@ansonlai/docx-redline-js 0.8.2 → 0.8.4

package/services/operation-heuristics.js

@@ -1,173 +1,197 @@
-import {
-    getParagraphText,
-    normalizeWhitespaceForTargeting
-} from '../core/paragraph-targeting.js';
-import {
-    getParagraphListInfo,
-    stripRedundantLeadingListMarkers
-} from '../core/list-targeting.js';
-
-const LIST_LINE_REGEX = /^(\s*)((?:\d+(?:\.\d+)*\.?|\((?:\d+|[a-zA-Z]|[ivxlcIVXLC]+)\)|[a-zA-Z]\.|[ivxlcIVXLC]+\.|[-*+\u2022]))\s+(.*)$/;
-const INLINE_LIST_MARKER_REGEX = /(?:^|\s)(?:\d+(?:\.\d+)*\.?|[A-Za-z]\.|[ivxlcIVXLC]+\.)\s+/g;
-
-function parseOutlineLevelFromMarker(marker) {
-    const normalized = String(marker || '').trim();
-    if (!/^\d+(?:\.\d+)+\.?$/.test(normalized)) return null;
-    const parts = normalized.replace(/\.$/, '').split('.');
-    return Math.max(0, parts.length - 1);
-}
-
-function parseModifiedListLines(modifiedText) {
-    const lines = String(modifiedText || '')
-        .split(/\r?\n/g)
-        .map(line => line.trimEnd())
-        .filter(line => line.trim().length > 0);
-    if (lines.length < 2) return null;
-
-    const parsed = [];
-    for (const line of lines) {
-        const markerMatch = line.match(LIST_LINE_REGEX);
-        if (!markerMatch) return null;
-        const marker = markerMatch[2];
-        const markerType = /^[-*+\u2022]$/.test(marker) ? 'bullet' : 'numbered';
-        parsed.push({
-            marker,
-            markerType,
-            level: Math.floor((markerMatch[1] || '').length / 2),
-            outlineLevel: markerType === 'numbered' ? parseOutlineLevelFromMarker(marker) : null,
-            text: stripRedundantLeadingListMarkers(markerMatch[3])
-        });
-    }
-    return parsed.length >= 2 ? parsed : null;
-}
-
-export function buildExplicitRangeInsertionEntries(explicitRangeParagraphs, modifiedText) {
-    if (!Array.isArray(explicitRangeParagraphs) || explicitRangeParagraphs.length === 0) return null;
-    const parsedLines = parseModifiedListLines(modifiedText);
-    if (!parsedLines) return null;
-
-    const originalTexts = explicitRangeParagraphs.map(paragraph =>
-        normalizeWhitespaceForTargeting(getParagraphText(paragraph))
-    );
-    const modifiedTexts = parsedLines.map(item => normalizeWhitespaceForTargeting(item.text));
-    if (originalTexts.some(text => !text) || modifiedTexts.some(text => !text)) return null;
-
-    const listInfos = explicitRangeParagraphs.map(paragraph => getParagraphListInfo(paragraph));
-    if (listInfos.some(info => !info || !info.numId)) return null;
-    const baselineNumId = String(listInfos[0].numId);
-    if (listInfos.some(info => String(info.numId) !== baselineNumId)) return null;
-
-    const matchedPairs = [];
-    let originalIndex = 0;
-    for (let modifiedIndex = 0; modifiedIndex < modifiedTexts.length && originalIndex < originalTexts.length; modifiedIndex += 1) {
-        if (modifiedTexts[modifiedIndex] === originalTexts[originalIndex]) {
-            matchedPairs.push({ originalIndex, modifiedIndex });
-            originalIndex += 1;
-        }
-    }
-    if (originalIndex !== originalTexts.length) return null;
-
-    const matchedModifiedIndexes = new Set(matchedPairs.map(pair => pair.modifiedIndex));
-    const insertedIndexes = [];
-    for (let idx = 0; idx < parsedLines.length; idx += 1) {
-        if (!matchedModifiedIndexes.has(idx)) insertedIndexes.push(idx);
-    }
-    if (insertedIndexes.length === 0) return null;
-
-    const baseIndentLevel = parsedLines[0]?.level || 0;
-    return insertedIndexes.map(modifiedIndex => {
-        const nextMatch = matchedPairs.find(pair => pair.modifiedIndex > modifiedIndex) || null;
-        const prevMatch = [...matchedPairs].reverse().find(pair => pair.modifiedIndex < modifiedIndex) || null;
-        const referenceMatch = nextMatch || prevMatch;
-        if (!referenceMatch) return null;
-        const referenceListInfo = listInfos[referenceMatch.originalIndex] || listInfos[0];
-        if (!referenceListInfo) return null;
-
-        const entry = parsedLines[modifiedIndex];
-        const relativeLevel = Math.max(0, entry.level - baseIndentLevel);
-        const explicitOutlineLevel = Number.isInteger(entry.outlineLevel) ? entry.outlineLevel : null;
-        return {
-            text: entry.text,
-            markerType: entry.markerType,
-            ilvl: explicitOutlineLevel != null
-                ? explicitOutlineLevel
-                : Math.max(0, (referenceListInfo.ilvl || 0) + relativeLevel),
-            numId: String(referenceListInfo.numId),
-            insertBeforeOriginalIndex: nextMatch ? nextMatch.originalIndex : null
-        };
-    }).filter(Boolean);
-}
-
-function countWords(text) {
-    return String(text || '').trim().split(/\s+/).filter(Boolean).length;
-}
-
-function hasMultipleInlineListMarkers(text) {
-    const source = String(text || '');
-    if (!source) return false;
-    let count = 0;
-    const regex = new RegExp(INLINE_LIST_MARKER_REGEX.source, INLINE_LIST_MARKER_REGEX.flags);
-    while (regex.exec(source)) {
-        count += 1;
-        if (count >= 2) return true;
-    }
-    return false;
-}
-
-export function deriveSingleParagraphListAdjacencyInsertion(currentParagraphText, modifiedText) {
-    const rawCurrent = String(currentParagraphText || '').trim();
-    const rawModified = String(modifiedText || '').trim();
-    if (!rawCurrent || !rawModified || rawModified === rawCurrent || rawModified.includes('\n')) return null;
-    const normalizedCurrent = normalizeWhitespaceForTargeting(rawCurrent);
-    const sanitizeCandidate = text => stripRedundantLeadingListMarkers(String(text || '').trim()).trim();
-    const buildCandidate = (position, text) => {
-        const cleanedText = sanitizeCandidate(text);
-        const cleanedNormalized = normalizeWhitespaceForTargeting(cleanedText);
-        if (!cleanedText || cleanedText === rawCurrent || countWords(cleanedText) < 6) return null;
-        if (hasMultipleInlineListMarkers(cleanedText)) return null;
-        if (normalizedCurrent && cleanedNormalized.includes(normalizedCurrent)) return null;
-        return { position, text: cleanedText };
-    };
-
-    if (rawModified.endsWith(rawCurrent)) {
-        const candidate = buildCandidate('before', rawModified.slice(0, -rawCurrent.length));
-        if (candidate) return candidate;
-    }
-    if (rawModified.startsWith(rawCurrent)) {
-        const candidate = buildCandidate('after', rawModified.slice(rawCurrent.length));
-        if (candidate) return candidate;
-    }
-
-    const normalizedModified = normalizeWhitespaceForTargeting(rawModified);
-    if (!normalizedCurrent || normalizedCurrent === normalizedModified) return null;
-    if (normalizedModified.endsWith(normalizedCurrent)) {
-        const candidate = buildCandidate('before', normalizedModified.slice(0, -normalizedCurrent.length));
-        if (candidate) return candidate;
-    }
-    if (normalizedModified.startsWith(normalizedCurrent)) {
-        const candidate = buildCandidate('after', normalizedModified.slice(normalizedCurrent.length));
-        if (candidate) return candidate;
-    }
-    return null;
-}
-
-export function deriveSingleParagraphPlainAdjacencyInsertion(currentParagraphText, modifiedText) {
-    const rawCurrent = String(currentParagraphText || '').trim();
-    const rawModified = String(modifiedText || '');
-    if (!rawCurrent || !rawModified || !rawModified.includes('\n')) return null;
-
-    const lines = rawModified.split(/\r?\n/g).map(line => String(line || '').trim()).filter(Boolean);
-    if (lines.length < 2) return null;
-    const normalize = value => normalizeWhitespaceForTargeting(String(value || ''));
-    const normalizedCurrent = normalize(rawCurrent);
-
-    if (normalize(lines[lines.length - 1]) === normalizedCurrent) {
-        const paragraphs = lines.slice(0, -1).filter(Boolean);
-        if (paragraphs.length > 0) return { position: 'before', paragraphs };
-    }
-    if (normalize(lines[0]) === normalizedCurrent) {
-        const paragraphs = lines.slice(1).filter(Boolean);
-        if (paragraphs.length > 0) return { position: 'after', paragraphs };
-    }
-    return null;
-}
+import {
+    getParagraphText,
+    normalizeWhitespaceForTargeting
+} from '../core/paragraph-targeting.js';
+import {
+    getParagraphListInfo,
+    stripRedundantLeadingListMarkers
+} from '../core/list-targeting.js';
+
+const LIST_LINE_REGEX = /^(\s*)((?:\d+(?:\.\d+)*\.?|\((?:\d+|[a-zA-Z]|[ivxlcIVXLC]+)\)|[a-zA-Z]\.|[ivxlcIVXLC]+\.|[-*+\u2022]))\s+(.*)$/;
+const INLINE_LIST_MARKER_REGEX = /(?:^|\s)(?:\d+(?:\.\d+)*\.?|[A-Za-z]\.|[ivxlcIVXLC]+\.)\s+/g;
+
+function parseOutlineLevelFromMarker(marker) {
+    const normalized = String(marker || '').trim();
+    if (!/^\d+(?:\.\d+)+\.?$/.test(normalized)) return null;
+    const parts = normalized.replace(/\.$/, '').split('.');
+    return Math.max(0, parts.length - 1);
+}
+
+function parseModifiedListLines(modifiedText) {
+    const lines = String(modifiedText || '')
+        .split(/\r?\n/g)
+        .map(line => line.trimEnd())
+        .filter(line => line.trim().length > 0);
+    if (lines.length < 2) return null;
+
+    const parsed = [];
+    for (const line of lines) {
+        const markerMatch = line.match(LIST_LINE_REGEX);
+        if (!markerMatch) return null;
+        const marker = markerMatch[2];
+        const markerType = /^[-*+\u2022]$/.test(marker) ? 'bullet' : 'numbered';
+        parsed.push({
+            marker,
+            markerType,
+            level: Math.floor((markerMatch[1] || '').length / 2),
+            outlineLevel: markerType === 'numbered' ? parseOutlineLevelFromMarker(marker) : null,
+            text: stripRedundantLeadingListMarkers(markerMatch[3])
+        });
+    }
+    return parsed.length >= 2 ? parsed : null;
+}
+
+export function buildExplicitRangeInsertionEntries(explicitRangeParagraphs, modifiedText) {
+    if (!Array.isArray(explicitRangeParagraphs) || explicitRangeParagraphs.length === 0) return null;
+    const parsedLines = parseModifiedListLines(modifiedText);
+    if (!parsedLines) return null;
+
+    const originalTexts = explicitRangeParagraphs.map(paragraph =>
+        normalizeWhitespaceForTargeting(getParagraphText(paragraph))
+    );
+    const modifiedTexts = parsedLines.map(item => normalizeWhitespaceForTargeting(item.text));
+    if (originalTexts.some(text => !text) || modifiedTexts.some(text => !text)) return null;
+
+    const listInfos = explicitRangeParagraphs.map(paragraph => getParagraphListInfo(paragraph));
+    if (listInfos.some(info => !info || !info.numId)) return null;
+    const baselineNumId = String(listInfos[0].numId);
+    if (listInfos.some(info => String(info.numId) !== baselineNumId)) return null;
+
+    const matchedPairs = [];
+    let originalIndex = 0;
+    for (let modifiedIndex = 0; modifiedIndex < modifiedTexts.length && originalIndex < originalTexts.length; modifiedIndex += 1) {
+        if (modifiedTexts[modifiedIndex] === originalTexts[originalIndex]) {
+            matchedPairs.push({ originalIndex, modifiedIndex });
+            originalIndex += 1;
+        }
+    }
+    if (originalIndex !== originalTexts.length) return null;
+
+    const matchedModifiedIndexes = new Set(matchedPairs.map(pair => pair.modifiedIndex));
+    const insertedIndexes = [];
+    for (let idx = 0; idx < parsedLines.length; idx += 1) {
+        if (!matchedModifiedIndexes.has(idx)) insertedIndexes.push(idx);
+    }
+    if (insertedIndexes.length === 0) return null;
+
+    const baseIndentLevel = parsedLines[0]?.level || 0;
+    return insertedIndexes.map(modifiedIndex => {
+        const nextMatch = matchedPairs.find(pair => pair.modifiedIndex > modifiedIndex) || null;
+        const prevMatch = [...matchedPairs].reverse().find(pair => pair.modifiedIndex < modifiedIndex) || null;
+        const referenceMatch = nextMatch || prevMatch;
+        if (!referenceMatch) return null;
+        const referenceListInfo = listInfos[referenceMatch.originalIndex] || listInfos[0];
+        if (!referenceListInfo) return null;
+
+        const entry = parsedLines[modifiedIndex];
+        const relativeLevel = Math.max(0, entry.level - baseIndentLevel);
+        const explicitOutlineLevel = Number.isInteger(entry.outlineLevel) ? entry.outlineLevel : null;
+        return {
+            text: entry.text,
+            markerType: entry.markerType,
+            ilvl: explicitOutlineLevel != null
+                ? explicitOutlineLevel
+                : Math.max(0, (referenceListInfo.ilvl || 0) + relativeLevel),
+            numId: String(referenceListInfo.numId),
+            insertBeforeOriginalIndex: nextMatch ? nextMatch.originalIndex : null
+        };
+    }).filter(Boolean);
+}
+
+function countWords(text) {
+    return String(text || '').trim().split(/\s+/).filter(Boolean).length;
+}
+
+function hasMultipleInlineListMarkers(text) {
+    const source = String(text || '');
+    if (!source) return false;
+    let count = 0;
+    const regex = new RegExp(INLINE_LIST_MARKER_REGEX.source, INLINE_LIST_MARKER_REGEX.flags);
+    while (regex.exec(source)) {
+        count += 1;
+        if (count >= 2) return true;
+    }
+    return false;
+}
+
+export function deriveSingleParagraphListAdjacencyInsertion(currentParagraphText, modifiedText) {
+    const rawCurrent = String(currentParagraphText || '').trim();
+    const rawModified = String(modifiedText || '').trim();
+    if (!rawCurrent || !rawModified || rawModified === rawCurrent || rawModified.includes('\n')) return null;
+    const normalizedCurrent = normalizeWhitespaceForTargeting(rawCurrent);
+    const sanitizeCandidate = text => stripRedundantLeadingListMarkers(String(text || '').trim()).trim();
+    const buildCandidate = (position, text) => {
+        const cleanedText = sanitizeCandidate(text);
+        const cleanedNormalized = normalizeWhitespaceForTargeting(cleanedText);
+        if (!cleanedText || cleanedText === rawCurrent || countWords(cleanedText) < 6) return null;
+        if (hasMultipleInlineListMarkers(cleanedText)) return null;
+        if (normalizedCurrent && cleanedNormalized.includes(normalizedCurrent)) return null;
+        return { position, text: cleanedText };
+    };
+
+    if (rawModified.endsWith(rawCurrent)) {
+        const candidate = buildCandidate('before', rawModified.slice(0, -rawCurrent.length));
+        if (candidate) return candidate;
+    }
+    if (rawModified.startsWith(rawCurrent)) {
+        const candidate = buildCandidate('after', rawModified.slice(rawCurrent.length));
+        if (candidate) return candidate;
+    }
+
+    const normalizedModified = normalizeWhitespaceForTargeting(rawModified);
+    if (!normalizedCurrent || normalizedCurrent === normalizedModified) return null;
+    if (normalizedModified.endsWith(normalizedCurrent)) {
+        const candidate = buildCandidate('before', normalizedModified.slice(0, -normalizedCurrent.length));
+        if (candidate) return candidate;
+    }
+    if (normalizedModified.startsWith(normalizedCurrent)) {
+        const candidate = buildCandidate('after', normalizedModified.slice(normalizedCurrent.length));
+        if (candidate) return candidate;
+    }
+    return null;
+}
+
+export function deriveSingleParagraphPlainAdjacencyInsertion(currentParagraphText, modifiedText) {
+    const rawCurrent = String(currentParagraphText || '').trim();
+    const rawModified = String(modifiedText || '');
+    if (!rawCurrent || !rawModified || !rawModified.includes('\n')) return null;
+
+    const lines = rawModified.split(/\r?\n/g).map(line => String(line || '').trim()).filter(Boolean);
+    if (lines.length < 2) return null;
+    const normalize = value => normalizeWhitespaceForTargeting(String(value || ''));
+    const normalizedCurrent = normalize(rawCurrent);
+
+    if (normalize(lines[lines.length - 1]) === normalizedCurrent) {
+        const paragraphs = lines.slice(0, -1).filter(Boolean);
+        if (paragraphs.length > 0) return { position: 'before', paragraphs };
+    }
+    if (normalize(lines[0]) === normalizedCurrent) {
+        const paragraphs = lines.slice(1).filter(Boolean);
+        if (paragraphs.length > 0) return { position: 'after', paragraphs };
+    }
+    return null;
+}
+
+/**
+ * Detects "retained paragraph followed by one Markdown table": a single leading
+ * paragraph line (inline Markdown allowed) then a contiguous pipe table with a
+ * separator row. The caller verifies that the paragraph line's plain text equals
+ * the target paragraph.
+ *
+ * @param {string} modifiedText - Requested replacement text
+ * @returns {{ paragraphMarkdown: string, tableText: string }|null}
+ */
+export function deriveParagraphThenTableAppend(modifiedText) {
+    const lines = String(modifiedText || '')
+        .split(/\r?\n/g)
+        .map(line => line.trimEnd())
+        .filter(line => line.trim().length > 0);
+    if (lines.length < 3) return null;
+    const tableStart = lines.findIndex(line => line.trim().startsWith('|'));
+    if (tableStart !== 1) return null;
+    const tableLines = lines.slice(1);
+    if (!tableLines.every(line => line.trim().startsWith('|'))) return null;
+    const separator = tableLines[1].replace(/\s+/g, '');
+    if (!/^\|:?-{3,}:?(\|:?-{3,}:?)+\|?$/.test(separator)) return null;
+    return { paragraphMarkdown: lines[0].trim(), tableText: tableLines.join('\n') };
+}