@pollium/vyva-convert 0.0.0-stage → 0.1.0

package/dist/import/titles.js

@@ -0,0 +1,6 @@
+//#region src/import/titles.ts
+const MAX_TITLE = 200;
+const UNTITLED = "Untitled";
+const titleOf = (text, filename) => (text.trim() || filename.replace(/\.[a-z0-9]+$/i, "").replace(/[_-]+/g, " ").trim() || UNTITLED).slice(0, MAX_TITLE);
+//#endregion
+export { titleOf };

package/dist/import.d.ts

@@ -0,0 +1,14 @@
+import { SaveImage } from "./import/images.js";
+import { Content } from "@pollium/vyva-core/content";
+//#region src/import.d.ts
+export declare const MAX_IMPORT_BYTES: number;
+export interface ImportOptions {
+  saveImage: SaveImage;
+  maxBytes?: number;
+}
+export interface ImportedContent {
+  title: string;
+  content: Content;
+}
+export declare const importFile: (data: Uint8Array, filename: string, { saveImage, maxBytes }: ImportOptions) => Promise<ImportedContent>;
+//#endregion

package/dist/import.js

@@ -0,0 +1,66 @@
+import { ConvertError } from "./errors.js";
+import { importDocx, importMarkdown, importPdf, importPlainText } from "./import/documents.js";
+import { importPptx } from "./import/pptx.js";
+import { unzipSync } from "fflate";
+//#region src/import.ts
+const MAX_IMPORT_BYTES = 26214400;
+const ZIP_SIGNATURE = 67324752;
+const PDF_SIGNATURE = "%PDF-";
+const TEXT_SAMPLE_BYTES = 65536;
+const BOM = /^/;
+const zipParts = (buffer) => {
+	if (buffer.length < 4 || buffer.readUInt32LE(0) !== ZIP_SIGNATURE) return null;
+	try {
+		const names = /* @__PURE__ */ new Set();
+		unzipSync(buffer, { filter: (file) => {
+			names.add(file.name);
+			return false;
+		} });
+		return names;
+	} catch {
+		return null;
+	}
+};
+const isUtf8Text = (buffer) => {
+	const sample = buffer.subarray(0, TEXT_SAMPLE_BYTES);
+	if (sample.includes(0)) return false;
+	try {
+		new TextDecoder("utf-8", { fatal: true }).decode(sample);
+		return true;
+	} catch {
+		return false;
+	}
+};
+const detect = (buffer, filename) => {
+	const parts = zipParts(buffer);
+	if (parts !== null) {
+		if (parts.has("word/document.xml")) return "docx";
+		return parts.has("ppt/presentation.xml") ? "pptx" : null;
+	}
+	if (buffer.subarray(0, 5).toString("latin1") === PDF_SIGNATURE) return "pdf";
+	if (!isUtf8Text(buffer)) return null;
+	const extension = filename.toLowerCase().split(".").at(-1) ?? "";
+	if (extension === "md" || extension === "markdown") return "markdown";
+	if (extension === "txt") return "text";
+	return null;
+};
+const convert = async (detected, buffer, filename, saveImage) => {
+	if (detected === "docx") return importDocx(buffer, filename, saveImage);
+	if (detected === "pptx") return importPptx(buffer, filename, saveImage);
+	if (detected === "pdf") return importPdf(buffer, filename);
+	const text = buffer.toString("utf8").replace(BOM, "");
+	return detected === "markdown" ? importMarkdown(text, filename) : importPlainText(text, filename);
+};
+const isEmpty = (content) => content.kind === "document" ? content.blocks.length === 0 : content.slides.length === 0;
+const importFile = async (data, filename, { saveImage, maxBytes = MAX_IMPORT_BYTES }) => {
+	if (data.length > maxBytes) throw new ConvertError("too-large", `The file is larger than ${maxBytes} bytes`);
+	const buffer = Buffer.from(data.buffer, data.byteOffset, data.byteLength);
+	const name = filename.replace(/[\\/]/g, "_").slice(-200);
+	const detected = detect(buffer, name);
+	if (detected === null) throw new ConvertError("unsupported", `${name} is not a file Vyva can import`);
+	const imported = await convert(detected, buffer, name, saveImage);
+	if (isEmpty(imported.content)) throw new ConvertError("empty", `${name} has nothing to import`);
+	return imported;
+};
+//#endregion
+export { MAX_IMPORT_BYTES, importFile };

package/dist/text-measure.d.ts

@@ -0,0 +1,4 @@
+import { FontFamily, TextStyle } from "@pollium/vyva-core/content";
+//#region src/text-measure.d.ts
+export declare const textHeight: (html: string, style: TextStyle, font: FontFamily, width: number) => number;
+//#endregion

package/dist/text-measure.js

@@ -0,0 +1,98 @@
+import { FONT_WIDTHS, MEASURED_CHARS } from "./font-widths.js";
+import { CODE_FONT } from "@pollium/vyva-core/code";
+//#region src/text-measure.ts
+const EM = 1e3;
+const BLOCK_GAP_EMS = .35;
+const LIST_INDENT_EMS = 1.15;
+const WIDTH_SAFETY = .97;
+const BLOCK_END = /<\/(?:p|li|h[1-6])>/i;
+const TAG = /<(\/?)([a-z0-9]+)[^>]*>|([^<]+)/gi;
+const BOLD_TAGS = /* @__PURE__ */ new Set([
+	"strong",
+	"b",
+	"h1",
+	"h2",
+	"h3",
+	"h4",
+	"h5",
+	"h6"
+]);
+const ENTITIES = {
+	amp: "&",
+	lt: "<",
+	gt: ">",
+	quot: "\"",
+	apos: "'",
+	nbsp: " ",
+	"#39": "'"
+};
+const INDEX = new Map([...MEASURED_CHARS].map((char, index) => [char, index]));
+const FALLBACK = INDEX.get("n") ?? 0;
+const decoded = (text) => text.replace(/&(#?\w+);/g, (entity, name) => ENTITIES[name] ?? (name.startsWith("#") ? String.fromCodePoint(Number(name.slice(1))) : entity));
+const glyphsOf = (html, font, isBold) => {
+	const glyphs = [];
+	let bold = 0;
+	let code = 0;
+	for (const [, closing, tag, text] of html.matchAll(TAG)) {
+		if (text !== void 0) {
+			for (const char of decoded(text).replace(/\s+/g, " ")) glyphs.push({
+				char,
+				font: code > 0 ? CODE_FONT : font,
+				isBold: isBold || bold > 0
+			});
+			continue;
+		}
+		const step = closing === "/" ? -1 : 1;
+		const name = tag.toLowerCase();
+		if (BOLD_TAGS.has(name)) bold = Math.max(0, bold + step);
+		else if (name === "code") code = Math.max(0, code + step);
+	}
+	return glyphs;
+};
+const blocksOf = (html, font, isBold) => html.split(BLOCK_END).flatMap((chunk) => {
+	const glyphs = glyphsOf(chunk, font, isBold);
+	while (glyphs[0]?.char === " ") glyphs.shift();
+	while (glyphs.at(-1)?.char === " ") glyphs.pop();
+	return glyphs.length === 0 ? [] : [{
+		glyphs,
+		isItem: /<li\b/i.test(chunk)
+	}];
+});
+const widthOf = (glyph, fontSize) => {
+	return FONT_WIDTHS[glyph.font][glyph.isBold ? "bold" : "regular"][INDEX.get(glyph.char) ?? FALLBACK] / EM * fontSize;
+};
+const linesOf = (glyphs, fontSize, width) => {
+	let lines = 1;
+	let line = 0;
+	let word = 0;
+	let space = 0;
+	const place = () => {
+		if (word > width) lines = Infinity;
+		if (line > 0 && line + space + word > width) {
+			lines += 1;
+			line = 0;
+		}
+		line += (line === 0 ? 0 : space) + word;
+		word = 0;
+	};
+	for (const glyph of glyphs) {
+		if (glyph.char !== " ") {
+			word += widthOf(glyph, fontSize);
+			continue;
+		}
+		place();
+		space = widthOf(glyph, fontSize);
+	}
+	place();
+	return lines;
+};
+const textHeight = (html, style, font, width) => {
+	const blocks = blocksOf(html, font, style.isBold);
+	return blocks.reduce((height, block, index) => {
+		const indent = block.isItem ? LIST_INDENT_EMS * style.fontSize : 0;
+		const gap = !block.isItem && index === blocks.length - 1 ? 0 : BLOCK_GAP_EMS * style.fontSize;
+		return height + linesOf(block.glyphs, style.fontSize, (width - indent) * WIDTH_SAFETY) * style.fontSize * style.lineHeight + gap;
+	}, 0);
+};
+//#endregion
+export { textHeight };

package/package.json

@@ -1,6 +1,52 @@
 {
   "name": "@pollium/vyva-convert",
-  "version": "0.0.0-stage",
-  "stub": true,
-  "description": "Temporary package placeholder for staged publishing"
+  "version": "0.1.0",
+  "description": "Server-side tools for Vyva content: clean what clients send, import Word, PowerPoint, PDF and Markdown files, and export to PDF, Word, PowerPoint and HTML.",
+  "type": "module",
+  "sideEffects": false,
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/Pollium/Vyva.git",
+    "directory": "packages/convert"
+  },
+  "files": [
+    "dist"
+  ],
+  "exports": {
+    "./*": {
+      "types": "./dist/*.d.ts",
+      "default": "./dist/*.js"
+    }
+  },
+  "publishConfig": {
+    "access": "public"
+  },
+  "dependencies": {
+    "@fontsource-variable/inter": "^5.3.0",
+    "@fontsource-variable/jetbrains-mono": "^5.3.0",
+    "@fontsource-variable/lora": "^5.3.0",
+    "@fontsource-variable/merriweather": "^5.3.0",
+    "@fontsource-variable/montserrat": "^5.3.0",
+    "@fontsource-variable/open-sans": "^5.3.0",
+    "@fontsource-variable/playfair-display": "^5.3.0",
+    "@fontsource-variable/roboto": "^5.3.0",
+    "docx": "^9.9.0",
+    "fflate": "^0.8.3",
+    "htmlparser2": "^10.1.0",
+    "mammoth": "^1.13.0",
+    "pptxgenjs": "^4.0.1",
+    "sanitize-html": "^2.18.0",
+    "unpdf": "^1.8.1",
+    "@pollium/vyva-core": "0.1.0"
+  },
+  "devDependencies": {
+    "@types/node": "^26.6.4",
+    "@types/sanitize-html": "^2.16.2"
+  },
+  "scripts": {
+    "build": "tsdown",
+    "typecheck": "tsc -p tsconfig.json",
+    "lint": "oxlint --deny-warnings",
+    "test": "vitest run"
+  }
 }