@liuhange/dsh-data-sensitivity-classification 2.0.1 → 2.0.2

package/package.json

@@ -1,39 +1,53 @@
-{
-    "name":  "@liuhange/dsh-data-sensitivity-classification",
-    "description":  "Data sensitivity classification plugin: auto-identify sensitivity levels (Public/Internal/Confidential/Secret) and recommend masking strategies",
-    "version": "2.0.1",
-    "publishConfig":  {
-                          "access":  "public"
-                      },
-    "repository":  {
-                       "type":  "git",
-                       "url":  "git+https://github.com/liuhange789/data-asset-inspector.git",
-                       "directory":  "packages/data-asset/data-sensitivity-classification"
-                   },
-    "type":  "module",
-    "main":  "lib/index.js",
-    "types":  "lib/types/index.d.ts",
-    "exports":  {
-                    ".":  {
-                              "types":  "./lib/types/index.d.ts",
-                              "default":  "./lib/index.js"
-                          },
-                    "./invariant":  {
-                                        "types":  "./lib/types/invariant.d.ts",
-                                        "default":  "./lib/invariant.js"
-                                    },
-                    "./src/*":  "./src/*",
-                    "./package.json":  "./package.json"
-                },
-    "files":  [
-                  "lib/index.js",
-                  "lib/invariant.js",
-                  "lib/types/**/*.d.ts"
-              ],
-    "license":  "MIT",
-    "peerDependencies":  {
-                             "@deepseek-ai/cordis":  "^0.1.0",
-                             "@deepseek-ai/dsh-tools":  "^0.1.0",
-                             "@liuhange/dsh-data-asset-shared":  "^2.0.0"
-                         }
-}
+{
+  "name": "@liuhange/dsh-data-sensitivity-classification",
+  "description": "Data sensitivity classification plugin: auto-identify sensitivity levels (Public/Internal/Confidential/Secret) and recommend masking strategies Also known as @deepseek-ai/dsh-data-sensitivity-classification.",
+  "version": "2.0.2",
+  "publishConfig": {
+    "access": "public"
+  },
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/liuhange789/data-asset-inspector.git",
+    "directory": "packages/data-asset/data-sensitivity-classification"
+  },
+  "type": "module",
+  "main": "lib/index.js",
+  "types": "lib/types/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./lib/types/index.d.ts",
+      "default": "./lib/index.js"
+    },
+    "./invariant": {
+      "types": "./lib/types/invariant.d.ts",
+      "default": "./lib/invariant.js"
+    },
+    "./src/*": "./src/*",
+    "./package.json": "./package.json"
+  },
+  "files": [
+    "lib/index.js",
+    "lib/invariant.js",
+    "lib/types/**/*.d.ts"
+  ],
+  "license": "MIT",
+  "scripts": {
+    "typecheck": "tsc --noEmit",
+    "test": "vitest run",
+    "build": "tsc"
+  },
+  "peerDependencies": {
+    "@deepseek-ai/cordis": "^4.0.0",
+    "@deepseek-ai/dsh-tools": "0.0.1-rc.1",
+    "@liuhange/dsh-data-asset-shared": "^2.0.0"
+  },
+  "keywords": [
+    "deepseek-harness",
+    "dsh-plugin",
+    "data-asset",
+    "data-asset-inspector",
+    "shujintong",
+    "数据资产"
+  ],
+  "homepage": "https://github.com/liuhange789/data-asset-inspector#readme"
+}

package/lib/index.js

@@ -1,345 +0,0 @@
-import { defineTool } from "@deepseek-ai/dsh-tools";
-import * as fs from "fs";
-import * as path from "path";
-import { AuditLogger, BusinessRulesLoader, FileFormatAdapter, PathValidator } from "@liuhange/dsh-data-asset-shared";
-//#region lib/types/defaultClassificationRules.js
-const defaultClassificationRules = {
-	fieldNameRules: [
-		{
-			name: "idCard",
-			pattern: "id_?card|身份证|identity",
-			level: "Secret"
-		},
-		{
-			name: "phone",
-			pattern: "phone|手机|mobile|tel",
-			level: "Confidential"
-		},
-		{
-			name: "email",
-			pattern: "email|邮箱|mail",
-			level: "Confidential"
-		},
-		{
-			name: "bankCard",
-			pattern: "bank_?card|银行卡|account",
-			level: "Secret"
-		},
-		{
-			name: "address",
-			pattern: "address|地址|addr",
-			level: "Internal"
-		},
-		{
-			name: "name",
-			pattern: "^name$|姓名|username",
-			level: "Internal"
-		},
-		{
-			name: "public",
-			pattern: "status|type|category|label|tag|level",
-			level: "Public"
-		}
-	],
-	fieldValueRules: [
-		{
-			name: "idCardValue",
-			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: "Secret"
-		},
-		{
-			name: "phoneValue",
-			pattern: "1[3-9]\\d{9}",
-			level: "Confidential"
-		},
-		{
-			name: "emailValue",
-			pattern: "[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\\.[a-zA-Z]{2,}",
-			level: "Confidential"
-		},
-		{
-			name: "bankCardValue",
-			pattern: "[1-9]\\d{15,18}",
-			level: "Secret"
-		}
-	],
-	levelMapping: {
-		Public: "none",
-		Internal: "partial",
-		Confidential: "full",
-		Secret: "encrypt"
-	},
-	defaultLevel: "Internal"
-};
-//#endregion
-//#region lib/types/fieldNameMatcher.js
-var FieldNameMatcher = class {
-	match(fieldName, rules) {
-		const invalidRules = [];
-		for (const rule of rules) try {
-			if (new RegExp(rule.pattern, "i").test(fieldName)) return {
-				result: {
-					matched: true,
-					name: rule.name,
-					pattern: rule.pattern,
-					level: rule.level
-				},
-				invalidRules
-			};
-		} catch {
-			invalidRules.push({
-				ruleName: rule.name,
-				reason: "正则编译失败"
-			});
-		}
-		return {
-			result: { matched: false },
-			invalidRules
-		};
-	}
-};
-//#endregion
-//#region lib/types/fieldValueMatcher.js
-var FieldValueMatcher = class {
-	match(sampleValues, rules) {
-		const invalidRules = [];
-		if (sampleValues.length === 0) return {
-			result: { matched: false },
-			invalidRules
-		};
-		for (const rule of rules) try {
-			const regex = new RegExp(rule.pattern);
-			if (sampleValues.some((v) => v !== null && v !== void 0 && regex.test(String(v)))) return {
-				result: {
-					matched: true,
-					name: rule.name,
-					pattern: rule.pattern,
-					level: rule.level
-				},
-				invalidRules
-			};
-		} catch {
-			invalidRules.push({
-				ruleName: rule.name,
-				reason: "正则编译失败"
-			});
-		}
-		return {
-			result: { matched: false },
-			invalidRules
-		};
-	}
-};
-//#endregion
-//#region lib/types/sensitivityClassifier.js
-var SensitivityClassifier = class {
-	fieldNameMatcher = new FieldNameMatcher();
-	fieldValueMatcher = new FieldValueMatcher();
-	classify(fields, fieldSamples, config) {
-		const classifications = [];
-		const allInvalidRules = [];
-		for (const fieldName of fields) {
-			const samples = fieldSamples[fieldName] ?? [];
-			const nameResult = this.fieldNameMatcher.match(fieldName, config.fieldNameRules);
-			allInvalidRules.push(...nameResult.invalidRules);
-			let level;
-			let identifiedBy;
-			if (nameResult.result.matched) {
-				level = nameResult.result.level ?? config.defaultLevel;
-				identifiedBy = {
-					ruleName: nameResult.result.name ?? "unknown",
-					matchType: "fieldName"
-				};
-			} else {
-				const valueResult = this.fieldValueMatcher.match(samples, config.fieldValueRules);
-				allInvalidRules.push(...valueResult.invalidRules);
-				if (valueResult.result.matched) {
-					level = valueResult.result.level ?? config.defaultLevel;
-					identifiedBy = {
-						ruleName: valueResult.result.name ?? "unknown",
-						matchType: "fieldValue"
-					};
-				} else {
-					level = config.defaultLevel;
-					identifiedBy = {
-						ruleName: "default",
-						matchType: "default"
-					};
-				}
-			}
-			classifications.push({
-				fieldName,
-				sensitivityLevel: level,
-				identifiedBy,
-				recommendedStrategy: "none"
-			});
-		}
-		return {
-			classifications,
-			invalidRules: allInvalidRules
-		};
-	}
-};
-//#endregion
-//#region lib/types/strategyRecommender.js
-var StrategyRecommender = class {
-	recommend(level, levelMapping) {
-		return levelMapping[level] ?? "partial";
-	}
-};
-//#endregion
-//#region lib/types/classificationReportGenerator.js
-var ClassificationReportGenerator = class {
-	async writeJson(filePath, data) {
-		const dir = path.dirname(filePath);
-		if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
-		fs.writeFileSync(filePath, JSON.stringify(data, null, 2), "utf-8");
-	}
-	async writeMarkdown(filePath, data, invalidRules) {
-		const dir = path.dirname(filePath);
-		if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
-		const lines = [];
-		lines.push("【数据敏感度分级报告】");
-		lines.push("");
-		lines.push(`> 分类时间: ${data.classifiedAt}`);
-		lines.push(`> 字段数: ${data.fields.length}`);
-		lines.push("");
-		lines.push("## 字段分级");
-		lines.push("");
-		lines.push("| 字段名 | 敏感度等级 | 识别依据 | 推荐策略 |");
-		lines.push("|--------|-----------|---------|---------|");
-		for (const field of data.fields) lines.push(`| ${field.fieldName} | ${field.sensitivityLevel} | ${field.identifiedBy.matchType}(${field.identifiedBy.ruleName}) | ${field.recommendedStrategy} |`);
-		lines.push("");
-		if (invalidRules.length > 0) {
-			lines.push("## 无效规则");
-			lines.push("");
-			for (const rule of invalidRules) lines.push(`- ${rule.ruleName}: ${rule.reason}`);
-			lines.push("");
-		}
-		lines.push("## 配置状态");
-		lines.push("");
-		lines.push(`- 配置状态: ${data.configStatus}`);
-		lines.push(`- 无效规则数: ${invalidRules.length}`);
-		fs.writeFileSync(filePath, lines.join("\n"), "utf-8");
-	}
-};
-//#endregion
-//#region lib/types/index.js
-const name = "data-sensitivity-classification";
-const inject = ["tools"];
-function parseRecords(lines, format) {
-	if (format === "csv" && lines.length > 0) {
-		const header = lines[0].split(",");
-		const records = [];
-		for (let i = 1; i < lines.length; i++) {
-			const values = lines[i].split(",");
-			const record = {};
-			for (let j = 0; j < header.length; j++) record[header[j]] = values[j] ?? "";
-			records.push(record);
-		}
-		return records;
-	}
-	if (format === "json") {
-		const records = [];
-		for (const line of lines) try {
-			const parsed = JSON.parse(line);
-			if (typeof parsed === "object" && parsed !== null) records.push(parsed);
-		} catch {}
-		return records;
-	}
-	return [];
-}
-function apply(ctx) {
-	const loader = new BusinessRulesLoader();
-	const formatAdapter = new FileFormatAdapter();
-	const pathValidator = new PathValidator();
-	const auditLogger = new AuditLogger();
-	const classifier = new SensitivityClassifier();
-	const recommender = new StrategyRecommender();
-	const reportGenerator = new ClassificationReportGenerator();
-	ctx.tools.register(defineTool({
-		name: "classify_sensitivity",
-		description: "自动识别数据字段敏感度等级（公开/内部/机密/绝密），并推荐相应脱敏策略",
-		parameters: {
-			filePath: {
-				type: "string",
-				required: true,
-				description: "待分级数据文件的路径"
-			},
-			outputPathDir: {
-				type: "string",
-				description: "报告输出目录（默认数据文件所在目录）"
-			},
-			sampleSize: {
-				type: "string",
-				description: "样本大小（默认100）"
-			}
-		},
-		output: {
-			schema: { type: "string" },
-			render: (_args, value) => [{
-				type: "text",
-				text: value
-			}]
-		},
-		async execute(args) {
-			const { config, status: configStatus } = loader.load();
-			const workingDir = process.cwd();
-			const filePath = args.filePath;
-			const outputPathDir = args.outputPathDir ?? path.dirname(path.resolve(filePath));
-			const sampleSize = parseInt(args.sampleSize ?? "100", 10);
-			try {
-				pathValidator.validate(filePath, workingDir);
-			} catch {
-				return `错误：路径不合法 - ${filePath}`;
-			}
-			const fullPath = path.resolve(filePath);
-			if (!fs.existsSync(fullPath)) return `错误：文件不存在 - ${fullPath}`;
-			const classificationConfig = config.sensitivityClassification ?? defaultClassificationRules;
-			const readResult = await formatAdapter.read(fullPath);
-			const records = parseRecords(readResult.lines, readResult.format);
-			const fieldNames = records.length > 0 ? Object.keys(records[0]) : [];
-			const fieldSamples = {};
-			for (const fieldName of fieldNames) {
-				const samples = [];
-				for (let i = 0; i < Math.min(records.length, sampleSize); i++) {
-					const value = records[i][fieldName];
-					if (value !== null && value !== void 0 && String(value).trim() !== "") samples.push(value);
-				}
-				fieldSamples[fieldName] = samples;
-			}
-			const { classifications, invalidRules } = classifier.classify(fieldNames, fieldSamples, classificationConfig);
-			const fieldsWithStrategy = classifications.map((c) => ({
-				...c,
-				recommendedStrategy: recommender.recommend(c.sensitivityLevel, classificationConfig.levelMapping)
-			}));
-			const reportJson = {
-				fields: fieldsWithStrategy,
-				configStatus,
-				classifiedAt: (/* @__PURE__ */ new Date()).toISOString()
-			};
-			const jsonReportPath = path.join(outputPathDir, "classification_report.json");
-			const markdownReportPath = path.join(outputPathDir, "classification_report.md");
-			await reportGenerator.writeJson(jsonReportPath, reportJson);
-			await reportGenerator.writeMarkdown(markdownReportPath, reportJson, invalidRules);
-			auditLogger.log({
-				pluginName: "data-sensitivity-classification",
-				operation: "classify_sensitivity",
-				inputPath: fullPath,
-				outputPath: jsonReportPath,
-				result: "SUCCESS"
-			});
-			return JSON.stringify({
-				jsonReportPath,
-				markdownReportPath,
-				report: reportJson,
-				fields: fieldsWithStrategy,
-				status: "SUCCESS",
-				configStatus
-			}, null, 2);
-		}
-	}));
-	console.log("[data-sensitivity-classification] 数据敏感度分级插件已加载");
-}
-//#endregion
-export { apply, inject, name };

package/lib/invariant.js

@@ -1,5 +0,0 @@
-//#region lib/types/invariant.js
-const invariant = "data-sensitivity-classification";
-function install() {}
-//#endregion
-export { install, invariant };

package/lib/types/classificationReportGenerator.d.ts

@@ -1,6 +0,0 @@
-import type { ClassificationReportJson, InvalidRule } from './types.js';
-export declare class ClassificationReportGenerator {
-    writeJson(filePath: string, data: ClassificationReportJson): Promise<void>;
-    writeMarkdown(filePath: string, data: ClassificationReportJson, invalidRules: InvalidRule[]): Promise<void>;
-}
-//# sourceMappingURL=classificationReportGenerator.d.ts.map

package/lib/types/defaultClassificationRules.d.ts

@@ -1,3 +0,0 @@
-import type { SensitivityClassificationConfig } from '@liuhange/dsh-data-asset-shared';
-export declare const defaultClassificationRules: SensitivityClassificationConfig;
-//# sourceMappingURL=defaultClassificationRules.d.ts.map

package/lib/types/fieldNameMatcher.d.ts

@@ -1,9 +0,0 @@
-import type { SensitivityClassificationConfig } from '@liuhange/dsh-data-asset-shared';
-import type { MatchResult, InvalidRule } from './types.js';
-export declare class FieldNameMatcher {
-    match(fieldName: string, rules: SensitivityClassificationConfig['fieldNameRules']): {
-        result: MatchResult;
-        invalidRules: InvalidRule[];
-    };
-}
-//# sourceMappingURL=fieldNameMatcher.d.ts.map

package/lib/types/fieldValueMatcher.d.ts

@@ -1,9 +0,0 @@
-import type { SensitivityClassificationConfig } from '@liuhange/dsh-data-asset-shared';
-import type { MatchResult, InvalidRule } from './types.js';
-export declare class FieldValueMatcher {
-    match(sampleValues: unknown[], rules: SensitivityClassificationConfig['fieldValueRules']): {
-        result: MatchResult;
-        invalidRules: InvalidRule[];
-    };
-}
-//# sourceMappingURL=fieldValueMatcher.d.ts.map

package/lib/types/index.d.ts

@@ -1,5 +0,0 @@
-import type { Context } from '@deepseek-ai/cordis';
-export declare const name = "data-sensitivity-classification";
-export declare const inject: string[];
-export declare function apply(ctx: Context): void;
-//# sourceMappingURL=index.d.ts.map

package/lib/types/invariant.d.ts

@@ -1,3 +0,0 @@
-export declare const invariant = "data-sensitivity-classification";
-export declare function install(): void;
-//# sourceMappingURL=invariant.d.ts.map

package/lib/types/sensitivityClassifier.d.ts

@@ -1,11 +0,0 @@
-import type { SensitivityClassificationConfig } from '@liuhange/dsh-data-asset-shared';
-import type { FieldClassification, InvalidRule } from './types.js';
-export declare class SensitivityClassifier {
-    private fieldNameMatcher;
-    private fieldValueMatcher;
-    classify(fields: string[], fieldSamples: Record<string, unknown[]>, config: SensitivityClassificationConfig): {
-        classifications: FieldClassification[];
-        invalidRules: InvalidRule[];
-    };
-}
-//# sourceMappingURL=sensitivityClassifier.d.ts.map

package/lib/types/strategyRecommender.d.ts

@@ -1,6 +0,0 @@
-import type { SensitivityClassificationConfig } from '@liuhange/dsh-data-asset-shared';
-import type { SensitivityLevel, RecommendedStrategy } from './types.js';
-export declare class StrategyRecommender {
-    recommend(level: SensitivityLevel, levelMapping: SensitivityClassificationConfig['levelMapping']): RecommendedStrategy;
-}
-//# sourceMappingURL=strategyRecommender.d.ts.map

package/lib/types/types.d.ts

@@ -1,40 +0,0 @@
-export type SensitivityLevel = 'Public' | 'Internal' | 'Confidential' | 'Secret';
-export type RecommendedStrategy = 'none' | 'partial' | 'full' | 'encrypt';
-export interface ClassifySensitivityParams {
-    filePath: string;
-    outputPathDir?: string;
-    sampleSize?: number;
-}
-export interface MatchResult {
-    matched: boolean;
-    name?: string;
-    pattern?: string;
-    level?: string;
-}
-export interface InvalidRule {
-    ruleName: string;
-    reason: string;
-}
-export interface FieldClassification {
-    fieldName: string;
-    sensitivityLevel: SensitivityLevel;
-    identifiedBy: {
-        ruleName: string;
-        matchType: string;
-    };
-    recommendedStrategy: RecommendedStrategy;
-}
-export interface ClassificationReportJson {
-    fields: FieldClassification[];
-    configStatus: string;
-    classifiedAt: string;
-}
-export interface ClassifySensitivityResult {
-    jsonReportPath: string;
-    markdownReportPath: string;
-    report: ClassificationReportJson;
-    fields: FieldClassification[];
-    status: 'SUCCESS' | 'FAILED';
-    configStatus: string;
-}
-//# sourceMappingURL=types.d.ts.map