@mulmocast/slide 0.4.1 → 0.5.0

package/lib/convert/pdfvision.js

@@ -0,0 +1,247 @@
+import * as fs from "fs";
+import * as path from "path";
+import { execSync } from "child_process";
+import { mulmoScriptSchema } from "mulmocast";
+import { resolveLang } from "../utils/lang.js";
+import { convertPdfToImages, extractTextFromPdf, writeMulmoScript } from "../utils/pdf.js";
+import { checkDependencies } from "../utils/dependencies.js";
+import { resolveVisionProvider, callVisionAPI, callTextLLM, } from "../utils/vision-provider.js";
+import { buildDocumentAnalysisPrompt, parseDocumentAnalysis, } from "../utils/document-analysis.js";
+import { buildNarrationPrompt, parseNarrationResponse } from "../utils/narration-generator.js";
+const CROP_PADDING_PERCENT = 5;
+const CROP_DPI = 600;
+const TRIM_BORDER_PX = 20;
+const getMagickCmd = () => {
+    return process.platform === "linux" ? "convert" : "magick";
+};
+const buildPageImages = (imagesDir, basename, pageCount) => {
+    return Array.from({ length: pageCount }, (_, i) => ({
+        path: path.join(imagesDir, `${basename}-${i}.png`),
+    })).filter((img) => fs.existsSync(img.path));
+};
+const sanitizeLabel = (label) => {
+    return label.replace(/[^a-zA-Z0-9_-]/g, "_").toLowerCase();
+};
+const getImageDimensions = (imagePath) => {
+    try {
+        const magick = getMagickCmd();
+        const identifyCmd = magick === "magick" ? "magick identify" : "identify";
+        const output = execSync(`${identifyCmd} -format "%w %h" "${imagePath}"`, { encoding: "utf-8" });
+        const [w, h] = output.trim().split(" ").map(Number);
+        return { width: w, height: h };
+    }
+    catch {
+        return null;
+    }
+};
+const convertPageHighRes = (pdfPath, page, outputPath) => {
+    try {
+        const magick = getMagickCmd();
+        const cmd = `${magick} -density ${CROP_DPI} -antialias "${pdfPath}[${page}]" -background white -alpha remove -quality 95 "${outputPath}"`;
+        execSync(cmd, { stdio: "pipe" });
+        return fs.existsSync(outputPath);
+    }
+    catch {
+        return false;
+    }
+};
+const addPadding = (bbox, padding) => {
+    const x = Math.max(0, bbox.x - padding);
+    const y = Math.max(0, bbox.y - padding);
+    const width = Math.min(100 - x, bbox.width + padding * 2);
+    const height = Math.min(100 - y, bbox.height + padding * 2);
+    return { x, y, width, height };
+};
+const cropFigure = (pageImagePath, outputPath, bbox) => {
+    try {
+        const dims = getImageDimensions(pageImagePath);
+        if (!dims)
+            return false;
+        const padded = addPadding(bbox, CROP_PADDING_PERCENT);
+        const cropX = Math.round((padded.x / 100) * dims.width);
+        const cropY = Math.round((padded.y / 100) * dims.height);
+        const cropW = Math.round((padded.width / 100) * dims.width);
+        const cropH = Math.round((padded.height / 100) * dims.height);
+        const magick = getMagickCmd();
+        const cropCmd = [
+            `${magick} "${pageImagePath}"`,
+            `-crop ${cropW}x${cropH}+${cropX}+${cropY} +repage`,
+            `-trim +repage`,
+            `-bordercolor white -border ${TRIM_BORDER_PX}`,
+            `"${outputPath}"`,
+        ].join(" ");
+        execSync(cropCmd, { stdio: "pipe" });
+        return fs.existsSync(outputPath);
+    }
+    catch {
+        return false;
+    }
+};
+const cropFigures = (analysis, imagesDir, basename, pdfPath) => {
+    const figureImageMap = new Map();
+    // Identify pages that need high-res conversion
+    const pagesWithFigures = new Set();
+    analysis.figures.forEach((figure) => {
+        if (figure.bbox && figure.label) {
+            pagesWithFigures.add(figure.page);
+        }
+    });
+    // Convert those pages at high DPI
+    const highResDir = path.join(imagesDir, "_highres");
+    if (pagesWithFigures.size > 0) {
+        if (!fs.existsSync(highResDir)) {
+            fs.mkdirSync(highResDir, { recursive: true });
+        }
+    }
+    const highResMap = new Map();
+    pagesWithFigures.forEach((page) => {
+        const highResPath = path.join(highResDir, `${basename}-${page}-hires.png`);
+        if (convertPageHighRes(pdfPath, page, highResPath)) {
+            highResMap.set(page, highResPath);
+            console.log(`  High-res (${CROP_DPI}dpi): page ${page}`);
+        }
+    });
+    // Crop figures from high-res images (fallback to standard images)
+    analysis.figures.forEach((figure) => {
+        if (!figure.bbox || !figure.label)
+            return;
+        const sourceImage = highResMap.get(figure.page) ?? path.join(imagesDir, `${basename}-${figure.page}.png`);
+        if (!fs.existsSync(sourceImage))
+            return;
+        const sanitized = sanitizeLabel(figure.label);
+        const croppedFilename = `${basename}-fig-${sanitized}.png`;
+        const croppedPath = path.join(imagesDir, croppedFilename);
+        if (cropFigure(sourceImage, croppedPath, figure.bbox)) {
+            figureImageMap.set(figure.label, `./images/${croppedFilename}`);
+            console.log(`  Cropped: ${figure.label} → ${croppedFilename}`);
+        }
+    });
+    // Clean up high-res temp images
+    if (fs.existsSync(highResDir)) {
+        fs.readdirSync(highResDir).forEach((f) => fs.unlinkSync(path.join(highResDir, f)));
+        fs.rmdirSync(highResDir);
+    }
+    return figureImageMap;
+};
+const analyzeDocument = async (provider, images, extractedTexts, lang) => {
+    console.log(`Analyzing document with ${provider} Vision API...`);
+    const prompt = buildDocumentAnalysisPrompt({
+        pageCount: images.length,
+        extractedTexts,
+        lang,
+    });
+    const response = await callVisionAPI(provider, { prompt, images });
+    return parseDocumentAnalysis(response);
+};
+const generateNarrations = async (provider, analysis, extractedTexts, lang) => {
+    console.log("Generating narration with text LLM...");
+    const prompt = buildNarrationPrompt({
+        documentAnalysis: analysis,
+        extractedTexts,
+        lang,
+    });
+    const response = await callTextLLM(provider, prompt);
+    const entries = parseNarrationResponse(response, analysis.slides.length);
+    return entries.map((e) => e.text);
+};
+const buildMulmoScript = (analysis, narrations, basename, lang, figureImageMap) => {
+    const beats = analysis.slides.map((slide, i) => {
+        const imagePage = slide.imagePage ?? slide.sourcePages[0] ?? 0;
+        const pageImagePath = `./images/${basename}-${imagePage}.png`;
+        const imagePath = slide.figureRef && figureImageMap.has(slide.figureRef)
+            ? figureImageMap.get(slide.figureRef)
+            : pageImagePath;
+        return {
+            text: narrations[i] || "",
+            image: {
+                type: "image",
+                source: {
+                    kind: "path",
+                    path: imagePath,
+                },
+            },
+        };
+    });
+    const mulmoScript = {
+        $mulmocast: { version: "1.1" },
+        lang,
+        beats,
+    };
+    const result = mulmoScriptSchema.safeParse(mulmoScript);
+    if (!result.success) {
+        console.error("MulmoScript validation failed:");
+        console.error(result.error.format());
+        throw new Error("Invalid MulmoScript generated");
+    }
+    return result.data;
+};
+export const convertPdfVision = async (options) => {
+    const { inputPath, provider: providerArg } = options;
+    const pdfFile = path.resolve(inputPath);
+    if (!fs.existsSync(pdfFile)) {
+        throw new Error(`File not found: ${pdfFile}`);
+    }
+    checkDependencies("pdf");
+    const provider = resolveVisionProvider(providerArg);
+    console.log(`Using Vision provider: ${provider}`);
+    const basename = path.basename(pdfFile, ".pdf");
+    const outputDir = path.join("scripts", basename);
+    const imagesDir = path.join(outputDir, "images");
+    if (!fs.existsSync(outputDir)) {
+        fs.mkdirSync(outputDir, { recursive: true });
+    }
+    // Step 1: Convert PDF to page images
+    console.log("Converting PDF to images...");
+    const { slideCount: pageCount } = convertPdfToImages({
+        pdfPath: pdfFile,
+        imagesDir,
+        basename,
+    });
+    // Step 2: Extract text
+    console.log("Extracting text from PDF...");
+    const pageTexts = await extractTextFromPdf(pdfFile);
+    const extractedTexts = [];
+    pageTexts.forEach((page) => {
+        extractedTexts[page.pageNumber] = page.text;
+    });
+    console.log(`Extracted text from ${pageTexts.length} pages`);
+    const resolvedLang = resolveLang(options.lang, extractedTexts.filter(Boolean));
+    // Save extracted texts
+    const hasExtractedText = extractedTexts.some((t) => t && t.length > 0);
+    let extractedTextsPath = null;
+    if (hasExtractedText) {
+        extractedTextsPath = path.join(outputDir, "extracted_texts.json");
+        fs.writeFileSync(extractedTextsPath, JSON.stringify(extractedTexts, null, 2));
+    }
+    // Step 3: Vision API - analyze document (1 API call)
+    const images = buildPageImages(imagesDir, basename, pageCount);
+    const analysis = await analyzeDocument(provider, images, extractedTexts, resolvedLang);
+    // Save analysis
+    const analysisPath = path.join(outputDir, "analysis.json");
+    fs.writeFileSync(analysisPath, JSON.stringify(analysis, null, 2));
+    console.log(`Document analysis saved: ${analysisPath}`);
+    console.log(`  Sections: ${analysis.sections.length}`);
+    console.log(`  Figures: ${analysis.figures.length}`);
+    console.log(`  Planned slides: ${analysis.slides.length}`);
+    // Step 4: Crop figures from high-res page images
+    console.log("Cropping figures from page images...");
+    const figureImageMap = cropFigures(analysis, imagesDir, basename, pdfFile);
+    console.log(`  Cropped ${figureImageMap.size} figures`);
+    // Step 5: Text LLM - generate narration (1 API call)
+    const narrations = await generateNarrations(provider, analysis, extractedTexts, resolvedLang);
+    // Step 6: Build and write MulmoScript
+    const mulmoScript = buildMulmoScript(analysis, narrations, basename, resolvedLang, figureImageMap);
+    const jsonPath = path.join(outputDir, `${basename}.json`);
+    writeMulmoScript(mulmoScript, jsonPath);
+    console.log(`\n✓ pdfvision conversion complete!`);
+    console.log(`  Provider: ${provider}`);
+    console.log(`  Pages: ${pageCount} → Slides: ${analysis.slides.length}`);
+    console.log(`  Output: ${jsonPath}`);
+    return {
+        mulmoScriptPath: jsonPath,
+        extractedTextsPath,
+        analysisPath,
+        slideCount: analysis.slides.length,
+    };
+};
+//# sourceMappingURL=pdfvision.js.map

package/lib/convert/pdfvision.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"pdfvision.js","sourceRoot":"","sources":["../../src/convert/pdfvision.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,MAAM,IAAI,CAAC;AACzB,OAAO,KAAK,IAAI,MAAM,MAAM,CAAC;AAC7B,OAAO,EAAE,QAAQ,EAAE,MAAM,eAAe,CAAC;AACzC,OAAO,EAAE,iBAAiB,EAAE,MAAM,WAAW,CAAC;AAE9C,OAAO,EAAE,WAAW,EAAsB,MAAM,kBAAkB,CAAC;AACnE,OAAO,EAAE,kBAAkB,EAAE,kBAAkB,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC;AAC3F,OAAO,EAAE,iBAAiB,EAAE,MAAM,0BAA0B,CAAC;AAC7D,OAAO,EACL,qBAAqB,EACrB,aAAa,EACb,WAAW,GAGZ,MAAM,6BAA6B,CAAC;AACrC,OAAO,EACL,2BAA2B,EAC3B,qBAAqB,GAGtB,MAAM,+BAA+B,CAAC;AACvC,OAAO,EAAE,oBAAoB,EAAE,sBAAsB,EAAE,MAAM,iCAAiC,CAAC;AAI/F,MAAM,oBAAoB,GAAG,CAAC,CAAC;AAC/B,MAAM,QAAQ,GAAG,GAAG,CAAC;AACrB,MAAM,cAAc,GAAG,EAAE,CAAC;AAe1B,MAAM,YAAY,GAAG,GAAW,EAAE;IAChC,OAAO,OAAO,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,QAAQ,CAAC;AAC7D,CAAC,CAAC;AAEF,MAAM,eAAe,GAAG,CAAC,SAAiB,EAAE,QAAgB,EAAE,SAAiB,EAAiB,EAAE;IAChG,OAAO,KAAK,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,SAAS,EAAE,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC;QAClD,IAAI,EAAE,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,GAAG,QAAQ,IAAI,CAAC,MAAM,CAAC;KACnD,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,EAAE,CAAC,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC;AAC/C,CAAC,CAAC;AAEF,MAAM,aAAa,GAAG,CAAC,KAAa,EAAU,EAAE;IAC9C,OAAO,KAAK,CAAC,OAAO,CAAC,iBAAiB,EAAE,GAAG,CAAC,CAAC,WAAW,EAAE,CAAC;AAC7D,CAAC,CAAC;AAEF,MAAM,kBAAkB,GAAG,CAAC,SAAiB,EAA4C,EAAE;IACzF,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,YAAY,EAAE,CAAC;QAC9B,MAAM,WAAW,GAAG,MAAM,KAAK,QAAQ,CAAC,CAAC,CAAC,iBAAiB,CAAC,CAAC,CAAC,UAAU,CAAC;QACzE,MAAM,MAAM,GAAG,QAAQ,CAAC,GAAG,WAAW,qBAAqB,SAAS,GAAG,EAAE,EAAE,QAAQ,EAAE,OAAO,EAAE,CAAC,CAAC;QAChG,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;QACpD,OAAO,EAAE,KAAK,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,CAAC;IACjC,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,kBAAkB,GAAG,CAAC,OAAe,EAAE,IAAY,EAAE,UAAkB,EAAW,EAAE;IACxF,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,YAAY,EAAE,CAAC;QAC9B,MAAM,GAAG,GAAG,GAAG,MAAM,aAAa,QAAQ,gBAAgB,OAAO,IAAI,IAAI,mDAAmD,UAAU,GAAG,CAAC;QAC1I,QAAQ,CAAC,GAAG,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;QACjC,OAAO,EAAE,CAAC,UAAU,CAAC,UAAU,CAAC,CAAC;IACnC,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,UAAU,GAAG,CACjB,IAA6D,EAC7D,OAAe,EAC0C,EAAE;IAC3D,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,GAAG,OAAO,CAAC,CAAC;IACxC,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,GAAG,OAAO,CAAC,CAAC;IACxC,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,EAAE,IAAI,CAAC,KAAK,GAAG,OAAO,GAAG,CAAC,CAAC,CAAC;IAC1D,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,EAAE,IAAI,CAAC,MAAM,GAAG,OAAO,GAAG,CAAC,CAAC,CAAC;IAC5D,OAAO,EAAE,CAAC,EAAE,CAAC,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC;AACjC,CAAC,CAAC;AAEF,MAAM,UAAU,GAAG,CACjB,aAAqB,EACrB,UAAkB,EAClB,IAA6D,EACpD,EAAE;IACX,IAAI,CAAC;QACH,MAAM,IAAI,GAAG,kBAAkB,CAAC,aAAa,CAAC,CAAC;QAC/C,IAAI,CAAC,IAAI;YAAE,OAAO,KAAK,CAAC;QAExB,MAAM,MAAM,GAAG,UAAU,CAAC,IAAI,EAAE,oBAAoB,CAAC,CAAC;QAEtD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC;QACxD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,CAAC,GAAG,GAAG,CAAC,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC;QACzD,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,KAAK,GAAG,GAAG,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC;QAC5D,MAAM,KAAK,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,MAAM,GAAG,GAAG,CAAC,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC;QAE9D,MAAM,MAAM,GAAG,YAAY,EAAE,CAAC;QAC9B,MAAM,OAAO,GAAG;YACd,GAAG,MAAM,KAAK,aAAa,GAAG;YAC9B,SAAS,KAAK,IAAI,KAAK,IAAI,KAAK,IAAI,KAAK,UAAU;YACnD,eAAe;YACf,8BAA8B,cAAc,EAAE;YAC9C,IAAI,UAAU,GAAG;SAClB,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACZ,QAAQ,CAAC,OAAO,EAAE,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;QACrC,OAAO,EAAE,CAAC,UAAU,CAAC,UAAU,CAAC,CAAC;IACnC,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC,CAAC;AAEF,MAAM,WAAW,GAAG,CAClB,QAA0B,EAC1B,SAAiB,EACjB,QAAgB,EAChB,OAAe,EACM,EAAE;IACvB,MAAM,cAAc,GAAG,IAAI,GAAG,EAAkB,CAAC;IAEjD,+CAA+C;IAC/C,MAAM,gBAAgB,GAAG,IAAI,GAAG,EAAU,CAAC;IAC3C,QAAQ,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,MAAkB,EAAE,EAAE;QAC9C,IAAI,MAAM,CAAC,IAAI,IAAI,MAAM,CAAC,KAAK,EAAE,CAAC;YAChC,gBAAgB,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC;QACpC,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,kCAAkC;IAClC,MAAM,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,UAAU,CAAC,CAAC;IACpD,IAAI,gBAAgB,CAAC,IAAI,GAAG,CAAC,EAAE,CAAC;QAC9B,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;YAC/B,EAAE,CAAC,SAAS,CAAC,UAAU,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;QAChD,CAAC;IACH,CAAC;IAED,MAAM,UAAU,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC7C,gBAAgB,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,EAAE;QAChC,MAAM,WAAW,GAAG,IAAI,CAAC,IAAI,CAAC,UAAU,EAAE,GAAG,QAAQ,IAAI,IAAI,YAAY,CAAC,CAAC;QAC3E,IAAI,kBAAkB,CAAC,OAAO,EAAE,IAAI,EAAE,WAAW,CAAC,EAAE,CAAC;YACnD,UAAU,CAAC,GAAG,CAAC,IAAI,EAAE,WAAW,CAAC,CAAC;YAClC,OAAO,CAAC,GAAG,CAAC,eAAe,QAAQ,cAAc,IAAI,EAAE,CAAC,CAAC;QAC3D,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,kEAAkE;IAClE,QAAQ,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,MAAkB,EAAE,EAAE;QAC9C,IAAI,CAAC,MAAM,CAAC,IAAI,IAAI,CAAC,MAAM,CAAC,KAAK;YAAE,OAAO;QAE1C,MAAM,WAAW,GACf,UAAU,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,GAAG,QAAQ,IAAI,MAAM,CAAC,IAAI,MAAM,CAAC,CAAC;QACxF,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,WAAW,CAAC;YAAE,OAAO;QAExC,MAAM,SAAS,GAAG,aAAa,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;QAC9C,MAAM,eAAe,GAAG,GAAG,QAAQ,QAAQ,SAAS,MAAM,CAAC;QAC3D,MAAM,WAAW,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,eAAe,CAAC,CAAC;QAE1D,IAAI,UAAU,CAAC,WAAW,EAAE,WAAW,EAAE,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC;YACtD,cAAc,CAAC,GAAG,CAAC,MAAM,CAAC,KAAK,EAAE,YAAY,eAAe,EAAE,CAAC,CAAC;YAChE,OAAO,CAAC,GAAG,CAAC,cAAc,MAAM,CAAC,KAAK,MAAM,eAAe,EAAE,CAAC,CAAC;QACjE,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,gCAAgC;IAChC,IAAI,EAAE,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;QAC9B,EAAE,CAAC,WAAW,CAAC,UAAU,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACnF,EAAE,CAAC,SAAS,CAAC,UAAU,CAAC,CAAC;IAC3B,CAAC;IAED,OAAO,cAAc,CAAC;AACxB,CAAC,CAAC;AAEF,MAAM,eAAe,GAAG,KAAK,EAC3B,QAAwB,EACxB,MAAqB,EACrB,cAAwB,EACxB,IAAmB,EACQ,EAAE;IAC7B,OAAO,CAAC,GAAG,CAAC,2BAA2B,QAAQ,gBAAgB,CAAC,CAAC;IAEjE,MAAM,MAAM,GAAG,2BAA2B,CAAC;QACzC,SAAS,EAAE,MAAM,CAAC,MAAM;QACxB,cAAc;QACd,IAAI;KACL,CAAC,CAAC;IAEH,MAAM,QAAQ,GAAG,MAAM,aAAa,CAAC,QAAQ,EAAE,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC,CAAC;IACnE,OAAO,qBAAqB,CAAC,QAAQ,CAAC,CAAC;AACzC,CAAC,CAAC;AAEF,MAAM,kBAAkB,GAAG,KAAK,EAC9B,QAAwB,EACxB,QAA0B,EAC1B,cAAwB,EACxB,IAAmB,EACA,EAAE;IACrB,OAAO,CAAC,GAAG,CAAC,uCAAuC,CAAC,CAAC;IAErD,MAAM,MAAM,GAAG,oBAAoB,CAAC;QAClC,gBAAgB,EAAE,QAAQ;QAC1B,cAAc;QACd,IAAI;KACL,CAAC,CAAC;IAEH,MAAM,QAAQ,GAAG,MAAM,WAAW,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;IACrD,MAAM,OAAO,GAAG,sBAAsB,CAAC,QAAQ,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;IACzE,OAAO,OAAO,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;AACpC,CAAC,CAAC;AAEF,MAAM,gBAAgB,GAAG,CACvB,QAA0B,EAC1B,UAAoB,EACpB,QAAgB,EAChB,IAAmB,EACnB,cAAmC,EACC,EAAE;IACtC,MAAM,KAAK,GAAG,QAAQ,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE;QAC7C,MAAM,SAAS,GAAG,KAAK,CAAC,SAAS,IAAI,KAAK,CAAC,WAAW,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;QAC/D,MAAM,aAAa,GAAG,YAAY,QAAQ,IAAI,SAAS,MAAM,CAAC;QAC9D,MAAM,SAAS,GACb,KAAK,CAAC,SAAS,IAAI,cAAc,CAAC,GAAG,CAAC,KAAK,CAAC,SAAS,CAAC;YACpD,CAAC,CAAC,cAAc,CAAC,GAAG,CAAC,KAAK,CAAC,SAAS,CAAE;YACtC,CAAC,CAAC,aAAa,CAAC;QAEpB,OAAO;YACL,IAAI,EAAE,UAAU,CAAC,CAAC,CAAC,IAAI,EAAE;YACzB,KAAK,EAAE;gBACL,IAAI,EAAE,OAAgB;gBACtB,MAAM,EAAE;oBACN,IAAI,EAAE,MAAe;oBACrB,IAAI,EAAE,SAAS;iBAChB;aACF;SACF,CAAC;IACJ,CAAC,CAAC,CAAC;IAEH,MAAM,WAAW,GAAqB;QACpC,UAAU,EAAE,EAAE,OAAO,EAAE,KAAK,EAAE;QAC9B,IAAI;QACJ,KAAK;KACN,CAAC;IAEF,MAAM,MAAM,GAAG,iBAAiB,CAAC,SAAS,CAAC,WAAW,CAAC,CAAC;IACxD,IAAI,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC;QACpB,OAAO,CAAC,KAAK,CAAC,gCAAgC,CAAC,CAAC;QAChD,OAAO,CAAC,KAAK,CAAC,MAAM,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC;QACrC,MAAM,IAAI,KAAK,CAAC,+BAA+B,CAAC,CAAC;IACnD,CAAC;IAED,OAAO,MAAM,CAAC,IAAI,CAAC;AACrB,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,gBAAgB,GAAG,KAAK,EACnC,OAAgC,EACC,EAAE;IACnC,MAAM,EAAE,SAAS,EAAE,QAAQ,EAAE,WAAW,EAAE,GAAG,OAAO,CAAC;IACrD,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC,SAAS,CAAC,CAAC;IAExC,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;QAC5B,MAAM,IAAI,KAAK,CAAC,mBAAmB,OAAO,EAAE,CAAC,CAAC;IAChD,CAAC;IAED,iBAAiB,CAAC,KAAK,CAAC,CAAC;IAEzB,MAAM,QAAQ,GAAG,qBAAqB,CAAC,WAAW,CAAC,CAAC;IACpD,OAAO,CAAC,GAAG,CAAC,0BAA0B,QAAQ,EAAE,CAAC,CAAC;IAElD,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;IAChD,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,QAAQ,CAAC,CAAC;IACjD,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,QAAQ,CAAC,CAAC;IAEjD,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,SAAS,CAAC,EAAE,CAAC;QAC9B,EAAE,CAAC,SAAS,CAAC,SAAS,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAC/C,CAAC;IAED,qCAAqC;IACrC,OAAO,CAAC,GAAG,CAAC,6BAA6B,CAAC,CAAC;IAC3C,MAAM,EAAE,UAAU,EAAE,SAAS,EAAE,GAAG,kBAAkB,CAAC;QACnD,OAAO,EAAE,OAAO;QAChB,SAAS;QACT,QAAQ;KACT,CAAC,CAAC;IAEH,uBAAuB;IACvB,OAAO,CAAC,GAAG,CAAC,6BAA6B,CAAC,CAAC;IAC3C,MAAM,SAAS,GAAG,MAAM,kBAAkB,CAAC,OAAO,CAAC,CAAC;IACpD,MAAM,cAAc,GAAa,EAAE,CAAC;IACpC,SAAS,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,EAAE;QACzB,cAAc,CAAC,IAAI,CAAC,UAAU,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC;IAC9C,CAAC,CAAC,CAAC;IACH,OAAO,CAAC,GAAG,CAAC,uBAAuB,SAAS,CAAC,MAAM,QAAQ,CAAC,CAAC;IAE7D,MAAM,YAAY,GAAG,WAAW,CAAC,OAAO,CAAC,IAAI,EAAE,cAAc,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC;IAE/E,uBAAuB;IACvB,MAAM,gBAAgB,GAAG,cAAc,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACvE,IAAI,kBAAkB,GAAkB,IAAI,CAAC;IAC7C,IAAI,gBAAgB,EAAE,CAAC;QACrB,kBAAkB,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,sBAAsB,CAAC,CAAC;QAClE,EAAE,CAAC,aAAa,CAAC,kBAAkB,EAAE,IAAI,CAAC,SAAS,CAAC,cAAc,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;IAChF,CAAC;IAED,qDAAqD;IACrD,MAAM,MAAM,GAAG,eAAe,CAAC,SAAS,EAAE,QAAQ,EAAE,SAAS,CAAC,CAAC;IAC/D,MAAM,QAAQ,GAAG,MAAM,eAAe,CAAC,QAAQ,EAAE,MAAM,EAAE,cAAc,EAAE,YAAY,CAAC,CAAC;IAEvF,gBAAgB;IAChB,MAAM,YAAY,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,eAAe,CAAC,CAAC;IAC3D,EAAE,CAAC,aAAa,CAAC,YAAY,EAAE,IAAI,CAAC,SAAS,CAAC,QAAQ,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC;IAClE,OAAO,CAAC,GAAG,CAAC,4BAA4B,YAAY,EAAE,CAAC,CAAC;IACxD,OAAO,CAAC,GAAG,CAAC,eAAe,QAAQ,CAAC,QAAQ,CAAC,MAAM,EAAE,CAAC,CAAC;IACvD,OAAO,CAAC,GAAG,CAAC,cAAc,QAAQ,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC;IACrD,OAAO,CAAC,GAAG,CAAC,qBAAqB,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;IAE3D,iDAAiD;IACjD,OAAO,CAAC,GAAG,CAAC,sCAAsC,CAAC,CAAC;IACpD,MAAM,cAAc,GAAG,WAAW,CAAC,QAAQ,EAAE,SAAS,EAAE,QAAQ,EAAE,OAAO,CAAC,CAAC;IAC3E,OAAO,CAAC,GAAG,CAAC,aAAa,cAAc,CAAC,IAAI,UAAU,CAAC,CAAC;IAExD,qDAAqD;IACrD,MAAM,UAAU,GAAG,MAAM,kBAAkB,CAAC,QAAQ,EAAE,QAAQ,EAAE,cAAc,EAAE,YAAY,CAAC,CAAC;IAE9F,sCAAsC;IACtC,MAAM,WAAW,GAAG,gBAAgB,CAClC,QAAQ,EACR,UAAU,EACV,QAAQ,EACR,YAAY,EACZ,cAAc,CACf,CAAC;IACF,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,GAAG,QAAQ,OAAO,CAAC,CAAC;IAC1D,gBAAgB,CAAC,WAAW,EAAE,QAAQ,CAAC,CAAC;IAExC,OAAO,CAAC,GAAG,CAAC,oCAAoC,CAAC,CAAC;IAClD,OAAO,CAAC,GAAG,CAAC,eAAe,QAAQ,EAAE,CAAC,CAAC;IACvC,OAAO,CAAC,GAAG,CAAC,YAAY,SAAS,cAAc,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,CAAC,CAAC;IACzE,OAAO,CAAC,GAAG,CAAC,aAAa,QAAQ,EAAE,CAAC,CAAC;IAErC,OAAO;QACL,eAAe,EAAE,QAAQ;QACzB,kBAAkB;QAClB,YAAY;QACZ,UAAU,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM;KACnC,CAAC;AACJ,CAAC,CAAC"}

package/lib/utils/document-analysis.d.ts

@@ -0,0 +1,43 @@
+import type { SupportedLang } from "./lang.js";
+export interface SectionInfo {
+    name: string;
+    pages: number[];
+    summary: string;
+}
+export interface BoundingBox {
+    x: number;
+    y: number;
+    width: number;
+    height: number;
+}
+export interface FigureInfo {
+    page: number;
+    type: "figure" | "table" | "chart" | "diagram";
+    label?: string;
+    description: string;
+    importance: "high" | "medium" | "low";
+    bbox?: BoundingBox;
+}
+export interface SlideSpec {
+    title: string;
+    section: string;
+    sourcePages: number[];
+    imagePage?: number;
+    figureRef?: string;
+    narrationHint: string;
+}
+export interface DocumentAnalysis {
+    title: string;
+    authors?: string;
+    sections: SectionInfo[];
+    figures: FigureInfo[];
+    slides: SlideSpec[];
+}
+export interface BuildAnalysisPromptOptions {
+    pageCount: number;
+    extractedTexts: string[];
+    lang: SupportedLang;
+}
+export declare const buildDocumentAnalysisPrompt: (options: BuildAnalysisPromptOptions) => string;
+export declare const parseDocumentAnalysis: (content: string) => DocumentAnalysis;
+//# sourceMappingURL=document-analysis.d.ts.map

package/lib/utils/document-analysis.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"document-analysis.d.ts","sourceRoot":"","sources":["../../src/utils/document-analysis.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,WAAW,CAAC;AAG/C,MAAM,WAAW,WAAW;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,MAAM,EAAE,CAAC;IAChB,OAAO,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,WAAW;IAC1B,CAAC,EAAE,MAAM,CAAC;IACV,CAAC,EAAE,MAAM,CAAC;IACV,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,UAAU;IACzB,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,QAAQ,GAAG,OAAO,GAAG,OAAO,GAAG,SAAS,CAAC;IAC/C,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,EAAE,MAAM,CAAC;IACpB,UAAU,EAAE,MAAM,GAAG,QAAQ,GAAG,KAAK,CAAC;IACtC,IAAI,CAAC,EAAE,WAAW,CAAC;CACpB;AAED,MAAM,WAAW,SAAS;IACxB,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,EAAE,MAAM,CAAC;IAChB,WAAW,EAAE,MAAM,EAAE,CAAC;IACtB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,aAAa,EAAE,MAAM,CAAC;CACvB;AAED,MAAM,WAAW,gBAAgB;IAC/B,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,OAAO,EAAE,UAAU,EAAE,CAAC;IACtB,MAAM,EAAE,SAAS,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,0BAA0B;IACzC,SAAS,EAAE,MAAM,CAAC;IAClB,cAAc,EAAE,MAAM,EAAE,CAAC;IACzB,IAAI,EAAE,aAAa,CAAC;CACrB;AAED,eAAO,MAAM,2BAA2B,GAAI,SAAS,0BAA0B,KAAG,MAwEjF,CAAC;AAEF,eAAO,MAAM,qBAAqB,GAAI,SAAS,MAAM,KAAG,gBA+CvD,CAAC"}

package/lib/utils/document-analysis.js

@@ -0,0 +1,118 @@
+import { getLanguageName, extractJsonFromResponse } from "./llm.js";
+export const buildDocumentAnalysisPrompt = (options) => {
+    const { pageCount, extractedTexts, lang } = options;
+    const languageName = getLanguageName(lang);
+    const textSummaries = extractedTexts
+        .map((text, i) => {
+        if (!text || text.trim().length === 0)
+            return `--- Page ${i} ---\n(no text)`;
+        const truncated = text.length > 2000 ? text.slice(0, 2000) + "..." : text;
+        return `--- Page ${i} ---\n${truncated}`;
+    })
+        .join("\n\n");
+    return `You are analyzing a PDF document to create an engaging presentation.
+
+The document has ${pageCount} pages. I'm showing you all pages as images and providing extracted text.
+
+Extracted text per page:
+${textSummaries}
+
+Analyze the document and create a presentation plan. Respond in JSON:
+
+{
+  "title": "document title",
+  "authors": "author names if identifiable",
+  "sections": [
+    {
+      "name": "section name",
+      "pages": [0, 1],
+      "summary": "brief section summary"
+    }
+  ],
+  "figures": [
+    {
+      "page": 0,
+      "type": "figure|table|chart|diagram",
+      "label": "Figure 1",
+      "description": "what the figure shows",
+      "importance": "high|medium|low",
+      "bbox": {"x": 10, "y": 30, "width": 80, "height": 40}
+    }
+  ],
+  "slides": [
+    {
+      "title": "slide title in ${languageName}",
+      "section": "section name",
+      "sourcePages": [0, 1],
+      "imagePage": 0,
+      "figureRef": "Figure 1",
+      "narrationHint": "key points to explain in this slide"
+    }
+  ]
+}
+
+Guidelines:
+- "sections": identify the logical structure of the document (intro, main sections, conclusion, etc.)
+- "figures": identify ALL figures, tables, charts, and diagrams. Mark important ones as "high"
+  - "bbox": bounding box as percentage of page dimensions (0-100). x = left edge %, y = top edge %, width and height in %.
+    IMPORTANT: err on the side of LARGER bounding boxes. Add 3-5% extra margin on all sides. It is much better to include a bit of surrounding whitespace than to cut off any part of the figure, its axis labels, legends, title, or caption.
+    Include the full figure with ALL labels, axis text, legends, and captions.
+    For a figure in the lower-left quadrant: {"x": 2, "y": 50, "width": 50, "height": 48}
+- "slides": create a presentation that explains the document to an audience
+  - NOT 1:1 with pages. Group related content, split dense pages
+  - Each important figure (high importance) should get its own slide
+  - "imagePage": which page image to show for this slide (0-based)
+  - "title": write in ${languageName}
+  - "narrationHint": describe what the presenter should explain (in English for clarity)
+  - Typical slide count: 8-15 slides for a 10-20 page document
+  - Include an introduction slide and a conclusion/summary slide
+- Skip appendix/reference pages unless they contain critical content
+- "figureRef": reference a figure label from the figures array when the slide focuses on that figure
+
+Respond ONLY with valid JSON.`;
+};
+export const parseDocumentAnalysis = (content) => {
+    const jsonStr = extractJsonFromResponse(content);
+    const parsed = JSON.parse(jsonStr);
+    const analysis = {
+        title: parsed.title ?? "Untitled",
+        authors: parsed.authors,
+        sections: (parsed.sections ?? []).map((s) => ({
+            name: String(s.name ?? ""),
+            pages: Array.isArray(s.pages) ? s.pages.map(Number) : [],
+            summary: String(s.summary ?? ""),
+        })),
+        figures: (parsed.figures ?? []).map((f) => {
+            const bbox = f.bbox;
+            const parsedBbox = bbox && bbox.x != null && bbox.y != null && bbox.width != null && bbox.height != null
+                ? {
+                    x: Number(bbox.x),
+                    y: Number(bbox.y),
+                    width: Number(bbox.width),
+                    height: Number(bbox.height),
+                }
+                : undefined;
+            return {
+                page: Number(f.page ?? 0),
+                type: String(f.type ?? "figure"),
+                label: f.label ? String(f.label) : undefined,
+                description: String(f.description ?? ""),
+                importance: String(f.importance ?? "medium"),
+                bbox: parsedBbox,
+            };
+        }),
+        slides: (parsed.slides ?? []).map((s) => ({
+            title: String(s.title ?? ""),
+            section: String(s.section ?? ""),
+            sourcePages: Array.isArray(s.sourcePages) ? s.sourcePages.map(Number) : [],
+            imagePage: s.imagePage != null ? Number(s.imagePage) : undefined,
+            figureRef: s.figureRef ? String(s.figureRef) : undefined,
+            narrationHint: String(s.narrationHint ?? ""),
+        })),
+    };
+    if (analysis.slides.length === 0) {
+        throw new Error("DocumentAnalysis has no slides");
+    }
+    return analysis;
+};
+//# sourceMappingURL=document-analysis.js.map

package/lib/utils/document-analysis.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"document-analysis.js","sourceRoot":"","sources":["../../src/utils/document-analysis.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,eAAe,EAAE,uBAAuB,EAAE,MAAM,UAAU,CAAC;AA+CpE,MAAM,CAAC,MAAM,2BAA2B,GAAG,CAAC,OAAmC,EAAU,EAAE;IACzF,MAAM,EAAE,SAAS,EAAE,cAAc,EAAE,IAAI,EAAE,GAAG,OAAO,CAAC;IACpD,MAAM,YAAY,GAAG,eAAe,CAAC,IAAI,CAAC,CAAC;IAE3C,MAAM,aAAa,GAAG,cAAc;SACjC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,EAAE,EAAE;QACf,IAAI,CAAC,IAAI,IAAI,IAAI,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC;YAAE,OAAO,YAAY,CAAC,iBAAiB,CAAC;QAC7E,MAAM,SAAS,GAAG,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,CAAC;QAC1E,OAAO,YAAY,CAAC,SAAS,SAAS,EAAE,CAAC;IAC3C,CAAC,CAAC;SACD,IAAI,CAAC,MAAM,CAAC,CAAC;IAEhB,OAAO;;mBAEU,SAAS;;;EAG1B,aAAa;;;;;;;;;;;;;;;;;;;;;;;;;;iCA0BkB,YAAY;;;;;;;;;;;;;;;;;;;;;wBAqBrB,YAAY;;;;;;;8BAON,CAAC;AAC/B,CAAC,CAAC;AAEF,MAAM,CAAC,MAAM,qBAAqB,GAAG,CAAC,OAAe,EAAoB,EAAE;IACzE,MAAM,OAAO,GAAG,uBAAuB,CAAC,OAAO,CAAC,CAAC;IACjD,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IAEnC,MAAM,QAAQ,GAAqB;QACjC,KAAK,EAAE,MAAM,CAAC,KAAK,IAAI,UAAU;QACjC,OAAO,EAAE,MAAM,CAAC,OAAO;QACvB,QAAQ,EAAE,CAAC,MAAM,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAA0B,EAAE,EAAE,CAAC,CAAC;YACrE,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC;YAC1B,KAAK,EAAE,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,EAAE;YACxD,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC,OAAO,IAAI,EAAE,CAAC;SACjC,CAAC,CAAC;QACH,OAAO,EAAE,CAAC,MAAM,CAAC,OAAO,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAA0B,EAAE,EAAE;YACjE,MAAM,IAAI,GAAG,CAAC,CAAC,IAA2C,CAAC;YAC3D,MAAM,UAAU,GACd,IAAI,IAAI,IAAI,CAAC,CAAC,IAAI,IAAI,IAAI,IAAI,CAAC,CAAC,IAAI,IAAI,IAAI,IAAI,CAAC,KAAK,IAAI,IAAI,IAAI,IAAI,CAAC,MAAM,IAAI,IAAI;gBACnF,CAAC,CAAC;oBACE,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC;oBACjB,CAAC,EAAE,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC;oBACjB,KAAK,EAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC;oBACzB,MAAM,EAAE,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC;iBAC5B;gBACH,CAAC,CAAC,SAAS,CAAC;YAChB,OAAO;gBACL,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC,CAAC;gBACzB,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,IAAI,IAAI,QAAQ,CAAuB;gBACtD,KAAK,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,SAAS;gBAC5C,WAAW,EAAE,MAAM,CAAC,CAAC,CAAC,WAAW,IAAI,EAAE,CAAC;gBACxC,UAAU,EAAE,MAAM,CAAC,CAAC,CAAC,UAAU,IAAI,QAAQ,CAA6B;gBACxE,IAAI,EAAE,UAAU;aACjB,CAAC;QACJ,CAAC,CAAC;QACF,MAAM,EAAE,CAAC,MAAM,CAAC,MAAM,IAAI,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAA0B,EAAE,EAAE,CAAC,CAAC;YACjE,KAAK,EAAE,MAAM,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YAC5B,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC,OAAO,IAAI,EAAE,CAAC;YAChC,WAAW,EAAE,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,EAAE;YAC1E,SAAS,EAAE,CAAC,CAAC,SAAS,IAAI,IAAI,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,SAAS;YAChE,SAAS,EAAE,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,SAAS;YACxD,aAAa,EAAE,MAAM,CAAC,CAAC,CAAC,aAAa,IAAI,EAAE,CAAC;SAC7C,CAAC,CAAC;KACJ,CAAC;IAEF,IAAI,QAAQ,CAAC,MAAM,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACjC,MAAM,IAAI,KAAK,CAAC,gCAAgC,CAAC,CAAC;IACpD,CAAC;IAED,OAAO,QAAQ,CAAC;AAClB,CAAC,CAAC"}

package/lib/utils/markdown-parser.d.ts

@@ -0,0 +1,28 @@
+/**
+ * Markdown Structure Parser
+ *
+ * Parses markdown into a structured representation for LLM-based
+ * presentation planning. Unlike the slide-splitting markdown plugins,
+ * this parser preserves the full document structure (heading hierarchy,
+ * element types) for intelligent beat allocation by an LLM.
+ */
+export interface MarkdownElement {
+    type: "text" | "table" | "mermaid" | "codeBlock" | "citation" | "image" | "list";
+    content: string;
+    lang?: string;
+    url?: string;
+    alt?: string;
+}
+export interface MarkdownSection {
+    id: string;
+    heading: string;
+    level: number;
+    elements: MarkdownElement[];
+    children: string[];
+}
+export interface ParsedMarkdown {
+    frontmatter: Record<string, string> | null;
+    sections: MarkdownSection[];
+}
+export declare const parseMarkdown: (markdown: string) => ParsedMarkdown;
+//# sourceMappingURL=markdown-parser.d.ts.map

package/lib/utils/markdown-parser.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"markdown-parser.d.ts","sourceRoot":"","sources":["../../src/utils/markdown-parser.ts"],"names":[],"mappings":"AAAA;;;;;;;GAOG;AAEH,MAAM,WAAW,eAAe;IAC9B,IAAI,EAAE,MAAM,GAAG,OAAO,GAAG,SAAS,GAAG,WAAW,GAAG,UAAU,GAAG,OAAO,GAAG,MAAM,CAAC;IACjF,OAAO,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,GAAG,CAAC,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,eAAe;IAC9B,EAAE,EAAE,MAAM,CAAC;IACX,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,QAAQ,EAAE,eAAe,EAAE,CAAC;IAC5B,QAAQ,EAAE,MAAM,EAAE,CAAC;CACpB;AAED,MAAM,WAAW,cAAc;IAC7B,WAAW,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,GAAG,IAAI,CAAC;IAC3C,QAAQ,EAAE,eAAe,EAAE,CAAC;CAC7B;AAwOD,eAAO,MAAM,aAAa,GAAI,UAAU,MAAM,KAAG,cAMhD,CAAC"}