOK,IAAI,CAACJ;IACpC,OAAOD;AACT;AAWA,SAASM,SAASC,KAAa;IAC7B,IAAIA,MAAMH,MAAM,KAAK,GAAG,OAAOI;IAC/B,MAAMC,QAAQF,KAAK,CAAC,EAAE;IACtB,MAAMG,OAAOH,KAAK,CAACA,MAAMH,MAAM,GAAG,EAAE;IACpC,OAAO;QAAEO,WAAWF,MAAME,SAAS;QAAEC,SAASF,KAAKE,OAAO;QAAEf,MAAMU,MAAMM,GAAG,CAAC,CAACC,IAAMA,EAAEjB,IAAI,EAAEkB,IAAI,CAAC;QAAOC,OAAOT,MAAMU,IAAI,CAAC,CAACH,IAAMA,EAAEE,KAAK;IAAE;AAC3I;AAEA,MAAME,iBAAiB3B,eAAe;AACtC,gGAAgG;AAChG,8DAA8D;AAC9D,MAAM4B,aAAa,IAAIC;AAEvB,SAASC,WAAWxB,IAAY,EAAEyB,WAAgC;IAChE,IAAIC,YAAYJ,WAAWK,GAAG,CAACF;IAC/B,IAAI,CAACC,WAAW;QACdA,YAAY,IAAIE,KAAKC,SAAS,CAAClB,WAAW;YAAEc;QAAY;QACxDH,WAAWQ,GAAG,CAACL,aAAaC;IAC9B;IACA,OAAOK,MAAMC,IAAI,CAACN,UAAUO,OAAO,CAACjC,OAAO,CAACkC,IAAMA,EAAED,OAAO;AAC7D;AAEA,gGAAgG;AAChG,2FAA2F;AAC3F,SAASE,KAAKC,QAAkB,EAAEC,OAAe;IAC/C,MAAMC,SAAmB,EAAE;IAC3B,IAAIlC,UAAU;IACd,IAAImC,SAAS;IACb,KAAK,MAAMN,WAAWG,SAAU;QAC9B,MAAMI,gBAAgB9C,eAAeuC;QACrC,IAAI7B,QAAQG,MAAM,GAAG,KAAKgC,SAASC,gBAAgBH,SAAS;YAC1DC,OAAO9B,IAAI,CAACJ;YACZA,UAAU;YACVmC,SAAS;QACX;QACAnC,WAAW6B;QACXM,UAAUC;IACZ;IACA,IAAIpC,QAAQG,MAAM,GAAG,GAAG+B,OAAO9B,IAAI,CAACJ;IACpC,OAAOkC;AACT;AAEA,8FAA8F;AAC9F,0GAA0G;AAC1G,SAASG,cAAczC,IAAY,EAAEqC,OAAe;IAClD,MAAMK,YAAYlB,WAAWxB,MAAM;IACnC,MAAM2C,MAAgB,EAAE;IACxB,IAAIvC,UAAU;IACd,IAAImC,SAAS;IACb,MAAMK,QAAQ;QACZ,IAAIxC,QAAQG,MAAM,GAAG,GAAG;YACtBoC,IAAInC,IAAI,CAACJ;YACTA,UAAU;YACVmC,SAAS;QACX;IACF;IACA,KAAK,MAAMM,YAAYH,UAAW;QAChC,MAAMI,iBAAiBpD,eAAemD;QACtC,IAAIC,iBAAiBT,SAAS;YAC5BO;YACAD,IAAInC,IAAI,IAAI2B,KAAKX,WAAWqB,UAAU,SAASR;YAC/C;QACF;QACA,IAAIjC,QAAQG,MAAM,GAAG,KAAKgC,SAASO,iBAAiBT,SAASO;QAC7DxC,WAAWyC;QACXN,UAAUO;IACZ;IACAF;IACA,OAAOD;AACT;AAEA,MAAMI,eAAuC,IAAIC,IAAI;IAAC;IAAQ;IAAS;CAAO;AAE9E,iGAAiG;AACjG,2EAA2E;AAC3E,SAASC,MAAMC,UAAoB,EAAEpC,SAAiB,EAAEC,OAAe,EAAEoC,SAAoB,EAAEC,QAA6B;IAC1H,MAAMpD,OACJ+C,aAAaM,GAAG,CAACF,cAAcC,aAAa,QACxCF,WAAWhC,IAAI,CAAC,QAChB1B,MAAM0D,WAAWhC,IAAI,CAAC,OACnBF,GAAG,CAAC,CAACsC,IAAM/D,YAAY+D,IACvBpC,IAAI,CAAC;IACd,OAAO;QAAEJ;QAAWC;QAASf;IAAK;AACpC;AAEA,kGAAkG;AAClG,yGAAyG;AACzG,SAASuD,cAAcL,UAAoB,EAAEpC,SAAiB,EAAEC,OAAe,EAAEsB,OAAe,EAAEc,SAAoB,EAAEC,QAA6B;IACnJ,MAAMnC,IAAIgC,MAAMC,YAAYpC,WAAWC,SAASoC,WAAWC;IAC3D,IAAIF,WAAW3C,MAAM,KAAK,KAAKb,eAAeuB,EAAEjB,IAAI,IAAIqC,SAAS;QAC/D,OAAOI,cAAcxB,EAAEjB,IAAI,EAAEqC,SAASrB,GAAG,CAAC,CAAChB,OAAU,CAAA;gBAAEc;gBAAWC;gBAASf;gBAAMmB,OAAO;YAAK,CAAA;IAC/F;IACA,OAAO;QAACF;KAAE;AACZ;AAEA,iGAAiG;AACjG,0GAA0G;AAC1G,SAASuC,mBAAmBC,KAAe,EAAE3C,SAAiB,EAAEC,OAAe,EAAEsB,OAAe,EAAEc,SAAoB,EAAEC,QAA6B,EAAEM,OAAO,CAAC;IAC7J,MAAMC,SAAiB,EAAE;IACzB,IAAIT,aAAuB,EAAE;IAC7B,IAAIU,aAAa9C;IACjB,IAAIyB,SAASmB;IACb,IAAK,IAAIG,OAAO/C,WAAW+C,QAAQ9C,SAAS8C,OAAQ;QAClD,MAAMC,WAAWL,KAAK,CAACI,OAAO,EAAE;QAChC,MAAME,MAAMb,WAAW3C,MAAM,GAAG,KAAKgC,SAAS,IAAIlB,iBAAiB;QACnE,MAAM2C,aAAatE,eAAeoE;QAClC,IAAIZ,WAAW3C,MAAM,GAAG,KAAKgC,SAASwB,MAAMC,aAAa3B,SAAS;YAChEsB,OAAOnD,IAAI,IAAI+C,cAAcL,YAAYU,YAAYC,OAAO,GAAGxB,SAASc,WAAWC;YACnFF,aAAa,EAAE;YACfX,SAAS;YACTqB,aAAaC;YACbX,WAAW1C,IAAI,CAACsD;YAChBvB,UAAUyB;YACV;QACF;QACAd,WAAW1C,IAAI,CAACsD;QAChBvB,UAAUwB,MAAMC;IAClB;IACA,IAAId,WAAW3C,MAAM,GAAG,GAAGoD,OAAOnD,IAAI,IAAI+C,cAAcL,YAAYU,YAAY7C,SAASsB,SAASc,WAAWC;IAC7G,OAAOO;AACT;AAEA,0FAA0F;AAC1F,8FAA8F;AAC9F,SAASM,WAAWC,WAAoB,EAAET,KAAe,EAAEU,QAAyB;IAClF,MAAM9B,UAAU8B,SAASpE,YAAY;IACrC,MAAMqE,UAAU/B,UAAUzC;IAC1B,MAAMyE,WAA4C,EAAE;IACpD,IAAI3D,QAAgB,EAAE;IACtB,IAAI4D,iBAAiB;IACrB,IAAI/B,SAAS;IAEb,SAASgC;QACP,MAAMC,QAAQ/D,SAASC;QACvB,IAAI8D,OAAOH,SAAS7D,IAAI,CAACgE;QACzB9D,QAAQ,EAAE;QACV4D,iBAAiB;QACjB/B,SAAS;IACX;IAEA,mFAAmF;IACnF,2FAA2F;IAC3F,SAASkC,QAAQzE,IAAY,EAAE0E,QAAgB,EAAE5D,SAAiB,EAAEC,OAAe;QACjFwB,UAAU,AAAC7B,CAAAA,MAAMH,MAAM,GAAG,IAAIc,iBAAiB,CAAA,IAAK3B,eAAegF;QACnEhE,MAAMF,IAAI,CAAC;YAAEM;YAAWC;YAASf;QAAK;IACxC;IAEA,KAAK,MAAMK,SAAS6D,YAAa;QAC/B,MAAMS,MAAMlB,MAAMmB,KAAK,CAACvE,MAAMS,SAAS,GAAG,GAAGT,MAAMU,OAAO,EAAEG,IAAI,CAAC;QACjE,MAAM2D,cAAcnF,eAAeiF;QAEnC,IAAIE,cAAcT,SAAS;YACzB,MAAMV,OAAOhD,MAAMH,MAAM,GAAG,IAAIgC,SAAS;YACzC,KAAK,MAAMtB,KAAKuC,mBAAmBC,OAAOpD,MAAMS,SAAS,EAAET,MAAMU,OAAO,EAAEsB,SAAShC,MAAMC,IAAI,EAAE6D,SAASnE,IAAI,EAAE0D,MAAO;gBACnHhD,MAAMF,IAAI,CAACS;gBACXsD;YACF;YACA;QACF;QAEA,MAAMO,YAAYvF,YAAYc;QAC9B,MAAMqE,WAAWP,SAASnE,IAAI,KAAK,QAAQ2E,MAAMG;QAEjD,IAAIzE,MAAMC,IAAI,KAAK,WAAW;YAC5BmE,QAAQK,WAAWJ,UAAUrE,MAAMS,SAAS,EAAET,MAAMU,OAAO;YAC3D;QACF;QAEA,wFAAwF;QACxF,wDAAwD;QACxD,MAAMgE,eAAerE,MAAMH,MAAM,GAAG,KAAKgC,SAASlB,iBAAiBwD,cAAcT;QACjF,IAAIW,cAAcR;QAElBE,QAAQK,WAAWJ,UAAUrE,MAAMS,SAAS,EAAET,MAAMU,OAAO;QAC3DuD;QAEA,IAAIA,kBAAkB3E,gBAAgB4E;IACxC;IACAA;IAEA,OAAOF;AACT;AAEA,8FAA8F;AAC9F,8CAA8C;AAC9C,OAAO,SAASG,MAAMtE,MAAe,EAAE8E,IAAY,EAAElF,IAAmB;IACtE,MAAMqE,WAAWtE,eAAeC;IAChC,MAAM2D,QAAQuB,KAAKC,KAAK,CAAC;IACzB,MAAMC,SAA0C,EAAE;IAClD,KAAK,MAAMC,SAASlF,YAAYC,QAASgF,OAAO1E,IAAI,IAAIyD,WAAWkB,OAAO1B,OAAOU;IACjF,+FAA+F;IAC/F,qHAAqH;IACrH,MAAMiB,SACJjB,SAASnE,IAAI,KAAK,QACdkF,OAAOlE,GAAG,CAAC,CAACqE,IACVA,EAAElE,KAAK,GACHkE,IACA;YACE,GAAGA,CAAC;YACJrF,MAAMyD,MACHmB,KAAK,CAACS,EAAEvE,SAAS,GAAG,GAAGuE,EAAEtE,OAAO,EAChCG,IAAI,CAAC,MACLoE,IAAI;QACT,KAENJ;IACN,mHAAmH;IACnH,OAAOE,OAAOG,MAAM,CAAC,CAACF,IAAMA,EAAErF,IAAI,CAACsF,IAAI,GAAG/E,MAAM,GAAG,GAAGS,GAAG,CAAC,CAACqE,IAAO,CAAA;YAAEvE,WAAWuE,EAAEvE,SAAS;YAAEC,SAASsE,EAAEtE,OAAO;YAAEf,MAAMqF,EAAErF,IAAI;QAAC,CAAA;AAC/H"}

package/dist/esm/chunk/index.js.map

@@ -1 +1 @@
-{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/chunk/index.ts"],"sourcesContent":["import { group } from './group.ts';\nimport { parse } from './parse.ts';\nimport type { Block, Chunk, ChunkOptions } from './types.ts';\n\n// Groups blocks a caller already parsed (e.g. scan/index.ts, sharing one parse with the FTS\n// text path) against the same body they came from, per opts (D1/D3).\nexport function chunkFromBlocks(blocks: Block[], body: string, opts?: ChunkOptions): Chunk[] {\n  return group(blocks, body, opts);\n}\n\n// Pure, deterministic function of file content; chunk semantics are version-stamped via\n// CHUNK_VERSION (./version.ts). Algorithm and evidence: BENCHMARKING.md, \"The chunking algorithm\".\nexport function chunk(body: string, opts?: ChunkOptions): Chunk[] {\n  return chunkFromBlocks(parse(body), body, opts);\n}\n\nexport { extractText, extractTexts } from './extract.ts';\nexport { parse } from './parse.ts';\nexport { DEFAULT_TARGET_TOKENS, estimateTokens } from './tokens.ts';\nexport type { Block, BlockType, Chunk, ChunkOptions } from './types.ts';\nexport { CHUNK_VERSION } from './version.ts';\n"],"names":["group","parse","chunkFromBlocks","blocks","body","opts","chunk","extractText","extractTexts","DEFAULT_TARGET_TOKENS","estimateTokens","CHUNK_VERSION"],"mappings":"AAAA,SAASA,KAAK,QAAQ,aAAa;AACnC,SAASC,KAAK,QAAQ,aAAa;AAGnC,4FAA4F;AAC5F,qEAAqE;AACrE,OAAO,SAASC,gBAAgBC,MAAe,EAAEC,IAAY,EAAEC,IAAmB;IAChF,OAAOL,MAAMG,QAAQC,MAAMC;AAC7B;AAEA,wFAAwF;AACxF,mGAAmG;AACnG,OAAO,SAASC,MAAMF,IAAY,EAAEC,IAAmB;IACrD,OAAOH,gBAAgBD,MAAMG,OAAOA,MAAMC;AAC5C;AAEA,SAASE,WAAW,EAAEC,YAAY,QAAQ,eAAe;AACzD,SAASP,KAAK,QAAQ,aAAa;AACnC,SAASQ,qBAAqB,EAAEC,cAAc,QAAQ,cAAc;AAEpE,SAASC,aAAa,QAAQ,eAAe"}
+{"version":3,"sources":["/Users/kevin/Dev/open-source/ai/sensemaking/src/chunk/index.ts"],"sourcesContent":["import { group } from './group.ts';\nimport { parse } from './parse.ts';\nimport type { Block, Chunk, ChunkOptions } from './types.ts';\n\n// Groups blocks a caller already parsed (e.g. scan/index.ts, sharing one parse with the FTS\n// text path) against the same body they came from, per opts (D1/D3).\nexport function chunkFromBlocks(blocks: Block[], body: string, opts?: ChunkOptions): Chunk[] {\n  return group(blocks, body, opts);\n}\n\n// Pure, deterministic function of file content; chunk semantics are version-stamped via\n// CHUNK_VERSION (./version.ts). Algorithm and evidence: BENCHMARKING.md, \"The chunking algorithm\".\nexport function chunk(body: string, opts?: ChunkOptions): Chunk[] {\n  return chunkFromBlocks(parse(body), body, opts);\n}\n\nexport { extractText, extractTexts } from './extract.ts';\nexport { parse } from './parse.ts';\nexport { DEFAULT_TARGET_TOKENS, estimateTokens } from './tokens.ts';\nexport type { Block, BlockType, Chunk, ChunkOptions } from './types.ts';\nexport { CHUNK_VERSION } from './version.ts';\n"],"names":["group","parse","chunkFromBlocks","blocks","body","opts","chunk","extractText","extractTexts","DEFAULT_TARGET_TOKENS","estimateTokens","CHUNK_VERSION"],"mappings":"AAAA,SAASA,KAAK,QAAQ,aAAa;AACnC,SAASC,KAAK,QAAQ,aAAa;AAGnC,4FAA4F;AAC5F,qEAAqE;AACrE,OAAO,SAASC,gBAAgBC,MAAe,EAAEC,IAAY,EAAEC,IAAmB;IAChF,OAAOL,MAAMG,QAAQC,MAAMC;AAC7B;AAEA,wFAAwF;AACxF,mGAAmG;AACnG,OAAO,SAASC,MAAMF,IAAY,EAAEC,IAAmB;IACrD,OAAOH,gBAAgBD,MAAMG,OAAOA,MAAMC;AAC5C;AAEA,SAASE,WAAW,EAAEC,YAAY,QAAQ,eAAe;AACzD,SAASP,KAAK,QAAQ,aAAa;AACnC,SAASQ,qBAAqB,EAAEC,cAAc,QAAQ,cAAc;AAEpE,SAASC,aAAa,QAAQ,eAAe"}