@farm.js/core 0.1.60 → 0.1.61

package/dist/chunk-T4TBK5NY.mjs

@@ -1,593 +0,0 @@
-import {
-  isFarmMetadataNoIndex
-} from "./chunk-KVW2M6PI.mjs";
-import {
-  matchesFarmMarkdownRoutePattern
-} from "./chunk-YSXTNSXS.mjs";
-import {
-  applyFarmBasePath
-} from "./chunk-SF3JB56B.mjs";
-
-// src/agent-crawlers.ts
-var FARM_AI_CRAWLERS = {
-  search: [
-    "OAI-SearchBot",
-    "ChatGPT-User",
-    "Claude-SearchBot",
-    "Claude-User",
-    "PerplexityBot",
-    "Perplexity-User"
-  ],
-  training: [
-    "GPTBot",
-    "ClaudeBot",
-    "Google-Extended",
-    "CCBot",
-    "Applebot-Extended",
-    "Meta-ExternalAgent",
-    "Bytespider"
-  ]
-};
-var LINE_BREAK = /[\r\n]/;
-function resolveFarmAgentCrawlers(input) {
-  if (!input) return { enabled: false, rules: [] };
-  if (input === true) return { enabled: true, rules: [] };
-  if (typeof input !== "object" || Array.isArray(input)) {
-    throw new TypeError("agent.crawlers must be true, false, or an options object.");
-  }
-  if (input.enabled === false) return { enabled: false, rules: [] };
-  for (const key of ["search", "training"]) {
-    const policy = input[key];
-    if (policy !== void 0 && policy !== "allow" && policy !== "block") {
-      throw new TypeError(`agent.crawlers.${key} must be "allow" or "block".`);
-    }
-  }
-  return {
-    enabled: true,
-    ...input.search ? { search: input.search } : {},
-    ...input.training ? { training: input.training } : {},
-    rules: resolveRobotsRules(input.rules),
-    ...input.sitemap !== void 0 ? { sitemap: resolveSitemapUrls(input.sitemap) } : {}
-  };
-}
-function resolveRobotsRules(value) {
-  if (value === void 0) return [];
-  if (!Array.isArray(value)) {
-    throw new TypeError("agent.crawlers.rules must be an array of robots.txt rules.");
-  }
-  return value.map((rule, index) => {
-    const name = `agent.crawlers.rules[${index}]`;
-    if (!rule || typeof rule !== "object" || Array.isArray(rule)) {
-      throw new TypeError(`${name} must be an object.`);
-    }
-    const userAgent = readDirectiveValues(rule.userAgent, `${name}.userAgent`);
-    if (userAgent.length === 0) {
-      throw new TypeError(`${name}.userAgent must name at least one user agent.`);
-    }
-    const allow = readDirectiveValues(rule.allow, `${name}.allow`);
-    const disallow = readDirectiveValues(rule.disallow, `${name}.disallow`);
-    if (rule.crawlDelay !== void 0 && (typeof rule.crawlDelay !== "number" || !Number.isFinite(rule.crawlDelay) || rule.crawlDelay < 0)) {
-      throw new TypeError(`${name}.crawlDelay must be a number of seconds.`);
-    }
-    return {
-      userAgent,
-      ...allow.length ? { allow } : {},
-      ...disallow.length ? { disallow } : {},
-      ...rule.crawlDelay !== void 0 ? { crawlDelay: rule.crawlDelay } : {}
-    };
-  });
-}
-function readDirectiveValues(value, name) {
-  if (value === void 0) return [];
-  const values = Array.isArray(value) ? value : [value];
-  for (const item of values) {
-    if (typeof item !== "string" || !item || LINE_BREAK.test(item)) {
-      throw new TypeError(`${name} must be a non-empty string or an array of them.`);
-    }
-  }
-  return values;
-}
-function resolveSitemapUrls(value) {
-  const values = Array.isArray(value) ? value : [value];
-  for (const item of values) {
-    if (typeof item !== "string" || !toAbsoluteHttpUrl(item) || LINE_BREAK.test(item)) {
-      throw new TypeError("agent.crawlers.sitemap must be an absolute URL or an array of them.");
-    }
-  }
-  return values;
-}
-function toAbsoluteHttpUrl(value) {
-  if (typeof value !== "string" && !(value instanceof URL)) return null;
-  try {
-    const url = new URL(String(value));
-    return url.protocol === "https:" || url.protocol === "http:" ? url : null;
-  } catch {
-    return null;
-  }
-}
-function createFarmAgentRobots(options) {
-  const { crawlers } = options;
-  const rules = [];
-  for (const group of ["search", "training"]) {
-    const policy = crawlers[group];
-    if (!policy) continue;
-    rules.push({
-      userAgent: [...FARM_AI_CRAWLERS[group]],
-      ...policy === "allow" ? { allow: "/" } : { disallow: "/" }
-    });
-  }
-  rules.push(...crawlers.rules);
-  const namesEveryAgent = crawlers.rules.some(
-    (rule) => (Array.isArray(rule.userAgent) ? rule.userAgent : [rule.userAgent]).includes("*")
-  );
-  if (!namesEveryAgent) rules.push({ userAgent: "*", allow: "/" });
-  const sitemap = crawlers.sitemap ?? resolveDefaultSitemap(options);
-  return { rules, ...sitemap.length > 0 ? { sitemap } : {} };
-}
-function resolveDefaultSitemap(options) {
-  if (!options.sitemapPath) return [];
-  const base = toAbsoluteHttpUrl(options.metadataBase);
-  if (!base) return [];
-  return [new URL(applyFarmBasePath(options.sitemapPath, options.basePath), base).href];
-}
-
-// src/agent-sitemap.ts
-function resolveFarmAgentSitemap(input) {
-  if (!input) return { enabled: false, include: [], exclude: [] };
-  if (input === true) return { enabled: true, include: [], exclude: [] };
-  if (typeof input !== "object") {
-    throw new TypeError("agent.sitemap must be true, false, or an options object.");
-  }
-  if (input.enabled === false) return { enabled: false, include: [], exclude: [] };
-  return {
-    enabled: true,
-    include: resolveRoutePatterns(input.include, "include"),
-    exclude: resolveRoutePatterns(input.exclude, "exclude")
-  };
-}
-function resolveRoutePatterns(value, key) {
-  if (value === void 0) return [];
-  if (!Array.isArray(value)) {
-    throw new TypeError(`agent.sitemap.${key} must be an array of route patterns.`);
-  }
-  return value.map((pattern, index) => {
-    if (typeof pattern !== "string" || !pattern.startsWith("/")) {
-      throw new TypeError(
-        `agent.sitemap.${key}[${index}] must be a route pattern starting with "/", such as "/admin/[...path]".`
-      );
-    }
-    return pattern;
-  });
-}
-function createFarmDefaultSitemap(sources, options) {
-  const { include, exclude } = options.config;
-  const base = toAbsoluteHttpOrigin(options.metadataBase) ?? options.origin;
-  const entries = /* @__PURE__ */ new Map();
-  for (const source of sources) {
-    const path = source.pattern;
-    if (path.includes("[") || path.includes("(") || entries.has(path)) continue;
-    if (include.length && !include.some((pattern) => matchesFarmMarkdownRoutePattern(pattern, path))) {
-      continue;
-    }
-    if (exclude.some((pattern) => matchesFarmMarkdownRoutePattern(pattern, path))) continue;
-    const module = isRecord(source.module) ? source.module : void 0;
-    if (isFarmMetadataNoIndex(module?.metadata)) continue;
-    if (module?.__farmRouteGuard) continue;
-    entries.set(path, {
-      url: new URL(applyFarmBasePath(path, options.basePath), base).href,
-      ...source.lastModified ? { lastModified: source.lastModified } : {}
-    });
-  }
-  return [...entries.entries()].sort(([left], [right]) => left.localeCompare(right)).map(([, entry]) => entry);
-}
-function toAbsoluteHttpOrigin(value) {
-  if (typeof value !== "string" && !(value instanceof URL)) return void 0;
-  try {
-    const url = new URL(String(value));
-    return url.protocol === "https:" || url.protocol === "http:" ? url.origin : void 0;
-  } catch {
-    return void 0;
-  }
-}
-function isRecord(value) {
-  return Boolean(value && typeof value === "object");
-}
-
-// src/preload.ts
-var DEFAULT_PRELOAD_CONFIG = {
-  mode: "enforce",
-  maxImages: 1,
-  maxFonts: 2
-};
-var reportedWarnings = /* @__PURE__ */ new Map();
-var nextWarningExpiry = Infinity;
-var PRELOAD_WARNING_TTL_MS = 6e4;
-var MAX_REPORTED_PRELOAD_WARNINGS = 256;
-function resolveFarmPerformanceConfig(config) {
-  return {
-    preload: {
-      mode: config?.preload?.mode === "warn" ? "warn" : "enforce",
-      maxImages: normalizeBudget(config?.preload?.maxImages, DEFAULT_PRELOAD_CONFIG.maxImages),
-      maxFonts: normalizeBudget(config?.preload?.maxFonts, DEFAULT_PRELOAD_CONFIG.maxFonts)
-    }
-  };
-}
-function manageFarmHtmlPreloads(html, config) {
-  const elements = findHtmlLinkElements(html);
-  const candidates = elements.flatMap((element, index) => {
-    const kind = getHtmlPreloadKind(element.value);
-    return kind ? [
-      {
-        index,
-        kind,
-        href: readHtmlAttribute(element.value, "href"),
-        highPriority: readHtmlAttribute(element.value, "fetchpriority")?.toLowerCase() === "high"
-      }
-    ] : [];
-  });
-  const budget = selectPreloadsWithinBudget(candidates, config);
-  if (config.mode === "warn" || budget.removed.size === 0) {
-    return { value: html, warnings: budget.warnings };
-  }
-  return {
-    value: removeHtmlLinkElements(html, elements, budget.removed),
-    warnings: budget.warnings
-  };
-}
-function manageFarmLinkHeaderPreloads(value, config) {
-  if (!value) return { value, warnings: [] };
-  const links = splitLinkHeader(value);
-  const candidates = links.flatMap((link, index) => {
-    const kind = getLinkHeaderPreloadKind(link);
-    return kind ? [
-      {
-        index,
-        kind,
-        href: getLinkHeaderTarget(link),
-        highPriority: getLinkHeaderParameter(link, "fetchpriority")?.toLowerCase() === "high"
-      }
-    ] : [];
-  });
-  const budget = selectPreloadsWithinBudget(candidates, config);
-  return {
-    value: config.mode === "enforce" ? links.filter((_, index) => !budget.removed.has(index)).join(", ") : value,
-    warnings: budget.warnings
-  };
-}
-function manageFarmDocumentPreloads(html, linkHeader, config) {
-  const elements = findHtmlLinkElements(html);
-  const links = linkHeader ? splitLinkHeader(linkHeader) : [];
-  const headerOffset = elements.length;
-  const candidates = [];
-  for (const [index, element] of elements.entries()) {
-    const kind = getHtmlPreloadKind(element.value);
-    if (!kind) continue;
-    candidates.push({
-      index,
-      kind,
-      href: readHtmlAttribute(element.value, "href"),
-      highPriority: readHtmlAttribute(element.value, "fetchpriority")?.toLowerCase() === "high"
-    });
-  }
-  for (const [index, link] of links.entries()) {
-    const kind = getLinkHeaderPreloadKind(link);
-    if (!kind) continue;
-    candidates.push({
-      index: headerOffset + index,
-      kind,
-      href: getLinkHeaderTarget(link),
-      highPriority: getLinkHeaderParameter(link, "fetchpriority")?.toLowerCase() === "high"
-    });
-  }
-  const budget = selectPreloadsWithinBudget(candidates, config);
-  if (config.mode === "warn" || budget.removed.size === 0) {
-    return { html, linkHeader, warnings: budget.warnings };
-  }
-  return {
-    html: removeHtmlLinkElements(html, elements, budget.removed),
-    linkHeader: links.filter((_, index) => !budget.removed.has(headerOffset + index)).join(", "),
-    warnings: budget.warnings
-  };
-}
-function reportFarmPreloadWarnings(warnings, context = "the rendered document") {
-  if (warnings.length === 0 && reportedWarnings.size === 0) return;
-  const now = Date.now();
-  if (now >= nextWarningExpiry) {
-    nextWarningExpiry = Infinity;
-    for (const [key, reportedAt] of reportedWarnings) {
-      const expiresAt = reportedAt + PRELOAD_WARNING_TTL_MS;
-      if (now >= expiresAt) reportedWarnings.delete(key);
-      else nextWarningExpiry = Math.min(nextWarningExpiry, expiresAt);
-    }
-  }
-  for (const warning of warnings) {
-    const key = `${context}:${warning.kind}:${warning.count}:${warning.budget}:${warning.removed}`;
-    const reportedAt = reportedWarnings.get(key);
-    if (reportedAt !== void 0 && now - reportedAt < PRELOAD_WARNING_TTL_MS) continue;
-    reportedWarnings.set(key, now);
-    nextWarningExpiry = Math.min(nextWarningExpiry, now + PRELOAD_WARNING_TTL_MS);
-    while (reportedWarnings.size > MAX_REPORTED_PRELOAD_WARNINGS) {
-      const oldest = reportedWarnings.keys().next().value;
-      if (oldest === void 0) break;
-      reportedWarnings.delete(oldest);
-    }
-    const action = warning.removed > 0 ? ` Removed ${warning.removed} lower-priority hint(s).` : "";
-    const recommendation = warning.kind === "image" ? ' Mark only the LCP image with `preload` or `fetchPriority="high"`.' : " Set `preload: false` on fonts that are not required above the fold.";
-    console.warn(
-      `[Farm.js] ${context} emitted ${warning.count} ${warning.kind} preload hints (budget: ${warning.budget}).${action}${recommendation}`
-    );
-  }
-}
-function selectPreloadsWithinBudget(candidates, config) {
-  const removed = /* @__PURE__ */ new Set();
-  const warnings = [];
-  for (const kind of ["image", "font"]) {
-    const resources = /* @__PURE__ */ new Map();
-    for (const candidate of candidates) {
-      if (candidate.kind !== kind) continue;
-      const key = candidate.href ?? `#${candidate.index}`;
-      const hints = resources.get(key);
-      if (hints) hints.push(candidate);
-      else resources.set(key, [candidate]);
-    }
-    const limit = kind === "image" ? config.maxImages : config.maxFonts;
-    if (resources.size <= limit) continue;
-    const ranked = [...resources.values()].map((hints) => ({
-      hints,
-      index: hints[0].index,
-      highPriority: hints.some((hint) => hint.highPriority)
-    })).sort(
-      (left, right) => Number(right.highPriority) - Number(left.highPriority) || left.index - right.index
-    );
-    if (config.mode === "enforce") {
-      for (const resource of ranked.slice(limit)) {
-        for (const hint of resource.hints) removed.add(hint.index);
-      }
-    }
-    warnings.push({
-      kind,
-      count: resources.size,
-      budget: limit,
-      removed: config.mode === "enforce" ? resources.size - limit : 0
-    });
-  }
-  return { removed, warnings };
-}
-function getHtmlPreloadKind(link) {
-  const rel = readHtmlAttribute(link, "rel")?.toLowerCase().split(/\s+/) || [];
-  if (!rel.includes("preload")) return void 0;
-  return normalizePreloadKind(readHtmlAttribute(link, "as"));
-}
-function getLinkHeaderPreloadKind(link) {
-  const relations = getLinkHeaderParameter(link, "rel")?.toLowerCase().split(/\s+/) ?? [];
-  if (!relations.includes("preload")) return void 0;
-  return normalizePreloadKind(getLinkHeaderParameter(link, "as"));
-}
-function getLinkHeaderTarget(link) {
-  const start = link.indexOf("<");
-  const end = link.indexOf(">", start + 1);
-  return start === -1 || end === -1 ? void 0 : link.slice(start + 1, end).trim();
-}
-function getLinkHeaderParameter(link, name) {
-  const uriEnd = link.indexOf(">");
-  if (uriEnd === -1) return void 0;
-  for (const parameter of splitLinkParameters(link.slice(uriEnd + 1))) {
-    const separator = parameter.indexOf("=");
-    const parameterName = (separator === -1 ? parameter : parameter.slice(0, separator)).trim().toLowerCase();
-    if (parameterName !== name.toLowerCase() || separator === -1) continue;
-    const rawValue = parameter.slice(separator + 1).trim();
-    if (rawValue.startsWith('"') && rawValue.endsWith('"')) {
-      return rawValue.slice(1, -1).replace(/\\([\\"])/g, "$1");
-    }
-    return rawValue;
-  }
-  return void 0;
-}
-function splitLinkParameters(value) {
-  const parameters = [];
-  let start = 0;
-  let quoted = false;
-  for (let index = 0; index < value.length; index += 1) {
-    const character = value[index];
-    if (character === '"' && !isEscaped(value, index)) quoted = !quoted;
-    else if (character === ";" && !quoted) {
-      const parameter2 = value.slice(start, index).trim();
-      if (parameter2) parameters.push(parameter2);
-      start = index + 1;
-    }
-  }
-  const parameter = value.slice(start).trim();
-  if (parameter) parameters.push(parameter);
-  return parameters;
-}
-function normalizePreloadKind(value) {
-  const normalized = value?.toLowerCase();
-  return normalized === "image" || normalized === "font" ? normalized : void 0;
-}
-function readHtmlAttribute(element, name) {
-  const match = element.match(
-    new RegExp(`\\s${name}\\s*=\\s*(?:"([^"]*)"|'([^']*)'|([^\\s>]+))`, "i")
-  );
-  return match?.[1] ?? match?.[2] ?? match?.[3];
-}
-function findHtmlLinkElements(html) {
-  const lastToken = typeof html === "string" ? findLastPreloadToken(html) : Infinity;
-  if (lastToken === -1) return [];
-  const elements = [];
-  let lowerHtml = foldHtmlTagCase(
-    lastToken === Infinity ? html : html.slice(0, lastToken + RAW_TEXT_TAG_LOOKAHEAD)
-  );
-  const rawTextTag = /<(script|style|template|textarea|title|noscript|svg)(?=[\s/>])/y;
-  const linkTag = /<link(?=[\s/>])/iy;
-  const ordinaryTag = /<\/?[A-Za-z][\w:-]*(?=[\s/>])/y;
-  let cursor = 0;
-  while (cursor < html.length) {
-    const start = html.indexOf("<", cursor);
-    if (start === -1 || start > lastToken) break;
-    if (html.startsWith("<!--", start)) {
-      const commentEnd = html.indexOf("-->", start + 4);
-      cursor = commentEnd === -1 ? html.length : commentEnd + 3;
-      continue;
-    }
-    rawTextTag.lastIndex = start;
-    const rawText = rawTextTag.exec(lowerHtml);
-    if (rawText?.[1]) {
-      const openingEnd = findHtmlTagEnd(html, start);
-      if (openingEnd === -1) break;
-      if (rawText[1] === "svg" && html[openingEnd - 2] === "/") {
-        cursor = openingEnd;
-        continue;
-      }
-      let closingStart = findHtmlClosingTag(lowerHtml, rawText[1], openingEnd);
-      if (closingStart === -1 && lowerHtml.length < html.length) {
-        lowerHtml = foldHtmlTagCase(html);
-        closingStart = findHtmlClosingTag(lowerHtml, rawText[1], openingEnd);
-      }
-      if (closingStart === -1) {
-        cursor = html.length;
-        continue;
-      }
-      const closingEnd = findHtmlTagEnd(html, closingStart);
-      cursor = closingEnd === -1 ? html.length : closingEnd;
-      continue;
-    }
-    linkTag.lastIndex = start;
-    if (linkTag.test(html)) {
-      const end = findHtmlTagEnd(html, start);
-      if (end === -1) break;
-      elements.push({ start, end, value: html.slice(start, end) });
-      cursor = end;
-      continue;
-    }
-    ordinaryTag.lastIndex = start;
-    if (ordinaryTag.test(html)) {
-      const end = findHtmlTagEnd(html, start);
-      cursor = end === -1 ? html.length : end;
-      continue;
-    }
-    cursor = start + 1;
-  }
-  return elements;
-}
-var RAW_TEXT_TAG_LOOKAHEAD = 16;
-var PRELOAD_TOKEN = /\bpreload\b/gi;
-function findLastPreloadToken(html) {
-  let last = -1;
-  PRELOAD_TOKEN.lastIndex = 0;
-  for (let match = PRELOAD_TOKEN.exec(html); match; match = PRELOAD_TOKEN.exec(html)) {
-    last = match.index;
-  }
-  return last;
-}
-function foldHtmlTagCase(html) {
-  const lowerHtml = html.toLowerCase();
-  return lowerHtml.length === html.length ? lowerHtml : html.replace(/[A-Z]/g, (letter) => letter.toLowerCase());
-}
-function findHtmlClosingTag(html, tagName, start) {
-  const prefix = `</${tagName}`;
-  let candidate = html.indexOf(prefix, start);
-  while (candidate !== -1) {
-    const boundary = html[candidate + prefix.length];
-    if (boundary === ">" || boundary === "/" || /\s/.test(boundary ?? "")) return candidate;
-    candidate = html.indexOf(prefix, candidate + prefix.length);
-  }
-  return -1;
-}
-function findHtmlTagEnd(html, start) {
-  let quote;
-  for (let index = start + 1; index < html.length; index += 1) {
-    const character = html[index];
-    if (quote) {
-      if (character === quote) quote = void 0;
-    } else if (character === '"' || character === "'") {
-      quote = character;
-    } else if (character === ">") {
-      return index + 1;
-    }
-  }
-  return -1;
-}
-function removeHtmlLinkElements(html, elements, removed) {
-  let cursor = 0;
-  let output = "";
-  for (const [index, element] of elements.entries()) {
-    if (!removed.has(index)) continue;
-    output += html.slice(cursor, element.start);
-    cursor = element.end;
-  }
-  return output + html.slice(cursor);
-}
-function splitLinkHeader(value) {
-  const links = [];
-  let start = 0;
-  let insideAngleBrackets = false;
-  let quoted = false;
-  for (let index = 0; index < value.length; index += 1) {
-    const character = value[index];
-    if (quoted) {
-      if (character === '"' && !isEscaped(value, index)) quoted = false;
-      continue;
-    }
-    if (character === '"') {
-      quoted = true;
-    } else if (character === "<") {
-      insideAngleBrackets = true;
-    } else if (character === ">") {
-      insideAngleBrackets = false;
-    } else if (character === "," && !insideAngleBrackets) {
-      links.push(value.slice(start, index).trim());
-      start = index + 1;
-    }
-  }
-  links.push(value.slice(start).trim());
-  return links.filter(Boolean);
-}
-function isEscaped(value, index) {
-  let backslashes = 0;
-  for (let cursor = index - 1; cursor >= 0 && value[cursor] === "\\"; cursor -= 1) {
-    backslashes += 1;
-  }
-  return backslashes % 2 === 1;
-}
-function normalizeBudget(value, fallback) {
-  return typeof value === "number" && Number.isFinite(value) && value >= 0 ? Math.floor(value) : fallback;
-}
-
-// src/preview-noindex.ts
-var FARM_PREVIEW_ROBOTS_TAG = "noindex, nofollow";
-function resolveFarmNoindexPreviews(input) {
-  if (input === void 0) return false;
-  if (typeof input !== "boolean") {
-    throw new TypeError("agent.noindexPreviews must be a boolean.");
-  }
-  return input;
-}
-function applyFarmPreviewRobotsTag(response) {
-  try {
-    response.headers.set("X-Robots-Tag", FARM_PREVIEW_ROBOTS_TAG);
-    return response;
-  } catch {
-    const headers = new Headers(response.headers);
-    headers.set("X-Robots-Tag", FARM_PREVIEW_ROBOTS_TAG);
-    return new Response(response.body, {
-      status: response.status,
-      statusText: response.statusText,
-      headers
-    });
-  }
-}
-
-export {
-  resolveFarmAgentCrawlers,
-  createFarmAgentRobots,
-  resolveFarmAgentSitemap,
-  createFarmDefaultSitemap,
-  resolveFarmPerformanceConfig,
-  manageFarmHtmlPreloads,
-  manageFarmLinkHeaderPreloads,
-  manageFarmDocumentPreloads,
-  reportFarmPreloadWarnings,
-  FARM_PREVIEW_ROBOTS_TAG,
-  resolveFarmNoindexPreviews,
-  applyFarmPreviewRobotsTag
-};