docx-diff-editor 1.0.41 → 1.0.43

package/README.md

@@ -6,8 +6,9 @@ A React component for DOCX document comparison with track changes visualization.
 
 - 📄 Compare two DOCX documents side by side
 - 🔍 Character-level diff with track changes
-- 📊 **Block-level diffing** for tables, lists, and images
-- ✅ Accept/reject individual changes
+- 📊 **Block-level structural diffing** for tables, lists, paragraphs, and images
+- 🔄 **Structure-aware merge** - inserted/deleted blocks appear in the editor with track marks
+- ✅ Accept/reject individual changes (both text and structural)
 - 🎨 Visual track changes (insert, delete, format)
 - 📋 **Structural Changes Pane** for table rows, list items, images
 - 🤖 Extract enriched change context for LLM processing
@@ -149,6 +150,9 @@ interface DocxDiffEditorRef {
 
   // Set document core properties (partial update)
   setProperties(properties: Partial<DocumentProperties>): Promise<boolean>;
+
+  // Parse HTML to ProseMirror JSON
+  parseHtml(html: string): Promise<ProseMirrorJSON>;
 }
 ```
 
@@ -251,6 +255,23 @@ await editorRef.current?.setProperties({
 });
 ```
 
+## Parsing HTML to JSON
+
+Convert HTML strings to ProseMirror JSON without visible rendering:
+
+```tsx
+// Using the ref method (requires editor to be initialized)
+const json = await editorRef.current?.parseHtml('<h1>Title</h1><p>Content here</p>');
+console.log(json); // { type: 'doc', content: [...] }
+
+// Use with other methods
+await editorRef.current?.updateContent(json);
+
+// Or use the standalone function (requires SuperDoc class)
+import { parseHtmlToJson } from 'docx-diff-editor';
+const json = await parseHtmlToJson(htmlString, SuperDocClass);
+```
+
 ## Customization
 
 ### CSS Variables
@@ -292,6 +313,18 @@ The component supports three types of track changes:
 
 When comparing documents with structural differences (tables, lists, images), a floating pane appears showing these changes with Accept/Reject controls.
 
+### How It Works
+
+The component uses a **structure-aware merge** approach:
+
+1. **Block alignment**: Documents are aligned at the block level (paragraphs, tables, lists) using content fingerprinting
+2. **Recursive merge**: Tables and lists are merged recursively (row-by-row, item-by-item)
+3. **Character-level diff**: Within matched blocks, character-level diffing is applied
+4. **Insert/delete marking**: New blocks get `trackInsert` marks; deleted blocks are preserved with `trackDelete` marks
+5. **Shared IDs**: Each structural change has a unique ID linking the track marks to the pane entry
+
+This means inserted tables, paragraphs, and list items actually appear in the editor (with green highlighting), and deleted content remains visible (with red strikethrough) until you accept or reject the changes.
+
 ### What's Detected
 
 | Change Type | Description |

package/dist/index.d.mts

@@ -362,6 +362,8 @@ interface DocxDiffEditorRef {
     getProperties(): Promise<DocumentProperties | null>;
     /** Set document core properties (partial update) */
     setProperties(properties: Partial<DocumentProperties>): Promise<boolean>;
+    /** Parse HTML string to ProseMirror JSON (uses hidden SuperDoc instance) */
+    parseHtml(html: string): Promise<ProseMirrorJSON>;
 }
 
 /**
@@ -420,6 +422,10 @@ declare function detectContentType(content: DocxContent): 'file' | 'html' | 'jso
  * Validate that content looks like ProseMirror JSON
  */
 declare function isProseMirrorJSON(content: unknown): boolean;
+/**
+ * Parse an HTML string into ProseMirror JSON using a hidden SuperDoc instance.
+ */
+declare function parseHtmlToJson(html: string, SuperDoc: SuperDocConstructor): Promise<ProseMirrorJSON>;
 /**
  * Parse a DOCX File into ProseMirror JSON using a hidden SuperDoc instance.
  */
@@ -710,4 +716,4 @@ declare function getBlankTemplateBlob(): Blob;
  */
 declare function isValidDocxFile(file: File): boolean;
 
-export { type AttrDiff, type AttributeChange, CSS_PREFIX, type ChangeLocation, type ComparisonResult, DEFAULT_AUTHOR, DEFAULT_SUPERDOC_USER, type DiffResult, type DiffSegment, type DocumentInfo, type DocumentProperties, type DocxContent, DocxDiffEditor, type DocxDiffEditorProps, type DocxDiffEditorRef, type EnrichedChange, type FingerprintedNode, type FormatChange, type FormatDetails, type HybridDiffResult, type NodeMatch, type ProseMirrorJSON, type ProseMirrorMark, type ProseMirrorNode, type StructuralChange, type StructuralChangeInfo, type StructuralChangeType, StructuralChangesPane, type StructuralPanePosition, type TrackChangeAuthor, alignDocuments, createTrackDeleteMark, createTrackFormatMark, createTrackInsertMark, DocxDiffEditor as default, detectContentType, diffDocuments, diffImages, diffLists, diffTables, extractEnrichedChanges, extractEnrichedChangesWithStructural, generateFingerprint, generateStructuralChangeSummary, getBlankTemplateBlob, getBlankTemplateFile, isAtomicNode, isImage, isList, isProseMirrorJSON, isTable, isValidDocxFile, mergeDocuments, parseDocxFile, processStructuralChanges };
+export { type AttrDiff, type AttributeChange, CSS_PREFIX, type ChangeLocation, type ComparisonResult, DEFAULT_AUTHOR, DEFAULT_SUPERDOC_USER, type DiffResult, type DiffSegment, type DocumentInfo, type DocumentProperties, type DocxContent, DocxDiffEditor, type DocxDiffEditorProps, type DocxDiffEditorRef, type EnrichedChange, type FingerprintedNode, type FormatChange, type FormatDetails, type HybridDiffResult, type NodeMatch, type ProseMirrorJSON, type ProseMirrorMark, type ProseMirrorNode, type StructuralChange, type StructuralChangeInfo, type StructuralChangeType, StructuralChangesPane, type StructuralPanePosition, type TrackChangeAuthor, alignDocuments, createTrackDeleteMark, createTrackFormatMark, createTrackInsertMark, DocxDiffEditor as default, detectContentType, diffDocuments, diffImages, diffLists, diffTables, extractEnrichedChanges, extractEnrichedChangesWithStructural, generateFingerprint, generateStructuralChangeSummary, getBlankTemplateBlob, getBlankTemplateFile, isAtomicNode, isImage, isList, isProseMirrorJSON, isTable, isValidDocxFile, mergeDocuments, parseDocxFile, parseHtmlToJson, processStructuralChanges };

package/dist/index.d.ts

@@ -362,6 +362,8 @@ interface DocxDiffEditorRef {
     getProperties(): Promise<DocumentProperties | null>;
     /** Set document core properties (partial update) */
     setProperties(properties: Partial<DocumentProperties>): Promise<boolean>;
+    /** Parse HTML string to ProseMirror JSON (uses hidden SuperDoc instance) */
+    parseHtml(html: string): Promise<ProseMirrorJSON>;
 }
 
 /**
@@ -420,6 +422,10 @@ declare function detectContentType(content: DocxContent): 'file' | 'html' | 'jso
  * Validate that content looks like ProseMirror JSON
  */
 declare function isProseMirrorJSON(content: unknown): boolean;
+/**
+ * Parse an HTML string into ProseMirror JSON using a hidden SuperDoc instance.
+ */
+declare function parseHtmlToJson(html: string, SuperDoc: SuperDocConstructor): Promise<ProseMirrorJSON>;
 /**
  * Parse a DOCX File into ProseMirror JSON using a hidden SuperDoc instance.
  */
@@ -710,4 +716,4 @@ declare function getBlankTemplateBlob(): Blob;
  */
 declare function isValidDocxFile(file: File): boolean;
 
-export { type AttrDiff, type AttributeChange, CSS_PREFIX, type ChangeLocation, type ComparisonResult, DEFAULT_AUTHOR, DEFAULT_SUPERDOC_USER, type DiffResult, type DiffSegment, type DocumentInfo, type DocumentProperties, type DocxContent, DocxDiffEditor, type DocxDiffEditorProps, type DocxDiffEditorRef, type EnrichedChange, type FingerprintedNode, type FormatChange, type FormatDetails, type HybridDiffResult, type NodeMatch, type ProseMirrorJSON, type ProseMirrorMark, type ProseMirrorNode, type StructuralChange, type StructuralChangeInfo, type StructuralChangeType, StructuralChangesPane, type StructuralPanePosition, type TrackChangeAuthor, alignDocuments, createTrackDeleteMark, createTrackFormatMark, createTrackInsertMark, DocxDiffEditor as default, detectContentType, diffDocuments, diffImages, diffLists, diffTables, extractEnrichedChanges, extractEnrichedChangesWithStructural, generateFingerprint, generateStructuralChangeSummary, getBlankTemplateBlob, getBlankTemplateFile, isAtomicNode, isImage, isList, isProseMirrorJSON, isTable, isValidDocxFile, mergeDocuments, parseDocxFile, processStructuralChanges };
+export { type AttrDiff, type AttributeChange, CSS_PREFIX, type ChangeLocation, type ComparisonResult, DEFAULT_AUTHOR, DEFAULT_SUPERDOC_USER, type DiffResult, type DiffSegment, type DocumentInfo, type DocumentProperties, type DocxContent, DocxDiffEditor, type DocxDiffEditorProps, type DocxDiffEditorRef, type EnrichedChange, type FingerprintedNode, type FormatChange, type FormatDetails, type HybridDiffResult, type NodeMatch, type ProseMirrorJSON, type ProseMirrorMark, type ProseMirrorNode, type StructuralChange, type StructuralChangeInfo, type StructuralChangeType, StructuralChangesPane, type StructuralPanePosition, type TrackChangeAuthor, alignDocuments, createTrackDeleteMark, createTrackFormatMark, createTrackInsertMark, DocxDiffEditor as default, detectContentType, diffDocuments, diffImages, diffLists, diffTables, extractEnrichedChanges, extractEnrichedChangesWithStructural, generateFingerprint, generateStructuralChangeSummary, getBlankTemplateBlob, getBlankTemplateFile, isAtomicNode, isImage, isList, isProseMirrorJSON, isTable, isValidDocxFile, mergeDocuments, parseDocxFile, parseHtmlToJson, processStructuralChanges };