webforai 2.1.1 → 4.0.0

package/dist/cjs/mdast-handlers/definition-list-handler.js

@@ -0,0 +1,110 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var definition_list_handler_exports = {};
+__export(definition_list_handler_exports, {
+  definitionListHandler: () => definitionListHandler
+});
+module.exports = __toCommonJS(definition_list_handler_exports);
+const hoistBoundaryWhitespace = (nodes) => {
+  const result = [];
+  for (const node of nodes) {
+    if (node.type !== "emphasis" && node.type !== "strong") {
+      result.push(node);
+      continue;
+    }
+    node.children = hoistBoundaryWhitespace(node.children);
+    const first = node.children[0];
+    if (first?.type === "text" && /^\s/.test(first.value)) {
+      result.push({ type: "text", value: " " });
+      first.value = first.value.replace(/^\s+/, "");
+    }
+    let trailing = false;
+    const last = node.children[node.children.length - 1];
+    if (last?.type === "text" && /\s$/.test(last.value)) {
+      last.value = last.value.replace(/\s+$/, "");
+      trailing = true;
+    }
+    node.children = node.children.filter((child) => !(child.type === "text" && child.value.length === 0));
+    if (node.children.length > 0) {
+      result.push(node);
+    }
+    if (trailing) {
+      result.push({ type: "text", value: " " });
+    }
+  }
+  return result;
+};
+const termsAndDefinitions = (parent) => {
+  const found = [];
+  for (const child of parent.children) {
+    if (child.type !== "element") {
+      continue;
+    }
+    if (child.tagName === "dt" || child.tagName === "dd") {
+      found.push(child);
+    } else if (child.tagName === "div") {
+      found.push(...termsAndDefinitions(child));
+    }
+  }
+  return found;
+};
+const definitionListHandler = (state, node) => {
+  const items = [];
+  let current;
+  const flush = () => {
+    if (!current) {
+      return;
+    }
+    const heading = {
+      type: "paragraph",
+      children: current.term.length > 0 ? [{ type: "strong", children: current.term }] : []
+    };
+    items.push({ type: "listItem", spread: false, children: [heading, ...current.definitions] });
+    current = void 0;
+  };
+  for (const element of termsAndDefinitions(node)) {
+    if (element.tagName === "dt") {
+      if (current && current.definitions.length > 0) {
+        flush();
+      }
+      const phrasing = hoistBoundaryWhitespace(state.all(element));
+      current = current ? { term: [...current.term, { type: "text", value: ", " }, ...phrasing], definitions: [] } : { term: phrasing, definitions: [] };
+      continue;
+    }
+    if (element.tagName === "dd") {
+      if (!current) {
+        current = { term: [], definitions: [] };
+      }
+      const content = state.toFlow(state.all(element));
+      current.definitions.push(...content);
+    }
+  }
+  flush();
+  if (items.length === 0) {
+    const fallback = state.toFlow(state.all(node));
+    return fallback.length > 0 ? fallback : void 0;
+  }
+  const result = { type: "list", ordered: false, spread: false, children: items };
+  state.patch(node, result);
+  return result;
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  definitionListHandler
+});

package/dist/cjs/mdast-handlers/math-handler.js

@@ -23,10 +23,19 @@ __export(math_handler_exports, {
 module.exports = __toCommonJS(math_handler_exports);
 var import_hast_util_to_html = require("hast-util-to-html");
 var import_mathml_to_latex = require("mathml-to-latex");
+var import_hast_fast = require("../utils/hast-fast");
+const readAnnotation = (node) => {
+  const alttext = (0, import_hast_fast.stringProperty)(node, "alttext");
+  if (!alttext) {
+    return void 0;
+  }
+  const unwrapped = alttext.trim().replace(/^\{\\displaystyle\s*([\s\S]*)\}$/, "$1").trim();
+  return unwrapped.length > 0 ? unwrapped : void 0;
+};
+const isDisplayMath = (node) => (0, import_hast_fast.stringProperty)(node, "display") === "block";
 const mathHandler = (state, node) => {
-  const mathMl = (0, import_hast_util_to_html.toHtml)(node);
-  const latex = import_mathml_to_latex.MathMLToLaTeX.convert(mathMl);
-  const result = { type: "inlineMath", value: latex };
+  const latex = readAnnotation(node) ?? import_mathml_to_latex.MathMLToLaTeX.convert((0, import_hast_util_to_html.toHtml)(node));
+  const result = isDisplayMath(node) ? { type: "math", value: latex, meta: null } : { type: "inlineMath", value: latex };
   state.patch(node, result);
   return result;
 };

package/dist/cjs/mdast-handlers/sup-sub-handler.js

@@ -0,0 +1,39 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var sup_sub_handler_exports = {};
+__export(sup_sub_handler_exports, {
+  supSubHandler: () => supSubHandler
+});
+module.exports = __toCommonJS(sup_sub_handler_exports);
+var import_hast_util_to_string = require("hast-util-to-string");
+const MAX_ANNOTATION_LENGTH = 32;
+const supSubHandler = (tagName) => (state, node) => {
+  const inner = (0, import_hast_util_to_string.toString)(node).trim();
+  if (inner.length === 0 || inner.length > MAX_ANNOTATION_LENGTH) {
+    return state.all(node);
+  }
+  const escaped = inner.replace(/&/g, "&amp;").replace(/</g, "&lt;").replace(/>/g, "&gt;");
+  const result = { type: "html", value: `<${tagName}>${escaped}</${tagName}>` };
+  state.patch(node, result);
+  return result;
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  supSubHandler
+});

package/dist/cjs/mdast-to-markdown.js

@@ -33,11 +33,8 @@ const DEFAULT_MDAST_TO_MARKDOWN_OPTIONS = {
 };
 const mdastToMarkdown = (mdast, options) => {
   const { baseUrl, ...toMarkdownOptions } = { ...DEFAULT_MDAST_TO_MARKDOWN_OPTIONS, ...options };
-  let markdown = (0, import_mdast_util_to_markdown.toMarkdown)((0, import_mdast_utils.warpRoot)(mdast), toMarkdownOptions).replace(/\*\*\*\*/g, "");
-  if (baseUrl) {
-    markdown = (0, import_link_replacer.linkReplacer)(markdown, baseUrl);
-  }
-  return markdown;
+  const tree = (0, import_mdast_utils.warpRoot)(mdast);
+  return (0, import_mdast_util_to_markdown.toMarkdown)(baseUrl ? (0, import_link_replacer.linkReplacer)(tree, baseUrl) : tree, toMarkdownOptions);
 };
 // Annotate the CommonJS export names for ESM import in node:
 0 && (module.exports = {

package/dist/cjs/metadata/index.js

@@ -0,0 +1,223 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var metadata_exports = {};
+__export(metadata_exports, {
+  extractMetadata: () => extractMetadata,
+  readJsonLd: () => readJsonLd,
+  toFrontmatter: () => toFrontmatter
+});
+module.exports = __toCommonJS(metadata_exports);
+var import_hast_fast = require("../utils/hast-fast");
+const readMetaTags = (tree) => {
+  const tags = /* @__PURE__ */ new Map();
+  for (const meta of (0, import_hast_fast.collectElements)(tree, (element) => element.tagName === "meta")) {
+    const key = (0, import_hast_fast.stringProperty)(meta, "property") ?? (0, import_hast_fast.stringProperty)(meta, "name") ?? (0, import_hast_fast.stringProperty)(meta, "itemprop");
+    const content = (0, import_hast_fast.stringProperty)(meta, "content");
+    if (!(key && content)) {
+      continue;
+    }
+    const normalized = key.toLowerCase();
+    if (!tags.has(normalized)) {
+      tags.set(normalized, content);
+    }
+  }
+  return tags;
+};
+const readJsonLd = (tree) => {
+  const entries = [];
+  for (const script of (0, import_hast_fast.collectElements)(tree, (element) => element.tagName === "script")) {
+    const type = (0, import_hast_fast.stringProperty)(script, "type");
+    if (type?.toLowerCase() !== "application/ld+json") {
+      continue;
+    }
+    const source = script.children.map((child) => child.type === "text" ? child.value : "").join("").trim();
+    if (source.length === 0) {
+      continue;
+    }
+    let parsed;
+    try {
+      parsed = JSON.parse(source);
+    } catch {
+      continue;
+    }
+    for (const entry of flattenJsonLd(parsed)) {
+      entries.push(entry);
+    }
+  }
+  return entries;
+};
+const flattenJsonLd = (value) => {
+  if (Array.isArray(value)) {
+    return value.flatMap(flattenJsonLd);
+  }
+  if (typeof value !== "object" || value === null) {
+    return [];
+  }
+  const record = value;
+  const graph = record["@graph"];
+  if (graph) {
+    return flattenJsonLd(graph);
+  }
+  return [record];
+};
+const ARTICLE_TYPES = /* @__PURE__ */ new Set([
+  "article",
+  "newsarticle",
+  "blogposting",
+  "techarticle",
+  "scholarlyarticle",
+  "report",
+  "webpage"
+]);
+const jsonLdType = (entry) => {
+  const type = entry["@type"];
+  if (typeof type === "string") {
+    return [type.toLowerCase()];
+  }
+  if (Array.isArray(type)) {
+    return type.filter((value) => typeof value === "string").map((value) => value.toLowerCase());
+  }
+  return [];
+};
+const readName = (value) => {
+  if (typeof value === "string") {
+    return value;
+  }
+  if (Array.isArray(value)) {
+    for (const entry of value) {
+      const name = readName(entry);
+      if (name) {
+        return name;
+      }
+    }
+    return void 0;
+  }
+  if (typeof value === "object" && value !== null) {
+    const name = value.name;
+    return typeof name === "string" ? name : void 0;
+  }
+  return void 0;
+};
+const asString = (value) => typeof value === "string" && value.length > 0 ? value : void 0;
+const MAX_TITLE_LENGTH = 160;
+const asTitle = (value) => {
+  const cleaned = clean(value);
+  return cleaned && cleaned.length <= MAX_TITLE_LENGTH ? cleaned : void 0;
+};
+const clean = (value) => {
+  if (value === void 0) {
+    return void 0;
+  }
+  const trimmed = value.replace(/\s+/g, " ").trim();
+  return trimmed.length > 0 ? trimmed : void 0;
+};
+const documentTitle = (tree) => {
+  const title = (0, import_hast_fast.findElement)(tree, (element) => element.tagName === "title");
+  if (!title) {
+    return void 0;
+  }
+  return title.children.map((child) => child.type === "text" ? child.value : "").join("");
+};
+const firstHeading = (tree) => {
+  const heading = (0, import_hast_fast.findElement)(tree, (element) => element.tagName === "h1");
+  if (!heading) {
+    return void 0;
+  }
+  let text = "";
+  (0, import_hast_fast.walk)(heading, (node) => {
+    if (node.type === "text") {
+      text += node.value;
+    }
+  });
+  return text;
+};
+const linkHref = (tree, rel) => {
+  const link = (0, import_hast_fast.findElement)(
+    tree,
+    (element) => element.tagName === "link" && ((0, import_hast_fast.stringProperty)(element, "rel") ?? "").toLowerCase().includes(rel)
+  );
+  return link ? (0, import_hast_fast.stringProperty)(link, "href") : void 0;
+};
+const documentLang = (tree) => {
+  const html = (0, import_hast_fast.findElement)(tree, (element) => element.tagName === "html");
+  if (html) {
+    const declared = (0, import_hast_fast.stringProperty)(html, "lang") ?? (0, import_hast_fast.stringProperty)(html, "xml:lang");
+    if (declared) {
+      return declared;
+    }
+  }
+  if ((0, import_hast_fast.isElement)(tree)) {
+    const own = (0, import_hast_fast.stringProperty)(tree, "lang");
+    if (own) {
+      return own;
+    }
+  }
+  const tagged = (0, import_hast_fast.findElement)(tree, (element) => Boolean((0, import_hast_fast.stringProperty)(element, "lang")));
+  return tagged ? (0, import_hast_fast.stringProperty)(tagged, "lang") : void 0;
+};
+const extractMetadata = (tree) => {
+  const meta = readMetaTags(tree);
+  const jsonLd = readJsonLd(tree);
+  const article = jsonLd.find((entry) => jsonLdType(entry).some((type) => ARTICLE_TYPES.has(type))) ?? jsonLd[0];
+  const get = (...keys) => {
+    for (const key of keys) {
+      const value = meta.get(key);
+      if (value) {
+        return value;
+      }
+    }
+    return void 0;
+  };
+  const title = asTitle(get("og:title", "twitter:title")) ?? asTitle(asString(article?.headline)) ?? asTitle(asString(article?.name)) ?? asTitle(documentTitle(tree)) ?? asTitle(firstHeading(tree));
+  return {
+    title,
+    description: clean(get("og:description", "twitter:description", "description")) ?? clean(asString(article?.description)),
+    author: clean(readName(article?.author)) ?? clean(get("author", "article:author", "twitter:creator")),
+    published: clean(asString(article?.datePublished)) ?? clean(get("article:published_time", "date", "pubdate")),
+    modified: clean(asString(article?.dateModified)) ?? clean(get("article:modified_time", "og:updated_time")),
+    siteName: clean(get("og:site_name", "application-name")) ?? clean(readName(article?.publisher)),
+    canonicalUrl: clean(linkHref(tree, "canonical")) ?? clean(get("og:url", "twitter:url")),
+    lang: clean(documentLang(tree)),
+    image: clean(get("og:image", "twitter:image", "twitter:image:src")),
+    type: clean(get("og:type"))
+  };
+};
+const yamlString = (value) => `"${value.replace(/\\/g, "\\\\").replace(/"/g, '\\"')}"`;
+const toFrontmatter = (metadata) => {
+  const lines = [];
+  for (const [key, value] of Object.entries(metadata)) {
+    if (typeof value === "string" && value.length > 0) {
+      lines.push(`${key}: ${yamlString(value)}`);
+    }
+  }
+  if (lines.length === 0) {
+    return "";
+  }
+  return `---
+${lines.join("\n")}
+---
+
+`;
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  extractMetadata,
+  readJsonLd,
+  toFrontmatter
+});

package/dist/cjs/normalize/index.js

@@ -0,0 +1,177 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var normalize_exports = {};
+__export(normalize_exports, {
+  bestFromSrcset: () => bestFromSrcset,
+  dedupeRenderedMath: () => dedupeRenderedMath,
+  normalizeHast: () => normalizeHast,
+  promoteAriaHeadings: () => promoteAriaHeadings,
+  resolveImageSources: () => resolveImageSources
+});
+module.exports = __toCommonJS(normalize_exports);
+var import_hast_fast = require("../utils/hast-fast");
+const LAZY_SRC_ATTRIBUTES = [
+  "data-src",
+  "data-original",
+  "data-lazy-src",
+  "data-original-src",
+  "data-actualsrc",
+  "data-hi-res-src",
+  "data-echo"
+];
+const PLACEHOLDER_SRC = /^$|^data:image\/|^about:blank$|spacer\.gif$|blank\.(gif|png)$/i;
+const readSrcset = (element, attribute) => {
+  const raw = (0, import_hast_fast.property)(element, attribute);
+  if (typeof raw === "string") {
+    return raw;
+  }
+  if (Array.isArray(raw)) {
+    return raw.join(", ");
+  }
+  return void 0;
+};
+const bestFromSrcset = (srcset) => {
+  let best;
+  for (const entry of srcset.split(",")) {
+    const parts = entry.trim().split(/\s+/);
+    const url = parts[0];
+    if (!url) {
+      continue;
+    }
+    const descriptor = parts[1];
+    let weight = 1;
+    if (descriptor?.endsWith("w")) {
+      weight = Number.parseFloat(descriptor) || 1;
+    } else if (descriptor?.endsWith("x")) {
+      weight = (Number.parseFloat(descriptor) || 1) * 1e3;
+    }
+    if (!best || weight > best.weight) {
+      best = { url, weight };
+    }
+  }
+  return best?.url;
+};
+const resolveImageSources = (tree) => {
+  (0, import_hast_fast.walk)(tree, (node) => {
+    if (!(0, import_hast_fast.isElement)(node) || node.tagName !== "img") {
+      return;
+    }
+    const current = (0, import_hast_fast.stringProperty)(node, "src");
+    if (current && !PLACEHOLDER_SRC.test(current)) {
+      return;
+    }
+    for (const attribute of LAZY_SRC_ATTRIBUTES) {
+      const candidate = (0, import_hast_fast.stringProperty)(node, attribute);
+      if (candidate && !PLACEHOLDER_SRC.test(candidate)) {
+        node.properties.src = candidate;
+        return;
+      }
+    }
+    for (const attribute of ["srcSet", "srcset", "data-srcset"]) {
+      const srcset = readSrcset(node, attribute);
+      const best = srcset ? bestFromSrcset(srcset) : void 0;
+      if (best) {
+        node.properties.src = best;
+        return;
+      }
+    }
+  });
+};
+const katexDuplicates = (node) => {
+  const classes = (0, import_hast_fast.classList)(node);
+  if (!(classes.includes("katex") || classes.includes("katex-display"))) {
+    return [];
+  }
+  const hasMathml = node.children.some(
+    (child) => (0, import_hast_fast.isElement)(child) && (0, import_hast_fast.classList)(child).includes("katex-mathml") && hasMathSource(child)
+  );
+  if (!hasMathml) {
+    return [];
+  }
+  return node.children.filter((child) => (0, import_hast_fast.isElement)(child) && (0, import_hast_fast.classList)(child).includes("katex-html"));
+};
+const mediawikiDuplicates = (node) => {
+  if (!(0, import_hast_fast.classList)(node).includes("mwe-math-element") || !hasMathSource(node)) {
+    return [];
+  }
+  return node.children.filter(
+    (child) => (0, import_hast_fast.isElement)(child) && (0, import_hast_fast.classList)(child).some((name) => name.startsWith("mwe-math-fallback-image"))
+  );
+};
+const mathjaxDuplicates = (node) => {
+  if (node.tagName !== "mjx-container") {
+    return [];
+  }
+  const hasMathml = node.children.some(
+    (child) => (0, import_hast_fast.isElement)(child) && child.tagName === "mjx-assistive-mml" && hasMathSource(child)
+  );
+  if (!hasMathml) {
+    return [];
+  }
+  return node.children.filter((child) => (0, import_hast_fast.isElement)(child) && child.tagName !== "mjx-assistive-mml");
+};
+const hasMathSource = (node) => Boolean((0, import_hast_fast.findElement)(node, (element) => element.tagName === "math" && element.children.length > 0));
+const dedupeRenderedMath = (tree) => {
+  const doomed = /* @__PURE__ */ new Set();
+  (0, import_hast_fast.walk)(tree, (node) => {
+    if (!(0, import_hast_fast.isElement)(node)) {
+      return;
+    }
+    for (const duplicate of [...katexDuplicates(node), ...mathjaxDuplicates(node), ...mediawikiDuplicates(node)]) {
+      doomed.add(duplicate);
+    }
+  });
+  if (doomed.size > 0) {
+    (0, import_hast_fast.pruneInPlace)(tree, (node) => !((0, import_hast_fast.isElement)(node) && doomed.has(node)));
+  }
+};
+const promoteAriaHeadings = (tree) => {
+  (0, import_hast_fast.walk)(tree, (node) => {
+    if (!(0, import_hast_fast.isElement)(node) || (0, import_hast_fast.stringProperty)(node, "role") !== "heading") {
+      return;
+    }
+    if (/^h[1-6]$/.test(node.tagName)) {
+      return;
+    }
+    const level = (0, import_hast_fast.numericProperty)(node, "aria-level") ?? 2;
+    const clamped = Math.min(6, Math.max(1, Math.round(level)));
+    node.tagName = `h${clamped}`;
+  });
+};
+const normalizeHast = (tree, options = {}) => {
+  const { images = true, math = true, ariaHeadings = true } = options;
+  if (images) {
+    resolveImageSources(tree);
+  }
+  if (math) {
+    dedupeRenderedMath(tree);
+  }
+  if (ariaHeadings) {
+    promoteAriaHeadings(tree);
+  }
+  return tree;
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  bestFromSrcset,
+  dedupeRenderedMath,
+  normalizeHast,
+  promoteAriaHeadings,
+  resolveImageSources
+});