@liuhange/dsh-data-asset-shared 2.0.0 → 2.0.2

package/README.md

@@ -1,4 +1,4 @@
-# @deepseek-ai/dsh-data-asset-shared
+# @liuhange/dsh-data-asset-shared
 
 Shared infrastructure for data assetization plugins.
 
@@ -13,7 +13,7 @@ Shared infrastructure for data assetization plugins.
 ## Usage
 
 ```typescript
-import { BusinessRulesLoader, FileFormatAdapter, ReportGenerator, PathValidator, AuditLogger } from '@deepseek-ai/dsh-data-asset-shared'
+import { BusinessRulesLoader, FileFormatAdapter, ReportGenerator, PathValidator, AuditLogger } from '@liuhange/dsh-data-asset-shared'
 
 const loader = new BusinessRulesLoader()
 const { config, status } = loader.load()

package/lib/index.js

@@ -1,415 +1,9 @@
-import * as fs from "fs";
-import * as path from "path";
-//#region lib/types/defaultBusinessRules.js
-const defaultBusinessRules = {
-	version: "1.0",
-	lastUpdated: "2026-08-17",
-	valueAssessment: {
-		highValue: {
-			keywords: [
-				"供应链",
-				"销售明细",
-				"设备工况",
-				"交易流水"
-			],
-			score: 5,
-			label: "★★★★★",
-			recommendation: "优先交易，建议入表"
-		},
-		mediumValue: {
-			keywords: [
-				"用户行为",
-				"产品数据",
-				"行业报告"
-			],
-			score: 3,
-			label: "★★★",
-			recommendation: "脱敏后交易"
-		},
-		lowValue: {
-			keywords: [
-				"日志",
-				"备份",
-				"临时"
-			],
-			score: 1,
-			label: "★",
-			recommendation: "评估处理成本"
-		},
-		defaultValue: {
-			score: 3,
-			label: "★★★",
-			recommendation: "需进一步评估"
-		}
-	},
-	sensitivePatterns: {
-		idCard: {
-			pattern: "[1-9]\\d{5}(18|19|20)?\\d{2}(0[1-9]|1[0-2])(0[1-9]|[12]\\d|3[01])\\d{3}[\\dXx]",
-			level: "FULL"
-		},
-		phone: {
-			pattern: "1[3-9]\\d{9}",
-			level: "PARTIAL"
-		},
-		bankCard: {
-			pattern: "[1-9]\\d{15,18}",
-			level: "PARTIAL"
-		},
-		email: {
-			pattern: "[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\\.[a-zA-Z]{2,}",
-			level: "PARTIAL"
-		}
-	},
-	maskingLevels: {
-		FULL: {
-			description: "完全匿名化",
-			applyTo: ["idCard"]
-		},
-		PARTIAL: {
-			description: "假名化，保留前3后4",
-			applyTo: ["phone", "bankCard"]
-		},
-		GENERALIZE: {
-			description: "泛化处理",
-			applyTo: ["email"]
-		}
-	},
-	cleaningRules: {
-		missingValueStrategy: {
-			government: "interpolation",
-			financial: "manual_review",
-			behavior: "mode_fill"
-		},
-		qualityThreshold: {
-			A: {
-				missingRate: .05,
-				errorRate: .01,
-				label: "优秀"
-			},
-			B: {
-				missingRate: .1,
-				errorRate: .03,
-				label: "良好"
-			},
-			C: {
-				missingRate: .2,
-				errorRate: .05,
-				label: "合格"
-			},
-			D: {
-				missingRate: .2,
-				errorRate: .05,
-				label: "不合格",
-				action: "不建议交易"
-			}
-		}
-	},
-	compliance: { policies: [
-		{
-			name: "数据产权登记工作指引",
-			doc: "国数综政策〔2026〕35号",
-			check: "数据来源可登记"
-		},
-		{
-			name: "数据流通安全治理",
-			doc: "发改数据〔2025〕18号",
-			check: "来源合法性"
-		},
-		{
-			name: "数据脱敏工具标准",
-			doc: "YD/T 6225-2024",
-			check: "脱敏效果达标"
-		}
-	] },
-	packaging: {
-		complianceStatements: [
-			"本数据产品已完成脱敏处理",
-			"不包含个人身份信息",
-			"数据来源合法",
-			"建议交易前完成数据产权登记"
-		],
-		pricingRules: [
-			{
-				valueLevel: "★★★★★",
-				suggestion: "根据数据质量、稀缺性和应用场景综合定价，建议高价交易"
-			},
-			{
-				valueLevel: "★★★★",
-				suggestion: "根据数据质量和应用场景定价，建议中高价交易"
-			},
-			{
-				valueLevel: "★★★",
-				suggestion: "根据数据质量和应用场景综合定价"
-			},
-			{
-				valueLevel: "★★",
-				suggestion: "建议低价交易或内部使用"
-			},
-			{
-				valueLevel: "★",
-				suggestion: "建议评估处理成本后决定是否交易"
-			}
-		],
-		defaultDescription: "适用于数据分析、模型训练、行业研究等场景",
-		version: "V1.0"
-	}
-};
-//#endregion
-//#region lib/types/businessRulesLoader.js
-var BusinessRulesLoader = class {
-	cachedResult = null;
-	configPath;
-	constructor(configPath) {
-		this.configPath = configPath ? path.resolve(configPath) : path.resolve(process.cwd(), "config", "business-rules.json");
-	}
-	load() {
-		if (this.cachedResult) return this.cachedResult;
-		const result = this.loadInternal();
-		this.cachedResult = result;
-		return result;
-	}
-	loadInternal() {
-		if (!fs.existsSync(this.configPath)) return this.createDefaultResult("DEFAULT_MISSING");
-		let raw;
-		try {
-			raw = fs.readFileSync(this.configPath, "utf-8");
-		} catch {
-			return this.createDefaultResult("DEFAULT_PARSE");
-		}
-		let parsed;
-		try {
-			parsed = JSON.parse(raw);
-		} catch {
-			return this.createDefaultResult("DEFAULT_PARSE");
-		}
-		if (!this.isBusinessRulesConfig(parsed)) return this.createDefaultResult("DEFAULT_PARSE");
-		if (!this.isVersionCompatible(parsed.version)) return this.createDefaultResult("DEFAULT_VERSION");
-		if (!this.hasRequiredNodes(parsed)) return this.createDefaultResult("DEFAULT_PARTIAL");
-		return {
-			config: parsed,
-			status: "CONFIG_LOADED"
-		};
-	}
-	createDefaultResult(status) {
-		return {
-			config: defaultBusinessRules,
-			status
-		};
-	}
-	isBusinessRulesConfig(value) {
-		return typeof value === "object" && value !== null && "version" in value && "valueAssessment" in value && "sensitivePatterns" in value && "cleaningRules" in value && "compliance" in value && "packaging" in value;
-	}
-	isVersionCompatible(version) {
-		if (typeof version !== "string") return false;
-		return version.startsWith("1.");
-	}
-	hasRequiredNodes(config) {
-		return config.valueAssessment !== void 0 && config.sensitivePatterns !== void 0 && config.maskingLevels !== void 0 && config.cleaningRules !== void 0 && config.compliance !== void 0 && config.packaging !== void 0;
-	}
-	clearCache() {
-		this.cachedResult = null;
-	}
-};
-//#endregion
-//#region lib/types/fileFormatAdapter.js
-var FileFormatAdapter = class {
-	detectFormat(filePath) {
-		const ext = path.extname(filePath).toLowerCase().slice(1);
-		if (ext === "csv" || ext === "json" || ext === "txt" || ext === "xlsx") return ext;
-		return "unknown";
-	}
-	async read(filePath) {
-		const format = this.detectFormat(filePath);
-		const fullPath = path.resolve(filePath);
-		if (!fs.existsSync(fullPath)) throw new Error(`错误：文件不存在 - ${fullPath}`);
-		const isLargeFile = fs.statSync(fullPath).size >= 10 * 1024 * 1024;
-		switch (format) {
-			case "csv": return this.readCsv(fullPath);
-			case "json": return this.readJson(fullPath);
-			case "xlsx": return this.readXlsx(fullPath);
-			default: return this.readTxt(fullPath, format, isLargeFile);
-		}
-	}
-	async write(filePath, lines, format) {
-		const fullPath = path.resolve(filePath);
-		const dir = path.dirname(fullPath);
-		if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
-		switch (format) {
-			case "csv":
-				this.writeCsv(fullPath, lines);
-				break;
-			case "json":
-				this.writeJson(fullPath, lines);
-				break;
-			case "xlsx":
-				this.writeXlsx(fullPath, lines);
-				break;
-			default:
-				fs.writeFileSync(fullPath, lines.join("\n"), "utf-8");
-				break;
-		}
-	}
-	readTxt(filePath, format, _isLargeFile) {
-		const content = fs.readFileSync(filePath, "utf-8");
-		return {
-			lines: content.split("\n").filter((line) => line.trim()),
-			format,
-			raw: content
-		};
-	}
-	readCsv(filePath) {
-		const content = fs.readFileSync(filePath, "utf-8");
-		return {
-			lines: content.split("\n").filter((line) => line.trim()),
-			format: "csv",
-			raw: content
-		};
-	}
-	readJson(filePath) {
-		const content = fs.readFileSync(filePath, "utf-8");
-		const parsed = JSON.parse(content);
-		if (Array.isArray(parsed)) return {
-			lines: parsed.map((item) => JSON.stringify(item)),
-			format: "json",
-			raw: parsed
-		};
-		return {
-			lines: [JSON.stringify(parsed)],
-			format: "json",
-			raw: parsed
-		};
-	}
-	readXlsx(filePath) {
-		const content = fs.readFileSync(filePath, "utf-8");
-		return {
-			lines: content.split("\n").filter((line) => line.trim()),
-			format: "xlsx",
-			raw: content
-		};
-	}
-	writeCsv(filePath, lines) {
-		fs.writeFileSync(filePath, lines.join("\n"), "utf-8");
-	}
-	writeJson(filePath, lines) {
-		const objects = lines.map((line) => {
-			try {
-				return JSON.parse(line);
-			} catch {
-				return line;
-			}
-		});
-		fs.writeFileSync(filePath, JSON.stringify(objects, null, 2), "utf-8");
-	}
-	writeXlsx(filePath, lines) {
-		fs.writeFileSync(filePath, lines.join("\n"), "utf-8");
-	}
-};
-//#endregion
-//#region lib/types/reportGenerator.js
-var ReportGenerator = class {
-	generateMaskingReport(params) {
-		const findingsText = Object.keys(params.fieldTypeCounts).length > 0 ? Object.entries(params.fieldTypeCounts).map(([type, count]) => `发现 ${count} 个${type}类型的敏感数据`).join("；") : "无";
-		return `【脱敏报告】
-- 处理文件：${params.inputPath}
-- 发现敏感字段：${findingsText}
-- 脱敏策略：${params.strategy}
-- 输出文件：${params.outputPath}
-- 状态：✅ 完成`;
-	}
-	generateCleaningReport(params) {
-		let report = `【数据清洗报告】
-- 文件：${params.inputPath}
-- 原始行数：${params.originalLineCount}
-- 去重后行数：${params.originalLineCount - params.duplicateRemoved}（删除 ${params.duplicateRemoved} 行重复）
-`;
-		if (params.standardizeApplied) report += `- 格式标准化：✅ 完成\n`;
-		report += `- 异常值检测：发现 ${params.anomalyTotalCount} 个异常\n`;
-		if (params.anomalies.length > 0) {
-			const anomalyText = params.anomalies.slice(0, 5).map((a) => `行${a.line}: ${a.reason}`).join("；");
-			report += `  异常详情：${anomalyText}`;
-			if (params.anomalyTotalCount > 5) report += `...等${params.anomalyTotalCount}项`;
-			report += `\n`;
-		}
-		report += `- 输出文件：${params.outputPath}\n- 状态：✅ 完成`;
-		return report;
-	}
-	generateInventoryReport(params) {
-		let report = `【数据资产盘点报告】\n- 目录：${params.directory}\n- 发现数据文件：${params.fileCount} 个\n\n`;
-		report += `| 文件名 | 大小 | 类型 | 价值评估 | 说明 |\n`;
-		report += `|--------|------|------|----------|------|\n`;
-		for (const item of params.assetItems) report += `| ${item.fileName} | ${item.size} | ${item.type} | ${item.valueAssessment} | ${item.description} |\n`;
-		report += `\n---\n【建议】\n`;
-		report += `1. 高价值数据（★★★★★）建议优先进行清洗和脱敏后交易\n`;
-		report += `2. 低价值数据建议评估是否值得投入处理成本\n`;
-		report += `3. 所有数据交易前必须完成合规审查\n`;
-		return report;
-	}
-	generatePackagingManual(params) {
-		const manual = params.productManual;
-		const complianceText = manual.complianceStatements.map((stmt) => `✅ ${stmt}`).join("\n");
-		return `========================================
-        数据产品说明书
-========================================
-
-【产品名称】${manual.productName}
-【版本】${manual.version}
-【生成日期】${manual.generatedDate}
-
-【数据概览】
-- 源文件：${manual.dataOverview.sourceFile}
-- 数据量：${manual.dataOverview.recordCount} 条记录
-- 文件大小：${manual.dataOverview.fileSizeKb.toFixed(1)} KB
-- 格式：${manual.dataOverview.format}
-
-【数据样例】（前5行）
-${manual.sampleLines.join("\n")}
-
-【使用场景】
-${manual.usageScenario}
-
-【合规声明】
-${complianceText}
-
-【定价建议】
-${manual.pricingSuggestion}
-
---- 本产品由数据资产化工具包自动生成 ---`;
-	}
-};
-//#endregion
-//#region lib/types/pathValidator.js
-var PathValidator = class {
-	validate(filePath, workingDir) {
-		const resolved = path.resolve(filePath);
-		if (filePath.includes("..")) throw new Error(`错误：路径不合法 - ${filePath}`);
-		const relative = path.relative(workingDir, resolved);
-		if (relative.startsWith("..") || path.isAbsolute(relative)) throw new Error(`错误：路径不合法 - ${filePath}`);
-		return resolved;
-	}
-	validateDirectory(dirPath, workingDir) {
-		return this.validate(dirPath, workingDir);
-	}
-};
-//#endregion
-//#region lib/types/auditLogger.js
-var AuditLogger = class {
-	logDir;
-	constructor(logDir) {
-		this.logDir = logDir ?? path.resolve(process.cwd(), "logs");
-	}
-	log(params) {
-		const timestamp = (/* @__PURE__ */ new Date()).toISOString();
-		const operator = params.operator ?? "system";
-		const logLine = `[${params.pluginName}] ${timestamp} | 操作: ${params.operation} | 操作人: ${operator} | 输入: ${params.inputPath} | 输出: ${params.outputPath} | 结果: ${params.result}`;
-		console.log(logLine);
-		if (params.writeToFile) this.writeToFile(logLine);
-	}
-	writeToFile(logLine) {
-		if (!fs.existsSync(this.logDir)) fs.mkdirSync(this.logDir, { recursive: true });
-		const date = (/* @__PURE__ */ new Date()).toISOString().split("T")[0];
-		const logFile = path.join(this.logDir, `audit-${date}.log`);
-		fs.appendFileSync(logFile, logLine + "\n", "utf-8");
-	}
-};
-//#endregion
-export { AuditLogger, BusinessRulesLoader, FileFormatAdapter, PathValidator, ReportGenerator, defaultBusinessRules };
+export * from './types.js';
+export { defaultBusinessRules } from './defaultBusinessRules.js';
+export { defaultQualityScoringConfig, defaultLineageConfig, defaultAdvancedMaskingConfig, defaultVisualizationConfig, defaultSensitivityClassificationConfig, defaultIncrementalSchedulingConfig, } from './defaultBusinessRules.js';
+export { BusinessRulesLoader } from './businessRulesLoader.js';
+export { FileFormatAdapter } from './fileFormatAdapter.js';
+export { ReportGenerator } from './reportGenerator.js';
+export { PathValidator } from './pathValidator.js';
+export { AuditLogger } from './auditLogger.js';
+//# sourceMappingURL=index.js.map

package/lib/invariant.js

@@ -1,5 +1,4 @@
-//#region lib/types/invariant.js
-const invariant = "data-asset-shared";
-function install() {}
-//#endregion
-export { install, invariant };
+export const invariant = 'data-asset-shared';
+export function install() {
+}
+//# sourceMappingURL=invariant.js.map

package/package.json

@@ -1,13 +1,13 @@
-{
+{
     "name":  "@liuhange/dsh-data-asset-shared",
     "description":  "Shared infrastructure for data assetization plugins: business rules loader, file format adapter, report generator, path validator, audit logger",
-    "version":  "2.0.0",
+    "version": "2.0.2",
     "publishConfig":  {
                           "access":  "public"
                       },
     "repository":  {
                        "type":  "git",
-                       "url":  "git+https://github.com/deepseek-ai/deepseek-harness.git",
+                       "url":  "git+https://github.com/liuhange789/data-asset-inspector.git",
                        "directory":  "packages/data-asset/data-asset-shared"
                    },
     "type":  "module",
@@ -31,12 +31,17 @@
                   "lib/types/**/*.d.ts"
               ],
     "license":  "MIT",
+    "scripts": {
+      "typecheck": "tsc --noEmit",
+      "test": "vitest run",
+      "build": "tsc"
+    },
     "dependencies":  {
                          "csv-parse":  "^5.5.0"
                      },
     "peerDependencies":  {
-                             "@deepseek-ai/cordis":  "^0.1.0",
-                             "@deepseek-ai/dsh-tools":  "^0.1.0",
-                             "@deepseek-ai/schemastery":  "^0.1.0"
+                             "@deepseek-ai/cordis": "^4.0.0",
+                             "@deepseek-ai/dsh-tools": "0.0.1-rc.1",
+                             "@deepseek-ai/schemastery": "^3.18.0"
                          }
 }