@cesarandreslopez/occ 0.1.1 → 0.2.0

package/dist/src/progress.d.ts

@@ -0,0 +1,10 @@
+export interface ProgressBar {
+    update(increment: number, detail?: string): void;
+    done(): void;
+}
+export interface ProgressOptions {
+    total: number;
+    label?: string;
+    enabled?: boolean;
+}
+export declare function createProgress({ total, label, enabled }: ProgressOptions): ProgressBar;

package/dist/src/progress.js

@@ -0,0 +1,38 @@
+import { basename } from 'node:path';
+const BAR_WIDTH = 20;
+function noop() { }
+const noopProgress = { update: noop, done: noop };
+export function createProgress({ total, label = 'Parsing', enabled = true }) {
+    if (!enabled || !process.stderr.isTTY || total === 0)
+        return noopProgress;
+    let completed = 0;
+    const startTime = Date.now();
+    const cols = process.stderr.columns || 80;
+    function render(detail) {
+        const pct = Math.round((completed / total) * 100);
+        const filled = Math.round((completed / total) * BAR_WIDTH);
+        const bar = '\u2588'.repeat(filled) + '\u2591'.repeat(BAR_WIDTH - filled);
+        let remaining = '';
+        if (completed > 0) {
+            const elapsed = (Date.now() - startTime) / 1000;
+            const eta = Math.round((elapsed / completed) * (total - completed));
+            remaining = ` ~${eta}s remaining`;
+        }
+        const name = detail ? ' ' + basename(detail) : '';
+        let line = `${label} [${bar}] ${completed}/${total} (${pct}%)${name}${remaining}`;
+        if (line.length > cols)
+            line = line.slice(0, cols);
+        line = line.padEnd(cols);
+        process.stderr.write('\r' + line);
+    }
+    return {
+        update(increment, detail) {
+            completed += increment;
+            render(detail);
+        },
+        done() {
+            process.stderr.write('\r' + ' '.repeat(cols) + '\r');
+        },
+    };
+}
+//# sourceMappingURL=progress.js.map

package/dist/src/progress.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"progress.js","sourceRoot":"","sources":["../../src/progress.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,WAAW,CAAC;AAErC,MAAM,SAAS,GAAG,EAAE,CAAC;AAarB,SAAS,IAAI,KAAI,CAAC;AAClB,MAAM,YAAY,GAAgB,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC;AAE/D,MAAM,UAAU,cAAc,CAAC,EAAE,KAAK,EAAE,KAAK,GAAG,SAAS,EAAE,OAAO,GAAG,IAAI,EAAmB;IAC1F,IAAI,CAAC,OAAO,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,KAAK,IAAI,KAAK,KAAK,CAAC;QAAE,OAAO,YAAY,CAAC;IAE1E,IAAI,SAAS,GAAG,CAAC,CAAC;IAClB,MAAM,SAAS,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;IAC7B,MAAM,IAAI,GAAG,OAAO,CAAC,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC;IAE1C,SAAS,MAAM,CAAC,MAAe;QAC7B,MAAM,GAAG,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,SAAS,GAAG,KAAK,CAAC,GAAG,GAAG,CAAC,CAAC;QAClD,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,SAAS,GAAG,KAAK,CAAC,GAAG,SAAS,CAAC,CAAC;QAC3D,MAAM,GAAG,GAAG,QAAQ,CAAC,MAAM,CAAC,MAAM,CAAC,GAAG,QAAQ,CAAC,MAAM,CAAC,SAAS,GAAG,MAAM,CAAC,CAAC;QAE1E,IAAI,SAAS,GAAG,EAAE,CAAC;QACnB,IAAI,SAAS,GAAG,CAAC,EAAE,CAAC;YAClB,MAAM,OAAO,GAAG,CAAC,IAAI,CAAC,GAAG,EAAE,GAAG,SAAS,CAAC,GAAG,IAAI,CAAC;YAChD,MAAM,GAAG,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,OAAO,GAAG,SAAS,CAAC,GAAG,CAAC,KAAK,GAAG,SAAS,CAAC,CAAC,CAAC;YACpE,SAAS,GAAG,KAAK,GAAG,aAAa,CAAC;QACpC,CAAC;QAED,MAAM,IAAI,GAAG,MAAM,CAAC,CAAC,CAAC,GAAG,GAAG,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;QAClD,IAAI,IAAI,GAAG,GAAG,KAAK,KAAK,GAAG,KAAK,SAAS,IAAI,KAAK,KAAK,GAAG,KAAK,IAAI,GAAG,SAAS,EAAE,CAAC;QAElF,IAAI,IAAI,CAAC,MAAM,GAAG,IAAI;YAAE,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC;QACnD,IAAI,GAAG,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC;QAEzB,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,IAAI,GAAG,IAAI,CAAC,CAAC;IACpC,CAAC;IAED,OAAO;QACL,MAAM,CAAC,SAAiB,EAAE,MAAe;YACvC,SAAS,IAAI,SAAS,CAAC;YACvB,MAAM,CAAC,MAAM,CAAC,CAAC;QACjB,CAAC;QACD,IAAI;YACF,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,IAAI,GAAG,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,CAAC;QACvD,CAAC;KACF,CAAC;AACJ,CAAC"}

package/dist/src/scc.d.ts

@@ -0,0 +1,28 @@
+export interface SccLanguage {
+    Name: string;
+    Count: number;
+    Lines: number;
+    Blank: number;
+    Comment: number;
+    Code: number;
+    Files?: SccFile[];
+    [key: string]: unknown;
+}
+export interface SccFile {
+    Filename?: string;
+    Location?: string;
+    Lines: number;
+    Blank: number;
+    Comment: number;
+    Code: number;
+    [key: string]: unknown;
+}
+export interface RunSccOptions {
+    byFile?: boolean;
+    excludeDir?: string[];
+    sort?: string;
+    ci?: boolean;
+    noGitignore?: boolean;
+}
+export declare function checkScc(): Promise<string>;
+export declare function runScc(sccBinary: string | null, directories: string[], options?: RunSccOptions): Promise<SccLanguage[]>;

package/dist/src/scc.js

@@ -0,0 +1,83 @@
+import { execFile } from 'node:child_process';
+import { promisify } from 'node:util';
+import { existsSync } from 'node:fs';
+import { fileURLToPath } from 'node:url';
+import path from 'node:path';
+import { OFFICE_EXTENSIONS } from './utils.js';
+const execFileAsync = promisify(execFile);
+function getVendoredSccPath() {
+    const __dirname = path.dirname(fileURLToPath(import.meta.url));
+    const binary = process.platform === 'win32' ? 'scc.exe' : 'scc';
+    // Works from both src/ (dev) and dist/src/ (built)
+    const candidate1 = path.join(__dirname, '..', 'vendor', binary);
+    const candidate2 = path.join(__dirname, '..', '..', 'vendor', binary);
+    if (existsSync(candidate1))
+        return candidate1;
+    return candidate2;
+}
+async function findScc() {
+    // Prefer vendored binary
+    const vendored = getVendoredSccPath();
+    if (existsSync(vendored))
+        return vendored;
+    // Fall back to PATH
+    try {
+        const cmd = process.platform === 'win32' ? 'where' : 'which';
+        const { stdout } = await execFileAsync(cmd, ['scc']);
+        return stdout.trim().split('\n')[0];
+    }
+    catch {
+        return null;
+    }
+}
+export async function checkScc() {
+    const binary = await findScc();
+    if (!binary) {
+        throw new Error('scc is required but not found.\n' +
+            'Run "npm install" to auto-download it, or install manually from https://github.com/boyter/scc');
+    }
+    return binary;
+}
+export async function runScc(sccBinary, directories, options = {}) {
+    const { byFile = false, excludeDir = [], sort, ci = false, noGitignore = false, } = options;
+    if (!sccBinary)
+        return [];
+    const args = ['--format', 'json'];
+    // Exclude office extensions
+    args.push('--exclude-ext', OFFICE_EXTENSIONS.join(','));
+    if (byFile)
+        args.push('--by-file');
+    if (ci)
+        args.push('--ci');
+    if (noGitignore)
+        args.push('--no-gitignore');
+    for (const dir of excludeDir) {
+        args.push('--exclude-dir', dir);
+    }
+    if (sort) {
+        const sortMap = { files: 'files', name: 'name', size: 'lines', words: 'lines' };
+        args.push('-s', sortMap[sort] || 'files');
+    }
+    // Add target directories
+    const dirs = directories.length > 0 ? directories : ['.'];
+    args.push(...dirs);
+    try {
+        const { stdout } = await execFileAsync(sccBinary, args, {
+            maxBuffer: 50 * 1024 * 1024,
+            timeout: 60000,
+        });
+        if (!stdout.trim())
+            return [];
+        return JSON.parse(stdout);
+    }
+    catch (err) {
+        const error = err;
+        if (error.code === 'ENOENT') {
+            throw new Error('scc binary not found.');
+        }
+        // scc exited non-zero or produced no output
+        process.stderr.write(`Warning: scc returned an error: ${error.message}\n`);
+        return [];
+    }
+}
+//# sourceMappingURL=scc.js.map

package/dist/src/scc.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"scc.js","sourceRoot":"","sources":["../../src/scc.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,oBAAoB,CAAC;AAC9C,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AACtC,OAAO,EAAE,UAAU,EAAE,MAAM,SAAS,CAAC;AACrC,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AACzC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,iBAAiB,EAAE,MAAM,YAAY,CAAC;AAE/C,MAAM,aAAa,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;AA+B1C,SAAS,kBAAkB;IACzB,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,aAAa,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;IAC/D,MAAM,MAAM,GAAG,OAAO,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,KAAK,CAAC;IAChE,mDAAmD;IACnD,MAAM,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,EAAE,QAAQ,EAAE,MAAM,CAAC,CAAC;IAChE,MAAM,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,EAAE,IAAI,EAAE,QAAQ,EAAE,MAAM,CAAC,CAAC;IACtE,IAAI,UAAU,CAAC,UAAU,CAAC;QAAE,OAAO,UAAU,CAAC;IAC9C,OAAO,UAAU,CAAC;AACpB,CAAC;AAED,KAAK,UAAU,OAAO;IACpB,yBAAyB;IACzB,MAAM,QAAQ,GAAG,kBAAkB,EAAE,CAAC;IACtC,IAAI,UAAU,CAAC,QAAQ,CAAC;QAAE,OAAO,QAAQ,CAAC;IAE1C,oBAAoB;IACpB,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,OAAO,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC;QAC7D,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,aAAa,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC,CAAC,CAAC;QACrD,OAAO,MAAM,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACtC,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,QAAQ;IAC5B,MAAM,MAAM,GAAG,MAAM,OAAO,EAAE,CAAC;IAC/B,IAAI,CAAC,MAAM,EAAE,CAAC;QACZ,MAAM,IAAI,KAAK,CACb,kCAAkC;YAClC,+FAA+F,CAChG,CAAC;IACJ,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,MAAM,CAAC,SAAwB,EAAE,WAAqB,EAAE,UAAyB,EAAE;IACvG,MAAM,EACJ,MAAM,GAAG,KAAK,EACd,UAAU,GAAG,EAAE,EACf,IAAI,EACJ,EAAE,GAAG,KAAK,EACV,WAAW,GAAG,KAAK,GACpB,GAAG,OAAO,CAAC;IAEZ,IAAI,CAAC,SAAS;QAAE,OAAO,EAAE,CAAC;IAE1B,MAAM,IAAI,GAAG,CAAC,UAAU,EAAE,MAAM,CAAC,CAAC;IAElC,4BAA4B;IAC5B,IAAI,CAAC,IAAI,CAAC,eAAe,EAAE,iBAAiB,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC;IAExD,IAAI,MAAM;QAAE,IAAI,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACnC,IAAI,EAAE;QAAE,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;IAC1B,IAAI,WAAW;QAAE,IAAI,CAAC,IAAI,CAAC,gBAAgB,CAAC,CAAC;IAE7C,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;QAC7B,IAAI,CAAC,IAAI,CAAC,eAAe,EAAE,GAAG,CAAC,CAAC;IAClC,CAAC;IAED,IAAI,IAAI,EAAE,CAAC;QACT,MAAM,OAAO,GAA2B,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,OAAO,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC;QACxG,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,OAAO,CAAC,IAAI,CAAC,IAAI,OAAO,CAAC,CAAC;IAC5C,CAAC;IAED,yBAAyB;IACzB,MAAM,IAAI,GAAG,WAAW,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IAC1D,IAAI,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,CAAC;IAEnB,IAAI,CAAC;QACH,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,aAAa,CAAC,SAAS,EAAE,IAAI,EAAE;YACtD,SAAS,EAAE,EAAE,GAAG,IAAI,GAAG,IAAI;YAC3B,OAAO,EAAE,KAAK;SACf,CAAC,CAAC;QAEH,IAAI,CAAC,MAAM,CAAC,IAAI,EAAE;YAAE,OAAO,EAAE,CAAC;QAE9B,OAAO,IAAI,CAAC,KAAK,CAAC,MAAM,CAAkB,CAAC;IAC7C,CAAC;IAAC,OAAO,GAAY,EAAE,CAAC;QACtB,MAAM,KAAK,GAAG,GAA4B,CAAC;QAC3C,IAAI,KAAK,CAAC,IAAI,KAAK,QAAQ,EAAE,CAAC;YAC5B,MAAM,IAAI,KAAK,CAAC,uBAAuB,CAAC,CAAC;QAC3C,CAAC;QACD,4CAA4C;QAC5C,OAAO,CAAC,MAAM,CAAC,KAAK,CAAC,mCAAmC,KAAK,CAAC,OAAO,IAAI,CAAC,CAAC;QAC3E,OAAO,EAAE,CAAC;IACZ,CAAC;AACH,CAAC"}

package/dist/src/stats.d.ts

@@ -0,0 +1,30 @@
+import type { ParseResult } from './types.js';
+export interface StatsRow {
+    fileType: string;
+    fileName?: string;
+    filePath?: string;
+    files: number;
+    size: number;
+    words: number;
+    pages: number;
+    paragraphs: number;
+    sheets: number;
+    rows: number;
+    cells: number;
+    slides: number;
+    [key: string]: string | number | boolean | undefined;
+}
+export interface ColumnVisibility {
+    [key: string]: boolean;
+}
+export interface AggregateResult {
+    rows: StatsRow[];
+    totals: StatsRow;
+    columns: ColumnVisibility;
+    mode: string;
+}
+export interface AggregateOptions {
+    byFile?: boolean;
+    sort?: string;
+}
+export declare function aggregate(results: ParseResult[], options?: AggregateOptions): AggregateResult;

package/dist/src/stats.js

@@ -0,0 +1,88 @@
+import path from 'node:path';
+import { METRIC_FIELDS, hasKey } from './utils.js';
+const SUM_FIELDS = ['files', ...METRIC_FIELDS, 'size'];
+export function aggregate(results, options = {}) {
+    const { byFile = false, sort = 'files' } = options;
+    if (byFile) {
+        return aggregateByFile(results, sort);
+    }
+    return aggregateByType(results, sort);
+}
+function aggregateByType(results, sort) {
+    const groups = {};
+    for (const r of results) {
+        const key = r.success ? r.fileType : 'Unreadable';
+        if (!groups[key]) {
+            const g = { fileType: key, files: 0, size: 0, words: 0, pages: 0, paragraphs: 0, sheets: 0, rows: 0, cells: 0, slides: 0 };
+            for (const f of METRIC_FIELDS) {
+                g[hasKey(f)] = false;
+            }
+            groups[key] = g;
+        }
+        const g = groups[key];
+        g.files++;
+        g.size += r.size || 0;
+        if (r.success && r.metrics) {
+            const m = r.metrics;
+            for (const f of METRIC_FIELDS) {
+                if (m[f] != null) {
+                    g[f] += m[f];
+                    g[hasKey(f)] = true;
+                }
+            }
+        }
+    }
+    return finalize(Object.values(groups), sort, 'grouped');
+}
+function aggregateByFile(results, sort) {
+    const rows = results.map(r => {
+        const row = {
+            fileType: r.success ? r.fileType : 'Unreadable',
+            fileName: path.basename(r.filePath),
+            filePath: r.filePath,
+            files: 1,
+            size: r.size || 0,
+            words: 0, pages: 0, paragraphs: 0, sheets: 0, rows: 0, cells: 0, slides: 0,
+        };
+        for (const f of METRIC_FIELDS) {
+            row[f] = r.metrics?.[f] || 0;
+            row[hasKey(f)] = r.metrics?.[f] != null;
+        }
+        return row;
+    });
+    return finalize(rows, sort, 'by-file');
+}
+function finalize(rows, sort, mode) {
+    sortRows(rows, sort);
+    const totals = computeTotals(rows);
+    const columns = detectColumns(rows);
+    return { rows, totals, columns, mode };
+}
+function sortRows(rows, sort) {
+    const sortFns = {
+        files: (a, b) => b.files - a.files,
+        name: (a, b) => (a.fileType || a.fileName || '').localeCompare(b.fileType || b.fileName || ''),
+        words: (a, b) => b.words - a.words,
+        size: (a, b) => b.size - a.size,
+    };
+    const fn = sortFns[sort] || sortFns.files;
+    rows.sort(fn);
+}
+function computeTotals(rows) {
+    const totals = { fileType: 'Total', files: 0, size: 0, words: 0, pages: 0, paragraphs: 0, sheets: 0, rows: 0, cells: 0, slides: 0 };
+    for (const f of SUM_FIELDS)
+        totals[f] = 0;
+    for (const r of rows) {
+        for (const f of SUM_FIELDS)
+            totals[f] += r[f];
+    }
+    return totals;
+}
+function detectColumns(rows) {
+    const columns = {};
+    for (const f of METRIC_FIELDS) {
+        columns[hasKey(f)] = rows.some(r => !!r[hasKey(f)]);
+    }
+    return columns;
+}
+//# sourceMappingURL=stats.js.map

package/dist/src/stats.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"stats.js","sourceRoot":"","sources":["../../src/stats.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,aAAa,EAAE,MAAM,EAAE,MAAM,YAAY,CAAC;AAoCnD,MAAM,UAAU,GAAG,CAAC,OAAO,EAAE,GAAG,aAAa,EAAE,MAAM,CAAU,CAAC;AAEhE,MAAM,UAAU,SAAS,CAAC,OAAsB,EAAE,UAA4B,EAAE;IAC9E,MAAM,EAAE,MAAM,GAAG,KAAK,EAAE,IAAI,GAAG,OAAO,EAAE,GAAG,OAAO,CAAC;IAEnD,IAAI,MAAM,EAAE,CAAC;QACX,OAAO,eAAe,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;IACxC,CAAC;IACD,OAAO,eAAe,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;AACxC,CAAC;AAED,SAAS,eAAe,CAAC,OAAsB,EAAE,IAAY;IAC3D,MAAM,MAAM,GAA6B,EAAE,CAAC;IAE5C,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,YAAY,CAAC;QAClD,IAAI,CAAC,MAAM,CAAC,GAAG,CAAC,EAAE,CAAC;YACjB,MAAM,CAAC,GAAa,EAAE,QAAQ,EAAE,GAAG,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,UAAU,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,CAAC;YACrI,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;gBAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC;YAAC,CAAC;YACxD,MAAM,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QAClB,CAAC;QACD,MAAM,CAAC,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC;QACtB,CAAC,CAAC,KAAK,EAAE,CAAC;QACV,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;QAEtB,IAAI,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,OAAO,EAAE,CAAC;YAC3B,MAAM,CAAC,GAAG,CAAC,CAAC,OAAO,CAAC;YACpB,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;gBAC9B,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC;oBAAE,CAAC,CAAC,CAAC,CAAY,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;oBAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC;gBAAC,CAAC;YACtE,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,QAAQ,CAAC,MAAM,CAAC,MAAM,CAAC,MAAM,CAAC,EAAE,IAAI,EAAE,SAAS,CAAC,CAAC;AAC1D,CAAC;AAED,SAAS,eAAe,CAAC,OAAsB,EAAE,IAAY;IAC3D,MAAM,IAAI,GAAe,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE;QACvC,MAAM,GAAG,GAAa;YACpB,QAAQ,EAAE,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,CAAC,YAAY;YAC/C,QAAQ,EAAE,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,QAAQ,CAAC;YACnC,QAAQ,EAAE,CAAC,CAAC,QAAQ;YACpB,KAAK,EAAE,CAAC;YACR,IAAI,EAAE,CAAC,CAAC,IAAI,IAAI,CAAC;YACjB,KAAK,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,UAAU,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC;SAC3E,CAAC;QACF,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;YAC9B,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;YAC7B,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;QAC1C,CAAC;QACD,OAAO,GAAG,CAAC;IACb,CAAC,CAAC,CAAC;IAEH,OAAO,QAAQ,CAAC,IAAI,EAAE,IAAI,EAAE,SAAS,CAAC,CAAC;AACzC,CAAC;AAED,SAAS,QAAQ,CAAC,IAAgB,EAAE,IAAY,EAAE,IAAY;IAC5D,QAAQ,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;IACrB,MAAM,MAAM,GAAG,aAAa,CAAC,IAAI,CAAC,CAAC;IACnC,MAAM,OAAO,GAAG,aAAa,CAAC,IAAI,CAAC,CAAC;IACpC,OAAO,EAAE,IAAI,EAAE,MAAM,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;AACzC,CAAC;AAED,SAAS,QAAQ,CAAC,IAAgB,EAAE,IAAY;IAC9C,MAAM,OAAO,GAAyD;QACpE,KAAK,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK;QAClC,IAAI,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,QAAQ,IAAI,CAAC,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC,aAAa,CAAC,CAAC,CAAC,QAAQ,IAAI,CAAC,CAAC,QAAQ,IAAI,EAAE,CAAC;QAC9F,KAAK,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK;QAClC,IAAI,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI;KAChC,CAAC;IACF,MAAM,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,OAAO,CAAC,KAAK,CAAC;IAC1C,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;AAChB,CAAC;AAED,SAAS,aAAa,CAAC,IAAgB;IACrC,MAAM,MAAM,GAAa,EAAE,QAAQ,EAAE,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,UAAU,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,CAAC;IAC9I,KAAK,MAAM,CAAC,IAAI,UAAU;QAAE,MAAM,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IAC1C,KAAK,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;QACrB,KAAK,MAAM,CAAC,IAAI,UAAU;YAAG,MAAM,CAAC,CAAC,CAAY,IAAK,CAAC,CAAC,CAAC,CAAY,CAAC;IACxE,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,aAAa,CAAC,IAAgB;IACrC,MAAM,OAAO,GAAqB,EAAE,CAAC;IACrC,KAAK,MAAM,CAAC,IAAI,aAAa,EAAE,CAAC;QAC9B,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACtD,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC"}

package/dist/src/structure/extract.d.ts

@@ -0,0 +1,7 @@
+import type { StructureNode, DocumentStructure } from './types.js';
+/** Main extraction pipeline: markdown content → DocumentStructure */
+export declare function extractFromMarkdown(content: string): DocumentStructure;
+/** Find which section a character range falls into */
+export declare function findChunkSection(structure: DocumentStructure, start: number, end: number): StructureNode | null;
+/** Get the content of a section from the original markdown */
+export declare function getSectionContent(content: string, node: StructureNode, includeChildren?: boolean): string;

package/dist/src/structure/extract.js

@@ -0,0 +1,176 @@
+/** Step 1: Extract headers from markdown content */
+function extractHeaders(content) {
+    const lines = content.split('\n');
+    const headers = [];
+    let charPosition = 0;
+    let inCodeBlock = false;
+    for (let i = 0; i < lines.length; i++) {
+        const line = lines[i];
+        if (line.trimStart().startsWith('```')) {
+            inCodeBlock = !inCodeBlock;
+        }
+        if (!inCodeBlock) {
+            const match = line.match(/^(#{1,6})\s+(.+)$/);
+            if (match) {
+                headers.push({
+                    title: match[2].trim(),
+                    level: match[1].length,
+                    lineNum: i + 1,
+                    startChar: charPosition,
+                });
+            }
+        }
+        charPosition += line.length + 1; // +1 for newline
+    }
+    return headers;
+}
+/** Step 2: Compute end positions — each section ends where the next begins */
+function computeEndPositions(headers, contentLength) {
+    return headers.map((h, i) => ({
+        ...h,
+        endChar: i < headers.length - 1 ? headers[i + 1].startChar : contentLength,
+    }));
+}
+/** Step 3: Build tree using a stack-based approach */
+function buildTree(headers) {
+    const rootNodes = [];
+    const stack = [];
+    let nodeCounter = 0;
+    for (const h of headers) {
+        nodeCounter++;
+        const node = {
+            nodeId: String(nodeCounter).padStart(4, '0'),
+            title: h.title,
+            level: h.level,
+            startChar: h.startChar,
+            endChar: h.endChar,
+            startLine: h.lineNum,
+            children: [],
+        };
+        // Pop stack until we find a parent (lower level)
+        while (stack.length > 0 && stack[stack.length - 1].level >= h.level) {
+            stack.pop();
+        }
+        if (stack.length === 0) {
+            rootNodes.push(node);
+        }
+        else {
+            const parent = stack[stack.length - 1].node;
+            node.parentNodeId = parent.nodeId;
+            parent.children.push(node);
+        }
+        stack.push({ node, level: h.level });
+    }
+    return rootNodes;
+}
+/** Step 4: Assign dotted structure codes recursively */
+function assignStructureCodes(nodes, prefix = '') {
+    for (let i = 0; i < nodes.length; i++) {
+        const code = prefix ? `${prefix}.${i + 1}` : String(i + 1);
+        nodes[i].structureCode = code;
+        assignStructureCodes(nodes[i].children, code);
+    }
+}
+/** Step 5: Extract page mappings from [Page N] markers in content */
+function extractPageMappings(content) {
+    const regex = /^\[Page\s+(\d+)\]/gm;
+    const mappings = [];
+    let match;
+    while ((match = regex.exec(content)) !== null) {
+        const pageNumber = parseInt(match[1], 10);
+        mappings.push({
+            pageNumber,
+            startChar: match.index,
+            endChar: content.length, // will be adjusted below
+        });
+    }
+    // Adjust end positions
+    for (let i = 0; i < mappings.length - 1; i++) {
+        mappings[i].endChar = mappings[i + 1].startChar;
+    }
+    return mappings;
+}
+/** Step 6: Map nodes to pages based on char positions */
+function mapNodesToPages(nodes, mappings) {
+    if (mappings.length === 0)
+        return;
+    function findPage(charPos) {
+        for (const m of mappings) {
+            if (charPos >= m.startChar && charPos < m.endChar) {
+                return m.pageNumber;
+            }
+        }
+        return undefined;
+    }
+    function mapRecursive(nodeList) {
+        for (const node of nodeList) {
+            node.startPage = findPage(node.startChar);
+            node.endPage = findPage(Math.max(node.endChar - 1, node.startChar));
+            mapRecursive(node.children);
+        }
+    }
+    mapRecursive(nodes);
+}
+/** Compute max depth of the tree */
+function computeMaxDepth(nodes) {
+    if (nodes.length === 0)
+        return 0;
+    let max = 0;
+    for (const node of nodes) {
+        const childDepth = computeMaxDepth(node.children);
+        max = Math.max(max, 1 + childDepth);
+    }
+    return max;
+}
+/** Count total nodes in tree */
+function countNodes(nodes) {
+    let count = 0;
+    for (const node of nodes) {
+        count += 1 + countNodes(node.children);
+    }
+    return count;
+}
+/** Main extraction pipeline: markdown content → DocumentStructure */
+export function extractFromMarkdown(content) {
+    const rawHeaders = extractHeaders(content);
+    if (rawHeaders.length === 0) {
+        return { rootNodes: [], pageMappings: [], totalNodes: 0, maxDepth: 0 };
+    }
+    const withEnds = computeEndPositions(rawHeaders, content.length);
+    const rootNodes = buildTree(withEnds);
+    assignStructureCodes(rootNodes);
+    const pageMappings = extractPageMappings(content);
+    mapNodesToPages(rootNodes, pageMappings);
+    return {
+        rootNodes,
+        pageMappings,
+        totalNodes: countNodes(rootNodes),
+        maxDepth: computeMaxDepth(rootNodes),
+    };
+}
+/** Find which section a character range falls into */
+export function findChunkSection(structure, start, end) {
+    let best = null;
+    function search(nodes) {
+        for (const node of nodes) {
+            if (start >= node.startChar && start < node.endChar) {
+                best = node;
+                search(node.children);
+            }
+        }
+    }
+    search(structure.rootNodes);
+    return best;
+}
+/** Get the content of a section from the original markdown */
+export function getSectionContent(content, node, includeChildren = true) {
+    if (includeChildren) {
+        return content.slice(node.startChar, node.endChar);
+    }
+    // Exclude children: get content from section start to first child start
+    if (node.children.length === 0) {
+        return content.slice(node.startChar, node.endChar);
+    }
+    return content.slice(node.startChar, node.children[0].startChar);
+}
+//# sourceMappingURL=extract.js.map

package/dist/src/structure/extract.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extract.js","sourceRoot":"","sources":["../../../src/structure/extract.ts"],"names":[],"mappings":"AASA,oDAAoD;AACpD,SAAS,cAAc,CAAC,OAAe;IACrC,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAClC,MAAM,OAAO,GAAgB,EAAE,CAAC;IAChC,IAAI,YAAY,GAAG,CAAC,CAAC;IACrB,IAAI,WAAW,GAAG,KAAK,CAAC;IAExB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QAEtB,IAAI,IAAI,CAAC,SAAS,EAAE,CAAC,UAAU,CAAC,KAAK,CAAC,EAAE,CAAC;YACvC,WAAW,GAAG,CAAC,WAAW,CAAC;QAC7B,CAAC;QAED,IAAI,CAAC,WAAW,EAAE,CAAC;YACjB,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,mBAAmB,CAAC,CAAC;YAC9C,IAAI,KAAK,EAAE,CAAC;gBACV,OAAO,CAAC,IAAI,CAAC;oBACX,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE;oBACtB,KAAK,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC,MAAM;oBACtB,OAAO,EAAE,CAAC,GAAG,CAAC;oBACd,SAAS,EAAE,YAAY;iBACxB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,YAAY,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,iBAAiB;IACpD,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,8EAA8E;AAC9E,SAAS,mBAAmB,CAAC,OAAoB,EAAE,aAAqB;IACtE,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;QAC5B,GAAG,CAAC;QACJ,OAAO,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,aAAa;KAC3E,CAAC,CAAC,CAAC;AACN,CAAC;AAED,sDAAsD;AACtD,SAAS,SAAS,CAAC,OAA+C;IAChE,MAAM,SAAS,GAAoB,EAAE,CAAC;IACtC,MAAM,KAAK,GAAkD,EAAE,CAAC;IAChE,IAAI,WAAW,GAAG,CAAC,CAAC;IAEpB,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,WAAW,EAAE,CAAC;QACd,MAAM,IAAI,GAAkB;YAC1B,MAAM,EAAE,MAAM,CAAC,WAAW,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC;YAC5C,KAAK,EAAE,CAAC,CAAC,KAAK;YACd,KAAK,EAAE,CAAC,CAAC,KAAK;YACd,SAAS,EAAE,CAAC,CAAC,SAAS;YACtB,OAAO,EAAE,CAAC,CAAC,OAAO;YAClB,SAAS,EAAE,CAAC,CAAC,OAAO;YACpB,QAAQ,EAAE,EAAE;SACb,CAAC;QAEF,iDAAiD;QACjD,OAAO,KAAK,CAAC,MAAM,GAAG,CAAC,IAAI,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,KAAK,IAAI,CAAC,CAAC,KAAK,EAAE,CAAC;YACpE,KAAK,CAAC,GAAG,EAAE,CAAC;QACd,CAAC;QAED,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACvB,SAAS,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACvB,CAAC;aAAM,CAAC;YACN,MAAM,MAAM,GAAG,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC;YAC5C,IAAI,CAAC,YAAY,GAAG,MAAM,CAAC,MAAM,CAAC;YAClC,MAAM,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC7B,CAAC;QAED,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC,KAAK,EAAE,CAAC,CAAC;IACvC,CAAC;IAED,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,wDAAwD;AACxD,SAAS,oBAAoB,CAAC,KAAsB,EAAE,MAAM,GAAG,EAAE;IAC/D,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;QACtC,MAAM,IAAI,GAAG,MAAM,CAAC,CAAC,CAAC,GAAG,MAAM,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QAC3D,KAAK,CAAC,CAAC,CAAC,CAAC,aAAa,GAAG,IAAI,CAAC;QAC9B,oBAAoB,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC;IAChD,CAAC;AACH,CAAC;AAED,qEAAqE;AACrE,SAAS,mBAAmB,CAAC,OAAe;IAC1C,MAAM,KAAK,GAAG,qBAAqB,CAAC;IACpC,MAAM,QAAQ,GAAkB,EAAE,CAAC;IACnC,IAAI,KAA6B,CAAC;IAElC,OAAO,CAAC,KAAK,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC9C,MAAM,UAAU,GAAG,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QAC1C,QAAQ,CAAC,IAAI,CAAC;YACZ,UAAU;YACV,SAAS,EAAE,KAAK,CAAC,KAAK;YACtB,OAAO,EAAE,OAAO,CAAC,MAAM,EAAE,yBAAyB;SACnD,CAAC,CAAC;IACL,CAAC;IAED,uBAAuB;IACvB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;QAC7C,QAAQ,CAAC,CAAC,CAAC,CAAC,OAAO,GAAG,QAAQ,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,SAAS,CAAC;IAClD,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,yDAAyD;AACzD,SAAS,eAAe,CAAC,KAAsB,EAAE,QAAuB;IACtE,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO;IAElC,SAAS,QAAQ,CAAC,OAAe;QAC/B,KAAK,MAAM,CAAC,IAAI,QAAQ,EAAE,CAAC;YACzB,IAAI,OAAO,IAAI,CAAC,CAAC,SAAS,IAAI,OAAO,GAAG,CAAC,CAAC,OAAO,EAAE,CAAC;gBAClD,OAAO,CAAC,CAAC,UAAU,CAAC;YACtB,CAAC;QACH,CAAC;QACD,OAAO,SAAS,CAAC;IACnB,CAAC;IAED,SAAS,YAAY,CAAC,QAAyB;QAC7C,KAAK,MAAM,IAAI,IAAI,QAAQ,EAAE,CAAC;YAC5B,IAAI,CAAC,SAAS,GAAG,QAAQ,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;YAC1C,IAAI,CAAC,OAAO,GAAG,QAAQ,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,GAAG,CAAC,EAAE,IAAI,CAAC,SAAS,CAAC,CAAC,CAAC;YACpE,YAAY,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;QAC9B,CAAC;IACH,CAAC;IAED,YAAY,CAAC,KAAK,CAAC,CAAC;AACtB,CAAC;AAED,oCAAoC;AACpC,SAAS,eAAe,CAAC,KAAsB;IAC7C,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IACjC,IAAI,GAAG,GAAG,CAAC,CAAC;IACZ,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,MAAM,UAAU,GAAG,eAAe,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;QAClD,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,GAAG,UAAU,CAAC,CAAC;IACtC,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,gCAAgC;AAChC,SAAS,UAAU,CAAC,KAAsB;IACxC,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,KAAK,IAAI,CAAC,GAAG,UAAU,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IACzC,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,qEAAqE;AACrE,MAAM,UAAU,mBAAmB,CAAC,OAAe;IACjD,MAAM,UAAU,GAAG,cAAc,CAAC,OAAO,CAAC,CAAC;IAE3C,IAAI,UAAU,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAC5B,OAAO,EAAE,SAAS,EAAE,EAAE,EAAE,YAAY,EAAE,EAAE,EAAE,UAAU,EAAE,CAAC,EAAE,QAAQ,EAAE,CAAC,EAAE,CAAC;IACzE,CAAC;IAED,MAAM,QAAQ,GAAG,mBAAmB,CAAC,UAAU,EAAE,OAAO,CAAC,MAAM,CAAC,CAAC;IACjE,MAAM,SAAS,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;IACtC,oBAAoB,CAAC,SAAS,CAAC,CAAC;IAEhC,MAAM,YAAY,GAAG,mBAAmB,CAAC,OAAO,CAAC,CAAC;IAClD,eAAe,CAAC,SAAS,EAAE,YAAY,CAAC,CAAC;IAEzC,OAAO;QACL,SAAS;QACT,YAAY;QACZ,UAAU,EAAE,UAAU,CAAC,SAAS,CAAC;QACjC,QAAQ,EAAE,eAAe,CAAC,SAAS,CAAC;KACrC,CAAC;AACJ,CAAC;AAED,sDAAsD;AACtD,MAAM,UAAU,gBAAgB,CAAC,SAA4B,EAAE,KAAa,EAAE,GAAW;IACvF,IAAI,IAAI,GAAyB,IAAI,CAAC;IAEtC,SAAS,MAAM,CAAC,KAAsB;QACpC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;YACzB,IAAI,KAAK,IAAI,IAAI,CAAC,SAAS,IAAI,KAAK,GAAG,IAAI,CAAC,OAAO,EAAE,CAAC;gBACpD,IAAI,GAAG,IAAI,CAAC;gBACZ,MAAM,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;YACxB,CAAC;QACH,CAAC;IACH,CAAC;IAED,MAAM,CAAC,SAAS,CAAC,SAAS,CAAC,CAAC;IAC5B,OAAO,IAAI,CAAC;AACd,CAAC;AAED,8DAA8D;AAC9D,MAAM,UAAU,iBAAiB,CAAC,OAAe,EAAE,IAAmB,EAAE,eAAe,GAAG,IAAI;IAC5F,IAAI,eAAe,EAAE,CAAC;QACpB,OAAO,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,OAAO,CAAC,CAAC;IACrD,CAAC;IAED,wEAAwE;IACxE,IAAI,IAAI,CAAC,QAAQ,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAC/B,OAAO,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,OAAO,CAAC,CAAC;IACrD,CAAC;IACD,OAAO,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,SAAS,EAAE,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC;AACnE,CAAC"}

package/dist/src/structure/index.d.ts

@@ -0,0 +1,3 @@
+export { extractFromMarkdown, findChunkSection, getSectionContent } from './extract.js';
+export type { StructureNode, DocumentStructure, PageMapping } from './types.js';
+export { flatten, getNodeById, getNodeByPath, toDict, fromDict } from './types.js';

package/dist/src/structure/index.js

@@ -0,0 +1,3 @@
+export { extractFromMarkdown, findChunkSection, getSectionContent } from './extract.js';
+export { flatten, getNodeById, getNodeByPath, toDict, fromDict } from './types.js';
+//# sourceMappingURL=index.js.map

package/dist/src/structure/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../../src/structure/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,mBAAmB,EAAE,gBAAgB,EAAE,iBAAiB,EAAE,MAAM,cAAc,CAAC;AAExF,OAAO,EAAE,OAAO,EAAE,WAAW,EAAE,aAAa,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,YAAY,CAAC"}

package/dist/src/structure/types.d.ts

@@ -0,0 +1,29 @@
+export interface PageMapping {
+    pageNumber: number;
+    startChar: number;
+    endChar: number;
+}
+export interface StructureNode {
+    nodeId: string;
+    title: string;
+    level: number;
+    startChar: number;
+    endChar: number;
+    startLine: number;
+    startPage?: number;
+    endPage?: number;
+    parentNodeId?: string;
+    structureCode?: string;
+    children: StructureNode[];
+}
+export interface DocumentStructure {
+    rootNodes: StructureNode[];
+    pageMappings: PageMapping[];
+    totalNodes: number;
+    maxDepth: number;
+}
+export declare function flatten(nodes: StructureNode[]): StructureNode[];
+export declare function getNodeById(nodes: StructureNode[], nodeId: string): StructureNode | undefined;
+export declare function getNodeByPath(nodes: StructureNode[], structureCode: string): StructureNode | undefined;
+export declare function toDict(structure: DocumentStructure): Record<string, unknown>;
+export declare function fromDict(data: Record<string, unknown>): DocumentStructure;

package/dist/src/structure/types.js

@@ -0,0 +1,72 @@
+export function flatten(nodes) {
+    const result = [];
+    function walk(list) {
+        for (const node of list) {
+            result.push(node);
+            walk(node.children);
+        }
+    }
+    walk(nodes);
+    return result;
+}
+export function getNodeById(nodes, nodeId) {
+    for (const node of flatten(nodes)) {
+        if (node.nodeId === nodeId)
+            return node;
+    }
+    return undefined;
+}
+export function getNodeByPath(nodes, structureCode) {
+    for (const node of flatten(nodes)) {
+        if (node.structureCode === structureCode)
+            return node;
+    }
+    return undefined;
+}
+export function toDict(structure) {
+    function nodeToDict(node) {
+        return {
+            nodeId: node.nodeId,
+            title: node.title,
+            level: node.level,
+            startChar: node.startChar,
+            endChar: node.endChar,
+            startLine: node.startLine,
+            startPage: node.startPage,
+            endPage: node.endPage,
+            parentNodeId: node.parentNodeId,
+            structureCode: node.structureCode,
+            children: node.children.map(nodeToDict),
+        };
+    }
+    return {
+        rootNodes: structure.rootNodes.map(nodeToDict),
+        pageMappings: structure.pageMappings,
+        totalNodes: structure.totalNodes,
+        maxDepth: structure.maxDepth,
+    };
+}
+export function fromDict(data) {
+    function nodeFromDict(d) {
+        return {
+            nodeId: d.nodeId,
+            title: d.title,
+            level: d.level,
+            startChar: d.startChar,
+            endChar: d.endChar,
+            startLine: d.startLine,
+            startPage: d.startPage,
+            endPage: d.endPage,
+            parentNodeId: d.parentNodeId,
+            structureCode: d.structureCode,
+            children: d.children.map(nodeFromDict),
+        };
+    }
+    return {
+        rootNodes: data.rootNodes.map(nodeFromDict),
+        pageMappings: data.pageMappings,
+        totalNodes: data.totalNodes,
+        maxDepth: data.maxDepth,
+    };
+}
+//# sourceMappingURL=types.js.map

package/dist/src/structure/types.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.js","sourceRoot":"","sources":["../../../src/structure/types.ts"],"names":[],"mappings":"AA2BA,MAAM,UAAU,OAAO,CAAC,KAAsB;IAC5C,MAAM,MAAM,GAAoB,EAAE,CAAC;IACnC,SAAS,IAAI,CAAC,IAAqB;QACjC,KAAK,MAAM,IAAI,IAAI,IAAI,EAAE,CAAC;YACxB,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAClB,IAAI,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;QACtB,CAAC;IACH,CAAC;IACD,IAAI,CAAC,KAAK,CAAC,CAAC;IACZ,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,KAAsB,EAAE,MAAc;IAChE,KAAK,MAAM,IAAI,IAAI,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;QAClC,IAAI,IAAI,CAAC,MAAM,KAAK,MAAM;YAAE,OAAO,IAAI,CAAC;IAC1C,CAAC;IACD,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,UAAU,aAAa,CAAC,KAAsB,EAAE,aAAqB;IACzE,KAAK,MAAM,IAAI,IAAI,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;QAClC,IAAI,IAAI,CAAC,aAAa,KAAK,aAAa;YAAE,OAAO,IAAI,CAAC;IACxD,CAAC;IACD,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,UAAU,MAAM,CAAC,SAA4B;IACjD,SAAS,UAAU,CAAC,IAAmB;QACrC,OAAO;YACL,MAAM,EAAE,IAAI,CAAC,MAAM;YACnB,KAAK,EAAE,IAAI,CAAC,KAAK;YACjB,KAAK,EAAE,IAAI,CAAC,KAAK;YACjB,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,SAAS,EAAE,IAAI,CAAC,SAAS;YACzB,OAAO,EAAE,IAAI,CAAC,OAAO;YACrB,YAAY,EAAE,IAAI,CAAC,YAAY;YAC/B,aAAa,EAAE,IAAI,CAAC,aAAa;YACjC,QAAQ,EAAE,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,UAAU,CAAC;SACxC,CAAC;IACJ,CAAC;IACD,OAAO;QACL,SAAS,EAAE,SAAS,CAAC,SAAS,CAAC,GAAG,CAAC,UAAU,CAAC;QAC9C,YAAY,EAAE,SAAS,CAAC,YAAY;QACpC,UAAU,EAAE,SAAS,CAAC,UAAU;QAChC,QAAQ,EAAE,SAAS,CAAC,QAAQ;KAC7B,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,QAAQ,CAAC,IAA6B;IACpD,SAAS,YAAY,CAAC,CAA0B;QAC9C,OAAO;YACL,MAAM,EAAE,CAAC,CAAC,MAAgB;YAC1B,KAAK,EAAE,CAAC,CAAC,KAAe;YACxB,KAAK,EAAE,CAAC,CAAC,KAAe;YACxB,SAAS,EAAE,CAAC,CAAC,SAAmB;YAChC,OAAO,EAAE,CAAC,CAAC,OAAiB;YAC5B,SAAS,EAAE,CAAC,CAAC,SAAmB;YAChC,SAAS,EAAE,CAAC,CAAC,SAA+B;YAC5C,OAAO,EAAE,CAAC,CAAC,OAA6B;YACxC,YAAY,EAAE,CAAC,CAAC,YAAkC;YAClD,aAAa,EAAE,CAAC,CAAC,aAAmC;YACpD,QAAQ,EAAG,CAAC,CAAC,QAAsC,CAAC,GAAG,CAAC,YAAY,CAAC;SACtE,CAAC;IACJ,CAAC;IACD,OAAO;QACL,SAAS,EAAG,IAAI,CAAC,SAAuC,CAAC,GAAG,CAAC,YAAY,CAAC;QAC1E,YAAY,EAAE,IAAI,CAAC,YAA6B;QAChD,UAAU,EAAE,IAAI,CAAC,UAAoB;QACrC,QAAQ,EAAE,IAAI,CAAC,QAAkB;KAClC,CAAC;AACJ,CAAC"}

package/dist/src/types.d.ts

@@ -0,0 +1,20 @@
+export interface FileEntry {
+    path: string;
+    size: number;
+}
+export interface SkippedEntry {
+    path: string;
+    reason: string;
+    size: number;
+}
+export interface ParserOutput {
+    fileType: string;
+    metrics: Record<string, number>;
+}
+export interface ParseResult {
+    filePath: string;
+    size: number;
+    success: boolean;
+    fileType: string;
+    metrics: Record<string, number> | null;
+}

package/dist/src/types.js

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=types.js.map

package/dist/src/types.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.js","sourceRoot":"","sources":["../../src/types.ts"],"names":[],"mappings":""}