aeo.js 0.0.2 → 0.0.4

package/dist/next.mjs

@@ -162,7 +162,7 @@ function detectFramework(projectRoot = process.cwd()) {
   };
 }
 function resolveConfig(config = {}) {
-  var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j, _k, _l, _m, _n, _o, _p, _q, _r, _s, _t, _u, _v, _w, _x;
+  var _a, _b, _c, _d, _e, _f, _g, _h, _i, _j, _k, _l, _m, _n, _o, _p, _q, _r, _s, _t, _u, _v, _w, _x, _y, _z, _A, _B, _C, _D, _E, _F, _G, _H, _I, _J, _K, _L, _M;
   const frameworkInfo = detectFramework();
   return {
     title: config.title || "My Site",
@@ -178,26 +178,43 @@ function resolveConfig(config = {}) {
       rawMarkdown: ((_d = config.generators) == null ? void 0 : _d.rawMarkdown) !== false,
       manifest: ((_e = config.generators) == null ? void 0 : _e.manifest) !== false,
       sitemap: ((_f = config.generators) == null ? void 0 : _f.sitemap) !== false,
-      aiIndex: ((_g = config.generators) == null ? void 0 : _g.aiIndex) !== false
+      aiIndex: ((_g = config.generators) == null ? void 0 : _g.aiIndex) !== false,
+      schema: ((_h = config.generators) == null ? void 0 : _h.schema) !== false
     },
     robots: {
-      allow: ((_h = config.robots) == null ? void 0 : _h.allow) || ["/"],
-      disallow: ((_i = config.robots) == null ? void 0 : _i.disallow) || [],
-      crawlDelay: ((_j = config.robots) == null ? void 0 : _j.crawlDelay) || 0,
-      sitemap: ((_k = config.robots) == null ? void 0 : _k.sitemap) || ""
+      allow: ((_i = config.robots) == null ? void 0 : _i.allow) || ["/"],
+      disallow: ((_j = config.robots) == null ? void 0 : _j.disallow) || [],
+      crawlDelay: ((_k = config.robots) == null ? void 0 : _k.crawlDelay) || 0,
+      sitemap: ((_l = config.robots) == null ? void 0 : _l.sitemap) || ""
+    },
+    schema: {
+      enabled: ((_m = config.schema) == null ? void 0 : _m.enabled) !== false,
+      organization: {
+        name: ((_o = (_n = config.schema) == null ? void 0 : _n.organization) == null ? void 0 : _o.name) || config.title || "My Site",
+        url: ((_q = (_p = config.schema) == null ? void 0 : _p.organization) == null ? void 0 : _q.url) || config.url || "https://example.com",
+        logo: ((_s = (_r = config.schema) == null ? void 0 : _r.organization) == null ? void 0 : _s.logo) || "",
+        sameAs: ((_u = (_t = config.schema) == null ? void 0 : _t.organization) == null ? void 0 : _u.sameAs) || []
+      },
+      defaultType: ((_v = config.schema) == null ? void 0 : _v.defaultType) || "WebPage"
+    },
+    og: {
+      enabled: ((_w = config.og) == null ? void 0 : _w.enabled) !== false,
+      image: ((_x = config.og) == null ? void 0 : _x.image) || "",
+      twitterHandle: ((_y = config.og) == null ? void 0 : _y.twitterHandle) || "",
+      type: ((_z = config.og) == null ? void 0 : _z.type) || "website"
     },
     widget: {
-      enabled: ((_l = config.widget) == null ? void 0 : _l.enabled) !== false,
-      position: ((_m = config.widget) == null ? void 0 : _m.position) || "bottom-right",
+      enabled: ((_A = config.widget) == null ? void 0 : _A.enabled) !== false,
+      position: ((_B = config.widget) == null ? void 0 : _B.position) || "bottom-right",
       theme: {
-        background: ((_o = (_n = config.widget) == null ? void 0 : _n.theme) == null ? void 0 : _o.background) || "rgba(18, 18, 24, 0.9)",
-        text: ((_q = (_p = config.widget) == null ? void 0 : _p.theme) == null ? void 0 : _q.text) || "#C0C0C5",
-        accent: ((_s = (_r = config.widget) == null ? void 0 : _r.theme) == null ? void 0 : _s.accent) || "#E8E8EA",
-        badge: ((_u = (_t = config.widget) == null ? void 0 : _t.theme) == null ? void 0 : _u.badge) || "#4ADE80"
+        background: ((_D = (_C = config.widget) == null ? void 0 : _C.theme) == null ? void 0 : _D.background) || "rgba(18, 18, 24, 0.9)",
+        text: ((_F = (_E = config.widget) == null ? void 0 : _E.theme) == null ? void 0 : _F.text) || "#C0C0C5",
+        accent: ((_H = (_G = config.widget) == null ? void 0 : _G.theme) == null ? void 0 : _H.accent) || "#E8E8EA",
+        badge: ((_J = (_I = config.widget) == null ? void 0 : _I.theme) == null ? void 0 : _J.badge) || "#4ADE80"
       },
-      humanLabel: ((_v = config.widget) == null ? void 0 : _v.humanLabel) || "Human",
-      aiLabel: ((_w = config.widget) == null ? void 0 : _w.aiLabel) || "AI",
-      showBadge: ((_x = config.widget) == null ? void 0 : _x.showBadge) !== false
+      humanLabel: ((_K = config.widget) == null ? void 0 : _K.humanLabel) || "Human",
+      aiLabel: ((_L = config.widget) == null ? void 0 : _L.aiLabel) || "AI",
+      showBadge: ((_M = config.widget) == null ? void 0 : _M.showBadge) !== false
     }
   };
 }
@@ -496,8 +513,8 @@ function generatePageMarkdownFiles(config) {
   const generated = [];
   const pages = config.pages || [];
   for (const page of pages) {
+    if (!page.content) continue;
     const pageTitle = page.title || (page.pathname === "/" ? config.title : void 0);
-    if (!page.content && !pageTitle) continue;
     let filename;
     if (page.pathname === "/") {
       filename = "index.md";
@@ -767,6 +784,140 @@ function generateAIIndex(config) {
   };
   return JSON.stringify(index, null, 2);
 }
+
+// src/core/schema.ts
+function generateSchema(config) {
+  const output = generateSchemaObjects(config);
+  return JSON.stringify(output, null, 2);
+}
+function generateSchemaObjects(config) {
+  const siteSchemas = generateSiteSchemas(config);
+  const pageSchemas = {};
+  for (const page of config.pages) {
+    const schemas = generatePageSchemas(page, config);
+    if (schemas.length > 0) {
+      pageSchemas[page.pathname] = schemas;
+    }
+  }
+  return { site: siteSchemas, pages: pageSchemas };
+}
+function generateSiteSchemas(config) {
+  const schemas = [];
+  schemas.push({
+    "@context": "https://schema.org",
+    "@type": "WebSite",
+    name: config.title,
+    description: config.description || void 0,
+    url: config.url
+  });
+  const org = config.schema.organization;
+  if (org.name || org.sameAs.length > 0) {
+    const orgSchema = {
+      "@context": "https://schema.org",
+      "@type": "Organization",
+      name: org.name,
+      url: org.url
+    };
+    if (org.logo) orgSchema.logo = org.logo;
+    if (org.sameAs.length > 0) orgSchema.sameAs = org.sameAs;
+    schemas.push(orgSchema);
+  }
+  return schemas;
+}
+function generatePageSchemas(page, config) {
+  const schemas = [];
+  const pageUrl = page.pathname === "/" ? config.url : `${config.url.replace(/\/$/, "")}${page.pathname}`;
+  const faqItems = detectFaqPatterns(page.content || "");
+  if (faqItems.length > 0) {
+    schemas.push({
+      "@context": "https://schema.org",
+      "@type": "FAQPage",
+      mainEntity: faqItems.map(({ question, answer }) => ({
+        "@type": "Question",
+        name: question,
+        acceptedAnswer: {
+          "@type": "Answer",
+          text: answer
+        }
+      }))
+    });
+  }
+  const pageType = config.schema.defaultType;
+  const pageSchema = {
+    "@context": "https://schema.org",
+    "@type": pageType,
+    name: page.title || config.title,
+    url: pageUrl
+  };
+  if (page.description) pageSchema.description = page.description;
+  if (pageType === "Article") {
+    pageSchema.headline = page.title || config.title;
+    pageSchema.author = {
+      "@type": "Organization",
+      name: config.schema.organization.name
+    };
+  }
+  schemas.push(pageSchema);
+  if (page.pathname !== "/") {
+    const breadcrumbs = generateBreadcrumbs(page.pathname, config);
+    if (breadcrumbs.length > 1) {
+      schemas.push({
+        "@context": "https://schema.org",
+        "@type": "BreadcrumbList",
+        itemListElement: breadcrumbs.map((crumb, i) => ({
+          "@type": "ListItem",
+          position: i + 1,
+          name: crumb.name,
+          item: crumb.url
+        }))
+      });
+    }
+  }
+  return schemas;
+}
+function generateBreadcrumbs(pathname, config) {
+  const baseUrl = config.url.replace(/\/$/, "");
+  const parts = pathname.split("/").filter(Boolean);
+  const crumbs = [
+    { name: "Home", url: baseUrl + "/" }
+  ];
+  let currentPath = "";
+  for (const part of parts) {
+    currentPath += "/" + part;
+    crumbs.push({
+      name: part.charAt(0).toUpperCase() + part.slice(1).replace(/-/g, " "),
+      url: baseUrl + currentPath
+    });
+  }
+  return crumbs;
+}
+function detectFaqPatterns(content) {
+  const items = [];
+  const lines = content.split("\n");
+  for (let i = 0; i < lines.length; i++) {
+    const line = lines[i].trim();
+    const headingMatch = line.match(/^#{1,6}\s+(.+\?)\s*$/);
+    if (headingMatch) {
+      const answerLines = [];
+      for (let j = i + 1; j < lines.length; j++) {
+        const nextLine = lines[j].trim();
+        if (!nextLine) {
+          if (answerLines.length > 0) break;
+          continue;
+        }
+        if (/^#{1,6}\s/.test(nextLine)) break;
+        answerLines.push(nextLine);
+      }
+      if (answerLines.length > 0) {
+        items.push({
+          question: headingMatch[1],
+          answer: answerLines.join(" ").slice(0, 500)
+        });
+      }
+    }
+  }
+  return items;
+}
 async function generateAEOFiles(configOrRoot, maybeConfig) {
   var _a;
   let config;
@@ -795,7 +946,7 @@ async function generateAEOFiles(configOrRoot, maybeConfig) {
   if (config.generators.llmsTxt) {
     try {
       const content = generateLlmsTxt(config);
-      writeFileSync(join(outDir, "llms.txt"), "\uFEFF" + content, "utf-8");
+      writeFileSync(join(outDir, "llms.txt"), content, "utf-8");
       files.push("llms.txt");
     } catch (e) {
       errors.push(`llms.txt: ${e.message}`);
@@ -804,7 +955,7 @@ async function generateAEOFiles(configOrRoot, maybeConfig) {
   if (config.generators.llmsFullTxt) {
     try {
       const content = generateLlmsFullTxt(config);
-      writeFileSync(join(outDir, "llms-full.txt"), "\uFEFF" + content, "utf-8");
+      writeFileSync(join(outDir, "llms-full.txt"), content, "utf-8");
       files.push("llms-full.txt");
     } catch (e) {
       errors.push(`llms-full.txt: ${e.message}`);
@@ -855,8 +1006,78 @@ async function generateAEOFiles(configOrRoot, maybeConfig) {
       errors.push(`ai-index.json: ${e.message}`);
     }
   }
+  if (config.generators.schema && config.schema.enabled) {
+    try {
+      const content = generateSchema(config);
+      writeFileSync(join(outDir, "schema.json"), content, "utf-8");
+      files.push("schema.json");
+    } catch (e) {
+      errors.push(`schema.json: ${e.message}`);
+    }
+  }
   return { files, errors };
 }
+
+// src/core/html-extract.ts
+function extractTextFromHtml(html) {
+  let text = html;
+  text = text.replace(/<script[\s\S]*?<\/script>/gi, "");
+  text = text.replace(/<style[\s\S]*?<\/style>/gi, "");
+  text = text.replace(/<svg[\s\S]*?<\/svg>/gi, "");
+  const mainMatch = text.match(/<main[^>]*>([\s\S]*)<\/main>/i);
+  if (mainMatch) {
+    text = mainMatch[1];
+  } else {
+    text = text.replace(/<nav[\s\S]*?<\/nav>/gi, "");
+    text = text.replace(/<header[\s\S]*?<\/header>/gi, "");
+    text = text.replace(/<footer[\s\S]*?<\/footer>/gi, "");
+  }
+  text = text.replace(/<a[^>]+href=["']([^"']*)["'][^>]*>([\s\S]*?)<\/a>/gi, (_, url, inner) => {
+    if (/<(?:h[1-6]|div|p|section)[^>]*>/i.test(inner)) {
+      const cleanInner = inner.replace(/<[^>]+>/g, " ").replace(/\s+/g, " ").trim();
+      return `
+[${cleanInner.slice(0, 120).trim()}](${url})
+`;
+    }
+    return `[${inner}](${url})`;
+  });
+  text = text.replace(/<h1[^>]*>([\s\S]*?)<\/h1>/gi, "\n\n## $1\n\n");
+  text = text.replace(/<h2[^>]*>([\s\S]*?)<\/h2>/gi, "\n\n## $1\n\n");
+  text = text.replace(/<h3[^>]*>([\s\S]*?)<\/h3>/gi, "\n\n### $1\n\n");
+  text = text.replace(/<h4[^>]*>([\s\S]*?)<\/h4>/gi, "\n\n#### $1\n\n");
+  text = text.replace(/<h5[^>]*>([\s\S]*?)<\/h5>/gi, "\n\n##### $1\n\n");
+  text = text.replace(/<h6[^>]*>([\s\S]*?)<\/h6>/gi, "\n\n###### $1\n\n");
+  text = text.replace(/<a[^>]+href=["']([^"']*)["'][^>]*>([\s\S]*?)<\/a>/gi, "[$2]($1)");
+  text = text.replace(/<(?:strong|b)[^>]*>([\s\S]*?)<\/(?:strong|b)>/gi, "**$1**");
+  text = text.replace(/<(?:em|i)[^>]*>([\s\S]*?)<\/(?:em|i)>/gi, "*$1*");
+  text = text.replace(/<li[^>]*>([\s\S]*?)<\/li>/gi, "\n- $1");
+  text = text.replace(/<blockquote[^>]*>([\s\S]*?)<\/blockquote>/gi, "\n\n> $1\n\n");
+  text = text.replace(/<hr[^>]*\/?>/gi, "\n\n---\n\n");
+  text = text.replace(/<br[^>]*\/?>/gi, "\n");
+  text = text.replace(/<\/p>/gi, "\n\n");
+  text = text.replace(/<p[^>]*>/gi, "");
+  text = text.replace(/<\/?(?:div|section|article|header|main|aside|figure|figcaption|table|thead|tbody|tr|td|th|ul|ol|dl|dt|dd)[^>]*>/gi, "\n");
+  text = text.replace(/<[^>]+>/g, "");
+  text = text.replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&quot;/g, '"').replace(/&#39;/g, "'").replace(/&nbsp;/g, " ").replace(/&copy;/g, "(c)");
+  text = text.replace(/[\u{1F1E0}-\u{1FAFF}\u{2600}-\u{27BF}\u{FE00}-\u{FE0F}\u{200D}\u{20E3}]/gu, "");
+  text = text.split("\n").map((l) => l.replace(/\s+/g, " ").trim()).join("\n");
+  text = text.replace(/\n{3,}/g, "\n\n");
+  text = text.replace(/\[[\s\n]+/g, "[").replace(/[\s\n]+\]/g, "]");
+  text = text.replace(/(#{2,6})\s*\n+\s*/g, "$1 ");
+  text = text.replace(/^#{2,6}\s*$/gm, "");
+  text = text.replace(/\n{3,}/g, "\n\n");
+  return text.trim().slice(0, 8e3);
+}
+function extractTitle2(html) {
+  var _a, _b;
+  const match = html.match(/<title>([^<]*)<\/title>/i);
+  if (!match) return void 0;
+  return ((_b = (_a = match[1]) == null ? void 0 : _a.split("|")[0]) == null ? void 0 : _b.trim()) || match[1];
+}
+function extractDescription(html) {
+  const match = html.match(/<meta\s+name=["']description["']\s+content=["']([^"']*)["']/i);
+  return match == null ? void 0 : match[1];
+}
 function scanNextPages(projectRoot) {
   const pages = [];
   for (const base of ["app", "src/app"]) {
@@ -971,7 +1192,7 @@ async function generateAeoMetadata(config) {
   if (process.env.NODE_ENV === "production") {
     await generateAEOFiles(resolvedConfig);
   }
-  return {
+  const metadata = {
     title: resolvedConfig.title,
     description: resolvedConfig.description,
     alternates: {
@@ -987,62 +1208,30 @@ async function generateAeoMetadata(config) {
       }
     }
   };
-}
-function extractText(html) {
-  let text = html;
-  text = text.replace(/<script[\s\S]*?<\/script>/gi, "");
-  text = text.replace(/<style[\s\S]*?<\/style>/gi, "");
-  text = text.replace(/<svg[\s\S]*?<\/svg>/gi, "");
-  const mainMatch = text.match(/<main[^>]*>([\s\S]*)<\/main>/i);
-  if (mainMatch) {
-    text = mainMatch[1];
-  } else {
-    text = text.replace(/<nav[\s\S]*?<\/nav>/gi, "");
-    text = text.replace(/<header[\s\S]*?<\/header>/gi, "");
-    text = text.replace(/<footer[\s\S]*?<\/footer>/gi, "");
+  if (resolvedConfig.og.enabled) {
+    metadata.openGraph = {
+      type: resolvedConfig.og.type,
+      title: resolvedConfig.title,
+      description: resolvedConfig.description,
+      url: resolvedConfig.url,
+      siteName: resolvedConfig.title,
+      ...resolvedConfig.og.image ? { images: [{ url: resolvedConfig.og.image }] } : {}
+    };
+    metadata.twitter = {
+      card: resolvedConfig.og.image ? "summary_large_image" : "summary",
+      title: resolvedConfig.title,
+      description: resolvedConfig.description,
+      ...resolvedConfig.og.twitterHandle ? { site: resolvedConfig.og.twitterHandle } : {},
+      ...resolvedConfig.og.image ? { images: [resolvedConfig.og.image] } : {}
+    };
   }
-  text = text.replace(/<a[^>]+href=["']([^"']*)["'][^>]*>([\s\S]*?)<\/a>/gi, (_, url, inner) => {
-    if (/<(?:h[1-6]|div|p|section)[^>]*>/i.test(inner)) {
-      const cleanInner = inner.replace(/<[^>]+>/g, " ").replace(/\s+/g, " ").trim();
-      return `
-[${cleanInner.slice(0, 120).trim()}](${url})
-`;
-    }
-    return `[${inner}](${url})`;
-  });
-  text = text.replace(/<h1[^>]*>([\s\S]*?)<\/h1>/gi, "\n\n## $1\n\n");
-  text = text.replace(/<h2[^>]*>([\s\S]*?)<\/h2>/gi, "\n\n## $1\n\n");
-  text = text.replace(/<h3[^>]*>([\s\S]*?)<\/h3>/gi, "\n\n### $1\n\n");
-  text = text.replace(/<h4[^>]*>([\s\S]*?)<\/h4>/gi, "\n\n#### $1\n\n");
-  text = text.replace(/<h5[^>]*>([\s\S]*?)<\/h5>/gi, "\n\n##### $1\n\n");
-  text = text.replace(/<h6[^>]*>([\s\S]*?)<\/h6>/gi, "\n\n###### $1\n\n");
-  text = text.replace(/<a[^>]+href=["']([^"']*)["'][^>]*>([\s\S]*?)<\/a>/gi, "[$2]($1)");
-  text = text.replace(/<(?:strong|b)[^>]*>([\s\S]*?)<\/(?:strong|b)>/gi, "**$1**");
-  text = text.replace(/<(?:em|i)[^>]*>([\s\S]*?)<\/(?:em|i)>/gi, "*$1*");
-  text = text.replace(/<li[^>]*>([\s\S]*?)<\/li>/gi, "\n- $1");
-  text = text.replace(/<blockquote[^>]*>([\s\S]*?)<\/blockquote>/gi, "\n\n> $1\n\n");
-  text = text.replace(/<hr[^>]*\/?>/gi, "\n\n---\n\n");
-  text = text.replace(/<br[^>]*\/?>/gi, "\n");
-  text = text.replace(/<\/p>/gi, "\n\n");
-  text = text.replace(/<p[^>]*>/gi, "");
-  text = text.replace(/<\/?(?:div|section|article|header|main|aside|figure|figcaption|table|thead|tbody|tr|td|th|ul|ol|dl|dt|dd)[^>]*>/gi, "\n");
-  text = text.replace(/<[^>]+>/g, "");
-  text = text.replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&quot;/g, '"').replace(/&#39;/g, "'").replace(/&nbsp;/g, " ").replace(/&copy;/g, "(c)");
-  text = text.replace(/[\u{1F1E0}-\u{1FAFF}\u{2600}-\u{27BF}\u{FE00}-\u{FE0F}\u{200D}\u{20E3}]/gu, "");
-  text = text.split("\n").map((l) => l.replace(/\s+/g, " ").trim()).join("\n");
-  text = text.replace(/\n{3,}/g, "\n\n");
-  text = text.replace(/\[[\s\n]+/g, "[").replace(/[\s\n]+\]/g, "]");
-  text = text.replace(/(#{2,6})\s*\n+\s*/g, "$1 ");
-  text = text.replace(/^#{2,6}\s*$/gm, "");
-  text = text.replace(/\n{3,}/g, "\n\n");
-  return text.trim().slice(0, 8e3);
+  return metadata;
 }
 function scanNextBuildOutput(projectRoot) {
   const pages = [];
   const serverAppDir = join(projectRoot, ".next", "server", "app");
   if (!existsSync(serverAppDir)) return pages;
   function walk(dir, basePath = "") {
-    var _a, _b;
     try {
       const entries = readdirSync(dir);
       for (const entry of entries) {
@@ -1052,14 +1241,14 @@ function scanNextBuildOutput(projectRoot) {
           walk(fullPath, `${basePath}/${entry}`);
         } else if (entry === "index.html") {
           const html = readFileSync(fullPath, "utf-8");
-          const titleMatch = html.match(/<title>([^<]*)<\/title>/i);
-          const descMatch = html.match(/<meta\s+name=["']description["']\s+content=["']([^"']*)["']/i);
-          const textContent = extractText(html);
+          const title = extractTitle2(html);
+          const description = extractDescription(html);
+          const textContent = extractTextFromHtml(html);
           const pathname = basePath || "/";
           pages.push({
             pathname,
-            title: (_b = (_a = titleMatch == null ? void 0 : titleMatch[1]) == null ? void 0 : _a.split("|")[0]) == null ? void 0 : _b.trim(),
-            description: descMatch == null ? void 0 : descMatch[1],
+            title,
+            description,
             content: textContent
           });
         }