@zonuexe/techbook-mcp 0.2.3 → 0.2.4

package/src/adapters/http/mock-client.ts

@@ -1,77 +0,0 @@
-import type { HttpClient, RequestOptions, HttpResponse } from "../../ports/http.js";
-
-export interface MockResponseData {
-  status: number;
-  body: string;
-  headers?: Record<string, string>;
-}
-
-class MockHttpResponse implements HttpResponse {
-  constructor(
-    private readonly data: MockResponseData,
-    private readonly requestUrl: string,
-  ) {}
-
-  get status(): number { return this.data.status; }
-  get url(): string { return this.requestUrl; }
-  async text(): Promise<string> { return this.data.body; }
-  header(name: string): string | null {
-    return this.data.headers?.[name.toLowerCase()] ?? null;
-  }
-}
-
-export class MockHttpClient implements HttpClient {
-  private readonly handlers = new Map<string, MockResponseData>();
-  private readonly postHandlers = new Map<string, MockResponseData>();
-  private readonly _calls: string[] = [];
-
-  /** GET: URL の前方一致でレスポンスを登録する */
-  addResponse(urlPrefix: string, data: MockResponseData): this {
-    this.handlers.set(urlPrefix, data);
-    return this;
-  }
-
-  /** POST: URL の前方一致でレスポンスを登録する */
-  addPostResponse(urlPrefix: string, data: MockResponseData): this {
-    this.postHandlers.set(urlPrefix, data);
-    return this;
-  }
-
-  get calls(): readonly string[] {
-    return this._calls;
-  }
-
-  async get(url: string, _options?: RequestOptions): Promise<HttpResponse> {
-    this._calls.push(url);
-
-    // 完全一致を優先
-    if (this.handlers.has(url)) {
-      return new MockHttpResponse(this.handlers.get(url)!, url);
-    }
-
-    // 前方一致
-    for (const [prefix, data] of this.handlers) {
-      if (url.startsWith(prefix)) {
-        return new MockHttpResponse(data, url);
-      }
-    }
-
-    throw new Error(`MockHttpClient: no handler for GET: ${url}`);
-  }
-
-  async post(url: string, _body: string, _options?: RequestOptions): Promise<HttpResponse> {
-    this._calls.push(url);
-
-    if (this.postHandlers.has(url)) {
-      return new MockHttpResponse(this.postHandlers.get(url)!, url);
-    }
-
-    for (const [prefix, data] of this.postHandlers) {
-      if (url.startsWith(prefix)) {
-        return new MockHttpResponse(data, url);
-      }
-    }
-
-    throw new Error(`MockHttpClient: no handler for POST: ${url}`);
-  }
-}

package/src/adapters/openbd.ts

@@ -1,142 +0,0 @@
-import type { PublisherDeps } from "../domain/publisher.js";
-import type { BookRecord } from "../domain/book.js";
-import { fetchText } from "./publishers/base.js";
-
-const OPENBD_API_URL = "https://api.openbd.jp/v1/get";
-
-// --- 型定義 ---
-
-interface OpenBDSummary {
-  isbn: string;
-  title: string;
-  publisher: string;
-  pubdate: string;   // "YYYYMMDD"
-  cover: string;     // "https://cover.openbd.jp/{isbn}.jpg"
-  author: string;
-}
-
-interface OpenBDTextContent {
-  TextType: string;  // "02": 短い説明, "03": 説明文, "04": 目次
-  ContentAudience: string;
-  Text: string;
-}
-
-interface OpenBDPrice {
-  PriceType: string;    // "03": 税込定価
-  PriceAmount: string;
-  CurrencyCode: string;
-}
-
-interface OpenBDHanmoto {
-  isbn: string;
-  storelink?: string;
-  [key: string]: unknown;
-}
-
-export interface OpenBDEntry {
-  summary: OpenBDSummary;
-  hanmoto?: OpenBDHanmoto;
-  onix: {
-    CollateralDetail?: {
-      TextContent?: OpenBDTextContent[];
-    };
-    ProductSupply?: {
-      SupplyDetail?: {
-        Price?: OpenBDPrice[];
-      };
-    };
-  };
-}
-
-// --- ユーティリティ ---
-
-function parsePubDate(pubdate: string): string | undefined {
-  if (!pubdate || pubdate.length < 8) return undefined;
-  return `${pubdate.slice(0, 4)}-${pubdate.slice(4, 6)}-${pubdate.slice(6, 8)}`;
-}
-
-function findTextByType(entry: OpenBDEntry, ...types: string[]): string | undefined {
-  const texts = entry.onix.CollateralDetail?.TextContent;
-  if (!texts) return undefined;
-  for (const type of types) {
-    const found = texts.find(t => t.TextType === type);
-    if (found?.Text) return found.Text;
-  }
-  return undefined;
-}
-
-function getTaxIncludedPrice(entry: OpenBDEntry): number | undefined {
-  const prices = entry.onix.ProductSupply?.SupplyDetail?.Price;
-  if (!prices) return undefined;
-  // PriceType "03" = 税込定価
-  const price = prices.find(p => p.PriceType === "03");
-  if (!price) return undefined;
-  const amount = parseInt(price.PriceAmount, 10);
-  return isNaN(amount) ? undefined : amount;
-}
-
-// --- 公開API ---
-
-/**
- * openBD API から複数ISBNの書誌情報を一括取得する。
- * @returns ISBNをキーとするMapを返す。該当なし・取得失敗のISBNは含まれない。
- */
-export async function fetchOpenBDBooks(
-  isbns: string[],
-  deps: PublisherDeps,
-): Promise<Map<string, OpenBDEntry>> {
-  if (isbns.length === 0) return new Map();
-
-  const url = `${OPENBD_API_URL}?isbn=${isbns.join(",")}`;
-  const text = await fetchText(url, deps);
-  const data: (OpenBDEntry | null)[] = JSON.parse(text);
-
-  const result = new Map<string, OpenBDEntry>();
-  for (let i = 0; i < isbns.length; i++) {
-    const entry = data[i];
-    if (entry !== null && entry !== undefined) {
-      result.set(isbns[i], entry);
-    }
-  }
-  return result;
-}
-
-/**
- * openBD エントリを BookRecord に変換する。
- * 出版社サイトから取得できない場合のフォールバック用。
- * url には hanmoto.storelink を使用し、なければ openBD API URL を使用する。
- */
-export function openBDEntryToBookRecord(entry: OpenBDEntry): BookRecord {
-  const { summary } = entry;
-  const storelink = entry.hanmoto?.storelink;
-
-  const authors = summary.author
-    ? summary.author.split(/[\/／、,，]/).map(a => a.trim()).filter(Boolean)
-    : [];
-
-  return {
-    title: summary.title,
-    authors,
-    publisher: summary.publisher,
-    isbn: summary.isbn,
-    publishedAt: parsePubDate(summary.pubdate),
-    url: storelink ?? `https://api.openbd.jp/v1/get?isbn=${summary.isbn}`,
-    price: getTaxIncludedPrice(entry),
-    coverImageUrl: summary.cover || undefined,
-    description: findTextByType(entry, "03", "02"),
-  };
-}
-
-/**
- * openBD の書誌情報で BookRecord の欠損フィールドを補完する。
- * 既存のフィールドは上書きしない。
- */
-export function enrichWithOpenBD(book: BookRecord, entry: OpenBDEntry): BookRecord {
-  return {
-    ...book,
-    publishedAt: book.publishedAt ?? parsePubDate(entry.summary.pubdate),
-    price: book.price ?? getTaxIncludedPrice(entry),
-    coverImageUrl: book.coverImageUrl ?? (entry.summary.cover || undefined),
-    description: book.description ?? findTextByType(entry, "03", "02"),
-  };
-}

package/src/adapters/publishers/base.ts

@@ -1,279 +0,0 @@
-import iconv from "iconv-lite";
-import type { PublisherDeps } from "../../domain/publisher.js";
-import type { EbookStore, DrmType } from "../../domain/book.js";
-import type { HtmlDocument } from "../../ports/html-parser.js";
-
-const DEFAULT_HEADERS = {
-  "User-Agent": "techbook-mcp/0.1.0 (+https://github.com/zonuexe/techbook-mcp; bibliographic search bot)",
-  "Accept": "text/html,application/xhtml+xml,application/json",
-  "Accept-Language": "ja,en;q=0.9",
-};
-
-export const CACHE_TTL_SECONDS = 3600; // 1時間
-export const ROBOTS_CACHE_TTL_SECONDS = 6 * 3600; // 6時間
-
-// --- robots.txt チェック ---
-
-/** robots.txt の1ルール */
-interface RobotsRule {
-  type: "allow" | "disallow";
-  path: string;
-}
-
-/** robots.txt のユーザーエージェントセクション */
-interface RobotsSection {
-  agents: string[];
-  rules: RobotsRule[];
-}
-
-/** robots.txt をパースしてセクション一覧を返す */
-function parseRobotsTxt(content: string): RobotsSection[] {
-  const sections: RobotsSection[] = [];
-  let current: RobotsSection | null = null;
-  let inAgentBlock = true;
-
-  for (const rawLine of content.split(/\r?\n/)) {
-    const trimmedRaw = rawLine.trim();
-    // 空行（コメント行ではない）のみセクションをリセット
-    if (!trimmedRaw || trimmedRaw.startsWith("#")) {
-      if (!trimmedRaw) {
-        current = null;
-        inAgentBlock = true;
-      }
-      continue;
-    }
-
-    const line = trimmedRaw.split("#")[0].trim();
-    if (!line) continue;
-
-    const colonIdx = line.indexOf(":");
-    if (colonIdx === -1) continue;
-
-    const key = line.slice(0, colonIdx).trim().toLowerCase();
-    const value = line.slice(colonIdx + 1).trim();
-
-    if (key === "user-agent") {
-      if (inAgentBlock && current !== null) {
-        // 同じセクションに複数のUser-agent行
-        current.agents.push(value.toLowerCase());
-      } else {
-        // 新しいセクション開始
-        current = { agents: [value.toLowerCase()], rules: [] };
-        sections.push(current);
-        inAgentBlock = true;
-      }
-    } else if (current !== null && (key === "allow" || key === "disallow")) {
-      inAgentBlock = false;
-      current.rules.push({ type: key, path: value });
-    }
-  }
-
-  return sections;
-}
-
-/** 指定ユーザーエージェントに適用されるルールを返す（固有エージェント優先、なければ * にフォールバック） */
-function getRulesForAgent(sections: RobotsSection[], agentToken: string): RobotsRule[] {
-  const lower = agentToken.toLowerCase();
-
-  for (const section of sections) {
-    if (section.agents.includes(lower)) return section.rules;
-  }
-  for (const section of sections) {
-    if (section.agents.includes("*")) return section.rules;
-  }
-  return [];
-}
-
-/** パスがルール一覧で許可されているか判定する（最長プレフィックス一致） */
-function isPathAllowed(path: string, rules: RobotsRule[]): boolean {
-  let bestMatch = { length: -1, allowed: true };
-
-  for (const rule of rules) {
-    if (!rule.path) continue; // 空の Disallow は「全許可」を意味するが不一致として扱う
-
-    if (path.startsWith(rule.path) && rule.path.length > bestMatch.length) {
-      bestMatch = { length: rule.path.length, allowed: rule.type === "allow" };
-    }
-  }
-
-  return bestMatch.allowed;
-}
-
-/**
- * 指定URLのオリジンの robots.txt を取得してアクセス可否を返す。
- * 取得結果は6時間キャッシュする。エラー時はアクセスを許可する（fail-open）。
- */
-export async function checkRobotsTxt(url: string, deps: PublisherDeps): Promise<boolean> {
-  const parsed = new URL(url);
-  const origin = `${parsed.protocol}//${parsed.host}`;
-  const cacheKey = `robots:${origin}`;
-
-  let content: string;
-  const cached = await deps.cache.get(cacheKey);
-
-  if (cached !== null) {
-    content = cached;
-  } else {
-    try {
-      const response = await deps.http.get(`${origin}/robots.txt`, { headers: DEFAULT_HEADERS });
-      content = response.status === 200 ? await response.text() : "";
-    } catch {
-      // robots.txt 取得失敗時はアクセスを許可する
-      content = "";
-    }
-    await deps.cache.set(cacheKey, content, ROBOTS_CACHE_TTL_SECONDS);
-  }
-
-  if (!content) return true;
-
-  const sections = parseRobotsTxt(content);
-  const rules = getRulesForAgent(sections, "techbook-mcp");
-  return isPathAllowed(parsed.pathname + parsed.search, rules);
-}
-
-export async function fetchText(
-  url: string,
-  deps: PublisherDeps,
-  extraHeaders?: Record<string, string>,
-): Promise<string> {
-  const cached = await deps.cache.get(url);
-  if (cached !== null) return cached;
-
-  const headers = extraHeaders
-    ? { ...DEFAULT_HEADERS, ...extraHeaders }
-    : DEFAULT_HEADERS;
-
-  const response = await deps.http.get(url, { headers });
-  if (response.status !== 200) {
-    throw new Error(`HTTP ${response.status}: ${url}`);
-  }
-
-  const text = await response.text();
-  await deps.cache.set(url, text, CACHE_TTL_SECONDS);
-  return text;
-}
-
-/** HTMLタグを除去する（gihyo APIのauthorフィールドのruby markup除去に使用） */
-export function stripHtmlTags(html: string): string {
-  return html.replace(/<[^>]+>/g, "");
-}
-
-/**
- * キーワードを EUC-JP でパーセントエンコードする。
- * born-digital・rutles など EUC-JP エンコードのみ受け付けるサイト向け。
- */
-export function encodeEucJp(text: string): string {
-  const bytes = iconv.encode(text, "euc-jp");
-  return Array.from(bytes)
-    .map(b => "%" + b.toString(16).toUpperCase().padStart(2, "0"))
-    .join("");
-}
-
-/**
- * "2026年3月25日" → "2026-03-25"
- * 1桁の月・日も対応する。
- */
-export function parseJapaneseDateToISO(text: string): string | undefined {
-  const m = text.match(/(\d{4})年(\d{1,2})月(\d{1,2})日/);
-  if (!m) return undefined;
-  return `${m[1]}-${m[2].padStart(2, "0")}-${m[3].padStart(2, "0")}`;
-}
-
-/**
- * 著者名末尾の役割語（著・訳・編・監修・監訳など）を除去して名前だけを返す。
- * 例: "Dan Vanderkam　著" → "Dan Vanderkam"
- */
-export function stripAuthorRole(name: string): string {
-  return name.replace(/[\u3000\s]*(著|訳|編|監修|監訳|著訳|著・訳|他)[\u3000\s]*$/, "").trim();
-}
-
-/** "¥3,960" や "3,300円（税込）" などから整数値を取り出す */
-export function parseJapanesePrice(text: string): number | undefined {
-  const match = text.match(/[\d,]+/);
-  if (!match) return undefined;
-  return parseInt(match[0].replace(/,/g, ""), 10);
-}
-
-/** 相対URLを絶対URLに解決する */
-export function resolveUrl(base: string, path: string): string {
-  return new URL(path, base).toString();
-}
-
-/**
- * HTMLテキストから Amazon ASIN を抽出する。
- * amazon.co.jp/dp/{ASIN}, /gp/product/{ASIN}, /o/ASIN/{ASIN} 形式に対応。
- */
-export function extractAsin(html: string): string | undefined {
-  const match = html.match(/amazon\.co\.jp\/(?:dp|gp\/product|o\/ASIN)\/([A-Z0-9]{10})/);
-  return match?.[1];
-}
-
-// --- 電子書籍ストア分類 ---
-
-interface StorePattern {
-  pattern: RegExp;
-  name: string;
-  drm: DrmType;
-}
-
-const EBOOK_STORE_PATTERNS: StorePattern[] = [
-  // DRM-free
-  { pattern: /techbookfest\.org\/product\//, name: "技術書典", drm: "free" },
-  { pattern: /oreilly\.co\.jp\/books\//, name: "オライリー・ジャパン", drm: "free" },
-  { pattern: /shop\.rutles\.net\//, name: "ラトルズ", drm: "free" },
-  { pattern: /peaks\.cc\/books\//, name: "PEAKS", drm: "free" },
-  { pattern: /optronics-ebook\.com\/products\//, name: "オプトロニクス社", drm: "free" },
-  { pattern: /gihyo\.jp\/dp\/ebook\//, name: "Gihyo Digital Publishing", drm: "social" },
-  { pattern: /seshop\.com\/product\//, name: "SEshop", drm: "social" },
-  { pattern: /book-tech\.com\/books\//, name: "BOOK TECH", drm: "social" },
-  { pattern: /wgn-obs\.shop-pro\.jp\/\?pid=/, name: "ボーンデジタル", drm: "social" },
-  // ソーシャルDRM (購入時生成IDまたは購入者情報を透かし刻印、技術的制限なし)
-  { pattern: /book\.mynavi\.jp\/manatee\//, name: "マナティ", drm: "social" },
-  { pattern: /www\.lambdanote\.com\/products\//, name: "ラムダノート", drm: "social" },
-  { pattern: /tatsu-zine\.com\/books\/(?!pub\/)/, name: "達人出版会", drm: "social" },
-  // ソーシャルDRM (購入者情報透かし入りPDF、技術的制限なし)
-  { pattern: /book\.impress\.co\.jp\/books\//, name: "インプレスブックス", drm: "social" },
-  // DRM-attached
-  { pattern: /saiensu\.co\.jp/, name: "サイエンス社", drm: "password_pdf" },
-  { pattern: /amazon\.co\.jp/, name: "Kindle", drm: "drm" },
-  { pattern: /kinokuniya\.co\.jp\/(?:kinoppystore|f\/dsg-08)/, name: "Kinoppy", drm: "drm" },
-  { pattern: /coop-ebook\.jp\/mem\//, name: "VarsityWave eBooks", drm: "drm" },
-  { pattern: /books\.rakuten\.co\.jp|rakuten\.kobo\.com|kobo\.com/, name: "楽天Kobo", drm: "drm" },
-  { pattern: /booklive\.jp/, name: "BookLive", drm: "drm" },
-  { pattern: /honto\.jp/, name: "honto", drm: "drm" },
-  { pattern: /bookwalker\.jp/, name: "BOOK☆WALKER", drm: "drm" },
-  { pattern: /ebookjapan\.yahoo\.co\.jp/, name: "eBookJapan", drm: "drm" },
-  { pattern: /store\.line\.me/, name: "LINEマンガ", drm: "drm" },
-];
-
-/** URLから電子書籍ストア情報を返す。未知のストアは null。 */
-export function classifyEbookStore(url: string): EbookStore | null {
-  for (const { pattern, name, drm } of EBOOK_STORE_PATTERNS) {
-    if (pattern.test(url)) {
-      return { name, url, drm };
-    }
-  }
-  return null;
-}
-
-/**
- * HTMLドキュメント内の全リンクを走査して電子書籍ストアを抽出する。
- * 同一ストアのURLが複数あれば最初の1件のみ返す。
- */
-export function extractEbookStoresFromDoc(doc: HtmlDocument): EbookStore[] {
-  const stores: EbookStore[] = [];
-  const seenNames = new Set<string>();
-
-  for (const link of doc.select("a[href]")) {
-    const href = link.attr("href");
-    if (!href) continue;
-
-    const store = classifyEbookStore(href);
-    if (store && !seenNames.has(store.name)) {
-      seenNames.add(store.name);
-      stores.push(store);
-    }
-  }
-
-  return stores;
-}

package/src/adapters/publishers/book-tech.ts

@@ -1,117 +0,0 @@
-import type { PublisherAdapter, PublisherDeps } from "../../domain/publisher.js";
-import type { BookRecord, SearchQuery } from "../../domain/book.js";
-import { fetchText, parseJapanesePrice, resolveUrl } from "./base.js";
-
-const BASE_URL = "https://book-tech.com";
-const SEARCH_URL = `${BASE_URL}/books`;
-// クエリパラメータキー（URLエンコード済み）
-const SEARCH_PARAM = "q%5Btitle_or_overview_or_identification_number_1_or_product_code_cont%5D";
-
-/** "2026/2/20" → "2026-02-20" */
-function parseDate(text: string): string | undefined {
-  const m = text.match(/(\d{4})\/(\d{1,2})\/(\d{1,2})/);
-  if (!m) return undefined;
-  return `${m[1]}-${m[2].padStart(2, "0")}-${m[3].padStart(2, "0")}`;
-}
-
-/** "(著)" などの役割語を末尾から除去する */
-function stripRole(name: string): string {
-  return name.replace(/\s*[（(][^)）]*[)）]\s*$/, "").trim();
-}
-
-export const bookTechAdapter: PublisherAdapter = {
-  id: "book-tech",
-  name: "BOOK TECH",
-  baseUrl: BASE_URL,
-
-  async search(query: SearchQuery, deps: PublisherDeps): Promise<BookRecord[]> {
-    const word = [query.title, query.author].filter(Boolean).join(" ");
-    if (!word) return [];
-
-    const url = `${SEARCH_URL}?${SEARCH_PARAM}=${encodeURIComponent(word)}`;
-    const html = await fetchText(url, deps);
-    const doc = deps.parser.parse(html);
-
-    const results: BookRecord[] = [];
-    const limit = query.limit ?? 10;
-
-    for (const item of doc.select("div.contents-index-item")) {
-      const linkEl = item.find("a.book-ribbon-link")[0];
-      const href = linkEl?.attr("href");
-      if (!href) continue;
-      const bookUrl = resolveUrl(BASE_URL, href);
-
-      const title = item.find(".contents-index-item-detail-title")[0]?.text().trim();
-      if (!title) continue;
-
-      const publisherEl = item.find("a[href*='publisher_relations']")[0];
-      const publisher = publisherEl?.text().trim() ?? "";
-
-      const authors = item.find("a[href*='author_relations']")
-        .map(el => stripRole(el.text().trim()))
-        .filter(Boolean);
-
-      const priceText = item.find(".contents-index-item-detail-price_include_tax")[0]?.text();
-      const price = priceText ? parseJapanesePrice(priceText) : undefined;
-
-      const dateText = item.find(".my-1")[0]?.text();
-      const publishedAt = dateText ? parseDate(dateText) : undefined;
-
-      const imgEl = item.find("img.thumb")[0];
-      const coverImageUrl = imgEl?.attr("src") ?? undefined;
-
-      results.push({
-        title,
-        authors,
-        publisher,
-        url: bookUrl,
-        price,
-        publishedAt,
-        coverImageUrl,
-        ebookStores: [{ name: "BOOK TECH", url: bookUrl, drm: "social" }],
-      });
-
-      if (results.length >= limit) break;
-    }
-
-    return results;
-  },
-
-  async getDetail(url: string, deps: PublisherDeps): Promise<BookRecord> {
-    const html = await fetchText(url, deps);
-    const doc = deps.parser.parse(html);
-
-    const title = doc.selectOne(".contents-book-about-title h1")?.text().trim() ?? "";
-
-    const publisherEl = doc.selectOne("a[href*='publisher_relations']");
-    const publisher = publisherEl?.text().trim() ?? "";
-
-    const authors = doc.select("a[href*='author_relations']")
-      .map(el => stripRole(el.text().trim()))
-      .filter(Boolean);
-
-    const priceText = doc.selectOne(".contents-book-item-detail-price_include_tax")?.text();
-    const price = priceText ? parseJapanesePrice(priceText) : undefined;
-
-    const dateText = doc.selectOne(".contents-book-about-publicationdate")?.text();
-    const publishedAt = dateText ? parseDate(dateText) : undefined;
-
-    const isbnText = doc.selectOne(".contents-book-about-id")?.text();
-    const isbn = isbnText?.match(/\d{13}/)?.[0];
-
-    const imgEl = doc.selectOne("img.thumb");
-    const coverImageUrl = imgEl?.attr("src") ?? undefined;
-
-    return {
-      title,
-      authors,
-      publisher,
-      url,
-      isbn,
-      price,
-      publishedAt,
-      coverImageUrl,
-      ebookStores: [{ name: "BOOK TECH", url, drm: "social" }],
-    };
-  },
-};