@zonuexe/techbook-mcp 0.2.2 → 0.2.4

package/src/adapters/publishers/gihyo.ts

@@ -1,120 +0,0 @@
-import type { PublisherAdapter, PublisherDeps } from "../../domain/publisher.js";
-import type { BookRecord, SearchQuery } from "../../domain/book.js";
-import { fetchText, stripHtmlTags, resolveUrl, extractAsin, extractEbookStoresFromDoc } from "./base.js";
-
-const BASE_URL = "https://gihyo.jp";
-
-// --- API レスポンス型 ---
-
-interface GihyoSearchResponse {
-  lname: string;
-  total: number;
-  list: Record<string, GihyoBookEntry>;
-  next: boolean;
-  query: string;
-}
-
-interface GihyoBookEntry {
-  series: string;
-  title: string;
-  subtitle: string;
-  /** キー: 役割 ("著", "監修" など)、値: { 著者名: markup } */
-  author: Record<string, Record<string, string>>;
-  /** [定価, 割引価格] */
-  price: [number, number];
-  stock: number;
-  /** ["YYYY.M.D", ""] */
-  release: [string, string];
-  /** 相対URL: "/book/YYYY/978-4-..." */
-  url: string;
-  upcoming: boolean;
-  support: boolean;
-  /** [thumb_url, width, height, full_url] */
-  cover: [string, number, number, string];
-}
-
-// --- 変換ヘルパー ---
-
-function parseAuthors(authorField: Record<string, Record<string, string>>): string[] {
-  return Object.values(authorField).flatMap(roleEntries =>
-    Object.keys(roleEntries).map(name => stripHtmlTags(name).trim()),
-  );
-}
-
-function parseReleaseDate(release: [string, string]): string | undefined {
-  const raw = release[0];
-  if (!raw) return undefined;
-  // "2025.9.29" → "2025-09-29"
-  const parts = raw.split(".");
-  if (parts.length !== 3) return raw;
-  const [y, m, d] = parts;
-  return `${y}-${m.padStart(2, "0")}-${d.padStart(2, "0")}`;
-}
-
-function entryToBookRecord(isbn: string, entry: GihyoBookEntry): BookRecord {
-  const title = entry.subtitle
-    ? `${entry.title} ${entry.subtitle}`
-    : entry.title;
-
-  return {
-    title,
-    authors: parseAuthors(entry.author),
-    publisher: "技術評論社",
-    publishedAt: parseReleaseDate(entry.release),
-    isbn: isbn.replace(/-/g, ""),
-    url: resolveUrl(BASE_URL, entry.url),
-    price: entry.price[0] > 0 ? entry.price[0] : undefined,
-    coverImageUrl: entry.cover[3] ? resolveUrl(BASE_URL, entry.cover[3]) : undefined,
-  };
-}
-
-// --- アダプター実装 ---
-
-export const gihyoAdapter: PublisherAdapter = {
-  id: "gihyo",
-  name: "技術評論社",
-  baseUrl: BASE_URL,
-
-  async search(query: SearchQuery, deps: PublisherDeps): Promise<BookRecord[]> {
-    const word = [query.title, query.author].filter(Boolean).join(" ");
-    if (!word) return [];
-
-    const limit = query.limit ?? 10;
-    const url = `${BASE_URL}/api_gh/site/search?search=${encodeURIComponent(word)}&limit=${limit}`;
-    const text = await fetchText(url, deps);
-    const data: GihyoSearchResponse = JSON.parse(text);
-
-    return Object.entries(data.list).map(([isbn, entry]) => entryToBookRecord(isbn, entry));
-  },
-
-  async getDetail(url: string, deps: PublisherDeps): Promise<BookRecord> {
-    // URL例: https://gihyo.jp/book/2022/978-4-297-12815-2
-    const isbnMatch = url.match(/\/(978-[\d-]+)\s*$/);
-    if (!isbnMatch) throw new Error(`URLからISBNを取得できません: ${url}`);
-
-    const isbn = isbnMatch[1];
-    const apiUrl = `${BASE_URL}/api_gh/site/search?search=${encodeURIComponent(isbn)}&limit=1`;
-
-    // JSONメタデータとebook store情報(HTML)を並列取得
-    const [apiText, htmlText] = await Promise.all([
-      fetchText(apiUrl, deps),
-      fetchText(url, deps),
-    ]);
-
-    const data: GihyoSearchResponse = JSON.parse(apiText);
-    const entry = data.list[isbn];
-    if (!entry) throw new Error(`書籍が見つかりません: ${isbn}`);
-
-    const base = entryToBookRecord(isbn, entry);
-
-    const doc = deps.parser.parse(htmlText);
-    const ebookStores = extractEbookStoresFromDoc(doc);
-    const asin = extractAsin(htmlText);
-
-    return {
-      ...base,
-      asin,
-      ebookStores: ebookStores.length > 0 ? ebookStores : undefined,
-    };
-  },
-};

package/src/adapters/publishers/impress.ts

@@ -1,103 +0,0 @@
-import type { PublisherAdapter, PublisherDeps } from "../../domain/publisher.js";
-import type { BookRecord, SearchQuery, DrmType } from "../../domain/book.js";
-import { fetchText, parseJapanesePrice, stripAuthorRole } from "./base.js";
-
-const BASE_URL = "https://book.impress.co.jp";
-
-/** "2026/1/22" → "2026-01-22" */
-function parseImpressDate(text: string): string | undefined {
-  const m = text.trim().match(/(\d{4})\/(\d{1,2})\/(\d{1,2})/);
-  if (!m) return undefined;
-  return `${m[1]}-${m[2].padStart(2, "0")}-${m[3].padStart(2, "0")}`;
-}
-
-/**
- * 著者文字列から著者名配列を返す。
- * 例: "山本康太　著" → ["山本康太"]
- * 複数著者は「、」または改行で区切られる。
- */
-function parseAuthors(text: string): string[] {
-  return text
-    .split(/[、,\n]/)
-    .map(s => stripAuthorRole(s.trim()))
-    .filter(Boolean);
-}
-
-/**
- * 電子書籍ガイドのテキストから DRM 種別を判定する。
- * 明示されていない場合はインプレスの公式方針に基づき social を返す。
- */
-function parseDrmType(text: string): DrmType {
-  if (/ソーシャルDRM/i.test(text)) return "social";
-  if (/DRM-?free|DRMフリー/i.test(text)) return "free";
-  if (/パスワード/i.test(text)) return "password_pdf";
-  return "social";
-}
-
-export const impressBooksAdapter: PublisherAdapter = {
-  id: "impress-books",
-  name: "インプレスブックス",
-  baseUrl: BASE_URL,
-
-  async search(_query: SearchQuery, _deps: PublisherDeps): Promise<BookRecord[]> {
-    // 検索ページは Google Custom Search Engine による JavaScript レンダリングのためスクレイピング不可
-    return [];
-  },
-
-  async getDetail(url: string, deps: PublisherDeps): Promise<BookRecord> {
-    const html = await fetchText(url, deps);
-    const doc = deps.parser.parse(html);
-
-    // タイトル（ページ内最初の h2）
-    const title = doc.selectOne("h2")?.text().trim() ?? "";
-
-    // dl.module-book-data の dt/dd を順序でペアリング
-    const dts = doc.select("dl.module-book-data dt");
-    const dds = doc.select("dl.module-book-data dd");
-    const bookDataMap = new Map<string, string>();
-    for (let i = 0; i < dts.length; i++) {
-      const key = dts[i].text().trim();
-      const val = dds[i]?.text().trim() ?? "";
-      if (key) bookDataMap.set(key, val);
-    }
-
-    const authors = parseAuthors(bookDataMap.get("著者") ?? "");
-    const isbn = bookDataMap.get("ISBN")?.replace(/\s/g, "") || undefined;
-    const publishedAt = parseImpressDate(bookDataMap.get("発売日") ?? "");
-
-    // カバー画像（img.ips.co.jp のプロトコル相対URLに https: を補完）
-    const coverSrc = doc.selectOne(".block-book-detail-img img")?.attr("src");
-    const coverImageUrl = coverSrc
-      ? (coverSrc.startsWith("//") ? `https:${coverSrc}` : coverSrc)
-      : undefined;
-
-    // 電子版価格・DRM
-    const ebookGuide = doc.selectOne(".module-e-book-buy-guide-txt");
-    const ebookBuyBtn = doc.selectOne(".module-e-book-buy-guide-btn a");
-
-    let price: number | undefined;
-    let drm: DrmType = "social";
-
-    if (ebookGuide) {
-      const priceText = ebookGuide.find(".module-e-book-price")[0]?.text();
-      if (priceText) price = parseJapanesePrice(priceText);
-      drm = parseDrmType(ebookGuide.text());
-    }
-
-    const ebookStores = ebookBuyBtn
-      ? [{ name: "インプレスブックス", url, drm }]
-      : [];
-
-    return {
-      title,
-      authors,
-      publisher: "インプレスブックス",
-      url,
-      isbn,
-      price,
-      publishedAt,
-      coverImageUrl,
-      ebookStores,
-    };
-  },
-};

package/src/adapters/publishers/lambdanote.ts

@@ -1,146 +0,0 @@
-import type { PublisherAdapter, PublisherDeps } from "../../domain/publisher.js";
-import type { BookRecord, SearchQuery } from "../../domain/book.js";
-import { fetchText, parseJapanesePrice, resolveUrl, extractAsin, extractEbookStoresFromDoc } from "./base.js";
-import type { EbookStore } from "../../domain/book.js";
-
-const BASE_URL = "https://www.lambdanote.com";
-
-function extractIsbn(text: string): string | undefined {
-  const match = text.match(/97[89]-[\d-]{10,}/);
-  return match ? match[0].replace(/-/g, "") : undefined;
-}
-
-/**
- * Shopify ページに埋め込まれた product JSON を取得する。
- * <script type="application/json" id="ProductJson-..."> または
- * data-product-json 属性を探す。
- */
-function parseShopifyProductJson(html: string): Record<string, unknown> | null {
-  const match = html.match(
-    /<script[^>]+type=["']application\/json["'][^>]*id=["']ProductJson[^"']*["'][^>]*>([\s\S]*?)<\/script>/i,
-  );
-  if (match?.[1]) {
-    try {
-      return JSON.parse(match[1]) as Record<string, unknown>;
-    } catch {
-      // fall through
-    }
-  }
-  return null;
-}
-
-export const lambdanoteAdapter: PublisherAdapter = {
-  id: "lambdanote",
-  name: "ラムダノート",
-  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 = `${BASE_URL}/search?q=${encodeURIComponent(word)}&type=product`;
-    const html = await fetchText(url, deps);
-    const doc = deps.parser.parse(html);
-
-    const results: BookRecord[] = [];
-
-    // Shopify Dawn テーマ系の検索結果セレクター
-    const items = doc.select(
-      ".search-results__list-item, li.search-result-product",
-    );
-
-    for (const item of items) {
-      const titleEl =
-        item.find(".card__heading a")[0] ??
-        item.find("h2 a")[0] ??
-        item.find("h3 a")[0] ??
-        item.find("a.full-unstyled-link")[0];
-
-      if (!titleEl) continue;
-
-      const title = titleEl.text();
-      const href = titleEl.attr("href");
-      if (!title || !href) continue;
-
-      const bookUrl = resolveUrl(BASE_URL, href);
-
-      const priceEl =
-        item.find(".price__regular .price-item")[0] ??
-        item.find(".price-item")[0] ??
-        item.find(".price")[0];
-      const price = priceEl ? parseJapanesePrice(priceEl.text()) : undefined;
-
-      const imgEl = item.find("img")[0];
-      const imgSrc = imgEl?.attr("src") ?? imgEl?.attr("data-src");
-      const coverImageUrl = imgSrc ? resolveUrl(BASE_URL, imgSrc) : undefined;
-
-      results.push({
-        title,
-        authors: [],
-        publisher: "ラムダノート",
-        url: bookUrl,
-        price,
-        coverImageUrl,
-      });
-    }
-
-    return results.slice(0, query.limit ?? 10);
-  },
-
-  async getDetail(url: string, deps: PublisherDeps): Promise<BookRecord> {
-    const html = await fetchText(url, deps);
-    const doc = deps.parser.parse(html);
-
-    const title =
-      doc.selectOne("h1.product__title")?.text() ??
-      doc.selectOne("h1")?.text() ??
-      "";
-
-    const priceEl =
-      doc.selectOne(".price__regular .price-item") ??
-      doc.selectOne(".price-item--regular") ??
-      doc.selectOne("[class*='price']");
-    const price = priceEl ? parseJapanesePrice(priceEl.text()) : undefined;
-
-    const descEl =
-      doc.selectOne(".product__description") ??
-      doc.selectOne("[class*='description']");
-    const description = descEl?.text() || undefined;
-
-    const isbn = extractIsbn(html);
-    const asin = extractAsin(html);
-
-    // vendor フィールドに著者が入っている場合がある
-    const productJson = parseShopifyProductJson(html);
-    const vendor = typeof productJson?.["vendor"] === "string" ? productJson["vendor"] : null;
-    const authors: string[] = vendor
-      ? vendor.split(/[,、/／]/).map((a: string) => a.trim()).filter(Boolean)
-      : [];
-
-    const imgEl =
-      doc.selectOne(".product__media img") ??
-      doc.selectOne(".product-featured-image") ??
-      doc.selectOne("[class*='product'] img");
-    const imgSrc = imgEl?.attr("src") ?? imgEl?.attr("data-src");
-    const coverImageUrl = imgSrc ? resolveUrl(BASE_URL, imgSrc) : undefined;
-
-    // ラムダノートは購入時生成の一意IDをPDF欄外に印字 (ソーシャルDRM)
-    const ebookStores: EbookStore[] = [
-      { name: "ラムダノート", url, drm: "social" },
-      ...extractEbookStoresFromDoc(doc).filter(s => s.name !== "ラムダノート"),
-    ];
-
-    return {
-      title,
-      authors,
-      publisher: "ラムダノート",
-      url,
-      price,
-      description,
-      isbn,
-      asin,
-      coverImageUrl,
-      ebookStores,
-    };
-  },
-};

package/src/adapters/publishers/manatee.ts

@@ -1,113 +0,0 @@
-import type { PublisherAdapter, PublisherDeps } from "../../domain/publisher.js";
-import type { BookRecord, SearchQuery } from "../../domain/book.js";
-import type { HtmlDocument } from "../../ports/html-parser.js";
-import { fetchText, parseJapanesePrice, resolveUrl } from "./base.js";
-
-const BASE_URL = "https://book.mynavi.jp/manatee";
-const BOOKS_URL = `${BASE_URL}/books/`;
-
-/**
- * `.attribute li` 内の著者リンクを配列にする。
- * 各 <a> のテキストが著者名（役割は括弧内テキストで付記されているが名前は <a> 内）。
- */
-function parseAuthors(doc: HtmlDocument): string[] {
-  return doc
-    .select(".attribute li a")
-    .map(el => el.text().trim())
-    .filter(Boolean);
-}
-
-export const manateeAdapter: PublisherAdapter = {
-  id: "manatee",
-  name: "マナティ",
-  baseUrl: BASE_URL,
-
-  async search(query: SearchQuery, deps: PublisherDeps): Promise<BookRecord[]> {
-    const word = [query.title, query.author].filter(Boolean).join(" ");
-    if (!word) return [];
-
-    const url = `${BOOKS_URL}?topics_keyword=${encodeURIComponent(word)}`;
-    const html = await fetchText(url, deps);
-    const doc = deps.parser.parse(html);
-
-    const results: BookRecord[] = [];
-    const limit = query.limit ?? 10;
-
-    // 書籍アイテムは <!-- item --> コメントで区切られた <li> 内
-    for (const li of doc.select("ul.category_list li")) {
-      const titleEl = li.find("dl.detail dt a")[0];
-      if (!titleEl) continue;
-
-      const title = titleEl.text().trim();
-      if (!title) continue;
-
-      const href = titleEl.attr("href");
-      if (!href) continue;
-      const bookUrl = href.startsWith("http") ? href : resolveUrl(BASE_URL, href);
-
-      const priceText = li.find(".detail__price")[0]?.text().trim();
-      const price = priceText ? parseJapanesePrice(priceText) : undefined;
-
-      const imgEl = li.find(".image img")[0];
-      const imgSrc = imgEl?.attr("src");
-      const coverImageUrl = imgSrc ? resolveUrl(BASE_URL, imgSrc) : undefined;
-
-      results.push({
-        title,
-        authors: [],
-        publisher: "マナティ",
-        url: bookUrl,
-        price,
-        coverImageUrl,
-        ebookStores: [{ name: "マナティ", 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("h1.title")?.text().trim() ?? "";
-
-    // `.intro` の最初の <p> が出版社名
-    const publisher = doc.selectOne(".intro p")?.text().trim() ?? "マナティ";
-
-    const authors = parseAuthors(doc);
-
-    // 購入形態テーブルの最初の行から価格取得
-    const priceText = doc.selectOne("#item_selectlist table .price")?.text().trim();
-    const price = priceText ? parseJapanesePrice(priceText) : undefined;
-
-    // 表紙画像の alt 属性が "ISBN.jpg" 形式（例: "9784839984274.jpg"）
-    const imgEl = doc.selectOne("div.item_meta_image img");
-    const imgSrc = imgEl?.attr("src");
-    const coverImageUrl = imgSrc ? resolveUrl(BASE_URL, imgSrc) : undefined;
-    const imgAlt = imgEl?.attr("alt") ?? "";
-    const isbn = imgAlt.replace(/\.jpg$/i, "").replace(/-/g, "") || undefined;
-
-    // 発売日: "発売日：2024-03-22"
-    const dateText = doc.selectOne("p.date")?.text().trim();
-    const publishedAt = dateText?.match(/(\d{4}-\d{2}-\d{2})/)?.[1] ?? undefined;
-
-    const descEl = doc.selectOne(".item_desc p:not(.date):not(.pages)");
-    const description = descEl?.text().trim() || undefined;
-
-    return {
-      title,
-      authors,
-      publisher,
-      url,
-      isbn,
-      price,
-      publishedAt,
-      description,
-      coverImageUrl,
-      ebookStores: [{ name: "マナティ", url, drm: "social" }],
-    };
-  },
-};

package/src/adapters/publishers/maruzen-publishing.ts

@@ -1,129 +0,0 @@
-import type { PublisherAdapter, PublisherDeps } from "../../domain/publisher.js";
-import type { BookRecord, SearchQuery } from "../../domain/book.js";
-import { fetchText, resolveUrl, extractEbookStoresFromDoc, parseJapaneseDateToISO, stripAuthorRole } from "./base.js";
-
-const BASE_URL = "https://www.maruzen-publishing.co.jp";
-const SEARCH_URL = `${BASE_URL}/search/`;
-
-/** 403回避のために自サイトをRefererとして付与する必要がある */
-const EXTRA_HEADERS = { Referer: `${BASE_URL}/` };
-
-/**
- * "ボリス・チェルニー 著　折山文哉 訳" → ["ボリス・チェルニー", "折山文哉"]
- * 役割語（著・訳・編・監訳・監修など）を除去する。
- */
-function parseAuthorsFromText(text: string): string[] {
-  return text.split(/[　\s]+(?=\S)/).map(stripAuthorRole).filter(Boolean);
-}
-
-/**
- * div.author 内の各リンクから役割語を除去して著者名のみ返す。
- */
-function parseAuthorLinks(authors: string[]): string[] {
-  return authors.map(stripAuthorRole).filter(Boolean);
-}
-
-/**
- * "2020/03/31" → "2020-03-31"
- * "2020年3月31日" → "2020-03-31"
- */
-function parseDate(text: string): string | undefined {
-  const m1 = text.match(/(\d{4})\/(\d{1,2})\/(\d{1,2})/);
-  if (m1) {
-    return `${m1[1]}-${m1[2].padStart(2, "0")}-${m1[3].padStart(2, "0")}`;
-  }
-  return parseJapaneseDateToISO(text);
-}
-
-export const maruzenPublishingAdapter: PublisherAdapter = {
-  id: "maruzen-publishing",
-  name: "丸善出版",
-  baseUrl: BASE_URL,
-
-  async search(query: SearchQuery, deps: PublisherDeps): Promise<BookRecord[]> {
-    const word = [query.title, query.author].filter(Boolean).join(" ");
-    if (!word) return [];
-
-    const url = `${SEARCH_URL}?search_keyword=${encodeURIComponent(word)}&format=1`;
-    const html = await fetchText(url, deps, EXTRA_HEADERS);
-    const doc = deps.parser.parse(html);
-
-    const results: BookRecord[] = [];
-    const limit = query.limit ?? 10;
-
-    for (const item of doc.select("div.booklist div.item")) {
-      const linkEl = item.find("div.ttl a")[0];
-      if (!linkEl) continue;
-
-      const title = linkEl.text().trim();
-      const href = linkEl.attr("href");
-      if (!title || !href) continue;
-
-      const bookUrl = href.startsWith("http") ? href : `${BASE_URL}${href}`;
-
-      const authorLinks = item.find("div.author a").map(el => el.text().trim());
-      const authors = parseAuthorLinks(authorLinks);
-
-      const imgEl = item.find("div.image img")[0];
-      const imgSrc = imgEl?.attr("src");
-      const coverImageUrl = imgSrc ? resolveUrl(BASE_URL, imgSrc) : undefined;
-
-      results.push({
-        title,
-        authors,
-        publisher: "丸善出版",
-        url: bookUrl,
-        coverImageUrl,
-      });
-
-      if (results.length >= limit) break;
-    }
-
-    return results;
-  },
-
-  async getDetail(url: string, deps: PublisherDeps): Promise<BookRecord> {
-    const html = await fetchText(url, deps, EXTRA_HEADERS);
-    const doc = deps.parser.parse(html);
-
-    // タイトルは <title> タグ、またはh1など（サイト構造に依存）
-    // div#bookData table の th/td からメタデータを取得
-    const table = doc.selectOne("div#bookData table");
-    let authors: string[] = [];
-    let publisher = "丸善出版";
-    let publishedAt: string | undefined;
-
-    if (table) {
-      for (const row of table.find("tr")) {
-        const th = row.find("th")[0]?.text().trim() ?? "";
-        const tdText = row.find("td")[0]?.text().trim() ?? "";
-
-        if (th === "著者") {
-          authors = parseAuthorsFromText(tdText);
-        } else if (th === "発行元") {
-          publisher = tdText || "丸善出版";
-        } else if (th === "出版年月日") {
-          publishedAt = parseDate(tdText);
-        }
-      }
-    }
-
-    // ページタイトルから書籍タイトルを取得（" | 丸善出版" を除去）
-    const pageTitle = doc.selectOne("title")?.text().trim() ?? "";
-    const title = pageTitle.replace(/\s*[|｜]\s*丸善出版.*$/, "").trim();
-
-    // 電子書籍ストア（kw.maruzen.co.jp は機関向けなので除外）
-    const ebookStores = extractEbookStoresFromDoc(doc).filter(
-      store => !store.url.includes("kw.maruzen.co.jp"),
-    );
-
-    return {
-      title,
-      authors,
-      publisher,
-      url,
-      publishedAt,
-      ebookStores: ebookStores.length > 0 ? ebookStores : undefined,
-    };
-  },
-};