speccore 8.3.121 → 8.3.127

package/dist/core/analyze-engine.js

@@ -12,7 +12,7 @@ exports.generateSpecsFromRequirements = generateSpecsFromRequirements;
  *
  * 输入:
  *   - 纯需求: --req docs/a.md docs/b.md
- *   - 纯代码: --src backend/src 20-frontend/src
+ *   - 纯代码: --src backend/src frontend/src
  *   - 需求+代码: 两者都指定
  *
  * 输出范围 (scope):
@@ -25,6 +25,8 @@ const path_1 = require("path");
 const logger_1 = require("../utils/logger");
 const task_utils_1 = require("../utils/task-utils");
 const code_scanner_1 = require("./code-scanner");
+// v8.3.125+: 同义词扩展，解决中英文/缩写匹配失败问题
+const synonyms_1 = require("../utils/synonyms");
 const ai_context_generator_1 = require("./ai-context-generator");
 const git_integration_1 = require("./git-integration");
 const rag_engine_1 = require("./rag-engine");
@@ -51,7 +53,7 @@ async function runAnalysis(input) {
         if (input.scope === 'iteration' && input.iteration) {
             const productReqDir = (0, path_1.join)(`Iteration-${input.iteration}`, '010-requirements');
             if (await (0, fs_extra_1.pathExists)(productReqDir)) {
-                // 递归扫描子目录 (backend/ 20-frontend/Web/ 等)
+                // 递归扫描子目录 ({端名}/ 等)
                 const scanDir = async (dir) => {
                     const entries = await (0, fs_extra_1.readdir)(dir, { withFileTypes: true });
                     for (const e of entries) {
@@ -160,9 +162,11 @@ async function runAnalysis(input) {
         logger_1.logger.debug('RAG 索引生成失败（非关键）:', e);
     }
     // ── 统一刷新代码索引 + 知识图谱（所有 scope）──
+    // v8.3.125+: 按迭代端名筛选源码路径，减少增量刷新范围
     try {
         logger_1.logger.info('   🔄 刷新代码索引...');
-        await (0, code_scanner_1.buildCodeIndex)(undefined, true);
+        const codeIndexScope = await (0, code_scanner_1.resolveIterationScope)(effectiveInput.iteration);
+        await (0, code_scanner_1.buildCodeIndex)(codeIndexScope, true);
         logger_1.logger.info('   ✅ 代码索引已刷新');
     }
     catch (e) {
@@ -203,16 +207,18 @@ async function analyzeRequirements(input) {
     const issues = scanCompleteness(fullContent);
     const archImpact = await analyzeArchitectureImpact(fullContent);
     // ── 需求分析时默认读取关联代码（避免分析不接地气 + 减少重复 token 消耗） ──
+    // v8.3.122+: 增加文件数和字节配额，确保深入阅读业务逻辑
     let sourceContents = {};
     const shouldReadSource = input.readSource !== false;
     if (shouldReadSource && (input.iteration || input.taskId)) {
-        const limit = input.depth === 'deep' ? 15 : (input.depth === 'quick' ? 3 : 8);
-        const maxBytes = input.depth === 'deep' ? 80000 : (input.depth === 'quick' ? 20000 : 40000);
+        const limit = input.depth === 'deep' ? 25 : (input.depth === 'quick' ? 5 : 12);
+        const maxBytes = input.depth === 'deep' ? 150000 : (input.depth === 'quick' ? 30000 : 70000);
         const rawMatches = await (0, code_scanner_1.findRelevantCode)(fullContent, limit, input.sourceScope, input.iteration, input.taskId);
-        sourceContents = await (0, code_scanner_1.readRelevantSource)(rawMatches, maxBytes);
+        sourceContents = await (0, code_scanner_1.readRelevantSource)(rawMatches, maxBytes, limit);
         if (Object.keys(sourceContents).length > 0) {
             const scopeHint = input.sourceScope ? ` (范围: ${input.sourceScope})` : '';
-            logger_1.logger.info(`   📖 需求分析已关联 ${Object.keys(sourceContents).length} 个源码文件${scopeHint}`);
+            const totalChars = Object.values(sourceContents).reduce((a, c) => a + c.length, 0);
+            logger_1.logger.info(`   📖 需求分析已关联 ${Object.keys(sourceContents).length} 个源码文件 (${Math.round(totalChars / 1000)}K 字符)${scopeHint}`);
         }
     }
     let outputPath;
@@ -339,15 +345,17 @@ async function analyzeCombined(input) {
     // 代码结构分析
     const fileStats = await scanSourceDirs(input.sources, input.depth);
     const apiInventory = await buildApiInventory(input.sources);
-    // 确保代码索引是最新的（解耦设计：总是全量索引，分析时按 scope 筛选）
-    if (await (0, code_scanner_1.isIndexStale)()) {
-        logger_1.logger.info('   🔍 索引已过期，自动重建全量代码索引...');
-        await (0, code_scanner_1.buildCodeIndex)(); // 总是全量建，不是按 scope
+    // 确保代码索引是最新的
+    // v8.3.125+: 按迭代端名筛选 scope + 增量刷新 + mtime 精确判断过期
+    const codeIndexScope = await (0, code_scanner_1.resolveIterationScope)(input.iteration);
+    if (await (0, code_scanner_1.isIndexStale)(codeIndexScope)) {
+        logger_1.logger.info('   🔍 索引已过期，自动重建代码索引...');
+        await (0, code_scanner_1.buildCodeIndex)(codeIndexScope, true);
         logger_1.logger.info('   ✅ 代码索引已更新（包含多端识别 + 模块分组 + git 联动分析）');
         logger_1.logger.info('   💡 提示：可手动运行 `speccore code-index --show` 查看索引摘要');
     }
     else {
-        logger_1.logger.info('   📚 使用缓存的代码索引（1 小时内已构建）');
+        logger_1.logger.info('   📚 使用缓存的代码索引（源码未变更）');
     }
     // ── AI 上下文生成: 替代关键词匹配（传入已读取的需求内容，避免重复 readFile） ──
     const aiContext = await (0, ai_context_generator_1.generateAIContext)({
@@ -364,13 +372,14 @@ async function analyzeCombined(input) {
     logger_1.logger.info(`   🔗 API: ${aiContext.totalApis} 个`);
     // 默认读取相关源码内容注入分析（除非明确关闭）
     // 解耦设计：从完整索引中按 sourceScope 筛选，不是按 scope 建索引
+    // v8.3.122+: 增加配额，确保深入阅读业务逻辑和公共模块
     let sourceContents = {};
     const shouldReadSource = input.readSource !== false;
     if (shouldReadSource) {
-        const limit = input.depth === 'deep' ? 20 : (input.depth === 'quick' ? 5 : 10);
-        const maxBytes = input.depth === 'deep' ? 120000 : (input.depth === 'quick' ? 30000 : 60000);
+        const limit = input.depth === 'deep' ? 30 : (input.depth === 'quick' ? 8 : 15);
+        const maxBytes = input.depth === 'deep' ? 200000 : (input.depth === 'quick' ? 40000 : 100000);
         const rawMatches = await (0, code_scanner_1.findRelevantCode)(fullReqContent, limit, input.sourceScope, input.iteration, input.taskId);
-        sourceContents = await (0, code_scanner_1.readRelevantSource)(rawMatches, maxBytes);
+        sourceContents = await (0, code_scanner_1.readRelevantSource)(rawMatches, maxBytes, limit);
         if (Object.keys(sourceContents).length > 0) {
             const scopeHint = input.sourceScope ? ` (范围: ${input.sourceScope})` : '';
             logger_1.logger.info(`   📖 已读取 ${Object.keys(sourceContents).length} 个源码文件${scopeHint} (${Object.values(sourceContents).reduce((a, c) => a + c.length, 0)} 字符)`);
@@ -835,12 +844,14 @@ function buildIterationReqReport(input, issues, archImpact, sourceContents) {
     if (archImpact.newDependencies.length > 0)
         r += `- [ ] 确认新增依赖的引入方案和排期\n`;
     // ── 代码关联分析（v6.8.0 新增：需求分析默认关联代码） ──
+    // v8.3.122+: 按深度动态调整代码展示配额，确保业务逻辑可见
     if (sourceContents && Object.keys(sourceContents).length > 0) {
         r += `\n---\n\n## 3.5 关联代码现状\n\n`;
         r += `> 以下源码文件与当前需求相关，供技术方案参考\n\n`;
+        const previewLimit = input.depth === 'deep' ? 3000 : (input.depth === 'quick' ? 800 : 1500);
         for (const [path, content] of Object.entries(sourceContents)) {
-            const preview = content.slice(0, 600).replace(/\n/g, '\n  ');
-            r += `### \`${path}\`\n\n\`\`\`${path.split('.').pop() || 'ts'}\n${preview}${content.length > 600 ? '\n  // ... 截断 ...' : ''}\n\`\`\`\n\n`;
+            const preview = content.slice(0, previewLimit).replace(/\n/g, '\n  ');
+            r += `### \`${path}\`\n\n\`\`\`${path.split('.').pop() || 'ts'}\n${preview}${content.length > previewLimit ? '\n  // ... 截断 ...' : ''}\n\`\`\`\n\n`;
         }
     }
     r += `\n---\n\n## 4. 技术方案（待填写）\n\n`;
@@ -1098,14 +1109,20 @@ async function buildAIEnhancedReport(input, scope, params) {
         if (readFiles.length > 0 || uncoveredFiles.length > 0) {
             r += `## 📚 源码分析清单\n\n`;
             // 已分析文件
+            // v8.3.122+: 增加代码内容预览，确保 AI 能看到业务逻辑
             if (readFiles.length > 0) {
                 r += `### ✅ 已分析（${readFiles.length} 个文件）\n\n`;
+                const previewLimit = input.depth === 'deep' ? 3000 : (input.depth === 'quick' ? 800 : 1500);
                 for (const [file, content] of Object.entries(sourceContents)) {
                     const lines = content.split('\n').length;
                     const size = (content.length / 1024).toFixed(1);
                     r += `- \`${file}\` (${lines} 行, ${size}KB)\n`;
+                    // 展示代码预览
+                    if (content.length > 0) {
+                        const preview = content.slice(0, previewLimit).replace(/\n/g, '\n  ');
+                        r += `\n  \`\`\`${file.split('.').pop() || 'ts'}\n  ${preview}${content.length > previewLimit ? '\n  // ... 截断 ...' : ''}\n  \`\`\`\n\n`;
+                    }
                 }
-                r += `\n`;
             }
             // 未覆盖文件（按目录分组）
             if (uncoveredFiles.length > 0) {
@@ -1182,7 +1199,10 @@ function icon(severity) {
         default: return '⚪';
     }
 }
-/** 按需求功能 × 端 分目录写入分析报告（按端提取差异化内容） */
+/** 按需求功能 × 端 分目录写入分析报告（按端提取差异化内容）
+ * v8.3.122+: 修复路径 — 从 020-specs/{platform}/ 改为 020-specs/{feature}/{platform}/
+ * 与 split 读取路径保持一致，确保端级分析文档能被正确关联
+ */
 async function writePerPlatform(iterDir, report, filename) {
     const reqDir = (0, path_1.join)(iterDir, '010-requirements');
     const specsBase = (0, path_1.join)(iterDir, '020-specs');
@@ -1197,16 +1217,14 @@ async function writePerPlatform(iterDir, report, filename) {
             return;
         // 平台列表从 CONSTITUTION 获取（不再硬编码默认值）
         const platforms = await detectPlatformsFromConstitution();
-        for (const platform of platforms) {
-            const platformDir = (0, path_1.join)(specsBase, platform);
-            await (0, fs_extra_1.ensureDir)(platformDir);
-            // 按端提取差异化内容（而非写入同一份报告）
-            const platformReport = extractPlatformContent(report, platform);
-            await (0, fs_extra_1.writeFile)((0, path_1.join)(platformDir, filename), platformReport);
-            for (const feature of features) {
-                // 每个 feature 文件也按端提取
+        for (const feature of features) {
+            for (const platform of platforms) {
+                // v8.3.122+: 输出到 020-specs/{feature}/{platform}/ 与 split 读取路径一致
+                // v8.3.125+: 使用增量合并，避免多次分析覆盖丢失
+                const platformDir = (0, path_1.join)(specsBase, feature, platform);
+                await (0, fs_extra_1.ensureDir)(platformDir);
                 const featureContent = extractFeatureForPlatform(report, feature, platform);
-                await (0, fs_extra_1.writeFile)((0, path_1.join)(platformDir, `${feature}.md`), featureContent);
+                await writeMergedFile((0, path_1.join)(platformDir, filename), featureContent);
             }
         }
     }
@@ -1264,8 +1282,8 @@ function extractPlatformContent(report, platform) {
 }
 /** 从报告中提取特定功能 × 特定端的内容 */
 function extractFeatureForPlatform(report, feature, platform) {
-    // 先提取该功能的段落
-    const featureKeywords = [feature, feature.replace(/-/g, ' ')];
+    // v8.3.125+: 使用同义词扩展，解决 "登录" 与 "auth" 等跨语言/缩写不匹配问题
+    const featureKeywords = [...(0, synonyms_1.expandSynonyms)((0, synonyms_1.extractNormalizedKeywords)(feature))];
     const lines = report.split('\n');
     const sections = [];
     let current = { heading: '', content: [] };
@@ -1285,9 +1303,10 @@ function extractFeatureForPlatform(report, feature, platform) {
         sections.push({ heading: current.heading, content: current.content });
     }
     // 先按功能名匹配，再按端关键词过滤
+    // v8.3.125+: 使用同义词扩展后的关键词集合匹配
     const featureSections = sections.filter(s => {
         const text = (s.heading + ' ' + s.content.join(' ')).toLowerCase();
-        return featureKeywords.some(kw => text.includes(kw.toLowerCase()));
+        return featureKeywords.some(kw => text.includes(kw));
     });
     if (featureSections.length > 0) {
         const platformKws = PLATFORM_KEYWORDS[platform] || [platform];
@@ -1300,9 +1319,41 @@ function extractFeatureForPlatform(report, feature, platform) {
     }
     return `# ${feature} (${platform})\n\n_该功能在 ${platform} 端暂无独立分析内容，请参考完整报告。_\n`;
 }
+/** 从报告中提取特定功能模块的内容（不限制端，用于模块总览） */
+function extractFeatureContent(report, feature) {
+    // v8.3.125+: 使用同义词扩展
+    const featureKeywords = [...(0, synonyms_1.expandSynonyms)((0, synonyms_1.extractNormalizedKeywords)(feature))];
+    const lines = report.split('\n');
+    const sections = [];
+    let current = { heading: '', content: [] };
+    for (const line of lines) {
+        const hm = line.match(/^(#{1,4})\s+(.+)/);
+        if (hm) {
+            if (current.heading || current.content.length > 0) {
+                sections.push({ heading: current.heading, content: current.content });
+            }
+            current = { heading: line, content: [] };
+        }
+        else {
+            current.content.push(line);
+        }
+    }
+    if (current.heading || current.content.length > 0) {
+        sections.push({ heading: current.heading, content: current.content });
+    }
+    const matched = sections.filter(s => {
+        const text = (s.heading + ' ' + s.content.join(' ')).toLowerCase();
+        return featureKeywords.some(kw => text.includes(kw));
+    });
+    if (matched.length > 0) {
+        return matched.map(s => s.heading + '\n' + s.content.join('\n')).join('\n\n');
+    }
+    return '';
+}
 /**
- * 局部分析：只分析单个功能模块，写入 020-specs/features/{feature}.md
+ * 局部分析：只分析单个功能模块，写入 020-specs/{feature}/overview/
  * 用于 --feature 模式，避免全量重跑
+ * v8.3.122+: 修复路径与规范一致
  */
 async function analyzeSingleFeature(iterDir, featureName) {
     const reqDir = (0, path_1.join)(iterDir, '010-requirements');
@@ -1346,10 +1397,10 @@ async function analyzeSingleFeature(iterDir, featureName) {
         r += `_未检测到明显架构影响_\n\n`;
     }
     r += `## 3. 需求原文\n\n${content}\n`;
-    // 写入 020-specs/features/
-    const featuresDir = (0, path_1.join)(iterDir, '020-specs', 'features');
-    await (0, fs_extra_1.ensureDir)(featuresDir);
-    const outputPath = (0, path_1.join)(featuresDir, `${featureName}.md`);
+    // v8.3.122+: 写入 020-specs/{feature}/overview/ 与规范一致
+    const overviewDir = (0, path_1.join)(iterDir, '020-specs', featureName, 'overview');
+    await (0, fs_extra_1.ensureDir)(overviewDir);
+    const outputPath = (0, path_1.join)(overviewDir, 'ANALYSIS.md');
     await (0, fs_extra_1.writeFile)(outputPath, r);
     return { outputPath, report: r };
 }
@@ -1402,10 +1453,13 @@ async function analyzeSingleTypedDoc(iterDir, docPath) {
     await (0, fs_extra_1.writeFile)(outputPath, r);
     return { outputPath, report: r };
 }
-/** 按功能模块写入 020-specs/features/ */
+/** 按功能模块写入模块总览（提取该功能模块在所有端中的内容）
+ * v8.3.122+: 修复路径 — 从 020-specs/features/{feature}.md 改为 020-specs/{feature}/overview/
+ * 与 split 读取路径保持一致，避免整份报告冗余复制
+ */
 async function writePerFeature(iterDir, report, filename) {
     const reqDir = (0, path_1.join)(iterDir, '010-requirements');
-    const featuresDir = (0, path_1.join)(iterDir, '020-specs', 'features');
+    const specsBase = (0, path_1.join)(iterDir, '020-specs');
     // 非 feature 型目录的排除名单
     const EXCLUDED_DIRS = new Set(['sources', 'assets', 'converted', 'staging', 'bugs', 'refactors', 'research', 'prototypes']);
     try {
@@ -1416,16 +1470,18 @@ async function writePerFeature(iterDir, report, filename) {
             .map(e => e.name);
         if (features.length === 0)
             return;
-        await (0, fs_extra_1.ensureDir)(featuresDir);
         for (const feature of features) {
-            // 读取该 feature 的 README.md 作为头部，拼接完整报告
-            const featureReqPath = (0, path_1.join)(reqDir, feature, 'README.md');
-            let featureHeader = '';
-            if (await (0, fs_extra_1.pathExists)(featureReqPath)) {
-                featureHeader = `# ${feature} — 需求分析\n\n> 来源: 010-requirements/features/${feature}/README.md\n\n---\n\n`;
+            // 提取该功能模块的专属内容（不限制端）
+            let featureContent = extractFeatureContent(report, feature);
+            if (featureContent.trim().length === 0) {
+                // 无匹配内容时回退：写入提示说明
+                featureContent = `# ${feature} — 模块总览\n\n_未从分析报告中提取到该模块的专属内容，请参考完整报告或检查需求文档命名。_\n`;
             }
-            const featureContent = featureHeader + report;
-            await (0, fs_extra_1.writeFile)((0, path_1.join)(featuresDir, `${feature}.md`), featureContent);
+            // v8.3.122+: 输出到 020-specs/{feature}/overview/ 与 split 读取路径一致
+            // v8.3.125+: 使用增量合并，避免多次分析覆盖丢失
+            const overviewDir = (0, path_1.join)(specsBase, feature, 'overview');
+            await (0, fs_extra_1.ensureDir)(overviewDir);
+            await writeMergedFile((0, path_1.join)(overviewDir, filename), featureContent);
         }
     }
     catch { }
@@ -1716,17 +1772,21 @@ async function detectPlatformsFromConstitution() {
             const lines = content.split('\n');
             // ── Layer 1: 表格「对应端」/「对应需求端」列（旧版回退）──
             let headerRowIndex = -1;
-            let headerCells = [];
+            let platformColIdx = -1;
+            let identifierColIdx = -1;
             for (let i = 0; i < lines.length; i++) {
                 const cells = lines[i].split('|').map((c) => c.trim()).filter(Boolean);
-                const platformColIdx = cells.findIndex((c) => c.includes('对应端') || c.includes('对应需求端'));
-                if (platformColIdx >= 0) {
+                const pIdx = cells.findIndex((c) => c.includes('对应端') || c.includes('对应需求端'));
+                if (pIdx >= 0) {
                     headerRowIndex = i;
-                    headerCells = cells;
+                    platformColIdx = pIdx;
+                    identifierColIdx = cells.findIndex((c) => c.includes('工程标识') || c.includes('工程'));
+                    if (identifierColIdx < 0)
+                        identifierColIdx = 0;
                     break;
                 }
             }
-            if (headerRowIndex >= 0) {
+            if (headerRowIndex >= 0 && platformColIdx >= 0) {
                 const platforms = [];
                 const backendPlatforms = [];
                 const seen = new Set();
@@ -1742,8 +1802,8 @@ async function detectPlatformsFromConstitution() {
                     if (!line.startsWith('|'))
                         break;
                     const cells = line.split('|').map((c) => c.trim()).filter(Boolean);
-                    const projectName = cells[0] || '';
-                    const platformChinese = cells[5] || cells[cells.length - 1] || '';
+                    const projectName = identifierColIdx >= 0 && cells.length > identifierColIdx ? cells[identifierColIdx] : '';
+                    const platformChinese = cells.length > platformColIdx ? cells[platformColIdx] : '';
                     // 跳过空值和占位符（如「待填写」）
                     if (!platformChinese || !projectName)
                         continue;
@@ -3057,4 +3117,108 @@ function extractStatusEnumsFromContent(platformContent) {
     }
     return enums;
 }
+// ═══════════════════════════════════════════════
+// v8.3.125+: 增量合并 — 避免多次分析覆盖丢失
+// ═══════════════════════════════════════════════
+/**
+ * 按 Markdown 标题合并文档内容
+ * v8.3.122+: 新内容中存在的章节替换旧章节，新章节追加，旧文档中未变动的章节保留
+ * v8.3.125+: 从 analyze.ts 迁移到 analyze-engine.ts，供 writePerPlatform/writePerFeature 使用
+ */
+function mergeDocumentContent(existing, newContent) {
+    if (!existing.trim())
+        return newContent;
+    // 解析文档为标题块
+    function parseSections(text) {
+        const sections = new Map();
+        const lines = text.split('\n');
+        let currentHeading = '';
+        let currentLevel = 0;
+        let currentBody = [];
+        let headerLines = []; // 文档开头无标题的内容
+        for (const line of lines) {
+            const hm = line.match(/^(#{1,3})\s+(.+)/);
+            if (hm) {
+                if (currentHeading) {
+                    sections.set(currentHeading, {
+                        heading: currentHeading,
+                        level: currentLevel,
+                        body: currentBody.join('\n'),
+                    });
+                }
+                else if (currentBody.length > 0) {
+                    headerLines = [...currentBody];
+                }
+                currentHeading = hm[2].trim();
+                currentLevel = hm[1].length;
+                currentBody = [line];
+            }
+            else {
+                currentBody.push(line);
+            }
+        }
+        if (currentHeading) {
+            sections.set(currentHeading, {
+                heading: currentHeading,
+                level: currentLevel,
+                body: currentBody.join('\n'),
+            });
+        }
+        // 保存文档头部内容（如 frontmatter、标题等）
+        if (headerLines.length > 0) {
+            sections.set('__HEADER__', { heading: '__HEADER__', level: 0, body: headerLines.join('\n') });
+        }
+        return sections;
+    }
+    const oldSections = parseSections(existing);
+    const newSections = parseSections(newContent);
+    // 构建合并结果：保留旧文档的头部，按旧文档顺序输出
+    // 对于每个旧章节：如果新文档有同名章节，用新内容替换；否则保留旧内容
+    // 新文档中有但旧文档中没有的章节，追加到最后
+    const result = [];
+    // 先输出头部
+    const header = oldSections.get('__HEADER__');
+    if (header) {
+        result.push(header.body);
+        oldSections.delete('__HEADER__');
+    }
+    newSections.delete('__HEADER__');
+    // 收集旧文档中存在的章节标题顺序
+    const oldOrder = [...oldSections.keys()];
+    const addedNew = new Set();
+    for (const heading of oldOrder) {
+        if (newSections.has(heading)) {
+            result.push(newSections.get(heading).body);
+            addedNew.add(heading);
+        }
+        else {
+            result.push(oldSections.get(heading).body);
+        }
+    }
+    // 追加新文档中旧文档没有的章节
+    for (const [heading, section] of newSections) {
+        if (!addedNew.has(heading)) {
+            result.push(section.body);
+        }
+    }
+    return result.join('\n\n');
+}
+/**
+ * 安全写入文件：如果文件已存在，先读取并与新内容按章节合并
+ * v8.3.125+: 替代直接 writeFile，避免多次分析覆盖丢失
+ */
+async function writeMergedFile(filePath, newContent) {
+    try {
+        if (await (0, fs_extra_1.pathExists)(filePath)) {
+            const existing = await (0, fs_extra_1.readFile)(filePath, 'utf-8');
+            const merged = mergeDocumentContent(existing, newContent);
+            await (0, fs_extra_1.writeFile)(filePath, merged);
+            return;
+        }
+    }
+    catch {
+        // 读取或合并失败，回退到直接写入
+    }
+    await (0, fs_extra_1.writeFile)(filePath, newContent);
+}
 //# sourceMappingURL=analyze-engine.js.map