@nsyan/db 1.0.0

package/src/core/scan/parsers.ts

@@ -0,0 +1,149 @@
+// scan/parsers.ts —— 按文件类型解析：.env / .properties / yml+yaml（Spring 结构优先）/ docker-compose.yml / 通用 URL 正则
+
+export function parseEnv(text: string): Record<string, string> {
+  const out: Record<string, string> = {};
+  for (const line of text.split("\n")) {
+    const t = line.trim();
+    if (!t || t.startsWith("#")) continue;
+    const eq = t.indexOf("=");
+    if (eq <= 0) continue;
+    const key = t.slice(0, eq).trim();
+    let val = t.slice(eq + 1).trim();
+    if ((val.startsWith('"') && val.endsWith('"')) || (val.startsWith("'") && val.endsWith("'"))) {
+      val = val.slice(1, -1);
+    }
+    out[key] = val;
+  }
+  return out;
+}
+
+export function parseProperties(text: string): Record<string, string> {
+  const out: Record<string, string> = {};
+  for (const line of text.split("\n")) {
+    const t = line.trim();
+    if (!t || t.startsWith("#") || t.startsWith("!")) continue;
+    const m = /^([^=:!]+)[=:](.*)$/.exec(t);
+    if (!m) continue;
+    out[m[1].trim()] = m[2].trim();
+  }
+  return out;
+}
+
+/**
+ * 极简 YAML 子集解析：嵌套 map + 标量（够 Spring 配置用）。
+ * 不支持多行标量/锚点/复杂流式结构——遇到即跳过该行。
+ */
+export function parseSimpleYaml(text: string): Record<string, unknown> {
+  interface YamlLine { indent: number; key: string; value: string | null; }
+  const lines: YamlLine[] = [];
+  for (const raw of text.split("\n")) {
+    if (!raw.trim() || raw.trim().startsWith("#")) continue;
+    if (raw.trim().startsWith("- ")) continue; // 列表项：Spring 配置子树用不到，交给 compose 解析器
+    const indent = raw.length - raw.trimStart().length;
+    const content = raw.trim();
+    const colon = content.indexOf(":");
+    if (colon < 0) continue; // 多行标量等不支持，跳过
+    const key = content.slice(0, colon).trim().replace(/^["']|["']$/g, "");
+    let value: string | null = content.slice(colon + 1).trim();
+    if (value === "" || value === "|" || value === ">") value = null;
+    else value = value.replace(/^["']|["']$/g, "");
+    lines.push({ indent, key, value });
+  }
+  const root: Record<string, unknown> = {};
+  const stack: { indent: number; obj: Record<string, unknown> }[] = [{ indent: -1, obj: root }];
+  for (const ln of lines) {
+    while (stack.length > 1 && ln.indent <= stack[stack.length - 1].indent) stack.pop();
+    const parent = stack[stack.length - 1].obj;
+    if (ln.value !== null) {
+      parent[ln.key] = ln.value;
+    } else {
+      const child: Record<string, unknown> = {};
+      parent[ln.key] = child;
+      stack.push({ indent: ln.indent, obj: child });
+    }
+  }
+  return root;
+}
+
+// ── docker-compose.yml ────────────────────────────
+
+export interface ComposeService {
+  name: string;
+  image?: string;
+  env: Record<string, string>;
+  ports: string[];
+}
+
+export function parseCompose(text: string): ComposeService[] {
+  const services: ComposeService[] = [];
+  let cur: ComposeService | null = null;
+  let block: { kind: "environment" | "ports"; indent: number } | null = null;
+  let servicesIndent = -1;
+
+  for (const raw of text.split("\n")) {
+    const t = raw.trim();
+    if (!t || t.startsWith("#")) continue;
+    const indent = raw.length - raw.trimStart().length;
+
+    if (t.startsWith("- ")) {
+      if (!cur || !block || indent <= block.indent) continue;
+      const item = t.slice(2).trim().replace(/^["']|["']$/g, "");
+      if (block.kind === "ports") {
+        cur.ports.push(item);
+      } else {
+        const eq = item.indexOf("=");
+        if (eq > 0) cur.env[item.slice(0, eq).trim()] = item.slice(eq + 1).trim();
+      }
+      continue;
+    }
+
+    if (block && indent <= block.indent) block = null;
+
+    const colon = t.indexOf(":");
+    if (colon < 0) continue;
+    const key = t.slice(0, colon).trim().replace(/^["']|["']$/g, "");
+    const value = t.slice(colon + 1).trim();
+
+    if (key === "services" && value === "") {
+      servicesIndent = indent;
+      cur = null;
+      block = null;
+      continue;
+    }
+    // service 名行：services 的直接子级（缩进 = servicesIndent + 2）且无内联值
+    if (servicesIndent >= 0 && indent === servicesIndent + 2 && value === "") {
+      cur = { name: key, image: undefined, env: {}, ports: [] };
+      services.push(cur);
+      block = null;
+      continue;
+    }
+    if (cur) {
+      if (key === "image" && value) {
+        cur.image = value.replace(/^["']|["']$/g, "");
+      } else if (key === "environment") {
+        if (value === "") block = { kind: "environment", indent };
+        else {
+          const eq = value.indexOf("=");
+          if (eq > 0) cur.env[value.slice(0, eq).trim()] = value.slice(eq + 1).trim().replace(/^["']|["']$/g, "");
+        }
+      } else if (key === "ports") {
+        block = value === "" ? { kind: "ports", indent } : null;
+      } else if (block?.kind === "environment" && indent > block.indent && value !== "") {
+        cur.env[key] = value.replace(/^["']|["']$/g, "");
+      }
+    }
+  }
+  return services;
+}
+
+// ── 通用 URL 正则（全文件扫描兜底）──────────────────
+
+const URL_RE = /(?:jdbc:(?:postgresql|mysql|oracle|dm|hive2)|rediss?|postgresql|mysql):\/\/[^\s"'<>`]+|https?:\/\/[^\s"'<>`]*:9200[^\s"'<>`]*/g;
+
+export function extractUrls(text: string): string[] {
+  const out = new Set<string>();
+  for (const m of text.matchAll(URL_RE)) {
+    out.add(m[0].replace(/[),.;\]]+$/, ""));
+  }
+  return [...out];
+}

package/src/core/scan/placeholders.ts

@@ -0,0 +1,24 @@
+// scan/placeholders.ts —— Spring 占位符 ${KEY:default} 解析：default → 同目录 .env → 进程 env → 标「待补」
+const PH_RE = /^\$\{([^:}]+)(?::((?:.|\n)*))?\}$/;
+
+export interface PlaceholderResult {
+  resolved: boolean;
+  value?: string;
+}
+
+export function resolvePlaceholder(value: string, localEnv?: Record<string, string>): PlaceholderResult {
+  const m = PH_RE.exec(value.trim());
+  if (!m) return { resolved: true, value };
+  const key = m[1];
+  const hasDefault = m[2] !== undefined;
+
+  // 有非空 default → 直接用
+  if (hasDefault && m[2] !== "") return { resolved: true, value: m[2] };
+
+  // 无 default 或 default 为空串：查同目录 .env → 进程 env
+  const fromEnv = localEnv?.[key] ?? process.env[key];
+  if (fromEnv !== undefined && fromEnv !== "") return { resolved: true, value: fromEnv };
+
+  // 缺省：标待补（空串 default 同样视为待补，如 ${REDIS_PASSWORD:}）
+  return { resolved: false };
+}

package/src/core/scan/scoring.ts

@@ -0,0 +1,25 @@
+// scan/scoring.ts —— 置信度：docker-compose/.env/application.yml 高权重；*test*/*example*/*.md/logs 降权或排除
+// 注：排除/降权只看相对扫描根的路径——固件本身在 test/ 下不受影响。
+
+const SOURCE_WEIGHT: Array<[RegExp, number]> = [
+  [/application[^/]*\.ya?ml$/i, 0.9],
+  [/docker-compose[^/]*\.ya?ml$/i, 0.85],
+  [/^docker-compose[^/]*\.ya?ml$/i, 0.85],
+  [/\.env(\.[^/]*)?$/i, 0.8],
+  [/application[^/]*\.properties$/i, 0.7],
+  [/\.properties$/i, 0.6],
+];
+
+const FALLBACK_WEIGHT = 0.5; // 通用 URL 正则扫出的候选
+
+/** 命中测试/示例/文档/日志路径的文件整体排除，不产候选。 */
+export function isExcluded(relPath: string): boolean {
+  return /(^|\/)(test|tests|spec|specs|example|examples|docs?)(\/|$)/i.test(relPath)
+    || /\.md$/i.test(relPath)
+    || /\.(log|bak)$/i.test(relPath);
+}
+
+export function sourceWeight(relPath: string): number {
+  for (const [re, w] of SOURCE_WEIGHT) if (re.test(relPath)) return w;
+  return FALLBACK_WEIGHT;
+}

package/src/core/scan/spring.ts

@@ -0,0 +1,73 @@
+// scan/spring.ts —— Spring 专属：datasource / data.redis / elasticsearch 键映射 + profile 分组
+import type { DbTypeId } from "../types.js";
+
+export type SpringGroup = "datasource" | "redis" | "es";
+
+export interface RawDbConfig {
+  group: SpringGroup;
+  dialectId?: DbTypeId;   // 有 URL 时由 candidates 层经 registry 判定
+  url?: string;
+  host?: string;
+  port?: string;
+  username?: string;
+  password?: string;
+  database?: string;
+  dbIndex?: string;
+  profile: string;
+  file: string;
+  weight: number;         // scoring 来源权重
+}
+
+export function profileOf(filename: string): string {
+  const m = /application-([A-Za-z0-9_-]+)\.(?:yml|yaml|properties)$/i.exec(filename);
+  return m ? m[1] : "default";
+}
+
+type Field = "url" | "username" | "password" | "host" | "port" | "database" | "dbIndex";
+
+const SPRING_KEYS: Array<[string, Field]> = [
+  ["spring.datasource.url", "url"],
+  ["spring.datasource.jdbc-url", "url"],
+  ["spring.datasource.username", "username"],
+  ["spring.datasource.password", "password"],
+  ["spring.data.redis.host", "host"],
+  ["spring.data.redis.port", "port"],
+  ["spring.data.redis.password", "password"],
+  ["spring.data.redis.database", "dbIndex"],
+  ["spring.redis.host", "host"],           // Spring Boot 2.x 旧前缀
+  ["spring.redis.port", "port"],
+  ["spring.redis.password", "password"],
+  ["spring.redis.database", "dbIndex"],
+  ["spring.elasticsearch.uris", "url"],
+  ["spring.elasticsearch.url", "url"],
+  ["spring.elasticsearch.username", "username"],
+  ["spring.elasticsearch.password", "password"],
+  ["spring.data.elasticsearch.uris", "url"],
+  ["spring.data.elasticsearch.username", "username"],
+  ["spring.data.elasticsearch.password", "password"],
+];
+
+function groupOf(key: string): SpringGroup {
+  if (key.includes("redis")) return "redis";
+  if (key.includes("elasticsearch")) return "es";
+  return "datasource";
+}
+
+/** 扁平化 dotted 键 → 按组聚合成 RawDbConfig[] */
+export function springKeysToRaw(dotted: Record<string, string>, file: string, weight: number): RawDbConfig[] {
+  const groups = new Map<SpringGroup, RawDbConfig>();
+  for (const [key, field] of SPRING_KEYS) {
+    const v = dotted[key];
+    if (v === undefined || v === "") continue;
+    const g = groupOf(key);
+    const raw = groups.get(g) ?? { group: g, profile: profileOf(basename(file)), file, weight };
+    (raw as Record<string, unknown>)[field] = v;
+    groups.set(g, raw);
+  }
+  return [...groups.values()];
+}
+
+function basename(p: string): string {
+  const i = Math.max(p.lastIndexOf("/"), p.lastIndexOf("\\"));
+  return i >= 0 ? p.slice(i + 1) : p;
+}

package/src/core/scan/walker.ts

@@ -0,0 +1,47 @@
+// scan/walker.ts —— 目录漫步：默认忽略常见无关目录；resolveRoot 越界拒绝（Spec §8.5 红线）
+import { readdirSync, statSync, type Dirent } from "node:fs";
+import { isAbsolute, join, relative, resolve } from "node:path";
+
+const IGNORED_DIRS = new Set([
+  "node_modules", ".git", "target", "dist", "venv", "logs", "docs",
+  "build", "coverage", "__pycache__", ".idea", ".vscode",
+]);
+
+const MAX_FILES = 2000;
+const MAX_DEPTH = 8;
+
+/**
+ * 把输入路径解析为绝对路径；越界（cwd 子树之外、`..` 上跳、cwd 本身除外）
+ * 直接抛 "scan path out of scope"——与「密码不进上下文」同级别的红线。
+ */
+export function resolveRoot(input: string): string {
+  const cwd = process.cwd();
+  const abs = isAbsolute(input) ? resolve(input) : resolve(cwd, input);
+  const rel = relative(cwd, abs);
+  if (rel.startsWith("..") || isAbsolute(rel)) {
+    throw new Error(`scan path out of scope: ${input}`);
+  }
+  return abs;
+}
+
+/** 递归收集文本候选文件（忽略无关目录与隐藏目录，上限 MAX_FILES）。 */
+export function walk(root: string): string[] {
+  const out: string[] = [];
+  const visit = (dir: string, depth: number): void => {
+    if (depth > MAX_DEPTH || out.length >= MAX_FILES) return;
+    let entries: Dirent[];
+    try { entries = readdirSync(dir, { withFileTypes: true }); } catch { return; }
+    for (const e of entries) {
+      if (IGNORED_DIRS.has(e.name)) continue;
+      if (e.isDirectory() && e.name.startsWith(".")) continue;
+      const p = join(dir, e.name);
+      if (e.isDirectory()) visit(p, depth + 1);
+      else if (e.isFile()) {
+        try { if (statSync(p).isFile()) out.push(p); } catch { /* ignore */ }
+        if (out.length >= MAX_FILES) return;
+      }
+    }
+  };
+  visit(root, 0);
+  return out;
+}

package/src/core/sql-text.ts

@@ -0,0 +1,72 @@
+// src/core/sql-text.ts
+export function stripComments(sql: string): string {
+  let out = "";
+  let i = 0;
+  let q: string | null = null;
+  while (i < sql.length) {
+    const ch = sql[i];
+    if (q) {
+      out += ch;
+      if (ch === q) {
+        if (sql[i + 1] === q) { out += sql[i + 1]; i += 2; continue; } // '' 转义
+        q = null;
+      }
+      i++;
+      continue;
+    }
+    if (ch === "'" || ch === '"' || ch === "`") { q = ch; out += ch; i++; continue; }
+    if (ch === "-" && sql[i + 1] === "-") { while (i < sql.length && sql[i] !== "\n") i++; continue; }
+    if (ch === "/" && sql[i + 1] === "*") {
+      i += 2;
+      while (i < sql.length && !(sql[i] === "*" && sql[i + 1] === "/")) i++;
+      i += 2;
+      continue;
+    }
+    out += ch; i++;
+  }
+  return out;
+}
+
+export function splitStatements(sql: string): string[] {
+  // 注：先 stripComments 再切分——注释文本不进入语句，注释内分号自然消失；
+  // 这里只处理字符串字面量（含 '' 双写转义）内的分号
+  const clean = stripComments(sql);
+  const stmts: string[] = [];
+  let cur = "";
+  let q: string | null = null;
+  for (let i = 0; i < clean.length; i++) {
+    const ch = clean[i];
+    const nx = clean[i + 1];
+    if (q) {
+      cur += ch;
+      if (ch === q) {
+        if (nx === q) { cur += nx; i++; continue; }
+        q = null;
+      }
+      continue;
+    }
+    if (ch === "'" || ch === '"' || ch === "`") { q = ch; cur += ch; continue; }
+    if (ch === ";") {
+      if (cur.trim()) stmts.push(cur.trim());
+      cur = "";
+      continue;
+    }
+    cur += ch;
+  }
+  if (cur.trim()) stmts.push(cur.trim());
+  return stmts;
+}
+
+const WRITE_RE = /^\s*(INSERT|UPDATE|DELETE|DROP|CREATE|ALTER|TRUNCATE|REPLACE|LOAD|MERGE|EXEC|EXECUTE|CALL|MSCK|CACHE|REFRESH)\b/i;
+const WITH_DML_RE = /^\s*WITH\b[\s\S]*\b(INSERT|UPDATE|DELETE|MERGE)\b/i;
+const FOR_UPDATE_RE = /\bFOR\s+UPDATE\b/i;
+const DROP_TABLE_RE = /^\s*DROP\s+(TABLE|DATABASE)\b/i;
+
+export function isWriteStatement(stmt: string): boolean {
+  const clean = stripComments(stmt);
+  return WRITE_RE.test(clean) || WITH_DML_RE.test(clean) || FOR_UPDATE_RE.test(clean);
+}
+
+export function isDropStatement(stmt: string): boolean {
+  return DROP_TABLE_RE.test(stripComments(stmt));
+}

package/src/core/types.ts

@@ -0,0 +1,88 @@
+// core/types.ts
+export type DbTypeId = "postgresql" | "mysql" | "oracle" | "dm"
+  | "redis" | "elasticsearch" | "hive" | "spark";
+export type DbFamily = "relational" | "kv" | "search" | "bigdata";
+
+export interface ConnConfig {
+  id: string; name: string; type: DbTypeId;
+  description?: string; host?: string; port?: number;
+  username?: string; password?: string;
+  /** 家族语义：关系型=库名 / DM=schema / Redis=不用（用 dbIndex）/
+      ES=默认 index / Hive-Spark=database 名 */
+  database?: string;
+  dbIndex?: number;                 // Redis 库号
+  apiKey?: string;                  // ES 预留（二期）
+  options?: Record<string, string>; // Hive/Spark 会话变量等（旧文件 extraParams 读取时并入此字段）
+  isDefault?: boolean;              // 默认连接标记（§11.1，G 阶段接线）
+  createdAt: string;
+}
+
+export interface ParsedTarget {
+  host: string; port: number;
+  username?: string; password?: string;
+  database?: string; dbIndex?: number; ssl?: boolean;
+}
+
+export interface DbConnection { type: DbTypeId; client: unknown; close(): Promise<void>; }
+export interface ExecOpts { readonly: boolean; maxRows: number; timeoutSec: number; }
+
+// scan/candidates.ts
+export type CandidateStatus = "ready" | "incomplete" | "encrypted" | "exists";
+export interface Candidate {
+  status: CandidateStatus;
+  dialectId: DbTypeId;              // 由方言 fingerprints 匹配得出
+  partial: Partial<ConnConfig>;     // 已抽到的字段
+  missing: string[];                // 待补字段名（incomplete 时）
+  source: { file: string; profile?: string; confidence: number };
+}
+
+// ── 以下从 src/db.ts 原样搬入（字段不变） ──────────
+
+export interface TableInfo {
+  schema: string;
+  name: string;
+  type: string;
+  description: string;
+}
+
+export interface ColumnInfo {
+  name: string;
+  type: string;
+  nullable: boolean;
+  default: string | null;
+  primaryKey: boolean;
+  comment: string;
+}
+
+export interface QueryResult {
+  success: boolean;
+  columns?: string[];
+  rows?: unknown[][];
+  rowCount?: number;
+  duration?: string;
+  error?: string;
+  /** 结果是否因达到 maxRows 而截断（Task 8 加法字段，Task 11 导出显示可复用） */
+  truncated?: boolean;
+}
+
+export interface ListTablesResult {
+  success: boolean;
+  tables?: TableInfo[];
+  count?: number;
+  error?: string;
+}
+
+export interface DescribeTableResult {
+  success: boolean;
+  columns?: ColumnInfo[];
+  count?: number;
+  error?: string;
+}
+
+export interface TestConnectionResult {
+  success: boolean;
+  version?: string;
+  latency?: string;
+  error?: string;
+  warning?: string;   // ES 未知大版本等非致命版本警告（Task 7，Spec §12）
+}

package/src/core/whitelist.ts

@@ -0,0 +1,4 @@
+// core/whitelist.ts —— 命令白名单匹配工具（KV/搜索共用）
+export function matchCommand(cmd: string, list: string[]): boolean {
+  return list.includes(cmd.trim().toUpperCase());
+}