@ansonlai/docx-redline-js 0.8.3 → 0.9.0

package/services/table-reconciliation.js

@@ -7,8 +7,9 @@
 import { computeWordLevelDiffOps } from '../pipeline/diff-engine.js';
 import { splitRunsAtDiffBoundaries, applyPatches } from '../pipeline/patching.js';
 import { serializeToOoxml } from '../pipeline/serialization.js';
-import { createRevisionMetadata, escapeXml, RunKind } from '../core/types.js';
-import { preprocessMarkdown } from '../pipeline/markdown-processor.js';
+import { createRevisionMetadata, escapeXml, RunKind } from '../core/types.js';
+import { preprocessMarkdown } from '../pipeline/markdown-processor.js';
+import { paragraphMarkPPrXml } from '../engine/paragraph-mark-revisions.js';
 
 /**
  * Generates a new w:tbl OOXML structure from Markdown table data.
@@ -18,14 +19,16 @@ import { preprocessMarkdown } from '../pipeline/markdown-processor.js';
  * @param {Object} options - { generateRedlines, author }
  * @returns {string} Complete w:tbl OOXML
  */
-export function generateTableOoxml(tableData, options = {}) {
-    const {
-        generateRedlines = false,
-        author = 'AI',
-        revisionIdAllocator = null,
-        trackAsBlock = false
-    } = options;
-    const tableInsertMeta = generateRedlines && trackAsBlock ? createRevisionMetadata(author, revisionIdAllocator) : null;
+export function generateTableOoxml(tableData, options = {}) {
+    const {
+        generateRedlines = false,
+        author = 'AI',
+        revisionIdAllocator = null,
+        trackAsBlock = false
+    } = options;
+    // trackAsBlock: Word-native inserted table = tracked-inserted rows (trPr/ins), inserted cell runs and
+    // tracked-inserted cell paragraph marks. (w:ins around w:tbl is not a table insertion to Word.)
+    const trackRows = generateRedlines && trackAsBlock;
 
     // Determine number of columns
     const numCols = tableData.headers?.length || (tableData.rows?.[0]?.length || 1);
@@ -65,7 +68,7 @@ export function generateTableOoxml(tableData, options = {}) {
 
             // Build run model for the cell
             const runModel = [{
-                kind: generateRedlines && !trackAsBlock ? RunKind.INSERTION : RunKind.TEXT,
+                kind: generateRedlines ? RunKind.INSERTION : RunKind.TEXT,
                 text: cleanText,
                 rPrXml: isHeaderRow ? '<w:rPr><w:b/></w:rPr>' : '',
                 author,
@@ -73,11 +76,12 @@ export function generateTableOoxml(tableData, options = {}) {
                 endOffset: cleanText.length
             }];
 
-            const runsOoxml = serializeToOoxml(runModel, null, formatHints, {
-                author,
-                generateRedlines,
-                revisionIdAllocator
-            });
+            const cellPPr = trackRows ? paragraphMarkPPrXml('ins', { author, revisionIdAllocator }) : null;
+            const runsOoxml = serializeToOoxml(runModel, cellPPr, formatHints, {
+                author,
+                generateRedlines,
+                revisionIdAllocator
+            });
 
             // Cell properties
             const tcPr = '<w:tcPr><w:tcW w:w="0" w:type="auto"/></w:tcPr>';
@@ -85,24 +89,20 @@ export function generateTableOoxml(tableData, options = {}) {
             cellsXml += `<w:tc>${tcPr}${runsOoxml}</w:tc>`;
         }
 
-        // Keep each logical record intact across page boundaries. When a
-        // Markdown header is present, also mark it as a repeating Word table
-        // header so continuation pages retain their column labels.
-        const trPr = isHeaderRow
-            ? '<w:trPr><w:tblHeader/><w:cantSplit/></w:trPr>'
-            : '<w:trPr><w:cantSplit/></w:trPr>';
-        rowsXml += `<w:tr>${trPr}${cellsXml}</w:tr>`;
+        // Keep each logical record intact across page boundaries. When a
+        // Markdown header is present, also mark it as a repeating Word table
+        // header so continuation pages retain their column labels.
+        const trPr = isHeaderRow
+            ? '<w:trPr><w:tblHeader/><w:cantSplit/></w:trPr>'
+            : '<w:trPr><w:cantSplit/></w:trPr>';
+        const rowMeta = trackRows ? createRevisionMetadata(author, revisionIdAllocator) : null;
+        const trackedTrPr = rowMeta
+            ? trPr.replace('</w:trPr>', `<w:ins w:id="${rowMeta.id}" w:author="${escapeXml(rowMeta.author)}" w:date="${rowMeta.date}"/></w:trPr>`)
+            : trPr;
+        rowsXml += `<w:tr>${trackedTrPr}${cellsXml}</w:tr>`;
     }
 
-    // Build the table
-    let tableXml = `<w:tbl>${tblPr}${tblGrid}${rowsXml}</w:tbl>`;
-
-    // Wrap entire table in w:ins if generating redlines
-    if (tableInsertMeta) {
-        tableXml = `<w:ins w:id="${tableInsertMeta.id}" w:author="${escapeXml(tableInsertMeta.author)}" w:date="${tableInsertMeta.date}">${tableXml}</w:ins>`;
-    }
-
-    return tableXml;
+    return `<w:tbl>${tblPr}${tblGrid}${rowsXml}</w:tbl>`;
 }
 
 /**
@@ -151,8 +151,9 @@ export function diffTablesWithVirtualGrid(oldGrid, newTableData) {
             if (col > oldCell.gridCol) continue;
 
             if (newCellText !== undefined) {
-                const oldText = oldCell.getText();
-                if (oldText !== newCellText) {
+                // Markdown cannot represent a cell's trailing empty paragraphs, so they do not count as a change.
+                const oldText = trimTrailingNewlines(oldCell.getText());
+                if (oldText !== trimTrailingNewlines(newCellText)) {
                     operations.push({
                         type: 'cell_modify',
                         gridRow: row,
@@ -177,7 +178,7 @@ export function diffTablesWithVirtualGrid(oldGrid, newTableData) {
  * @returns {string} Reconciled w:tbl OOXML
  */
 export function serializeVirtualGridToOoxml(grid, operations, options) {
-    const { generateRedlines, author, revisionIdAllocator = null } = options;
+    const { generateRedlines, author, revisionIdAllocator = null } = options;
     const opIndex = buildTableOperationIndex(operations);
 
     let rowsXml = '';
@@ -219,7 +220,7 @@ export function serializeVirtualGridToOoxml(grid, operations, options) {
         let trPr = grid.trPrList[row] || '<w:trPr/>';
 
         if (rowDeleteOp && generateRedlines) {
-            const metadata = createRevisionMetadata(author, revisionIdAllocator);
+            const metadata = createRevisionMetadata(author, revisionIdAllocator);
             const delMark = `<w:del w:id="${metadata.id}" w:author="${escapeXml(metadata.author)}" w:date="${metadata.date}"/>`;
             if (trPr.includes('</w:trPr>')) {
                 trPr = trPr.replace('</w:trPr>', `${delMark}</w:trPr>`);
@@ -235,7 +236,7 @@ export function serializeVirtualGridToOoxml(grid, operations, options) {
     const insertOps = opIndex.rowInsertOperations;
     for (const op of insertOps) {
         let cellsXml = '';
-        const rowInsertMeta = generateRedlines ? createRevisionMetadata(author, revisionIdAllocator) : null;
+        const rowInsertMeta = generateRedlines ? createRevisionMetadata(author, revisionIdAllocator) : null;
 
         for (const cellText of op.cells) {
             const { cleanText, formatHints } = preprocessMarkdown(cellText);
@@ -249,11 +250,11 @@ export function serializeVirtualGridToOoxml(grid, operations, options) {
                 endOffset: cleanText.length
             }];
 
-            const runsOoxml = serializeToOoxml(runModel, null, formatHints, {
-                author,
-                generateRedlines,
-                revisionIdAllocator
-            });
+            const runsOoxml = serializeToOoxml(runModel, null, formatHints, {
+                author,
+                generateRedlines,
+                revisionIdAllocator
+            });
             // serializeToOoxml already returns one or more w:p blocks.
             cellsXml += `<w:tc><w:tcPr><w:tcW w:w="0" w:type="auto"/></w:tcPr>${runsOoxml}</w:tc>`;
         }
@@ -309,8 +310,61 @@ function buildTableOperationIndex(operations) {
     };
 }
 
+function trimTrailingNewlines(text) {
+    return typeof text === 'string' ? text.replace(/\n+$/, '') : text;
+}
+
+/**
+ * Reconciles a multi-paragraph cell paragraph by paragraph so unchanged paragraphs (including
+ * empty ones the Markdown cell text cannot express) are preserved exactly.
+ */
+function reconcileMultiBlockCellContent(cell, newText, options) {
+    const { generateRedlines, author, revisionIdAllocator = null } = options;
+    const lines = String(newText ?? '').split('\n');
+    // Trailing blank lines carry no content; never create inserted paragraphs from them.
+    while (lines.length > cell.blocks.length && lines[lines.length - 1] === '') lines.pop();
+    let out = '';
+
+    cell.blocks.forEach((block, index) => {
+        const line = lines[index];
+        if (line === undefined || line === block.acceptedText) {
+            if (line === undefined && block.acceptedText !== '') {
+                // Markdown dropped this paragraph's text: track the text deletion.
+                const diffOps = computeWordLevelDiffOps(block.acceptedText, '');
+                const model = applyPatches(splitRunsAtDiffBoundaries(block.runModel, diffOps), diffOps, {
+                    generateRedlines, author, formatHints: []
+                });
+                out += serializeToOoxml(model, block.pPr, [], { author, generateRedlines, revisionIdAllocator });
+            } else {
+                out += serializeToOoxml(block.runModel, block.pPr, [], {});
+            }
+            return;
+        }
+        const { cleanText, formatHints } = preprocessMarkdown(line);
+        const diffOps = computeWordLevelDiffOps(block.acceptedText, cleanText);
+        const model = applyPatches(splitRunsAtDiffBoundaries(block.runModel, diffOps), diffOps, {
+            generateRedlines, author, formatHints
+        });
+        out += serializeToOoxml(model, block.pPr, formatHints, { author, generateRedlines, revisionIdAllocator });
+    });
+
+    for (let i = cell.blocks.length; i < lines.length; i++) {
+        const { cleanText, formatHints } = preprocessMarkdown(lines[i]);
+        const runModel = [{
+            kind: generateRedlines ? RunKind.INSERTION : RunKind.TEXT,
+            text: cleanText, rPrXml: '', author, startOffset: 0, endOffset: cleanText.length
+        }];
+        const pPr = generateRedlines ? paragraphMarkPPrXml('ins', { author, revisionIdAllocator }) : null;
+        out += serializeToOoxml(runModel, pPr, formatHints, { author, generateRedlines, revisionIdAllocator });
+    }
+    return out;
+}
+
 function reconcileCellContent(cell, newText, options) {
-    const { generateRedlines, author, revisionIdAllocator = null } = options;
+    if (cell.blocks.length > 1) {
+        return reconcileMultiBlockCellContent(cell, newText, options);
+    }
+    const { generateRedlines, author, revisionIdAllocator = null } = options;
     const { cleanText, formatHints } = preprocessMarkdown(newText);
 
     // For now, satisfy with single block or join blocks
@@ -327,11 +381,11 @@ function reconcileCellContent(cell, newText, options) {
         formatHints
     });
 
-    const runsOoxml = serializeToOoxml(patchedModel, baseBlock.pPr, formatHints, {
-        author,
-        generateRedlines,
-        revisionIdAllocator
-    });
+    const runsOoxml = serializeToOoxml(patchedModel, baseBlock.pPr, formatHints, {
+        author,
+        generateRedlines,
+        revisionIdAllocator
+    });
     return runsOoxml;
 }
 
@@ -343,7 +397,7 @@ function serializeCellBlocks(blocks) {
     }).join('');
 }
 
-function buildTcXml(cell, content, _options) {
+function buildTcXml(cell, content, _options) {
     let tcPr = cell.tcPrXml;
 
     // Ensure gridSpan is preserved