@aglyn/plugins-ai 1.0.0-beta.217 → 1.0.0-beta.219

package/src/lib/runtime/seo-page-facts.js

@@ -1,150 +0,0 @@
-/**
- * @license
- * Copyright 2026 Aglyn LLC
- *
- * Licensed under the Apache License, Version 2.0 (the "License");
- * you may not use this file except in compliance with the License.
- * You may obtain a copy of the License at
- *
- *   http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an "AS IS" BASIS,
- * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
- * See the License for the specific language governing permissions and
- * limitations under the License.
- */ import { buildPageMarkdown, headingLevelOf, isPageChromeNode, pageContentRootId } from "@aglyn/aglyn/app-utils/page-markdown";
-/**
- * What a page SAYS, read off its node tree (AGL-2910): the text a search
- * listing is written from, the headings, and the images — each heading and
- * image by node id, because a content fix addresses the element it changes.
- *
- * The text is the page's Markdown representation, the one an agent is
- * served, so what the listing summarizes is what a reader of the page gets.
- * The walk starts where that serializer starts — the content region — and
- * skips what it skips, through the same chrome predicate: a heading in the
- * site's navigation is not the page's heading.
- */ /** How much of a page's text a prompt carries. A listing summarizes; it does not need the page. */ export const AI_SEO_PAGE_TEXT_MAX_CHARS = 2000;
-/** How much text around an image describes it. */ const IMAGE_CONTEXT_MAX_CHARS = 160;
-/** The deepest a walk goes, and it stops at a cycle — a node map is data. */ const MAX_DEPTH = 64;
-/** The text components whose heading level `headingLevelOf` reads. */ const TEXT_COMPONENTS = new Set([
-    'muiTypography',
-    'muiInlineText',
-    'muiListItemText'
-]);
-const collapse = (value)=>value.replace(/\s+/g, ' ').trim();
-/** Author HTML to its text, for a heading stored as rich text. */ function htmlText(html) {
-    return collapse(html.replace(/<[^>]*>/g, ' ').replace(/&nbsp;/g, ' ').replace(/&amp;/g, '&'));
-}
-/** A text component's own text, and whether it is plain (a fix can rewrite it). */ function textOf(props) {
-    const html = props['html'];
-    if (typeof html === 'string' && html.trim()) return {
-        text: htmlText(html),
-        plain: false
-    };
-    const children = props['children'];
-    return typeof children === 'string' ? {
-        text: collapse(children),
-        plain: true
-    } : {
-        text: '',
-        plain: true
-    };
-}
-const childIdsOf = (node)=>Array.isArray(node == null ? void 0 : node.nodes) ? (node == null ? void 0 : node.nodes).filter((id)=>typeof id === 'string') : [];
-/** Read the facts a listing and an audit need from one page's node map. */ export function aiSeoPageFacts(nodes, options = {}) {
-    var _options_maxChars;
-    const empty = {
-        text: '',
-        wordCount: 0,
-        headings: [],
-        h1s: [],
-        images: [],
-        imagesMissingAlt: [],
-        contentRootId: null
-    };
-    if (!nodes) return empty;
-    const contentRootId = pageContentRootId(nodes, options.rootId);
-    if (!contentRootId) return empty;
-    const markdown = buildPageMarkdown({
-        nodes,
-        rootId: options.rootId
-    });
-    const text = markdown.trim().slice(0, (_options_maxChars = options.maxChars) != null ? _options_maxChars : AI_SEO_PAGE_TEXT_MAX_CHARS);
-    const headings = [];
-    const images = [];
-    let lastText = '';
-    const seen = new Set();
-    const visit = (id, depth)=>{
-        var _node_props, _node_componentId, _props_html;
-        if (depth > MAX_DEPTH || seen.has(id)) return;
-        const node = nodes[id];
-        if (!node) return;
-        seen.add(id);
-        if (isPageChromeNode(node)) return;
-        const props = (_node_props = node.props) != null ? _node_props : {};
-        const componentId = String((_node_componentId = node.componentId) != null ? _node_componentId : '');
-        if (TEXT_COMPONENTS.has(componentId)) {
-            const { text: own, plain } = textOf(props);
-            const level = headingLevelOf(props);
-            if (own) {
-                if (level > 0) headings.push({
-                    nodeId: id,
-                    level,
-                    text: own,
-                    editable: plain
-                });
-                lastText = own;
-            }
-        } else if (componentId === 'markdown' && typeof props['content'] === 'string') {
-            for (const line of String(props['content']).split('\n')){
-                const match = /^(#{1,6})\s+(.+)$/.exec(line.trim());
-                if (match) {
-                    headings.push({
-                        nodeId: null,
-                        level: match[1].length,
-                        text: collapse(match[2]),
-                        editable: false
-                    });
-                }
-            }
-            lastText = collapse(String(props['content'])).slice(0, IMAGE_CONTEXT_MAX_CHARS);
-        } else if (componentId === 'custom-html' && typeof ((_props_html = props['html']) != null ? _props_html : props['children']) === 'string') {
-            var _props_html1;
-            const html = String((_props_html1 = props['html']) != null ? _props_html1 : props['children']);
-            for (const match of html.matchAll(/<h([1-6])[^>]*>([\s\S]*?)<\/h\1>/gi)){
-                headings.push({
-                    nodeId: null,
-                    level: Number(match[1]),
-                    text: htmlText(match[2]),
-                    editable: false
-                });
-            }
-        } else if (componentId === 'image') {
-            var _props_src;
-            const src = String((_props_src = props['src']) != null ? _props_src : '').trim();
-            if (src) {
-                var _props_alt;
-                images.push({
-                    nodeId: id,
-                    src,
-                    alt: collapse(String((_props_alt = props['alt']) != null ? _props_alt : '')),
-                    context: lastText.slice(0, IMAGE_CONTEXT_MAX_CHARS)
-                });
-            }
-        }
-        for (const childId of childIdsOf(node))visit(childId, depth + 1);
-    };
-    visit(contentRootId, 0);
-    return {
-        text,
-        wordCount: text ? text.split(/\s+/).filter(Boolean).length : 0,
-        headings,
-        h1s: headings.filter((heading)=>heading.level === 1),
-        images,
-        imagesMissingAlt: images.filter((image)=>!image.alt),
-        contentRootId
-    };
-}
-
-//# sourceMappingURL=seo-page-facts.js.map

package/src/lib/runtime/seo-page-facts.js.map

@@ -1 +0,0 @@
-{"version":3,"sources":["../../../../../../../libs/plugins/ai/src/lib/runtime/seo-page-facts.ts"],"sourcesContent":["/**\n * @license\n * Copyright 2026 Aglyn LLC\n *\n * Licensed under the Apache License, Version 2.0 (the \"License\");\n * you may not use this file except in compliance with the License.\n * You may obtain a copy of the License at\n *\n *   http://www.apache.org/licenses/LICENSE-2.0\n *\n * Unless required by applicable law or agreed to in writing, software\n * distributed under the License is distributed on an \"AS IS\" BASIS,\n * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n * See the License for the specific language governing permissions and\n * limitations under the License.\n */\n\nimport {\n  buildPageMarkdown,\n  headingLevelOf,\n  isPageChromeNode,\n  pageContentRootId,\n  type PageMarkdownNode,\n  type PageMarkdownNodes,\n} from '@aglyn/aglyn/app-utils/page-markdown'\n\n/**\n * What a page SAYS, read off its node tree (AGL-2910): the text a search\n * listing is written from, the headings, and the images — each heading and\n * image by node id, because a content fix addresses the element it changes.\n *\n * The text is the page's Markdown representation, the one an agent is\n * served, so what the listing summarizes is what a reader of the page gets.\n * The walk starts where that serializer starts — the content region — and\n * skips what it skips, through the same chrome predicate: a heading in the\n * site's navigation is not the page's heading.\n */\n\n/** How much of a page's text a prompt carries. A listing summarizes; it does not need the page. */\nexport const AI_SEO_PAGE_TEXT_MAX_CHARS = 2_000\n\n/** How much text around an image describes it. */\nconst IMAGE_CONTEXT_MAX_CHARS = 160\n\n/** The deepest a walk goes, and it stops at a cycle — a node map is data. */\nconst MAX_DEPTH = 64\n\n/** The text components whose heading level `headingLevelOf` reads. */\nconst TEXT_COMPONENTS: ReadonlySet<string> = new Set(['muiTypography', 'muiInlineText', 'muiListItemText'])\n\nexport interface AiSeoHeading {\n  /** The element, or `null` for a heading inside rich text a fix cannot address. */\n  nodeId: string | null\n  level: number\n  text: string\n  /** Whether a content fix can change it: a text component with plain text. */\n  editable: boolean\n}\n\nexport interface AiSeoImage {\n  nodeId: string\n  src: string\n  alt: string\n  /** The nearest heading or text before it, which is what the image illustrates. */\n  context: string\n}\n\nexport interface AiSeoPageFacts {\n  /** The page's Markdown body, capped. */\n  text: string\n  wordCount: number\n  headings: AiSeoHeading[]\n  h1s: AiSeoHeading[]\n  images: AiSeoImage[]\n  imagesMissingAlt: AiSeoImage[]\n  /** The node the content region starts at; a new heading goes first inside it. */\n  contentRootId: string | null\n}\n\nconst collapse = (value: string): string => value.replace(/\\s+/g, ' ').trim()\n\n/** Author HTML to its text, for a heading stored as rich text. */\nfunction htmlText(html: string): string {\n  return collapse(html.replace(/<[^>]*>/g, ' ').replace(/&nbsp;/g, ' ').replace(/&amp;/g, '&'))\n}\n\n/** A text component's own text, and whether it is plain (a fix can rewrite it). */\nfunction textOf(props: Record<string, unknown>): { text: string; plain: boolean } {\n  const html = props['html']\n  if (typeof html === 'string' && html.trim()) return { text: htmlText(html), plain: false }\n  const children = props['children']\n  return typeof children === 'string' ? { text: collapse(children), plain: true } : { text: '', plain: true }\n}\n\nconst childIdsOf = (node: PageMarkdownNode | undefined): string[] =>\n  Array.isArray(node?.nodes) ? (node?.nodes as unknown[]).filter((id): id is string => typeof id === 'string') : []\n\n/** Read the facts a listing and an audit need from one page's node map. */\nexport function aiSeoPageFacts(\n  nodes: PageMarkdownNodes | null | undefined,\n  options: { rootId?: string; maxChars?: number } = {},\n): AiSeoPageFacts {\n  const empty: AiSeoPageFacts = {\n    text: '',\n    wordCount: 0,\n    headings: [],\n    h1s: [],\n    images: [],\n    imagesMissingAlt: [],\n    contentRootId: null,\n  }\n  if (!nodes) return empty\n  const contentRootId = pageContentRootId(nodes, options.rootId)\n  if (!contentRootId) return empty\n\n  const markdown = buildPageMarkdown({ nodes, rootId: options.rootId })\n  const text = markdown.trim().slice(0, options.maxChars ?? AI_SEO_PAGE_TEXT_MAX_CHARS)\n  const headings: AiSeoHeading[] = []\n  const images: AiSeoImage[] = []\n  let lastText = ''\n\n  const seen = new Set<string>()\n  const visit = (id: string, depth: number): void => {\n    if (depth > MAX_DEPTH || seen.has(id)) return\n    const node = nodes[id]\n    if (!node) return\n    seen.add(id)\n    if (isPageChromeNode(node)) return\n    const props = (node.props ?? {}) as Record<string, unknown>\n    const componentId = String(node.componentId ?? '')\n    if (TEXT_COMPONENTS.has(componentId)) {\n      const { text: own, plain } = textOf(props)\n      const level = headingLevelOf(props)\n      if (own) {\n        if (level > 0) headings.push({ nodeId: id, level, text: own, editable: plain })\n        lastText = own\n      }\n    } else if (componentId === 'markdown' && typeof props['content'] === 'string') {\n      for (const line of String(props['content']).split('\\n')) {\n        const match = /^(#{1,6})\\s+(.+)$/.exec(line.trim())\n        if (match) {\n          headings.push({ nodeId: null, level: match[1].length, text: collapse(match[2]), editable: false })\n        }\n      }\n      lastText = collapse(String(props['content'])).slice(0, IMAGE_CONTEXT_MAX_CHARS)\n    } else if (componentId === 'custom-html' && typeof (props['html'] ?? props['children']) === 'string') {\n      const html = String(props['html'] ?? props['children'])\n      for (const match of html.matchAll(/<h([1-6])[^>]*>([\\s\\S]*?)<\\/h\\1>/gi)) {\n        headings.push({ nodeId: null, level: Number(match[1]), text: htmlText(match[2]), editable: false })\n      }\n    } else if (componentId === 'image') {\n      const src = String(props['src'] ?? '').trim()\n      if (src) {\n        images.push({\n          nodeId: id,\n          src,\n          alt: collapse(String(props['alt'] ?? '')),\n          context: lastText.slice(0, IMAGE_CONTEXT_MAX_CHARS),\n        })\n      }\n    }\n    for (const childId of childIdsOf(node)) visit(childId, depth + 1)\n  }\n  visit(contentRootId, 0)\n\n  return {\n    text,\n    wordCount: text ? text.split(/\\s+/).filter(Boolean).length : 0,\n    headings,\n    h1s: headings.filter((heading) => heading.level === 1),\n    images,\n    imagesMissingAlt: images.filter((image) => !image.alt),\n    contentRootId,\n  }\n}\n"],"names":["buildPageMarkdown","headingLevelOf","isPageChromeNode","pageContentRootId","AI_SEO_PAGE_TEXT_MAX_CHARS","IMAGE_CONTEXT_MAX_CHARS","MAX_DEPTH","TEXT_COMPONENTS","Set","collapse","value","replace","trim","htmlText","html","textOf","props","text","plain","children","childIdsOf","node","Array","isArray","nodes","filter","id","aiSeoPageFacts","options","empty","wordCount","headings","h1s","images","imagesMissingAlt","contentRootId","rootId","markdown","slice","maxChars","lastText","seen","visit","depth","has","add","componentId","String","own","level","push","nodeId","editable","line","split","match","exec","length","matchAll","Number","src","alt","context","childId","Boolean","heading","image"],"mappings":"AAAA;;;;;;;;;;;;;;;CAeC,GAED,SACEA,iBAAiB,EACjBC,cAAc,EACdC,gBAAgB,EAChBC,iBAAiB,QAGZ,uCAAsC;AAE7C;;;;;;;;;;CAUC,GAED,iGAAiG,GACjG,OAAO,MAAMC,6BAA6B,KAAK;AAE/C,gDAAgD,GAChD,MAAMC,0BAA0B;AAEhC,2EAA2E,GAC3E,MAAMC,YAAY;AAElB,oEAAoE,GACpE,MAAMC,kBAAuC,IAAIC,IAAI;IAAC;IAAiB;IAAiB;CAAkB;AA+B1G,MAAMC,WAAW,CAACC,QAA0BA,MAAMC,OAAO,CAAC,QAAQ,KAAKC,IAAI;AAE3E,gEAAgE,GAChE,SAASC,SAASC,IAAY;IAC5B,OAAOL,SAASK,KAAKH,OAAO,CAAC,YAAY,KAAKA,OAAO,CAAC,WAAW,KAAKA,OAAO,CAAC,UAAU;AAC1F;AAEA,iFAAiF,GACjF,SAASI,OAAOC,KAA8B;IAC5C,MAAMF,OAAOE,KAAK,CAAC,OAAO;IAC1B,IAAI,OAAOF,SAAS,YAAYA,KAAKF,IAAI,IAAI,OAAO;QAAEK,MAAMJ,SAASC;QAAOI,OAAO;IAAM;IACzF,MAAMC,WAAWH,KAAK,CAAC,WAAW;IAClC,OAAO,OAAOG,aAAa,WAAW;QAAEF,MAAMR,SAASU;QAAWD,OAAO;IAAK,IAAI;QAAED,MAAM;QAAIC,OAAO;IAAK;AAC5G;AAEA,MAAME,aAAa,CAACC,OAClBC,MAAMC,OAAO,CAACF,wBAAAA,KAAMG,KAAK,IAAI,CAACH,wBAAAA,KAAMG,KAAK,EAAeC,MAAM,CAAC,CAACC,KAAqB,OAAOA,OAAO,YAAY,EAAE;AAEnH,yEAAyE,GACzE,OAAO,SAASC,eACdH,KAA2C,EAC3CI,UAAkD,CAAC,CAAC;QAgBdA;IAdtC,MAAMC,QAAwB;QAC5BZ,MAAM;QACNa,WAAW;QACXC,UAAU,EAAE;QACZC,KAAK,EAAE;QACPC,QAAQ,EAAE;QACVC,kBAAkB,EAAE;QACpBC,eAAe;IACjB;IACA,IAAI,CAACX,OAAO,OAAOK;IACnB,MAAMM,gBAAgBhC,kBAAkBqB,OAAOI,QAAQQ,MAAM;IAC7D,IAAI,CAACD,eAAe,OAAON;IAE3B,MAAMQ,WAAWrC,kBAAkB;QAAEwB;QAAOY,QAAQR,QAAQQ,MAAM;IAAC;IACnE,MAAMnB,OAAOoB,SAASzB,IAAI,GAAG0B,KAAK,CAAC,IAAGV,oBAAAA,QAAQW,QAAQ,YAAhBX,oBAAoBxB;IAC1D,MAAM2B,WAA2B,EAAE;IACnC,MAAME,SAAuB,EAAE;IAC/B,IAAIO,WAAW;IAEf,MAAMC,OAAO,IAAIjC;IACjB,MAAMkC,QAAQ,CAAChB,IAAYiB;YAMVtB,aACYA,mBAgByBL;QAtBpD,IAAI2B,QAAQrC,aAAamC,KAAKG,GAAG,CAAClB,KAAK;QACvC,MAAML,OAAOG,KAAK,CAACE,GAAG;QACtB,IAAI,CAACL,MAAM;QACXoB,KAAKI,GAAG,CAACnB;QACT,IAAIxB,iBAAiBmB,OAAO;QAC5B,MAAML,SAASK,cAAAA,KAAKL,KAAK,YAAVK,cAAc,CAAC;QAC9B,MAAMyB,cAAcC,QAAO1B,oBAAAA,KAAKyB,WAAW,YAAhBzB,oBAAoB;QAC/C,IAAId,gBAAgBqC,GAAG,CAACE,cAAc;YACpC,MAAM,EAAE7B,MAAM+B,GAAG,EAAE9B,KAAK,EAAE,GAAGH,OAAOC;YACpC,MAAMiC,QAAQhD,eAAee;YAC7B,IAAIgC,KAAK;gBACP,IAAIC,QAAQ,GAAGlB,SAASmB,IAAI,CAAC;oBAAEC,QAAQzB;oBAAIuB;oBAAOhC,MAAM+B;oBAAKI,UAAUlC;gBAAM;gBAC7EsB,WAAWQ;YACb;QACF,OAAO,IAAIF,gBAAgB,cAAc,OAAO9B,KAAK,CAAC,UAAU,KAAK,UAAU;YAC7E,KAAK,MAAMqC,QAAQN,OAAO/B,KAAK,CAAC,UAAU,EAAEsC,KAAK,CAAC,MAAO;gBACvD,MAAMC,QAAQ,oBAAoBC,IAAI,CAACH,KAAKzC,IAAI;gBAChD,IAAI2C,OAAO;oBACTxB,SAASmB,IAAI,CAAC;wBAAEC,QAAQ;wBAAMF,OAAOM,KAAK,CAAC,EAAE,CAACE,MAAM;wBAAExC,MAAMR,SAAS8C,KAAK,CAAC,EAAE;wBAAGH,UAAU;oBAAM;gBAClG;YACF;YACAZ,WAAW/B,SAASsC,OAAO/B,KAAK,CAAC,UAAU,GAAGsB,KAAK,CAAC,GAAGjC;QACzD,OAAO,IAAIyC,gBAAgB,iBAAiB,SAAQ9B,cAAAA,KAAK,CAAC,OAAO,YAAbA,cAAiBA,KAAK,CAAC,WAAW,MAAM,UAAU;gBAChFA;YAApB,MAAMF,OAAOiC,QAAO/B,eAAAA,KAAK,CAAC,OAAO,YAAbA,eAAiBA,KAAK,CAAC,WAAW;YACtD,KAAK,MAAMuC,SAASzC,KAAK4C,QAAQ,CAAC,sCAAuC;gBACvE3B,SAASmB,IAAI,CAAC;oBAAEC,QAAQ;oBAAMF,OAAOU,OAAOJ,KAAK,CAAC,EAAE;oBAAGtC,MAAMJ,SAAS0C,KAAK,CAAC,EAAE;oBAAGH,UAAU;gBAAM;YACnG;QACF,OAAO,IAAIN,gBAAgB,SAAS;gBACf9B;YAAnB,MAAM4C,MAAMb,QAAO/B,aAAAA,KAAK,CAAC,MAAM,YAAZA,aAAgB,IAAIJ,IAAI;YAC3C,IAAIgD,KAAK;oBAIgB5C;gBAHvBiB,OAAOiB,IAAI,CAAC;oBACVC,QAAQzB;oBACRkC;oBACAC,KAAKpD,SAASsC,QAAO/B,aAAAA,KAAK,CAAC,MAAM,YAAZA,aAAgB;oBACrC8C,SAAStB,SAASF,KAAK,CAAC,GAAGjC;gBAC7B;YACF;QACF;QACA,KAAK,MAAM0D,WAAW3C,WAAWC,MAAOqB,MAAMqB,SAASpB,QAAQ;IACjE;IACAD,MAAMP,eAAe;IAErB,OAAO;QACLlB;QACAa,WAAWb,OAAOA,KAAKqC,KAAK,CAAC,OAAO7B,MAAM,CAACuC,SAASP,MAAM,GAAG;QAC7D1B;QACAC,KAAKD,SAASN,MAAM,CAAC,CAACwC,UAAYA,QAAQhB,KAAK,KAAK;QACpDhB;QACAC,kBAAkBD,OAAOR,MAAM,CAAC,CAACyC,QAAU,CAACA,MAAML,GAAG;QACrD1B;IACF;AACF"}