@unisim/doc 0.1.0

package/README.md

@@ -0,0 +1,91 @@
+# @unisim/doc
+
+The suite's document stack. Reads Word, OpenDocument, RTF, HTML, Markdown, CSV,
+JSON and plain text into one model, and writes that model back out as PDF, text,
+HTML, Markdown, CSV or JSON.
+
+**No runtime dependencies.** That is not incidental — one of its two consumers
+(Universal Converter) is sold on being offline from first visit with nothing
+fetched, so a dependency here would cost that product its pitch.
+
+```ts
+import { readToDoc, convertDocument, DEFAULT_DOC_SETTINGS } from '@unisim/doc'
+
+// The model, for a caller that wants to inspect or combine documents
+const doc = await readToDoc(file)
+
+// Or straight to an output
+const { blob, ext, notices, pages } = await convertDocument(file, {
+  ...DEFAULT_DOC_SETTINGS,
+  format: 'pdf',
+})
+```
+
+`convertDocument` hands back the target **extension**, not a filename. Naming
+the output file is the caller's job on purpose: the app knows what the source
+was called and which of its own naming rules apply, and a library offering a
+second opinion is how you end up chasing a `.png.png`.
+
+## Why it exists
+
+It existed twice. On 2026-08-13 two sessions that could not see each other
+shipped the same 1,200–1,500 lines on the same afternoon — a ZIP reader, a
+namespace-agnostic XML layer, an OOXML walker, an ODF walker, and a
+tokenise → wrap → paginate layout engine — into Universal PDF and Universal
+Converter. Both files ended up carrying a header saying *"a fix to one is a fix
+to neither."*
+
+This is Converter's version, because it was the better one: a model designed as
+a hub (nine readers × six writers) rather than grown out of a Markdown renderer,
+with heading level 4, real image blocks, `underline`/`strike` on a run, and
+`DocNotice` — a channel for reporting what a conversion **lost**, which the
+other had no equivalent of. Universal PDF's model had exactly one thing this
+lacked, `pagesetup`, which was merged in.
+
+## What it does not do
+
+It re-typesets a document; it does not reproduce the page. Fonts, columns,
+headers and footers, text boxes and floating shapes do not survive. Anything a
+reader has to drop is reported through `RichDoc.notices` rather than discovered
+later — the failure worth designing against is someone assuming the output is a
+copy.
+
+## Two writers, on purpose
+
+The PDF writer here goes through `pdfcore` — base-14 fonts, its own AFM tables,
+no dependency. A consumer that already carries pdf-lib for other reasons
+(Universal PDF does, for signing, redaction and forms) gains nothing by dropping
+it, and can take the readers and the model from here while keeping its own
+writer. The seam is that `write/pdf.ts` depends only on `pdfcore`'s small
+surface — `addPage` / `drawText` / `drawRect` / `drawImage` / `widthOfText`.
+
+## Notes for anyone editing it
+
+- **Imports carry a `.js` extension.** Not `.ts` (which needs
+  `allowImportingTsExtensions`, only legal with `noEmit`) and not extensionless
+  (emitted verbatim, so `dist/index.js` would import `./convert` and Node would
+  refuse it). The first build of this package made exactly that mistake.
+- **The model is the intersection of what these formats can express, not the
+  union.** A DOCX can hold a text box rotated 3° over a chart; Markdown cannot.
+  Anything outside the vocabulary is flattened on the way *in*, by the reader
+  that knows what it meant.
+- **`pagesetup` is an instruction, not a property**, because that is how the
+  source formats model it — orientation belongs to a section, and a section runs
+  until the next break. ⚠️ The two office formats express it in **opposite
+  directions**: OOXML's `w:sectPr` sits on the **last** paragraph of the section
+  it describes, ODF's `style:master-page-name` on the **first** paragraph of the
+  new one. Reading them the same way puts every change one page out.
+
+## Tests
+
+```
+npm test
+```
+
+Covers what is DOM-free: the hand-written PDF byte layer (including that the
+xref offsets actually point at their objects — a reader that repairs a damaged
+table will open a broken file, so "it opened in Chrome" proves nothing), pages
+of differing sizes in one document, and the `setSize` guard rail. The rest is
+DOM-bound and is covered by the consuming apps' browser suites — in particular
+Universal PDF's `e2e/office-import.e2e.mjs`, which drives real `.docx`/`.odt`
+fixtures and asserts on the page sizes pdf.js reads back out.

package/dist/convert.d.ts

@@ -0,0 +1,54 @@
+import type { DocNotice, RichDoc } from './model.js';
+import { type PdfSettings } from './write/pdfSettings.js';
+export type { PdfSettings, PaperSize, PageMargin, FontChoice } from './write/pdfSettings.js';
+export { DEFAULT_PDF_SETTINGS } from './write/pdfSettings.js';
+export type { DocNotice, RichDoc } from './model.js';
+export type DocFormat = 'pdf' | 'txt' | 'html' | 'md' | 'csv' | 'json';
+export interface DocSettings {
+    format: DocFormat;
+    pdf: PdfSettings;
+    /** CSV → JSON: read "42" as a number rather than a string. */
+    inferTypes: boolean;
+}
+export declare const DEFAULT_DOC_SETTINGS: DocSettings;
+/** What came out, plus anything the person should be told about it. */
+/**
+ * What a conversion produced.
+ *
+ * ⚠️ Carries the target EXTENSION, not a filename. Naming the output file is
+ * the caller's job and deliberately not this package's: the app knows what the
+ * source was called and which of its own naming rules apply, and a library that
+ * guessed would be a second opinion on a question that has already caused a
+ * `.png.png` hunt once. Callers pass this to their own `withExtension`.
+ */
+export interface DocResult {
+    blob: Blob;
+    /** Target extension without the dot — 'pdf', 'txt', 'html', 'md', 'csv', 'json'. */
+    ext: DocFormat;
+    notices: DocNotice[];
+    /** Page count, for the formats that have pages. */
+    pages?: number;
+}
+/**
+ * Input extensions, grouped by the shape of what they produce.
+ *
+ * `tabular` sources also yield a `Grid`, which is the only thing that can
+ * become CSV or JSON. Everything else is a document, and asking for CSV from a
+ * Word file is refused by the UI rather than answered with a one-column sheet
+ * of its paragraphs.
+ */
+export declare const DOC_INPUT_EXTS: readonly ["docx", "doc", "odt", "rtf", "txt", "text", "log", "md", "markdown", "html", "htm", "csv", "tsv", "json"];
+export declare function isTabular(ext: string): boolean;
+/** The targets a given input can actually reach. */
+export declare function targetsFor(ext: string): DocFormat[];
+/** Targets reachable by every file in a mixed queue. */
+export declare function commonTargets(exts: readonly string[]): DocFormat[];
+/**
+ * Read a file to a RichDoc, whichever kind it is.
+ *
+ * The one entry point the "join into one PDF" export needs — it wants the model
+ * and not a converted file, so that it can concatenate before writing once.
+ */
+export declare function readToDoc(file: File): Promise<RichDoc>;
+export declare function convertDocument(file: File, settings: DocSettings, onProgress?: (fraction: number) => void): Promise<DocResult>;
+//# sourceMappingURL=convert.d.ts.map

package/dist/convert.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"convert.d.ts","sourceRoot":"","sources":["../src/convert.ts"],"names":[],"mappings":"AA4BA,OAAO,KAAK,EAAE,SAAS,EAAQ,OAAO,EAAE,MAAM,YAAY,CAAA;AAI1D,OAAO,EAAwB,KAAK,WAAW,EAAE,MAAM,wBAAwB,CAAA;AAE/E,YAAY,EAAE,WAAW,EAAE,SAAS,EAAE,UAAU,EAAE,UAAU,EAAE,MAAM,wBAAwB,CAAA;AAC5F,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAA;AAC7D,YAAY,EAAE,SAAS,EAAE,OAAO,EAAE,MAAM,YAAY,CAAA;AAEpD,MAAM,MAAM,SAAS,GAAG,KAAK,GAAG,KAAK,GAAG,MAAM,GAAG,IAAI,GAAG,KAAK,GAAG,MAAM,CAAA;AAEtE,MAAM,WAAW,WAAW;IAC1B,MAAM,EAAE,SAAS,CAAA;IACjB,GAAG,EAAE,WAAW,CAAA;IAChB,8DAA8D;IAC9D,UAAU,EAAE,OAAO,CAAA;CACpB;AAED,eAAO,MAAM,oBAAoB,EAAE,WAIlC,CAAA;AAED,uEAAuE;AACvE;;;;;;;;GAQG;AACH,MAAM,WAAW,SAAS;IACxB,IAAI,EAAE,IAAI,CAAA;IACV,oFAAoF;IACpF,GAAG,EAAE,SAAS,CAAA;IACd,OAAO,EAAE,SAAS,EAAE,CAAA;IACpB,mDAAmD;IACnD,KAAK,CAAC,EAAE,MAAM,CAAA;CACf;AAID;;;;;;;GAOG;AACH,eAAO,MAAM,cAAc,qHAGjB,CAAA;AAIV,wBAAgB,SAAS,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAE9C;AAED,oDAAoD;AACpD,wBAAgB,UAAU,CAAC,GAAG,EAAE,MAAM,GAAG,SAAS,EAAE,CAKnD;AAED,wDAAwD;AACxD,wBAAgB,aAAa,CAAC,IAAI,EAAE,SAAS,MAAM,EAAE,GAAG,SAAS,EAAE,CAKlE;AAOD;;;;;GAKG;AACH,wBAAsB,SAAS,CAAC,IAAI,EAAE,IAAI,GAAG,OAAO,CAAC,OAAO,CAAC,CAE5D;AAsDD,wBAAsB,eAAe,CACnC,IAAI,EAAE,IAAI,EACV,QAAQ,EAAE,WAAW,EACrB,UAAU,CAAC,EAAE,CAAC,QAAQ,EAAE,MAAM,KAAK,IAAI,GACtC,OAAO,CAAC,SAAS,CAAC,CAqEpB"}

package/dist/convert.js

@@ -0,0 +1,216 @@
+// ---------------------------------------------------------------------------
+// The Files pipeline: read anything → RichDoc (or a Grid) → write anything.
+//
+// This file is the only place that knows which inputs pair with which outputs.
+// Everything either side of it is ignorant of the other, which is what keeps a
+// tenth input format to one new file.
+//
+// EVERY READER IS LAZILY IMPORTED. Together they are a fair amount of code —
+// a CFB reader, an RTF tokeniser, two XML dialects — and somebody who opened
+// the app to convert a JPEG should not download any of it. Dynamic `import()`
+// puts each in its own chunk, fetched the first time a file of that type is
+// converted, and never on the audio, images or video tabs. That is the same
+// bargain LAME strikes, except these are already in the bundle rather than
+// coming from npm, so it works offline: the service worker precaches them.
+//
+// ⚠️ THIS STACK EXISTS TWICE IN THE SUITE. Universal PDF shipped its own office
+// import — `unzip.ts`, `officeXml.ts`, `docxToBlocks.ts`, `odtToBlocks.ts`,
+// `blockPdf.ts` — at 16:24 on 2026-08-13 (commit c1f3faa), the same afternoon
+// this landed, from a parallel session neither could see. The ZIP reader, the
+// XML helpers, the OOXML/ODF walkers and the layout engine all overlap.
+//
+// They are not identical: this side reads .doc, .rtf, .csv, .json, .txt, .md and
+// .html, embeds images, and writes with NO dependency; that side refuses .doc and
+// .rtf and renders through pdf-lib. Consolidating them into @unisim/sdk is an
+// open backlog item ("Universal Converter" in backlog-unisim.md). Until it is
+// done, A FIX HERE IS A FIX TO NEITHER — check the twin.
+// ---------------------------------------------------------------------------
+// ⚠️ From `pdfSettings`, NOT from `write/pdf`. A static import of the renderer
+// here would cancel the dynamic one below and pull the whole layout engine and
+// its font tables into the main bundle — see that file's header.
+import { DEFAULT_PDF_SETTINGS } from './write/pdfSettings.js';
+export { DEFAULT_PDF_SETTINGS } from './write/pdfSettings.js';
+export const DEFAULT_DOC_SETTINGS = {
+    format: 'pdf',
+    pdf: DEFAULT_PDF_SETTINGS,
+    inferTypes: true,
+};
+// ── What reads what ──────────────────────────────────────────────────────────
+/**
+ * Input extensions, grouped by the shape of what they produce.
+ *
+ * `tabular` sources also yield a `Grid`, which is the only thing that can
+ * become CSV or JSON. Everything else is a document, and asking for CSV from a
+ * Word file is refused by the UI rather than answered with a one-column sheet
+ * of its paragraphs.
+ */
+export const DOC_INPUT_EXTS = [
+    'docx', 'doc', 'odt', 'rtf', 'txt', 'text', 'log', 'md', 'markdown',
+    'html', 'htm', 'csv', 'tsv', 'json',
+];
+const TABULAR_EXTS = new Set(['csv', 'tsv', 'json']);
+export function isTabular(ext) {
+    return TABULAR_EXTS.has(ext);
+}
+/** The targets a given input can actually reach. */
+export function targetsFor(ext) {
+    if (ext === 'csv' || ext === 'tsv')
+        return ['pdf', 'json', 'html', 'md', 'txt'];
+    if (ext === 'json')
+        return ['pdf', 'csv', 'html', 'md', 'txt'];
+    // A document has no grid behind it, so CSV and JSON are not on offer.
+    return ['pdf', 'txt', 'html', 'md'];
+}
+/** Targets reachable by every file in a mixed queue. */
+export function commonTargets(exts) {
+    if (!exts.length)
+        return ['pdf', 'txt', 'html', 'md'];
+    return exts
+        .map(targetsFor)
+        .reduce((shared, next) => shared.filter((f) => next.includes(f)));
+}
+/**
+ * Read a file to a RichDoc, whichever kind it is.
+ *
+ * The one entry point the "join into one PDF" export needs — it wants the model
+ * and not a converted file, so that it can concatenate before writing once.
+ */
+export async function readToDoc(file) {
+    return (await read(file, extensionOf(file.name))).doc;
+}
+async function read(file, ext) {
+    switch (ext) {
+        case 'docx': {
+            const { readDocx } = await import('./read/docx.js');
+            return { doc: await readDocx(file), grid: null };
+        }
+        case 'doc': {
+            const { readDoc } = await import('./read/doc.js');
+            return { doc: await readDoc(file), grid: null };
+        }
+        case 'odt': {
+            const { readOdt } = await import('./read/odt.js');
+            return { doc: await readOdt(file), grid: null };
+        }
+        case 'rtf': {
+            const { readRtf } = await import('./read/rtf.js');
+            return { doc: await readRtf(file), grid: null };
+        }
+        case 'md':
+        case 'markdown': {
+            const { readMarkdown } = await import('./read/markdown.js');
+            return { doc: await readMarkdown(file), grid: null };
+        }
+        case 'html':
+        case 'htm': {
+            const { readHtml } = await import('./read/html.js');
+            return { doc: await readHtml(file), grid: null };
+        }
+        case 'csv':
+        case 'tsv': {
+            const { readCsv } = await import('./read/csv.js');
+            const { doc, grid } = await readCsv(file);
+            return { doc, grid };
+        }
+        case 'json': {
+            const { readJson } = await import('./read/json.js');
+            const { doc, grid } = await readJson(file);
+            return { doc, grid };
+        }
+        case 'txt':
+        case 'text':
+        case 'log': {
+            const { readText } = await import('./read/text.js');
+            return { doc: await readText(file), grid: null };
+        }
+        default:
+            throw new Error(`${ext ? ext.toUpperCase() : 'That file type'} isn’t a document this can read.`);
+    }
+}
+// ── The conversion ───────────────────────────────────────────────────────────
+export async function convertDocument(file, settings, onProgress) {
+    const ext = extensionOf(file.name);
+    onProgress?.(0.05);
+    const { doc, grid } = await read(file, ext);
+    // Reading is most of the work for the formats that need unzipping and
+    // decoding; writing is fast. The split reflects that rather than pretending
+    // to know a percentage it cannot measure.
+    onProgress?.(0.65);
+    const notices = [...doc.notices];
+    switch (settings.format) {
+        case 'pdf': {
+            const { docToPdf } = await import('./write/pdf.js');
+            const result = await docToPdf(doc, settings.pdf);
+            if (result.dropped.length) {
+                // Named, not counted. "12 characters could not be written" sends
+                // somebody hunting; showing them the actual glyphs makes it obvious
+                // at a glance that it was the Greek in one footnote.
+                const sample = result.dropped.slice(0, 12).join(' ');
+                notices.push({
+                    severity: 'loss',
+                    message: `Some characters aren’t in the PDF’s built-in fonts and came out as “?”: ${sample}` +
+                        (result.dropped.length > 12 ? ` and ${result.dropped.length - 12} more.` : '.') +
+                        ' Those alphabets need a font embedded in the file, which this converter doesn’t do.',
+                });
+            }
+            onProgress?.(1);
+            return { blob: result.blob, ext: 'pdf', notices, pages: result.pages };
+        }
+        case 'txt': {
+            const { docToText } = await import('./write/text.js');
+            onProgress?.(1);
+            return { blob: textBlob(docToText(doc), 'text/plain'), ext: 'txt', notices };
+        }
+        case 'html': {
+            const { docToHtml } = await import('./write/text.js');
+            onProgress?.(1);
+            return { blob: textBlob(docToHtml(doc), 'text/html'), ext: 'html', notices };
+        }
+        case 'md': {
+            const { docToMarkdown } = await import('./write/text.js');
+            onProgress?.(1);
+            return { blob: textBlob(docToMarkdown(doc), 'text/markdown'), ext: 'md', notices };
+        }
+        case 'csv': {
+            if (!grid)
+                throw new Error('Only a spreadsheet-shaped file can become a CSV.');
+            const { gridToCsv } = await import('./write/data.js');
+            onProgress?.(1);
+            return { blob: csvBlob(gridToCsv(grid)), ext: 'csv', notices };
+        }
+        case 'json': {
+            if (!grid)
+                throw new Error('Only a spreadsheet-shaped file can become JSON.');
+            const { gridToJson } = await import('./write/data.js');
+            onProgress?.(1);
+            return {
+                blob: textBlob(gridToJson(grid, { inferTypes: settings.inferTypes }), 'application/json'),
+                ext: 'json',
+                notices,
+            };
+        }
+    }
+}
+function textBlob(text, type) {
+    return new Blob([text], { type: `${type};charset=utf-8` });
+}
+/**
+ * A CSV, and the ONE place a byte-order mark is written.
+ *
+ * ⚠️ Excel opens a UTF-8 CSV as Windows-1252 unless the file starts with a BOM,
+ * so "Ré" arrives as "Ré" for anybody who double-clicks the result — and
+ * double-clicking it is what a CSV is for. The BOM is NOT put on the other text
+ * targets, and specifically not on JSON: RFC 8259 forbids one, and `JSON.parse`
+ * throws on a string that has it, so the tidy-looking "put it on everything"
+ * rule would hand back JSON that will not load.
+ */
+function csvBlob(text) {
+    return new Blob(['', text], { type: 'text/csv;charset=utf-8' });
+}
+function extensionOf(filename) {
+    const dot = filename.lastIndexOf('.');
+    if (dot < 1 || dot === filename.length - 1)
+        return '';
+    return filename.slice(dot + 1).toLowerCase();
+}
+//# sourceMappingURL=convert.js.map

package/dist/convert.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"convert.js","sourceRoot":"","sources":["../src/convert.ts"],"names":[],"mappings":"AAAA,8EAA8E;AAC9E,4EAA4E;AAC5E,EAAE;AACF,+EAA+E;AAC/E,+EAA+E;AAC/E,sCAAsC;AACtC,EAAE;AACF,6EAA6E;AAC7E,6EAA6E;AAC7E,8EAA8E;AAC9E,4EAA4E;AAC5E,4EAA4E;AAC5E,2EAA2E;AAC3E,2EAA2E;AAC3E,EAAE;AACF,gFAAgF;AAChF,4EAA4E;AAC5E,8EAA8E;AAC9E,8EAA8E;AAC9E,wEAAwE;AACxE,EAAE;AACF,iFAAiF;AACjF,kFAAkF;AAClF,8EAA8E;AAC9E,8EAA8E;AAC9E,yDAAyD;AACzD,8EAA8E;AAG9E,+EAA+E;AAC/E,+EAA+E;AAC/E,iEAAiE;AACjE,OAAO,EAAE,oBAAoB,EAAoB,MAAM,wBAAwB,CAAA;AAG/E,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAA;AAY7D,MAAM,CAAC,MAAM,oBAAoB,GAAgB;IAC/C,MAAM,EAAE,KAAK;IACb,GAAG,EAAE,oBAAoB;IACzB,UAAU,EAAE,IAAI;CACjB,CAAA;AAqBD,gFAAgF;AAEhF;;;;;;;GAOG;AACH,MAAM,CAAC,MAAM,cAAc,GAAG;IAC5B,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,UAAU;IACnE,MAAM,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM;CAC3B,CAAA;AAEV,MAAM,YAAY,GAAG,IAAI,GAAG,CAAC,CAAC,KAAK,EAAE,KAAK,EAAE,MAAM,CAAC,CAAC,CAAA;AAEpD,MAAM,UAAU,SAAS,CAAC,GAAW;IACnC,OAAO,YAAY,CAAC,GAAG,CAAC,GAAG,CAAC,CAAA;AAC9B,CAAC;AAED,oDAAoD;AACpD,MAAM,UAAU,UAAU,CAAC,GAAW;IACpC,IAAI,GAAG,KAAK,KAAK,IAAI,GAAG,KAAK,KAAK;QAAE,OAAO,CAAC,KAAK,EAAE,MAAM,EAAE,MAAM,EAAE,IAAI,EAAE,KAAK,CAAC,CAAA;IAC/E,IAAI,GAAG,KAAK,MAAM;QAAE,OAAO,CAAC,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,KAAK,CAAC,CAAA;IAC9D,sEAAsE;IACtE,OAAO,CAAC,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,CAAC,CAAA;AACrC,CAAC;AAED,wDAAwD;AACxD,MAAM,UAAU,aAAa,CAAC,IAAuB;IACnD,IAAI,CAAC,IAAI,CAAC,MAAM;QAAE,OAAO,CAAC,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,CAAC,CAAA;IACrD,OAAO,IAAI;SACR,GAAG,CAAC,UAAU,CAAC;SACf,MAAM,CAAC,CAAC,MAAM,EAAE,IAAI,EAAE,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAA;AACrE,CAAC;AAOD;;;;;GAKG;AACH,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,IAAU;IACxC,OAAO,CAAC,MAAM,IAAI,CAAC,IAAI,EAAE,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,GAAG,CAAA;AACvD,CAAC;AAED,KAAK,UAAU,IAAI,CAAC,IAAU,EAAE,GAAW;IACzC,QAAQ,GAAG,EAAE,CAAC;QACZ,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,MAAM,CAAC,gBAAgB,CAAC,CAAA;YACnD,OAAO,EAAE,GAAG,EAAE,MAAM,QAAQ,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QAClD,CAAC;QACD,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,eAAe,CAAC,CAAA;YACjD,OAAO,EAAE,GAAG,EAAE,MAAM,OAAO,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QACjD,CAAC;QACD,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,eAAe,CAAC,CAAA;YACjD,OAAO,EAAE,GAAG,EAAE,MAAM,OAAO,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QACjD,CAAC;QACD,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,eAAe,CAAC,CAAA;YACjD,OAAO,EAAE,GAAG,EAAE,MAAM,OAAO,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QACjD,CAAC;QACD,KAAK,IAAI,CAAC;QACV,KAAK,UAAU,CAAC,CAAC,CAAC;YAChB,MAAM,EAAE,YAAY,EAAE,GAAG,MAAM,MAAM,CAAC,oBAAoB,CAAC,CAAA;YAC3D,OAAO,EAAE,GAAG,EAAE,MAAM,YAAY,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QACtD,CAAC;QACD,KAAK,MAAM,CAAC;QACZ,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,MAAM,CAAC,gBAAgB,CAAC,CAAA;YACnD,OAAO,EAAE,GAAG,EAAE,MAAM,QAAQ,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QAClD,CAAC;QACD,KAAK,KAAK,CAAC;QACX,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,OAAO,EAAE,GAAG,MAAM,MAAM,CAAC,eAAe,CAAC,CAAA;YACjD,MAAM,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,MAAM,OAAO,CAAC,IAAI,CAAC,CAAA;YACzC,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,CAAA;QACtB,CAAC;QACD,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,MAAM,CAAC,gBAAgB,CAAC,CAAA;YACnD,MAAM,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,MAAM,QAAQ,CAAC,IAAI,CAAC,CAAA;YAC1C,OAAO,EAAE,GAAG,EAAE,IAAI,EAAE,CAAA;QACtB,CAAC;QACD,KAAK,KAAK,CAAC;QACX,KAAK,MAAM,CAAC;QACZ,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,MAAM,CAAC,gBAAgB,CAAC,CAAA;YACnD,OAAO,EAAE,GAAG,EAAE,MAAM,QAAQ,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAA;QAClD,CAAC;QACD;YACE,MAAM,IAAI,KAAK,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC,CAAC,CAAC,gBAAgB,kCAAkC,CAAC,CAAA;IACpG,CAAC;AACH,CAAC;AAED,gFAAgF;AAEhF,MAAM,CAAC,KAAK,UAAU,eAAe,CACnC,IAAU,EACV,QAAqB,EACrB,UAAuC;IAEvC,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;IAClC,UAAU,EAAE,CAAC,IAAI,CAAC,CAAA;IAElB,MAAM,EAAE,GAAG,EAAE,IAAI,EAAE,GAAG,MAAM,IAAI,CAAC,IAAI,EAAE,GAAG,CAAC,CAAA;IAC3C,sEAAsE;IACtE,4EAA4E;IAC5E,0CAA0C;IAC1C,UAAU,EAAE,CAAC,IAAI,CAAC,CAAA;IAElB,MAAM,OAAO,GAAG,CAAC,GAAG,GAAG,CAAC,OAAO,CAAC,CAAA;IAEhC,QAAQ,QAAQ,CAAC,MAAM,EAAE,CAAC;QACxB,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,QAAQ,EAAE,GAAG,MAAM,MAAM,CAAC,gBAAgB,CAAC,CAAA;YACnD,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,GAAG,EAAE,QAAQ,CAAC,GAAG,CAAC,CAAA;YAChD,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC;gBAC1B,iEAAiE;gBACjE,oEAAoE;gBACpE,qDAAqD;gBACrD,MAAM,MAAM,GAAG,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;gBACpD,OAAO,CAAC,IAAI,CAAC;oBACX,QAAQ,EAAE,MAAM;oBAChB,OAAO,EACL,2EAA2E,MAAM,EAAE;wBACnF,CAAC,MAAM,CAAC,OAAO,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC,CAAC,QAAQ,MAAM,CAAC,OAAO,CAAC,MAAM,GAAG,EAAE,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC;wBAC/E,qFAAqF;iBACxF,CAAC,CAAA;YACJ,CAAC;YACD,UAAU,EAAE,CAAC,CAAC,CAAC,CAAA;YACf,OAAO,EAAE,IAAI,EAAE,MAAM,CAAC,IAAI,EAAE,GAAG,EAAE,KAAK,EAAE,OAAO,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,EAAE,CAAA;QACxE,CAAC;QAED,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,iBAAiB,CAAC,CAAA;YACrD,UAAU,EAAE,CAAC,CAAC,CAAC,CAAA;YACf,OAAO,EAAE,IAAI,EAAE,QAAQ,CAAC,SAAS,CAAC,GAAG,CAAC,EAAE,YAAY,CAAC,EAAE,GAAG,EAAE,KAAK,EAAE,OAAO,EAAE,CAAA;QAC9E,CAAC;QAED,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,iBAAiB,CAAC,CAAA;YACrD,UAAU,EAAE,CAAC,CAAC,CAAC,CAAA;YACf,OAAO,EAAE,IAAI,EAAE,QAAQ,CAAC,SAAS,CAAC,GAAG,CAAC,EAAE,WAAW,CAAC,EAAE,GAAG,EAAE,MAAM,EAAE,OAAO,EAAE,CAAA;QAC9E,CAAC;QAED,KAAK,IAAI,CAAC,CAAC,CAAC;YACV,MAAM,EAAE,aAAa,EAAE,GAAG,MAAM,MAAM,CAAC,iBAAiB,CAAC,CAAA;YACzD,UAAU,EAAE,CAAC,CAAC,CAAC,CAAA;YACf,OAAO,EAAE,IAAI,EAAE,QAAQ,CAAC,aAAa,CAAC,GAAG,CAAC,EAAE,eAAe,CAAC,EAAE,GAAG,EAAE,IAAI,EAAE,OAAO,EAAE,CAAA;QACpF,CAAC;QAED,KAAK,KAAK,CAAC,CAAC,CAAC;YACX,IAAI,CAAC,IAAI;gBAAE,MAAM,IAAI,KAAK,CAAC,kDAAkD,CAAC,CAAA;YAC9E,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,iBAAiB,CAAC,CAAA;YACrD,UAAU,EAAE,CAAC,CAAC,CAAC,CAAA;YACf,OAAO,EAAE,IAAI,EAAE,OAAO,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC,EAAE,GAAG,EAAE,KAAK,EAAE,OAAO,EAAE,CAAA;QAChE,CAAC;QAED,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,IAAI,CAAC,IAAI;gBAAE,MAAM,IAAI,KAAK,CAAC,iDAAiD,CAAC,CAAA;YAC7E,MAAM,EAAE,UAAU,EAAE,GAAG,MAAM,MAAM,CAAC,iBAAiB,CAAC,CAAA;YACtD,UAAU,EAAE,CAAC,CAAC,CAAC,CAAA;YACf,OAAO;gBACL,IAAI,EAAE,QAAQ,CAAC,UAAU,CAAC,IAAI,EAAE,EAAE,UAAU,EAAE,QAAQ,CAAC,UAAU,EAAE,CAAC,EAAE,kBAAkB,CAAC;gBACzF,GAAG,EAAE,MAAM;gBACX,OAAO;aACR,CAAA;QACH,CAAC;IACH,CAAC;AACH,CAAC;AAED,SAAS,QAAQ,CAAC,IAAY,EAAE,IAAY;IAC1C,OAAO,IAAI,IAAI,CAAC,CAAC,IAAI,CAAC,EAAE,EAAE,IAAI,EAAE,GAAG,IAAI,gBAAgB,EAAE,CAAC,CAAA;AAC5D,CAAC;AAED;;;;;;;;;GASG;AACH,SAAS,OAAO,CAAC,IAAY;IAC3B,OAAO,IAAI,IAAI,CAAC,CAAC,GAAG,EAAE,IAAI,CAAC,EAAE,EAAE,IAAI,EAAE,wBAAwB,EAAE,CAAC,CAAA;AAClE,CAAC;AAED,SAAS,WAAW,CAAC,QAAgB;IACnC,MAAM,GAAG,GAAG,QAAQ,CAAC,WAAW,CAAC,GAAG,CAAC,CAAA;IACrC,IAAI,GAAG,GAAG,CAAC,IAAI,GAAG,KAAK,QAAQ,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,EAAE,CAAA;IACrD,OAAO,QAAQ,CAAC,KAAK,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,WAAW,EAAE,CAAA;AAC9C,CAAC"}

package/dist/image.d.ts

@@ -0,0 +1,17 @@
+import { type PdfImage } from './pdfcore.js';
+export type PdfPageSource = PdfImage;
+/**
+ * Draw a file to a canvas and read it back as JPEG.
+ *
+ * The white fill is not cosmetic: a canvas starts transparent, and drawing a
+ * PNG with alpha onto it and asking for JPEG gives BLACK where the transparency
+ * was on some browsers and white on others. Filling first makes it white
+ * everywhere, which is what somebody printing a page expects.
+ */
+export declare function imageToJpeg(file: Blob, quality: number, maxEdge: number | null): Promise<PdfPageSource>;
+/**
+ * Assemble pages into a PDF — one image per page, laid out at 72 dpi so a pixel
+ * is a point and the page is exactly the size of the picture on it.
+ */
+export declare function imagesToPdf(pages: readonly PdfPageSource[], title: string): Blob;
+//# sourceMappingURL=image.d.ts.map

package/dist/image.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"image.d.ts","sourceRoot":"","sources":["../src/image.ts"],"names":[],"mappings":"AAoBA,OAAO,EAAe,KAAK,QAAQ,EAAE,MAAM,cAAc,CAAA;AAEzD,MAAM,MAAM,aAAa,GAAG,QAAQ,CAAA;AAEpC;;;;;;;GAOG;AACH,wBAAsB,WAAW,CAAC,IAAI,EAAE,IAAI,EAAE,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,IAAI,GAAG,OAAO,CAAC,aAAa,CAAC,CAuB7G;AAED;;;GAGG;AACH,wBAAgB,WAAW,CAAC,KAAK,EAAE,SAAS,aAAa,EAAE,EAAE,KAAK,EAAE,MAAM,GAAG,IAAI,CAQhF"}

package/dist/image.js

@@ -0,0 +1,66 @@
+// Re-encoding a picture for embedding in a PDF.
+//
+// Lifted out of Universal Converter's `pdf.ts` when the document stack moved
+// here, because the document WRITER needs it (a `{ kind: 'image' }` block) and
+// the rest of that file — one-full-bleed-page-per-image — is a different
+// product feature that stayed in the app.
+//
+// EVERY EMBEDDED IMAGE IS A JPEG, AND THAT IS A REAL TRADE. `/DCTDecode` is the
+// one filter every reader supports and the only one that lets the compressed
+// bytes go in untouched. The cost is that it is LOSSY (a PNG screenshot of text
+// softens slightly) and that TRANSPARENCY IS FLATTENED onto white, because JPEG
+// has no alpha. The alternative, `/FlateDecode` with raw RGB, needs a deflate
+// encoder and would turn a 2 MP photo into a 20 MB page.
+// ⚠️ `.js`, not `.ts` and not extensionless — the TypeScript-ESM convention
+// that @unisim/media already follows, and it is load-bearing twice over. This
+// package EMITS, so `allowImportingTsExtensions` (which needs noEmit) is out;
+// and extensionless specifiers are emitted VERBATIM, so `dist/index.js` would
+// import './convert' and Node would refuse it with ERR_MODULE_NOT_FOUND. That
+// is not hypothetical: it is what the first build of this package did.
+import { PdfDocument } from './pdfcore.js';
+/**
+ * Draw a file to a canvas and read it back as JPEG.
+ *
+ * The white fill is not cosmetic: a canvas starts transparent, and drawing a
+ * PNG with alpha onto it and asking for JPEG gives BLACK where the transparency
+ * was on some browsers and white on others. Filling first makes it white
+ * everywhere, which is what somebody printing a page expects.
+ */
+export async function imageToJpeg(file, quality, maxEdge) {
+    const bitmap = await createImageBitmap(file);
+    let { width, height } = bitmap;
+    if (maxEdge && Math.max(width, height) > maxEdge) {
+        const scale = maxEdge / Math.max(width, height);
+        width = Math.max(1, Math.round(width * scale));
+        height = Math.max(1, Math.round(height * scale));
+    }
+    const canvas = document.createElement('canvas');
+    canvas.width = width;
+    canvas.height = height;
+    const ctx = canvas.getContext('2d');
+    if (!ctx)
+        throw new Error('This browser would not give us a canvas to draw on.');
+    ctx.fillStyle = '#ffffff';
+    ctx.fillRect(0, 0, width, height);
+    ctx.drawImage(bitmap, 0, 0, width, height);
+    bitmap.close?.();
+    const blob = await new Promise((resolve) => canvas.toBlob(resolve, 'image/jpeg', quality));
+    if (!blob)
+        throw new Error('This browser could not encode a JPEG.');
+    return { jpeg: new Uint8Array(await blob.arrayBuffer()), width, height };
+}
+/**
+ * Assemble pages into a PDF — one image per page, laid out at 72 dpi so a pixel
+ * is a point and the page is exactly the size of the picture on it.
+ */
+export function imagesToPdf(pages, title) {
+    if (!pages.length)
+        throw new Error('A PDF needs at least one page.');
+    const pdf = new PdfDocument();
+    for (const image of pages) {
+        const page = pdf.addPage(image.width, image.height);
+        page.drawImage(image, { x: 0, y: 0, width: image.width, height: image.height });
+    }
+    return pdf.save({ title });
+}
+//# sourceMappingURL=image.js.map

package/dist/image.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"image.js","sourceRoot":"","sources":["../src/image.ts"],"names":[],"mappings":"AAAA,gDAAgD;AAChD,EAAE;AACF,6EAA6E;AAC7E,+EAA+E;AAC/E,yEAAyE;AACzE,0CAA0C;AAC1C,EAAE;AACF,gFAAgF;AAChF,6EAA6E;AAC7E,gFAAgF;AAChF,gFAAgF;AAChF,8EAA8E;AAC9E,yDAAyD;AAEzD,4EAA4E;AAC5E,8EAA8E;AAC9E,8EAA8E;AAC9E,8EAA8E;AAC9E,8EAA8E;AAC9E,uEAAuE;AACvE,OAAO,EAAE,WAAW,EAAiB,MAAM,cAAc,CAAA;AAIzD;;;;;;;GAOG;AACH,MAAM,CAAC,KAAK,UAAU,WAAW,CAAC,IAAU,EAAE,OAAe,EAAE,OAAsB;IACnF,MAAM,MAAM,GAAG,MAAM,iBAAiB,CAAC,IAAI,CAAC,CAAA;IAC5C,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,GAAG,MAAM,CAAA;IAC9B,IAAI,OAAO,IAAI,IAAI,CAAC,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,GAAG,OAAO,EAAE,CAAC;QACjD,MAAM,KAAK,GAAG,OAAO,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,CAAA;QAC/C,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,KAAK,GAAG,KAAK,CAAC,CAAC,CAAA;QAC9C,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,KAAK,CAAC,CAAC,CAAA;IAClD,CAAC;IAED,MAAM,MAAM,GAAG,QAAQ,CAAC,aAAa,CAAC,QAAQ,CAAC,CAAA;IAC/C,MAAM,CAAC,KAAK,GAAG,KAAK,CAAA;IACpB,MAAM,CAAC,MAAM,GAAG,MAAM,CAAA;IACtB,MAAM,GAAG,GAAG,MAAM,CAAC,UAAU,CAAC,IAAI,CAAC,CAAA;IACnC,IAAI,CAAC,GAAG;QAAE,MAAM,IAAI,KAAK,CAAC,qDAAqD,CAAC,CAAA;IAChF,GAAG,CAAC,SAAS,GAAG,SAAS,CAAA;IACzB,GAAG,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,CAAA;IACjC,GAAG,CAAC,SAAS,CAAC,MAAM,EAAE,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,MAAM,CAAC,CAAA;IAC1C,MAAM,CAAC,KAAK,EAAE,EAAE,CAAA;IAEhB,MAAM,IAAI,GAAG,MAAM,IAAI,OAAO,CAAc,CAAC,OAAO,EAAE,EAAE,CACtD,MAAM,CAAC,MAAM,CAAC,OAAO,EAAE,YAAY,EAAE,OAAO,CAAC,CAAC,CAAA;IAChD,IAAI,CAAC,IAAI;QAAE,MAAM,IAAI,KAAK,CAAC,uCAAuC,CAAC,CAAA;IACnE,OAAO,EAAE,IAAI,EAAE,IAAI,UAAU,CAAC,MAAM,IAAI,CAAC,WAAW,EAAE,CAAC,EAAE,KAAK,EAAE,MAAM,EAAE,CAAA;AAC1E,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,WAAW,CAAC,KAA+B,EAAE,KAAa;IACxE,IAAI,CAAC,KAAK,CAAC,MAAM;QAAE,MAAM,IAAI,KAAK,CAAC,gCAAgC,CAAC,CAAA;IACpE,MAAM,GAAG,GAAG,IAAI,WAAW,EAAE,CAAA;IAC7B,KAAK,MAAM,KAAK,IAAI,KAAK,EAAE,CAAC;QAC1B,MAAM,IAAI,GAAG,GAAG,CAAC,OAAO,CAAC,KAAK,CAAC,KAAK,EAAE,KAAK,CAAC,MAAM,CAAC,CAAA;QACnD,IAAI,CAAC,SAAS,CAAC,KAAK,EAAE,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,KAAK,CAAC,KAAK,EAAE,MAAM,EAAE,KAAK,CAAC,MAAM,EAAE,CAAC,CAAA;IACjF,CAAC;IACD,OAAO,GAAG,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,CAAC,CAAA;AAC5B,CAAC"}

package/dist/index.d.ts

@@ -0,0 +1,10 @@
+export type { Block, DocNotice, Grid, ListItem, RichDoc, Run, } from './model.js';
+export { DEFAULT_DOC_SETTINGS, DEFAULT_PDF_SETTINGS, DOC_INPUT_EXTS, commonTargets, convertDocument, isTabular, readToDoc, targetsFor, } from './convert.js';
+export type { DocFormat, DocResult, DocSettings, FontChoice, PageMargin, PaperSize, PdfSettings, } from './convert.js';
+export { mergeToPdf, readForMerge } from './merge.js';
+export type { MergePart } from './merge.js';
+export { PdfDocument, PdfPage, rgb, widthOfText } from './pdfcore.js';
+export type { FontId, PdfImage, Rgb } from './pdfcore.js';
+export { imageToJpeg, imagesToPdf } from './image.js';
+export type { PdfPageSource } from './image.js';
+//# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AA2CA,YAAY,EACV,KAAK,EACL,SAAS,EACT,IAAI,EACJ,QAAQ,EACR,OAAO,EACP,GAAG,GACJ,MAAM,YAAY,CAAA;AAGnB,OAAO,EACL,oBAAoB,EACpB,oBAAoB,EACpB,cAAc,EACd,aAAa,EACb,eAAe,EACf,SAAS,EACT,SAAS,EACT,UAAU,GACX,MAAM,cAAc,CAAA;AACrB,YAAY,EACV,SAAS,EACT,SAAS,EACT,WAAW,EACX,UAAU,EACV,UAAU,EACV,SAAS,EACT,WAAW,GACZ,MAAM,cAAc,CAAA;AAGrB,OAAO,EAAE,UAAU,EAAE,YAAY,EAAE,MAAM,YAAY,CAAA;AACrD,YAAY,EAAE,SAAS,EAAE,MAAM,YAAY,CAAA;AAK3C,OAAO,EAAE,WAAW,EAAE,OAAO,EAAE,GAAG,EAAE,WAAW,EAAE,MAAM,cAAc,CAAA;AACrE,YAAY,EAAE,MAAM,EAAE,QAAQ,EAAE,GAAG,EAAE,MAAM,cAAc,CAAA;AACzD,OAAO,EAAE,WAAW,EAAE,WAAW,EAAE,MAAM,YAAY,CAAA;AACrD,YAAY,EAAE,aAAa,EAAE,MAAM,YAAY,CAAA"}

package/dist/index.js

@@ -0,0 +1,51 @@
+// ---------------------------------------------------------------------------
+// @unisim/doc — the suite's document stack.
+//
+// WHY THIS PACKAGE EXISTS
+// -----------------------
+// It existed twice. On 2026-08-13, two sessions that could not see each other
+// shipped the same 1,200–1,500 lines on the same afternoon: Universal PDF's
+// `unzip.ts` / `officeXml.ts` / `docxToBlocks.ts` / `odtToBlocks.ts` /
+// `blockPdf.ts`, and Universal Converter's `doc/` tree. A ZIP reader, a
+// namespace-agnostic XML layer, an OOXML walker, an ODF walker, and a
+// tokenise → wrap → paginate layout engine — twice, and diverging from the day
+// they landed. Both files carried a header saying "a fix to one is a fix to
+// neither".
+//
+// This is Converter's version, because Converter's was ahead on substance: a
+// model designed as a hub (nine readers × six writers) rather than grown out of
+// a Markdown renderer, with heading level 4, real image blocks, `underline` and
+// `strike` on a run, and `DocNotice` — a channel for reporting what a
+// conversion LOST, which the other had no equivalent of. It also fixed defects
+// the original still had: smart quotes and dashes survive (Universal PDF's
+// `sanitize()` flattened '’' to "'" and '—' to '-', though both are WinAnsi and
+// encode fine), and a hard line break inside a paragraph is honoured.
+//
+// Universal PDF's model had exactly one thing this lacked — `pagesetup`, added
+// 2026-08-20 for per-page orientation on import — so the merge is a superset.
+//
+// ONE MODEL AND ONE ENGINE, BUT TWO WRITERS ON PURPOSE
+// ----------------------------------------------------
+// The obvious next step, unifying the PDF writer too, is the wrong one. This
+// package writes PDFs through `pdfcore` — its own base-14 fonts and AFM tables,
+// no dependency — because Universal Converter's whole pitch is
+// offline-from-first-visit with nothing fetched. Universal PDF carries pdf-lib
+// anyway for signing, redaction and forms, so it gains nothing by dropping it.
+// A consumer that would rather render through pdf-lib can take the readers and
+// the model from here and keep its own writer; the seam is that `write/pdf.ts`
+// depends only on `pdfcore`'s small surface (addPage / drawText / drawRect /
+// drawImage / widthOfText), not on the rest of the package.
+//
+// NO RUNTIME DEPENDENCIES, and it should stay that way — that constraint is the
+// product for one of its two consumers.
+// ---------------------------------------------------------------------------
+// ---- Reading, writing, and the orchestration over both ---------------------
+export { DEFAULT_DOC_SETTINGS, DEFAULT_PDF_SETTINGS, DOC_INPUT_EXTS, commonTargets, convertDocument, isTabular, readToDoc, targetsFor, } from './convert.js';
+// ---- Joining several documents into one -----------------------------------
+export { mergeToPdf, readForMerge } from './merge.js';
+// ---- The PDF primitives, for a consumer that wants them directly ----------
+// Universal Converter's "images → one PDF" feature builds pages itself rather
+// than going through the document model, so the writer surface is public.
+export { PdfDocument, PdfPage, rgb, widthOfText } from './pdfcore.js';
+export { imageToJpeg, imagesToPdf } from './image.js';
+//# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,8EAA8E;AAC9E,4CAA4C;AAC5C,EAAE;AACF,0BAA0B;AAC1B,0BAA0B;AAC1B,8EAA8E;AAC9E,4EAA4E;AAC5E,uEAAuE;AACvE,wEAAwE;AACxE,sEAAsE;AACtE,+EAA+E;AAC/E,4EAA4E;AAC5E,YAAY;AACZ,EAAE;AACF,6EAA6E;AAC7E,gFAAgF;AAChF,gFAAgF;AAChF,sEAAsE;AACtE,+EAA+E;AAC/E,2EAA2E;AAC3E,gFAAgF;AAChF,sEAAsE;AACtE,EAAE;AACF,+EAA+E;AAC/E,8EAA8E;AAC9E,EAAE;AACF,uDAAuD;AACvD,uDAAuD;AACvD,6EAA6E;AAC7E,gFAAgF;AAChF,+DAA+D;AAC/D,+EAA+E;AAC/E,+EAA+E;AAC/E,+EAA+E;AAC/E,+EAA+E;AAC/E,6EAA6E;AAC7E,4DAA4D;AAC5D,EAAE;AACF,gFAAgF;AAChF,wCAAwC;AACxC,8EAA8E;AAY9E,+EAA+E;AAC/E,OAAO,EACL,oBAAoB,EACpB,oBAAoB,EACpB,cAAc,EACd,aAAa,EACb,eAAe,EACf,SAAS,EACT,SAAS,EACT,UAAU,GACX,MAAM,cAAc,CAAA;AAWrB,8EAA8E;AAC9E,OAAO,EAAE,UAAU,EAAE,YAAY,EAAE,MAAM,YAAY,CAAA;AAGrD,8EAA8E;AAC9E,8EAA8E;AAC9E,0EAA0E;AAC1E,OAAO,EAAE,WAAW,EAAE,OAAO,EAAE,GAAG,EAAE,WAAW,EAAE,MAAM,cAAc,CAAA;AAErE,OAAO,EAAE,WAAW,EAAE,WAAW,EAAE,MAAM,YAAY,CAAA"}

package/dist/merge.d.ts

@@ -0,0 +1,14 @@
+import { type RichDoc } from './model.js';
+import type { PdfSettings } from './write/pdfSettings.js';
+export interface MergePart {
+    name: string;
+    doc: RichDoc;
+}
+/** Read one file to a RichDoc, whichever kind it is. */
+export declare function readForMerge(file: File): Promise<MergePart>;
+export declare function mergeToPdf(parts: readonly MergePart[], settings: PdfSettings): Promise<{
+    blob: Blob;
+    pages: number;
+    notices: RichDoc['notices'];
+}>;
+//# sourceMappingURL=merge.d.ts.map

package/dist/merge.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"merge.d.ts","sourceRoot":"","sources":["../src/merge.ts"],"names":[],"mappings":"AAeA,OAAO,EAAa,KAAK,OAAO,EAAE,MAAM,YAAY,CAAA;AAEpD,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,wBAAwB,CAAA;AAEzD,MAAM,WAAW,SAAS;IACxB,IAAI,EAAE,MAAM,CAAA;IACZ,GAAG,EAAE,OAAO,CAAA;CACb;AAED,wDAAwD;AACxD,wBAAsB,YAAY,CAAC,IAAI,EAAE,IAAI,GAAG,OAAO,CAAC,SAAS,CAAC,CAGjE;AAED,wBAAsB,UAAU,CAC9B,KAAK,EAAE,SAAS,SAAS,EAAE,EAC3B,QAAQ,EAAE,WAAW,GACpB,OAAO,CAAC;IAAE,IAAI,EAAE,IAAI,CAAC;IAAC,KAAK,EAAE,MAAM,CAAC;IAAC,OAAO,EAAE,OAAO,CAAC,SAAS,CAAC,CAAA;CAAE,CAAC,CAgCrE"}

package/dist/merge.js

@@ -0,0 +1,51 @@
+// ---------------------------------------------------------------------------
+// Several documents → one PDF.
+//
+// Deliberately NOT a second pipeline. Each file goes through the same reader it
+// would on its own, the resulting RichDocs are concatenated into one, and that
+// goes through the same writer with the same settings. The only thing this file
+// adds is the join: a page break between documents and a heading naming each.
+//
+// The alternative — writing each to its own PDF and stitching the byte streams
+// — needs a PDF *parser* to renumber every object, which is the thing
+// `pdfcore.ts` explicitly does not do. Joining at the model level is both less
+// code and a better result: the page numbers run 1..n across the whole thing
+// rather than restarting per document.
+// ---------------------------------------------------------------------------
+import { addNotice } from './model.js';
+import { docToPdf } from './write/pdf.js';
+/** Read one file to a RichDoc, whichever kind it is. */
+export async function readForMerge(file) {
+    const { readToDoc } = await import('./index.js');
+    return { name: file.name, doc: await readToDoc(file) };
+}
+export async function mergeToPdf(parts, settings) {
+    if (!parts.length)
+        throw new Error('There are no documents to join.');
+    const merged = { blocks: [], notices: [], title: 'Documents' };
+    parts.forEach((part, index) => {
+        // A page break BEFORE each document except the first: putting one after
+        // each instead leaves a blank final page.
+        if (index > 0)
+            merged.blocks.push({ kind: 'pagebreak' });
+        // The filename as a heading. Without it a joined PDF is one undifferentiated
+        // run of text and there is no way to tell where one document ended.
+        merged.blocks.push({
+            kind: 'heading',
+            level: 1,
+            runs: [{ text: part.name.replace(/\.[^.]+$/, '') }],
+        });
+        merged.blocks.push(...part.doc.blocks);
+        for (const notice of part.doc.notices) {
+            // Prefixed with the filename: in a joined batch "only its text could be
+            // read" is useless without knowing which of the six it refers to.
+            addNotice(merged, `${part.name}: ${notice.message}`, notice.severity);
+        }
+    });
+    const result = await docToPdf(merged, settings);
+    if (result.dropped.length) {
+        addNotice(merged, `Some characters aren’t in the PDF’s built-in fonts and came out as “?”: ${result.dropped.slice(0, 12).join(' ')}`);
+    }
+    return { blob: result.blob, pages: result.pages, notices: merged.notices };
+}
+//# sourceMappingURL=merge.js.map

package/dist/merge.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"merge.js","sourceRoot":"","sources":["../src/merge.ts"],"names":[],"mappings":"AAAA,8EAA8E;AAC9E,+BAA+B;AAC/B,EAAE;AACF,gFAAgF;AAChF,+EAA+E;AAC/E,gFAAgF;AAChF,8EAA8E;AAC9E,EAAE;AACF,+EAA+E;AAC/E,sEAAsE;AACtE,+EAA+E;AAC/E,6EAA6E;AAC7E,uCAAuC;AACvC,8EAA8E;AAE9E,OAAO,EAAE,SAAS,EAAgB,MAAM,YAAY,CAAA;AACpD,OAAO,EAAE,QAAQ,EAAE,MAAM,gBAAgB,CAAA;AAQzC,wDAAwD;AACxD,MAAM,CAAC,KAAK,UAAU,YAAY,CAAC,IAAU;IAC3C,MAAM,EAAE,SAAS,EAAE,GAAG,MAAM,MAAM,CAAC,YAAY,CAAC,CAAA;IAChD,OAAO,EAAE,IAAI,EAAE,IAAI,CAAC,IAAI,EAAE,GAAG,EAAE,MAAM,SAAS,CAAC,IAAI,CAAC,EAAE,CAAA;AACxD,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,UAAU,CAC9B,KAA2B,EAC3B,QAAqB;IAErB,IAAI,CAAC,KAAK,CAAC,MAAM;QAAE,MAAM,IAAI,KAAK,CAAC,iCAAiC,CAAC,CAAA;IAErE,MAAM,MAAM,GAAY,EAAE,MAAM,EAAE,EAAE,EAAE,OAAO,EAAE,EAAE,EAAE,KAAK,EAAE,WAAW,EAAE,CAAA;IAEvE,KAAK,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE;QAC5B,wEAAwE;QACxE,0CAA0C;QAC1C,IAAI,KAAK,GAAG,CAAC;YAAE,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,CAAC,CAAA;QACxD,6EAA6E;QAC7E,oEAAoE;QACpE,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC;YACjB,IAAI,EAAE,SAAS;YACf,KAAK,EAAE,CAAC;YACR,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,IAAI,CAAC,IAAI,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,EAAE,CAAC;SACpD,CAAC,CAAA;QACF,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,MAAM,CAAC,CAAA;QACtC,KAAK,MAAM,MAAM,IAAI,IAAI,CAAC,GAAG,CAAC,OAAO,EAAE,CAAC;YACtC,wEAAwE;YACxE,kEAAkE;YAClE,SAAS,CAAC,MAAM,EAAE,GAAG,IAAI,CAAC,IAAI,KAAK,MAAM,CAAC,OAAO,EAAE,EAAE,MAAM,CAAC,QAAQ,CAAC,CAAA;QACvE,CAAC;IACH,CAAC,CAAC,CAAA;IAEF,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,MAAM,EAAE,QAAQ,CAAC,CAAA;IAC/C,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC;QAC1B,SAAS,CACP,MAAM,EACN,2EAA2E,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CACnH,CAAA;IACH,CAAC;IACD,OAAO,EAAE,IAAI,EAAE,MAAM,CAAC,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,KAAK,EAAE,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,CAAA;AAC5E,CAAC"}