officeparser 7.2.0 → 7.2.2

package/dist/cli.js

@@ -5,24 +5,26 @@
  *
  * Allows running officeparser from the command line:
  *   npx officeparser file.docx
- *   officeparser file.docx --toText=true
- *   officeparser file.docx --ocr=true --extractAttachments=true
+ *   officeparser file.docx --to=text
+ *   officeparser file.docx --ocr --extractAttachments
  *
- * Options (--key=value):
- *   --format=json|text|md|html|csv|rtf|pdf|chunks  Convert AST to specified format
+ * Options (--key=value, --key value, or bare flags):
+ *   --to=json|text|md|html|csv|rtf|pdf|chunks  Convert AST to specified format (default: json)
  *   --output=path             Save result to a file
- *   --toText=true             Legacy flag for plain text output
- *   --ocr=true                Enable OCR for images
- *   --ocrLanguage=eng         OCR language (default: eng)
- *   --extractAttachments=true Extract embedded attachments
- *   --ignoreNotes=true        Ignore footnotes/endnotes
- *   --ignoreComments=true     Ignore inline comments
- *   --ignoreHeadersAndFooters=true Ignore headers and footers
- *   --ignoreSlideMasters=true Ignore slide masters
- *   --ignoreInternalLinks=true Ignore internal links
- *   --putNotesAtLast=true     Move notes to end of document
- *   --includeRawContent=true  Include raw content in AST
- *   --outputErrorToConsole=true  Log errors to console
+ *   --fileType=docx|xlsx|...  Override file type detection
+ *   --ocr                     Enable OCR for images (default: false)
+ *   --ocrConfig.language=eng  OCR language (default: eng)
+ *   --extractAttachments      Extract embedded attachments (default: false)
+ *   --ignoreNotes             Ignore footnotes/endnotes/speaker notes (default: false)
+ *   --ignoreComments          Ignore inline comments (default: false)
+ *   --ignoreHeadersAndFooters Ignore headers and footers (default: false)
+ *   --ignoreSlideMasters      Ignore slide masters (default: false)
+ *   --ignoreInternalLinks     Ignore internal links (default: false)
+ *   --includeRawContent       Include raw content in AST (default: false)
+ *   --serializeRawContent     Include stringified XML in metadata (default: true)
+ *   --preserveXmlWhitespace   Keep raw formatting space (default: false)
+ *   --includeBreakNodes       Include break nodes (DOCX only, default: false)
+ *   --verbose                 Show full error stack traces and warning logs
  */
 var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
     if (k2 === undefined) k2 = k;
@@ -63,69 +65,219 @@ const OfficeGenerator_js_1 = require("./OfficeGenerator.js");
 const fs = __importStar(require("fs"));
 const args = process.argv.slice(2);
 let fileArg;
-let toText = false;
+let showHelp = false;
+let toFlagOption;
+let formatFlagOption;
+let toTextOption;
 let verbose = false;
-let outputFormat;
 let outputFile;
-const configArgs = [];
-function isConfigOption(arg) {
-    return arg.startsWith('--') && arg.includes('=');
-}
-args.forEach(arg => {
-    if (isConfigOption(arg)) {
-        configArgs.push(arg);
-    }
-    else if (!fileArg) {
-        fileArg = arg;
+// Parser and Generator configuration objects that will be populated by command line options.
+const config = {};
+const generatorConfig = {};
+// Known boolean configurations to validate user input against.
+const knownParserBooleans = new Set([
+    'ocr', 'extractAttachments', 'ignoreNotes', 'ignoreComments',
+    'ignoreHeadersAndFooters', 'ignoreSlideMasters', 'ignoreInternalLinks',
+    'includeRawContent', 'serializeRawContent', 'preserveXmlWhitespace', 'includeBreakNodes'
+]);
+const knownGeneratorBooleans = new Set([
+    'includeFormatting', 'generateIds', 'renderMetadata', 'includeImages', 'includeCharts', 'ignoreInternalLinks'
+]);
+// Prefixes used to identify configurations targeted for the generator instead of the parser.
+const generatorPrefixes = [
+    'generatorConfig.', 'htmlConfig.', 'csvConfig.', 'textConfig.', 'mdConfig.', 'pdfConfig.', 'rtfConfig.', 'chunksConfig.'
+];
+// Trackers to detect if deprecated/legacy options were used to log helpful warnings.
+let usedFormat = false;
+let usedToText = false;
+let usedOcrLanguage = false;
+let usedPutNotesAtLast = false;
+let usedOutputErrorToConsole = false;
+// Parse the arguments list
+for (let i = 0; i < args.length; i++) {
+    const arg = args[i];
+    // Help flags trigger immediate termination of parsing and print help
+    if (arg === '-h' || arg === '--help') {
+        showHelp = true;
+        break;
     }
-});
-if (fileArg) {
-    const config = {};
-    configArgs.forEach(arg => {
-        const [key, value] = arg.split('=');
-        const cleanKey = key.replace('--', '');
-        const lowerValue = value.toLowerCase();
+    if (arg.startsWith('--')) {
+        let cleanKey;
+        let val;
+        // Support --key=value syntax
+        if (arg.includes('=')) {
+            const idx = arg.indexOf('=');
+            cleanKey = arg.slice(2, idx);
+            val = arg.slice(idx + 1);
+        }
+        // Support negation shorthand: --no-ocr sets ocr to false
+        else if (arg.startsWith('--no-')) {
+            cleanKey = arg.slice(5);
+            val = 'false';
+        }
+        // Support space-separated options or bare flags
+        else {
+            cleanKey = arg.slice(2);
+            const isKnownBoolean = knownParserBooleans.has(cleanKey) ||
+                knownGeneratorBooleans.has(cleanKey) ||
+                cleanKey === 'verbose' ||
+                cleanKey === 'toText' ||
+                cleanKey === 'outputErrorToConsole';
+            const isNextBool = i + 1 < args.length &&
+                (args[i + 1].toLowerCase() === 'true' || args[i + 1].toLowerCase() === 'false');
+            // If the next arg is not another option flag, treat it as the value (e.g., --to html)
+            // But if this is a known boolean, only consume the next argument if it is a valid boolean string.
+            if (i + 1 < args.length && !args[i + 1].startsWith('-') && (!isKnownBoolean || isNextBool)) {
+                val = args[i + 1];
+                i++;
+            }
+            // Bare presence implies true (e.g., --ocr is true)
+            else {
+                val = 'true';
+            }
+        }
+        // Parse boolean strings to raw boolean types
+        const lowerValue = val.toLowerCase();
         const boolValue = lowerValue === 'true' ? true : (lowerValue === 'false' ? false : undefined);
-        const knownBooleans = new Set([
-            'toText', 'ocr', 'extractAttachments', 'ignoreNotes', 'ignoreComments',
-            'ignoreHeadersAndFooters', 'ignoreSlideMasters', 'ignoreInternalLinks',
-            'putNotesAtLast', 'includeRawContent', 'outputErrorToConsole',
-            'serializeRawContent', 'preserveXmlWhitespace', 'includeBreakNodes', 'verbose'
-        ]);
-        if (cleanKey === 'format') {
-            outputFormat = value;
+        // Map core CLI options to variables
+        if (cleanKey === 'to') {
+            toFlagOption = val;
+        }
+        else if (cleanKey === 'format') {
+            formatFlagOption = val;
+            usedFormat = true;
         }
         else if (cleanKey === 'output') {
-            outputFile = value;
+            outputFile = val;
+        }
+        else if (cleanKey === 'toText') {
+            toTextOption = boolValue !== undefined ? boolValue : true;
+            usedToText = true;
+        }
+        else if (cleanKey === 'verbose') {
+            verbose = boolValue !== undefined ? boolValue : true;
+        }
+        else if (cleanKey === 'ocrLanguage') {
+            config.ocrLanguage = val;
+            usedOcrLanguage = true;
+        }
+        else if (cleanKey === 'putNotesAtLast') {
+            config.putNotesAtLast = boolValue !== undefined ? boolValue : true;
+            usedPutNotesAtLast = true;
+        }
+        else if (cleanKey === 'outputErrorToConsole') {
+            verbose = boolValue !== undefined ? boolValue : true;
+            usedOutputErrorToConsole = true;
         }
         else {
-            if (boolValue !== undefined) {
-                if (cleanKey === 'toText')
-                    toText = boolValue;
-                else if (cleanKey === 'verbose') {
-                    verbose = boolValue;
-                    if (verbose)
-                        config.outputErrorToConsole = true;
+            // Check if the flag belongs to generatorConfig or a specific sub-generator (e.g., htmlConfig)
+            const isGeneratorOption = knownGeneratorBooleans.has(cleanKey) || generatorPrefixes.some(pref => cleanKey.startsWith(pref));
+            const target = isGeneratorOption ? generatorConfig : config;
+            let path = cleanKey;
+            // Strip generatorConfig prefix to flatten it onto the generatorConfig object
+            if (isGeneratorOption && cleanKey.startsWith('generatorConfig.')) {
+                path = cleanKey.slice('generatorConfig.'.length);
+            }
+            // Support nested dot-notation parsing (e.g., --ocrConfig.language=fra)
+            if (path.includes('.')) {
+                const parts = path.split('.');
+                let current = target;
+                for (let j = 0; j < parts.length - 1; j++) {
+                    const part = parts[j];
+                    if (!current[part])
+                        current[part] = {};
+                    current = current[part];
+                }
+                const lastPart = parts[parts.length - 1];
+                current[lastPart] = boolValue !== undefined ? boolValue : val;
+            }
+            // Flat key assignment
+            else {
+                if (boolValue !== undefined) {
+                    target[path] = boolValue;
+                }
+                else if (knownParserBooleans.has(path) || (isGeneratorOption && knownGeneratorBooleans.has(path))) {
+                    console.warn(`Invalid boolean value for --${cleanKey}: ${val}. Using default.`);
                 }
                 else {
-                    // @ts-ignore
-                    config[cleanKey] = boolValue;
+                    target[path] = val;
                 }
             }
-            else if (knownBooleans.has(cleanKey)) {
-                console.warn(`Invalid boolean value for --${cleanKey}: ${value}. Using default.`);
-            }
-            else {
-                // @ts-ignore
-                config[cleanKey] = value;
+        }
+    }
+    else {
+        // First positional argument that is not a flag is treated as the input file path
+        if (!fileArg) {
+            fileArg = arg;
+        }
+    }
+}
+if (fileArg && !showHelp) {
+    // Resolve output format prioritizing: --to > --format > --toText
+    let outputFormat;
+    if (toFlagOption) {
+        outputFormat = toFlagOption;
+    }
+    else if (formatFlagOption) {
+        outputFormat = formatFlagOption;
+    }
+    else if (toTextOption === true) {
+        outputFormat = 'text';
+    }
+    // Display warning messages for any deprecated CLI options used
+    if (usedFormat) {
+        console.warn('Warning: --format is deprecated. Use --to instead.');
+    }
+    if (usedToText) {
+        console.warn('Warning: --toText is deprecated. Use --to=text instead.');
+    }
+    if (usedOcrLanguage) {
+        console.warn('Warning: --ocrLanguage is deprecated. Use --ocrConfig.language instead.');
+    }
+    if (usedPutNotesAtLast) {
+        console.warn('Warning: --putNotesAtLast is deprecated and will be ignored by all parsers.');
+    }
+    if (usedOutputErrorToConsole) {
+        console.warn('Warning: --outputErrorToConsole is deprecated. Use --verbose instead.');
+    }
+    // Intercept parser warning callbacks to format and print issues when verbose is enabled
+    const originalOnWarning = config.onWarning;
+    config.onWarning = (issue) => {
+        if (verbose) {
+            const severity = issue.type === 'error' ? 'Error' : 'Warning';
+            console.error(`[OfficeParser ${severity}] [${issue.code}]: ${issue.message}`);
+            if (issue.details) {
+                console.error(issue.details);
             }
         }
-    });
+        if (originalOnWarning)
+            originalOnWarning(issue);
+    };
+    // Propagate newlineDelimiter and csvDelimiter if configured flatly but not in generator sub-configs
+    if (config.newlineDelimiter !== undefined) {
+        if (!generatorConfig.textConfig)
+            generatorConfig.textConfig = {};
+        if (generatorConfig.textConfig.newlineDelimiter === undefined) {
+            generatorConfig.textConfig.newlineDelimiter = config.newlineDelimiter;
+        }
+    }
+    if (config.csvDelimiter !== undefined) {
+        if (!generatorConfig.csvConfig)
+            generatorConfig.csvConfig = {};
+        if (generatorConfig.csvConfig.columnDelimiter === undefined) {
+            generatorConfig.csvConfig.columnDelimiter = config.csvDelimiter;
+        }
+    }
+    // Run the main parser
     OfficeParser_js_1.OfficeParser.parseOffice(fileArg, config)
         .then(async (ast) => {
         let output;
-        if (outputFormat) {
-            const result = await OfficeGenerator_js_1.OfficeGenerator.generate(ast, outputFormat);
+        // Generate JSON output or convert AST using OfficeGenerator
+        if (outputFormat === 'json') {
+            output = JSON.stringify(ast, null, 2);
+        }
+        else if (outputFormat) {
+            const result = await OfficeGenerator_js_1.OfficeGenerator.generate(ast, outputFormat, generatorConfig);
             if (Array.isArray(result.value)) {
                 output = JSON.stringify(result.value, null, 2);
             }
@@ -133,12 +285,10 @@ if (fileArg) {
                 output = result.value;
             }
         }
-        else if (toText) {
-            output = ast.toText();
-        }
         else {
             output = JSON.stringify(ast, null, 2);
         }
+        // Write generated output to output file or print to standard output
         if (outputFile) {
             if (output instanceof Uint8Array) {
                 fs.writeFileSync(outputFile, output);
@@ -163,15 +313,16 @@ if (fileArg) {
         }
     })
         .catch(async (err) => {
+        // Handle and display parsing error messages
         console.error(`Error parsing file "${fileArg}":`);
         if (verbose) {
             console.error(err);
         }
         else {
             console.error(err.message || err);
-            console.error('Use --verbose=true for full stack trace.');
+            console.error('Use --verbose for full stack trace.');
         }
-        // Ensure OCR workers are terminated even on error
+        // Ensure OCR workers are terminated even on error to prevent process hang
         if (config.ocr) {
             await OfficeParser_js_1.OfficeParser.terminateOcr();
         }
@@ -179,32 +330,46 @@ if (fileArg) {
     });
 }
 else {
-    console.log('Usage: officeparser <file> [--option=value]');
+    console.log('Usage: officeparser <file> [options]');
     console.log('');
     console.log('Options:');
-    console.log('  --format=json|md|html|rtf|csv|text|pdf|chunks  Convert to specified format');
-    console.log('  --output=file.ext            Save output to file instead of stdout');
-    console.log('  --toText=true                Output plain text instead of JSON AST (legacy)');
-    console.log('  --ocr=true                   Enable OCR for images');
-    console.log('  --ocrLanguage=eng            OCR language (default: eng)');
-    console.log('  --extractAttachments=true    Extract embedded attachments');
-    console.log('  --ignoreNotes=true           Ignore footnotes/endnotes');
-    console.log('  --ignoreComments=true        Ignore inline comments');
-    console.log('  --ignoreHeadersAndFooters=true Ignore headers and footers');
-    console.log('  --ignoreSlideMasters=true    Ignore slide masters');
-    console.log('  --ignoreInternalLinks=true   Ignore internal links');
-    console.log('  --putNotesAtLast=true        Move notes to end of document');
-    console.log('  --includeRawContent=true     Include raw content in AST');
-    console.log('  --serializeRawContent=true   Serialize raw XML content (default: true)');
-    console.log('  --preserveXmlWhitespace=true Preserve whitespace in serialized XML (default: false)');
-    console.log('  --includeBreakNodes=false    Include break nodes (DOCX only, default: false)');
-    console.log('  --verbose=true               Show full error stack traces');
+    console.log('  --to=json|text|md|html|pdf|csv|rtf|chunks   Target conversion format (default: json)');
+    console.log('  --output=file.ext                           Save output to file instead of stdout');
+    console.log('  --fileType=docx|xlsx|pptx|odt|...           Explicitly override input file type detection');
+    console.log('  --ocr                                       Enable OCR for images (default: false)');
+    console.log('  --ocrConfig.language=eng                    OCR language (default: eng)');
+    console.log('  --extractAttachments                        Extract embedded attachments (default: false)');
+    console.log('  --ignoreNotes                               Ignore footnotes/endnotes/speaker notes (default: false)');
+    console.log('  --ignoreComments                            Ignore inline comments (default: false)');
+    console.log('  --ignoreHeadersAndFooters                   Ignore headers and footers (default: false)');
+    console.log('  --ignoreSlideMasters                        Ignore slide masters (default: false)');
+    console.log('  --ignoreInternalLinks                       Ignore internal links (default: false)');
+    console.log('  --includeRawContent                         Include raw content in AST (default: false)');
+    console.log('  --serializeRawContent                       Serialize raw XML content (default: true)');
+    console.log('  --preserveXmlWhitespace                     Keep raw formatting space (default: false)');
+    console.log('  --includeBreakNodes                         Include break nodes (DOCX only, default: false)');
+    console.log('  --verbose                                   Show full error stack traces and warning logs');
+    console.log('  --newlineDelimiter=string                   Delimiter string between blocks/lines (default: \\n)');
+    console.log('  --csvDelimiter=char                         Custom CSV delimiter (default: ,)');
+    console.log('');
+    console.log('High-Value Generator Options:');
+    console.log('  --includeFormatting                         Include font formatting like bold/italic (default: true)');
+    console.log('  --renderMetadata                            Render metadata in output content (default: false)');
+    console.log('  --htmlConfig.containerWidth=value           HTML container width (auto | px | % | vw etc., default: auto)');
+    console.log('');
+    console.log('Advanced Nested Config Examples:');
+    console.log('  --pdfConfig.format=Letter                   Configure Puppeteer PDF format (A4 | Letter | Legal etc.)');
+    console.log('  --chunksConfig.strategy=fixed-size          Chunking strategy (fixed-size | document-structure | semantic)');
+    console.log('');
+    console.log('Format Syntax:');
+    console.log('  Flags can be written as --flag (presence implies true), --no-flag (negation),');
+    console.log('  --flag=value, or --flag value.');
     console.log('');
     console.log('Examples:');
     console.log('  officeparser document.docx');
-    console.log('  officeparser document.docx --format=html --output=doc.html');
-    console.log('  officeparser document.docx --format=md');
-    console.log('  officeparser report.pdf --ocr=true --format=text');
-    console.log('  officeparser data.xlsx --format=csv --output=data.csv');
-    console.log('  officeparser complex.docx --serializeRawContent=false --includeRawContent=true');
+    console.log('  officeparser document.docx --to html --output doc.html');
+    console.log('  officeparser document.docx --to md');
+    console.log('  officeparser report.pdf --ocr --ocrConfig.language eng --to text');
+    console.log('  officeparser data.xlsx --to csv --output data.csv --csvDelimiter ";"');
+    console.log('  officeparser image_doc --fileType docx --to json');
 }

package/dist/defaults.js

@@ -59,6 +59,10 @@ exports.DEFAULT_OFFICE_PARSER_CONFIG = {
     ignoreInternalLinks: false,
     fileType: null,
     csvDelimiter: ',',
+    decompressionLimits: {
+        maxUncompressedBytes: 512 * 1024 * 1024,
+        maxZipEntries: 10000,
+    },
 };
 /**
  * Default configuration for HTML generation.
@@ -119,7 +123,7 @@ const DEFAULT_MD_GENERATOR_CONFIG = {
  */
 const DEFAULT_TEXT_GENERATOR_CONFIG = {
     newlineDelimiter: '\n',
-    preserveLayout: false,
+    preserveLayout: true,
 };
 /**
  * Default configuration for Fixed-Size chunking.

package/dist/generators/BaseGenerator.d.ts

@@ -10,6 +10,7 @@ export declare abstract class BaseGenerator<D extends UniversalGeneratorFormat =
     protected ast: OfficeParserAST;
     protected messages: OfficeIssue[];
     protected styleMapper: StyleMapper;
+    protected collectedNotes: OfficeContentNode[];
     constructor(destination: D, ast: OfficeParserAST, config?: GeneratorConfig<D> | FullGeneratorConfig);
     /**
      * Retrieves the semantic mapping for a node, respecting the includeFormatting flag.

package/dist/generators/BaseGenerator.js

@@ -14,6 +14,7 @@ class BaseGenerator {
     ast;
     messages = [];
     styleMapper;
+    collectedNotes = [];
     constructor(destination, ast, config) {
         this.destination = destination;
         this.config = (0, configUtils_js_1.resolveGeneratorConfig)(destination, ast.config, config);
@@ -65,7 +66,18 @@ class BaseGenerator {
                 childrenOutput += await this.processNodeRecursive(child, processor);
             }
         }
-        return await processor(node, childrenOutput);
+        if (node.notes && node.notes.length > 0) {
+            if (node.type !== 'slide') {
+                this.collectedNotes.push(...node.notes);
+            }
+        }
+        let result = await processor(node, childrenOutput);
+        if (node.type === 'slide' && node.notes && node.notes.length > 0) {
+            for (const note of node.notes) {
+                result += await this.processNodeRecursive(note, processor);
+            }
+        }
+        return result;
     }
     /**
      * Helper to generate a unique ID (slug) from text.

package/dist/generators/ChunkingGenerator.js

@@ -214,10 +214,17 @@ class ChunkingGenerator extends BaseGenerator_js_1.BaseGenerator {
             || splitBy === 'sheet' && node.type === 'sheet';
         if (isForcedSplit) {
             // Process children of the container as individual chunks within the boundary
-            if (node.children) {
+            if (node.children || node.notes) {
                 const innerChunks = [];
-                for (const child of node.children) {
-                    await this.processNodeForStructure(child, config, 'paragraph', maxChunkSize, measure, innerChunks, contextStack);
+                if (node.children) {
+                    for (const child of node.children) {
+                        await this.processNodeForStructure(child, config, 'paragraph', maxChunkSize, measure, innerChunks, contextStack);
+                    }
+                }
+                if (node.notes) {
+                    for (const note of node.notes) {
+                        await this.processNodeForStructure(note, config, 'paragraph', maxChunkSize, measure, innerChunks, contextStack);
+                    }
                 }
                 for (const ic of innerChunks) {
                     ic.metadata.slideNumber = contextStack.slideNumber;
@@ -268,12 +275,17 @@ class ChunkingGenerator extends BaseGenerator_js_1.BaseGenerator {
             }
             return;
         }
-        // Recurse into children for container nodes
+        // Recurse into children and notes for container nodes
         if (node.children) {
             for (const child of node.children) {
                 await this.processNodeForStructure(child, config, splitBy, maxChunkSize, measure, chunks, contextStack);
             }
         }
+        if (node.notes) {
+            for (const note of node.notes) {
+                await this.processNodeForStructure(note, config, splitBy, maxChunkSize, measure, chunks, contextStack);
+            }
+        }
     }
     isStructuralBoundary(node, splitBy) {
         if (splitBy === 'heading')
@@ -524,6 +536,10 @@ class ChunkingGenerator extends BaseGenerator_js_1.BaseGenerator {
                 for (const child of node.children)
                     await walk(child);
             }
+            if (node.notes) {
+                for (const note of node.notes)
+                    await walk(note);
+            }
         };
         for (const node of this.ast.content)
             await walk(node);
@@ -568,6 +584,10 @@ class ChunkingGenerator extends BaseGenerator_js_1.BaseGenerator {
                 for (const child of node.children)
                     await walk(child);
             }
+            if (node.notes) {
+                for (const note of node.notes)
+                    await walk(note);
+            }
         };
         for (const node of this.ast.content)
             await walk(node);
@@ -624,7 +644,7 @@ class ChunkingGenerator extends BaseGenerator_js_1.BaseGenerator {
                     let timerId;
                     const timeoutPromise = new Promise((_, reject) => {
                         timerId = setTimeout(() => {
-                            reject(new Error(`Embedding call timed out after ${timeoutMs}ms`));
+                            reject((0, errorUtils_js_1.getOfficeError)(types_js_1.OfficeErrorType.EMBEDDING_TIMEOUT, this.config, timeoutMs));
                         }, timeoutMs);
                     });
                     return Promise.race([call, timeoutPromise]).finally(() => {

package/dist/generators/HtmlGenerator.js

@@ -27,7 +27,14 @@ class HtmlGenerator extends BaseGenerator_js_1.BaseGenerator {
             containerClass = 'presentation-container';
         else if (isPdf)
             containerClass = 'pdf-container';
-        const bodyContent = await this.processNodeArray(this.ast.content);
+        let bodyContent = await this.processNodeArray(this.ast.content);
+        if (this.collectedNotes.length > 0) {
+            let notesHtml = '';
+            for (const note of this.collectedNotes) {
+                notesHtml += await this.processNodeRecursive(note, this.nodeProcessor.bind(this));
+            }
+            bodyContent += `\n<div class="document-notes-section">\n<hr class="page-break">\n${notesHtml}\n</div>\n`;
+        }
         const metadataBlock = this.config.renderMetadata ? this.renderMetadataSummary() : '';
         let title = 'Document';
         let metaTags = '';
@@ -385,9 +392,19 @@ class HtmlGenerator extends BaseGenerator_js_1.BaseGenerator {
             // Fallback for nodes that have text property but no children (e.g. simple paragraphs)
             childrenOutput = this.escape(node.text);
         }
-        const result = await processor(node, childrenOutput);
+        if (node.notes && node.notes.length > 0) {
+            if (node.type !== 'slide') {
+                this.collectedNotes.push(...node.notes);
+            }
+        }
+        let result = await processor(node, childrenOutput);
         if (isTable)
             this.tableNestingLevel--;
+        if (node.type === 'slide' && node.notes && node.notes.length > 0) {
+            for (const note of node.notes) {
+                result += await this.processNodeRecursive(note, processor);
+            }
+        }
         return result;
     }
     /**
@@ -452,7 +469,7 @@ class HtmlGenerator extends BaseGenerator_js_1.BaseGenerator {
                         src = `data:${attachment.mimeType || 'image/png'};base64,${attachment.data}`;
                     }
                 }
-                const img = `<img src="${src}" alt="${this.escape(node.text || meta?.altText || '')}"${className}${mappedAttrs}${styleAttr}>`;
+                const img = `<img src="${this.escape(src)}" alt="${this.escape(node.text || meta?.altText || '')}"${className}${mappedAttrs}${styleAttr}>`;
                 const content = this.config.includeFormatting ? `<div class="image-container">${img}<div class="caption">${this.escape(attachmentName || '')}</div></div>` : img;
                 return `${extraAnchors}<div${idAttr}>${content}</div>`;
             }

package/dist/generators/MarkdownGenerator.js

@@ -218,6 +218,11 @@ class MarkdownGenerator extends BaseGenerator_js_1.BaseGenerator {
                     const anchors = this.renderAnchors(node.metadata);
                     return `\n---\n\n${anchors}${anchors ? '\n' : ''}${childrenOutput}\n\n`;
                 }
+                case 'note': {
+                    const meta = node.metadata;
+                    const typeLabel = meta?.noteType === 'footnote' ? 'Footnote' : (meta?.noteType === 'endnote' ? 'Endnote' : 'Note');
+                    return `> **${typeLabel}:** ${childrenOutput.trim()}\n\n`;
+                }
                 default:
                     return childrenOutput;
             }
@@ -241,6 +246,13 @@ class MarkdownGenerator extends BaseGenerator_js_1.BaseGenerator {
             }
             output += result;
         }
+        if (this.collectedNotes.length > 0) {
+            let notesMd = '\n\n---\n\n### Notes\n\n';
+            for (const note of this.collectedNotes) {
+                notesMd += await this.processNodeRecursive(note, processor);
+            }
+            output += notesMd;
+        }
         return {
             value: (output + '\n\n' + this.hoistedContent.join('\n\n')).trim(),
             messages: this.messages
@@ -267,7 +279,18 @@ class MarkdownGenerator extends BaseGenerator_js_1.BaseGenerator {
                 childrenOutput += await this.processNodeRecursive(child, processor);
             }
         }
-        return await processor(node, childrenOutput);
+        if (node.notes && node.notes.length > 0) {
+            if (node.type !== 'slide') {
+                this.collectedNotes.push(...node.notes);
+            }
+        }
+        let result = await processor(node, childrenOutput);
+        if (node.type === 'slide' && node.notes && node.notes.length > 0) {
+            for (const note of node.notes) {
+                result += await this.processNodeRecursive(note, processor);
+            }
+        }
+        return result;
     }
     /**
      * Merges adjacent text nodes with identical formatting and metadata.
@@ -284,9 +307,17 @@ class MarkdownGenerator extends BaseGenerator_js_1.BaseGenerator {
                 current.text = (current.text || '') + (node.text || '');
                 if (current.rawContent && node.rawContent)
                     current.rawContent += node.rawContent;
+                if (node.notes && node.notes.length > 0) {
+                    if (!current.notes)
+                        current.notes = [];
+                    current.notes.push(...node.notes);
+                }
             }
             else {
                 current = { ...node }; // Clone
+                if (node.notes) {
+                    current.notes = [...node.notes];
+                }
                 result.push(current);
             }
         }