sensemaking 0.16.0 → 0.17.0

package/dist/cjs/scan/list.js

@@ -0,0 +1,154 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", {
+    value: true
+});
+function _export(target, all) {
+    for(var name in all)Object.defineProperty(target, name, {
+        enumerable: true,
+        get: Object.getOwnPropertyDescriptor(all, name).get
+    });
+}
+_export(exports, {
+    get listFiles () {
+        return listFiles;
+    },
+    get toPosixPath () {
+        return toPosixPath;
+    }
+});
+var _nodefs = require("node:fs");
+var _nodepath = require("node:path");
+var _indexts = require("../config/index.js");
+function _array_like_to_array(arr, len) {
+    if (len == null || len > arr.length) len = arr.length;
+    for(var i = 0, arr2 = new Array(len); i < len; i++)arr2[i] = arr[i];
+    return arr2;
+}
+function _array_without_holes(arr) {
+    if (Array.isArray(arr)) return _array_like_to_array(arr);
+}
+function _iterable_to_array(iter) {
+    if (typeof Symbol !== "undefined" && iter[Symbol.iterator] != null || iter["@@iterator"] != null) {
+        return Array.from(iter);
+    }
+}
+function _non_iterable_spread() {
+    throw new TypeError("Invalid attempt to spread non-iterable instance.\nIn order to be iterable, non-array objects must have a [Symbol.iterator]() method.");
+}
+function _to_consumable_array(arr) {
+    return _array_without_holes(arr) || _iterable_to_array(arr) || _unsupported_iterable_to_array(arr) || _non_iterable_spread();
+}
+function _unsupported_iterable_to_array(o, minLen) {
+    if (!o) return;
+    if (typeof o === "string") return _array_like_to_array(o, minLen);
+    var n = Object.prototype.toString.call(o).slice(8, -1);
+    if (n === "Object" && o.constructor) n = o.constructor.name;
+    if (n === "Map" || n === "Set") return Array.from(n);
+    if (n === "Arguments" || /^(?:Ui|I)nt(?:8|16|32)(?:Clamped)?Array$/.test(n)) return _array_like_to_array(o, minLen);
+}
+function toPosixPath(relPath) {
+    var separator = arguments.length > 1 && arguments[1] !== void 0 ? arguments[1] : _nodepath.sep;
+    return separator === '\\' ? relPath.split(separator).join('/') : relPath;
+}
+// Every command pays listFiles before it answers (the freshness check stats each file), so
+// per-file work here is the hottest path in the package. Everything derivable from the config
+// alone is computed once, above the loop.
+var NO_THROW = {
+    throwIfNoEntry: false
+};
+function listFiles(cfg, baseDir) {
+    var coverage = new Map();
+    var posixNeeded = _nodepath.sep === '\\';
+    var _iteratorNormalCompletion = true, _didIteratorError = false, _iteratorError = undefined;
+    try {
+        for(var _iterator = (0, _indexts.presetNames)(cfg)[Symbol.iterator](), _step; !(_iteratorNormalCompletion = (_step = _iterator.next()).done); _iteratorNormalCompletion = true){
+            var name = _step.value;
+            var preset = cfg.presets[name];
+            var _iteratorNormalCompletion1 = true, _didIteratorError1 = false, _iteratorError1 = undefined;
+            try {
+                for(var _iterator1 = (0, _nodefs.globSync)(preset.include, {
+                    cwd: baseDir,
+                    exclude: preset.exclude
+                })[Symbol.iterator](), _step1; !(_iteratorNormalCompletion1 = (_step1 = _iterator1.next()).done); _iteratorNormalCompletion1 = true){
+                    var matched = _step1.value;
+                    var _coverage_get;
+                    var relPath = posixNeeded ? toPosixPath(matched) : matched;
+                    var set = (_coverage_get = coverage.get(relPath)) !== null && _coverage_get !== void 0 ? _coverage_get : new Set();
+                    set.add(name);
+                    coverage.set(relPath, set);
+                }
+            } catch (err) {
+                _didIteratorError1 = true;
+                _iteratorError1 = err;
+            } finally{
+                try {
+                    if (!_iteratorNormalCompletion1 && _iterator1.return != null) {
+                        _iterator1.return();
+                    }
+                } finally{
+                    if (_didIteratorError1) {
+                        throw _iteratorError1;
+                    }
+                }
+            }
+        }
+    } catch (err) {
+        _didIteratorError = true;
+        _iteratorError = err;
+    } finally{
+        try {
+            if (!_iteratorNormalCompletion && _iterator.return != null) {
+                _iterator.return();
+            }
+        } finally{
+            if (_didIteratorError) {
+                throw _iteratorError;
+            }
+        }
+    }
+    // Which presets want vectors is a property of the config, not of any file.
+    var embedding = (0, _indexts.embedEnabled)(cfg);
+    var vectorPresets = embedding ? new Set((0, _indexts.presetNames)(cfg).filter(function(name) {
+        return (0, _indexts.presetHasSignal)(cfg, name, 'vectors');
+    })) : null;
+    var files = [];
+    var _iteratorNormalCompletion2 = true, _didIteratorError2 = false, _iteratorError2 = undefined;
+    try {
+        for(var _iterator2 = _to_consumable_array(coverage.keys()).sort()[Symbol.iterator](), _step2; !(_iteratorNormalCompletion2 = (_step2 = _iterator2.next()).done); _iteratorNormalCompletion2 = true){
+            var relPath1 = _step2.value;
+            var absPath = (0, _nodepath.join)(baseDir, relPath1); // join re-applies the platform separator for fs calls
+            // node:fs glob matches directories and dangling symlinks; fast-glob returned neither, so
+            // one stat filters both back out (throwIfNoEntry keeps a dangling link from throwing).
+            var st = (0, _nodefs.statSync)(absPath, NO_THROW);
+            if (!(st === null || st === void 0 ? void 0 : st.isFile())) continue;
+            var presets = _to_consumable_array(coverage.get(relPath1)).sort();
+            var embed = vectorPresets !== null && presets.some(function(name) {
+                return vectorPresets.has(name);
+            });
+            files.push({
+                relPath: relPath1,
+                absPath: absPath,
+                mtimeMs: st.mtimeMs,
+                ctimeMs: st.birthtimeMs,
+                size: st.size,
+                presets: presets,
+                embed: embed
+            });
+        }
+    } catch (err) {
+        _didIteratorError2 = true;
+        _iteratorError2 = err;
+    } finally{
+        try {
+            if (!_iteratorNormalCompletion2 && _iterator2.return != null) {
+                _iterator2.return();
+            }
+        } finally{
+            if (_didIteratorError2) {
+                throw _iteratorError2;
+            }
+        }
+    }
+    return files;
+}
+/* CJS INTEROP */ if (exports.__esModule && exports.default) { try { Object.defineProperty(exports.default, '__esModule', { value: true }); for (var key in exports) { exports.default[key] = exports[key]; } } catch (_) {}; module.exports = exports.default; }

package/dist/cjs/scan/list.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/scan/list.ts"],"sourcesContent":["import { globSync, statSync } from 'node:fs';\nimport { join, sep } from 'node:path';\nimport type { Config } from '../config/index.ts';\nimport { embedEnabled, presetHasSignal, presetNames } from '../config/index.ts';\n\nexport interface FileStat {\n  relPath: string;\n  absPath: string;\n  mtimeMs: number;\n  ctimeMs: number; // filesystem birthtime; a clone or copy resets it, like _mtime\n  size: number;\n  presets: string[]; // every declared preset covering this file (>= 1; union, overlap allowed)\n  embed: boolean; // true iff a model is named and some covering preset's signals include vectors\n}\n\n// Presets are views, not partitions: they overlap freely, and a file's covering set (not one\n// owner) drives indexing. Globs resolve relative to baseDir; unmatched files are not indexed.\nexport function toPosixPath(relPath: string, separator: string = sep): string {\n  return separator === '\\\\' ? relPath.split(separator).join('/') : relPath;\n}\n\n// Every command pays listFiles before it answers (the freshness check stats each file), so\n// per-file work here is the hottest path in the package. Everything derivable from the config\n// alone is computed once, above the loop.\nconst NO_THROW = { throwIfNoEntry: false } as const;\n\nexport function listFiles(cfg: Config, baseDir: string): FileStat[] {\n  const coverage = new Map<string, Set<string>>();\n  const posixNeeded = sep === '\\\\';\n  for (const name of presetNames(cfg)) {\n    const preset = cfg.presets[name];\n    for (const matched of globSync(preset.include, { cwd: baseDir, exclude: preset.exclude })) {\n      const relPath = posixNeeded ? toPosixPath(matched) : matched;\n      const set = coverage.get(relPath) ?? new Set<string>();\n      set.add(name);\n      coverage.set(relPath, set);\n    }\n  }\n\n  // Which presets want vectors is a property of the config, not of any file.\n  const embedding = embedEnabled(cfg);\n  const vectorPresets = embedding ? new Set(presetNames(cfg).filter((name) => presetHasSignal(cfg, name, 'vectors'))) : null;\n\n  const files: FileStat[] = [];\n  for (const relPath of [...coverage.keys()].sort()) {\n    const absPath = join(baseDir, relPath); // join re-applies the platform separator for fs calls\n    // node:fs glob matches directories and dangling symlinks; fast-glob returned neither, so\n    // one stat filters both back out (throwIfNoEntry keeps a dangling link from throwing).\n    const st = statSync(absPath, NO_THROW);\n    if (!st?.isFile()) continue;\n    const presets = [...(coverage.get(relPath) as Set<string>)].sort();\n    const embed = vectorPresets !== null && presets.some((name) => vectorPresets.has(name));\n    files.push({ relPath, absPath, mtimeMs: st.mtimeMs, ctimeMs: st.birthtimeMs, size: st.size, presets, embed });\n  }\n  return files;\n}\n"],"names":["listFiles","toPosixPath","relPath","separator","sep","split","join","NO_THROW","throwIfNoEntry","cfg","baseDir","coverage","Map","posixNeeded","presetNames","name","preset","presets","globSync","include","cwd","exclude","matched","set","get","Set","add","embedding","embedEnabled","vectorPresets","filter","presetHasSignal","files","keys","sort","absPath","st","statSync","isFile","embed","some","has","push","mtimeMs","ctimeMs","birthtimeMs","size"],"mappings":";;;;;;;;;;;QA0BgBA;eAAAA;;QATAC;eAAAA;;;sBAjBmB;wBACT;uBAEiC;;;;;;;;;;;;;;;;;;;;;;;;;;;;AAcpD,SAASA,YAAYC,OAAe;QAAEC,YAAAA,iEAAoBC,aAAG;IAClE,OAAOD,cAAc,OAAOD,QAAQG,KAAK,CAACF,WAAWG,IAAI,CAAC,OAAOJ;AACnE;AAEA,2FAA2F;AAC3F,8FAA8F;AAC9F,0CAA0C;AAC1C,IAAMK,WAAW;IAAEC,gBAAgB;AAAM;AAElC,SAASR,UAAUS,GAAW,EAAEC,OAAe;IACpD,IAAMC,WAAW,IAAIC;IACrB,IAAMC,cAAcT,aAAG,KAAK;QACvB,kCAAA,2BAAA;;QAAL,QAAK,YAAcU,IAAAA,oBAAW,EAACL,yBAA1B,SAAA,6BAAA,QAAA,yBAAA,iCAAgC;YAAhC,IAAMM,OAAN;YACH,IAAMC,SAASP,IAAIQ,OAAO,CAACF,KAAK;gBAC3B,mCAAA,4BAAA;;gBAAL,QAAK,aAAiBG,IAAAA,gBAAQ,EAACF,OAAOG,OAAO,EAAE;oBAAEC,KAAKV;oBAASW,SAASL,OAAOK,OAAO;gBAAC,uBAAlF,UAAA,8BAAA,SAAA,0BAAA,kCAAsF;oBAAtF,IAAMC,UAAN;wBAESX;oBADZ,IAAMT,UAAUW,cAAcZ,YAAYqB,WAAWA;oBACrD,IAAMC,OAAMZ,gBAAAA,SAASa,GAAG,CAACtB,sBAAbS,2BAAAA,gBAAyB,IAAIc;oBACzCF,IAAIG,GAAG,CAACX;oBACRJ,SAASY,GAAG,CAACrB,SAASqB;gBACxB;;gBALK;gBAAA;;;yBAAA,8BAAA;wBAAA;;;wBAAA;8BAAA;;;;QAMP;;QARK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAUL,2EAA2E;IAC3E,IAAMI,YAAYC,IAAAA,qBAAY,EAACnB;IAC/B,IAAMoB,gBAAgBF,YAAY,IAAIF,IAAIX,IAAAA,oBAAW,EAACL,KAAKqB,MAAM,CAAC,SAACf;eAASgB,IAAAA,wBAAe,EAACtB,KAAKM,MAAM;UAAe;IAEtH,IAAMiB,QAAoB,EAAE;QACvB,mCAAA,4BAAA;;QAAL,QAAK,aAAiB,AAAC,qBAAGrB,SAASsB,IAAI,IAAIC,IAAI,uBAA1C,UAAA,8BAAA,SAAA,0BAAA,kCAA8C;YAA9C,IAAMhC,WAAN;YACH,IAAMiC,UAAU7B,IAAAA,cAAI,EAACI,SAASR,WAAU,sDAAsD;YAC9F,yFAAyF;YACzF,uFAAuF;YACvF,IAAMkC,KAAKC,IAAAA,gBAAQ,EAACF,SAAS5B;YAC7B,IAAI,EAAC6B,eAAAA,yBAAAA,GAAIE,MAAM,KAAI;YACnB,IAAMrB,UAAU,AAAC,qBAAIN,SAASa,GAAG,CAACtB,WAA0BgC,IAAI;YAChE,IAAMK,QAAQV,kBAAkB,QAAQZ,QAAQuB,IAAI,CAAC,SAACzB;uBAASc,cAAcY,GAAG,CAAC1B;;YACjFiB,MAAMU,IAAI,CAAC;gBAAExC,SAAAA;gBAASiC,SAAAA;gBAASQ,SAASP,GAAGO,OAAO;gBAAEC,SAASR,GAAGS,WAAW;gBAAEC,MAAMV,GAAGU,IAAI;gBAAE7B,SAAAA;gBAASsB,OAAAA;YAAM;QAC7G;;QATK;QAAA;;;iBAAA,8BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAUL,OAAOP;AACT"}

package/dist/cjs/text/segment.d.cts

@@ -0,0 +1,3 @@
+export declare const UNSPACED_SCRIPTS = "\\p{scx=Han}\\p{scx=Hiragana}\\p{scx=Katakana}\\p{scx=Thai}\\p{scx=Khmer}\\p{scx=Lao}\\p{scx=Myanmar}";
+export declare function segmentField(text: string): string;
+export declare function segmentMatch(terms: string): string;

package/dist/cjs/text/segment.d.ts

@@ -0,0 +1,3 @@
+export declare const UNSPACED_SCRIPTS = "\\p{scx=Han}\\p{scx=Hiragana}\\p{scx=Katakana}\\p{scx=Thai}\\p{scx=Khmer}\\p{scx=Lao}\\p{scx=Myanmar}";
+export declare function segmentField(text: string): string;
+export declare function segmentMatch(terms: string): string;

package/dist/cjs/{segment.js → text/segment.js}

@@ -1,3 +1,14 @@
+// Word boundaries FTS5's tokenizers cannot find on their own.
+//
+// Segmentation is context-free: seg(query) appears inside seg(document) whenever the query is
+// a substring of the document. That is the contract -- substring semantics, what `grep` and
+// `LIKE '%..%'` give -- and it is what makes index and query agree on every input, always.
+// Intl.Segmenter's word mode is context-dependent (`东京都政府` splits as `东 | 京都 | 政府`
+// while the query `东京` splits as one word) and is rejected for that reason. Its grapheme
+// mode has no such dependency -- UAX #29 cluster boundaries never look past adjacent
+// characters -- so grapheme mode is used below; no dictionary, no word mode.
+// Scripts written without word spaces: a closed set of writing systems.
+// Script_Extensions, not Script: the katakana long-vowel mark ー is Script=Common.
 "use strict";
 Object.defineProperty(exports, "__esModule", {
     value: true
@@ -9,6 +20,9 @@ function _export(target, all) {
     });
 }
 _export(exports, {
+    get UNSPACED_SCRIPTS () {
+        return UNSPACED_SCRIPTS;
+    },
     get segmentField () {
         return segmentField;
     },
@@ -16,22 +30,11 @@ _export(exports, {
         return segmentMatch;
     }
 });
-// Word boundaries FTS5's tokenizers cannot find on their own.
-//
-// Segmentation is context-free: seg(query) appears inside seg(document) whenever the query is
-// a substring of the document. That is the contract -- substring semantics, what `grep` and
-// `LIKE '%..%'` give -- and it is what makes index and query agree on every input, always.
-// Intl.Segmenter's word mode is context-dependent (`东京都政府` splits as `东 | 京都 | 政府`
-// while the query `东京` splits as one word) and is rejected for that reason. Its grapheme
-// mode has no such dependency -- UAX #29 cluster boundaries never look past adjacent
-// characters -- so grapheme mode is used below; no dictionary, no word mode.
-// Scripts written without word spaces: a closed set of writing systems.
-// Script_Extensions, not Script: the katakana long-vowel mark ー is Script=Common.
-var SCRIPTS = '\\p{scx=Han}\\p{scx=Hiragana}\\p{scx=Katakana}\\p{scx=Thai}\\p{scx=Khmer}\\p{scx=Lao}\\p{scx=Myanmar}';
+var UNSPACED_SCRIPTS = '\\p{scx=Han}\\p{scx=Hiragana}\\p{scx=Katakana}\\p{scx=Thai}\\p{scx=Khmer}\\p{scx=Lao}\\p{scx=Myanmar}';
 // A run is script BASE characters with their combining marks attached; a bare mark after a
 // Latin letter (decomposed é) never starts one.
-var RUN = new RegExp("((?:[".concat(SCRIPTS, "]\\p{M}*)+)"), 'gu');
-var HAS_RUN = new RegExp("[".concat(SCRIPTS, "]"), 'u');
+var RUN = new RegExp("((?:[".concat(UNSPACED_SCRIPTS, "]\\p{M}*)+)"), 'gu');
+var HAS_RUN = new RegExp("[".concat(UNSPACED_SCRIPTS, "]"), 'u');
 // Grapheme clusters, ECMA-402/UAX #29: base char plus its marks, ZWJ sequences, Hangul jamo.
 // Built once (construction cost amortizes) and used for both index and query splitting.
 var GRAPHEME_SEGMENTER = new Intl.Segmenter(undefined, {

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

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/text/segment.ts"],"sourcesContent":["// Word boundaries FTS5's tokenizers cannot find on their own.\n//\n// Segmentation is context-free: seg(query) appears inside seg(document) whenever the query is\n// a substring of the document. That is the contract -- substring semantics, what `grep` and\n// `LIKE '%..%'` give -- and it is what makes index and query agree on every input, always.\n// Intl.Segmenter's word mode is context-dependent (`东京都政府` splits as `东 | 京都 | 政府`\n// while the query `东京` splits as one word) and is rejected for that reason. Its grapheme\n// mode has no such dependency -- UAX #29 cluster boundaries never look past adjacent\n// characters -- so grapheme mode is used below; no dictionary, no word mode.\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 = new RegExp(`((?:[${UNSPACED_SCRIPTS}]\\\\p{M}*)+)`, 'gu');\nconst HAS_RUN = new RegExp(`[${UNSPACED_SCRIPTS}]`, 'u');\n// Grapheme clusters, ECMA-402/UAX #29: base char plus its marks, ZWJ sequences, Hangul jamo.\n// Built once (construction cost amortizes) and used for both index and query splitting.\nconst GRAPHEME_SEGMENTER = new Intl.Segmenter(undefined, { granularity: 'grapheme' });\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  return Array.from(GRAPHEME_SEGMENTER.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 (the common case, paid for by\n// nothing). Otherwise each run explodes into its graphemes, barrier-delimited from its\n// 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// An unqualified run's target: raw title/summary/text drop punctuation as unicode61's token\n// separator, so two adjacent single-grapheme phrase positions match across punctuation the same\n// as across a real gap (`数数` vs `数。数`) -- a false positive only the barriered `_seg` columns\n// are safe from. FTS5's column-set filter, not parens+OR, so the group still composes under\n// AND/OR/NOT/juxtaposition exactly like the single-column qualifier form below.\nconst SIDECAR_COLUMNS = '{title_seg summary_seg text_seg}:';\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): string {\n  return splitOnPunctuation(run)\n    .map((g) => `${columnPrefix}${g.length > 1 ? `\"${g.join(' ')}\"` : g[0]}`)\n    .join(' ');\n}\n\n// Query side: each unspaced run becomes phrases of its graphemes, matching how segmentField\n// indexed it. A qualifier ahead of such a run maps to its `_seg` column; unqualified maps to all\n// three (SIDECAR_COLUMNS), since raw columns cannot express the contract. An author's own quoted\n// phrase is their explicit escape hatch to FTS5's native syntax, and passes through byte-identical.\nexport function segmentMatch(terms: string): string {\n  if (!HAS_RUN.test(terms)) return terms;\n  let out = '';\n  let quoted = false;\n  const pieces = terms.split(RUN); // split keeps captured runs at odd indices\n  for (let i = 0; i < pieces.length; i++) {\n    if (i % 2 === 0) {\n      for (const ch of pieces[i]) if (ch === '\"') quoted = !quoted;\n      out += pieces[i];\n      continue;\n    }\n    if (quoted) {\n      out += pieces[i]; // an author's phrase is matched as written\n      continue;\n    }\n    const m = out.match(QUALIFIER);\n    if (m) {\n      out = `${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  }\n  return out;\n}\n"],"names":["UNSPACED_SCRIPTS","segmentField","segmentMatch","RUN","RegExp","HAS_RUN","GRAPHEME_SEGMENTER","Intl","Segmenter","undefined","granularity","PUNCTUATION","BARRIER","graphemes","run","Array","from","segment","s","splitOnPunctuation","groups","g","test","push","length","filter","text","out","replace","body","map","join","trim","QUALIFIER","SIDECAR_COLUMNS","runQuery","columnPrefix","terms","quoted","pieces","split","i","ch","m","match","slice","index"],"mappings":"AAAA,8DAA8D;AAC9D,EAAE;AACF,8FAA8F;AAC9F,4FAA4F;AAC5F,2FAA2F;AAC3F,mFAAmF;AACnF,yFAAyF;AACzF,qFAAqF;AACrF,6EAA6E;AAE7E,wEAAwE;AACxE,kFAAkF;;;;;;;;;;;;QACrEA;eAAAA;;QAiCGC;eAAAA;;QAkCAC;eAAAA;;;AAnET,IAAMF,mBAAmB;AAChC,2FAA2F;AAC3F,gDAAgD;AAChD,IAAMG,MAAM,IAAIC,OAAO,AAAC,QAAwB,OAAjBJ,kBAAiB,gBAAc;AAC9D,IAAMK,UAAU,IAAID,OAAO,AAAC,IAAoB,OAAjBJ,kBAAiB,MAAI;AACpD,6FAA6F;AAC7F,wFAAwF;AACxF,IAAMM,qBAAqB,IAAIC,KAAKC,SAAS,CAACC,WAAW;IAAEC,aAAa;AAAW;AACnF,8FAA8F;AAC9F,8FAA8F;AAC9F,IAAMC,cAAc;AACpB,6FAA6F;AAC7F,0FAA0F;AAC1F,IAAMC,UAAU;AAEhB,SAASC,UAAUC,GAAW;IAC5B,OAAOC,MAAMC,IAAI,CAACV,mBAAmBW,OAAO,CAACH,MAAM,SAACI;eAAMA,EAAED,OAAO;;AACrE;AAEA,8FAA8F;AAC9F,0FAA0F;AAC1F,SAASE,mBAAmBL,GAAW;IACrC,IAAMM,SAAqB;QAAC,EAAE;KAAC;QAC1B,kCAAA,2BAAA;;QAAL,QAAK,YAAWP,UAAUC,yBAArB,SAAA,6BAAA,QAAA,yBAAA,iCAA2B;YAA3B,IAAMO,IAAN;YACH,IAAIV,YAAYW,IAAI,CAACD,IAAID,OAAOG,IAAI,CAAC,EAAE;iBAClCH,MAAM,CAACA,OAAOI,MAAM,GAAG,EAAE,CAACD,IAAI,CAACF;QACtC;;QAHK;QAAA;;;iBAAA,6BAAA;gBAAA;;;gBAAA;sBAAA;;;;IAIL,OAAOD,OAAOK,MAAM,CAAC,SAACJ;eAAMA,EAAEG,MAAM,GAAG;;AACzC;AAKO,SAASvB,aAAayB,IAAY;IACvC,IAAI,CAACrB,QAAQiB,IAAI,CAACI,OAAO,OAAO;IAChC,IAAMC,MAAMD,KAAKE,OAAO,CAACzB,KAAK,SAACW;QAC7B,IAAMe,OAAOV,mBAAmBL,KAC7BgB,GAAG,CAAC,SAACT;mBAAMA,EAAEU,IAAI,CAAC;WAClBA,IAAI,CAAC,AAAC,IAAW,OAARnB,SAAQ;QACpB,OAAO,AAAC,IAAciB,OAAXjB,SAAQ,KAAWA,OAARiB,MAAK,KAAW,OAARjB,SAAQ;IACxC;IACA,OAAOe,IAAIC,OAAO,CAAC,QAAQ,KAAKI,IAAI;AACtC;AAEA,0FAA0F;AAC1F,wFAAwF;AACxF,IAAMC,YAAY;AAElB,4FAA4F;AAC5F,gGAAgG;AAChG,6FAA6F;AAC7F,4FAA4F;AAC5F,gFAAgF;AAChF,IAAMC,kBAAkB;AAExB,4FAA4F;AAC5F,yFAAyF;AACzF,SAASC,SAASrB,GAAW,EAAEsB,YAAoB;IACjD,OAAOjB,mBAAmBL,KACvBgB,GAAG,CAAC,SAACT;eAAM,AAAC,GAAiBA,OAAfe,cAAwD,OAAzCf,EAAEG,MAAM,GAAG,IAAI,AAAC,IAAe,OAAZH,EAAEU,IAAI,CAAC,MAAK,OAAKV,CAAC,CAAC,EAAE;OACrEU,IAAI,CAAC;AACV;AAMO,SAAS7B,aAAamC,KAAa;IACxC,IAAI,CAAChC,QAAQiB,IAAI,CAACe,QAAQ,OAAOA;IACjC,IAAIV,MAAM;IACV,IAAIW,SAAS;IACb,IAAMC,SAASF,MAAMG,KAAK,CAACrC,MAAM,2CAA2C;IAC5E,IAAK,IAAIsC,IAAI,GAAGA,IAAIF,OAAOf,MAAM,EAAEiB,IAAK;QACtC,IAAIA,IAAI,MAAM,GAAG;gBACV,kCAAA,2BAAA;;gBAAL,QAAK,YAAYF,MAAM,CAACE,EAAE,qBAArB,SAAA,6BAAA,QAAA,yBAAA;oBAAA,IAAMC,KAAN;oBAAuB,IAAIA,OAAO,KAAKJ,SAAS,CAACA;;;gBAAjD;gBAAA;;;yBAAA,6BAAA;wBAAA;;;wBAAA;8BAAA;;;;YACLX,OAAOY,MAAM,CAACE,EAAE;YAChB;QACF;QACA,IAAIH,QAAQ;YACVX,OAAOY,MAAM,CAACE,EAAE,EAAE,2CAA2C;YAC7D;QACF;QACA,IAAME,IAAIhB,IAAIiB,KAAK,CAACX;QACpB,IAAIU,GAAG;YACLhB,MAAM,AAAC,GAA0BgB,OAAxBhB,IAAIkB,KAAK,CAAC,GAAGF,EAAEG,KAAK,GAAWX,OAAPQ,CAAC,CAAC,EAAE,EAA8C,OAA3CR,SAASI,MAAM,CAACE,EAAE,EAAE,AAAC,GAASE,OAAPA,CAAC,CAAC,EAAE,EAAQ,OAALA,CAAC,CAAC,EAAE,EAAC;QAC7E,OAAO;YACLhB,OAAOQ,SAASI,MAAM,CAACE,EAAE,EAAEP;QAC7B;IACF;IACA,OAAOP;AACT"}

package/dist/cjs/text/strip.d.cts

@@ -0,0 +1,3 @@
+import type { Block } from '../chunk/index.js';
+export declare function textFromBlocks(blocks: Block[]): string;
+export declare function stripText(value: string): string;

package/dist/cjs/text/strip.d.ts

@@ -0,0 +1,3 @@
+import type { Block } from '../chunk/index.js';
+export declare function textFromBlocks(blocks: Block[]): string;
+export declare function stripText(value: string): string;

package/dist/cjs/text/strip.js

@@ -0,0 +1,32 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", {
+    value: true
+});
+function _export(target, all) {
+    for(var name in all)Object.defineProperty(target, name, {
+        enumerable: true,
+        get: Object.getOwnPropertyDescriptor(all, name).get
+    });
+}
+_export(exports, {
+    get stripText () {
+        return stripText;
+    },
+    get textFromBlocks () {
+        return textFromBlocks;
+    }
+});
+var _indexts = require("../chunk/index.js");
+var _frontmatterts = require("../scan/frontmatter.js");
+function textFromBlocks(blocks) {
+    var text = blocks.map(function(block) {
+        return (0, _indexts.extractText)(block.node);
+    }).filter(function(s) {
+        return s.length > 0;
+    }).join('\n');
+    return (0, _frontmatterts.normalizeText)(text);
+}
+function stripText(value) {
+    return textFromBlocks((0, _indexts.parse)(value));
+}
+/* CJS INTEROP */ if (exports.__esModule && exports.default) { try { Object.defineProperty(exports.default, '__esModule', { value: true }); for (var key in exports) { exports.default[key] = exports[key]; } } catch (_) {}; module.exports = exports.default; }

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

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/text/strip.ts"],"sourcesContent":["import type { Block } from '../chunk/index.ts';\nimport { extractText, parse } from '../chunk/index.ts';\nimport { normalizeText } from '../scan/frontmatter.ts';\n\n// Flat prose from blocks a caller already parsed, via the same extractText 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  const text = blocks\n    .map((block) => extractText(block.node))\n    .filter((s) => s.length > 0)\n    .join('\\n');\n  return normalizeText(text);\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","text","map","block","extractText","node","filter","s","length","join","normalizeText","value","parse"],"mappings":";;;;;;;;;;;QAegBA;eAAAA;;QATAC;eAAAA;;;uBALmB;6BACL;AAIvB,SAASA,eAAeC,MAAe;IAC5C,IAAMC,OAAOD,OACVE,GAAG,CAAC,SAACC;eAAUC,IAAAA,oBAAW,EAACD,MAAME,IAAI;OACrCC,MAAM,CAAC,SAACC;eAAMA,EAAEC,MAAM,GAAG;OACzBC,IAAI,CAAC;IACR,OAAOC,IAAAA,4BAAa,EAACT;AACvB;AAGO,SAASH,UAAUa,KAAa;IACrC,OAAOZ,eAAea,IAAAA,cAAK,EAACD;AAC9B"}

package/dist/esm/chunk/extract.d.ts

@@ -0,0 +1,2 @@
+import type { Nodes } from 'mdast';
+export declare function extractText(node: Nodes): string;

package/dist/esm/chunk/extract.js

@@ -0,0 +1,114 @@
+// mdast has no wikilink, embed, or %%comment%% node -- literal text resolved by regex passes
+// over non-code text (code/inlineCode is placeholder-held and spliced back verbatim, below).
+const COMMENT_PAIR = /%%[\s\S]*?%%/g;
+// An embed always yields its target, whatever follows a pipe (a resize suffix, never an alias);
+// must run before the wikilink regex below, bang included, or that regex matches first.
+const EMBED = /!\[\[([^\]]+)\]\]/g;
+const WIKILINK = /\[\[([^\]]+)\]\]/g;
+// A trailing " ^id" is an Obsidian block anchor, valid only at a line's end; a caret elsewhere
+// (mid-line) is ordinary text and is left alone.
+const BLOCK_ID = / \^[A-Za-z0-9-]+$/gm;
+// A callout marker is only ever the first thing in a blockquote's text (Obsidian's grammar),
+// so this anchors to blockquote output alone -- never applied to prose in general.
+const CALLOUT_MARKER = /^\[!\w[\w-]*\][+-]?[ \t]?/;
+// An html node's value is raw HTML, block or inline: a full <!-- --> comment is dropped, any
+// remaining tags are stripped, and the text a browser would still render survives.
+const HTML_COMMENT = /<!--[\s\S]*?-->/g;
+const HTML_TAG = /<[^>]+>/g;
+function stripHtml(value) {
+    return value.replace(HTML_COMMENT, '').replace(HTML_TAG, '');
+}
+// A `#anchor` keeps its text (hash dropped), unlike links.ts's parseWikilinkInner (the resolver
+// authority), which discards it entirely -- replicated here since that return value doesn't fit.
+function anchorText(base) {
+    const hashIdx = base.indexOf('#');
+    if (hashIdx === -1) return base.trim();
+    const target = base.slice(0, hashIdx).trim();
+    const anchor = base.slice(hashIdx + 1).trim();
+    return target ? `${target} ${anchor}` : anchor;
+}
+function wikilinkText(inner) {
+    const pipeIdx = inner.indexOf('|');
+    return pipeIdx === -1 ? anchorText(inner) : inner.slice(pipeIdx + 1).trim();
+}
+function embedText(inner) {
+    const pipeIdx = inner.indexOf('|');
+    return anchorText(pipeIdx === -1 ? inner : inner.slice(0, pipeIdx));
+}
+function resolveFlavor(text) {
+    return text.replace(COMMENT_PAIR, '').replace(EMBED, (_, inner)=>embedText(inner)).replace(WIKILINK, (_, inner)=>wikilinkText(inner)).replace(BLOCK_ID, '');
+}
+// Sibling blocks (list items, blockquote content, footnote bodies) joined one per line;
+// empties dropped so a blank child never widens the gap between its neighbors.
+function joinBlocks(nodes, code) {
+    return nodes.map((n)=>extractNode(n, code)).filter((s)=>s.length > 0).join('\n');
+}
+// Inline content (headings, paragraphs, emphasis, links) concatenated with no separator: the
+// source text nodes already carry whatever spacing separates them.
+function joinInline(nodes, code) {
+    return nodes.map((n)=>extractNode(n, code)).join('');
+}
+// U+E000 (Private Use Area) never appears in real markdown text, so it is a collision-free
+// placeholder delimiter, spliced back to the literal code value after flavor resolution.
+function codePlaceholder(value, code) {
+    const idx = code.push(value) - 1;
+    return `\uE000${idx}\uE000`;
+}
+// A GFM autolink's display text is its own target, missing only the scheme
+// mdast-util-gfm-autolink-literal fills into node.url (http:// for www., mailto: for an email).
+const SCHEME_PREFIX = /^[a-z][a-z0-9+.-]*:\/{0,2}/i;
+function isAutolink(url, text) {
+    return url === text || url.replace(SCHEME_PREFIX, '') === text;
+}
+function extractNode(node, code) {
+    switch(node.type){
+        case 'root':
+        case 'list':
+        case 'listItem':
+        case 'footnoteDefinition':
+            return joinBlocks(node.children, code);
+        case 'blockquote':
+            return joinBlocks(node.children, code).replace(CALLOUT_MARKER, '');
+        case 'link':
+            {
+                const text = joinInline(node.children, code);
+                return isAutolink(node.url, text) ? '' : text;
+            }
+        case 'heading':
+        case 'paragraph':
+        case 'linkReference':
+        case 'emphasis':
+        case 'strong':
+        case 'delete':
+            return joinInline(node.children, code);
+        case 'table':
+            return node.children.map((n)=>extractNode(n, code)).join('\n');
+        case 'tableRow':
+            return node.children.map((n)=>extractNode(n, code)).join(' ');
+        case 'tableCell':
+            return joinInline(node.children, code);
+        case 'image':
+        case 'imageReference':
+            var _node_alt;
+            return (_node_alt = node.alt) !== null && _node_alt !== void 0 ? _node_alt : '';
+        case 'text':
+            return node.value;
+        case 'inlineCode':
+        case 'code':
+            return codePlaceholder(node.value, code);
+        case 'html':
+            return stripHtml(node.value);
+        case 'break':
+            return ' ';
+        default:
+            return '';
+    }
+}
+const CODE_PLACEHOLDER = /\uE000(\d+)\uE000/g;
+// Plain text of one mdast node (or a whole tree): heading/list/table structure is kept as text,
+// markup (emphasis, link targets, task and callout markers) is dropped. Pure, synchronous.
+export function extractText(node) {
+    const code = [];
+    const resolved = resolveFlavor(extractNode(node, code));
+    return code.length === 0 ? resolved : resolved.replace(CODE_PLACEHOLDER, (_, i)=>code[Number(i)]);
+}

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

@@ -0,0 +1 @@
+{"version":3,"sources":["/Users/kevin/Dev/OpenSource/ai/sensemaking/src/chunk/extract.ts"],"sourcesContent":["import type { Nodes } from 'mdast';\n\n// mdast has no wikilink, embed, or %%comment%% node -- literal text resolved by regex passes\n// over non-code text (code/inlineCode is placeholder-held and spliced back verbatim, below).\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-]+$/gm;\n// A callout marker is only ever the first thing in a blockquote's text (Obsidian's grammar),\n// so this anchors to blockquote output alone -- never applied to prose in general.\nconst CALLOUT_MARKER = /^\\[!\\w[\\w-]*\\][+-]?[ \\t]?/;\n// An html node's value is raw HTML, block or inline: a full <!-- --> comment is dropped, any\n// remaining tags are stripped, and the text a browser would still render survives.\nconst HTML_COMMENT = /<!--[\\s\\S]*?-->/g;\nconst HTML_TAG = /<[^>]+>/g;\n\nfunction stripHtml(value: string): string {\n  return value.replace(HTML_COMMENT, '').replace(HTML_TAG, '');\n}\n\n// A `#anchor` keeps its text (hash dropped), unlike links.ts's parseWikilinkInner (the resolver\n// authority), which discards it entirely -- replicated here since that return value doesn't fit.\nfunction anchorText(base: string): string {\n  const hashIdx = base.indexOf('#');\n  if (hashIdx === -1) return base.trim();\n  const target = base.slice(0, hashIdx).trim();\n  const anchor = base.slice(hashIdx + 1).trim();\n  return target ? `${target} ${anchor}` : anchor;\n}\n\nfunction wikilinkText(inner: string): string {\n  const pipeIdx = inner.indexOf('|');\n  return pipeIdx === -1 ? anchorText(inner) : inner.slice(pipeIdx + 1).trim();\n}\n\nfunction embedText(inner: string): string {\n  const pipeIdx = inner.indexOf('|');\n  return anchorText(pipeIdx === -1 ? inner : inner.slice(0, pipeIdx));\n}\n\nfunction resolveFlavor(text: string): string {\n  return text\n    .replace(COMMENT_PAIR, '')\n    .replace(EMBED, (_, inner: string) => embedText(inner))\n    .replace(WIKILINK, (_, inner: string) => wikilinkText(inner))\n    .replace(BLOCK_ID, '');\n}\n\n// Sibling blocks (list items, blockquote content, footnote bodies) joined one per line;\n// empties dropped so a blank child never widens the gap between its neighbors.\nfunction joinBlocks(nodes: Nodes[], code: string[]): string {\n  return nodes\n    .map((n) => extractNode(n, code))\n    .filter((s) => s.length > 0)\n    .join('\\n');\n}\n\n// Inline content (headings, paragraphs, emphasis, links) concatenated with no separator: the\n// source text nodes already carry whatever spacing separates them.\nfunction joinInline(nodes: Nodes[], code: string[]): string {\n  return nodes.map((n) => extractNode(n, code)).join('');\n}\n\n// U+E000 (Private Use Area) never appears in real markdown text, so it is a collision-free\n// placeholder delimiter, spliced back to the literal code value after flavor resolution.\nfunction codePlaceholder(value: string, code: string[]): string {\n  const idx = code.push(value) - 1;\n  return `\\uE000${idx}\\uE000`;\n}\n\n// A GFM autolink's display text is its own target, missing only the scheme\n// mdast-util-gfm-autolink-literal fills into node.url (http:// for www., mailto: for an email).\nconst SCHEME_PREFIX = /^[a-z][a-z0-9+.-]*:\\/{0,2}/i;\nfunction isAutolink(url: string, text: string): boolean {\n  return url === text || url.replace(SCHEME_PREFIX, '') === text;\n}\n\nfunction extractNode(node: Nodes, code: string[]): string {\n  switch (node.type) {\n    case 'root':\n    case 'list':\n    case 'listItem':\n    case 'footnoteDefinition':\n      return joinBlocks(node.children, code);\n    case 'blockquote':\n      return joinBlocks(node.children, code).replace(CALLOUT_MARKER, '');\n    case 'link': {\n      const text = joinInline(node.children, code);\n      return isAutolink(node.url, text) ? '' : text;\n    }\n    case 'heading':\n    case 'paragraph':\n    case 'linkReference':\n    case 'emphasis':\n    case 'strong':\n    case 'delete':\n      return joinInline(node.children, code);\n    case 'table':\n      return node.children.map((n) => extractNode(n, code)).join('\\n');\n    case 'tableRow':\n      return node.children.map((n) => extractNode(n, code)).join(' ');\n    case 'tableCell':\n      return joinInline(node.children, code);\n    case 'image':\n    case 'imageReference':\n      return node.alt ?? '';\n    case 'text':\n      return node.value;\n    case 'inlineCode':\n    case 'code':\n      return codePlaceholder(node.value, code);\n    case 'html':\n      return stripHtml(node.value);\n    case 'break':\n      return ' ';\n    default:\n      return '';\n  }\n}\n\nconst CODE_PLACEHOLDER = /\\uE000(\\d+)\\uE000/g;\n\n// Plain text of one mdast node (or a whole tree): heading/list/table structure is kept as text,\n// markup (emphasis, link targets, task and callout markers) is dropped. Pure, synchronous.\nexport function extractText(node: Nodes): string {\n  const code: string[] = [];\n  const resolved = resolveFlavor(extractNode(node, code));\n  return code.length === 0 ? resolved : resolved.replace(CODE_PLACEHOLDER, (_, i: string) => code[Number(i)]);\n}\n"],"names":["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","_","joinBlocks","nodes","code","map","n","extractNode","filter","s","length","join","joinInline","codePlaceholder","idx","push","SCHEME_PREFIX","isAutolink","url","node","type","children","alt","CODE_PLACEHOLDER","extractText","resolved","i","Number"],"mappings":"AAEA,6FAA6F;AAC7F,6FAA6F;AAC7F,MAAMA,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,6FAA6F;AAC7F,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,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,wFAAwF;AACxF,+EAA+E;AAC/E,SAASsB,WAAWC,KAAc,EAAEC,IAAc;IAChD,OAAOD,MACJE,GAAG,CAAC,CAACC,IAAMC,YAAYD,GAAGF,OAC1BI,MAAM,CAAC,CAACC,IAAMA,EAAEC,MAAM,GAAG,GACzBC,IAAI,CAAC;AACV;AAEA,6FAA6F;AAC7F,mEAAmE;AACnE,SAASC,WAAWT,KAAc,EAAEC,IAAc;IAChD,OAAOD,MAAME,GAAG,CAAC,CAACC,IAAMC,YAAYD,GAAGF,OAAOO,IAAI,CAAC;AACrD;AAEA,2FAA2F;AAC3F,yFAAyF;AACzF,SAASE,gBAAgB5B,KAAa,EAAEmB,IAAc;IACpD,MAAMU,MAAMV,KAAKW,IAAI,CAAC9B,SAAS;IAC/B,OAAO,CAAC,MAAM,EAAE6B,IAAI,MAAM,CAAC;AAC7B;AAEA,2EAA2E;AAC3E,gGAAgG;AAChG,MAAME,gBAAgB;AACtB,SAASC,WAAWC,GAAW,EAAElB,IAAY;IAC3C,OAAOkB,QAAQlB,QAAQkB,IAAIhC,OAAO,CAAC8B,eAAe,QAAQhB;AAC5D;AAEA,SAASO,YAAYY,IAAW,EAAEf,IAAc;IAC9C,OAAQe,KAAKC,IAAI;QACf,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;YACH,OAAOlB,WAAWiB,KAAKE,QAAQ,EAAEjB;QACnC,KAAK;YACH,OAAOF,WAAWiB,KAAKE,QAAQ,EAAEjB,MAAMlB,OAAO,CAACL,gBAAgB;QACjE,KAAK;YAAQ;gBACX,MAAMmB,OAAOY,WAAWO,KAAKE,QAAQ,EAAEjB;gBACvC,OAAOa,WAAWE,KAAKD,GAAG,EAAElB,QAAQ,KAAKA;YAC3C;QACA,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;QACL,KAAK;YACH,OAAOY,WAAWO,KAAKE,QAAQ,EAAEjB;QACnC,KAAK;YACH,OAAOe,KAAKE,QAAQ,CAAChB,GAAG,CAAC,CAACC,IAAMC,YAAYD,GAAGF,OAAOO,IAAI,CAAC;QAC7D,KAAK;YACH,OAAOQ,KAAKE,QAAQ,CAAChB,GAAG,CAAC,CAACC,IAAMC,YAAYD,GAAGF,OAAOO,IAAI,CAAC;QAC7D,KAAK;YACH,OAAOC,WAAWO,KAAKE,QAAQ,EAAEjB;QACnC,KAAK;QACL,KAAK;gBACIe;YAAP,QAAOA,YAAAA,KAAKG,GAAG,cAARH,uBAAAA,YAAY;QACrB,KAAK;YACH,OAAOA,KAAKlC,KAAK;QACnB,KAAK;QACL,KAAK;YACH,OAAO4B,gBAAgBM,KAAKlC,KAAK,EAAEmB;QACrC,KAAK;YACH,OAAOpB,UAAUmC,KAAKlC,KAAK;QAC7B,KAAK;YACH,OAAO;QACT;YACE,OAAO;IACX;AACF;AAEA,MAAMsC,mBAAmB;AAEzB,gGAAgG;AAChG,2FAA2F;AAC3F,OAAO,SAASC,YAAYL,IAAW;IACrC,MAAMf,OAAiB,EAAE;IACzB,MAAMqB,WAAW1B,cAAcQ,YAAYY,MAAMf;IACjD,OAAOA,KAAKM,MAAM,KAAK,IAAIe,WAAWA,SAASvC,OAAO,CAACqC,kBAAkB,CAACtB,GAAGyB,IAActB,IAAI,CAACuB,OAAOD,GAAG;AAC5G"}

package/dist/esm/chunk/group.d.ts

@@ -0,0 +1,4 @@
+import type { Block, Chunk, ChunkOptions } from './types.js';
+export declare const DEFAULT_TARGET_TOKENS = 500;
+export declare function estimateTokens(text: string): number;
+export declare function group(blocks: Block[], body: string, opts?: ChunkOptions): Chunk[];